Skip to content

Conversation

@vmoens
Copy link
Collaborator

@vmoens vmoens commented Dec 18, 2025

Stack from ghstack (oldest at bottom):

[ghstack-poisoned]
vmoens added a commit that referenced this pull request Dec 18, 2025
ghstack-source-id: 0016ea2
Pull-Request: #3265
@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/3265

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

❌ 6 New Failures, 3 Cancelled Jobs

As of commit 7b645f5 with merge base 5c5992d (image):

NEW FAILURES - The following jobs have failed:

CANCELLED JOBS - The following jobs were cancelled. Please retry:

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

@vmoens vmoens mentioned this pull request Dec 18, 2025
@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]
vmoens added a commit that referenced this pull request Dec 18, 2025
ghstack-source-id: 81a3c13
Pull-Request: #3265
@vmoens vmoens added the Tests Incomplete or broken unit tests label Dec 18, 2025
@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: $\large\color{#35bf28}30$. 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] 82.6210μs 80.9920μs 12.3469 KOps/s 11.6016 KOps/s $\textbf{\color{#35bf28}+6.42\%}$
test_tensor_to_bytestream_speed[torch.save] 0.1409ms 0.1404ms 7.1222 KOps/s 6.7831 KOps/s $\color{#35bf28}+5.00\%$
test_tensor_to_bytestream_speed[untyped_storage] 0.1203s 0.1200s 8.3359 Ops/s 8.0423 Ops/s $\color{#35bf28}+3.65\%$
test_tensor_to_bytestream_speed[numpy] 2.6467μs 2.6432μs 378.3235 KOps/s 364.0917 KOps/s $\color{#35bf28}+3.91\%$
test_tensor_to_bytestream_speed[safetensors] 39.0410μs 38.6474μs 25.8750 KOps/s 24.8851 KOps/s $\color{#35bf28}+3.98\%$
test_simple 0.5585s 0.5547s 1.8029 Ops/s 1.7163 Ops/s $\textbf{\color{#35bf28}+5.04\%}$
test_transformed 1.1337s 1.1307s 0.8844 Ops/s 0.8431 Ops/s $\color{#35bf28}+4.90\%$
test_serial 1.6989s 1.6928s 0.5907 Ops/s 0.5699 Ops/s $\color{#35bf28}+3.66\%$
test_parallel 1.2319s 1.1893s 0.8408 Ops/s 0.8859 Ops/s $\textbf{\color{#d91a1a}-5.09\%}$
test_step_mdp_speed[True-True-True-True-True] 0.1603ms 43.8530μs 22.8035 KOps/s 22.1992 KOps/s $\color{#35bf28}+2.72\%$
test_step_mdp_speed[True-True-True-True-False] 56.0110μs 24.7439μs 40.4139 KOps/s 40.5520 KOps/s $\color{#d91a1a}-0.34\%$
test_step_mdp_speed[True-True-True-False-True] 58.3110μs 24.9133μs 40.1393 KOps/s 40.0465 KOps/s $\color{#35bf28}+0.23\%$
test_step_mdp_speed[True-True-True-False-False] 43.0210μs 13.8248μs 72.3340 KOps/s 73.4895 KOps/s $\color{#d91a1a}-1.57\%$
test_step_mdp_speed[True-True-False-True-True] 93.8210μs 46.8405μs 21.3490 KOps/s 21.2605 KOps/s $\color{#35bf28}+0.42\%$
test_step_mdp_speed[True-True-False-True-False] 55.3300μs 27.2586μs 36.6856 KOps/s 36.6878 KOps/s $-0.01\%$
test_step_mdp_speed[True-True-False-False-True] 73.4910μs 27.1958μs 36.7704 KOps/s 36.0210 KOps/s $\color{#35bf28}+2.08\%$
test_step_mdp_speed[True-True-False-False-False] 51.1910μs 16.3908μs 61.0100 KOps/s 60.7558 KOps/s $\color{#35bf28}+0.42\%$
test_step_mdp_speed[True-False-True-True-True] 83.1120μs 50.3418μs 19.8642 KOps/s 19.8524 KOps/s $\color{#35bf28}+0.06\%$
test_step_mdp_speed[True-False-True-True-False] 65.0010μs 29.6392μs 33.7391 KOps/s 33.1438 KOps/s $\color{#35bf28}+1.80\%$
test_step_mdp_speed[True-False-True-False-True] 98.4220μs 27.2679μs 36.6732 KOps/s 36.5754 KOps/s $\color{#35bf28}+0.27\%$
test_step_mdp_speed[True-False-True-False-False] 44.2210μs 16.4565μs 60.7664 KOps/s 60.4768 KOps/s $\color{#35bf28}+0.48\%$
test_step_mdp_speed[True-False-False-True-True] 94.2610μs 52.2983μs 19.1211 KOps/s 19.0317 KOps/s $\color{#35bf28}+0.47\%$
test_step_mdp_speed[True-False-False-True-False] 60.8010μs 32.9003μs 30.3948 KOps/s 30.6772 KOps/s $\color{#d91a1a}-0.92\%$
test_step_mdp_speed[True-False-False-False-True] 58.5510μs 29.7976μs 33.5597 KOps/s 33.1552 KOps/s $\color{#35bf28}+1.22\%$
test_step_mdp_speed[True-False-False-False-False] 60.8510μs 19.2322μs 51.9961 KOps/s 52.2455 KOps/s $\color{#d91a1a}-0.48\%$
test_step_mdp_speed[False-True-True-True-True] 81.7310μs 49.5110μs 20.1975 KOps/s 19.9973 KOps/s $\color{#35bf28}+1.00\%$
test_step_mdp_speed[False-True-True-True-False] 61.2610μs 30.1957μs 33.1173 KOps/s 32.7613 KOps/s $\color{#35bf28}+1.09\%$
test_step_mdp_speed[False-True-True-False-True] 2.3762ms 31.7884μs 31.4580 KOps/s 31.1539 KOps/s $\color{#35bf28}+0.98\%$
test_step_mdp_speed[False-True-True-False-False] 44.1510μs 18.3850μs 54.3923 KOps/s 54.2617 KOps/s $\color{#35bf28}+0.24\%$
test_step_mdp_speed[False-True-False-True-True] 0.1342ms 52.3548μs 19.1004 KOps/s 18.8249 KOps/s $\color{#35bf28}+1.46\%$
test_step_mdp_speed[False-True-False-True-False] 68.5710μs 33.1901μs 30.1295 KOps/s 30.4301 KOps/s $\color{#d91a1a}-0.99\%$
test_step_mdp_speed[False-True-False-False-True] 64.5510μs 33.8078μs 29.5789 KOps/s 29.4660 KOps/s $\color{#35bf28}+0.38\%$
test_step_mdp_speed[False-True-False-False-False] 49.7910μs 20.7714μs 48.1430 KOps/s 47.8150 KOps/s $\color{#35bf28}+0.69\%$
test_step_mdp_speed[False-False-True-True-True] 88.4410μs 55.2418μs 18.1022 KOps/s 18.2519 KOps/s $\color{#d91a1a}-0.82\%$
test_step_mdp_speed[False-False-True-True-False] 85.3010μs 35.3643μs 28.2771 KOps/s 28.1052 KOps/s $\color{#35bf28}+0.61\%$
test_step_mdp_speed[False-False-True-False-True] 56.4210μs 33.5740μs 29.7849 KOps/s 29.9181 KOps/s $\color{#d91a1a}-0.45\%$
test_step_mdp_speed[False-False-True-False-False] 46.3700μs 20.7876μs 48.1055 KOps/s 48.7600 KOps/s $\color{#d91a1a}-1.34\%$
test_step_mdp_speed[False-False-False-True-True] 0.1032ms 57.5387μs 17.3796 KOps/s 17.6140 KOps/s $\color{#d91a1a}-1.33\%$
test_step_mdp_speed[False-False-False-True-False] 71.7510μs 38.5111μs 25.9665 KOps/s 26.1093 KOps/s $\color{#d91a1a}-0.55\%$
test_step_mdp_speed[False-False-False-False-True] 68.0610μs 35.6047μs 28.0862 KOps/s 27.9886 KOps/s $\color{#35bf28}+0.35\%$
test_step_mdp_speed[False-False-False-False-False] 53.7910μs 23.7027μs 42.1893 KOps/s 43.4770 KOps/s $\color{#d91a1a}-2.96\%$
test_non_tensor_env_rollout_speed[1000-single-True] 0.8791s 0.7767s 1.2874 Ops/s 1.2777 Ops/s $\color{#35bf28}+0.76\%$
test_non_tensor_env_rollout_speed[1000-single-False] 0.7423s 0.6401s 1.5623 Ops/s 1.5540 Ops/s $\color{#35bf28}+0.53\%$
test_non_tensor_env_rollout_speed[1000-serial-no-buffers-True] 1.7792s 1.7045s 0.5867 Ops/s 0.5847 Ops/s $\color{#35bf28}+0.34\%$
test_non_tensor_env_rollout_speed[1000-serial-no-buffers-False] 1.5564s 1.4755s 0.6777 Ops/s 0.6738 Ops/s $\color{#35bf28}+0.58\%$
test_non_tensor_env_rollout_speed[1000-serial-buffers-True] 2.0377s 1.9521s 0.5123 Ops/s 0.5088 Ops/s $\color{#35bf28}+0.67\%$
test_non_tensor_env_rollout_speed[1000-serial-buffers-False] 1.8008s 1.7225s 0.5806 Ops/s 0.5757 Ops/s $\color{#35bf28}+0.84\%$
test_non_tensor_env_rollout_speed[1000-parallel-no-buffers-True] 4.8276s 4.7046s 0.2126 Ops/s 0.2105 Ops/s $\color{#35bf28}+0.98\%$
test_non_tensor_env_rollout_speed[1000-parallel-no-buffers-False] 4.6534s 4.5079s 0.2218 Ops/s 0.2209 Ops/s $\color{#35bf28}+0.44\%$
test_non_tensor_env_rollout_speed[1000-parallel-buffers-True] 2.0693s 1.9870s 0.5033 Ops/s 0.4962 Ops/s $\color{#35bf28}+1.43\%$
test_non_tensor_env_rollout_speed[1000-parallel-buffers-False] 1.7662s 1.6859s 0.5931 Ops/s 0.5729 Ops/s $\color{#35bf28}+3.54\%$
test_values[generalized_advantage_estimate-True-True] 10.4621ms 10.3350ms 96.7589 Ops/s 94.7436 Ops/s $\color{#35bf28}+2.13\%$
test_values[vec_generalized_advantage_estimate-True-True] 19.7622ms 17.5974ms 56.8265 Ops/s 56.9106 Ops/s $\color{#d91a1a}-0.15\%$
test_values[td0_return_estimate-False-False] 0.2318ms 0.1317ms 7.5948 KOps/s 7.4736 KOps/s $\color{#35bf28}+1.62\%$
test_values[td1_return_estimate-False-False] 28.6357ms 28.3555ms 35.2665 Ops/s 34.0846 Ops/s $\color{#35bf28}+3.47\%$
test_values[vec_td1_return_estimate-False-False] 18.4149ms 17.5800ms 56.8829 Ops/s 56.3951 Ops/s $\color{#35bf28}+0.86\%$
test_values[td_lambda_return_estimate-True-False] 42.7852ms 42.2186ms 23.6862 Ops/s 23.1372 Ops/s $\color{#35bf28}+2.37\%$
test_values[vec_td_lambda_return_estimate-True-False] 17.8329ms 17.5600ms 56.9475 Ops/s 57.0402 Ops/s $\color{#d91a1a}-0.16\%$
test_gae_speed[generalized_advantage_estimate-False-1-512] 9.2423ms 9.1336ms 109.4857 Ops/s 106.4825 Ops/s $\color{#35bf28}+2.82\%$
test_gae_speed[vec_generalized_advantage_estimate-True-1-512] 1.7965ms 1.5744ms 635.1603 Ops/s 648.7945 Ops/s $\color{#d91a1a}-2.10\%$
test_gae_speed[vec_generalized_advantage_estimate-False-1-512] 0.5239ms 0.4290ms 2.3311 KOps/s 2.3083 KOps/s $\color{#35bf28}+0.99\%$
test_gae_speed[vec_generalized_advantage_estimate-True-32-512] 34.7963ms 34.0677ms 29.3533 Ops/s 28.7540 Ops/s $\color{#35bf28}+2.08\%$
test_gae_speed[vec_generalized_advantage_estimate-False-32-512] 1.9234ms 1.7171ms 582.3650 Ops/s 580.9294 Ops/s $\color{#35bf28}+0.25\%$
test_dqn_speed[False-None] 1.5411ms 1.4079ms 710.2622 Ops/s 697.1303 Ops/s $\color{#35bf28}+1.88\%$
test_dqn_speed[False-backward] 2.3417ms 1.9416ms 515.0357 Ops/s 503.5286 Ops/s $\color{#35bf28}+2.29\%$
test_dqn_speed[True-None] 0.6233ms 0.5365ms 1.8638 KOps/s 1.7267 KOps/s $\textbf{\color{#35bf28}+7.94\%}$
test_dqn_speed[True-backward] 1.0202ms 0.9861ms 1.0141 KOps/s 892.8450 Ops/s $\textbf{\color{#35bf28}+13.58\%}$
test_dqn_speed[reduce-overhead-None] 0.6066ms 0.5241ms 1.9079 KOps/s 1.8358 KOps/s $\color{#35bf28}+3.93\%$
test_dqn_speed[reduce-overhead-backward] 1.1185ms 0.9915ms 1.0086 KOps/s 888.4230 Ops/s $\textbf{\color{#35bf28}+13.52\%}$
test_ddpg_speed[False-None] 3.5323ms 2.8936ms 345.5850 Ops/s 328.7598 Ops/s $\textbf{\color{#35bf28}+5.12\%}$
test_ddpg_speed[False-backward] 4.1792ms 4.0865ms 244.7060 Ops/s 239.2557 Ops/s $\color{#35bf28}+2.28\%$
test_ddpg_speed[True-None] 1.5136ms 1.3945ms 717.0929 Ops/s 689.7135 Ops/s $\color{#35bf28}+3.97\%$
test_ddpg_speed[True-backward] 2.4690ms 2.4019ms 416.3358 Ops/s 369.3070 Ops/s $\textbf{\color{#35bf28}+12.73\%}$
test_ddpg_speed[reduce-overhead-None] 1.4710ms 1.3762ms 726.6220 Ops/s 693.9917 Ops/s $\color{#35bf28}+4.70\%$
test_ddpg_speed[reduce-overhead-backward] 2.4844ms 2.3840ms 419.4631 Ops/s 384.8089 Ops/s $\textbf{\color{#35bf28}+9.01\%}$
test_sac_speed[False-None] 9.3112ms 8.0882ms 123.6370 Ops/s 122.0029 Ops/s $\color{#35bf28}+1.34\%$
test_sac_speed[False-backward] 11.7989ms 11.4545ms 87.3017 Ops/s 87.1973 Ops/s $\color{#35bf28}+0.12\%$
test_sac_speed[True-None] 2.2655ms 2.1329ms 468.8483 Ops/s 443.5114 Ops/s $\textbf{\color{#35bf28}+5.71\%}$
test_sac_speed[True-backward] 4.1423ms 4.0024ms 249.8479 Ops/s 207.4073 Ops/s $\textbf{\color{#35bf28}+20.46\%}$
test_sac_speed[reduce-overhead-None] 2.3270ms 2.1560ms 463.8291 Ops/s 449.8955 Ops/s $\color{#35bf28}+3.10\%$
test_sac_speed[reduce-overhead-backward] 4.2576ms 4.0745ms 245.4310 Ops/s 211.2428 Ops/s $\textbf{\color{#35bf28}+16.18\%}$
test_redq_speed[False-None] 15.0853ms 10.5500ms 94.7871 Ops/s 96.8210 Ops/s $\color{#d91a1a}-2.10\%$
test_redq_speed[False-backward] 18.5479ms 17.9032ms 55.8560 Ops/s 56.6710 Ops/s $\color{#d91a1a}-1.44\%$
test_redq_speed[True-None] 4.9811ms 4.5687ms 218.8821 Ops/s 226.1197 Ops/s $\color{#d91a1a}-3.20\%$
test_redq_speed[True-backward] 10.2723ms 9.8737ms 101.2787 Ops/s 100.6977 Ops/s $\color{#35bf28}+0.58\%$
test_redq_speed[reduce-overhead-None] 4.9040ms 4.4958ms 222.4305 Ops/s 235.2831 Ops/s $\textbf{\color{#d91a1a}-5.46\%}$
test_redq_speed[reduce-overhead-backward] 10.1754ms 9.9093ms 100.9153 Ops/s 92.0591 Ops/s $\textbf{\color{#35bf28}+9.62\%}$
test_redq_deprec_speed[False-None] 13.7060ms 11.1263ms 89.8773 Ops/s 89.2706 Ops/s $\color{#35bf28}+0.68\%$
test_redq_deprec_speed[False-backward] 16.5522ms 15.9515ms 62.6901 Ops/s 62.5529 Ops/s $\color{#35bf28}+0.22\%$
test_redq_deprec_speed[True-None] 4.1236ms 3.6283ms 275.6108 Ops/s 260.6462 Ops/s $\textbf{\color{#35bf28}+5.74\%}$
test_redq_deprec_speed[True-backward] 8.0047ms 7.7361ms 129.2645 Ops/s 126.4538 Ops/s $\color{#35bf28}+2.22\%$
test_redq_deprec_speed[reduce-overhead-None] 4.0151ms 3.6173ms 276.4511 Ops/s 270.9565 Ops/s $\color{#35bf28}+2.03\%$
test_redq_deprec_speed[reduce-overhead-backward] 8.0535ms 7.6770ms 130.2585 Ops/s 119.3192 Ops/s $\textbf{\color{#35bf28}+9.17\%}$
test_td3_speed[False-None] 8.2974ms 8.0739ms 123.8553 Ops/s 123.7740 Ops/s $\color{#35bf28}+0.07\%$
test_td3_speed[False-backward] 11.4265ms 10.9541ms 91.2899 Ops/s 91.1528 Ops/s $\color{#35bf28}+0.15\%$
test_td3_speed[True-None] 1.8674ms 1.8398ms 543.5487 Ops/s 538.9454 Ops/s $\color{#35bf28}+0.85\%$
test_td3_speed[True-backward] 3.7785ms 3.6490ms 274.0498 Ops/s 268.0400 Ops/s $\color{#35bf28}+2.24\%$
test_td3_speed[reduce-overhead-None] 1.8358ms 1.7981ms 556.1286 Ops/s 549.7248 Ops/s $\color{#35bf28}+1.16\%$
test_td3_speed[reduce-overhead-backward] 3.8219ms 3.6843ms 271.4184 Ops/s 233.4395 Ops/s $\textbf{\color{#35bf28}+16.27\%}$
test_cql_speed[False-None] 30.6527ms 27.0113ms 37.0215 Ops/s 38.6535 Ops/s $\color{#d91a1a}-4.22\%$
test_cql_speed[False-backward] 35.9926ms 35.3436ms 28.2937 Ops/s 28.2702 Ops/s $\color{#35bf28}+0.08\%$
test_cql_speed[True-None] 12.7338ms 12.4393ms 80.3901 Ops/s 78.7881 Ops/s $\color{#35bf28}+2.03\%$
test_cql_speed[True-backward] 18.7765ms 18.4258ms 54.2717 Ops/s 56.8377 Ops/s $\color{#d91a1a}-4.51\%$
test_cql_speed[reduce-overhead-None] 12.8800ms 12.6097ms 79.3040 Ops/s 79.3315 Ops/s $\color{#d91a1a}-0.03\%$
test_cql_speed[reduce-overhead-backward] 19.0694ms 18.6699ms 53.5620 Ops/s 56.2657 Ops/s $\color{#d91a1a}-4.81\%$
test_a2c_speed[False-None] 5.9485ms 5.4526ms 183.4000 Ops/s 177.6685 Ops/s $\color{#35bf28}+3.23\%$
test_a2c_speed[False-backward] 12.2136ms 11.9117ms 83.9513 Ops/s 81.8584 Ops/s $\color{#35bf28}+2.56\%$
test_a2c_speed[True-None] 4.0575ms 3.7275ms 268.2777 Ops/s 259.9910 Ops/s $\color{#35bf28}+3.19\%$
test_a2c_speed[True-backward] 8.8470ms 8.6337ms 115.8246 Ops/s 110.0228 Ops/s $\textbf{\color{#35bf28}+5.27\%}$
test_a2c_speed[reduce-overhead-None] 4.0907ms 3.7119ms 269.4046 Ops/s 262.0971 Ops/s $\color{#35bf28}+2.79\%$
test_a2c_speed[reduce-overhead-backward] 8.9469ms 8.7237ms 114.6304 Ops/s 111.7242 Ops/s $\color{#35bf28}+2.60\%$
test_ppo_speed[False-None] 5.9530ms 5.7702ms 173.3054 Ops/s 167.4143 Ops/s $\color{#35bf28}+3.52\%$
test_ppo_speed[False-backward] 12.9797ms 12.5536ms 79.6587 Ops/s 78.7703 Ops/s $\color{#35bf28}+1.13\%$
test_ppo_speed[True-None] 3.7918ms 3.6232ms 275.9990 Ops/s 270.0079 Ops/s $\color{#35bf28}+2.22\%$
test_ppo_speed[True-backward] 8.7207ms 8.4679ms 118.0935 Ops/s 116.3476 Ops/s $\color{#35bf28}+1.50\%$
test_ppo_speed[reduce-overhead-None] 3.7347ms 3.6169ms 276.4829 Ops/s 270.3395 Ops/s $\color{#35bf28}+2.27\%$
test_ppo_speed[reduce-overhead-backward] 8.9321ms 8.6816ms 115.1866 Ops/s 111.9804 Ops/s $\color{#35bf28}+2.86\%$
test_reinforce_speed[False-None] 6.1613ms 4.5932ms 217.7149 Ops/s 213.7109 Ops/s $\color{#35bf28}+1.87\%$
test_reinforce_speed[False-backward] 7.6335ms 7.4318ms 134.5573 Ops/s 132.2330 Ops/s $\color{#35bf28}+1.76\%$
test_reinforce_speed[True-None] 3.1006ms 2.9034ms 344.4271 Ops/s 340.5902 Ops/s $\color{#35bf28}+1.13\%$
test_reinforce_speed[True-backward] 8.2179ms 7.8826ms 126.8609 Ops/s 122.8095 Ops/s $\color{#35bf28}+3.30\%$
test_reinforce_speed[reduce-overhead-None] 3.0433ms 2.8851ms 346.6057 Ops/s 344.4406 Ops/s $\color{#35bf28}+0.63\%$
test_reinforce_speed[reduce-overhead-backward] 8.1702ms 7.9151ms 126.3411 Ops/s 118.7034 Ops/s $\textbf{\color{#35bf28}+6.43\%}$
test_iql_speed[False-None] 21.7195ms 19.6460ms 50.9010 Ops/s 50.6132 Ops/s $\color{#35bf28}+0.57\%$
test_iql_speed[False-backward] 34.8495ms 30.7419ms 32.5289 Ops/s 32.8274 Ops/s $\color{#d91a1a}-0.91\%$
test_iql_speed[True-None] 8.8532ms 8.5676ms 116.7189 Ops/s 114.4911 Ops/s $\color{#35bf28}+1.95\%$
test_iql_speed[True-backward] 17.4947ms 16.8143ms 59.4730 Ops/s 59.7387 Ops/s $\color{#d91a1a}-0.44\%$
test_iql_speed[reduce-overhead-None] 9.0162ms 8.6534ms 115.5620 Ops/s 111.2759 Ops/s $\color{#35bf28}+3.85\%$
test_iql_speed[reduce-overhead-backward] 17.5963ms 17.2225ms 58.0637 Ops/s 58.0964 Ops/s $\color{#d91a1a}-0.06\%$
test_rb_sample[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 6.2597ms 6.0267ms 165.9272 Ops/s 164.7303 Ops/s $\color{#35bf28}+0.73\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 0.9801ms 0.3082ms 3.2449 KOps/s 3.0797 KOps/s $\textbf{\color{#35bf28}+5.36\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.6199ms 0.3217ms 3.1089 KOps/s 2.8827 KOps/s $\textbf{\color{#35bf28}+7.85\%}$
test_rb_sample[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 5.9266ms 5.7285ms 174.5668 Ops/s 174.4649 Ops/s $\color{#35bf28}+0.06\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 0.7843s 0.8198ms 1.2199 KOps/s 3.3661 KOps/s $\textbf{\color{#d91a1a}-63.76\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.4873ms 0.2621ms 3.8160 KOps/s 3.3518 KOps/s $\textbf{\color{#35bf28}+13.85\%}$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-sampler6-10000] 1.4725ms 1.2653ms 790.3436 Ops/s 718.5792 Ops/s $\textbf{\color{#35bf28}+9.99\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-sampler7-10000] 1.4580ms 1.1845ms 844.2570 Ops/s 766.8069 Ops/s $\textbf{\color{#35bf28}+10.10\%}$
test_rb_sample[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 6.1444ms 6.0110ms 166.3619 Ops/s 167.8079 Ops/s $\color{#d91a1a}-0.86\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 1.1073ms 0.4381ms 2.2828 KOps/s 2.0468 KOps/s $\textbf{\color{#35bf28}+11.53\%}$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.6191ms 0.4147ms 2.4116 KOps/s 2.2314 KOps/s $\textbf{\color{#35bf28}+8.08\%}$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 5.9453ms 5.8638ms 170.5383 Ops/s 170.4715 Ops/s $\color{#35bf28}+0.04\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 2.0715ms 0.2855ms 3.5025 KOps/s 3.0650 KOps/s $\textbf{\color{#35bf28}+14.28\%}$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.3973ms 0.2632ms 3.7987 KOps/s 3.1828 KOps/s $\textbf{\color{#35bf28}+19.35\%}$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 6.0155ms 5.8191ms 171.8489 Ops/s 173.2845 Ops/s $\color{#d91a1a}-0.83\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 2.0366ms 0.3439ms 2.9076 KOps/s 3.1722 KOps/s $\textbf{\color{#d91a1a}-8.34\%}$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.5524ms 0.3229ms 3.0967 KOps/s 3.7772 KOps/s $\textbf{\color{#d91a1a}-18.02\%}$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 6.0933ms 5.9949ms 166.8091 Ops/s 167.2892 Ops/s $\color{#d91a1a}-0.29\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 2.0720ms 0.4829ms 2.0710 KOps/s 2.2425 KOps/s $\textbf{\color{#d91a1a}-7.64\%}$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.7447ms 0.4433ms 2.2558 KOps/s 2.0566 KOps/s $\textbf{\color{#35bf28}+9.69\%}$
test_rb_populate[TensorDictReplayBuffer-ListStorage-RandomSampler-400] 0.6738s 18.4766ms 54.1225 Ops/s 196.6341 Ops/s $\textbf{\color{#d91a1a}-72.48\%}$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-400] 10.7273ms 2.0538ms 486.9095 Ops/s 403.3768 Ops/s $\textbf{\color{#35bf28}+20.71\%}$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-400] 8.2017ms 1.2723ms 785.9871 Ops/s 794.7871 Ops/s $\color{#d91a1a}-1.11\%$
test_rb_populate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-400] 7.5031ms 5.2179ms 191.6485 Ops/s 49.6039 Ops/s $\textbf{\color{#35bf28}+286.36\%}$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-400] 9.3166ms 2.1304ms 469.3951 Ops/s 501.8654 Ops/s $\textbf{\color{#d91a1a}-6.47\%}$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-400] 7.0642ms 1.2521ms 798.6675 Ops/s 811.8298 Ops/s $\color{#d91a1a}-1.62\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-ListStorage-None-400] 6.8491ms 5.2730ms 189.6461 Ops/s 186.5536 Ops/s $\color{#35bf28}+1.66\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-400] 12.2289ms 2.1712ms 460.5800 Ops/s 442.4902 Ops/s $\color{#35bf28}+4.09\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-400] 6.0496ms 1.3401ms 746.2161 Ops/s 704.1429 Ops/s $\textbf{\color{#35bf28}+5.98\%}$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-True] 36.0716ms 34.2087ms 29.2323 Ops/s 29.2183 Ops/s $\color{#35bf28}+0.05\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-False] 19.5885ms 17.7880ms 56.2178 Ops/s 55.6777 Ops/s $\color{#35bf28}+0.97\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-True] 37.8402ms 35.1826ms 28.4232 Ops/s 28.6424 Ops/s $\color{#d91a1a}-0.77\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-False] 19.4401ms 17.9635ms 55.6686 Ops/s 55.7752 Ops/s $\color{#d91a1a}-0.19\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-True] 41.8491ms 37.3854ms 26.7484 Ops/s 27.0242 Ops/s $\color{#d91a1a}-1.02\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-False] 20.5876ms 19.4324ms 51.4606 Ops/s 51.7408 Ops/s $\color{#d91a1a}-0.54\%$

[ghstack-poisoned]
vmoens added a commit that referenced this pull request Dec 18, 2025
ghstack-source-id: 05dfa8b
Pull-Request: #3265
vmoens added a commit that referenced this pull request Dec 18, 2025
ghstack-source-id: 05dfa8b
Pull-Request: #3265
@vmoens vmoens merged commit 4cadc85 into gh/vmoens/174/base Dec 18, 2025
39 of 48 checks passed
@vmoens vmoens deleted the gh/vmoens/174/head branch December 18, 2025 16:13
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. Tests Incomplete or broken unit tests

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants