Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[Feature] Prevent loading existing mmap files in storages if they already exist #2438

Merged
merged 1 commit into from
Sep 17, 2024

Conversation

vmoens
Copy link
Contributor

@vmoens vmoens commented Sep 16, 2024

[ghstack-poisoned]
Copy link

pytorch-bot bot commented Sep 16, 2024

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/rl/2438

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

❌ 3 New Failures, 8 Unrelated Failures

As of commit e4c6c20 with merge base 36545af (image):

NEW FAILURES - The following jobs have failed:

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.

vmoens added a commit that referenced this pull request Sep 16, 2024
…eady exist

ghstack-source-id: 63bcb1e0420620d5dcd2b73d8e0a5b3bf137c8e1
Pull Request resolved: #2438
@facebook-github-bot facebook-github-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 Sep 16, 2024
@vmoens vmoens added the enhancement New feature or request label Sep 16, 2024
Copy link

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

Total Benchmarks: 91. Improved: $\large\color{#35bf28}10$. Worsened: $\large\color{#d91a1a}2$.

Expand to view detailed results
Name Max Mean Ops Ops on Repo HEAD Change
test_single 69.2737ms 60.4216ms 16.5504 Ops/s 16.6552 Ops/s $\color{#d91a1a}-0.63\%$
test_sync 41.8773ms 34.1663ms 29.2686 Ops/s 27.5518 Ops/s $\textbf{\color{#35bf28}+6.23\%}$
test_async 55.5498ms 31.2056ms 32.0455 Ops/s 31.8765 Ops/s $\color{#35bf28}+0.53\%$
test_simple 0.5053s 0.4259s 2.3482 Ops/s 2.4404 Ops/s $\color{#d91a1a}-3.78\%$
test_transformed 0.6649s 0.5899s 1.6953 Ops/s 1.7300 Ops/s $\color{#d91a1a}-2.01\%$
test_serial 1.2894s 1.2667s 0.7894 Ops/s 0.7806 Ops/s $\color{#35bf28}+1.14\%$
test_parallel 1.1860s 1.1216s 0.8916 Ops/s 0.8682 Ops/s $\color{#35bf28}+2.69\%$
test_step_mdp_speed[True-True-True-True-True] 0.2483ms 28.3029μs 35.3321 KOps/s 35.9151 KOps/s $\color{#d91a1a}-1.62\%$
test_step_mdp_speed[True-True-True-True-False] 48.8410μs 16.8402μs 59.3816 KOps/s 60.9793 KOps/s $\color{#d91a1a}-2.62\%$
test_step_mdp_speed[True-True-True-False-True] 53.2100μs 16.2397μs 61.5776 KOps/s 63.1348 KOps/s $\color{#d91a1a}-2.47\%$
test_step_mdp_speed[True-True-True-False-False] 40.8760μs 9.6475μs 103.6534 KOps/s 107.3286 KOps/s $\color{#d91a1a}-3.42\%$
test_step_mdp_speed[True-True-False-True-True] 63.4190μs 30.2531μs 33.0545 KOps/s 34.0496 KOps/s $\color{#d91a1a}-2.92\%$
test_step_mdp_speed[True-True-False-True-False] 69.4630μs 18.6866μs 53.5142 KOps/s 56.2056 KOps/s $\color{#d91a1a}-4.79\%$
test_step_mdp_speed[True-True-False-False-True] 73.3370μs 17.9777μs 55.6245 KOps/s 56.8792 KOps/s $\color{#d91a1a}-2.21\%$
test_step_mdp_speed[True-True-False-False-False] 45.8960μs 11.3149μs 88.3793 KOps/s 90.3840 KOps/s $\color{#d91a1a}-2.22\%$
test_step_mdp_speed[True-False-True-True-True] 78.8780μs 32.0962μs 31.1564 KOps/s 31.8493 KOps/s $\color{#d91a1a}-2.18\%$
test_step_mdp_speed[True-False-True-True-False] 74.1190μs 20.4089μs 48.9982 KOps/s 50.3868 KOps/s $\color{#d91a1a}-2.76\%$
test_step_mdp_speed[True-False-True-False-True] 51.2660μs 18.1168μs 55.1973 KOps/s 56.5697 KOps/s $\color{#d91a1a}-2.43\%$
test_step_mdp_speed[True-False-True-False-False] 43.2810μs 11.4421μs 87.3966 KOps/s 90.9017 KOps/s $\color{#d91a1a}-3.86\%$
test_step_mdp_speed[True-False-False-True-True] 83.4460μs 33.4317μs 29.9117 KOps/s 29.8215 KOps/s $\color{#35bf28}+0.30\%$
test_step_mdp_speed[True-False-False-True-False] 69.0790μs 22.1053μs 45.2380 KOps/s 47.6689 KOps/s $\textbf{\color{#d91a1a}-5.10\%}$
test_step_mdp_speed[True-False-False-False-True] 60.7240μs 19.6723μs 50.8329 KOps/s 52.5170 KOps/s $\color{#d91a1a}-3.21\%$
test_step_mdp_speed[True-False-False-False-False] 39.0730μs 13.0732μs 76.4924 KOps/s 80.3716 KOps/s $\color{#d91a1a}-4.83\%$
test_step_mdp_speed[False-True-True-True-True] 66.6240μs 31.8606μs 31.3868 KOps/s 31.6120 KOps/s $\color{#d91a1a}-0.71\%$
test_step_mdp_speed[False-True-True-True-False] 72.6360μs 20.4133μs 48.9877 KOps/s 50.9624 KOps/s $\color{#d91a1a}-3.87\%$
test_step_mdp_speed[False-True-True-False-True] 85.8910μs 21.0879μs 47.4206 KOps/s 48.9400 KOps/s $\color{#d91a1a}-3.10\%$
test_step_mdp_speed[False-True-True-False-False] 54.4020μs 12.7560μs 78.3946 KOps/s 81.5455 KOps/s $\color{#d91a1a}-3.86\%$
test_step_mdp_speed[False-True-False-True-True] 0.1035ms 33.5316μs 29.8226 KOps/s 30.4253 KOps/s $\color{#d91a1a}-1.98\%$
test_step_mdp_speed[False-True-False-True-False] 76.3430μs 21.7747μs 45.9248 KOps/s 47.1199 KOps/s $\color{#d91a1a}-2.54\%$
test_step_mdp_speed[False-True-False-False-True] 2.7692ms 22.3201μs 44.8026 KOps/s 45.2065 KOps/s $\color{#d91a1a}-0.89\%$
test_step_mdp_speed[False-True-False-False-False] 44.0430μs 14.2696μs 70.0792 KOps/s 72.4161 KOps/s $\color{#d91a1a}-3.23\%$
test_step_mdp_speed[False-False-True-True-True] 90.0690μs 34.9675μs 28.5980 KOps/s 28.5941 KOps/s $\color{#35bf28}+0.01\%$
test_step_mdp_speed[False-False-True-True-False] 62.7180μs 23.3654μs 42.7983 KOps/s 43.7752 KOps/s $\color{#d91a1a}-2.23\%$
test_step_mdp_speed[False-False-True-False-True] 64.9220μs 22.5022μs 44.4400 KOps/s 45.2187 KOps/s $\color{#d91a1a}-1.72\%$
test_step_mdp_speed[False-False-True-False-False] 38.3720μs 14.4439μs 69.2333 KOps/s 72.7621 KOps/s $\color{#d91a1a}-4.85\%$
test_step_mdp_speed[False-False-False-True-True] 0.1128ms 36.3610μs 27.5020 KOps/s 27.9916 KOps/s $\color{#d91a1a}-1.75\%$
test_step_mdp_speed[False-False-False-True-False] 88.7660μs 24.8528μs 40.2369 KOps/s 41.2370 KOps/s $\color{#d91a1a}-2.43\%$
test_step_mdp_speed[False-False-False-False-True] 75.4110μs 23.6535μs 42.2770 KOps/s 44.1087 KOps/s $\color{#d91a1a}-4.15\%$
test_step_mdp_speed[False-False-False-False-False] 53.4000μs 15.7760μs 63.3874 KOps/s 65.7965 KOps/s $\color{#d91a1a}-3.66\%$
test_values[generalized_advantage_estimate-True-True] 10.0815ms 9.5496ms 104.7169 Ops/s 102.6616 Ops/s $\color{#35bf28}+2.00\%$
test_values[vec_generalized_advantage_estimate-True-True] 39.6010ms 33.8617ms 29.5319 Ops/s 29.2978 Ops/s $\color{#35bf28}+0.80\%$
test_values[td0_return_estimate-False-False] 0.2348ms 0.1747ms 5.7240 KOps/s 5.1900 KOps/s $\textbf{\color{#35bf28}+10.29\%}$
test_values[td1_return_estimate-False-False] 25.7492ms 23.9721ms 41.7152 Ops/s 40.5057 Ops/s $\color{#35bf28}+2.99\%$
test_values[vec_td1_return_estimate-False-False] 35.9164ms 33.6134ms 29.7500 Ops/s 29.5501 Ops/s $\color{#35bf28}+0.68\%$
test_values[td_lambda_return_estimate-True-False] 34.8013ms 34.3939ms 29.0749 Ops/s 28.5613 Ops/s $\color{#35bf28}+1.80\%$
test_values[vec_td_lambda_return_estimate-True-False] 35.4295ms 33.5322ms 29.8221 Ops/s 29.1321 Ops/s $\color{#35bf28}+2.37\%$
test_gae_speed[generalized_advantage_estimate-False-1-512] 8.5356ms 8.3141ms 120.2778 Ops/s 117.9450 Ops/s $\color{#35bf28}+1.98\%$
test_gae_speed[vec_generalized_advantage_estimate-True-1-512] 2.1505ms 1.8743ms 533.5425 Ops/s 497.8504 Ops/s $\textbf{\color{#35bf28}+7.17\%}$
test_gae_speed[vec_generalized_advantage_estimate-False-1-512] 0.4379ms 0.3539ms 2.8253 KOps/s 2.7834 KOps/s $\color{#35bf28}+1.50\%$
test_gae_speed[vec_generalized_advantage_estimate-True-32-512] 41.0349ms 40.0793ms 24.9506 Ops/s 21.2170 Ops/s $\textbf{\color{#35bf28}+17.60\%}$
test_gae_speed[vec_generalized_advantage_estimate-False-32-512] 4.0178ms 3.0460ms 328.3028 Ops/s 325.3634 Ops/s $\color{#35bf28}+0.90\%$
test_dqn_speed 1.7687ms 1.2969ms 771.0990 Ops/s 758.7540 Ops/s $\color{#35bf28}+1.63\%$
test_ddpg_speed 3.4756ms 2.7052ms 369.6589 Ops/s 358.7066 Ops/s $\color{#35bf28}+3.05\%$
test_sac_speed 9.7579ms 8.1034ms 123.4054 Ops/s 121.1334 Ops/s $\color{#35bf28}+1.88\%$
test_redq_speed 14.5687ms 12.9737ms 77.0790 Ops/s 75.5090 Ops/s $\color{#35bf28}+2.08\%$
test_redq_deprec_speed 14.6437ms 13.2202ms 75.6418 Ops/s 73.8574 Ops/s $\color{#35bf28}+2.42\%$
test_td3_speed 11.9550ms 8.4269ms 118.6671 Ops/s 120.0827 Ops/s $\color{#d91a1a}-1.18\%$
test_cql_speed 37.1553ms 35.7042ms 28.0079 Ops/s 27.6258 Ops/s $\color{#35bf28}+1.38\%$
test_a2c_speed 8.8474ms 7.7080ms 129.7353 Ops/s 135.6302 Ops/s $\color{#d91a1a}-4.35\%$
test_ppo_speed 9.4308ms 7.7936ms 128.3098 Ops/s 124.2668 Ops/s $\color{#35bf28}+3.25\%$
test_reinforce_speed 7.5583ms 6.6864ms 149.5573 Ops/s 148.1568 Ops/s $\color{#35bf28}+0.95\%$
test_iql_speed 33.7023ms 32.4257ms 30.8398 Ops/s 30.8283 Ops/s $\color{#35bf28}+0.04\%$
test_rb_sample[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 5.7478ms 5.3273ms 187.7111 Ops/s 178.8059 Ops/s $\color{#35bf28}+4.98\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 0.9333ms 0.4847ms 2.0629 KOps/s 1.9539 KOps/s $\textbf{\color{#35bf28}+5.58\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.6981ms 0.4597ms 2.1752 KOps/s 2.1002 KOps/s $\color{#35bf28}+3.57\%$
test_rb_sample[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 6.3392ms 5.2116ms 191.8800 Ops/s 177.1249 Ops/s $\textbf{\color{#35bf28}+8.33\%}$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 0.8598ms 0.4842ms 2.0651 KOps/s 1.9871 KOps/s $\color{#35bf28}+3.93\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.6566ms 0.4560ms 2.1931 KOps/s 2.1622 KOps/s $\color{#35bf28}+1.43\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-sampler6-10000] 1.7743ms 1.5977ms 625.8995 Ops/s 614.5825 Ops/s $\color{#35bf28}+1.84\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-sampler7-10000] 2.4294ms 1.5151ms 660.0030 Ops/s 637.1544 Ops/s $\color{#35bf28}+3.59\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 5.8156ms 5.4840ms 182.3488 Ops/s 173.4833 Ops/s $\textbf{\color{#35bf28}+5.11\%}$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 2.3450ms 0.6213ms 1.6095 KOps/s 1.5508 KOps/s $\color{#35bf28}+3.78\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.8367ms 0.5943ms 1.6825 KOps/s 1.6508 KOps/s $\color{#35bf28}+1.92\%$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 5.6069ms 5.2986ms 188.7279 Ops/s 181.0963 Ops/s $\color{#35bf28}+4.21\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 1.4105ms 0.4850ms 2.0620 KOps/s 2.0163 KOps/s $\color{#35bf28}+2.26\%$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.6682ms 0.4649ms 2.1512 KOps/s 2.1459 KOps/s $\color{#35bf28}+0.24\%$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 8.5848ms 5.3293ms 187.6415 Ops/s 187.2526 Ops/s $\color{#35bf28}+0.21\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 0.7988ms 0.4822ms 2.0740 KOps/s 2.0162 KOps/s $\color{#35bf28}+2.87\%$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.6570ms 0.4646ms 2.1525 KOps/s 2.1093 KOps/s $\color{#35bf28}+2.05\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 6.2428ms 5.6213ms 177.8934 Ops/s 176.4456 Ops/s $\color{#35bf28}+0.82\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 1.4086ms 0.6225ms 1.6064 KOps/s 1.5585 KOps/s $\color{#35bf28}+3.07\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.7741ms 0.5955ms 1.6793 KOps/s 1.6381 KOps/s $\color{#35bf28}+2.52\%$
test_rb_populate[TensorDictReplayBuffer-ListStorage-RandomSampler-400] 0.1431s 6.7421ms 148.3228 Ops/s 106.3414 Ops/s $\textbf{\color{#35bf28}+39.48\%}$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-400] 0.1332s 15.6995ms 63.6961 Ops/s 74.3841 Ops/s $\textbf{\color{#d91a1a}-14.37\%}$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-400] 1.7253ms 1.2051ms 829.8135 Ops/s 754.2606 Ops/s $\textbf{\color{#35bf28}+10.02\%}$
test_rb_populate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-400] 0.1244s 6.3899ms 156.4976 Ops/s 146.2610 Ops/s $\textbf{\color{#35bf28}+7.00\%}$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-400] 17.3040ms 13.0728ms 76.4949 Ops/s 75.6620 Ops/s $\color{#35bf28}+1.10\%$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-400] 1.7362ms 1.2221ms 818.2748 Ops/s 802.5884 Ops/s $\color{#35bf28}+1.95\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-ListStorage-None-400] 0.1274s 6.5692ms 152.2252 Ops/s 146.1808 Ops/s $\color{#35bf28}+4.13\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-400] 17.8614ms 13.2202ms 75.6419 Ops/s 74.5047 Ops/s $\color{#35bf28}+1.53\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-400] 1.8755ms 1.3597ms 735.4407 Ops/s 701.4669 Ops/s $\color{#35bf28}+4.84\%$

Copy link

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

Total Benchmarks: 94. Improved: $\large\color{#35bf28}5$. Worsened: $\large\color{#d91a1a}1$.

Expand to view detailed results
Name Max Mean Ops Ops on Repo HEAD Change
test_single 0.1036s 0.1033s 9.6851 Ops/s 9.6666 Ops/s $\color{#35bf28}+0.19\%$
test_sync 94.3686ms 90.8973ms 11.0014 Ops/s 10.7554 Ops/s $\color{#35bf28}+2.29\%$
test_async 0.1717s 87.9595ms 11.3689 Ops/s 11.4624 Ops/s $\color{#d91a1a}-0.82\%$
test_single_pixels 0.1110s 0.1098s 9.1086 Ops/s 9.1557 Ops/s $\color{#d91a1a}-0.51\%$
test_sync_pixels 72.3348ms 71.2438ms 14.0363 Ops/s 13.9421 Ops/s $\color{#35bf28}+0.68\%$
test_async_pixels 0.1316s 67.3069ms 14.8573 Ops/s 14.9632 Ops/s $\color{#d91a1a}-0.71\%$
test_simple 0.7343s 0.7337s 1.3630 Ops/s 1.3387 Ops/s $\color{#35bf28}+1.82\%$
test_transformed 0.9656s 0.9649s 1.0364 Ops/s 1.0263 Ops/s $\color{#35bf28}+0.98\%$
test_serial 2.1700s 2.0967s 0.4769 Ops/s 0.4754 Ops/s $\color{#35bf28}+0.32\%$
test_parallel 1.8933s 1.8617s 0.5371 Ops/s 0.5377 Ops/s $\color{#d91a1a}-0.10\%$
test_step_mdp_speed[True-True-True-True-True] 0.4247ms 38.3390μs 26.0831 KOps/s 26.1601 KOps/s $\color{#d91a1a}-0.29\%$
test_step_mdp_speed[True-True-True-True-False] 55.7310μs 21.6564μs 46.1757 KOps/s 46.3566 KOps/s $\color{#d91a1a}-0.39\%$
test_step_mdp_speed[True-True-True-False-True] 48.6010μs 21.7541μs 45.9683 KOps/s 46.2056 KOps/s $\color{#d91a1a}-0.51\%$
test_step_mdp_speed[True-True-True-False-False] 0.3956ms 12.3215μs 81.1592 KOps/s 79.9372 KOps/s $\color{#35bf28}+1.53\%$
test_step_mdp_speed[True-True-False-True-True] 0.4383ms 40.5355μs 24.6697 KOps/s 24.6439 KOps/s $\color{#35bf28}+0.10\%$
test_step_mdp_speed[True-True-False-True-False] 66.8710μs 23.7098μs 42.1767 KOps/s 41.8301 KOps/s $\color{#35bf28}+0.83\%$
test_step_mdp_speed[True-True-False-False-True] 0.4117ms 23.5654μs 42.4352 KOps/s 42.3349 KOps/s $\color{#35bf28}+0.24\%$
test_step_mdp_speed[True-True-False-False-False] 0.3987ms 14.6353μs 68.3279 KOps/s 69.0991 KOps/s $\color{#d91a1a}-1.12\%$
test_step_mdp_speed[True-False-True-True-True] 0.4479ms 42.6705μs 23.4354 KOps/s 23.7353 KOps/s $\color{#d91a1a}-1.26\%$
test_step_mdp_speed[True-False-True-True-False] 57.3210μs 25.9211μs 38.5785 KOps/s 38.6015 KOps/s $\color{#d91a1a}-0.06\%$
test_step_mdp_speed[True-False-True-False-True] 0.4096ms 23.8202μs 41.9811 KOps/s 42.0393 KOps/s $\color{#d91a1a}-0.14\%$
test_step_mdp_speed[True-False-True-False-False] 0.4081ms 14.5621μs 68.6716 KOps/s 68.5969 KOps/s $\color{#35bf28}+0.11\%$
test_step_mdp_speed[True-False-False-True-True] 96.4920μs 44.8695μs 22.2869 KOps/s 22.1911 KOps/s $\color{#35bf28}+0.43\%$
test_step_mdp_speed[True-False-False-True-False] 0.4165ms 27.7915μs 35.9823 KOps/s 35.4201 KOps/s $\color{#35bf28}+1.59\%$
test_step_mdp_speed[True-False-False-False-True] 0.4103ms 25.6764μs 38.9462 KOps/s 39.4869 KOps/s $\color{#d91a1a}-1.37\%$
test_step_mdp_speed[True-False-False-False-False] 0.4096ms 16.6867μs 59.9279 KOps/s 60.4778 KOps/s $\color{#d91a1a}-0.91\%$
test_step_mdp_speed[False-True-True-True-True] 0.4252ms 42.5235μs 23.5164 KOps/s 23.5253 KOps/s $\color{#d91a1a}-0.04\%$
test_step_mdp_speed[False-True-True-True-False] 54.2310μs 25.7289μs 38.8669 KOps/s 38.3036 KOps/s $\color{#35bf28}+1.47\%$
test_step_mdp_speed[False-True-True-False-True] 0.4179ms 26.9668μs 37.0827 KOps/s 37.5560 KOps/s $\color{#d91a1a}-1.26\%$
test_step_mdp_speed[False-True-True-False-False] 0.3993ms 16.0117μs 62.4542 KOps/s 61.4440 KOps/s $\color{#35bf28}+1.64\%$
test_step_mdp_speed[False-True-False-True-True] 0.4310ms 44.6628μs 22.3900 KOps/s 22.7017 KOps/s $\color{#d91a1a}-1.37\%$
test_step_mdp_speed[False-True-False-True-False] 59.8810μs 27.9842μs 35.7344 KOps/s 37.1068 KOps/s $\color{#d91a1a}-3.70\%$
test_step_mdp_speed[False-True-False-False-True] 92.0480ms 33.0399μs 30.2665 KOps/s 34.6312 KOps/s $\textbf{\color{#d91a1a}-12.60\%}$
test_step_mdp_speed[False-True-False-False-False] 0.4086ms 18.1715μs 55.0313 KOps/s 55.0394 KOps/s $\color{#d91a1a}-0.01\%$
test_step_mdp_speed[False-False-True-True-True] 0.4395ms 46.3917μs 21.5556 KOps/s 21.6173 KOps/s $\color{#d91a1a}-0.29\%$
test_step_mdp_speed[False-False-True-True-False] 0.4189ms 30.0654μs 33.2608 KOps/s 33.5199 KOps/s $\color{#d91a1a}-0.77\%$
test_step_mdp_speed[False-False-True-False-True] 59.0210μs 29.0721μs 34.3972 KOps/s 34.7451 KOps/s $\color{#d91a1a}-1.00\%$
test_step_mdp_speed[False-False-True-False-False] 0.4044ms 18.0463μs 55.4129 KOps/s 54.8533 KOps/s $\color{#35bf28}+1.02\%$
test_step_mdp_speed[False-False-False-True-True] 0.4324ms 48.2246μs 20.7363 KOps/s 20.8915 KOps/s $\color{#d91a1a}-0.74\%$
test_step_mdp_speed[False-False-False-True-False] 0.4173ms 32.2660μs 30.9924 KOps/s 30.9881 KOps/s $\color{#35bf28}+0.01\%$
test_step_mdp_speed[False-False-False-False-True] 60.7210μs 29.9577μs 33.3804 KOps/s 32.4161 KOps/s $\color{#35bf28}+2.97\%$
test_step_mdp_speed[False-False-False-False-False] 0.3975ms 19.8871μs 50.2837 KOps/s 49.3043 KOps/s $\color{#35bf28}+1.99\%$
test_values[generalized_advantage_estimate-True-True] 25.0860ms 24.5065ms 40.8055 Ops/s 40.5472 Ops/s $\color{#35bf28}+0.64\%$
test_values[vec_generalized_advantage_estimate-True-True] 0.1055s 2.9964ms 333.7380 Ops/s 337.0118 Ops/s $\color{#d91a1a}-0.97\%$
test_values[td0_return_estimate-False-False] 93.0020μs 66.5816μs 15.0192 KOps/s 14.9511 KOps/s $\color{#35bf28}+0.46\%$
test_values[td1_return_estimate-False-False] 56.1395ms 54.9978ms 18.1826 Ops/s 17.9698 Ops/s $\color{#35bf28}+1.18\%$
test_values[vec_td1_return_estimate-False-False] 1.2742ms 1.0693ms 935.2176 Ops/s 928.8472 Ops/s $\color{#35bf28}+0.69\%$
test_values[td_lambda_return_estimate-True-False] 87.6776ms 87.0652ms 11.4856 Ops/s 11.3205 Ops/s $\color{#35bf28}+1.46\%$
test_values[vec_td_lambda_return_estimate-True-False] 1.3169ms 1.0661ms 937.9756 Ops/s 932.8489 Ops/s $\color{#35bf28}+0.55\%$
test_gae_speed[generalized_advantage_estimate-False-1-512] 24.8040ms 24.4971ms 40.8211 Ops/s 40.7225 Ops/s $\color{#35bf28}+0.24\%$
test_gae_speed[vec_generalized_advantage_estimate-True-1-512] 0.9263ms 0.7179ms 1.3929 KOps/s 1.3829 KOps/s $\color{#35bf28}+0.72\%$
test_gae_speed[vec_generalized_advantage_estimate-False-1-512] 0.7612ms 0.6565ms 1.5233 KOps/s 1.5092 KOps/s $\color{#35bf28}+0.93\%$
test_gae_speed[vec_generalized_advantage_estimate-True-32-512] 1.8111ms 1.4608ms 684.5421 Ops/s 682.6364 Ops/s $\color{#35bf28}+0.28\%$
test_gae_speed[vec_generalized_advantage_estimate-False-32-512] 0.7515ms 0.6721ms 1.4878 KOps/s 1.4732 KOps/s $\color{#35bf28}+0.99\%$
test_dqn_speed 3.0020ms 1.3109ms 762.8379 Ops/s 767.2371 Ops/s $\color{#d91a1a}-0.57\%$
test_ddpg_speed 2.9036ms 2.6701ms 374.5202 Ops/s 375.6145 Ops/s $\color{#d91a1a}-0.29\%$
test_sac_speed 8.1121ms 7.6879ms 130.0744 Ops/s 130.0626 Ops/s $+0.01\%$
test_redq_speed 14.6148ms 10.0752ms 99.2536 Ops/s 100.3964 Ops/s $\color{#d91a1a}-1.14\%$
test_redq_deprec_speed 11.1087ms 10.6483ms 93.9114 Ops/s 94.2158 Ops/s $\color{#d91a1a}-0.32\%$
test_td3_speed 8.0027ms 7.7799ms 128.5366 Ops/s 128.5526 Ops/s $\color{#d91a1a}-0.01\%$
test_cql_speed 27.6618ms 24.8561ms 40.2316 Ops/s 39.7865 Ops/s $\color{#35bf28}+1.12\%$
test_a2c_speed 5.7452ms 5.3057ms 188.4756 Ops/s 183.3457 Ops/s $\color{#35bf28}+2.80\%$
test_ppo_speed 6.0290ms 5.6430ms 177.2104 Ops/s 173.1433 Ops/s $\color{#35bf28}+2.35\%$
test_reinforce_speed 4.5684ms 4.3943ms 227.5655 Ops/s 222.0176 Ops/s $\color{#35bf28}+2.50\%$
test_iql_speed 19.6344ms 18.8384ms 53.0831 Ops/s 52.6398 Ops/s $\color{#35bf28}+0.84\%$
test_rb_sample[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 6.8514ms 6.7128ms 148.9695 Ops/s 148.0930 Ops/s $\color{#35bf28}+0.59\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 0.4737ms 0.2345ms 4.2638 KOps/s 4.2858 KOps/s $\color{#d91a1a}-0.51\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.3863ms 0.2123ms 4.7106 KOps/s 4.7673 KOps/s $\color{#d91a1a}-1.19\%$
test_rb_sample[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 6.9987ms 6.5883ms 151.7851 Ops/s 150.1137 Ops/s $\color{#35bf28}+1.11\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 1.2964ms 0.2328ms 4.2953 KOps/s 4.3546 KOps/s $\color{#d91a1a}-1.36\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.4001ms 0.2084ms 4.7975 KOps/s 4.8207 KOps/s $\color{#d91a1a}-0.48\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-sampler6-10000] 1.3801ms 1.2268ms 815.1457 Ops/s 797.2343 Ops/s $\color{#35bf28}+2.25\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-sampler7-10000] 1.3011ms 1.1364ms 879.9929 Ops/s 873.1508 Ops/s $\color{#35bf28}+0.78\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 7.1535ms 6.8206ms 146.6152 Ops/s 145.7527 Ops/s $\color{#35bf28}+0.59\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 0.7941ms 0.3757ms 2.6620 KOps/s 2.3771 KOps/s $\textbf{\color{#35bf28}+11.98\%}$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.5180ms 0.3544ms 2.8213 KOps/s 2.2375 KOps/s $\textbf{\color{#35bf28}+26.09\%}$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 6.8659ms 6.6783ms 149.7385 Ops/s 147.5567 Ops/s $\color{#35bf28}+1.48\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 1.1615ms 0.2390ms 4.1836 KOps/s 4.2788 KOps/s $\color{#d91a1a}-2.23\%$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.3527ms 0.2128ms 4.6998 KOps/s 4.7283 KOps/s $\color{#d91a1a}-0.60\%$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 7.1237ms 6.6047ms 151.4083 Ops/s 150.8360 Ops/s $\color{#35bf28}+0.38\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 1.2106ms 0.2329ms 4.2944 KOps/s 2.7872 KOps/s $\textbf{\color{#35bf28}+54.08\%}$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.6334ms 0.2110ms 4.7391 KOps/s 4.5749 KOps/s $\color{#35bf28}+3.59\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 7.2664ms 6.8577ms 145.8214 Ops/s 145.5172 Ops/s $\color{#35bf28}+0.21\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 1.3624ms 0.3801ms 2.6306 KOps/s 2.6270 KOps/s $\color{#35bf28}+0.14\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.5056ms 0.3549ms 2.8174 KOps/s 2.8104 KOps/s $\color{#35bf28}+0.25\%$
test_rb_populate[TensorDictReplayBuffer-ListStorage-RandomSampler-400] 0.1388s 7.8570ms 127.2744 Ops/s 120.1290 Ops/s $\textbf{\color{#35bf28}+5.95\%}$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-400] 21.4249ms 15.8826ms 62.9621 Ops/s 62.1912 Ops/s $\color{#35bf28}+1.24\%$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-400] 2.1725ms 1.0259ms 974.7574 Ops/s 999.6565 Ops/s $\color{#d91a1a}-2.49\%$
test_rb_populate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-400] 0.1292s 7.6958ms 129.9411 Ops/s 128.6503 Ops/s $\color{#35bf28}+1.00\%$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-400] 0.1387s 18.4450ms 54.2153 Ops/s 53.4986 Ops/s $\color{#35bf28}+1.34\%$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-400] 1.9941ms 0.9791ms 1.0214 KOps/s 928.0930 Ops/s $\textbf{\color{#35bf28}+10.05\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-ListStorage-None-400] 0.1289s 7.8726ms 127.0235 Ops/s 127.0846 Ops/s $\color{#d91a1a}-0.05\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-400] 21.3337ms 16.1986ms 61.7336 Ops/s 61.1999 Ops/s $\color{#35bf28}+0.87\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-400] 2.2776ms 1.1798ms 847.6195 Ops/s 842.4689 Ops/s $\color{#35bf28}+0.61\%$

@vmoens vmoens merged commit e4c6c20 into gh/vmoens/26/base Sep 17, 2024
63 of 74 checks passed
vmoens added a commit that referenced this pull request Sep 17, 2024
…eady exist

ghstack-source-id: 63bcb1e0420620d5dcd2b73d8e0a5b3bf137c8e1
Pull Request resolved: #2438
@vmoens vmoens deleted the gh/vmoens/26/head branch September 17, 2024 00:33
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. enhancement New feature or request
Projects
None yet
Development

Successfully merging this pull request may close these issues.

[BUG] Memory-Mapped Storage fails to zero out missing tensors when reusing scratch dir path
2 participants