Skip to content

Conversation

@vmoens
Copy link
Collaborator

@vmoens vmoens commented Jan 21, 2026

Summary

This is a test PR to verify the fixes for PR #3343 before pushing to the contributor's branch.

Fixes:

  • Use PENDULUM_VERSIONED() instead of hardcoded "Pendulum-v1" to support older gym versions that only have "Pendulum-v0"
  • Fixes test_gym_kwargs_preserved_with_seed failing on older gym versions with: gym.error.DeprecatedEnv: Env Pendulum-v1 not found (valid versions include ['Pendulum-v0'])

Test plan

  • CI passes for unittests-gym
  • CI passes for tests-olddeps
  • CI passes for tests-gpu

Once CI passes, the changes will be pushed to ParamThakkar123:add/num-envs-gym.

ParamThakkar123 and others added 19 commits January 19, 2026 16:17
- Remove incorrect CUDA skip decorator (test doesn't use CUDA)
- Use Pendulum-v1 instead of CartPole-v1 because CartPole can terminate
  early due to pole falling, especially with frame_skip=4, causing the
  rollout assertion to fail (expected 5 steps, got 3)
Use PENDULUM_VERSIONED() instead of hardcoded "Pendulum-v1" to support
older gym versions that only have "Pendulum-v0".
@pytorch-bot
Copy link

pytorch-bot bot commented Jan 21, 2026

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/rl/3367

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

✅ You can merge normally! (3 Unrelated Failures)

As of commit 9cbd720 with merge base 9dcca77 (image):

BROKEN TRUNK - The following jobs 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 Jan 21, 2026
- Fix import ordering
- Add missing space after ==
- Add newline at end of file
- Remove trailing whitespace
@github-actions
Copy link

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

Total Benchmarks: 153. Improved: $\large\color{#35bf28}17$. 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] 79.4632μs 78.4041μs 12.7544 KOps/s 12.6634 KOps/s $\color{#35bf28}+0.72\%$
test_tensor_to_bytestream_speed[torch.save] 0.1435ms 0.1427ms 7.0062 KOps/s 6.8618 KOps/s $\color{#35bf28}+2.10\%$
test_tensor_to_bytestream_speed[untyped_storage] 0.1093s 0.1091s 9.1693 Ops/s 9.2655 Ops/s $\color{#d91a1a}-1.04\%$
test_tensor_to_bytestream_speed[numpy] 2.4526μs 2.4497μs 408.2117 KOps/s 414.0257 KOps/s $\color{#d91a1a}-1.40\%$
test_tensor_to_bytestream_speed[safetensors] 38.2706μs 37.5182μs 26.6537 KOps/s 25.6145 KOps/s $\color{#35bf28}+4.06\%$
test_simple 0.6538s 0.5639s 1.7734 Ops/s 1.7673 Ops/s $\color{#35bf28}+0.35\%$
test_transformed 1.2200s 1.1271s 0.8872 Ops/s 0.8845 Ops/s $\color{#35bf28}+0.31\%$
test_serial 1.7700s 1.6757s 0.5968 Ops/s 0.5955 Ops/s $\color{#35bf28}+0.20\%$
test_parallel 1.1875s 1.1025s 0.9070 Ops/s 0.8519 Ops/s $\textbf{\color{#35bf28}+6.47\%}$
test_step_mdp_speed[True-True-True-True-True] 0.1975ms 43.3772μs 23.0536 KOps/s 22.4519 KOps/s $\color{#35bf28}+2.68\%$
test_step_mdp_speed[True-True-True-True-False] 0.4445ms 24.5811μs 40.6817 KOps/s 40.2339 KOps/s $\color{#35bf28}+1.11\%$
test_step_mdp_speed[True-True-True-False-True] 0.4503ms 24.4651μs 40.8745 KOps/s 40.3990 KOps/s $\color{#35bf28}+1.18\%$
test_step_mdp_speed[True-True-True-False-False] 39.4410μs 13.3919μs 74.6721 KOps/s 72.7131 KOps/s $\color{#35bf28}+2.69\%$
test_step_mdp_speed[True-True-False-True-True] 0.4687ms 47.6555μs 20.9839 KOps/s 21.3961 KOps/s $\color{#d91a1a}-1.93\%$
test_step_mdp_speed[True-True-False-True-False] 0.4490ms 26.9904μs 37.0502 KOps/s 36.5612 KOps/s $\color{#35bf28}+1.34\%$
test_step_mdp_speed[True-True-False-False-True] 62.9410μs 27.2907μs 36.6425 KOps/s 36.2879 KOps/s $\color{#35bf28}+0.98\%$
test_step_mdp_speed[True-True-False-False-False] 0.4492ms 16.2480μs 61.5460 KOps/s 60.2421 KOps/s $\color{#35bf28}+2.16\%$
test_step_mdp_speed[True-False-True-True-True] 0.4706ms 49.4069μs 20.2401 KOps/s 19.7748 KOps/s $\color{#35bf28}+2.35\%$
test_step_mdp_speed[True-False-True-True-False] 0.4474ms 29.8032μs 33.5535 KOps/s 33.4860 KOps/s $\color{#35bf28}+0.20\%$
test_step_mdp_speed[True-False-True-False-True] 62.3310μs 27.5061μs 36.3555 KOps/s 36.3653 KOps/s $\color{#d91a1a}-0.03\%$
test_step_mdp_speed[True-False-True-False-False] 0.4486ms 16.0788μs 62.1938 KOps/s 60.8920 KOps/s $\color{#35bf28}+2.14\%$
test_step_mdp_speed[True-False-False-True-True] 0.4803ms 52.3049μs 19.1187 KOps/s 18.7949 KOps/s $\color{#35bf28}+1.72\%$
test_step_mdp_speed[True-False-False-True-False] 0.4590ms 32.6239μs 30.6524 KOps/s 30.8446 KOps/s $\color{#d91a1a}-0.62\%$
test_step_mdp_speed[True-False-False-False-True] 61.4410μs 29.8774μs 33.4702 KOps/s 32.9947 KOps/s $\color{#35bf28}+1.44\%$
test_step_mdp_speed[True-False-False-False-False] 0.4445ms 18.5582μs 53.8847 KOps/s 52.4787 KOps/s $\color{#35bf28}+2.68\%$
test_step_mdp_speed[False-True-True-True-True] 0.4738ms 47.7960μs 20.9223 KOps/s 19.6236 KOps/s $\textbf{\color{#35bf28}+6.62\%}$
test_step_mdp_speed[False-True-True-True-False] 51.2610μs 29.7059μs 33.6633 KOps/s 33.0414 KOps/s $\color{#35bf28}+1.88\%$
test_step_mdp_speed[False-True-True-False-True] 0.4563ms 30.7581μs 32.5118 KOps/s 31.6159 KOps/s $\color{#35bf28}+2.83\%$
test_step_mdp_speed[False-True-True-False-False] 0.4401ms 17.6019μs 56.8121 KOps/s 55.4371 KOps/s $\color{#35bf28}+2.48\%$
test_step_mdp_speed[False-True-False-True-True] 2.7568ms 51.8817μs 19.2746 KOps/s 18.9863 KOps/s $\color{#35bf28}+1.52\%$
test_step_mdp_speed[False-True-False-True-False] 55.1510μs 32.3998μs 30.8644 KOps/s 30.3830 KOps/s $\color{#35bf28}+1.58\%$
test_step_mdp_speed[False-True-False-False-True] 0.4569ms 33.7859μs 29.5982 KOps/s 29.1913 KOps/s $\color{#35bf28}+1.39\%$
test_step_mdp_speed[False-True-False-False-False] 0.4416ms 20.6617μs 48.3987 KOps/s 48.2519 KOps/s $\color{#35bf28}+0.30\%$
test_step_mdp_speed[False-False-True-True-True] 84.6420μs 54.0505μs 18.5012 KOps/s 17.8720 KOps/s $\color{#35bf28}+3.52\%$
test_step_mdp_speed[False-False-True-True-False] 0.4575ms 34.8333μs 28.7082 KOps/s 27.9716 KOps/s $\color{#35bf28}+2.63\%$
test_step_mdp_speed[False-False-True-False-True] 0.4601ms 33.3249μs 30.0076 KOps/s 29.1225 KOps/s $\color{#35bf28}+3.04\%$
test_step_mdp_speed[False-False-True-False-False] 0.4407ms 20.3078μs 49.2421 KOps/s 48.3633 KOps/s $\color{#35bf28}+1.82\%$
test_step_mdp_speed[False-False-False-True-True] 0.1035ms 57.1499μs 17.4979 KOps/s 17.6467 KOps/s $\color{#d91a1a}-0.84\%$
test_step_mdp_speed[False-False-False-True-False] 0.4831ms 37.4179μs 26.7252 KOps/s 26.6289 KOps/s $\color{#35bf28}+0.36\%$
test_step_mdp_speed[False-False-False-False-True] 0.5022ms 35.7896μs 27.9410 KOps/s 27.8365 KOps/s $\color{#35bf28}+0.38\%$
test_step_mdp_speed[False-False-False-False-False] 0.4453ms 23.2136μs 43.0783 KOps/s 43.5717 KOps/s $\color{#d91a1a}-1.13\%$
test_non_tensor_env_rollout_speed[1000-single-True] 0.8439s 0.7687s 1.3010 Ops/s 1.3264 Ops/s $\color{#d91a1a}-1.92\%$
test_non_tensor_env_rollout_speed[1000-single-False] 0.7175s 0.6175s 1.6193 Ops/s 1.6085 Ops/s $\color{#35bf28}+0.67\%$
test_non_tensor_env_rollout_speed[1000-serial-no-buffers-True] 1.7040s 1.6272s 0.6145 Ops/s 0.6062 Ops/s $\color{#35bf28}+1.38\%$
test_non_tensor_env_rollout_speed[1000-serial-no-buffers-False] 1.4728s 1.4044s 0.7121 Ops/s 0.7013 Ops/s $\color{#35bf28}+1.54\%$
test_non_tensor_env_rollout_speed[1000-serial-buffers-True] 1.9470s 1.8614s 0.5372 Ops/s 0.5263 Ops/s $\color{#35bf28}+2.08\%$
test_non_tensor_env_rollout_speed[1000-serial-buffers-False] 1.7296s 1.6559s 0.6039 Ops/s 0.5952 Ops/s $\color{#35bf28}+1.46\%$
test_non_tensor_env_rollout_speed[1000-parallel-no-buffers-True] 4.6983s 4.6388s 0.2156 Ops/s 0.2180 Ops/s $\color{#d91a1a}-1.13\%$
test_non_tensor_env_rollout_speed[1000-parallel-no-buffers-False] 4.4841s 4.4059s 0.2270 Ops/s 0.2269 Ops/s $\color{#35bf28}+0.02\%$
test_non_tensor_env_rollout_speed[1000-parallel-buffers-True] 1.9599s 1.8879s 0.5297 Ops/s 0.5146 Ops/s $\color{#35bf28}+2.93\%$
test_non_tensor_env_rollout_speed[1000-parallel-buffers-False] 1.7130s 1.6338s 0.6121 Ops/s 0.6080 Ops/s $\color{#35bf28}+0.66\%$
test_values[generalized_advantage_estimate-True-True] 10.8990ms 10.4791ms 95.4283 Ops/s 96.0677 Ops/s $\color{#d91a1a}-0.67\%$
test_values[vec_generalized_advantage_estimate-True-True] 22.3070ms 18.4088ms 54.3218 Ops/s 87.3767 Ops/s $\textbf{\color{#d91a1a}-37.83\%}$
test_values[td0_return_estimate-False-False] 0.2305ms 0.1272ms 7.8628 KOps/s 7.8655 KOps/s $\color{#d91a1a}-0.03\%$
test_values[td1_return_estimate-False-False] 29.4255ms 28.2496ms 35.3987 Ops/s 35.3856 Ops/s $\color{#35bf28}+0.04\%$
test_values[vec_td1_return_estimate-False-False] 22.0208ms 18.4451ms 54.2150 Ops/s 87.5711 Ops/s $\textbf{\color{#d91a1a}-38.09\%}$
test_values[td_lambda_return_estimate-True-False] 44.1731ms 41.9624ms 23.8309 Ops/s 23.7749 Ops/s $\color{#35bf28}+0.24\%$
test_values[vec_td_lambda_return_estimate-True-False] 21.0169ms 18.1727ms 55.0275 Ops/s 87.6476 Ops/s $\textbf{\color{#d91a1a}-37.22\%}$
test_gae_speed[generalized_advantage_estimate-False-1-512] 9.3239ms 9.1949ms 108.7555 Ops/s 109.5796 Ops/s $\color{#d91a1a}-0.75\%$
test_gae_speed[vec_generalized_advantage_estimate-True-1-512] 1.9517ms 1.5211ms 657.4010 Ops/s 666.1125 Ops/s $\color{#d91a1a}-1.31\%$
test_gae_speed[vec_generalized_advantage_estimate-False-1-512] 0.5492ms 0.4371ms 2.2880 KOps/s 2.3074 KOps/s $\color{#d91a1a}-0.84\%$
test_gae_speed[vec_generalized_advantage_estimate-True-32-512] 35.0914ms 34.1359ms 29.2947 Ops/s 32.5912 Ops/s $\textbf{\color{#d91a1a}-10.11\%}$
test_gae_speed[vec_generalized_advantage_estimate-False-32-512] 2.1469ms 1.7592ms 568.4402 Ops/s 573.0745 Ops/s $\color{#d91a1a}-0.81\%$
test_dqn_speed[False-None] 1.5630ms 1.4420ms 693.4594 Ops/s 698.3164 Ops/s $\color{#d91a1a}-0.70\%$
test_dqn_speed[False-backward] 2.3984ms 1.9952ms 501.1910 Ops/s 509.2667 Ops/s $\color{#d91a1a}-1.59\%$
test_dqn_speed[True-None] 0.9469ms 0.5355ms 1.8673 KOps/s 1.8556 KOps/s $\color{#35bf28}+0.63\%$
test_dqn_speed[True-backward] 1.1491ms 0.9927ms 1.0074 KOps/s 846.3740 Ops/s $\textbf{\color{#35bf28}+19.02\%}$
test_dqn_speed[reduce-overhead-None] 0.8385ms 0.5148ms 1.9426 KOps/s 1.7864 KOps/s $\textbf{\color{#35bf28}+8.74\%}$
test_ddpg_speed[False-None] 3.5440ms 2.8064ms 356.3323 Ops/s 347.1768 Ops/s $\color{#35bf28}+2.64\%$
test_ddpg_speed[False-backward] 4.1518ms 4.0413ms 247.4435 Ops/s 242.2176 Ops/s $\color{#35bf28}+2.16\%$
test_ddpg_speed[True-None] 1.7903ms 1.3839ms 722.5764 Ops/s 710.6356 Ops/s $\color{#35bf28}+1.68\%$
test_ddpg_speed[True-backward] 2.4354ms 2.3662ms 422.6173 Ops/s 348.2200 Ops/s $\textbf{\color{#35bf28}+21.37\%}$
test_ddpg_speed[reduce-overhead-None] 1.7670ms 1.3670ms 731.5071 Ops/s 707.4413 Ops/s $\color{#35bf28}+3.40\%$
test_sac_speed[False-None] 9.2126ms 8.1005ms 123.4493 Ops/s 121.6066 Ops/s $\color{#35bf28}+1.52\%$
test_sac_speed[False-backward] 11.9514ms 11.2417ms 88.9549 Ops/s 86.1067 Ops/s $\color{#35bf28}+3.31\%$
test_sac_speed[True-None] 2.5384ms 2.1289ms 469.7315 Ops/s 454.7184 Ops/s $\color{#35bf28}+3.30\%$
test_sac_speed[True-backward] 4.1128ms 3.9900ms 250.6279 Ops/s 218.7796 Ops/s $\textbf{\color{#35bf28}+14.56\%}$
test_sac_speed[reduce-overhead-None] 2.3116ms 2.1043ms 475.2086 Ops/s 457.6707 Ops/s $\color{#35bf28}+3.83\%$
test_redq_speed[False-None] 15.7917ms 11.1731ms 89.5010 Ops/s 94.5570 Ops/s $\textbf{\color{#d91a1a}-5.35\%}$
test_redq_speed[False-backward] 18.7152ms 17.5217ms 57.0720 Ops/s 55.7412 Ops/s $\color{#35bf28}+2.39\%$
test_redq_speed[True-None] 4.7169ms 4.4244ms 226.0172 Ops/s 227.0818 Ops/s $\color{#d91a1a}-0.47\%$
test_redq_speed[True-backward] 9.9700ms 9.6557ms 103.5654 Ops/s 102.4038 Ops/s $\color{#35bf28}+1.13\%$
test_redq_speed[reduce-overhead-None] 4.8571ms 4.5021ms 222.1170 Ops/s 230.7123 Ops/s $\color{#d91a1a}-3.73\%$
test_redq_deprec_speed[False-None] 11.7954ms 11.0572ms 90.4387 Ops/s 88.7438 Ops/s $\color{#35bf28}+1.91\%$
test_redq_deprec_speed[False-backward] 16.3606ms 15.8487ms 63.0968 Ops/s 62.2144 Ops/s $\color{#35bf28}+1.42\%$
test_redq_deprec_speed[True-None] 3.9063ms 3.6542ms 273.6591 Ops/s 270.6035 Ops/s $\color{#35bf28}+1.13\%$
test_redq_deprec_speed[True-backward] 8.0613ms 7.5930ms 131.7008 Ops/s 127.9397 Ops/s $\color{#35bf28}+2.94\%$
test_redq_deprec_speed[reduce-overhead-None] 3.9271ms 3.5852ms 278.9213 Ops/s 282.2476 Ops/s $\color{#d91a1a}-1.18\%$
test_td3_speed[False-None] 8.4275ms 8.2486ms 121.2327 Ops/s 124.5757 Ops/s $\color{#d91a1a}-2.68\%$
test_td3_speed[False-backward] 11.6928ms 10.9888ms 91.0021 Ops/s 91.3415 Ops/s $\color{#d91a1a}-0.37\%$
test_td3_speed[True-None] 1.9240ms 1.8547ms 539.1682 Ops/s 546.0736 Ops/s $\color{#d91a1a}-1.26\%$
test_td3_speed[True-backward] 3.8837ms 3.6902ms 270.9867 Ops/s 228.7596 Ops/s $\textbf{\color{#35bf28}+18.46\%}$
test_td3_speed[reduce-overhead-None] 1.8494ms 1.7635ms 567.0657 Ops/s 553.4165 Ops/s $\color{#35bf28}+2.47\%$
test_cql_speed[False-None] 29.2217ms 25.9674ms 38.5098 Ops/s 39.0216 Ops/s $\color{#d91a1a}-1.31\%$
test_cql_speed[False-backward] 39.1466ms 35.5323ms 28.1434 Ops/s 28.1101 Ops/s $\color{#35bf28}+0.12\%$
test_cql_speed[True-None] 13.1698ms 12.6757ms 78.8912 Ops/s 82.4902 Ops/s $\color{#d91a1a}-4.36\%$
test_cql_speed[True-backward] 18.7899ms 18.0824ms 55.3025 Ops/s 56.6008 Ops/s $\color{#d91a1a}-2.29\%$
test_cql_speed[reduce-overhead-None] 15.2686ms 12.4010ms 80.6387 Ops/s 82.2732 Ops/s $\color{#d91a1a}-1.99\%$
test_a2c_speed[False-None] 5.7908ms 5.3085ms 188.3780 Ops/s 187.7176 Ops/s $\color{#35bf28}+0.35\%$
test_a2c_speed[False-backward] 11.9135ms 11.7070ms 85.4189 Ops/s 84.9728 Ops/s $\color{#35bf28}+0.53\%$
test_a2c_speed[True-None] 3.8401ms 3.7005ms 270.2355 Ops/s 257.6587 Ops/s $\color{#35bf28}+4.88\%$
test_a2c_speed[True-backward] 8.6723ms 8.4661ms 118.1179 Ops/s 110.7384 Ops/s $\textbf{\color{#35bf28}+6.66\%}$
test_a2c_speed[reduce-overhead-None] 3.9176ms 3.6828ms 271.5309 Ops/s 271.8917 Ops/s $\color{#d91a1a}-0.13\%$
test_ppo_speed[False-None] 5.9902ms 5.7546ms 173.7739 Ops/s 170.5920 Ops/s $\color{#35bf28}+1.87\%$
test_ppo_speed[False-backward] 13.0123ms 12.1831ms 82.0811 Ops/s 81.1346 Ops/s $\color{#35bf28}+1.17\%$
test_ppo_speed[True-None] 3.7855ms 3.5812ms 279.2342 Ops/s 272.6582 Ops/s $\color{#35bf28}+2.41\%$
test_ppo_speed[True-backward] 8.3938ms 8.2373ms 121.3987 Ops/s 115.9289 Ops/s $\color{#35bf28}+4.72\%$
test_ppo_speed[reduce-overhead-None] 3.7876ms 3.5602ms 280.8859 Ops/s 276.5765 Ops/s $\color{#35bf28}+1.56\%$
test_reinforce_speed[False-None] 4.6416ms 4.4370ms 225.3753 Ops/s 217.5343 Ops/s $\color{#35bf28}+3.60\%$
test_reinforce_speed[False-backward] 7.4502ms 7.1669ms 139.5301 Ops/s 134.8018 Ops/s $\color{#35bf28}+3.51\%$
test_reinforce_speed[True-None] 3.0344ms 2.8376ms 352.4082 Ops/s 341.8967 Ops/s $\color{#35bf28}+3.07\%$
test_reinforce_speed[True-backward] 7.8896ms 7.6497ms 130.7238 Ops/s 129.4497 Ops/s $\color{#35bf28}+0.98\%$
test_reinforce_speed[reduce-overhead-None] 3.0103ms 2.8357ms 352.6507 Ops/s 344.5325 Ops/s $\color{#35bf28}+2.36\%$
test_iql_speed[False-None] 24.9474ms 19.9328ms 50.1685 Ops/s 50.1904 Ops/s $\color{#d91a1a}-0.04\%$
test_iql_speed[False-backward] 37.4430ms 30.8897ms 32.3732 Ops/s 32.4489 Ops/s $\color{#d91a1a}-0.23\%$
test_iql_speed[True-None] 9.3898ms 8.4885ms 117.8061 Ops/s 113.8595 Ops/s $\color{#35bf28}+3.47\%$
test_iql_speed[True-backward] 16.6712ms 16.3785ms 61.0556 Ops/s 59.4819 Ops/s $\color{#35bf28}+2.65\%$
test_iql_speed[reduce-overhead-None] 8.9497ms 8.5270ms 117.2747 Ops/s 114.6471 Ops/s $\color{#35bf28}+2.29\%$
test_rb_sample[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 7.4964ms 5.9069ms 169.2938 Ops/s 171.0134 Ops/s $\color{#d91a1a}-1.01\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 0.4978ms 0.3275ms 3.0534 KOps/s 3.1174 KOps/s $\color{#d91a1a}-2.05\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.5352ms 0.3508ms 2.8506 KOps/s 2.7028 KOps/s $\textbf{\color{#35bf28}+5.47\%}$
test_rb_sample[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 5.7891ms 5.6048ms 178.4172 Ops/s 178.8544 Ops/s $\color{#d91a1a}-0.24\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 2.0328ms 0.2849ms 3.5095 KOps/s 2.7663 KOps/s $\textbf{\color{#35bf28}+26.86\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.5239ms 0.3226ms 3.0996 KOps/s 3.0824 KOps/s $\color{#35bf28}+0.56\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-sampler6-10000] 1.6692ms 1.3300ms 751.8769 Ops/s 695.4689 Ops/s $\textbf{\color{#35bf28}+8.11\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-sampler7-10000] 1.3350ms 1.1722ms 853.0727 Ops/s 741.3414 Ops/s $\textbf{\color{#35bf28}+15.07\%}$
test_rb_sample[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 9.6029ms 5.8738ms 170.2481 Ops/s 173.4249 Ops/s $\color{#d91a1a}-1.83\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 2.0788ms 0.4705ms 2.1254 KOps/s 2.2280 KOps/s $\color{#d91a1a}-4.60\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.6763ms 0.4251ms 2.3525 KOps/s 2.4511 KOps/s $\color{#d91a1a}-4.02\%$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 5.6747ms 5.5875ms 178.9707 Ops/s 176.0069 Ops/s $\color{#35bf28}+1.68\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 1.6834ms 0.3289ms 3.0405 KOps/s 2.8798 KOps/s $\textbf{\color{#35bf28}+5.58\%}$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.5529ms 0.3074ms 3.2533 KOps/s 3.0629 KOps/s $\textbf{\color{#35bf28}+6.22\%}$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 5.7460ms 5.5557ms 179.9958 Ops/s 179.7913 Ops/s $\color{#35bf28}+0.11\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 1.6031ms 0.3300ms 3.0307 KOps/s 3.0396 KOps/s $\color{#d91a1a}-0.29\%$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.5755ms 0.3011ms 3.3209 KOps/s 3.1046 KOps/s $\textbf{\color{#35bf28}+6.96\%}$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 6.2751ms 5.7636ms 173.5029 Ops/s 170.3588 Ops/s $\color{#35bf28}+1.85\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 1.1486ms 0.4834ms 2.0688 KOps/s 1.7779 KOps/s $\textbf{\color{#35bf28}+16.36\%}$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.6666ms 0.4994ms 2.0025 KOps/s 2.0135 KOps/s $\color{#d91a1a}-0.55\%$
test_rb_populate[TensorDictReplayBuffer-ListStorage-RandomSampler-400] 0.5859s 16.5686ms 60.3553 Ops/s 200.6177 Ops/s $\textbf{\color{#d91a1a}-69.92\%}$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-400] 11.4106ms 1.9746ms 506.4438 Ops/s 506.2006 Ops/s $\color{#35bf28}+0.05\%$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-400] 1.0899ms 0.8672ms 1.1532 KOps/s 1.1477 KOps/s $\color{#35bf28}+0.47\%$
test_rb_populate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-400] 6.2626ms 4.8611ms 205.7131 Ops/s 198.3486 Ops/s $\color{#35bf28}+3.71\%$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-400] 12.9049ms 2.0091ms 497.7342 Ops/s 547.5207 Ops/s $\textbf{\color{#d91a1a}-9.09\%}$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-400] 3.4499ms 1.1232ms 890.2751 Ops/s 1.1236 KOps/s $\textbf{\color{#d91a1a}-20.77\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-ListStorage-None-400] 0.5130s 15.3480ms 65.1552 Ops/s 58.1588 Ops/s $\textbf{\color{#35bf28}+12.03\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-400] 7.9125ms 2.0051ms 498.7283 Ops/s 517.4236 Ops/s $\color{#d91a1a}-3.61\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-400] 11.1647ms 1.4027ms 712.8922 Ops/s 970.4350 Ops/s $\textbf{\color{#d91a1a}-26.54\%}$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-True] 35.7821ms 33.8031ms 29.5831 Ops/s 29.2981 Ops/s $\color{#35bf28}+0.97\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-False] 19.3653ms 17.6971ms 56.5063 Ops/s 56.3679 Ops/s $\color{#35bf28}+0.25\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-True] 38.2154ms 34.8926ms 28.6593 Ops/s 28.2831 Ops/s $\color{#35bf28}+1.33\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-False] 19.3841ms 17.9391ms 55.7441 Ops/s 54.2230 Ops/s $\color{#35bf28}+2.81\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-True] 38.1862ms 36.3767ms 27.4901 Ops/s 27.0987 Ops/s $\color{#35bf28}+1.44\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-False] 20.6895ms 19.2099ms 52.0564 Ops/s 50.4981 Ops/s $\color{#35bf28}+3.09\%$

@github-actions
Copy link

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

Total Benchmarks: 148. Improved: $\large\color{#35bf28}10$. 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] 80.6550μs 79.9370μs 12.5098 KOps/s 12.5268 KOps/s $\color{#d91a1a}-0.14\%$
test_tensor_to_bytestream_speed[torch.save] 0.1400ms 0.1393ms 7.1768 KOps/s 7.0768 KOps/s $\color{#35bf28}+1.41\%$
test_tensor_to_bytestream_speed[untyped_storage] 0.1087s 0.1080s 9.2630 Ops/s 9.1477 Ops/s $\color{#35bf28}+1.26\%$
test_tensor_to_bytestream_speed[numpy] 2.4237μs 2.4210μs 413.0442 KOps/s 402.2458 KOps/s $\color{#35bf28}+2.68\%$
test_tensor_to_bytestream_speed[safetensors] 41.5517μs 41.3523μs 24.1824 KOps/s 26.7558 KOps/s $\textbf{\color{#d91a1a}-9.62\%}$
test_simple 0.9134s 0.8221s 1.2164 Ops/s 1.2331 Ops/s $\color{#d91a1a}-1.36\%$
test_transformed 1.5318s 1.4391s 0.6949 Ops/s 0.6948 Ops/s $+0.01\%$
test_serial 2.3996s 2.3123s 0.4325 Ops/s 0.4348 Ops/s $\color{#d91a1a}-0.55\%$
test_parallel 2.1295s 1.9901s 0.5025 Ops/s 0.5126 Ops/s $\color{#d91a1a}-1.98\%$
test_step_mdp_speed[True-True-True-True-True] 0.2991ms 45.5470μs 21.9554 KOps/s 22.0954 KOps/s $\color{#d91a1a}-0.63\%$
test_step_mdp_speed[True-True-True-True-False] 52.4910μs 24.9965μs 40.0055 KOps/s 39.9796 KOps/s $\color{#35bf28}+0.06\%$
test_step_mdp_speed[True-True-True-False-True] 56.9010μs 25.1770μs 39.7188 KOps/s 40.4695 KOps/s $\color{#d91a1a}-1.85\%$
test_step_mdp_speed[True-True-True-False-False] 44.0700μs 13.8443μs 72.2319 KOps/s 72.8005 KOps/s $\color{#d91a1a}-0.78\%$
test_step_mdp_speed[True-True-False-True-True] 85.9410μs 47.9541μs 20.8533 KOps/s 20.6653 KOps/s $\color{#35bf28}+0.91\%$
test_step_mdp_speed[True-True-False-True-False] 58.5310μs 28.0082μs 35.7038 KOps/s 35.7627 KOps/s $\color{#d91a1a}-0.16\%$
test_step_mdp_speed[True-True-False-False-True] 56.6410μs 28.0295μs 35.6767 KOps/s 35.4423 KOps/s $\color{#35bf28}+0.66\%$
test_step_mdp_speed[True-True-False-False-False] 45.5210μs 16.6563μs 60.0374 KOps/s 59.0601 KOps/s $\color{#35bf28}+1.65\%$
test_step_mdp_speed[True-False-True-True-True] 94.3920μs 50.6315μs 19.7505 KOps/s 19.5944 KOps/s $\color{#35bf28}+0.80\%$
test_step_mdp_speed[True-False-True-True-False] 61.5810μs 30.1456μs 33.1724 KOps/s 32.4211 KOps/s $\color{#35bf28}+2.32\%$
test_step_mdp_speed[True-False-True-False-True] 66.1910μs 27.3786μs 36.5249 KOps/s 36.1512 KOps/s $\color{#35bf28}+1.03\%$
test_step_mdp_speed[True-False-True-False-False] 42.3400μs 16.4673μs 60.7266 KOps/s 60.8964 KOps/s $\color{#d91a1a}-0.28\%$
test_step_mdp_speed[True-False-False-True-True] 91.3320μs 52.7270μs 18.9656 KOps/s 18.7339 KOps/s $\color{#35bf28}+1.24\%$
test_step_mdp_speed[True-False-False-True-False] 59.3510μs 32.9129μs 30.3833 KOps/s 30.1368 KOps/s $\color{#35bf28}+0.82\%$
test_step_mdp_speed[True-False-False-False-True] 60.0710μs 30.2147μs 33.0965 KOps/s 33.3635 KOps/s $\color{#d91a1a}-0.80\%$
test_step_mdp_speed[True-False-False-False-False] 45.1310μs 19.2633μs 51.9123 KOps/s 51.9100 KOps/s $+0.00\%$
test_step_mdp_speed[False-True-True-True-True] 88.6710μs 49.5836μs 20.1680 KOps/s 19.7862 KOps/s $\color{#35bf28}+1.93\%$
test_step_mdp_speed[False-True-True-True-False] 72.7610μs 30.4566μs 32.8336 KOps/s 32.5055 KOps/s $\color{#35bf28}+1.01\%$
test_step_mdp_speed[False-True-True-False-True] 67.5010μs 32.1842μs 31.0711 KOps/s 31.3322 KOps/s $\color{#d91a1a}-0.83\%$
test_step_mdp_speed[False-True-True-False-False] 54.3710μs 18.2327μs 54.8466 KOps/s 54.6773 KOps/s $\color{#35bf28}+0.31\%$
test_step_mdp_speed[False-True-False-True-True] 2.6988ms 53.6425μs 18.6419 KOps/s 18.7954 KOps/s $\color{#d91a1a}-0.82\%$
test_step_mdp_speed[False-True-False-True-False] 72.5320μs 33.5742μs 29.7848 KOps/s 29.6135 KOps/s $\color{#35bf28}+0.58\%$
test_step_mdp_speed[False-True-False-False-True] 63.7910μs 35.1301μs 28.4656 KOps/s 29.2114 KOps/s $\color{#d91a1a}-2.55\%$
test_step_mdp_speed[False-True-False-False-False] 49.7510μs 21.1949μs 47.1811 KOps/s 47.7937 KOps/s $\color{#d91a1a}-1.28\%$
test_step_mdp_speed[False-False-True-True-True] 95.0710μs 55.8622μs 17.9012 KOps/s 18.0000 KOps/s $\color{#d91a1a}-0.55\%$
test_step_mdp_speed[False-False-True-True-False] 63.3010μs 36.0873μs 27.7106 KOps/s 27.7768 KOps/s $\color{#d91a1a}-0.24\%$
test_step_mdp_speed[False-False-True-False-True] 91.6620μs 34.8074μs 28.7295 KOps/s 29.1311 KOps/s $\color{#d91a1a}-1.38\%$
test_step_mdp_speed[False-False-True-False-False] 54.2610μs 21.0132μs 47.5892 KOps/s 48.4879 KOps/s $\color{#d91a1a}-1.85\%$
test_step_mdp_speed[False-False-False-True-True] 98.3210μs 58.8357μs 16.9965 KOps/s 17.2672 KOps/s $\color{#d91a1a}-1.57\%$
test_step_mdp_speed[False-False-False-True-False] 72.5520μs 38.7812μs 25.7857 KOps/s 26.3027 KOps/s $\color{#d91a1a}-1.97\%$
test_step_mdp_speed[False-False-False-False-True] 70.6710μs 36.7054μs 27.2439 KOps/s 27.8217 KOps/s $\color{#d91a1a}-2.08\%$
test_step_mdp_speed[False-False-False-False-False] 60.1010μs 23.3887μs 42.7556 KOps/s 43.2936 KOps/s $\color{#d91a1a}-1.24\%$
test_non_tensor_env_rollout_speed[1000-single-True] 0.8657s 0.7922s 1.2622 Ops/s 1.3077 Ops/s $\color{#d91a1a}-3.47\%$
test_non_tensor_env_rollout_speed[1000-single-False] 0.7297s 0.6336s 1.5782 Ops/s 1.5768 Ops/s $\color{#35bf28}+0.09\%$
test_non_tensor_env_rollout_speed[1000-serial-no-buffers-True] 1.7515s 1.6763s 0.5966 Ops/s 0.5987 Ops/s $\color{#d91a1a}-0.36\%$
test_non_tensor_env_rollout_speed[1000-serial-no-buffers-False] 1.5361s 1.4574s 0.6862 Ops/s 0.6889 Ops/s $\color{#d91a1a}-0.39\%$
test_non_tensor_env_rollout_speed[1000-serial-buffers-True] 2.0001s 1.9233s 0.5199 Ops/s 0.5216 Ops/s $\color{#d91a1a}-0.31\%$
test_non_tensor_env_rollout_speed[1000-serial-buffers-False] 1.7789s 1.7041s 0.5868 Ops/s 0.5912 Ops/s $\color{#d91a1a}-0.73\%$
test_non_tensor_env_rollout_speed[1000-parallel-no-buffers-True] 4.7778s 4.6940s 0.2130 Ops/s 0.2153 Ops/s $\color{#d91a1a}-1.04\%$
test_non_tensor_env_rollout_speed[1000-parallel-no-buffers-False] 4.4966s 4.4120s 0.2267 Ops/s 0.2240 Ops/s $\color{#35bf28}+1.20\%$
test_non_tensor_env_rollout_speed[1000-parallel-buffers-True] 2.0198s 1.9636s 0.5093 Ops/s 0.5121 Ops/s $\color{#d91a1a}-0.55\%$
test_non_tensor_env_rollout_speed[1000-parallel-buffers-False] 1.7460s 1.6837s 0.5939 Ops/s 0.6054 Ops/s $\color{#d91a1a}-1.89\%$
test_values[generalized_advantage_estimate-True-True] 22.2794ms 20.3270ms 49.1956 Ops/s 49.5326 Ops/s $\color{#d91a1a}-0.68\%$
test_values[vec_generalized_advantage_estimate-True-True] 0.1335s 3.5883ms 278.6832 Ops/s 271.1529 Ops/s $\color{#35bf28}+2.78\%$
test_values[td0_return_estimate-False-False] 0.1083ms 84.2832μs 11.8648 KOps/s 12.0654 KOps/s $\color{#d91a1a}-1.66\%$
test_values[td1_return_estimate-False-False] 51.9807ms 49.1877ms 20.3303 Ops/s 20.7130 Ops/s $\color{#d91a1a}-1.85\%$
test_values[vec_td1_return_estimate-False-False] 1.3130ms 1.0866ms 920.2733 Ops/s 920.2963 Ops/s $-0.00\%$
test_values[td_lambda_return_estimate-True-False] 85.3806ms 80.1876ms 12.4708 Ops/s 12.5905 Ops/s $\color{#d91a1a}-0.95\%$
test_values[vec_td_lambda_return_estimate-True-False] 1.3199ms 1.0901ms 917.3654 Ops/s 922.7434 Ops/s $\color{#d91a1a}-0.58\%$
test_gae_speed[generalized_advantage_estimate-False-1-512] 22.3513ms 21.5747ms 46.3505 Ops/s 49.0641 Ops/s $\textbf{\color{#d91a1a}-5.53\%}$
test_gae_speed[vec_generalized_advantage_estimate-True-1-512] 1.0165ms 0.7527ms 1.3286 KOps/s 1.3355 KOps/s $\color{#d91a1a}-0.52\%$
test_gae_speed[vec_generalized_advantage_estimate-False-1-512] 0.7358ms 0.6764ms 1.4785 KOps/s 1.4444 KOps/s $\color{#35bf28}+2.36\%$
test_gae_speed[vec_generalized_advantage_estimate-True-32-512] 1.5998ms 1.4922ms 670.1473 Ops/s 672.0975 Ops/s $\color{#d91a1a}-0.29\%$
test_gae_speed[vec_generalized_advantage_estimate-False-32-512] 0.7931ms 0.7302ms 1.3695 KOps/s 1.4405 KOps/s $\color{#d91a1a}-4.94\%$
test_dqn_speed[False-None] 1.6276ms 1.5384ms 650.0415 Ops/s 648.6059 Ops/s $\color{#35bf28}+0.22\%$
test_dqn_speed[False-backward] 2.4149ms 2.1770ms 459.3486 Ops/s 459.8529 Ops/s $\color{#d91a1a}-0.11\%$
test_dqn_speed[True-None] 0.6550ms 0.5779ms 1.7303 KOps/s 1.7138 KOps/s $\color{#35bf28}+0.96\%$
test_dqn_speed[True-backward] 1.3109ms 1.2364ms 808.8190 Ops/s 880.4284 Ops/s $\textbf{\color{#d91a1a}-8.13\%}$
test_dqn_speed[reduce-overhead-None] 0.6760ms 0.6020ms 1.6611 KOps/s 1.6092 KOps/s $\color{#35bf28}+3.22\%$
test_ddpg_speed[False-None] 3.3862ms 2.9289ms 341.4297 Ops/s 346.6141 Ops/s $\color{#d91a1a}-1.50\%$
test_ddpg_speed[False-backward] 4.7296ms 4.2924ms 232.9698 Ops/s 240.0854 Ops/s $\color{#d91a1a}-2.96\%$
test_ddpg_speed[True-None] 1.5721ms 1.3631ms 733.6076 Ops/s 733.9047 Ops/s $\color{#d91a1a}-0.04\%$
test_ddpg_speed[True-backward] 2.9547ms 2.6044ms 383.9651 Ops/s 403.4449 Ops/s $\color{#d91a1a}-4.83\%$
test_ddpg_speed[reduce-overhead-None] 1.4866ms 1.3848ms 722.1047 Ops/s 718.6880 Ops/s $\color{#35bf28}+0.48\%$
test_sac_speed[False-None] 8.8300ms 8.3646ms 119.5514 Ops/s 119.6975 Ops/s $\color{#d91a1a}-0.12\%$
test_sac_speed[False-backward] 11.9345ms 11.4931ms 87.0090 Ops/s 88.7854 Ops/s $\color{#d91a1a}-2.00\%$
test_sac_speed[True-None] 2.0889ms 1.8964ms 527.3217 Ops/s 528.4743 Ops/s $\color{#d91a1a}-0.22\%$
test_sac_speed[True-backward] 4.1725ms 3.7478ms 266.8237 Ops/s 278.9937 Ops/s $\color{#d91a1a}-4.36\%$
test_sac_speed[reduce-overhead-None] 19.0468ms 10.6319ms 94.0566 Ops/s 96.8153 Ops/s $\color{#d91a1a}-2.85\%$
test_redq_deprec_speed[False-None] 9.8742ms 9.2973ms 107.5577 Ops/s 107.2271 Ops/s $\color{#35bf28}+0.31\%$
test_redq_deprec_speed[False-backward] 13.0730ms 12.6834ms 78.8430 Ops/s 80.4764 Ops/s $\color{#d91a1a}-2.03\%$
test_redq_deprec_speed[True-None] 2.7279ms 2.6236ms 381.1503 Ops/s 385.1137 Ops/s $\color{#d91a1a}-1.03\%$
test_redq_deprec_speed[True-backward] 4.8730ms 4.4791ms 223.2581 Ops/s 224.9204 Ops/s $\color{#d91a1a}-0.74\%$
test_redq_deprec_speed[reduce-overhead-None] 15.5330ms 9.6157ms 103.9966 Ops/s 87.4454 Ops/s $\textbf{\color{#35bf28}+18.93\%}$
test_td3_speed[False-None] 8.3900ms 8.2075ms 121.8399 Ops/s 119.6882 Ops/s $\color{#35bf28}+1.80\%$
test_td3_speed[False-backward] 11.4534ms 10.7670ms 92.8761 Ops/s 91.4794 Ops/s $\color{#35bf28}+1.53\%$
test_td3_speed[True-None] 1.7802ms 1.7477ms 572.1646 Ops/s 578.8839 Ops/s $\color{#d91a1a}-1.16\%$
test_td3_speed[True-backward] 3.8809ms 3.4267ms 291.8253 Ops/s 306.6035 Ops/s $\color{#d91a1a}-4.82\%$
test_td3_speed[reduce-overhead-None] 54.6299ms 23.9809ms 41.6999 Ops/s 40.8580 Ops/s $\color{#35bf28}+2.06\%$
test_cql_speed[False-None] 17.6026ms 17.2724ms 57.8959 Ops/s 57.8211 Ops/s $\color{#35bf28}+0.13\%$
test_cql_speed[False-backward] 23.5491ms 22.7585ms 43.9397 Ops/s 44.4029 Ops/s $\color{#d91a1a}-1.04\%$
test_cql_speed[True-None] 3.6292ms 3.5013ms 285.6056 Ops/s 285.8959 Ops/s $\color{#d91a1a}-0.10\%$
test_cql_speed[True-backward] 6.3120ms 5.8567ms 170.7434 Ops/s 174.8158 Ops/s $\color{#d91a1a}-2.33\%$
test_cql_speed[reduce-overhead-None] 18.4127ms 11.6048ms 86.1710 Ops/s 85.4700 Ops/s $\color{#35bf28}+0.82\%$
test_a2c_speed[False-None] 3.4594ms 3.2461ms 308.0653 Ops/s 309.7721 Ops/s $\color{#d91a1a}-0.55\%$
test_a2c_speed[False-backward] 6.8847ms 6.3818ms 156.6952 Ops/s 164.3942 Ops/s $\color{#d91a1a}-4.68\%$
test_a2c_speed[True-None] 1.5263ms 1.3653ms 732.4304 Ops/s 725.4426 Ops/s $\color{#35bf28}+0.96\%$
test_a2c_speed[True-backward] 3.3145ms 3.2154ms 311.0049 Ops/s 322.8584 Ops/s $\color{#d91a1a}-3.67\%$
test_a2c_speed[reduce-overhead-None] 1.0762ms 0.9636ms 1.0378 KOps/s 1.0441 KOps/s $\color{#d91a1a}-0.61\%$
test_ppo_speed[False-None] 4.1347ms 3.8561ms 259.3305 Ops/s 261.0272 Ops/s $\color{#d91a1a}-0.65\%$
test_ppo_speed[False-backward] 7.5274ms 7.1264ms 140.3243 Ops/s 139.7710 Ops/s $\color{#35bf28}+0.40\%$
test_ppo_speed[True-None] 1.5495ms 1.4305ms 699.0572 Ops/s 690.7955 Ops/s $\color{#35bf28}+1.20\%$
test_ppo_speed[True-backward] 3.7754ms 3.4199ms 292.4065 Ops/s 290.8268 Ops/s $\color{#35bf28}+0.54\%$
test_ppo_speed[reduce-overhead-None] 1.1333ms 1.0186ms 981.7440 Ops/s 947.6558 Ops/s $\color{#35bf28}+3.60\%$
test_reinforce_speed[False-None] 2.4170ms 2.2724ms 440.0726 Ops/s 436.6690 Ops/s $\color{#35bf28}+0.78\%$
test_reinforce_speed[False-backward] 3.7302ms 3.3960ms 294.4682 Ops/s 291.4633 Ops/s $\color{#35bf28}+1.03\%$
test_reinforce_speed[True-None] 1.3957ms 1.2870ms 777.0221 Ops/s 747.6739 Ops/s $\color{#35bf28}+3.93\%$
test_reinforce_speed[True-backward] 3.2706ms 3.2027ms 312.2334 Ops/s 308.5969 Ops/s $\color{#35bf28}+1.18\%$
test_reinforce_speed[reduce-overhead-None] 0.4636s 10.1336ms 98.6819 Ops/s 95.3281 Ops/s $\color{#35bf28}+3.52\%$
test_iql_speed[False-None] 9.9562ms 9.4326ms 106.0155 Ops/s 105.1858 Ops/s $\color{#35bf28}+0.79\%$
test_iql_speed[False-backward] 13.8225ms 13.3645ms 74.8252 Ops/s 74.3145 Ops/s $\color{#35bf28}+0.69\%$
test_iql_speed[True-None] 2.4730ms 2.2867ms 437.3100 Ops/s 438.6174 Ops/s $\color{#d91a1a}-0.30\%$
test_iql_speed[True-backward] 5.6677ms 5.1620ms 193.7241 Ops/s 197.8853 Ops/s $\color{#d91a1a}-2.10\%$
test_iql_speed[reduce-overhead-None] 0.7098s 12.4455ms 80.3501 Ops/s 98.3628 Ops/s $\textbf{\color{#d91a1a}-18.31\%}$
test_rb_sample[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 7.5372ms 6.1260ms 163.2378 Ops/s 162.8691 Ops/s $\color{#35bf28}+0.23\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 0.6366ms 0.3528ms 2.8343 KOps/s 2.9750 KOps/s $\color{#d91a1a}-4.73\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.6242ms 0.3807ms 2.6265 KOps/s 3.0912 KOps/s $\textbf{\color{#d91a1a}-15.03\%}$
test_rb_sample[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 6.2249ms 5.9193ms 168.9376 Ops/s 172.4118 Ops/s $\color{#d91a1a}-2.02\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 0.6413ms 0.3011ms 3.3208 KOps/s 3.4821 KOps/s $\color{#d91a1a}-4.63\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.6847ms 0.2959ms 3.3795 KOps/s 3.6975 KOps/s $\textbf{\color{#d91a1a}-8.60\%}$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-sampler6-10000] 1.5028ms 1.2925ms 773.7075 Ops/s 772.9979 Ops/s $\color{#35bf28}+0.09\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-sampler7-10000] 1.4757ms 1.2201ms 819.5942 Ops/s 829.3995 Ops/s $\color{#d91a1a}-1.18\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 10.0696ms 6.1953ms 161.4117 Ops/s 166.7759 Ops/s $\color{#d91a1a}-3.22\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 0.7237ms 0.4538ms 2.2037 KOps/s 1.8273 KOps/s $\textbf{\color{#35bf28}+20.60\%}$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.8104ms 0.4253ms 2.3512 KOps/s 1.9531 KOps/s $\textbf{\color{#35bf28}+20.39\%}$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 6.0969ms 5.9500ms 168.0676 Ops/s 169.4126 Ops/s $\color{#d91a1a}-0.79\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 2.1937ms 0.3253ms 3.0744 KOps/s 3.3487 KOps/s $\textbf{\color{#d91a1a}-8.19\%}$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.5982ms 0.2865ms 3.4899 KOps/s 2.9552 KOps/s $\textbf{\color{#35bf28}+18.09\%}$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 6.1448ms 5.8860ms 169.8950 Ops/s 170.5504 Ops/s $\color{#d91a1a}-0.38\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 2.3889ms 0.3457ms 2.8928 KOps/s 3.4495 KOps/s $\textbf{\color{#d91a1a}-16.14\%}$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.4889ms 0.2918ms 3.4270 KOps/s 3.6878 KOps/s $\textbf{\color{#d91a1a}-7.07\%}$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 6.1751ms 6.0405ms 165.5496 Ops/s 165.6625 Ops/s $\color{#d91a1a}-0.07\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 2.3829ms 0.4900ms 2.0407 KOps/s 2.1021 KOps/s $\color{#d91a1a}-2.92\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.7789ms 0.4103ms 2.4370 KOps/s 2.3142 KOps/s $\textbf{\color{#35bf28}+5.30\%}$
test_rb_populate[TensorDictReplayBuffer-ListStorage-RandomSampler-400] 6.5223ms 5.0101ms 199.5956 Ops/s 49.1917 Ops/s $\textbf{\color{#35bf28}+305.75\%}$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-400] 11.6726ms 2.3268ms 429.7667 Ops/s 512.8333 Ops/s $\textbf{\color{#d91a1a}-16.20\%}$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-400] 1.0889ms 0.9113ms 1.0973 KOps/s 756.2301 Ops/s $\textbf{\color{#35bf28}+45.11\%}$
test_rb_populate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-400] 0.5878s 16.7205ms 59.8068 Ops/s 196.6993 Ops/s $\textbf{\color{#d91a1a}-69.59\%}$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-400] 4.0066ms 1.8815ms 531.4863 Ops/s 551.1366 Ops/s $\color{#d91a1a}-3.57\%$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-400] 2.1192ms 1.1431ms 874.7973 Ops/s 1.0930 KOps/s $\textbf{\color{#d91a1a}-19.97\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-ListStorage-None-400] 8.0495ms 5.2422ms 190.7611 Ops/s 189.0516 Ops/s $\color{#35bf28}+0.90\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-400] 4.4382ms 1.9186ms 521.2098 Ops/s 490.8434 Ops/s $\textbf{\color{#35bf28}+6.19\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-400] 1.2435ms 1.0634ms 940.4021 Ops/s 732.1732 Ops/s $\textbf{\color{#35bf28}+28.44\%}$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-True] 37.8108ms 34.0797ms 29.3430 Ops/s 28.5829 Ops/s $\color{#35bf28}+2.66\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-False] 18.7099ms 17.3787ms 57.5418 Ops/s 55.6263 Ops/s $\color{#35bf28}+3.44\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-True] 37.1226ms 35.0192ms 28.5557 Ops/s 27.7238 Ops/s $\color{#35bf28}+3.00\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-False] 19.7041ms 17.8219ms 56.1107 Ops/s 53.8053 Ops/s $\color{#35bf28}+4.28\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-True] 37.2934ms 36.7799ms 27.1887 Ops/s 26.5778 Ops/s $\color{#35bf28}+2.30\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-False] 20.5787ms 19.2680ms 51.8995 Ops/s 49.3250 Ops/s $\textbf{\color{#35bf28}+5.22\%}$

@vmoens vmoens closed this Jan 21, 2026
@vmoens vmoens deleted the fix-num-envs-gym-tests branch January 21, 2026 13:37
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.

3 participants