Skip to content

[CI] Add flaky test tracking system#3408

Merged
vmoens merged 1 commit intomainfrom
flaky-ci
Jan 29, 2026
Merged

[CI] Add flaky test tracking system#3408
vmoens merged 1 commit intomainfrom
flaky-ci

Conversation

@vmoens
Copy link
Collaborator

@vmoens vmoens commented Jan 29, 2026

Summary

Adds a GitHub-native flaky CI test tracking system to identify, track, and visualize flaky tests over time on the main branch.

Features

  • Structured test output: Configures pytest to emit JSON results via pytest-json-report
  • Artifact collection: Uploads test result artifacts with CI metadata on every CI run
  • Daily aggregation: Scheduled workflow that analyzes recent CI runs to identify flaky tests
  • Flaky detection: Identifies tests with failure rates between 5-95% with configurable thresholds
  • Reports: Generates JSON and Markdown reports with flaky test details
  • Dashboard: Static HTML dashboard deployed to GitHub Pages with Chart.js visualization
  • Badge: Dynamic shields.io badge showing flaky test count
  • Notifications: Creates/updates a GitHub Issue with daily flaky test summaries

Workflows Analyzed

  • test-linux.yml - Main unit tests
  • test-linux-libs.yml - Library integrations (brax, gym, envpool, etc.)
  • test-linux-habitat.yml - Habitat environment tests
  • test-linux-llm.yml - LLM/RLHF tests
  • test-linux-sota.yml - SOTA implementation tests

Files Added

File Description
.github/workflows/flaky-test-tracker.yml Daily scheduled workflow for aggregation
.github/scripts/analyze_flaky_tests.py Main analysis script with GitHub API integration
.github/scripts/generate_dashboard.py Dashboard HTML generator
.github/unittest/helpers/upload_test_results.py Test result processor with metadata
.github/unittest/helpers/pytest_json_report.sh Reusable helper for JSON reporting

Configuration

Flaky test detection thresholds (configurable in analyze_flaky_tests.py):

  • Minimum failure rate: 5%
  • Maximum failure rate: 95%
  • Minimum failures: 2
  • Minimum executions: 3

Outputs

  • Badge: https://pytorch.github.io/rl/flaky/badge.json
  • Dashboard: https://pytorch.github.io/rl/flaky/
  • JSON Report: https://pytorch.github.io/rl/flaky/flaky-tests.json
  • Markdown Report: https://pytorch.github.io/rl/flaky/flaky-tests.md

Test plan

  • Verify workflow runs successfully on schedule or manual trigger
  • Verify test result JSON files are generated during CI runs
  • Verify dashboard is deployed to GitHub Pages
  • Verify badge displays correctly in README

Add a GitHub-native flaky CI test tracking system that:

- Configures pytest to emit JSON results via pytest-json-report
- Uploads test result artifacts with CI metadata on every run
- Runs a daily aggregation workflow that analyzes recent CI runs
- Identifies flaky tests based on failure rate heuristics
- Generates JSON/Markdown reports and a static dashboard
- Deploys to GitHub Pages with dynamic shields.io badge
- Creates/updates a GitHub Issue for notifications

Workflows analyzed:
- test-linux.yml (main unit tests)
- test-linux-libs.yml (library integrations)
- test-linux-habitat.yml (Habitat environment)
- test-linux-llm.yml (LLM/RLHF tests)
- test-linux-sota.yml (SOTA implementations)

Files added:
- .github/workflows/flaky-test-tracker.yml
- .github/scripts/analyze_flaky_tests.py
- .github/scripts/generate_dashboard.py
- .github/unittest/helpers/upload_test_results.py
- .github/unittest/helpers/pytest_json_report.sh
@pytorch-bot
Copy link

pytorch-bot bot commented Jan 29, 2026

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/rl/3408

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

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 29, 2026
@github-actions github-actions bot added the CI Has to do with CI setup (e.g. wheels & builds, tests...) label Jan 29, 2026
@vmoens vmoens merged commit e09ee4f into main Jan 29, 2026
94 of 113 checks passed
@github-actions
Copy link
Contributor

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

Total Benchmarks: 148. Improved: $\large\color{#35bf28}11$. Worsened: $\large\color{#d91a1a}11$.

Expand to view detailed results
Name Max Mean Ops Ops on Repo HEAD Change
test_tensor_to_bytestream_speed[pickle] 81.5882μs 80.5716μs 12.4113 KOps/s 12.4949 KOps/s $\color{#d91a1a}-0.67\%$
test_tensor_to_bytestream_speed[torch.save] 0.1413ms 0.1402ms 7.1336 KOps/s 7.2063 KOps/s $\color{#d91a1a}-1.01\%$
test_tensor_to_bytestream_speed[untyped_storage] 0.1297s 0.1291s 7.7486 Ops/s 8.2322 Ops/s $\textbf{\color{#d91a1a}-5.87\%}$
test_tensor_to_bytestream_speed[numpy] 2.6056μs 2.5862μs 386.6682 KOps/s 399.2190 KOps/s $\color{#d91a1a}-3.14\%$
test_tensor_to_bytestream_speed[safetensors] 38.9249μs 37.5879μs 26.6043 KOps/s 27.6907 KOps/s $\color{#d91a1a}-3.92\%$
test_simple 0.9139s 0.8230s 1.2151 Ops/s 1.2216 Ops/s $\color{#d91a1a}-0.53\%$
test_transformed 1.5452s 1.4528s 0.6883 Ops/s 0.6871 Ops/s $\color{#35bf28}+0.18\%$
test_serial 2.3227s 2.3194s 0.4311 Ops/s 0.4350 Ops/s $\color{#d91a1a}-0.88\%$
test_parallel 2.0667s 1.9974s 0.5006 Ops/s 0.5209 Ops/s $\color{#d91a1a}-3.88\%$
test_step_mdp_speed[True-True-True-True-True] 0.3481ms 44.2892μs 22.5789 KOps/s 22.3057 KOps/s $\color{#35bf28}+1.22\%$
test_step_mdp_speed[True-True-True-True-False] 56.2000μs 24.8907μs 40.1756 KOps/s 39.6036 KOps/s $\color{#35bf28}+1.44\%$
test_step_mdp_speed[True-True-True-False-True] 57.8810μs 24.6436μs 40.5785 KOps/s 39.5310 KOps/s $\color{#35bf28}+2.65\%$
test_step_mdp_speed[True-True-True-False-False] 42.3210μs 13.8435μs 72.2362 KOps/s 71.7815 KOps/s $\color{#35bf28}+0.63\%$
test_step_mdp_speed[True-True-False-True-True] 0.1042ms 48.3805μs 20.6695 KOps/s 20.6361 KOps/s $\color{#35bf28}+0.16\%$
test_step_mdp_speed[True-True-False-True-False] 79.7210μs 27.2535μs 36.6925 KOps/s 35.6593 KOps/s $\color{#35bf28}+2.90\%$
test_step_mdp_speed[True-True-False-False-True] 64.6610μs 27.6157μs 36.2113 KOps/s 35.7566 KOps/s $\color{#35bf28}+1.27\%$
test_step_mdp_speed[True-True-False-False-False] 59.1810μs 16.7262μs 59.7864 KOps/s 59.6045 KOps/s $\color{#35bf28}+0.31\%$
test_step_mdp_speed[True-False-True-True-True] 0.1169ms 50.9915μs 19.6111 KOps/s 20.1101 KOps/s $\color{#d91a1a}-2.48\%$
test_step_mdp_speed[True-False-True-True-False] 71.5710μs 30.6354μs 32.6420 KOps/s 33.2906 KOps/s $\color{#d91a1a}-1.95\%$
test_step_mdp_speed[True-False-True-False-True] 61.0210μs 27.5415μs 36.3088 KOps/s 36.0257 KOps/s $\color{#35bf28}+0.79\%$
test_step_mdp_speed[True-False-True-False-False] 54.6800μs 16.7465μs 59.7141 KOps/s 60.5012 KOps/s $\color{#d91a1a}-1.30\%$
test_step_mdp_speed[True-False-False-True-True] 94.1610μs 52.6170μs 19.0053 KOps/s 18.7881 KOps/s $\color{#35bf28}+1.16\%$
test_step_mdp_speed[True-False-False-True-False] 72.3910μs 32.6573μs 30.6210 KOps/s 30.1912 KOps/s $\color{#35bf28}+1.42\%$
test_step_mdp_speed[True-False-False-False-True] 0.1052ms 28.7905μs 34.7337 KOps/s 32.5590 KOps/s $\textbf{\color{#35bf28}+6.68\%}$
test_step_mdp_speed[True-False-False-False-False] 52.9510μs 18.9888μs 52.6627 KOps/s 51.8144 KOps/s $\color{#35bf28}+1.64\%$
test_step_mdp_speed[False-True-True-True-True] 0.1160ms 48.4762μs 20.6287 KOps/s 19.9047 KOps/s $\color{#35bf28}+3.64\%$
test_step_mdp_speed[False-True-True-True-False] 64.7910μs 29.8648μs 33.4842 KOps/s 32.6290 KOps/s $\color{#35bf28}+2.62\%$
test_step_mdp_speed[False-True-True-False-True] 68.7420μs 30.8973μs 32.3652 KOps/s 31.1329 KOps/s $\color{#35bf28}+3.96\%$
test_step_mdp_speed[False-True-True-False-False] 61.9810μs 18.0469μs 55.4112 KOps/s 55.1607 KOps/s $\color{#35bf28}+0.45\%$
test_step_mdp_speed[False-True-False-True-True] 2.6288ms 52.9491μs 18.8860 KOps/s 18.6465 KOps/s $\color{#35bf28}+1.28\%$
test_step_mdp_speed[False-True-False-True-False] 0.1016ms 32.7466μs 30.5376 KOps/s 29.9816 KOps/s $\color{#35bf28}+1.85\%$
test_step_mdp_speed[False-True-False-False-True] 98.9520μs 33.8301μs 29.5595 KOps/s 29.9057 KOps/s $\color{#d91a1a}-1.16\%$
test_step_mdp_speed[False-True-False-False-False] 50.8700μs 20.6129μs 48.5133 KOps/s 48.3277 KOps/s $\color{#35bf28}+0.38\%$
test_step_mdp_speed[False-False-True-True-True] 98.8810μs 56.0460μs 17.8425 KOps/s 17.8023 KOps/s $\color{#35bf28}+0.23\%$
test_step_mdp_speed[False-False-True-True-False] 82.7710μs 35.8989μs 27.8560 KOps/s 27.6854 KOps/s $\color{#35bf28}+0.62\%$
test_step_mdp_speed[False-False-True-False-True] 83.5910μs 33.9623μs 29.4444 KOps/s 29.3260 KOps/s $\color{#35bf28}+0.40\%$
test_step_mdp_speed[False-False-True-False-False] 60.7310μs 20.4744μs 48.8416 KOps/s 49.1300 KOps/s $\color{#d91a1a}-0.59\%$
test_step_mdp_speed[False-False-False-True-True] 0.1008ms 57.7167μs 17.3260 KOps/s 17.2627 KOps/s $\color{#35bf28}+0.37\%$
test_step_mdp_speed[False-False-False-True-False] 97.1820μs 38.5775μs 25.9218 KOps/s 26.2818 KOps/s $\color{#d91a1a}-1.37\%$
test_step_mdp_speed[False-False-False-False-True] 76.4110μs 35.8306μs 27.9091 KOps/s 27.9627 KOps/s $\color{#d91a1a}-0.19\%$
test_step_mdp_speed[False-False-False-False-False] 90.7910μs 23.1702μs 43.1589 KOps/s 43.4623 KOps/s $\color{#d91a1a}-0.70\%$
test_non_tensor_env_rollout_speed[1000-single-True] 0.7492s 0.7470s 1.3386 Ops/s 1.2884 Ops/s $\color{#35bf28}+3.89\%$
test_non_tensor_env_rollout_speed[1000-single-False] 0.7254s 0.6304s 1.5864 Ops/s 1.5664 Ops/s $\color{#35bf28}+1.27\%$
test_non_tensor_env_rollout_speed[1000-serial-no-buffers-True] 1.7490s 1.6736s 0.5975 Ops/s 0.5957 Ops/s $\color{#35bf28}+0.31\%$
test_non_tensor_env_rollout_speed[1000-serial-no-buffers-False] 1.5306s 1.4525s 0.6885 Ops/s 0.6843 Ops/s $\color{#35bf28}+0.61\%$
test_non_tensor_env_rollout_speed[1000-serial-buffers-True] 2.0027s 1.9273s 0.5188 Ops/s 0.5175 Ops/s $\color{#35bf28}+0.26\%$
test_non_tensor_env_rollout_speed[1000-serial-buffers-False] 1.7772s 1.7014s 0.5878 Ops/s 0.5848 Ops/s $\color{#35bf28}+0.51\%$
test_non_tensor_env_rollout_speed[1000-parallel-no-buffers-True] 4.7768s 4.6249s 0.2162 Ops/s 0.2163 Ops/s $\color{#d91a1a}-0.05\%$
test_non_tensor_env_rollout_speed[1000-parallel-no-buffers-False] 4.6616s 4.4927s 0.2226 Ops/s 0.2219 Ops/s $\color{#35bf28}+0.31\%$
test_non_tensor_env_rollout_speed[1000-parallel-buffers-True] 2.1035s 1.9843s 0.5040 Ops/s 0.5156 Ops/s $\color{#d91a1a}-2.26\%$
test_non_tensor_env_rollout_speed[1000-parallel-buffers-False] 1.7370s 1.6620s 0.6017 Ops/s 0.6022 Ops/s $\color{#d91a1a}-0.09\%$
test_values[generalized_advantage_estimate-True-True] 21.0348ms 20.6558ms 48.4125 Ops/s 49.5002 Ops/s $\color{#d91a1a}-2.20\%$
test_values[vec_generalized_advantage_estimate-True-True] 0.1443s 3.8178ms 261.9330 Ops/s 252.5185 Ops/s $\color{#35bf28}+3.73\%$
test_values[td0_return_estimate-False-False] 0.1215ms 85.5750μs 11.6857 KOps/s 11.8723 KOps/s $\color{#d91a1a}-1.57\%$
test_values[td1_return_estimate-False-False] 49.5309ms 49.1447ms 20.3481 Ops/s 20.6997 Ops/s $\color{#d91a1a}-1.70\%$
test_values[vec_td1_return_estimate-False-False] 1.3220ms 1.0989ms 909.9633 Ops/s 917.2290 Ops/s $\color{#d91a1a}-0.79\%$
test_values[td_lambda_return_estimate-True-False] 81.3602ms 81.0496ms 12.3381 Ops/s 12.6110 Ops/s $\color{#d91a1a}-2.16\%$
test_values[vec_td_lambda_return_estimate-True-False] 1.3063ms 1.0940ms 914.0672 Ops/s 919.4696 Ops/s $\color{#d91a1a}-0.59\%$
test_gae_speed[generalized_advantage_estimate-False-1-512] 21.0919ms 20.8110ms 48.0515 Ops/s 49.1298 Ops/s $\color{#d91a1a}-2.19\%$
test_gae_speed[vec_generalized_advantage_estimate-True-1-512] 1.0641ms 0.7695ms 1.2995 KOps/s 1.3080 KOps/s $\color{#d91a1a}-0.65\%$
test_gae_speed[vec_generalized_advantage_estimate-False-1-512] 0.7324ms 0.6854ms 1.4589 KOps/s 1.4730 KOps/s $\color{#d91a1a}-0.95\%$
test_gae_speed[vec_generalized_advantage_estimate-True-32-512] 1.5426ms 1.5015ms 666.0188 Ops/s 668.9286 Ops/s $\color{#d91a1a}-0.44\%$
test_gae_speed[vec_generalized_advantage_estimate-False-32-512] 0.7505ms 0.7015ms 1.4255 KOps/s 1.4387 KOps/s $\color{#d91a1a}-0.91\%$
test_dqn_speed[False-None] 1.6335ms 1.5445ms 647.4425 Ops/s 646.7712 Ops/s $\color{#35bf28}+0.10\%$
test_dqn_speed[False-backward] 2.4558ms 2.1971ms 455.1446 Ops/s 455.3982 Ops/s $\color{#d91a1a}-0.06\%$
test_dqn_speed[True-None] 0.6154ms 0.5387ms 1.8562 KOps/s 1.8379 KOps/s $\color{#35bf28}+0.99\%$
test_dqn_speed[True-backward] 1.2083ms 1.1809ms 846.8128 Ops/s 934.1345 Ops/s $\textbf{\color{#d91a1a}-9.35\%}$
test_dqn_speed[reduce-overhead-None] 0.6415ms 0.5598ms 1.7865 KOps/s 1.6767 KOps/s $\textbf{\color{#35bf28}+6.55\%}$
test_ddpg_speed[False-None] 3.3172ms 2.9229ms 342.1239 Ops/s 344.0852 Ops/s $\color{#d91a1a}-0.57\%$
test_ddpg_speed[False-backward] 4.7292ms 4.3438ms 230.2130 Ops/s 238.6651 Ops/s $\color{#d91a1a}-3.54\%$
test_ddpg_speed[True-None] 1.3652ms 1.2847ms 778.3747 Ops/s 765.8593 Ops/s $\color{#35bf28}+1.63\%$
test_ddpg_speed[True-backward] 2.5115ms 2.4628ms 406.0431 Ops/s 425.9239 Ops/s $\color{#d91a1a}-4.67\%$
test_ddpg_speed[reduce-overhead-None] 1.4072ms 1.3186ms 758.3892 Ops/s 759.4107 Ops/s $\color{#d91a1a}-0.13\%$
test_sac_speed[False-None] 8.8537ms 8.4096ms 118.9124 Ops/s 118.4554 Ops/s $\color{#35bf28}+0.39\%$
test_sac_speed[False-backward] 12.5083ms 11.7849ms 84.8544 Ops/s 87.0676 Ops/s $\color{#d91a1a}-2.54\%$
test_sac_speed[True-None] 1.8728ms 1.7873ms 559.5091 Ops/s 560.5391 Ops/s $\color{#d91a1a}-0.18\%$
test_sac_speed[True-backward] 3.6048ms 3.5370ms 282.7228 Ops/s 293.3166 Ops/s $\color{#d91a1a}-3.61\%$
test_sac_speed[reduce-overhead-None] 18.6295ms 10.6070ms 94.2772 Ops/s 93.3845 Ops/s $\color{#35bf28}+0.96\%$
test_redq_deprec_speed[False-None] 10.2705ms 9.4064ms 106.3103 Ops/s 105.5847 Ops/s $\color{#35bf28}+0.69\%$
test_redq_deprec_speed[False-backward] 13.4658ms 12.8270ms 77.9604 Ops/s 79.3665 Ops/s $\color{#d91a1a}-1.77\%$
test_redq_deprec_speed[True-None] 2.5857ms 2.5114ms 398.1823 Ops/s 395.7674 Ops/s $\color{#35bf28}+0.61\%$
test_redq_deprec_speed[True-backward] 4.5028ms 4.2698ms 234.2049 Ops/s 243.7804 Ops/s $\color{#d91a1a}-3.93\%$
test_redq_deprec_speed[reduce-overhead-None] 15.5413ms 9.6226ms 103.9217 Ops/s 105.7092 Ops/s $\color{#d91a1a}-1.69\%$
test_td3_speed[False-None] 8.6911ms 8.3282ms 120.0743 Ops/s 121.1898 Ops/s $\color{#d91a1a}-0.92\%$
test_td3_speed[False-backward] 11.6395ms 10.9742ms 91.1230 Ops/s 93.9312 Ops/s $\color{#d91a1a}-2.99\%$
test_td3_speed[True-None] 1.6286ms 1.5996ms 625.1603 Ops/s 627.6826 Ops/s $\color{#d91a1a}-0.40\%$
test_td3_speed[True-backward] 3.2510ms 3.1987ms 312.6236 Ops/s 326.0926 Ops/s $\color{#d91a1a}-4.13\%$
test_td3_speed[reduce-overhead-None] 55.5023ms 23.3103ms 42.8995 Ops/s 43.3858 Ops/s $\color{#d91a1a}-1.12\%$
test_cql_speed[False-None] 19.6478ms 17.4943ms 57.1614 Ops/s 57.6227 Ops/s $\color{#d91a1a}-0.80\%$
test_cql_speed[False-backward] 23.8111ms 23.2029ms 43.0981 Ops/s 44.1400 Ops/s $\color{#d91a1a}-2.36\%$
test_cql_speed[True-None] 3.5392ms 3.2704ms 305.7686 Ops/s 311.7069 Ops/s $\color{#d91a1a}-1.91\%$
test_cql_speed[True-backward] 5.7492ms 5.3389ms 187.3059 Ops/s 184.1514 Ops/s $\color{#35bf28}+1.71\%$
test_cql_speed[reduce-overhead-None] 18.3749ms 11.7438ms 85.1514 Ops/s 86.0292 Ops/s $\color{#d91a1a}-1.02\%$
test_a2c_speed[False-None] 4.3922ms 3.2924ms 303.7320 Ops/s 305.6855 Ops/s $\color{#d91a1a}-0.64\%$
test_a2c_speed[False-backward] 6.7401ms 6.2647ms 159.6240 Ops/s 154.4820 Ops/s $\color{#35bf28}+3.33\%$
test_a2c_speed[True-None] 1.4342ms 1.3230ms 755.8515 Ops/s 751.9632 Ops/s $\color{#35bf28}+0.52\%$
test_a2c_speed[True-backward] 3.0021ms 2.9131ms 343.2739 Ops/s 328.7429 Ops/s $\color{#35bf28}+4.42\%$
test_a2c_speed[reduce-overhead-None] 1.1556ms 0.9655ms 1.0357 KOps/s 1.0193 KOps/s $\color{#35bf28}+1.61\%$
test_ppo_speed[False-None] 4.0952ms 3.9189ms 255.1728 Ops/s 255.2195 Ops/s $\color{#d91a1a}-0.02\%$
test_ppo_speed[False-backward] 7.5360ms 7.1174ms 140.4998 Ops/s 136.9798 Ops/s $\color{#35bf28}+2.57\%$
test_ppo_speed[True-None] 1.5108ms 1.3947ms 717.0212 Ops/s 708.0938 Ops/s $\color{#35bf28}+1.26\%$
test_ppo_speed[True-backward] 3.4577ms 3.2258ms 310.0008 Ops/s 311.1484 Ops/s $\color{#d91a1a}-0.37\%$
test_ppo_speed[reduce-overhead-None] 1.0690ms 1.0190ms 981.3572 Ops/s 939.1581 Ops/s $\color{#35bf28}+4.49\%$
test_reinforce_speed[False-None] 2.4398ms 2.3062ms 433.6064 Ops/s 433.5622 Ops/s $\color{#35bf28}+0.01\%$
test_reinforce_speed[False-backward] 3.8857ms 3.4632ms 288.7510 Ops/s 292.5300 Ops/s $\color{#d91a1a}-1.29\%$
test_reinforce_speed[True-None] 1.3786ms 1.2357ms 809.2624 Ops/s 802.6682 Ops/s $\color{#35bf28}+0.82\%$
test_reinforce_speed[True-backward] 3.2091ms 3.0087ms 332.3725 Ops/s 330.3640 Ops/s $\color{#35bf28}+0.61\%$
test_reinforce_speed[reduce-overhead-None] 0.4605s 10.0793ms 99.2131 Ops/s 97.3842 Ops/s $\color{#35bf28}+1.88\%$
test_iql_speed[False-None] 10.3143ms 9.5632ms 104.5671 Ops/s 104.3476 Ops/s $\color{#35bf28}+0.21\%$
test_iql_speed[False-backward] 13.9695ms 13.6510ms 73.2549 Ops/s 74.5890 Ops/s $\color{#d91a1a}-1.79\%$
test_iql_speed[True-None] 2.2335ms 2.1317ms 469.1003 Ops/s 455.4741 Ops/s $\color{#35bf28}+2.99\%$
test_iql_speed[True-backward] 5.0304ms 4.7756ms 209.3985 Ops/s 206.1856 Ops/s $\color{#35bf28}+1.56\%$
test_iql_speed[reduce-overhead-None] 17.1663ms 10.1141ms 98.8716 Ops/s 75.8651 Ops/s $\textbf{\color{#35bf28}+30.33\%}$
test_rb_sample[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 6.1018ms 5.9631ms 167.6993 Ops/s 166.7937 Ops/s $\color{#35bf28}+0.54\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 1.0389ms 0.2863ms 3.4924 KOps/s 2.6329 KOps/s $\textbf{\color{#35bf28}+32.64\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.4942ms 0.2676ms 3.7367 KOps/s 2.7674 KOps/s $\textbf{\color{#35bf28}+35.03\%}$
test_rb_sample[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 6.1087ms 5.8324ms 171.4563 Ops/s 171.6447 Ops/s $\color{#d91a1a}-0.11\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 0.8104ms 0.2971ms 3.3655 KOps/s 3.2708 KOps/s $\color{#35bf28}+2.89\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.5452ms 0.3209ms 3.1166 KOps/s 3.3887 KOps/s $\textbf{\color{#d91a1a}-8.03\%}$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-sampler6-10000] 1.6114ms 1.2961ms 771.5567 Ops/s 677.7946 Ops/s $\textbf{\color{#35bf28}+13.83\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-sampler7-10000] 1.4330ms 1.1982ms 834.5827 Ops/s 734.3900 Ops/s $\textbf{\color{#35bf28}+13.64\%}$
test_rb_sample[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 6.1206ms 5.9922ms 166.8849 Ops/s 165.0214 Ops/s $\color{#35bf28}+1.13\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 1.0175ms 0.4862ms 2.0567 KOps/s 2.2582 KOps/s $\textbf{\color{#d91a1a}-8.92\%}$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.6728ms 0.4403ms 2.2712 KOps/s 2.4033 KOps/s $\textbf{\color{#d91a1a}-5.50\%}$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 5.8858ms 5.8252ms 171.6668 Ops/s 170.3478 Ops/s $\color{#35bf28}+0.77\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 2.0753ms 0.3064ms 3.2642 KOps/s 3.5346 KOps/s $\textbf{\color{#d91a1a}-7.65\%}$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.5815ms 0.3457ms 2.8927 KOps/s 3.7881 KOps/s $\textbf{\color{#d91a1a}-23.64\%}$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 6.1597ms 5.8363ms 171.3420 Ops/s 170.1916 Ops/s $\color{#35bf28}+0.68\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 0.7918ms 0.3296ms 3.0339 KOps/s 2.6724 KOps/s $\textbf{\color{#35bf28}+13.53\%}$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.6375ms 0.3735ms 2.6773 KOps/s 2.9338 KOps/s $\textbf{\color{#d91a1a}-8.74\%}$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 6.0856ms 5.9825ms 167.1530 Ops/s 165.3306 Ops/s $\color{#35bf28}+1.10\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 1.0410ms 0.5271ms 1.8972 KOps/s 2.0614 KOps/s $\textbf{\color{#d91a1a}-7.97\%}$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.7141ms 0.5182ms 1.9296 KOps/s 1.9712 KOps/s $\color{#d91a1a}-2.11\%$
test_rb_populate[TensorDictReplayBuffer-ListStorage-RandomSampler-400] 0.6077s 17.0982ms 58.4855 Ops/s 50.4679 Ops/s $\textbf{\color{#35bf28}+15.89\%}$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-400] 11.0781ms 1.9695ms 507.7401 Ops/s 495.9556 Ops/s $\color{#35bf28}+2.38\%$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-400] 7.4696ms 1.2436ms 804.1114 Ops/s 760.6846 Ops/s $\textbf{\color{#35bf28}+5.71\%}$
test_rb_populate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-400] 6.7244ms 5.0534ms 197.8848 Ops/s 193.4261 Ops/s $\color{#35bf28}+2.31\%$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-400] 11.9215ms 1.9399ms 515.4810 Ops/s 515.6271 Ops/s $\color{#d91a1a}-0.03\%$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-400] 1.0189ms 0.9027ms 1.1078 KOps/s 1.0781 KOps/s $\color{#35bf28}+2.76\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-ListStorage-None-400] 0.5524s 16.2051ms 61.7091 Ops/s 51.0400 Ops/s $\textbf{\color{#35bf28}+20.90\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-400] 12.6603ms 2.2790ms 438.7925 Ops/s 505.4726 Ops/s $\textbf{\color{#d91a1a}-13.19\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-400] 2.2941ms 1.2696ms 787.6629 Ops/s 914.1660 Ops/s $\textbf{\color{#d91a1a}-13.84\%}$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-True] 37.9769ms 36.1145ms 27.6897 Ops/s 27.3931 Ops/s $\color{#35bf28}+1.08\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-False] 20.1514ms 18.5201ms 53.9954 Ops/s 52.4007 Ops/s $\color{#35bf28}+3.04\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-True] 41.8062ms 37.0673ms 26.9780 Ops/s 26.6808 Ops/s $\color{#35bf28}+1.11\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-False] 20.2933ms 18.6503ms 53.6184 Ops/s 52.0161 Ops/s $\color{#35bf28}+3.08\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-True] 41.1255ms 38.8959ms 25.7096 Ops/s 25.6485 Ops/s $\color{#35bf28}+0.24\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-False] 21.6429ms 20.2773ms 49.3163 Ops/s 49.6543 Ops/s $\color{#d91a1a}-0.68\%$

@github-actions
Copy link
Contributor

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

Total Benchmarks: 153. Improved: $\large\color{#35bf28}8$. Worsened: $\large\color{#d91a1a}18$.

Expand to view detailed results
Name Max Mean Ops Ops on Repo HEAD Change
test_tensor_to_bytestream_speed[pickle] 83.1898μs 81.3893μs 12.2866 KOps/s 11.7650 KOps/s $\color{#35bf28}+4.43\%$
test_tensor_to_bytestream_speed[torch.save] 0.1461ms 0.1453ms 6.8830 KOps/s 7.0583 KOps/s $\color{#d91a1a}-2.48\%$
test_tensor_to_bytestream_speed[untyped_storage] 0.1378s 0.1373s 7.2834 Ops/s 9.2769 Ops/s $\textbf{\color{#d91a1a}-21.49\%}$
test_tensor_to_bytestream_speed[numpy] 2.8823μs 2.8746μs 347.8793 KOps/s 384.5755 KOps/s $\textbf{\color{#d91a1a}-9.54\%}$
test_tensor_to_bytestream_speed[safetensors] 38.9953μs 38.7923μs 25.7783 KOps/s 25.7764 KOps/s $+0.01\%$
test_simple 0.6745s 0.5817s 1.7192 Ops/s 1.7243 Ops/s $\color{#d91a1a}-0.30\%$
test_transformed 1.2753s 1.1814s 0.8465 Ops/s 0.8598 Ops/s $\color{#d91a1a}-1.55\%$
test_serial 1.8278s 1.7318s 0.5774 Ops/s 0.5946 Ops/s $\color{#d91a1a}-2.89\%$
test_parallel 1.2160s 1.1475s 0.8715 Ops/s 0.9070 Ops/s $\color{#d91a1a}-3.92\%$
test_step_mdp_speed[True-True-True-True-True] 0.3178ms 46.0666μs 21.7077 KOps/s 21.8994 KOps/s $\color{#d91a1a}-0.88\%$
test_step_mdp_speed[True-True-True-True-False] 59.8910μs 25.5657μs 39.1149 KOps/s 38.9215 KOps/s $\color{#35bf28}+0.50\%$
test_step_mdp_speed[True-True-True-False-True] 54.3710μs 25.5108μs 39.1991 KOps/s 39.1482 KOps/s $\color{#35bf28}+0.13\%$
test_step_mdp_speed[True-True-True-False-False] 47.6410μs 14.3615μs 69.6305 KOps/s 70.6652 KOps/s $\color{#d91a1a}-1.46\%$
test_step_mdp_speed[True-True-False-True-True] 89.6420μs 49.1795μs 20.3337 KOps/s 20.5663 KOps/s $\color{#d91a1a}-1.13\%$
test_step_mdp_speed[True-True-False-True-False] 56.6110μs 28.2918μs 35.3459 KOps/s 35.1131 KOps/s $\color{#35bf28}+0.66\%$
test_step_mdp_speed[True-True-False-False-True] 59.2910μs 28.8371μs 34.6776 KOps/s 34.4240 KOps/s $\color{#35bf28}+0.74\%$
test_step_mdp_speed[True-True-False-False-False] 45.5400μs 16.8927μs 59.1971 KOps/s 58.7889 KOps/s $\color{#35bf28}+0.69\%$
test_step_mdp_speed[True-False-True-True-True] 92.4020μs 51.9457μs 19.2509 KOps/s 19.2071 KOps/s $\color{#35bf28}+0.23\%$
test_step_mdp_speed[True-False-True-True-False] 65.0620μs 31.4404μs 31.8063 KOps/s 31.9379 KOps/s $\color{#d91a1a}-0.41\%$
test_step_mdp_speed[True-False-True-False-True] 62.5410μs 28.3417μs 35.2837 KOps/s 34.5968 KOps/s $\color{#35bf28}+1.99\%$
test_step_mdp_speed[True-False-True-False-False] 44.0610μs 17.0586μs 58.6216 KOps/s 58.3781 KOps/s $\color{#35bf28}+0.42\%$
test_step_mdp_speed[True-False-False-True-True] 96.4020μs 54.4556μs 18.3636 KOps/s 18.4650 KOps/s $\color{#d91a1a}-0.55\%$
test_step_mdp_speed[True-False-False-True-False] 62.7310μs 34.0283μs 29.3873 KOps/s 29.2871 KOps/s $\color{#35bf28}+0.34\%$
test_step_mdp_speed[True-False-False-False-True] 68.3220μs 30.7455μs 32.5251 KOps/s 32.2452 KOps/s $\color{#35bf28}+0.87\%$
test_step_mdp_speed[True-False-False-False-False] 51.6610μs 19.7267μs 50.6927 KOps/s 50.6241 KOps/s $\color{#35bf28}+0.14\%$
test_step_mdp_speed[False-True-True-True-True] 0.1102ms 50.9947μs 19.6099 KOps/s 19.3561 KOps/s $\color{#35bf28}+1.31\%$
test_step_mdp_speed[False-True-True-True-False] 58.4610μs 30.7999μs 32.4676 KOps/s 32.1991 KOps/s $\color{#35bf28}+0.83\%$
test_step_mdp_speed[False-True-True-False-True] 71.4310μs 32.2536μs 31.0043 KOps/s 31.2581 KOps/s $\color{#d91a1a}-0.81\%$
test_step_mdp_speed[False-True-True-False-False] 53.0010μs 18.8374μs 53.0859 KOps/s 53.3203 KOps/s $\color{#d91a1a}-0.44\%$
test_step_mdp_speed[False-True-False-True-True] 2.6616ms 55.2962μs 18.0844 KOps/s 18.3305 KOps/s $\color{#d91a1a}-1.34\%$
test_step_mdp_speed[False-True-False-True-False] 89.8020μs 33.4960μs 29.8543 KOps/s 29.3272 KOps/s $\color{#35bf28}+1.80\%$
test_step_mdp_speed[False-True-False-False-True] 67.5120μs 35.4657μs 28.1962 KOps/s 28.2927 KOps/s $\color{#d91a1a}-0.34\%$
test_step_mdp_speed[False-True-False-False-False] 63.1220μs 21.4996μs 46.5125 KOps/s 46.0529 KOps/s $\color{#35bf28}+1.00\%$
test_step_mdp_speed[False-False-True-True-True] 96.0620μs 58.2866μs 17.1566 KOps/s 17.2686 KOps/s $\color{#d91a1a}-0.65\%$
test_step_mdp_speed[False-False-True-True-False] 66.0310μs 37.1102μs 26.9468 KOps/s 26.9600 KOps/s $\color{#d91a1a}-0.05\%$
test_step_mdp_speed[False-False-True-False-True] 89.1820μs 35.8194μs 27.9178 KOps/s 28.6537 KOps/s $\color{#d91a1a}-2.57\%$
test_step_mdp_speed[False-False-True-False-False] 50.4210μs 21.6544μs 46.1801 KOps/s 46.5775 KOps/s $\color{#d91a1a}-0.85\%$
test_step_mdp_speed[False-False-False-True-True] 91.4920μs 59.7543μs 16.7352 KOps/s 16.7266 KOps/s $\color{#35bf28}+0.05\%$
test_step_mdp_speed[False-False-False-True-False] 84.9010μs 39.6587μs 25.2151 KOps/s 25.1288 KOps/s $\color{#35bf28}+0.34\%$
test_step_mdp_speed[False-False-False-False-True] 69.1620μs 38.1868μs 26.1870 KOps/s 26.5058 KOps/s $\color{#d91a1a}-1.20\%$
test_step_mdp_speed[False-False-False-False-False] 47.8910μs 24.3358μs 41.0917 KOps/s 41.3255 KOps/s $\color{#d91a1a}-0.57\%$
test_non_tensor_env_rollout_speed[1000-single-True] 0.7641s 0.7619s 1.3125 Ops/s 1.2746 Ops/s $\color{#35bf28}+2.97\%$
test_non_tensor_env_rollout_speed[1000-single-False] 0.7356s 0.6430s 1.5551 Ops/s 1.5567 Ops/s $\color{#d91a1a}-0.10\%$
test_non_tensor_env_rollout_speed[1000-serial-no-buffers-True] 1.7943s 1.7154s 0.5830 Ops/s 0.5885 Ops/s $\color{#d91a1a}-0.95\%$
test_non_tensor_env_rollout_speed[1000-serial-no-buffers-False] 1.5641s 1.4888s 0.6717 Ops/s 0.6809 Ops/s $\color{#d91a1a}-1.35\%$
test_non_tensor_env_rollout_speed[1000-serial-buffers-True] 2.0494s 1.9797s 0.5051 Ops/s 0.5154 Ops/s $\color{#d91a1a}-2.00\%$
test_non_tensor_env_rollout_speed[1000-serial-buffers-False] 1.8222s 1.7482s 0.5720 Ops/s 0.5814 Ops/s $\color{#d91a1a}-1.61\%$
test_non_tensor_env_rollout_speed[1000-parallel-no-buffers-True] 4.9667s 4.7789s 0.2093 Ops/s 0.2156 Ops/s $\color{#d91a1a}-2.92\%$
test_non_tensor_env_rollout_speed[1000-parallel-no-buffers-False] 4.5272s 4.4710s 0.2237 Ops/s 0.2235 Ops/s $\color{#35bf28}+0.05\%$
test_non_tensor_env_rollout_speed[1000-parallel-buffers-True] 2.0469s 1.9634s 0.5093 Ops/s 0.4991 Ops/s $\color{#35bf28}+2.04\%$
test_non_tensor_env_rollout_speed[1000-parallel-buffers-False] 1.8123s 1.7040s 0.5868 Ops/s 0.5920 Ops/s $\color{#d91a1a}-0.87\%$
test_values[generalized_advantage_estimate-True-True] 11.8339ms 11.4935ms 87.0054 Ops/s 93.5412 Ops/s $\textbf{\color{#d91a1a}-6.99\%}$
test_values[vec_generalized_advantage_estimate-True-True] 19.9795ms 17.8154ms 56.1313 Ops/s 56.5254 Ops/s $\color{#d91a1a}-0.70\%$
test_values[td0_return_estimate-False-False] 0.2403ms 0.1322ms 7.5641 KOps/s 7.4372 KOps/s $\color{#35bf28}+1.71\%$
test_values[td1_return_estimate-False-False] 31.6020ms 31.2223ms 32.0284 Ops/s 33.7950 Ops/s $\textbf{\color{#d91a1a}-5.23\%}$
test_values[vec_td1_return_estimate-False-False] 18.5949ms 17.9089ms 55.8381 Ops/s 56.1035 Ops/s $\color{#d91a1a}-0.47\%$
test_values[td_lambda_return_estimate-True-False] 47.5940ms 46.4095ms 21.5473 Ops/s 22.6688 Ops/s $\color{#d91a1a}-4.95\%$
test_values[vec_td_lambda_return_estimate-True-False] 19.5035ms 17.9229ms 55.7945 Ops/s 56.5089 Ops/s $\color{#d91a1a}-1.26\%$
test_gae_speed[generalized_advantage_estimate-False-1-512] 10.5194ms 10.4338ms 95.8421 Ops/s 105.8189 Ops/s $\textbf{\color{#d91a1a}-9.43\%}$
test_gae_speed[vec_generalized_advantage_estimate-True-1-512] 1.7361ms 1.5490ms 645.5886 Ops/s 678.7062 Ops/s $\color{#d91a1a}-4.88\%$
test_gae_speed[vec_generalized_advantage_estimate-False-1-512] 0.7532ms 0.4400ms 2.2730 KOps/s 2.2684 KOps/s $\color{#35bf28}+0.20\%$
test_gae_speed[vec_generalized_advantage_estimate-True-32-512] 34.7649ms 34.5707ms 28.9262 Ops/s 28.5989 Ops/s $\color{#35bf28}+1.14\%$
test_gae_speed[vec_generalized_advantage_estimate-False-32-512] 1.9092ms 1.7772ms 562.6681 Ops/s 577.6576 Ops/s $\color{#d91a1a}-2.59\%$
test_dqn_speed[False-None] 1.7874ms 1.4412ms 693.8855 Ops/s 700.2137 Ops/s $\color{#d91a1a}-0.90\%$
test_dqn_speed[False-backward] 2.3083ms 1.9708ms 507.3985 Ops/s 512.8849 Ops/s $\color{#d91a1a}-1.07\%$
test_dqn_speed[True-None] 0.7976ms 0.5494ms 1.8201 KOps/s 1.7594 KOps/s $\color{#35bf28}+3.45\%$
test_dqn_speed[True-backward] 1.0620ms 1.0196ms 980.7848 Ops/s 889.4819 Ops/s $\textbf{\color{#35bf28}+10.26\%}$
test_dqn_speed[reduce-overhead-None] 0.9011ms 0.5372ms 1.8614 KOps/s 1.7703 KOps/s $\textbf{\color{#35bf28}+5.14\%}$
test_ddpg_speed[False-None] 0.2011s 3.5545ms 281.3310 Ops/s 346.6818 Ops/s $\textbf{\color{#d91a1a}-18.85\%}$
test_ddpg_speed[False-backward] 4.3245ms 4.1785ms 239.3197 Ops/s 241.7055 Ops/s $\color{#d91a1a}-0.99\%$
test_ddpg_speed[True-None] 1.8180ms 1.4405ms 694.2192 Ops/s 701.0369 Ops/s $\color{#d91a1a}-0.97\%$
test_ddpg_speed[True-backward] 2.5194ms 2.4543ms 407.4478 Ops/s 382.3040 Ops/s $\textbf{\color{#35bf28}+6.58\%}$
test_ddpg_speed[reduce-overhead-None] 1.8031ms 1.4316ms 698.5217 Ops/s 701.1952 Ops/s $\color{#d91a1a}-0.38\%$
test_sac_speed[False-None] 9.0300ms 8.2536ms 121.1585 Ops/s 122.1525 Ops/s $\color{#d91a1a}-0.81\%$
test_sac_speed[False-backward] 11.9572ms 11.5292ms 86.7362 Ops/s 86.0901 Ops/s $\color{#35bf28}+0.75\%$
test_sac_speed[True-None] 2.5917ms 2.2035ms 453.8172 Ops/s 442.9531 Ops/s $\color{#35bf28}+2.45\%$
test_sac_speed[True-backward] 4.2083ms 4.1083ms 243.4122 Ops/s 236.0096 Ops/s $\color{#35bf28}+3.14\%$
test_sac_speed[reduce-overhead-None] 2.6260ms 2.1823ms 458.2416 Ops/s 446.2314 Ops/s $\color{#35bf28}+2.69\%$
test_redq_speed[False-None] 15.4652ms 10.8427ms 92.2280 Ops/s 94.4915 Ops/s $\color{#d91a1a}-2.40\%$
test_redq_speed[False-backward] 19.1643ms 18.2563ms 54.7756 Ops/s 56.3582 Ops/s $\color{#d91a1a}-2.81\%$
test_redq_speed[True-None] 4.8675ms 4.5182ms 221.3255 Ops/s 227.9170 Ops/s $\color{#d91a1a}-2.89\%$
test_redq_speed[True-backward] 10.1868ms 9.8685ms 101.3328 Ops/s 99.7165 Ops/s $\color{#35bf28}+1.62\%$
test_redq_speed[reduce-overhead-None] 4.9203ms 4.4844ms 222.9956 Ops/s 220.0864 Ops/s $\color{#35bf28}+1.32\%$
test_redq_deprec_speed[False-None] 11.8792ms 11.3151ms 88.3775 Ops/s 88.8342 Ops/s $\color{#d91a1a}-0.51\%$
test_redq_deprec_speed[False-backward] 17.4086ms 16.2322ms 61.6059 Ops/s 62.0047 Ops/s $\color{#d91a1a}-0.64\%$
test_redq_deprec_speed[True-None] 4.0276ms 3.7812ms 264.4662 Ops/s 266.9431 Ops/s $\color{#d91a1a}-0.93\%$
test_redq_deprec_speed[True-backward] 8.0133ms 7.7218ms 129.5030 Ops/s 127.9252 Ops/s $\color{#35bf28}+1.23\%$
test_redq_deprec_speed[reduce-overhead-None] 4.0930ms 3.7181ms 268.9515 Ops/s 265.6008 Ops/s $\color{#35bf28}+1.26\%$
test_td3_speed[False-None] 8.4097ms 8.2773ms 120.8126 Ops/s 122.5465 Ops/s $\color{#d91a1a}-1.41\%$
test_td3_speed[False-backward] 11.6552ms 11.2120ms 89.1899 Ops/s 90.4823 Ops/s $\color{#d91a1a}-1.43\%$
test_td3_speed[True-None] 1.9437ms 1.8857ms 530.3211 Ops/s 519.5591 Ops/s $\color{#35bf28}+2.07\%$
test_td3_speed[True-backward] 3.8699ms 3.7430ms 267.1632 Ops/s 269.6074 Ops/s $\color{#d91a1a}-0.91\%$
test_td3_speed[reduce-overhead-None] 1.8683ms 1.8352ms 544.9120 Ops/s 539.6428 Ops/s $\color{#35bf28}+0.98\%$
test_cql_speed[False-None] 31.4244ms 26.8968ms 37.1791 Ops/s 38.2002 Ops/s $\color{#d91a1a}-2.67\%$
test_cql_speed[False-backward] 39.9347ms 36.3246ms 27.5296 Ops/s 28.2302 Ops/s $\color{#d91a1a}-2.48\%$
test_cql_speed[True-None] 12.8047ms 12.3660ms 80.8671 Ops/s 80.6485 Ops/s $\color{#35bf28}+0.27\%$
test_cql_speed[True-backward] 18.8007ms 18.5253ms 53.9802 Ops/s 52.5995 Ops/s $\color{#35bf28}+2.62\%$
test_cql_speed[reduce-overhead-None] 15.7257ms 12.6648ms 78.9590 Ops/s 77.9687 Ops/s $\color{#35bf28}+1.27\%$
test_a2c_speed[False-None] 5.7304ms 5.4374ms 183.9111 Ops/s 185.0585 Ops/s $\color{#d91a1a}-0.62\%$
test_a2c_speed[False-backward] 12.3755ms 11.9487ms 83.6912 Ops/s 83.9828 Ops/s $\color{#d91a1a}-0.35\%$
test_a2c_speed[True-None] 4.0293ms 3.7714ms 265.1540 Ops/s 268.0954 Ops/s $\color{#d91a1a}-1.10\%$
test_a2c_speed[True-backward] 9.0277ms 8.7999ms 113.6382 Ops/s 114.8261 Ops/s $\color{#d91a1a}-1.03\%$
test_a2c_speed[reduce-overhead-None] 4.5284ms 3.7420ms 267.2397 Ops/s 263.7929 Ops/s $\color{#35bf28}+1.31\%$
test_ppo_speed[False-None] 6.4731ms 6.0890ms 164.2299 Ops/s 167.5200 Ops/s $\color{#d91a1a}-1.96\%$
test_ppo_speed[False-backward] 13.4447ms 12.8333ms 77.9222 Ops/s 78.3427 Ops/s $\color{#d91a1a}-0.54\%$
test_ppo_speed[True-None] 4.7149ms 3.6667ms 272.7215 Ops/s 271.7547 Ops/s $\color{#35bf28}+0.36\%$
test_ppo_speed[True-backward] 8.7329ms 8.5286ms 117.2526 Ops/s 114.5024 Ops/s $\color{#35bf28}+2.40\%$
test_ppo_speed[reduce-overhead-None] 3.7718ms 3.6082ms 277.1460 Ops/s 294.9494 Ops/s $\textbf{\color{#d91a1a}-6.04\%}$
test_reinforce_speed[False-None] 4.8396ms 4.5837ms 218.1666 Ops/s 228.2628 Ops/s $\color{#d91a1a}-4.42\%$
test_reinforce_speed[False-backward] 7.6565ms 7.4628ms 133.9978 Ops/s 135.8397 Ops/s $\color{#d91a1a}-1.36\%$
test_reinforce_speed[True-None] 3.7263ms 2.9531ms 338.6304 Ops/s 373.0998 Ops/s $\textbf{\color{#d91a1a}-9.24\%}$
test_reinforce_speed[True-backward] 8.1751ms 7.9123ms 126.3853 Ops/s 133.6620 Ops/s $\textbf{\color{#d91a1a}-5.44\%}$
test_reinforce_speed[reduce-overhead-None] 3.3188ms 2.9062ms 344.0894 Ops/s 376.9169 Ops/s $\textbf{\color{#d91a1a}-8.71\%}$
test_iql_speed[False-None] 23.4425ms 20.2544ms 49.3719 Ops/s 51.8607 Ops/s $\color{#d91a1a}-4.80\%$
test_iql_speed[False-backward] 31.6068ms 30.8798ms 32.3836 Ops/s 33.3651 Ops/s $\color{#d91a1a}-2.94\%$
test_iql_speed[True-None] 9.1551ms 8.5876ms 116.4470 Ops/s 122.2275 Ops/s $\color{#d91a1a}-4.73\%$
test_iql_speed[True-backward] 17.2623ms 16.8152ms 59.4699 Ops/s 63.7395 Ops/s $\textbf{\color{#d91a1a}-6.70\%}$
test_iql_speed[reduce-overhead-None] 9.0374ms 8.6095ms 116.1507 Ops/s 124.8511 Ops/s $\textbf{\color{#d91a1a}-6.97\%}$
test_rb_sample[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 6.3423ms 6.1184ms 163.4405 Ops/s 162.1200 Ops/s $\color{#35bf28}+0.81\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 2.2473ms 0.3595ms 2.7817 KOps/s 2.7749 KOps/s $\color{#35bf28}+0.24\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.6779ms 0.4167ms 2.3996 KOps/s 2.7374 KOps/s $\textbf{\color{#d91a1a}-12.34\%}$
test_rb_sample[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 6.1321ms 5.9082ms 169.2553 Ops/s 168.5526 Ops/s $\color{#35bf28}+0.42\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 2.3209ms 0.4252ms 2.3519 KOps/s 2.7868 KOps/s $\textbf{\color{#d91a1a}-15.60\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.6098ms 0.4080ms 2.4510 KOps/s 2.6777 KOps/s $\textbf{\color{#d91a1a}-8.47\%}$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-sampler6-10000] 1.7779ms 1.5019ms 665.8134 Ops/s 665.8500 Ops/s $-0.01\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-sampler7-10000] 1.7903ms 1.4255ms 701.5060 Ops/s 708.8661 Ops/s $\color{#d91a1a}-1.04\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 9.4885ms 6.2101ms 161.0284 Ops/s 163.5923 Ops/s $\color{#d91a1a}-1.57\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 1.8207ms 0.5058ms 1.9772 KOps/s 2.1569 KOps/s $\textbf{\color{#d91a1a}-8.33\%}$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.6406ms 0.4517ms 2.2140 KOps/s 2.1788 KOps/s $\color{#35bf28}+1.61\%$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-RandomSampler-4000] 6.0723ms 5.9563ms 167.8902 Ops/s 166.6730 Ops/s $\color{#35bf28}+0.73\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000] 2.0937ms 0.3391ms 2.9494 KOps/s 2.8412 KOps/s $\color{#35bf28}+3.81\%$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000] 0.5672ms 0.3760ms 2.6598 KOps/s 3.0835 KOps/s $\textbf{\color{#d91a1a}-13.74\%}$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000] 6.1717ms 5.9022ms 169.4297 Ops/s 167.9968 Ops/s $\color{#35bf28}+0.85\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000] 2.2276ms 0.2907ms 3.4396 KOps/s 3.5399 KOps/s $\color{#d91a1a}-2.83\%$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000] 0.6284ms 0.2713ms 3.6861 KOps/s 2.4498 KOps/s $\textbf{\color{#35bf28}+50.46\%}$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000] 6.2375ms 6.1121ms 163.6087 Ops/s 164.2291 Ops/s $\color{#d91a1a}-0.38\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000] 1.1918ms 0.5061ms 1.9759 KOps/s 2.0496 KOps/s $\color{#d91a1a}-3.59\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000] 0.6998ms 0.4772ms 2.0956 KOps/s 2.1719 KOps/s $\color{#d91a1a}-3.51\%$
test_rb_populate[TensorDictReplayBuffer-ListStorage-RandomSampler-400] 6.6678ms 5.1993ms 192.3319 Ops/s 50.5925 Ops/s $\textbf{\color{#35bf28}+280.16\%}$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-400] 10.9237ms 1.8929ms 528.2786 Ops/s 545.6110 Ops/s $\color{#d91a1a}-3.18\%$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-400] 1.1014ms 0.8989ms 1.1125 KOps/s 929.9040 Ops/s $\textbf{\color{#35bf28}+19.63\%}$
test_rb_populate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-400] 6.8023ms 5.1860ms 192.8281 Ops/s 191.1438 Ops/s $\color{#35bf28}+0.88\%$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-400] 3.9159ms 1.7927ms 557.8080 Ops/s 560.7438 Ops/s $\color{#d91a1a}-0.52\%$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-400] 6.4265ms 1.2270ms 815.0030 Ops/s 763.8170 Ops/s $\textbf{\color{#35bf28}+6.70\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-ListStorage-None-400] 0.5478s 16.2291ms 61.6179 Ops/s 58.0089 Ops/s $\textbf{\color{#35bf28}+6.22\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-400] 4.1508ms 1.9783ms 505.4849 Ops/s 527.5452 Ops/s $\color{#d91a1a}-4.18\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-400] 2.5297ms 1.1256ms 888.3866 Ops/s 959.7763 Ops/s $\textbf{\color{#d91a1a}-7.44\%}$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-True] 39.4080ms 37.0057ms 27.0229 Ops/s 27.2050 Ops/s $\color{#d91a1a}-0.67\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-False] 20.6244ms 18.8268ms 53.1158 Ops/s 55.1341 Ops/s $\color{#d91a1a}-3.66\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-True] 42.1929ms 38.0056ms 26.3119 Ops/s 26.7791 Ops/s $\color{#d91a1a}-1.74\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-False] 20.6111ms 19.0211ms 52.5731 Ops/s 54.6901 Ops/s $\color{#d91a1a}-3.87\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-True] 41.1948ms 39.5799ms 25.2653 Ops/s 25.2580 Ops/s $\color{#35bf28}+0.03\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-False] 21.7429ms 20.5137ms 48.7479 Ops/s 50.3334 Ops/s $\color{#d91a1a}-3.15\%$

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.

1 participant