[Feature] Trainer Infrastructure - Timing and Utilities #3188

vmoens · 2025-10-14T08:40:55Z

Stack from ghstack (oldest at bottom):

[ghstack-poisoned]

pytorch-bot · 2025-10-14T08:40:59Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/rl/3188

📄 Preview Python docs built from this PR

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

❌ 5 New Failures, 19 Pending, 1 Unrelated Failure

As of commit 4a09103 with merge base 92c20cd ():

NEW FAILURES - The following jobs have failed:

Generate documentation / build-docs (3.9, 12.8) / linux-job (gh)
RuntimeError: Command docker exec -t 5fd758574efdcf935bf62b8efbb56479321167e4d3e6aaef71556e10f15864ee /exec failed with exit code 1
Habitat Tests on Linux / tests (3.9, 12.8) / linux-job (gh)
RuntimeError: Command docker exec -t b132d7a2e355e030be987eecfc34615aada6908986e0942eb2e52e63d782530e /exec failed with exit code 1
Libs Tests on Linux / unittests-gym (3.9, 12.8) / linux-job (gh)
RuntimeError: Command docker exec -t 73bbf95fbac995563894cd010a612636e5134bf3dbdd820492a2b8540f303f38 /exec failed with exit code 1
Libs Tests on Linux / unittests-sklearn (3.10, 12.8) / linux-job (gh)
RuntimeError: Command docker exec -t fe66bee5c4335d0953210aaaa0055a71503204f8c6906d5ff06f7360f8cae155 /exec failed with exit code 2
SOTA Tests on Linux / tests (3.9, 12.8) / linux-job (gh)
RuntimeError: Command docker exec -t 06c74aea284623004af0cb23847329a763cbbee38ee4f28cb5afdf449b8f17ea /exec failed with exit code 1

BROKEN TRUNK - The following job failed but were present on the merge base:

👉 Rebase onto the `viable/strict` branch to avoid these failures

Unit-tests on Windows / unittests-cpu (3.10, windows.4xlarge, cpu) / windows-job (gh) (trunk failure)
Process completed with exit code 2.

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

[ghstack-poisoned]

github-actions · 2025-10-20T02:01:49Z

$\color{#D29922}\textsf{\Large&#x26A0;\kern{0.2cm}\normalsize Warning}$ Result of CPU Benchmark Tests

Total Benchmarks: 154. Improved: $\large\color{#35bf28}13$. Worsened: $\large\color{#d91a1a}7$.

Expand to view detailed results

Name	Max	Mean	Ops	Ops on Repo `HEAD`	Change
test_tensor_to_bytestream_speed[pickle]	81.9489μs	80.7821μs	12.3790 KOps/s	12.0883 KOps/s	$\color{#35bf28}+2.40\%$
test_tensor_to_bytestream_speed[torch.save]	0.1392ms	0.1380ms	7.2483 KOps/s	7.1275 KOps/s	$\color{#35bf28}+1.70\%$
test_tensor_to_bytestream_speed[untyped_storage]	0.1173s	0.1171s	8.5397 Ops/s	8.8809 Ops/s	$\color{#d91a1a}-3.84\%$
test_tensor_to_bytestream_speed[numpy]	2.8085μs	2.7990μs	357.2726 KOps/s	360.4763 KOps/s	$\color{#d91a1a}-0.89\%$
test_tensor_to_bytestream_speed[safetensors]	44.0547μs	43.7583μs	22.8528 KOps/s	23.2094 KOps/s	$\color{#d91a1a}-1.54\%$
test_simple	0.5456s	0.5450s	1.8350 Ops/s	1.7538 Ops/s	$\color{#35bf28}+4.63\%$
test_transformed	1.2140s	1.1202s	0.8927 Ops/s	0.8812 Ops/s	$\color{#35bf28}+1.30\%$
test_serial	1.6521s	1.6490s	0.6064 Ops/s	0.5913 Ops/s	$\color{#35bf28}+2.56\%$
test_parallel	1.1986s	1.1014s	0.9079 Ops/s	0.8936 Ops/s	$\color{#35bf28}+1.60\%$
test_step_mdp_speed[True-True-True-True-True]	0.1583ms	45.6370μs	21.9120 KOps/s	22.6713 KOps/s	$\color{#d91a1a}-3.35\%$
test_step_mdp_speed[True-True-True-True-False]	79.8740μs	25.2500μs	39.6039 KOps/s	39.5782 KOps/s	$\color{#35bf28}+0.06\%$
test_step_mdp_speed[True-True-True-False-True]	62.5230μs	26.1368μs	38.2602 KOps/s	39.6535 KOps/s	$\color{#d91a1a}-3.51\%$
test_step_mdp_speed[True-True-True-False-False]	57.2730μs	14.1349μs	70.7469 KOps/s	71.4294 KOps/s	$\color{#d91a1a}-0.96\%$
test_step_mdp_speed[True-True-False-True-True]	0.1015ms	49.4865μs	20.2075 KOps/s	20.5721 KOps/s	$\color{#d91a1a}-1.77\%$
test_step_mdp_speed[True-True-False-True-False]	68.1130μs	28.4139μs	35.1941 KOps/s	35.3917 KOps/s	$\color{#d91a1a}-0.56\%$
test_step_mdp_speed[True-True-False-False-True]	61.8330μs	28.6336μs	34.9240 KOps/s	35.3192 KOps/s	$\color{#d91a1a}-1.12\%$
test_step_mdp_speed[True-True-False-False-False]	49.2820μs	17.1387μs	58.3476 KOps/s	59.4699 KOps/s	$\color{#d91a1a}-1.89\%$
test_step_mdp_speed[True-False-True-True-True]	97.8150μs	52.7643μs	18.9522 KOps/s	19.6882 KOps/s	$\color{#d91a1a}-3.74\%$
test_step_mdp_speed[True-False-True-True-False]	69.0430μs	31.3108μs	31.9378 KOps/s	32.3067 KOps/s	$\color{#d91a1a}-1.14\%$
test_step_mdp_speed[True-False-True-False-True]	63.9630μs	29.1203μs	34.3403 KOps/s	35.4443 KOps/s	$\color{#d91a1a}-3.11\%$
test_step_mdp_speed[True-False-True-False-False]	46.6430μs	17.0788μs	58.5523 KOps/s	59.1391 KOps/s	$\color{#d91a1a}-0.99\%$
test_step_mdp_speed[True-False-False-True-True]	92.0640μs	54.9543μs	18.1969 KOps/s	18.6314 KOps/s	$\color{#d91a1a}-2.33\%$
test_step_mdp_speed[True-False-False-True-False]	66.7040μs	33.5686μs	29.7897 KOps/s	29.7962 KOps/s	$\color{#d91a1a}-0.02\%$
test_step_mdp_speed[True-False-False-False-True]	66.6430μs	31.5566μs	31.6891 KOps/s	32.0772 KOps/s	$\color{#d91a1a}-1.21\%$
test_step_mdp_speed[True-False-False-False-False]	51.2130μs	19.6025μs	51.0138 KOps/s	51.9703 KOps/s	$\color{#d91a1a}-1.84\%$
test_step_mdp_speed[False-True-True-True-True]	93.4250μs	52.1964μs	19.1584 KOps/s	19.5085 KOps/s	$\color{#d91a1a}-1.79\%$
test_step_mdp_speed[False-True-True-True-False]	62.9830μs	30.9711μs	32.2882 KOps/s	32.1071 KOps/s	$\color{#35bf28}+0.56\%$
test_step_mdp_speed[False-True-True-False-True]	2.3322ms	33.2089μs	30.1124 KOps/s	31.0600 KOps/s	$\color{#d91a1a}-3.05\%$
test_step_mdp_speed[False-True-True-False-False]	51.5720μs	19.1002μs	52.3553 KOps/s	53.6912 KOps/s	$\color{#d91a1a}-2.49\%$
test_step_mdp_speed[False-True-False-True-True]	92.9050μs	54.8424μs	18.2341 KOps/s	18.6437 KOps/s	$\color{#d91a1a}-2.20\%$
test_step_mdp_speed[False-True-False-True-False]	72.1040μs	33.5163μs	29.8363 KOps/s	29.5197 KOps/s	$\color{#35bf28}+1.07\%$
test_step_mdp_speed[False-True-False-False-True]	71.5030μs	35.5057μs	28.1645 KOps/s	28.5616 KOps/s	$\color{#d91a1a}-1.39\%$
test_step_mdp_speed[False-True-False-False-False]	57.7720μs	21.5308μs	46.4451 KOps/s	46.8586 KOps/s	$\color{#d91a1a}-0.88\%$
test_step_mdp_speed[False-False-True-True-True]	90.7240μs	57.7399μs	17.3191 KOps/s	17.8240 KOps/s	$\color{#d91a1a}-2.83\%$
test_step_mdp_speed[False-False-True-True-False]	76.3030μs	36.7077μs	27.2422 KOps/s	27.6076 KOps/s	$\color{#d91a1a}-1.32\%$
test_step_mdp_speed[False-False-True-False-True]	70.6040μs	35.6088μs	28.0829 KOps/s	28.9265 KOps/s	$\color{#d91a1a}-2.92\%$
test_step_mdp_speed[False-False-True-False-False]	63.2440μs	21.6016μs	46.2928 KOps/s	47.0469 KOps/s	$\color{#d91a1a}-1.60\%$
test_step_mdp_speed[False-False-False-True-True]	0.1170ms	60.5115μs	16.5258 KOps/s	17.1221 KOps/s	$\color{#d91a1a}-3.48\%$
test_step_mdp_speed[False-False-False-True-False]	67.8740μs	39.0876μs	25.5836 KOps/s	26.0007 KOps/s	$\color{#d91a1a}-1.60\%$
test_step_mdp_speed[False-False-False-False-True]	0.1169ms	37.2701μs	26.8312 KOps/s	27.1644 KOps/s	$\color{#d91a1a}-1.23\%$
test_step_mdp_speed[False-False-False-False-False]	66.5830μs	24.0418μs	41.5942 KOps/s	42.5475 KOps/s	$\color{#d91a1a}-2.24\%$
test_values[generalized_advantage_estimate-True-True]	10.9861ms	10.3397ms	96.7144 Ops/s	105.9574 Ops/s	$\textbf{\color{#d91a1a}-8.72\%}$
test_values[vec_generalized_advantage_estimate-True-True]	21.5888ms	17.8063ms	56.1599 Ops/s	57.0165 Ops/s	$\color{#d91a1a}-1.50\%$
test_values[td0_return_estimate-False-False]	0.2366ms	0.1360ms	7.3549 KOps/s	7.2272 KOps/s	$\color{#35bf28}+1.77\%$
test_values[td1_return_estimate-False-False]	28.0287ms	27.3026ms	36.6266 Ops/s	37.9213 Ops/s	$\color{#d91a1a}-3.41\%$
test_values[vec_td1_return_estimate-False-False]	18.7145ms	17.6671ms	56.6024 Ops/s	57.1265 Ops/s	$\color{#d91a1a}-0.92\%$
test_values[td_lambda_return_estimate-True-False]	41.3156ms	39.8251ms	25.1098 Ops/s	25.1433 Ops/s	$\color{#d91a1a}-0.13\%$
test_values[vec_td_lambda_return_estimate-True-False]	22.2532ms	18.0069ms	55.5342 Ops/s	56.3432 Ops/s	$\color{#d91a1a}-1.44\%$
test_gae_speed[generalized_advantage_estimate-False-1-512]	8.2437ms	8.0867ms	123.6601 Ops/s	122.0487 Ops/s	$\color{#35bf28}+1.32\%$
test_gae_speed[vec_generalized_advantage_estimate-True-1-512]	1.8868ms	1.4759ms	677.5521 Ops/s	672.9413 Ops/s	$\color{#35bf28}+0.69\%$
test_gae_speed[vec_generalized_advantage_estimate-False-1-512]	0.5600ms	0.4140ms	2.4156 KOps/s	2.4570 KOps/s	$\color{#d91a1a}-1.69\%$
test_gae_speed[vec_generalized_advantage_estimate-True-32-512]	34.8906ms	34.1754ms	29.2608 Ops/s	28.9677 Ops/s	$\color{#35bf28}+1.01\%$
test_gae_speed[vec_generalized_advantage_estimate-False-32-512]	1.9462ms	1.7529ms	570.4742 Ops/s	589.8765 Ops/s	$\color{#d91a1a}-3.29\%$
test_dqn_speed[False-None]	1.4592ms	1.3816ms	723.7993 Ops/s	705.2635 Ops/s	$\color{#35bf28}+2.63\%$
test_dqn_speed[False-backward]	1.9416ms	1.8915ms	528.6925 Ops/s	520.5543 Ops/s	$\color{#35bf28}+1.56\%$
test_dqn_speed[True-None]	1.0168ms	0.5290ms	1.8903 KOps/s	1.9432 KOps/s	$\color{#d91a1a}-2.72\%$
test_dqn_speed[True-backward]	1.0171ms	0.9609ms	1.0407 KOps/s	903.7037 Ops/s	$\textbf{\color{#35bf28}+15.16\%}$
test_dqn_speed[reduce-overhead-None]	0.9927ms	0.5148ms	1.9423 KOps/s	1.9471 KOps/s	$\color{#d91a1a}-0.25\%$
test_dqn_speed[reduce-overhead-backward]	0.9765ms	0.9444ms	1.0589 KOps/s	1.0376 KOps/s	$\color{#35bf28}+2.06\%$
test_ddpg_speed[False-None]	3.2589ms	2.8380ms	352.3602 Ops/s	348.4727 Ops/s	$\color{#35bf28}+1.12\%$
test_ddpg_speed[False-backward]	4.1319ms	4.0414ms	247.4366 Ops/s	245.0635 Ops/s	$\color{#35bf28}+0.97\%$
test_ddpg_speed[True-None]	1.5248ms	1.3947ms	717.0217 Ops/s	693.0376 Ops/s	$\color{#35bf28}+3.46\%$
test_ddpg_speed[True-backward]	2.8550ms	2.3677ms	422.3557 Ops/s	412.9036 Ops/s	$\color{#35bf28}+2.29\%$
test_ddpg_speed[reduce-overhead-None]	1.5463ms	1.3988ms	714.8962 Ops/s	714.4398 Ops/s	$\color{#35bf28}+0.06\%$
test_ddpg_speed[reduce-overhead-backward]	2.4306ms	2.3465ms	426.1684 Ops/s	417.0065 Ops/s	$\color{#35bf28}+2.20\%$
test_sac_speed[False-None]	8.2538ms	7.8018ms	128.1748 Ops/s	126.8954 Ops/s	$\color{#35bf28}+1.01\%$
test_sac_speed[False-backward]	11.4941ms	11.0471ms	90.5217 Ops/s	89.9931 Ops/s	$\color{#35bf28}+0.59\%$
test_sac_speed[True-None]	2.2737ms	2.1047ms	475.1168 Ops/s	464.3307 Ops/s	$\color{#35bf28}+2.32\%$
test_sac_speed[True-backward]	4.1149ms	3.9349ms	254.1350 Ops/s	240.7261 Ops/s	$\textbf{\color{#35bf28}+5.57\%}$
test_sac_speed[reduce-overhead-None]	2.2544ms	2.0721ms	482.6066 Ops/s	454.0521 Ops/s	$\textbf{\color{#35bf28}+6.29\%}$
test_sac_speed[reduce-overhead-backward]	4.4007ms	4.0121ms	249.2437 Ops/s	242.9181 Ops/s	$\color{#35bf28}+2.60\%$
test_redq_speed[False-None]	12.9190ms	10.1892ms	98.1429 Ops/s	95.5865 Ops/s	$\color{#35bf28}+2.67\%$
test_redq_speed[False-backward]	19.9649ms	17.5978ms	56.8252 Ops/s	55.2456 Ops/s	$\color{#35bf28}+2.86\%$
test_redq_speed[True-None]	6.5929ms	4.4127ms	226.6182 Ops/s	225.8360 Ops/s	$\color{#35bf28}+0.35\%$
test_redq_speed[True-backward]	10.0430ms	9.6732ms	103.3786 Ops/s	99.4901 Ops/s	$\color{#35bf28}+3.91\%$
test_redq_speed[reduce-overhead-None]	4.8204ms	4.4075ms	226.8859 Ops/s	230.9616 Ops/s	$\color{#d91a1a}-1.76\%$
test_redq_speed[reduce-overhead-backward]	10.1412ms	9.9014ms	100.9962 Ops/s	101.0293 Ops/s	$\color{#d91a1a}-0.03\%$
test_redq_deprec_speed[False-None]	11.3399ms	10.8214ms	92.4093 Ops/s	90.3529 Ops/s	$\color{#35bf28}+2.28\%$
test_redq_deprec_speed[False-backward]	16.2369ms	15.6780ms	63.7837 Ops/s	63.1841 Ops/s	$\color{#35bf28}+0.95\%$
test_redq_deprec_speed[True-None]	3.7576ms	3.5996ms	277.8111 Ops/s	274.7580 Ops/s	$\color{#35bf28}+1.11\%$
test_redq_deprec_speed[True-backward]	8.4121ms	7.5600ms	132.2754 Ops/s	124.4188 Ops/s	$\textbf{\color{#35bf28}+6.31\%}$
test_redq_deprec_speed[reduce-overhead-None]	3.9520ms	3.5543ms	281.3502 Ops/s	275.6518 Ops/s	$\color{#35bf28}+2.07\%$
test_redq_deprec_speed[reduce-overhead-backward]	7.7632ms	7.4987ms	133.3570 Ops/s	127.6410 Ops/s	$\color{#35bf28}+4.48\%$
test_td3_speed[False-None]	7.9642ms	7.7973ms	128.2501 Ops/s	127.5609 Ops/s	$\color{#35bf28}+0.54\%$
test_td3_speed[False-backward]	11.2636ms	10.7168ms	93.3115 Ops/s	93.0879 Ops/s	$\color{#35bf28}+0.24\%$
test_td3_speed[True-None]	1.8585ms	1.8133ms	551.4901 Ops/s	544.2661 Ops/s	$\color{#35bf28}+1.33\%$
test_td3_speed[True-backward]	3.7832ms	3.6167ms	276.4971 Ops/s	266.9857 Ops/s	$\color{#35bf28}+3.56\%$
test_td3_speed[reduce-overhead-None]	1.8041ms	1.7517ms	570.8597 Ops/s	556.3471 Ops/s	$\color{#35bf28}+2.61\%$
test_td3_speed[reduce-overhead-backward]	3.7325ms	3.6354ms	275.0758 Ops/s	265.8834 Ops/s	$\color{#35bf28}+3.46\%$
test_cql_speed[False-None]	26.0735ms	25.3556ms	39.4390 Ops/s	38.2680 Ops/s	$\color{#35bf28}+3.06\%$
test_cql_speed[False-backward]	35.4644ms	34.6686ms	28.8445 Ops/s	28.9774 Ops/s	$\color{#d91a1a}-0.46\%$
test_cql_speed[True-None]	12.6716ms	12.1798ms	82.1032 Ops/s	77.8576 Ops/s	$\textbf{\color{#35bf28}+5.45\%}$
test_cql_speed[True-backward]	18.4171ms	18.0043ms	55.5422 Ops/s	53.2955 Ops/s	$\color{#35bf28}+4.22\%$
test_cql_speed[reduce-overhead-None]	13.0021ms	12.3938ms	80.6857 Ops/s	80.7475 Ops/s	$\color{#d91a1a}-0.08\%$
test_cql_speed[reduce-overhead-backward]	18.4319ms	18.1013ms	55.2448 Ops/s	54.5310 Ops/s	$\color{#35bf28}+1.31\%$
test_a2c_speed[False-None]	5.5280ms	5.3432ms	187.1540 Ops/s	188.1219 Ops/s	$\color{#d91a1a}-0.51\%$
test_a2c_speed[False-backward]	11.9388ms	11.7641ms	85.0046 Ops/s	84.8159 Ops/s	$\color{#35bf28}+0.22\%$
test_a2c_speed[True-None]	3.7895ms	3.6796ms	271.7683 Ops/s	267.1732 Ops/s	$\color{#35bf28}+1.72\%$
test_a2c_speed[True-backward]	8.6690ms	8.4005ms	119.0412 Ops/s	112.1460 Ops/s	$\textbf{\color{#35bf28}+6.15\%}$
test_a2c_speed[reduce-overhead-None]	4.0532ms	3.7076ms	269.7134 Ops/s	271.5691 Ops/s	$\color{#d91a1a}-0.68\%$
test_a2c_speed[reduce-overhead-backward]	9.1445ms	8.8147ms	113.4466 Ops/s	113.3676 Ops/s	$\color{#35bf28}+0.07\%$
test_ppo_speed[False-None]	6.1248ms	5.8717ms	170.3094 Ops/s	170.6653 Ops/s	$\color{#d91a1a}-0.21\%$
test_ppo_speed[False-backward]	12.9125ms	12.6127ms	79.2853 Ops/s	81.5841 Ops/s	$\color{#d91a1a}-2.82\%$
test_ppo_speed[True-None]	3.7348ms	3.6194ms	276.2862 Ops/s	275.5039 Ops/s	$\color{#35bf28}+0.28\%$
test_ppo_speed[True-backward]	8.6030ms	8.3943ms	119.1288 Ops/s	113.1229 Ops/s	$\textbf{\color{#35bf28}+5.31\%}$
test_ppo_speed[reduce-overhead-None]	3.7238ms	3.6002ms	277.7590 Ops/s	275.5893 Ops/s	$\color{#35bf28}+0.79\%$
test_ppo_speed[reduce-overhead-backward]	8.8631ms	8.6319ms	115.8498 Ops/s	114.7101 Ops/s	$\color{#35bf28}+0.99\%$
test_reinforce_speed[False-None]	4.7374ms	4.5597ms	219.3136 Ops/s	218.8050 Ops/s	$\color{#35bf28}+0.23\%$
test_reinforce_speed[False-backward]	7.6737ms	7.3369ms	136.2966 Ops/s	137.2166 Ops/s	$\color{#d91a1a}-0.67\%$
test_reinforce_speed[True-None]	3.0144ms	2.8497ms	350.9162 Ops/s	346.6442 Ops/s	$\color{#35bf28}+1.23\%$
test_reinforce_speed[True-backward]	7.9599ms	7.5825ms	131.8821 Ops/s	116.6526 Ops/s	$\textbf{\color{#35bf28}+13.06\%}$
test_reinforce_speed[reduce-overhead-None]	3.2242ms	2.8599ms	349.6569 Ops/s	347.0266 Ops/s	$\color{#35bf28}+0.76\%$
test_reinforce_speed[reduce-overhead-backward]	8.2841ms	7.7747ms	128.6230 Ops/s	123.9685 Ops/s	$\color{#35bf28}+3.75\%$
test_iql_speed[False-None]	23.5273ms	20.0118ms	49.9704 Ops/s	47.4573 Ops/s	$\textbf{\color{#35bf28}+5.30\%}$
test_iql_speed[False-backward]	30.6527ms	30.0626ms	33.2639 Ops/s	32.6251 Ops/s	$\color{#35bf28}+1.96\%$
test_iql_speed[True-None]	8.8996ms	8.4707ms	118.0542 Ops/s	116.2771 Ops/s	$\color{#35bf28}+1.53\%$
test_iql_speed[True-backward]	16.7989ms	16.4264ms	60.8775 Ops/s	59.2668 Ops/s	$\color{#35bf28}+2.72\%$
test_iql_speed[reduce-overhead-None]	9.2429ms	8.5278ms	117.2635 Ops/s	113.8745 Ops/s	$\color{#35bf28}+2.98\%$
test_iql_speed[reduce-overhead-backward]	17.3403ms	16.8518ms	59.3407 Ops/s	58.5258 Ops/s	$\color{#35bf28}+1.39\%$
test_rb_sample[TensorDictReplayBuffer-ListStorage-RandomSampler-4000]	7.5368ms	6.0597ms	165.0241 Ops/s	165.2813 Ops/s	$\color{#d91a1a}-0.16\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000]	0.5435ms	0.3184ms	3.1410 KOps/s	3.1035 KOps/s	$\color{#35bf28}+1.21\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000]	0.5589ms	0.3029ms	3.3010 KOps/s	3.1705 KOps/s	$\color{#35bf28}+4.12\%$
test_rb_sample[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000]	6.0830ms	5.7620ms	173.5516 Ops/s	174.0535 Ops/s	$\color{#d91a1a}-0.29\%$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000]	1.9982ms	0.3446ms	2.9023 KOps/s	3.0915 KOps/s	$\textbf{\color{#d91a1a}-6.12\%}$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000]	0.8163ms	0.3289ms	3.0403 KOps/s	3.5167 KOps/s	$\textbf{\color{#d91a1a}-13.55\%}$
test_rb_sample[TensorDictReplayBuffer-LazyMemmapStorage-sampler6-10000]	1.6230ms	1.3439ms	744.0845 Ops/s	728.7858 Ops/s	$\color{#35bf28}+2.10\%$
test_rb_sample[TensorDictReplayBuffer-LazyTensorStorage-sampler7-10000]	1.6472ms	1.2595ms	793.9428 Ops/s	777.8792 Ops/s	$\color{#35bf28}+2.07\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000]	9.4010ms	6.0096ms	166.4000 Ops/s	169.8226 Ops/s	$\color{#d91a1a}-2.02\%$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000]	1.7728ms	0.5005ms	1.9982 KOps/s	2.1688 KOps/s	$\textbf{\color{#d91a1a}-7.87\%}$
test_rb_sample[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000]	0.7324ms	0.4619ms	2.1651 KOps/s	2.2888 KOps/s	$\textbf{\color{#d91a1a}-5.40\%}$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-RandomSampler-4000]	5.9091ms	5.7258ms	174.6474 Ops/s	173.2759 Ops/s	$\color{#35bf28}+0.79\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-10000]	0.7392ms	0.2759ms	3.6248 KOps/s	2.9239 KOps/s	$\textbf{\color{#35bf28}+23.97\%}$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-10000]	0.4927ms	0.2943ms	3.3974 KOps/s	3.3836 KOps/s	$\color{#35bf28}+0.41\%$
test_rb_iterate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-4000]	5.9459ms	5.6252ms	177.7708 Ops/s	175.9916 Ops/s	$\color{#35bf28}+1.01\%$
test_rb_iterate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-10000]	1.7148ms	0.2773ms	3.6066 KOps/s	2.9746 KOps/s	$\textbf{\color{#35bf28}+21.25\%}$
test_rb_iterate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-10000]	0.4424ms	0.2550ms	3.9221 KOps/s	3.0433 KOps/s	$\textbf{\color{#35bf28}+28.87\%}$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-ListStorage-None-4000]	6.4500ms	5.8483ms	170.9909 Ops/s	168.3994 Ops/s	$\color{#35bf28}+1.54\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-10000]	2.0820ms	0.4324ms	2.3128 KOps/s	2.2057 KOps/s	$\color{#35bf28}+4.86\%$
test_rb_iterate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-10000]	0.8522ms	0.4197ms	2.3827 KOps/s	2.4472 KOps/s	$\color{#d91a1a}-2.63\%$
test_rb_populate[TensorDictReplayBuffer-ListStorage-RandomSampler-400]	6.7297ms	5.0642ms	197.4650 Ops/s	196.5575 Ops/s	$\color{#35bf28}+0.46\%$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-RandomSampler-400]	9.8701ms	2.2982ms	435.1250 Ops/s	493.7832 Ops/s	$\textbf{\color{#d91a1a}-11.88\%}$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-RandomSampler-400]	6.5126ms	1.1529ms	867.3902 Ops/s	854.8696 Ops/s	$\color{#35bf28}+1.46\%$
test_rb_populate[TensorDictReplayBuffer-ListStorage-SamplerWithoutReplacement-400]	0.5196s	15.3883ms	64.9845 Ops/s	192.7346 Ops/s	$\textbf{\color{#d91a1a}-66.28\%}$
test_rb_populate[TensorDictReplayBuffer-LazyMemmapStorage-SamplerWithoutReplacement-400]	5.6755ms	1.9755ms	506.2045 Ops/s	493.7782 Ops/s	$\color{#35bf28}+2.52\%$
test_rb_populate[TensorDictReplayBuffer-LazyTensorStorage-SamplerWithoutReplacement-400]	10.1247ms	1.1904ms	840.0720 Ops/s	854.1152 Ops/s	$\color{#d91a1a}-1.64\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-ListStorage-None-400]	8.6514ms	5.3430ms	187.1610 Ops/s	56.4402 Ops/s	$\textbf{\color{#35bf28}+231.61\%}$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyMemmapStorage-None-400]	8.6887ms	2.1920ms	456.1966 Ops/s	467.7500 Ops/s	$\color{#d91a1a}-2.47\%$
test_rb_populate[TensorDictPrioritizedReplayBuffer-LazyTensorStorage-None-400]	1.9662ms	1.0198ms	980.6287 Ops/s	999.8175 Ops/s	$\color{#d91a1a}-1.92\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-True]	35.4040ms	32.8856ms	30.4084 Ops/s	30.6459 Ops/s	$\color{#d91a1a}-0.77\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-10000-10000-100-False]	18.8210ms	17.2898ms	57.8376 Ops/s	57.5389 Ops/s	$\color{#35bf28}+0.52\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-True]	36.2382ms	33.8590ms	29.5342 Ops/s	29.6455 Ops/s	$\color{#d91a1a}-0.38\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-100000-10000-100-False]	19.6914ms	17.9091ms	55.8377 Ops/s	56.2205 Ops/s	$\color{#d91a1a}-0.68\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-True]	37.8839ms	35.9456ms	27.8199 Ops/s	28.6006 Ops/s	$\color{#d91a1a}-2.73\%$
test_rb_extend_sample[ReplayBuffer-LazyTensorStorage-RandomSampler-1000000-10000-100-False]	20.1904ms	18.9693ms	52.7168 Ops/s	52.1843 Ops/s	$\color{#35bf28}+1.02\%$

[ghstack-poisoned]

Update

f553b66

[ghstack-poisoned]

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 Oct 14, 2025

vmoens added 7 commits October 16, 2025 15:00

Update

bda4145

[ghstack-poisoned]

Update

be2c94b

[ghstack-poisoned]

Update

876de4f

[ghstack-poisoned]

Update

0ad3f7a

[ghstack-poisoned]

Update

019f157

[ghstack-poisoned]

Update

db37e8c

[ghstack-poisoned]

Update

84b6f37

[ghstack-poisoned]

vmoens mentioned this pull request Oct 18, 2025

[Quality] Fix flaky test #3211

Open

vmoens added 4 commits October 18, 2025 19:00

Update

475044d

[ghstack-poisoned]

Update

2dc2e22

[ghstack-poisoned]

Update

c64e7f2

[ghstack-poisoned]

Update

9bad747

[ghstack-poisoned]

vmoens added the enhancement New feature or request label Oct 20, 2025

Update

a05a91d

[ghstack-poisoned]

vmoens mentioned this pull request Oct 20, 2025

[CI] Fix benchmarks for LLMs #3212

Merged

Update

c78b8d2

[ghstack-poisoned]

vmoens mentioned this pull request Oct 20, 2025

[BugFix] Fix GRPO tests and runs #3213

Merged

Update

3579eaa

[ghstack-poisoned]

vmoens added 6 commits October 19, 2025 20:56

Update

30d4e53

[ghstack-poisoned]

Update

e97f430

[ghstack-poisoned]

Update

495a051

[ghstack-poisoned]

Update

82ba9b9

[ghstack-poisoned]

Update

d30b5f4

[ghstack-poisoned]

Update

4a09103

[ghstack-poisoned]

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

[Feature] Trainer Infrastructure - Timing and Utilities #3188

[Feature] Trainer Infrastructure - Timing and Utilities #3188

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

[Feature] Trainer Infrastructure - Timing and Utilities #3188

Are you sure you want to change the base?

[Feature] Trainer Infrastructure - Timing and Utilities #3188

Conversation

Uh oh!

Uh oh!

Uh oh!

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/rl/3188

❌ 5 New Failures, 19 Pending, 1 Unrelated Failure

Uh oh!

Uh oh!

$\color{#D29922}\textsf{\Large&amp;#x26A0;\kern{0.2cm}\normalsize Warning}$ Result of CPU Benchmark Tests

Total Benchmarks: 154. Improved: $\large\color{#35bf28}13$. Worsened: $\large\color{#d91a1a}7$.

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

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