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/3263

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

❌ 4 New Failures, 11 Pending

As of commit 8e1dd7e with merge base 546a1b7 (image):

NEW FAILURES - The following jobs have failed:

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: $\large\color{#35bf28}24$. Worsened: $\large\color{#d91a1a}9$.

Expand to view detailed results
Name Max Mean Ops Ops on Repo HEAD Change
test_tensor_to_bytestream_speed[pickle] 86.1573μs 85.1728μs 11.7408 KOps/s 12.3809 KOps/s $\textbf{\color{#d91a1a}-5.17\%}$
test_tensor_to_bytestream_speed[torch.save] 0.1387ms 0.1372ms 7.2885 KOps/s 7.2444 KOps/s $\color{#35bf28}+0.61\%$
test_tensor_to_bytestream_speed[untyped_storage] 0.1247s 0.1241s 8.0575 Ops/s 8.1136 Ops/s $\color{#d91a1a}-0.69\%$
test_tensor_to_bytestream_speed[numpy] 2.7648μs 2.7530μs 363.2453 KOps/s 365.0801 KOps/s $\color{#d91a1a}-0.50\%$
test_tensor_to_bytestream_speed[safetensors] 39.7242μs 38.0236μs 26.2995 KOps/s 27.5655 KOps/s $\color{#d91a1a}-4.59\%$
test_simple 0.5504s 0.5446s 1.8363 Ops/s 1.7695 Ops/s $\color{#35bf28}+3.78\%$
test_transformed 1.1038s 1.1025s 0.9070 Ops/s 0.8758 Ops/s $\color{#35bf28}+3.57\%$
test_serial 1.6455s 1.6434s 0.6085 Ops/s 0.5910 Ops/s $\color{#35bf28}+2.96\%$
test_parallel 1.3346s 1.1489s 0.8704 Ops/s 0.8320 Ops/s $\color{#35bf28}+4.62\%$
test_step_mdp_speed[True-True-True-True-True] 0.1355ms 43.8642μs 22.7976 KOps/s 22.2851 KOps/s $\color{#35bf28}+2.30\%$
test_step_mdp_speed[True-True-True-True-False] 48.8500μs 24.3877μs 41.0042 KOps/s 40.4674 KOps/s $\color{#35bf28}+1.33\%$
test_step_mdp_speed[True-True-True-False-True] 57.8710μs 24.3636μs 41.0448 KOps/s 40.4586 KOps/s $\color{#35bf28}+1.45\%$
test_step_mdp_speed[True-True-True-False-False] 38.3800μs 13.4615μs 74.2859 KOps/s 73.0385 KOps/s $\color{#35bf28}+1.71\%$
test_step_mdp_speed[True-True-False-True-True] 81.1920μs 46.0493μs 21.7159 KOps/s 21.2052 KOps/s $\color{#35bf28}+2.41\%$
test_step_mdp_speed[True-True-False-True-False] 88.6920μs 26.7641μs 37.3635 KOps/s 35.8640 KOps/s $\color{#35bf28}+4.18\%$
test_step_mdp_speed[True-True-False-False-True] 70.1810μs 27.2109μs 36.7499 KOps/s 36.9645 KOps/s $\color{#d91a1a}-0.58\%$
test_step_mdp_speed[True-True-False-False-False] 44.7510μs 15.9690μs 62.6212 KOps/s 60.5460 KOps/s $\color{#35bf28}+3.43\%$
test_step_mdp_speed[True-False-True-True-True] 77.2510μs 49.0359μs 20.3932 KOps/s 20.1427 KOps/s $\color{#35bf28}+1.24\%$
test_step_mdp_speed[True-False-True-True-False] 56.8610μs 29.4177μs 33.9931 KOps/s 32.9177 KOps/s $\color{#35bf28}+3.27\%$
test_step_mdp_speed[True-False-True-False-True] 52.8210μs 26.5446μs 37.6725 KOps/s 36.1861 KOps/s $\color{#35bf28}+4.11\%$
test_step_mdp_speed[True-False-True-False-False] 42.0900μs 16.2023μs 61.7195 KOps/s 61.4703 KOps/s $\color{#35bf28}+0.41\%$
test_step_mdp_speed[True-False-False-True-True] 0.1062ms 51.3681μs 19.4673 KOps/s 19.3821 KOps/s $\color{#35bf28}+0.44\%$
test_step_mdp_speed[True-False-False-True-False] 76.1520μs 32.1570μs 31.0974 KOps/s 30.4317 KOps/s $\color{#35bf28}+2.19\%$
test_step_mdp_speed[True-False-False-False-True] 58.5620μs 29.1244μs 34.3354 KOps/s 33.6545 KOps/s $\color{#35bf28}+2.02\%$
test_step_mdp_speed[True-False-False-False-False] 54.0610μs 18.9002μs 52.9095 KOps/s 52.0348 KOps/s $\color{#35bf28}+1.68\%$
test_step_mdp_speed[False-True-True-True-True] 0.1406ms 48.8559μs 20.4683 KOps/s 20.0561 KOps/s $\color{#35bf28}+2.06\%$
test_step_mdp_speed[False-True-True-True-False] 62.9410μs 29.3563μs 34.0643 KOps/s 33.0695 KOps/s $\color{#35bf28}+3.01\%$
test_step_mdp_speed[False-True-True-False-True] 2.4923ms 31.3894μs 31.8579 KOps/s 31.8919 KOps/s $\color{#d91a1a}-0.11\%$
test_step_mdp_speed[False-True-True-False-False] 49.6610μs 17.8030μs 56.1703 KOps/s 55.2808 KOps/s $\color{#35bf28}+1.61\%$
test_step_mdp_speed[False-True-False-True-True] 88.8510μs 51.5769μs 19.3885 KOps/s 19.3667 KOps/s $\color{#35bf28}+0.11\%$
test_step_mdp_speed[False-True-False-True-False] 89.1420μs 31.9168μs 31.3315 KOps/s 30.5615 KOps/s $\color{#35bf28}+2.52\%$
test_step_mdp_speed[False-True-False-False-True] 58.1410μs 32.7620μs 30.5232 KOps/s 29.6482 KOps/s $\color{#35bf28}+2.95\%$
test_step_mdp_speed[False-True-False-False-False] 53.5010μs 20.1841μs 49.5439 KOps/s 48.9731 KOps/s $\color{#35bf28}+1.17\%$
test_step_mdp_speed[False-False-True-True-True] 97.3920μs 54.2661μs 18.4277 KOps/s 18.0693 KOps/s $\color{#35bf28}+1.98\%$
test_step_mdp_speed[False-False-True-True-False] 85.9910μs 34.5703μs 28.9266 KOps/s 28.4910 KOps/s $\color{#35bf28}+1.53\%$
test_step_mdp_speed[False-False-True-False-True] 68.1120μs 32.6147μs 30.6610 KOps/s 29.4321 KOps/s $\color{#35bf28}+4.18\%$
test_step_mdp_speed[False-False-True-False-False] 58.1810μs 19.9541μs 50.1150 KOps/s 48.1987 KOps/s $\color{#35bf28}+3.98\%$
test_step_mdp_speed[False-False-False-True-True] 0.1326ms 56.1765μs 17.8010 KOps/s 17.3541 KOps/s $\color{#35bf28}+2.58\%$
test_step_mdp_speed[False-False-False-True-False] 75.9610μs 37.2964μs 26.8123 KOps/s 26.2824 KOps/s $\color{#35bf28}+2.02\%$
test_step_mdp_speed[False-False-False-False-True] 68.1720μs 35.0453μs 28.5345 KOps/s 28.5057 KOps/s $\color{#35bf28}+0.10\%$
test_step_mdp_speed[False-False-False-False-False] 49.9310μs 22.7106μs 44.0324 KOps/s 42.9652 KOps/s $\color{#35bf28}+2.48\%$
test_non_tensor_env_rollout_speed[1000-single-True] 0.8630s 0.7627s 1.3111 Ops/s 1.2979 Ops/s $\color{#35bf28}+1.02\%$
test_non_tensor_env_rollout_speed[1000-single-False] 0.7284s 0.6298s 1.5879 Ops/s 1.5730 Ops/s $\color{#35bf28}+0.94\%$
test_non_tensor_env_rollout_speed[1000-serial-no-buffers-True] 1.7400s 1.6636s 0.6011 Ops/s 0.5972 Ops/s $\color{#35bf28}+0.66\%$
test_non_tensor_env_rollout_speed[1000-serial-no-buffers-False] 1.5366s 1.4558s 0.6869 Ops/s 0.6873 Ops/s $\color{#d91a1a}-0.06\%$
test_non_tensor_env_rollout_speed[1000-serial-buffers-True] 2.0133s 1.9324s 0.5175 Ops/s 0.5219 Ops/s $\color{#d91a1a}-0.84\%$
test_non_tensor_env_rollout_speed[1000-serial-buffers-False] 1.8105s 1.7180s 0.5821 Ops/s 0.5906 Ops/s $\color{#d91a1a}-1.45\%$
test_non_tensor_env_rollout_speed[1000-parallel-no-buffers-True] 4.7097s 4.6174s 0.2166 Ops/s 0.2178 Ops/s $\color{#d91a1a}-0.55\%$
test_non_tensor_env_rollout_speed[1000-parallel-no-buffers-False] 4.4726s 4.4443s 0.2250 Ops/s 0.2271 Ops/s $\color{#d91a1a}-0.91\%$
test_non_tensor_env_rollout_speed[1000-parallel-buffers-True] 2.0276s 1.9508s 0.5126 Ops/s 0.5064 Ops/s $\color{#35bf28}+1.23\%$
test_non_tensor_env_rollout_speed[1000-parallel-buffers-False] 1.7557s 1.6560s 0.6039 Ops/s 0.5923 Ops/s $\color{#35bf28}+1.95\%$
test_values[generalized_advantage_estimate-True-True] 10.7315ms 10.0779ms 99.2265 Ops/s 100.3740 Ops/s $\color{#d91a1a}-1.14\%$
test_values[vec_generalized_advantage_estimate-True-True] 20.5419ms 17.7707ms 56.2724 Ops/s 89.7547 Ops/s $\textbf{\color{#d91a1a}-37.30\%}$
test_values[td0_return_estimate-False-False] 0.2430ms 0.1288ms 7.7658 KOps/s 7.7040 KOps/s $\color{#35bf28}+0.80\%$
test_values[td1_return_estimate-False-False] 27.7941ms 26.6218ms 37.5632 Ops/s 37.7413 Ops/s $\color{#d91a1a}-0.47\%$
test_values[vec_td1_return_estimate-False-False] 18.6549ms 17.6112ms 56.7819 Ops/s 88.7898 Ops/s $\textbf{\color{#d91a1a}-36.05\%}$
test_values[td_lambda_return_estimate-True-False] 39.8824ms 39.2798ms 25.4584 Ops/s 25.3976 Ops/s $\color{#35bf28}+0.24\%$
test_values[vec_td_lambda_return_estimate-True-False] 18.5489ms 17.6373ms 56.6981 Ops/s 87.4963 Ops/s $\textbf{\color{#d91a1a}-35.20\%}$
test_gae_speed[generalized_advantage_estimate-False-1-512] 8.9997ms 8.8832ms 112.5718 Ops/s 113.4648 Ops/s $\color{#d91a1a}-0.79\%$
test_gae_speed[vec_generalized_advantage_estimate-True-1-512] 1.7757ms 1.5199ms 657.9426 Ops/s 652.6873 Ops/s $\color{#35bf28}+0.81\%$
test_gae_speed[vec_generalized_advantage_estimate-False-1-512] 0.5354ms 0.4174ms 2.3959 KOps/s 2.4281 KOps/s $\color{#d91a1a}-1.33\%$
test_gae_speed[vec_generalized_advantage_estimate-True-32-512] 38.7572ms 35.1364ms 28.4605 Ops/s 33.4168 Ops/s $\textbf{\color{#d91a1a}-14.83\%}$
test_gae_speed[vec_generalized_advantage_estimate-False-32-512] 1.8823ms 1.7069ms 585.8563 Ops/s 581.7277 Ops/s $\color{#35bf28}+0.71\%$
test_dqn_speed[False-None] 1.5250ms 1.3692ms 730.3536 Ops/s 719.3758 Ops/s $\color{#35bf28}+1.53\%$
test_dqn_speed[False-backward] 1.9254ms 1.8748ms 533.3859 Ops/s 527.0601 Ops/s $\color{#35bf28}+1.20\%$
test_dqn_speed[True-None] 0.7007ms 0.5218ms 1.9164 KOps/s 1.8762 KOps/s $\color{#35bf28}+2.14\%$
test_dqn_speed[True-backward] 1.0154ms 0.9679ms 1.0332 KOps/s 1.0330 KOps/s $\color{#35bf28}+0.01\%$
test_dqn_speed[reduce-overhead-None] 0.8731ms 0.5180ms 1.9305 KOps/s 1.9158 KOps/s $\color{#35bf28}+0.77\%$
test_dqn_speed[reduce-overhead-backward] 1.0877ms 0.9667ms 1.0344 KOps/s 866.4296 Ops/s $\textbf{\color{#35bf28}+19.39\%}$
test_ddpg_speed[False-None] 3.4255ms 2.8002ms 357.1114 Ops/s 351.9481 Ops/s $\color{#35bf28}+1.47\%$
test_ddpg_speed[False-backward] 4.1302ms 3.9937ms 250.3930 Ops/s 251.1309 Ops/s $\color{#d91a1a}-0.29\%$
test_ddpg_speed[True-None] 1.7992ms 1.3734ms 728.1348 Ops/s 722.3354 Ops/s $\color{#35bf28}+0.80\%$
test_ddpg_speed[True-backward] 2.4546ms 2.3564ms 424.3686 Ops/s 402.0956 Ops/s $\textbf{\color{#35bf28}+5.54\%}$
test_ddpg_speed[reduce-overhead-None] 1.8020ms 1.3710ms 729.3981 Ops/s 716.6346 Ops/s $\color{#35bf28}+1.78\%$
test_ddpg_speed[reduce-overhead-backward] 2.4220ms 2.3328ms 428.6614 Ops/s 354.5383 Ops/s $\textbf{\color{#35bf28}+20.91\%}$
test_sac_speed[False-None] 9.1186ms 7.8244ms 127.8050 Ops/s 125.3273 Ops/s $\color{#35bf28}+1.98\%$
test_sac_speed[False-backward] 11.3820ms 10.9788ms 91.0843 Ops/s 90.3296 Ops/s $\color{#35bf28}+0.84\%$
test_sac_speed[True-None] 2.2969ms 2.1242ms 470.7736 Ops/s 463.7606 Ops/s $\color{#35bf28}+1.51\%$
test_sac_speed[True-backward] 4.0917ms 3.9770ms 251.4438 Ops/s 247.7881 Ops/s $\color{#35bf28}+1.48\%$
test_sac_speed[reduce-overhead-None] 2.3377ms 2.1637ms 462.1732 Ops/s 460.7950 Ops/s $\color{#35bf28}+0.30\%$
test_sac_speed[reduce-overhead-backward] 4.2158ms 4.0103ms 249.3604 Ops/s 247.7698 Ops/s $\color{#35bf28}+0.64\%$
test_redq_speed[False-None] 13.3430ms 10.3946ms 96.2041 Ops/s 96.0053 Ops/s $\color{#35bf28}+0.21\%$
test_redq_speed[False-backward] 22.3729ms 17.9534ms 55.6998 Ops/s 56.6852 Ops/s $\color{#d91a1a}-1.74\%$
test_redq_speed[True-None] 5.3438ms 4.5293ms 220.7840 Ops/s 216.3145 Ops/s $\color{#35bf28}+2.07\%$
test_redq_speed[True-backward] 10.2918ms 9.9081ms 100.9270 Ops/s 91.3690 Ops/s $\textbf{\color{#35bf28}+10.46\%}$
test_redq_speed[reduce-overhead-None] 4.6433ms 4.4448ms 224.9834 Ops/s 221.8314 Ops/s $\color{#35bf28}+1.42\%$
test_redq_speed[reduce-overhead-backward] 10.3805ms 10.0727ms 99.2784 Ops/s 100.6683 Ops/s $\color{#d91a1a}-1.38\%$
test_redq_deprec_speed[False-None] 13.6622ms 10.9592ms 91.2473 Ops/s 92.8838 Ops/s $\color{#d91a1a}-1.76\%$
test_redq_deprec_speed[False-backward] 16.2156ms 15.8334ms 63.1578 Ops/s 65.1774 Ops/s $\color{#d91a1a}-3.10\%$
test_redq_deprec_speed[True-None] 4.2235ms 3.7117ms 269.4211 Ops/s 250.7049 Ops/s $\textbf{\color{#35bf28}+7.47\%}$
test_redq_deprec_speed[True-backward] 7.9983ms 7.6814ms 130.1853 Ops/s 120.2609 Ops/s $\textbf{\color{#35bf28}+8.25\%}$
test_redq_deprec_speed[reduce-overhead-None] 3.9642ms 3.6998ms 270.2862 Ops/s 259.3231 Ops/s $\color{#35bf28}+4.23\%$
test_redq_deprec_speed[reduce-overhead-backward] 7.8841ms 7.6776ms 130.2495 Ops/s 121.8959 Ops/s $\textbf{\color{#35bf28}+6.85\%}$
test_td3_speed[False-None] 8.7060ms 7.8905ms 126.7342 Ops/s 120.2134 Ops/s $\textbf{\color{#35bf28}+5.42\%}$
test_td3_speed[False-backward] 11.2229ms 10.7105ms 93.3664 Ops/s 93.1168 Ops/s $\color{#35bf28}+0.27\%$
test_td3_speed[True-None] 1.8652ms 1.8347ms 545.0555 Ops/s 538.5274 Ops/s $\color{#35bf28}+1.21\%$
test_td3_speed[True-backward] 3.8347ms 3.6740ms 272.1821 Ops/s 241.4854 Ops/s $\textbf{\color{#35bf28}+12.71\%}$
test_td3_speed[reduce-overhead-None] 1.8482ms 1.8156ms 550.7846 Ops/s 543.4382 Ops/s $\color{#35bf28}+1.35\%$
test_td3_speed[reduce-overhead-backward] 3.7680ms 3.6614ms 273.1233 Ops/s 277.1136 Ops/s $\color{#d91a1a}-1.44\%$
test_cql_speed[False-None] 29.9331ms 26.5765ms 37.6273 Ops/s 37.5433 Ops/s $\color{#35bf28}+0.22\%$
test_cql_speed[False-backward] 38.8314ms 35.0802ms 28.5061 Ops/s 28.3153 Ops/s $\color{#35bf28}+0.67\%$
test_cql_speed[True-None] 13.0904ms 12.3403ms 81.0350 Ops/s 79.2025 Ops/s $\color{#35bf28}+2.31\%$
test_cql_speed[True-backward] 18.7524ms 18.3274ms 54.5632 Ops/s 52.8064 Ops/s $\color{#35bf28}+3.33\%$
test_cql_speed[reduce-overhead-None] 12.8535ms 12.5447ms 79.7151 Ops/s 79.5595 Ops/s $\color{#35bf28}+0.20\%$
test_cql_speed[reduce-overhead-backward] 18.8329ms 18.4737ms 54.1310 Ops/s 54.5082 Ops/s $\color{#d91a1a}-0.69\%$
test_a2c_speed[False-None] 5.5632ms 5.3180ms 188.0409 Ops/s 183.5663 Ops/s $\color{#35bf28}+2.44\%$
test_a2c_speed[False-backward] 12.0442ms 11.7372ms 85.1989 Ops/s 83.4848 Ops/s $\color{#35bf28}+2.05\%$
test_a2c_speed[True-None] 3.8959ms 3.6979ms 270.4229 Ops/s 255.5047 Ops/s $\textbf{\color{#35bf28}+5.84\%}$
test_a2c_speed[True-backward] 8.8628ms 8.5495ms 116.9653 Ops/s 110.9044 Ops/s $\textbf{\color{#35bf28}+5.46\%}$
test_a2c_speed[reduce-overhead-None] 3.9041ms 3.7076ms 269.7184 Ops/s 270.3075 Ops/s $\color{#d91a1a}-0.22\%$
test_a2c_speed[reduce-overhead-backward] 9.0895ms 8.7786ms 113.9134 Ops/s 110.8682 Ops/s $\color{#35bf28}+2.75\%$
test_ppo_speed[False-None] 6.0853ms 5.8322ms 171.4608 Ops/s 175.5849 Ops/s $\color{#d91a1a}-2.35\%$
test_ppo_speed[False-backward] 12.5582ms 12.1504ms 82.3017 Ops/s 80.6656 Ops/s $\color{#35bf28}+2.03\%$
test_ppo_speed[True-None] 3.7321ms 3.5758ms 279.6580 Ops/s 263.1590 Ops/s $\textbf{\color{#35bf28}+6.27\%}$
test_ppo_speed[True-backward] 8.7621ms 8.4991ms 117.6592 Ops/s 116.9868 Ops/s $\color{#35bf28}+0.57\%$
test_ppo_speed[reduce-overhead-None] 3.7915ms 3.6048ms 277.4117 Ops/s 272.8120 Ops/s $\color{#35bf28}+1.69\%$
test_ppo_speed[reduce-overhead-backward] 8.9067ms 8.6540ms 115.5537 Ops/s 113.3812 Ops/s $\color{#35bf28}+1.92\%$
test_reinforce_speed[False-None] 7.2341ms 4.5288ms 220.8103 Ops/s 215.8580 Ops/s $\color{#35bf28}+2.29\%$
test_reinforce_speed[False-backward] 7.5215ms 7.2593ms 137.7541 Ops/s 133.8328 Ops/s $\color{#35bf28}+2.93\%$
test_reinforce_speed[True-None] 3.8966ms 2.8803ms 347.1850 Ops/s 335.0856 Ops/s $\color{#35bf28}+3.61\%$
test_reinforce_speed[True-backward] 8.0835ms 7.7273ms 129.4106 Ops/s 128.0140 Ops/s $\color{#35bf28}+1.09\%$
test_reinforce_speed[reduce-overhead-None] 3.1157ms 2.8594ms 349.7235 Ops/s 336.2779 Ops/s $\color{#35bf28}+4.00\%$
test_reinforce_speed[reduce-overhead-backward] 8.2079ms 7.9890ms 125.1718 Ops/s 118.9015 Ops/s $\textbf{\color{#35bf28}+5.27\%}$
test_iql_speed[False-None] 26.3658ms 20.1167ms 49.7099 Ops/s 50.2543 Ops/s $\color{#d91a1a}-1.08\%$
test_iql_speed[False-backward] 36.4402ms 30.7134ms 32.5590 Ops/s 32.7896 Ops/s $\color{#d91a1a}-0.70\%$
test_iql_speed[True-None] 9.3257ms 8.6536ms 115.5585 Ops/s 111.6056 Ops/s $\color{#35bf28}+3.54\%$
test_iql_speed[True-backward] 17.4748ms 16.7585ms 59.6713 Ops/s 62.1174 Ops/s $\color{#d91a1a}-3.94\%$
test_iql_speed[reduce-overhead-None] 8.9929ms 8.7027ms 114.9066 Ops/s 121.8293 Ops/s $\textbf{\color{#d91a1a}-5.68\%}$
test_iql_speed[reduce-overhead-backward] 17.8599ms 17.3712ms 57.5665 Ops/s 62.3598 Ops/s $\textbf{\color{#d91a1a}-7.69\%}$
test_rb_sample[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 7.6930ms 5.8656ms 170.4846 Ops/s 169.2871 Ops/s $\color{#35bf28}+0.71\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 0.9665ms 0.3474ms 2.8782 KOps/s 2.9057 KOps/s $\color{#d91a1a}-0.94\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.7410ms 0.3571ms 2.8003 KOps/s 2.8228 KOps/s $\color{#d91a1a}-0.79\%$
test_rb_sample[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 5.8057ms 5.6090ms 178.2851 Ops/s 179.5567 Ops/s $\color{#d91a1a}-0.71\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 0.7429ms 0.3474ms 2.8788 KOps/s 1.2265 KOps/s $\textbf{\color{#35bf28}+134.70\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.7023ms 0.2609ms 3.8326 KOps/s 2.8377 KOps/s $\textbf{\color{#35bf28}+35.06\%}$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-sampler6-10000] 1.4933ms 1.2479ms 801.3466 Ops/s 715.0688 Ops/s $\textbf{\color{#35bf28}+12.07\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-sampler7-10000] 1.4127ms 1.1659ms 857.6772 Ops/s 773.4185 Ops/s $\textbf{\color{#35bf28}+10.89\%}$
test_rb_sample[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 5.8626ms 5.7297ms 174.5293 Ops/s 169.7031 Ops/s $\color{#35bf28}+2.84\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 1.3242ms 0.4575ms 2.1856 KOps/s 2.1083 KOps/s $\color{#35bf28}+3.67\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.6487ms 0.4203ms 2.3794 KOps/s 2.1715 KOps/s $\textbf{\color{#35bf28}+9.57\%}$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 5.8703ms 5.5882ms 178.9491 Ops/s 173.7471 Ops/s $\color{#35bf28}+2.99\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 1.0924ms 0.2986ms 3.3495 KOps/s 3.2441 KOps/s $\color{#35bf28}+3.25\%$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.5606ms 0.2814ms 3.5538 KOps/s 3.7325 KOps/s $\color{#d91a1a}-4.79\%$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 6.0126ms 5.6651ms 176.5201 Ops/s 175.6171 Ops/s $\color{#35bf28}+0.51\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 0.9867ms 0.3205ms 3.1200 KOps/s 2.9937 KOps/s $\color{#35bf28}+4.22\%$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.5399ms 0.3425ms 2.9201 KOps/s 2.6815 KOps/s $\textbf{\color{#35bf28}+8.90\%}$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 5.9505ms 5.8203ms 171.8123 Ops/s 170.2226 Ops/s $\color{#35bf28}+0.93\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 0.7773ms 0.4495ms 2.2248 KOps/s 1.9418 KOps/s $\textbf{\color{#35bf28}+14.57\%}$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.6983ms 0.4740ms 2.1096 KOps/s 2.2674 KOps/s $\textbf{\color{#d91a1a}-6.96\%}$
test_rb_populate[TensorDictReplayBuffer-ListStorage-RandomSampler-400] 6.6509ms 5.0250ms 199.0044 Ops/s 199.1328 Ops/s $\color{#d91a1a}-0.06\%$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-400] 11.0294ms 2.4139ms 414.2649 Ops/s 418.3561 Ops/s $\color{#d91a1a}-0.98\%$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-400] 2.1950ms 1.1365ms 879.9168 Ops/s 835.4635 Ops/s $\textbf{\color{#35bf28}+5.32\%}$
test_rb_populate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-400] 0.6840s 18.8389ms 53.0816 Ops/s 50.4309 Ops/s $\textbf{\color{#35bf28}+5.26\%}$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-400] 3.5275ms 1.4735ms 678.6683 Ops/s 498.9603 Ops/s $\textbf{\color{#35bf28}+36.02\%}$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-400] 12.6996ms 1.3049ms 766.3307 Ops/s 838.6766 Ops/s $\textbf{\color{#d91a1a}-8.63\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-ListStorage-None-400] 7.4630ms 5.2644ms 189.9559 Ops/s 193.2756 Ops/s $\color{#d91a1a}-1.72\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-400] 8.5832ms 2.2069ms 453.1218 Ops/s 443.5089 Ops/s $\color{#35bf28}+2.17\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-400] 7.9218ms 1.3978ms 715.3928 Ops/s 738.5643 Ops/s $\color{#d91a1a}-3.14\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-True] 35.7374ms 33.4914ms 29.8584 Ops/s 30.0132 Ops/s $\color{#d91a1a}-0.52\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-False] 19.0882ms 17.4116ms 57.4329 Ops/s 59.1691 Ops/s $\color{#d91a1a}-2.93\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-True] 36.6418ms 34.4652ms 29.0148 Ops/s 28.6176 Ops/s $\color{#35bf28}+1.39\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-False] 19.7929ms 17.8084ms 56.1532 Ops/s 58.2232 Ops/s $\color{#d91a1a}-3.56\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-True] 41.6207ms 36.3589ms 27.5035 Ops/s 19.5157 Ops/s $\textbf{\color{#35bf28}+40.93\%}$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-False] 20.8650ms 19.1925ms 52.1037 Ops/s 51.9844 Ops/s $\color{#35bf28}+0.23\%$

[ghstack-poisoned]
[ghstack-poisoned]
vmoens added a commit that referenced this pull request Dec 18, 2025
ghstack-source-id: 0b6d9a1
Pull-Request: #3263
@vmoens vmoens merged commit 8e1dd7e into gh/vmoens/172/base Dec 18, 2025
96 of 104 checks passed
@vmoens vmoens deleted the gh/vmoens/172/head branch December 18, 2025 14:24
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