Skip to content

Conversation

@vmoens
Copy link
Collaborator

@vmoens vmoens commented Dec 31, 2025

[ghstack-poisoned]
@pytorch-bot
Copy link

pytorch-bot bot commented Dec 31, 2025

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/rl/3283

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

❗ 1 Active SEVs

There are 1 currently active SEVs. If your PR is affected, please view them below:

❌ 4 New Failures, 1 Unrelated Failure

As of commit dc95e15 with merge base 7866d11 (image):

NEW FAILURES - The following jobs have failed:

BROKEN TRUNK - The following job 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.

@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 31, 2025
@github-actions
Copy link

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

Total Benchmarks: 164. Improved: $\large\color{#35bf28}34$. Worsened: $\large\color{#d91a1a}12$.

Expand to view detailed results
Name Max Mean Ops Ops on Repo HEAD Change
test_tensor_to_bytestream_speed[pickle] 82.1078μs 81.3722μs 12.2892 KOps/s 12.3869 KOps/s $\color{#d91a1a}-0.79\%$
test_tensor_to_bytestream_speed[torch.save] 0.1421ms 0.1414ms 7.0711 KOps/s 7.1637 KOps/s $\color{#d91a1a}-1.29\%$
test_tensor_to_bytestream_speed[untyped_storage] 0.1246s 0.1236s 8.0918 Ops/s 7.9480 Ops/s $\color{#35bf28}+1.81\%$
test_tensor_to_bytestream_speed[numpy] 2.6828μs 2.6700μs 374.5352 KOps/s 366.5454 KOps/s $\color{#35bf28}+2.18\%$
test_tensor_to_bytestream_speed[safetensors] 41.6463μs 39.5291μs 25.2978 KOps/s 26.4824 KOps/s $\color{#d91a1a}-4.47\%$
test_simple 0.5585s 0.5520s 1.8117 Ops/s 1.7385 Ops/s $\color{#35bf28}+4.21\%$
test_transformed 1.1303s 1.1266s 0.8876 Ops/s 0.8639 Ops/s $\color{#35bf28}+2.75\%$
test_serial 1.6728s 1.6601s 0.6024 Ops/s 0.5835 Ops/s $\color{#35bf28}+3.24\%$
test_parallel 1.1388s 1.0983s 0.9105 Ops/s 0.8867 Ops/s $\color{#35bf28}+2.68\%$
test_step_mdp_speed[True-True-True-True-True] 0.3188ms 44.8504μs 22.2963 KOps/s 22.4118 KOps/s $\color{#d91a1a}-0.52\%$
test_step_mdp_speed[True-True-True-True-False] 49.4800μs 24.7881μs 40.3420 KOps/s 39.7842 KOps/s $\color{#35bf28}+1.40\%$
test_step_mdp_speed[True-True-True-False-True] 57.3300μs 25.2998μs 39.5259 KOps/s 40.5322 KOps/s $\color{#d91a1a}-2.48\%$
test_step_mdp_speed[True-True-True-False-False] 47.5300μs 13.6904μs 73.0440 KOps/s 71.4030 KOps/s $\color{#35bf28}+2.30\%$
test_step_mdp_speed[True-True-False-True-True] 0.1015ms 47.5685μs 21.0223 KOps/s 21.0648 KOps/s $\color{#d91a1a}-0.20\%$
test_step_mdp_speed[True-True-False-True-False] 69.7210μs 27.1954μs 36.7709 KOps/s 35.1561 KOps/s $\color{#35bf28}+4.59\%$
test_step_mdp_speed[True-True-False-False-True] 62.2610μs 27.5641μs 36.2791 KOps/s 36.0017 KOps/s $\color{#35bf28}+0.77\%$
test_step_mdp_speed[True-True-False-False-False] 56.0810μs 16.3346μs 61.2198 KOps/s 58.9865 KOps/s $\color{#35bf28}+3.79\%$
test_step_mdp_speed[True-False-True-True-True] 0.1322ms 50.5461μs 19.7839 KOps/s 20.0113 KOps/s $\color{#d91a1a}-1.14\%$
test_step_mdp_speed[True-False-True-True-False] 90.7510μs 29.6069μs 33.7759 KOps/s 32.0640 KOps/s $\textbf{\color{#35bf28}+5.34\%}$
test_step_mdp_speed[True-False-True-False-True] 58.7200μs 27.5041μs 36.3582 KOps/s 36.2300 KOps/s $\color{#35bf28}+0.35\%$
test_step_mdp_speed[True-False-True-False-False] 56.1100μs 16.3833μs 61.0379 KOps/s 60.2909 KOps/s $\color{#35bf28}+1.24\%$
test_step_mdp_speed[True-False-False-True-True] 94.5910μs 51.5289μs 19.4066 KOps/s 18.9039 KOps/s $\color{#35bf28}+2.66\%$
test_step_mdp_speed[True-False-False-True-False] 65.9210μs 33.0282μs 30.2772 KOps/s 29.6673 KOps/s $\color{#35bf28}+2.06\%$
test_step_mdp_speed[True-False-False-False-True] 72.3910μs 29.8126μs 33.5429 KOps/s 33.0226 KOps/s $\color{#35bf28}+1.58\%$
test_step_mdp_speed[True-False-False-False-False] 47.2600μs 18.6618μs 53.5854 KOps/s 51.6127 KOps/s $\color{#35bf28}+3.82\%$
test_step_mdp_speed[False-True-True-True-True] 91.7410μs 50.7653μs 19.6985 KOps/s 19.8444 KOps/s $\color{#d91a1a}-0.74\%$
test_step_mdp_speed[False-True-True-True-False] 67.3510μs 29.8078μs 33.5483 KOps/s 32.1594 KOps/s $\color{#35bf28}+4.32\%$
test_step_mdp_speed[False-True-True-False-True] 2.3699ms 31.2535μs 31.9964 KOps/s 31.1638 KOps/s $\color{#35bf28}+2.67\%$
test_step_mdp_speed[False-True-True-False-False] 47.6810μs 18.0014μs 55.5512 KOps/s 54.8471 KOps/s $\color{#35bf28}+1.28\%$
test_step_mdp_speed[False-True-False-True-True] 91.7910μs 52.1215μs 19.1859 KOps/s 19.1462 KOps/s $\color{#35bf28}+0.21\%$
test_step_mdp_speed[False-True-False-True-False] 72.9000μs 32.8310μs 30.4590 KOps/s 29.5310 KOps/s $\color{#35bf28}+3.14\%$
test_step_mdp_speed[False-True-False-False-True] 65.5910μs 33.6560μs 29.7124 KOps/s 29.0479 KOps/s $\color{#35bf28}+2.29\%$
test_step_mdp_speed[False-True-False-False-False] 60.0000μs 20.3739μs 49.0823 KOps/s 46.5013 KOps/s $\textbf{\color{#35bf28}+5.55\%}$
test_step_mdp_speed[False-False-True-True-True] 88.6710μs 55.1597μs 18.1292 KOps/s 17.7509 KOps/s $\color{#35bf28}+2.13\%$
test_step_mdp_speed[False-False-True-True-False] 63.2600μs 35.2985μs 28.3298 KOps/s 27.6529 KOps/s $\color{#35bf28}+2.45\%$
test_step_mdp_speed[False-False-True-False-True] 74.8510μs 33.7267μs 29.6501 KOps/s 29.0293 KOps/s $\color{#35bf28}+2.14\%$
test_step_mdp_speed[False-False-True-False-False] 45.7510μs 20.3332μs 49.1806 KOps/s 46.5230 KOps/s $\textbf{\color{#35bf28}+5.71\%}$
test_step_mdp_speed[False-False-False-True-True] 92.5010μs 56.7970μs 17.6066 KOps/s 17.0543 KOps/s $\color{#35bf28}+3.24\%$
test_step_mdp_speed[False-False-False-True-False] 77.8610μs 37.4815μs 26.6798 KOps/s 25.4726 KOps/s $\color{#35bf28}+4.74\%$
test_step_mdp_speed[False-False-False-False-True] 71.7310μs 35.6279μs 28.0679 KOps/s 27.2454 KOps/s $\color{#35bf28}+3.02\%$
test_step_mdp_speed[False-False-False-False-False] 54.5000μs 22.7692μs 43.9190 KOps/s 41.5245 KOps/s $\textbf{\color{#35bf28}+5.77\%}$
test_non_tensor_env_rollout_speed[1000-single-True] 0.8825s 0.7833s 1.2767 Ops/s 1.2751 Ops/s $\color{#35bf28}+0.13\%$
test_non_tensor_env_rollout_speed[1000-single-False] 0.7405s 0.6430s 1.5552 Ops/s 1.5505 Ops/s $\color{#35bf28}+0.31\%$
test_non_tensor_env_rollout_speed[1000-serial-no-buffers-True] 1.7902s 1.7095s 0.5850 Ops/s 0.5857 Ops/s $\color{#d91a1a}-0.12\%$
test_non_tensor_env_rollout_speed[1000-serial-no-buffers-False] 1.5624s 1.4801s 0.6756 Ops/s 0.6744 Ops/s $\color{#35bf28}+0.17\%$
test_non_tensor_env_rollout_speed[1000-serial-buffers-True] 2.0262s 1.9465s 0.5137 Ops/s 0.5107 Ops/s $\color{#35bf28}+0.60\%$
test_non_tensor_env_rollout_speed[1000-serial-buffers-False] 1.8026s 1.7232s 0.5803 Ops/s 0.5747 Ops/s $\color{#35bf28}+0.97\%$
test_non_tensor_env_rollout_speed[1000-parallel-no-buffers-True] 4.8193s 4.6643s 0.2144 Ops/s 0.2122 Ops/s $\color{#35bf28}+1.03\%$
test_non_tensor_env_rollout_speed[1000-parallel-no-buffers-False] 4.5286s 4.4353s 0.2255 Ops/s 0.2208 Ops/s $\color{#35bf28}+2.11\%$
test_non_tensor_env_rollout_speed[1000-parallel-buffers-True] 2.0799s 2.0409s 0.4900 Ops/s 0.5034 Ops/s $\color{#d91a1a}-2.66\%$
test_non_tensor_env_rollout_speed[1000-parallel-buffers-False] 1.7628s 1.6814s 0.5947 Ops/s 0.5846 Ops/s $\color{#35bf28}+1.74\%$
test_values[generalized_advantage_estimate-True-True] 10.5894ms 10.2416ms 97.6412 Ops/s 94.2332 Ops/s $\color{#35bf28}+3.62\%$
test_values[vec_generalized_advantage_estimate-True-True] 14.1223ms 11.5359ms 86.6858 Ops/s 55.5219 Ops/s $\textbf{\color{#35bf28}+56.13\%}$
test_values[td0_return_estimate-False-False] 0.2220ms 0.1256ms 7.9612 KOps/s 7.7481 KOps/s $\color{#35bf28}+2.75\%$
test_values[td1_return_estimate-False-False] 27.7962ms 27.4906ms 36.3761 Ops/s 35.2027 Ops/s $\color{#35bf28}+3.33\%$
test_values[vec_td1_return_estimate-False-False] 12.3262ms 11.3794ms 87.8780 Ops/s 55.7755 Ops/s $\textbf{\color{#35bf28}+57.56\%}$
test_values[td_lambda_return_estimate-True-False] 41.7423ms 40.6415ms 24.6054 Ops/s 23.8533 Ops/s $\color{#35bf28}+3.15\%$
test_values[vec_td_lambda_return_estimate-True-False] 12.1011ms 11.1949ms 89.3261 Ops/s 56.3511 Ops/s $\textbf{\color{#35bf28}+58.52\%}$
test_gae_speed[generalized_advantage_estimate-False-1-512] 9.2338ms 9.1411ms 109.3961 Ops/s 105.7856 Ops/s $\color{#35bf28}+3.41\%$
test_gae_speed[vec_generalized_advantage_estimate-True-1-512] 1.7296ms 1.5011ms 666.1738 Ops/s 652.2077 Ops/s $\color{#35bf28}+2.14\%$
test_gae_speed[vec_generalized_advantage_estimate-False-1-512] 0.7288ms 0.4149ms 2.4103 KOps/s 2.3753 KOps/s $\color{#35bf28}+1.47\%$
test_gae_speed[vec_generalized_advantage_estimate-True-32-512] 30.2233ms 29.0447ms 34.4297 Ops/s 30.4247 Ops/s $\textbf{\color{#35bf28}+13.16\%}$
test_gae_speed[vec_generalized_advantage_estimate-False-32-512] 2.2786ms 1.8103ms 552.3928 Ops/s 559.0227 Ops/s $\color{#d91a1a}-1.19\%$
test_dqn_speed[False-None] 1.7985ms 1.4006ms 713.9720 Ops/s 704.0158 Ops/s $\color{#35bf28}+1.41\%$
test_dqn_speed[False-backward] 1.9480ms 1.9069ms 524.3994 Ops/s 521.7681 Ops/s $\color{#35bf28}+0.50\%$
test_dqn_speed[True-None] 0.7380ms 0.5600ms 1.7856 KOps/s 1.8732 KOps/s $\color{#d91a1a}-4.67\%$
test_dqn_speed[True-backward] 1.0051ms 0.9620ms 1.0395 KOps/s 991.6986 Ops/s $\color{#35bf28}+4.82\%$
test_dqn_speed[reduce-overhead-None] 0.6836ms 0.5224ms 1.9141 KOps/s 1.8856 KOps/s $\color{#35bf28}+1.51\%$
test_dqn_speed[reduce-overhead-backward] 1.0145ms 0.9488ms 1.0539 KOps/s 870.3476 Ops/s $\textbf{\color{#35bf28}+21.09\%}$
test_ddpg_speed[False-None] 3.1318ms 2.8497ms 350.9090 Ops/s 340.9231 Ops/s $\color{#35bf28}+2.93\%$
test_ddpg_speed[False-backward] 4.1440ms 4.0297ms 248.1565 Ops/s 246.9372 Ops/s $\color{#35bf28}+0.49\%$
test_ddpg_speed[True-None] 1.7693ms 1.3580ms 736.3884 Ops/s 723.2975 Ops/s $\color{#35bf28}+1.81\%$
test_ddpg_speed[True-backward] 2.3566ms 2.3151ms 431.9417 Ops/s 426.8608 Ops/s $\color{#35bf28}+1.19\%$
test_ddpg_speed[reduce-overhead-None] 1.5094ms 1.3614ms 734.5564 Ops/s 718.9637 Ops/s $\color{#35bf28}+2.17\%$
test_ddpg_speed[reduce-overhead-backward] 2.4664ms 2.2947ms 435.7893 Ops/s 355.8665 Ops/s $\textbf{\color{#35bf28}+22.46\%}$
test_sac_speed[False-None] 8.3015ms 7.8340ms 127.6488 Ops/s 122.1960 Ops/s $\color{#35bf28}+4.46\%$
test_sac_speed[False-backward] 11.5503ms 11.0523ms 90.4789 Ops/s 89.0544 Ops/s $\color{#35bf28}+1.60\%$
test_sac_speed[True-None] 2.3603ms 2.0789ms 481.0135 Ops/s 463.1826 Ops/s $\color{#35bf28}+3.85\%$
test_sac_speed[True-backward] 4.2413ms 3.9465ms 253.3888 Ops/s 240.7205 Ops/s $\textbf{\color{#35bf28}+5.26\%}$
test_sac_speed[reduce-overhead-None] 2.6942ms 2.0565ms 486.2584 Ops/s 462.6340 Ops/s $\textbf{\color{#35bf28}+5.11\%}$
test_sac_speed[reduce-overhead-backward] 4.4132ms 3.9452ms 253.4702 Ops/s 236.0340 Ops/s $\textbf{\color{#35bf28}+7.39\%}$
test_redq_speed[False-None] 10.8985ms 10.0961ms 99.0478 Ops/s 93.6446 Ops/s $\textbf{\color{#35bf28}+5.77\%}$
test_redq_speed[False-backward] 18.0300ms 17.2440ms 57.9910 Ops/s 55.1534 Ops/s $\textbf{\color{#35bf28}+5.14\%}$
test_redq_speed[True-None] 4.6439ms 4.2646ms 234.4899 Ops/s 228.2773 Ops/s $\color{#35bf28}+2.72\%$
test_redq_speed[True-backward] 9.3996ms 9.0350ms 110.6807 Ops/s 105.6412 Ops/s $\color{#35bf28}+4.77\%$
test_redq_speed[reduce-overhead-None] 4.6189ms 4.1651ms 240.0904 Ops/s 229.3335 Ops/s $\color{#35bf28}+4.69\%$
test_redq_speed[reduce-overhead-backward] 9.8909ms 9.3199ms 107.2969 Ops/s 102.4513 Ops/s $\color{#35bf28}+4.73\%$
test_redq_deprec_speed[False-None] 11.6522ms 10.7952ms 92.6337 Ops/s 91.8392 Ops/s $\color{#35bf28}+0.87\%$
test_redq_deprec_speed[False-backward] 16.1416ms 15.5329ms 64.3793 Ops/s 64.0901 Ops/s $\color{#35bf28}+0.45\%$
test_redq_deprec_speed[True-None] 3.8279ms 3.4632ms 288.7478 Ops/s 258.1179 Ops/s $\textbf{\color{#35bf28}+11.87\%}$
test_redq_deprec_speed[True-backward] 7.4439ms 7.2080ms 138.7344 Ops/s 127.6033 Ops/s $\textbf{\color{#35bf28}+8.72\%}$
test_redq_deprec_speed[reduce-overhead-None] 3.5071ms 3.4120ms 293.0831 Ops/s 280.7503 Ops/s $\color{#35bf28}+4.39\%$
test_redq_deprec_speed[reduce-overhead-backward] 7.5628ms 7.2617ms 137.7092 Ops/s 121.0459 Ops/s $\textbf{\color{#35bf28}+13.77\%}$
test_td3_speed[False-None] 8.0181ms 7.8936ms 126.6851 Ops/s 124.1646 Ops/s $\color{#35bf28}+2.03\%$
test_td3_speed[False-backward] 11.2584ms 10.7384ms 93.1233 Ops/s 91.8514 Ops/s $\color{#35bf28}+1.38\%$
test_td3_speed[True-None] 1.9687ms 1.7732ms 563.9505 Ops/s 554.1984 Ops/s $\color{#35bf28}+1.76\%$
test_td3_speed[True-backward] 3.6450ms 3.5432ms 282.2275 Ops/s 273.4033 Ops/s $\color{#35bf28}+3.23\%$
test_td3_speed[reduce-overhead-None] 1.7978ms 1.7515ms 570.9507 Ops/s 559.7642 Ops/s $\color{#35bf28}+2.00\%$
test_td3_speed[reduce-overhead-backward] 3.7015ms 3.5506ms 281.6454 Ops/s 223.4535 Ops/s $\textbf{\color{#35bf28}+26.04\%}$
test_cql_speed[False-None] 28.5977ms 25.8703ms 38.6544 Ops/s 38.4144 Ops/s $\color{#35bf28}+0.62\%$
test_cql_speed[False-backward] 37.6958ms 35.2136ms 28.3981 Ops/s 28.6026 Ops/s $\color{#d91a1a}-0.71\%$
test_cql_speed[True-None] 12.4334ms 11.9920ms 83.3888 Ops/s 77.7281 Ops/s $\textbf{\color{#35bf28}+7.28\%}$
test_cql_speed[True-backward] 18.2056ms 17.6736ms 56.5814 Ops/s 56.2870 Ops/s $\color{#35bf28}+0.52\%$
test_cql_speed[reduce-overhead-None] 12.7138ms 12.2994ms 81.3049 Ops/s 80.8912 Ops/s $\color{#35bf28}+0.51\%$
test_cql_speed[reduce-overhead-backward] 17.9799ms 17.4711ms 57.2374 Ops/s 54.8652 Ops/s $\color{#35bf28}+4.32\%$
test_a2c_speed[False-None] 5.6127ms 5.4273ms 184.2523 Ops/s 180.4504 Ops/s $\color{#35bf28}+2.11\%$
test_a2c_speed[False-backward] 12.0182ms 11.6515ms 85.8260 Ops/s 84.6194 Ops/s $\color{#35bf28}+1.43\%$
test_a2c_speed[True-None] 4.1201ms 3.6703ms 272.4546 Ops/s 265.5133 Ops/s $\color{#35bf28}+2.61\%$
test_a2c_speed[True-backward] 8.7574ms 8.4899ms 117.7870 Ops/s 114.3519 Ops/s $\color{#35bf28}+3.00\%$
test_a2c_speed[reduce-overhead-None] 4.2291ms 3.6832ms 271.5046 Ops/s 269.1549 Ops/s $\color{#35bf28}+0.87\%$
test_a2c_speed[reduce-overhead-backward] 9.1213ms 8.6918ms 115.0508 Ops/s 94.1087 Ops/s $\textbf{\color{#35bf28}+22.25\%}$
test_ppo_speed[False-None] 6.1134ms 5.8030ms 172.3252 Ops/s 139.2905 Ops/s $\textbf{\color{#35bf28}+23.72\%}$
test_ppo_speed[False-backward] 12.8142ms 12.2247ms 81.8017 Ops/s 69.3023 Ops/s $\textbf{\color{#35bf28}+18.04\%}$
test_ppo_speed[True-None] 3.9306ms 3.5823ms 279.1495 Ops/s 270.1167 Ops/s $\color{#35bf28}+3.34\%$
test_ppo_speed[True-backward] 8.4612ms 8.3031ms 120.4369 Ops/s 109.6121 Ops/s $\textbf{\color{#35bf28}+9.88\%}$
test_ppo_speed[reduce-overhead-None] 3.7203ms 3.5270ms 283.5266 Ops/s 276.7221 Ops/s $\color{#35bf28}+2.46\%$
test_ppo_speed[reduce-overhead-backward] 8.7375ms 8.5758ms 116.6077 Ops/s 112.3523 Ops/s $\color{#35bf28}+3.79\%$
test_reinforce_speed[False-None] 4.7703ms 4.5855ms 218.0787 Ops/s 217.0243 Ops/s $\color{#35bf28}+0.49\%$
test_reinforce_speed[False-backward] 7.5377ms 7.3243ms 136.5327 Ops/s 135.0828 Ops/s $\color{#35bf28}+1.07\%$
test_reinforce_speed[True-None] 3.2186ms 2.7923ms 358.1327 Ops/s 335.2859 Ops/s $\textbf{\color{#35bf28}+6.81\%}$
test_reinforce_speed[True-backward] 8.0819ms 7.6226ms 131.1891 Ops/s 129.7535 Ops/s $\color{#35bf28}+1.11\%$
test_reinforce_speed[reduce-overhead-None] 3.0177ms 2.8124ms 355.5722 Ops/s 346.5215 Ops/s $\color{#35bf28}+2.61\%$
test_reinforce_speed[reduce-overhead-backward] 7.9460ms 7.7188ms 129.5544 Ops/s 119.2337 Ops/s $\textbf{\color{#35bf28}+8.66\%}$
test_iql_speed[False-None] 25.5463ms 20.2465ms 49.3912 Ops/s 48.7295 Ops/s $\color{#35bf28}+1.36\%$
test_iql_speed[False-backward] 30.5177ms 30.0371ms 33.2922 Ops/s 32.8832 Ops/s $\color{#35bf28}+1.24\%$
test_iql_speed[True-None] 8.7981ms 8.4466ms 118.3906 Ops/s 115.1860 Ops/s $\color{#35bf28}+2.78\%$
test_iql_speed[True-backward] 16.6226ms 16.3355ms 61.2163 Ops/s 60.5190 Ops/s $\color{#35bf28}+1.15\%$
test_iql_speed[reduce-overhead-None] 9.0621ms 8.4909ms 117.7730 Ops/s 114.0149 Ops/s $\color{#35bf28}+3.30\%$
test_iql_speed[reduce-overhead-backward] 17.1872ms 16.8171ms 59.4632 Ops/s 58.4971 Ops/s $\color{#35bf28}+1.65\%$
test_rb_sample[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 8.1469ms 6.2116ms 160.9900 Ops/s 163.7489 Ops/s $\color{#d91a1a}-1.68\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 0.5772ms 0.3727ms 2.6829 KOps/s 3.1586 KOps/s $\textbf{\color{#d91a1a}-15.06\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.6425ms 0.3506ms 2.8521 KOps/s 3.4127 KOps/s $\textbf{\color{#d91a1a}-16.43\%}$
test_rb_sample[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 6.2128ms 5.8933ms 169.6848 Ops/s 170.3799 Ops/s $\color{#d91a1a}-0.41\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 1.0369ms 0.3267ms 3.0607 KOps/s 3.3530 KOps/s $\textbf{\color{#d91a1a}-8.72\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.5648ms 0.3174ms 3.1503 KOps/s 3.1957 KOps/s $\color{#d91a1a}-1.42\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-sampler6-10000] 1.7414ms 1.4267ms 700.9404 Ops/s 774.4578 Ops/s $\textbf{\color{#d91a1a}-9.49\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-sampler7-10000] 1.6820ms 1.3434ms 744.3682 Ops/s 794.6014 Ops/s $\textbf{\color{#d91a1a}-6.32\%}$
test_rb_sample[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 10.0827ms 6.1599ms 162.3414 Ops/s 165.4051 Ops/s $\color{#d91a1a}-1.85\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 2.3034ms 0.4807ms 2.0802 KOps/s 2.2028 KOps/s $\textbf{\color{#d91a1a}-5.56\%}$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.7056ms 0.4749ms 2.1057 KOps/s 2.2585 KOps/s $\textbf{\color{#d91a1a}-6.77\%}$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 6.0747ms 5.8505ms 170.9261 Ops/s 169.2106 Ops/s $\color{#35bf28}+1.01\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 0.7586ms 0.3510ms 2.8493 KOps/s 2.9524 KOps/s $\color{#d91a1a}-3.49\%$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.6669ms 0.3099ms 3.2268 KOps/s 2.9142 KOps/s $\textbf{\color{#35bf28}+10.73\%}$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 6.0589ms 5.7422ms 174.1501 Ops/s 170.2435 Ops/s $\color{#35bf28}+2.29\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 0.7336ms 0.3171ms 3.1533 KOps/s 2.9048 KOps/s $\textbf{\color{#35bf28}+8.55\%}$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.5840ms 0.2843ms 3.5170 KOps/s 3.8051 KOps/s $\textbf{\color{#d91a1a}-7.57\%}$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 6.1120ms 5.9195ms 168.9343 Ops/s 165.5756 Ops/s $\color{#35bf28}+2.03\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 1.9932ms 0.4669ms 2.1417 KOps/s 2.2707 KOps/s $\textbf{\color{#d91a1a}-5.68\%}$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.5964ms 0.4104ms 2.4365 KOps/s 2.3989 KOps/s $\color{#35bf28}+1.57\%$
test_rb_populate[TensorDictReplayBuffer-ListStorage-RandomSampler-400] 6.5863ms 5.0240ms 199.0428 Ops/s 197.3371 Ops/s $\color{#35bf28}+0.86\%$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-400] 8.3427ms 2.0885ms 478.8183 Ops/s 407.8615 Ops/s $\textbf{\color{#35bf28}+17.40\%}$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-400] 8.2880ms 1.2218ms 818.4394 Ops/s 924.2481 Ops/s $\textbf{\color{#d91a1a}-11.45\%}$
test_rb_populate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-400] 6.8272ms 5.0393ms 198.4413 Ops/s 53.2507 Ops/s $\textbf{\color{#35bf28}+272.65\%}$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-400] 8.6305ms 2.0697ms 483.1562 Ops/s 695.2910 Ops/s $\textbf{\color{#d91a1a}-30.51\%}$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-400] 7.0956ms 1.1976ms 835.0040 Ops/s 776.9930 Ops/s $\textbf{\color{#35bf28}+7.47\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-ListStorage-None-400] 0.6157s 17.4348ms 57.3565 Ops/s 187.2122 Ops/s $\textbf{\color{#d91a1a}-69.36\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-400] 12.2227ms 2.1230ms 471.0289 Ops/s 446.8677 Ops/s $\textbf{\color{#35bf28}+5.41\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-400] 1.1798ms 1.0137ms 986.5057 Ops/s 751.2824 Ops/s $\textbf{\color{#35bf28}+31.31\%}$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-True] 37.1926ms 34.2661ms 29.1834 Ops/s 28.6648 Ops/s $\color{#35bf28}+1.81\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-False] 19.3442ms 17.6404ms 56.6880 Ops/s 54.7289 Ops/s $\color{#35bf28}+3.58\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-True] 36.6475ms 34.5979ms 28.9035 Ops/s 28.0128 Ops/s $\color{#35bf28}+3.18\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-False] 19.2607ms 17.6558ms 56.6387 Ops/s 53.1006 Ops/s $\textbf{\color{#35bf28}+6.66\%}$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-True] 38.8289ms 36.9371ms 27.0731 Ops/s 26.6684 Ops/s $\color{#35bf28}+1.52\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-False] 20.9063ms 19.4647ms 51.3752 Ops/s 50.6348 Ops/s $\color{#35bf28}+1.46\%$

vmoens added a commit that referenced this pull request Jan 1, 2026
@vmoens vmoens merged commit dc95e15 into gh/vmoens/169/base Jan 1, 2026
102 of 109 checks passed
@vmoens vmoens deleted the gh/vmoens/169/head branch January 1, 2026 07:25
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