Skip to content

Conversation

@vmoens
Copy link
Collaborator

@vmoens vmoens commented Dec 31, 2025

Stack from ghstack (oldest at bottom):

[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/3286

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

As of commit 1456121 with merge base 7866d11 (image):

NEW FAILURES - The following jobs have failed:

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

vmoens added a commit that referenced this pull request Dec 31, 2025
ghstack-source-id: 2e1839f
Pull-Request: #3286
@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

github-actions bot commented Dec 31, 2025

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

Total Benchmarks: 164. Improved: $\large\color{#35bf28}23$. Worsened: $\large\color{#d91a1a}6$.

Expand to view detailed results
Name Max Mean Ops Ops on Repo HEAD Change
test_tensor_to_bytestream_speed[pickle] 83.9984μs 82.0294μs 12.1908 KOps/s 12.2188 KOps/s $\color{#d91a1a}-0.23\%$
test_tensor_to_bytestream_speed[torch.save] 0.1397ms 0.1392ms 7.1857 KOps/s 7.1154 KOps/s $\color{#35bf28}+0.99\%$
test_tensor_to_bytestream_speed[untyped_storage] 0.1202s 0.1200s 8.3358 Ops/s 8.2009 Ops/s $\color{#35bf28}+1.65\%$
test_tensor_to_bytestream_speed[numpy] 2.6505μs 2.6396μs 378.8417 KOps/s 367.0778 KOps/s $\color{#35bf28}+3.20\%$
test_tensor_to_bytestream_speed[safetensors] 41.6740μs 41.4557μs 24.1221 KOps/s 26.3680 KOps/s $\textbf{\color{#d91a1a}-8.52\%}$
test_simple 0.5618s 0.5598s 1.7862 Ops/s 1.7356 Ops/s $\color{#35bf28}+2.92\%$
test_transformed 1.1554s 1.1425s 0.8753 Ops/s 0.8711 Ops/s $\color{#35bf28}+0.47\%$
test_serial 1.6896s 1.6761s 0.5966 Ops/s 0.5928 Ops/s $\color{#35bf28}+0.65\%$
test_parallel 1.2748s 1.2460s 0.8026 Ops/s 0.8267 Ops/s $\color{#d91a1a}-2.92\%$
test_step_mdp_speed[True-True-True-True-True] 0.3080ms 47.4214μs 21.0875 KOps/s 22.1381 KOps/s $\color{#d91a1a}-4.75\%$
test_step_mdp_speed[True-True-True-True-False] 49.7600μs 25.6838μs 38.9351 KOps/s 39.6071 KOps/s $\color{#d91a1a}-1.70\%$
test_step_mdp_speed[True-True-True-False-True] 55.7600μs 25.6433μs 38.9965 KOps/s 38.6430 KOps/s $\color{#35bf28}+0.91\%$
test_step_mdp_speed[True-True-True-False-False] 41.6500μs 14.6409μs 68.3016 KOps/s 70.9462 KOps/s $\color{#d91a1a}-3.73\%$
test_step_mdp_speed[True-True-False-True-True] 80.4910μs 48.9481μs 20.4298 KOps/s 20.4096 KOps/s $\color{#35bf28}+0.10\%$
test_step_mdp_speed[True-True-False-True-False] 52.5410μs 28.3063μs 35.3279 KOps/s 35.5326 KOps/s $\color{#d91a1a}-0.58\%$
test_step_mdp_speed[True-True-False-False-True] 57.3510μs 28.4439μs 35.1569 KOps/s 35.0973 KOps/s $\color{#35bf28}+0.17\%$
test_step_mdp_speed[True-True-False-False-False] 36.2110μs 17.1875μs 58.1819 KOps/s 59.6041 KOps/s $\color{#d91a1a}-2.39\%$
test_step_mdp_speed[True-False-True-True-True] 86.4310μs 52.1379μs 19.1799 KOps/s 19.4857 KOps/s $\color{#d91a1a}-1.57\%$
test_step_mdp_speed[True-False-True-True-False] 63.8110μs 31.1571μs 32.0954 KOps/s 32.7648 KOps/s $\color{#d91a1a}-2.04\%$
test_step_mdp_speed[True-False-True-False-True] 64.5310μs 28.2608μs 35.3847 KOps/s 35.8326 KOps/s $\color{#d91a1a}-1.25\%$
test_step_mdp_speed[True-False-True-False-False] 82.4310μs 17.2278μs 58.0456 KOps/s 59.3980 KOps/s $\color{#d91a1a}-2.28\%$
test_step_mdp_speed[True-False-False-True-True] 97.7110μs 52.9588μs 18.8826 KOps/s 18.8001 KOps/s $\color{#35bf28}+0.44\%$
test_step_mdp_speed[True-False-False-True-False] 72.3710μs 33.2518μs 30.0735 KOps/s 30.2893 KOps/s $\color{#d91a1a}-0.71\%$
test_step_mdp_speed[True-False-False-False-True] 56.3010μs 30.3182μs 32.9835 KOps/s 32.9276 KOps/s $\color{#35bf28}+0.17\%$
test_step_mdp_speed[True-False-False-False-False] 47.7510μs 19.7033μs 50.7530 KOps/s 51.0907 KOps/s $\color{#d91a1a}-0.66\%$
test_step_mdp_speed[False-True-True-True-True] 0.1117ms 50.9534μs 19.6258 KOps/s 19.6226 KOps/s $\color{#35bf28}+0.02\%$
test_step_mdp_speed[False-True-True-True-False] 58.0900μs 30.6385μs 32.6387 KOps/s 32.1169 KOps/s $\color{#35bf28}+1.62\%$
test_step_mdp_speed[False-True-True-False-True] 2.3948ms 32.3638μs 30.8987 KOps/s 30.5109 KOps/s $\color{#35bf28}+1.27\%$
test_step_mdp_speed[False-True-True-False-False] 50.5110μs 19.0954μs 52.3687 KOps/s 53.1001 KOps/s $\color{#d91a1a}-1.38\%$
test_step_mdp_speed[False-True-False-True-True] 81.7610μs 53.1829μs 18.8030 KOps/s 18.6618 KOps/s $\color{#35bf28}+0.76\%$
test_step_mdp_speed[False-True-False-True-False] 67.1110μs 33.5893μs 29.7714 KOps/s 29.9639 KOps/s $\color{#d91a1a}-0.64\%$
test_step_mdp_speed[False-True-False-False-True] 94.7710μs 33.5310μs 29.8232 KOps/s 28.8094 KOps/s $\color{#35bf28}+3.52\%$
test_step_mdp_speed[False-True-False-False-False] 63.3210μs 21.0163μs 47.5821 KOps/s 47.1237 KOps/s $\color{#35bf28}+0.97\%$
test_step_mdp_speed[False-False-True-True-True] 88.5510μs 55.4470μs 18.0353 KOps/s 17.7592 KOps/s $\color{#35bf28}+1.55\%$
test_step_mdp_speed[False-False-True-True-False] 61.4000μs 35.6639μs 28.0396 KOps/s 27.6972 KOps/s $\color{#35bf28}+1.24\%$
test_step_mdp_speed[False-False-True-False-True] 63.3800μs 33.8877μs 29.5092 KOps/s 29.1913 KOps/s $\color{#35bf28}+1.09\%$
test_step_mdp_speed[False-False-True-False-False] 82.6910μs 20.8306μs 48.0062 KOps/s 47.6410 KOps/s $\color{#35bf28}+0.77\%$
test_step_mdp_speed[False-False-False-True-True] 93.0010μs 58.2279μs 17.1739 KOps/s 16.9849 KOps/s $\color{#35bf28}+1.11\%$
test_step_mdp_speed[False-False-False-True-False] 69.2810μs 38.2377μs 26.1522 KOps/s 25.8625 KOps/s $\color{#35bf28}+1.12\%$
test_step_mdp_speed[False-False-False-False-True] 71.8610μs 36.3807μs 27.4871 KOps/s 26.9605 KOps/s $\color{#35bf28}+1.95\%$
test_step_mdp_speed[False-False-False-False-False] 49.9410μs 23.2722μs 42.9698 KOps/s 42.5925 KOps/s $\color{#35bf28}+0.89\%$
test_non_tensor_env_rollout_speed[1000-single-True] 0.8977s 0.7976s 1.2538 Ops/s 1.2807 Ops/s $\color{#d91a1a}-2.11\%$
test_non_tensor_env_rollout_speed[1000-single-False] 0.7483s 0.6559s 1.5246 Ops/s 1.5582 Ops/s $\color{#d91a1a}-2.16\%$
test_non_tensor_env_rollout_speed[1000-serial-no-buffers-True] 1.7910s 1.7082s 0.5854 Ops/s 0.5866 Ops/s $\color{#d91a1a}-0.20\%$
test_non_tensor_env_rollout_speed[1000-serial-no-buffers-False] 1.5663s 1.4848s 0.6735 Ops/s 0.6757 Ops/s $\color{#d91a1a}-0.32\%$
test_non_tensor_env_rollout_speed[1000-serial-buffers-True] 2.0396s 1.9511s 0.5125 Ops/s 0.5136 Ops/s $\color{#d91a1a}-0.20\%$
test_non_tensor_env_rollout_speed[1000-serial-buffers-False] 1.8149s 1.7294s 0.5782 Ops/s 0.5816 Ops/s $\color{#d91a1a}-0.57\%$
test_non_tensor_env_rollout_speed[1000-parallel-no-buffers-True] 4.9040s 4.7309s 0.2114 Ops/s 0.2179 Ops/s $\color{#d91a1a}-3.00\%$
test_non_tensor_env_rollout_speed[1000-parallel-no-buffers-False] 4.5392s 4.4448s 0.2250 Ops/s 0.2207 Ops/s $\color{#35bf28}+1.94\%$
test_non_tensor_env_rollout_speed[1000-parallel-buffers-True] 2.0703s 2.0199s 0.4951 Ops/s 0.5061 Ops/s $\color{#d91a1a}-2.18\%$
test_non_tensor_env_rollout_speed[1000-parallel-buffers-False] 1.7759s 1.6925s 0.5908 Ops/s 0.5909 Ops/s $-0.01\%$
test_values[generalized_advantage_estimate-True-True] 11.2730ms 10.3121ms 96.9732 Ops/s 98.7071 Ops/s $\color{#d91a1a}-1.76\%$
test_values[vec_generalized_advantage_estimate-True-True] 13.3379ms 11.1349ms 89.8076 Ops/s 90.6484 Ops/s $\color{#d91a1a}-0.93\%$
test_values[td0_return_estimate-False-False] 0.2421ms 0.1333ms 7.5029 KOps/s 7.5948 KOps/s $\color{#d91a1a}-1.21\%$
test_values[td1_return_estimate-False-False] 28.4595ms 27.9573ms 35.7689 Ops/s 36.2946 Ops/s $\color{#d91a1a}-1.45\%$
test_values[vec_td1_return_estimate-False-False] 19.1128ms 11.6548ms 85.8015 Ops/s 89.7044 Ops/s $\color{#d91a1a}-4.35\%$
test_values[td_lambda_return_estimate-True-False] 43.6400ms 41.6498ms 24.0097 Ops/s 24.6958 Ops/s $\color{#d91a1a}-2.78\%$
test_values[vec_td_lambda_return_estimate-True-False] 11.6117ms 11.2640ms 88.7783 Ops/s 90.0785 Ops/s $\color{#d91a1a}-1.44\%$
test_gae_speed[generalized_advantage_estimate-False-1-512] 9.3434ms 9.0829ms 110.0971 Ops/s 111.6723 Ops/s $\color{#d91a1a}-1.41\%$
test_gae_speed[vec_generalized_advantage_estimate-True-1-512] 1.7106ms 1.5183ms 658.6384 Ops/s 663.7972 Ops/s $\color{#d91a1a}-0.78\%$
test_gae_speed[vec_generalized_advantage_estimate-False-1-512] 0.5198ms 0.4172ms 2.3967 KOps/s 2.3268 KOps/s $\color{#35bf28}+3.00\%$
test_gae_speed[vec_generalized_advantage_estimate-True-32-512] 30.7588ms 25.4849ms 39.2390 Ops/s 41.0954 Ops/s $\color{#d91a1a}-4.52\%$
test_gae_speed[vec_generalized_advantage_estimate-False-32-512] 2.1756ms 1.7491ms 571.7204 Ops/s 581.6291 Ops/s $\color{#d91a1a}-1.70\%$
test_dqn_speed[False-None] 1.7982ms 1.4314ms 698.6087 Ops/s 681.9665 Ops/s $\color{#35bf28}+2.44\%$
test_dqn_speed[False-backward] 2.0674ms 1.9669ms 508.4092 Ops/s 503.1916 Ops/s $\color{#35bf28}+1.04\%$
test_dqn_speed[True-None] 0.9449ms 0.5409ms 1.8486 KOps/s 1.8645 KOps/s $\color{#d91a1a}-0.86\%$
test_dqn_speed[True-backward] 1.0611ms 0.9920ms 1.0081 KOps/s 923.4789 Ops/s $\textbf{\color{#35bf28}+9.16\%}$
test_dqn_speed[reduce-overhead-None] 0.9619ms 0.5355ms 1.8674 KOps/s 1.8830 KOps/s $\color{#d91a1a}-0.83\%$
test_dqn_speed[reduce-overhead-backward] 1.0420ms 0.9810ms 1.0194 KOps/s 1.0091 KOps/s $\color{#35bf28}+1.02\%$
test_ddpg_speed[False-None] 3.3206ms 2.8846ms 346.6689 Ops/s 340.2888 Ops/s $\color{#35bf28}+1.87\%$
test_ddpg_speed[False-backward] 4.2035ms 4.1162ms 242.9403 Ops/s 238.0876 Ops/s $\color{#35bf28}+2.04\%$
test_ddpg_speed[True-None] 1.7977ms 1.3991ms 714.7395 Ops/s 702.2985 Ops/s $\color{#35bf28}+1.77\%$
test_ddpg_speed[True-backward] 2.4512ms 2.3994ms 416.7771 Ops/s 411.9284 Ops/s $\color{#35bf28}+1.18\%$
test_ddpg_speed[reduce-overhead-None] 1.8361ms 1.3908ms 719.0004 Ops/s 707.6146 Ops/s $\color{#35bf28}+1.61\%$
test_ddpg_speed[reduce-overhead-backward] 2.4797ms 2.3681ms 422.2755 Ops/s 345.1494 Ops/s $\textbf{\color{#35bf28}+22.35\%}$
test_sac_speed[False-None] 8.4973ms 7.9605ms 125.6206 Ops/s 121.1728 Ops/s $\color{#35bf28}+3.67\%$
test_sac_speed[False-backward] 11.7017ms 11.2445ms 88.9320 Ops/s 86.5308 Ops/s $\color{#35bf28}+2.78\%$
test_sac_speed[True-None] 2.9459ms 2.2060ms 453.3008 Ops/s 457.4558 Ops/s $\color{#d91a1a}-0.91\%$
test_sac_speed[True-backward] 4.1798ms 4.0180ms 248.8813 Ops/s 227.5528 Ops/s $\textbf{\color{#35bf28}+9.37\%}$
test_sac_speed[reduce-overhead-None] 2.5096ms 2.1228ms 471.0682 Ops/s 462.0636 Ops/s $\color{#35bf28}+1.95\%$
test_sac_speed[reduce-overhead-backward] 4.1119ms 3.9824ms 251.1063 Ops/s 229.1172 Ops/s $\textbf{\color{#35bf28}+9.60\%}$
test_redq_speed[False-None] 10.9471ms 10.3327ms 96.7803 Ops/s 94.9736 Ops/s $\color{#35bf28}+1.90\%$
test_redq_speed[False-backward] 18.6749ms 17.7705ms 56.2730 Ops/s 56.2174 Ops/s $\color{#35bf28}+0.10\%$
test_redq_speed[True-None] 4.4940ms 4.2507ms 235.2580 Ops/s 215.9251 Ops/s $\textbf{\color{#35bf28}+8.95\%}$
test_redq_speed[True-backward] 9.8164ms 9.5270ms 104.9651 Ops/s 101.7467 Ops/s $\color{#35bf28}+3.16\%$
test_redq_speed[reduce-overhead-None] 4.6154ms 4.1745ms 239.5471 Ops/s 218.3156 Ops/s $\textbf{\color{#35bf28}+9.73\%}$
test_redq_speed[reduce-overhead-backward] 9.8957ms 9.6080ms 104.0801 Ops/s 102.8033 Ops/s $\color{#35bf28}+1.24\%$
test_redq_deprec_speed[False-None] 11.6713ms 11.0605ms 90.4122 Ops/s 88.7595 Ops/s $\color{#35bf28}+1.86\%$
test_redq_deprec_speed[False-backward] 16.3989ms 15.9289ms 62.7791 Ops/s 61.7045 Ops/s $\color{#35bf28}+1.74\%$
test_redq_deprec_speed[True-None] 4.1537ms 3.6493ms 274.0252 Ops/s 270.6261 Ops/s $\color{#35bf28}+1.26\%$
test_redq_deprec_speed[True-backward] 7.8436ms 7.5489ms 132.4695 Ops/s 133.9362 Ops/s $\color{#d91a1a}-1.10\%$
test_redq_deprec_speed[reduce-overhead-None] 3.9037ms 3.6969ms 270.4962 Ops/s 254.7810 Ops/s $\textbf{\color{#35bf28}+6.17\%}$
test_redq_deprec_speed[reduce-overhead-backward] 7.7167ms 7.5438ms 132.5588 Ops/s 125.9723 Ops/s $\textbf{\color{#35bf28}+5.23\%}$
test_td3_speed[False-None] 8.1727ms 8.0295ms 124.5414 Ops/s 122.1858 Ops/s $\color{#35bf28}+1.93\%$
test_td3_speed[False-backward] 11.3719ms 10.9445ms 91.3704 Ops/s 90.6023 Ops/s $\color{#35bf28}+0.85\%$
test_td3_speed[True-None] 1.8773ms 1.8144ms 551.1585 Ops/s 547.7006 Ops/s $\color{#35bf28}+0.63\%$
test_td3_speed[True-backward] 3.7764ms 3.6213ms 276.1407 Ops/s 276.5910 Ops/s $\color{#d91a1a}-0.16\%$
test_td3_speed[reduce-overhead-None] 1.8397ms 1.7999ms 555.5782 Ops/s 553.5245 Ops/s $\color{#35bf28}+0.37\%$
test_td3_speed[reduce-overhead-backward] 3.7331ms 3.6698ms 272.4942 Ops/s 265.4913 Ops/s $\color{#35bf28}+2.64\%$
test_cql_speed[False-None] 30.2381ms 26.0700ms 38.3583 Ops/s 38.6910 Ops/s $\color{#d91a1a}-0.86\%$
test_cql_speed[False-backward] 35.5596ms 35.0104ms 28.5630 Ops/s 27.8875 Ops/s $\color{#35bf28}+2.42\%$
test_cql_speed[True-None] 16.2399ms 12.4569ms 80.2771 Ops/s 78.3510 Ops/s $\color{#35bf28}+2.46\%$
test_cql_speed[True-backward] 18.3417ms 17.8599ms 55.9915 Ops/s 54.1826 Ops/s $\color{#35bf28}+3.34\%$
test_cql_speed[reduce-overhead-None] 13.0249ms 12.5134ms 79.9142 Ops/s 78.8265 Ops/s $\color{#35bf28}+1.38\%$
test_cql_speed[reduce-overhead-backward] 18.6222ms 18.0871ms 55.2879 Ops/s 53.9568 Ops/s $\color{#35bf28}+2.47\%$
test_a2c_speed[False-None] 5.6636ms 5.4572ms 183.2430 Ops/s 179.1772 Ops/s $\color{#35bf28}+2.27\%$
test_a2c_speed[False-backward] 12.5210ms 11.9567ms 83.6354 Ops/s 81.5924 Ops/s $\color{#35bf28}+2.50\%$
test_a2c_speed[True-None] 4.1385ms 3.6711ms 272.3967 Ops/s 256.7748 Ops/s $\textbf{\color{#35bf28}+6.08\%}$
test_a2c_speed[True-backward] 8.8862ms 8.6017ms 116.2558 Ops/s 115.6625 Ops/s $\color{#35bf28}+0.51\%$
test_a2c_speed[reduce-overhead-None] 3.8610ms 3.6927ms 270.8036 Ops/s 270.6467 Ops/s $\color{#35bf28}+0.06\%$
test_a2c_speed[reduce-overhead-backward] 8.8743ms 8.7111ms 114.7964 Ops/s 110.8668 Ops/s $\color{#35bf28}+3.54\%$
test_ppo_speed[False-None] 6.1588ms 5.8878ms 169.8435 Ops/s 165.5064 Ops/s $\color{#35bf28}+2.62\%$
test_ppo_speed[False-backward] 12.9822ms 12.6262ms 79.2004 Ops/s 77.6746 Ops/s $\color{#35bf28}+1.96\%$
test_ppo_speed[True-None] 3.8508ms 3.6302ms 275.4687 Ops/s 270.6278 Ops/s $\color{#35bf28}+1.79\%$
test_ppo_speed[True-backward] 11.2446ms 8.9485ms 111.7503 Ops/s 115.3090 Ops/s $\color{#d91a1a}-3.09\%$
test_ppo_speed[reduce-overhead-None] 3.7077ms 3.5764ms 279.6139 Ops/s 271.2700 Ops/s $\color{#35bf28}+3.08\%$
test_ppo_speed[reduce-overhead-backward] 9.6175ms 8.7564ms 114.2019 Ops/s 112.8829 Ops/s $\color{#35bf28}+1.17\%$
test_reinforce_speed[False-None] 4.8216ms 4.5537ms 219.6027 Ops/s 209.2986 Ops/s $\color{#35bf28}+4.92\%$
test_reinforce_speed[False-backward] 7.7249ms 7.4689ms 133.8891 Ops/s 130.9218 Ops/s $\color{#35bf28}+2.27\%$
test_reinforce_speed[True-None] 3.0700ms 2.8892ms 346.1166 Ops/s 335.2149 Ops/s $\color{#35bf28}+3.25\%$
test_reinforce_speed[True-backward] 7.9103ms 7.6607ms 130.5361 Ops/s 108.6434 Ops/s $\textbf{\color{#35bf28}+20.15\%}$
test_reinforce_speed[reduce-overhead-None] 3.0987ms 2.8839ms 346.7474 Ops/s 343.9487 Ops/s $\color{#35bf28}+0.81\%$
test_reinforce_speed[reduce-overhead-backward] 8.1260ms 7.9327ms 126.0606 Ops/s 121.0732 Ops/s $\color{#35bf28}+4.12\%$
test_iql_speed[False-None] 25.7838ms 20.2681ms 49.3386 Ops/s 49.0634 Ops/s $\color{#35bf28}+0.56\%$
test_iql_speed[False-backward] 36.5496ms 30.7443ms 32.5264 Ops/s 32.3721 Ops/s $\color{#35bf28}+0.48\%$
test_iql_speed[True-None] 8.9073ms 8.5444ms 117.0362 Ops/s 111.2513 Ops/s $\textbf{\color{#35bf28}+5.20\%}$
test_iql_speed[True-backward] 16.9602ms 16.6945ms 59.8999 Ops/s 58.6019 Ops/s $\color{#35bf28}+2.21\%$
test_iql_speed[reduce-overhead-None] 9.0600ms 8.6675ms 115.3734 Ops/s 108.2115 Ops/s $\textbf{\color{#35bf28}+6.62\%}$
test_iql_speed[reduce-overhead-backward] 17.6766ms 17.2836ms 57.8582 Ops/s 57.2660 Ops/s $\color{#35bf28}+1.03\%$
test_rb_sample[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 8.0117ms 6.0747ms 164.6180 Ops/s 163.4512 Ops/s $\color{#35bf28}+0.71\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 0.5579ms 0.3294ms 3.0355 KOps/s 2.7616 KOps/s $\textbf{\color{#35bf28}+9.92\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.5117ms 0.2740ms 3.6492 KOps/s 2.8675 KOps/s $\textbf{\color{#35bf28}+27.26\%}$
test_rb_sample[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 6.0688ms 5.7829ms 172.9231 Ops/s 170.7529 Ops/s $\color{#35bf28}+1.27\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 2.5557ms 0.2833ms 3.5301 KOps/s 2.9513 KOps/s $\textbf{\color{#35bf28}+19.61\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.4967ms 0.2637ms 3.7915 KOps/s 3.2833 KOps/s $\textbf{\color{#35bf28}+15.48\%}$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-sampler6-10000] 1.4820ms 1.2667ms 789.4704 Ops/s 728.5962 Ops/s $\textbf{\color{#35bf28}+8.35\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-sampler7-10000] 1.3262ms 1.1846ms 844.1766 Ops/s 780.5040 Ops/s $\textbf{\color{#35bf28}+8.16\%}$
test_rb_sample[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 9.5934ms 6.0932ms 164.1180 Ops/s 166.7186 Ops/s $\color{#d91a1a}-1.56\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 1.0912ms 0.4758ms 2.1019 KOps/s 2.2647 KOps/s $\textbf{\color{#d91a1a}-7.19\%}$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.7238ms 0.4932ms 2.0277 KOps/s 2.3647 KOps/s $\textbf{\color{#d91a1a}-14.25\%}$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 5.8272ms 5.7417ms 174.1637 Ops/s 169.8007 Ops/s $\color{#35bf28}+2.57\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 0.7322ms 0.2854ms 3.5033 KOps/s 3.4394 KOps/s $\color{#35bf28}+1.86\%$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.4838ms 0.2648ms 3.7769 KOps/s 3.7292 KOps/s $\color{#35bf28}+1.28\%$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 5.9209ms 5.6224ms 177.8593 Ops/s 171.9748 Ops/s $\color{#35bf28}+3.42\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 0.6267ms 0.2832ms 3.5308 KOps/s 2.8994 KOps/s $\textbf{\color{#35bf28}+21.78\%}$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.4679ms 0.2667ms 3.7491 KOps/s 3.0609 KOps/s $\textbf{\color{#35bf28}+22.48\%}$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 6.0028ms 5.9154ms 169.0508 Ops/s 167.3481 Ops/s $\color{#35bf28}+1.02\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 2.2825ms 0.4394ms 2.2760 KOps/s 2.1458 KOps/s $\textbf{\color{#35bf28}+6.07\%}$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.6319ms 0.4217ms 2.3712 KOps/s 2.3391 KOps/s $\color{#35bf28}+1.37\%$
test_rb_populate[TensorDictReplayBuffer-ListStorage-RandomSampler-400] 6.4654ms 5.0277ms 198.8979 Ops/s 194.8820 Ops/s $\color{#35bf28}+2.06\%$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-400] 7.8710ms 2.3168ms 431.6299 Ops/s 415.6990 Ops/s $\color{#35bf28}+3.83\%$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-400] 8.0390ms 1.2590ms 794.2724 Ops/s 861.1135 Ops/s $\textbf{\color{#d91a1a}-7.76\%}$
test_rb_populate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-400] 7.1186ms 5.0930ms 196.3483 Ops/s 52.3888 Ops/s $\textbf{\color{#35bf28}+274.79\%}$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-400] 0.6161s 14.4793ms 69.0643 Ops/s 690.3321 Ops/s $\textbf{\color{#d91a1a}-90.00\%}$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-400] 7.3459ms 1.2218ms 818.4367 Ops/s 909.1419 Ops/s $\textbf{\color{#d91a1a}-9.98\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-ListStorage-None-400] 7.0243ms 5.2881ms 189.1024 Ops/s 188.5264 Ops/s $\color{#35bf28}+0.31\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-400] 3.9766ms 1.7555ms 569.6288 Ops/s 444.7776 Ops/s $\textbf{\color{#35bf28}+28.07\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-400] 7.5048ms 1.4058ms 711.3404 Ops/s 707.7314 Ops/s $\color{#35bf28}+0.51\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-True] 36.0752ms 33.5392ms 29.8158 Ops/s 29.4080 Ops/s $\color{#35bf28}+1.39\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-False] 19.6079ms 17.9448ms 55.7266 Ops/s 56.4136 Ops/s $\color{#d91a1a}-1.22\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-True] 36.8752ms 34.6187ms 28.8861 Ops/s 28.5822 Ops/s $\color{#35bf28}+1.06\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-False] 19.5288ms 18.1021ms 55.2423 Ops/s 55.8573 Ops/s $\color{#d91a1a}-1.10\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-True] 43.1199ms 36.4842ms 27.4091 Ops/s 27.1173 Ops/s $\color{#35bf28}+1.08\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-False] 20.4720ms 19.5137ms 51.2461 Ops/s 51.6886 Ops/s $\color{#d91a1a}-0.86\%$

@vmoens vmoens added the CI Has to do with CI setup (e.g. wheels & builds, tests...) label Dec 31, 2025
[ghstack-poisoned]
vmoens added a commit that referenced this pull request Dec 31, 2025
ghstack-source-id: 7c92a6d
Pull-Request: #3286
[ghstack-poisoned]
vmoens added a commit that referenced this pull request Dec 31, 2025
ghstack-source-id: d5fae60
Pull-Request: #3286
[ghstack-poisoned]
@vmoens vmoens mentioned this pull request Jan 1, 2026
[ghstack-poisoned]
vmoens added a commit that referenced this pull request Jan 1, 2026
ghstack-source-id: 34cd3ab
Pull-Request: #3286

amend

ghstack-source-id: 34cd3ab
Pull-Request: #3287
[ghstack-poisoned]
vmoens added a commit that referenced this pull request Jan 1, 2026
ghstack-source-id: b6179e0
Pull-Request: #3286

amend

ghstack-source-id: b6179e0
Pull-Request: #3287
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

CI Has to do with CI setup (e.g. wheels & builds, tests...) 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