Skip to page content
SAVRN
Search Plan a project

Dataset

grpo_dppo_qwen3_4B_maxrl_polaris1000_knk_rollouts

by Fahim Tajwar ftajwar/grpo_dppo_qwen3_4B_maxrl_polaris1000_knk_rollouts

Training rollouts of one verl run, one parquet per trainer step (data/step-.parquet), rows in the order the trainer consumed them; the rollout.n samples of a prompt are contiguous and share uid.

Rows1,056,768
Configurations1
Size10.2 GB
License—
AccessPublicly accessible
Monthly Downloads396

Dataset Card

Training rollouts of one verl run, one parquet per trainer step (data/step-.parquet), rows in the order the trainer consumed them; the rollout.n samples of a prompt are contiguous and share uid. prompttokenids / responsetokenids are the exact token ids (special tokens kept, no padding) and rolloutlogprobs the sampling-time log-prob of every response token. input / output are decoded with skipspecialtokens=True and are for reading only. score is the raw reward.

Excerpt from the card by Fahim Tajwar.

Structure

default 1,056,768 rows

SplitRowsSize
train1,056,76833.4 GB
inputstringoutputstringgtsstringscorefloat64stepint64accfloat64prompt_token_idslistresponse_token_idslistrollout_log_probslistuidstring

Details

Repository
ftajwar/grpo_dppo_qwen3_4B_maxrl_polaris1000_knk_rollouts
Publisher
Fahim Tajwar
Task category
Not stated by the source
Tags
Not stated by the source
Size category
Not stated by the source
Languages
Not stated by the source
Revision
6c48a69e15684a75269fa4d0017d87fa34fe9a99
Last updated
2026-10-08

Files

259 files, 10.2 GB in total.

Data257 files · 10.2 GB
Documentation1 file · 597 B
Repository1 file · 2.5 KB
Every file
FileTypeSizeSHA-256
data/step-0001.parquetData27.6 MBf8ca4b79292b
data/step-0002.parquetData27.7 MBd4bfe72678f0
data/step-0003.parquetData26.8 MB520128a10205
data/step-0004.parquetData28.7 MBbf3a181f4ffb
data/step-0005.parquetData27.8 MBc1cd5b3159cd
data/step-0006.parquetData27.4 MB4c37c74a2438
data/step-0007.parquetData28.5 MB423a15cdd3cc
data/step-0008.parquetData28.3 MBcbec55155b7c
data/step-0009.parquetData28.9 MB37ed1b94a387
data/step-0010.parquetData28.6 MBb31a741b9ba5
data/step-0011.parquetData28.8 MB6f9f69927251
data/step-0012.parquetData28.2 MB7c7a047c1f08
data/step-0013.parquetData30.3 MBc8f9ae9fd66a
data/step-0014.parquetData30.3 MBaf4700a2c0ed
data/step-0015.parquetData30.6 MBf902deb73735
data/step-0016.parquetData31.5 MB87af83e48057
data/step-0017.parquetData31.8 MB59e8f3eb2cc2
data/step-0018.parquetData30.7 MB481e23425b2e
data/step-0019.parquetData32.1 MB8d2ca1903a2d
data/step-0020.parquetData32.8 MBdecfaa1cb9c2
data/step-0021.parquetData33.4 MB4d200729e633
data/step-0022.parquetData33.2 MB2d9978373687
data/step-0023.parquetData33.7 MBc85364ffffcf
data/step-0024.parquetData34.1 MB1064062dd38c
data/step-0025.parquetData36.0 MBfacd371ddc70
data/step-0026.parquetData37.0 MB3689d90a1c1a
data/step-0027.parquetData36.8 MB28aaf2e106ec
data/step-0028.parquetData36.0 MBe748cacb0546
data/step-0029.parquetData38.6 MB4bb22105e599
data/step-0030.parquetData39.2 MBca062c0f51cc
data/step-0031.parquetData37.3 MBc8ece9016d40
data/step-0032.parquetData40.3 MB5fdbf93e3ab1
data/step-0033.parquetData39.3 MB05d2a9e3fc9d
data/step-0034.parquetData39.7 MB2db0b62a567a
data/step-0035.parquetData40.5 MBda6c21be941b
data/step-0036.parquetData40.4 MBac0e5bd42fad
data/step-0037.parquetData40.6 MB15967bb8fd6d
data/step-0038.parquetData41.7 MB75dd9eadc3d1
data/step-0039.parquetData41.5 MBb84fb419576c
data/step-0040.parquetData40.4 MBec5e34d64056
data/step-0041.parquetData41.2 MB05c941e26071
data/step-0042.parquetData43.1 MB41f535e124c8
data/step-0043.parquetData42.6 MB795c37d3e712
data/step-0044.parquetData42.7 MB3c53662ef1fa
data/step-0045.parquetData43.0 MBb417d1837c9b
data/step-0046.parquetData42.1 MBc488d8c2baad
data/step-0047.parquetData41.7 MBf9de95a7f583
data/step-0048.parquetData43.0 MBc8a8edfa998d
data/step-0049.parquetData43.9 MBd47abb2f8f4e
data/step-0050.parquetData43.7 MB72d5905f0b49
data/step-0051.parquetData43.4 MBacc3b8452083
data/step-0052.parquetData44.2 MB58517fd58e5e
data/step-0053.parquetData46.6 MB2adb67774fe0
data/step-0054.parquetData45.1 MB804f2e8876a6
data/step-0055.parquetData44.9 MBf8514f5419ac
data/step-0056.parquetData44.6 MB26ac8231685c
data/step-0057.parquetData45.6 MB38e6fcb09882
data/step-0058.parquetData44.0 MB88aebcc2c6e4
data/step-0059.parquetData43.7 MB713f3d32cf49
data/step-0060.parquetData45.5 MB2f31d1c79968
data/step-0061.parquetData46.5 MB72db74cc64dc
data/step-0062.parquetData44.2 MB80da49f339d4
data/step-0063.parquetData46.8 MB76e616f1c546
data/step-0064.parquetData48.0 MB77e8e2bef587
data/step-0065.parquetData46.8 MB70f59321463c
data/step-0066.parquetData45.9 MBf43e2cb455c4
data/step-0067.parquetData46.9 MB3bbdb5642868
data/step-0068.parquetData47.3 MB30fb645e5c17
data/step-0069.parquetData45.2 MB8dea47f656a9
data/step-0070.parquetData47.6 MB03d6935c9bc2
data/step-0071.parquetData47.6 MB3e207254b049
data/step-0072.parquetData47.9 MB64544df2afd6
data/step-0073.parquetData46.8 MB46623502c74b
data/step-0074.parquetData48.1 MB897d1063b350
data/step-0075.parquetData47.2 MB9d99e315e0ac
data/step-0076.parquetData47.0 MBdb991fcbc0fb
data/step-0077.parquetData46.7 MBcab88cc1b960
data/step-0078.parquetData46.9 MB76bcf481bcc5
data/step-0079.parquetData47.9 MB4257d3e59ee7
data/step-0080.parquetData48.8 MB966640c7306f
data/step-0081.parquetData47.3 MBde106c5d6543
data/step-0082.parquetData47.9 MB2392e7d6dfd9
data/step-0083.parquetData48.9 MBdb5499b2faae
data/step-0084.parquetData47.6 MBf8a6a2a61564
data/step-0085.parquetData47.5 MB5f113bfece9b
data/step-0086.parquetData47.3 MB06495627df4a
data/step-0087.parquetData45.6 MB84fc1b934ec9
data/step-0088.parquetData45.5 MBb6aae118019d
data/step-0089.parquetData48.8 MBd47e335833ef
data/step-0090.parquetData48.6 MBa7879d5fa925
data/step-0091.parquetData46.8 MBca47389c508e
data/step-0092.parquetData47.8 MB50684fb068c3
data/step-0093.parquetData47.2 MB419ecd3dcc6f
data/step-0094.parquetData47.7 MB1e09b2ba697e
data/step-0095.parquetData49.0 MBf8c5fe3d045a
data/step-0096.parquetData47.9 MB2ec64b59889d
data/step-0097.parquetData49.0 MB41d4dbd91018
data/step-0098.parquetData46.1 MBf36a17fef35a
data/step-0099.parquetData46.5 MB4c3a41f2372c
data/step-0100.parquetData45.9 MB3d989579d224
data/step-0101.parquetData46.5 MB90b161403654
data/step-0102.parquetData46.1 MBad637c0e7846
data/step-0103.parquetData44.9 MB4aff95fd4641
data/step-0104.parquetData45.7 MB2bfd6c2eb0af
data/step-0105.parquetData45.6 MBc53bf097d8f1
data/step-0106.parquetData45.1 MB8660dc63d0f0
data/step-0107.parquetData45.0 MB305b0d57a703
data/step-0108.parquetData44.5 MB3ffaed41a422
data/step-0109.parquetData45.4 MBec67a7353682
data/step-0110.parquetData43.4 MB8ba42c6fc908
data/step-0111.parquetData43.7 MB8d37e43133bb
data/step-0112.parquetData45.7 MBf604739c831a
data/step-0113.parquetData43.2 MBc8a999b64ba6
data/step-0114.parquetData43.5 MB3a8997f62e3d
data/step-0115.parquetData43.9 MB1248cfed48f3
data/step-0116.parquetData44.8 MB3e864d9a0186
data/step-0117.parquetData44.9 MB4ba613747039
data/step-0118.parquetData42.7 MBaeb43c63daf8
data/step-0119.parquetData43.7 MBb7dd09be4489
data/step-0120.parquetData41.6 MBa7f269cdfa76
data/step-0121.parquetData44.6 MB202b9269d3a6
data/step-0122.parquetData43.2 MB3947a8f0eead
data/step-0123.parquetData43.4 MB46335c0f659f
data/step-0124.parquetData42.7 MB055e42c9ec12
data/step-0125.parquetData42.4 MBe9cdfa6ccbbb
data/step-0126.parquetData42.3 MBa97661465b93
data/step-0127.parquetData42.2 MB15ad3786299d
data/step-0128.parquetData42.8 MBde89d4c15f9d
data/step-0129.parquetData43.0 MBadb240322bc4
data/step-0130.parquetData41.7 MB0646a43c2c5d
data/step-0131.parquetData42.9 MB77f8562f48f3
data/step-0132.parquetData44.7 MB64612aaf8594
data/step-0133.parquetData43.1 MBede3738ff431
data/step-0134.parquetData41.3 MB3258975db513
data/step-0135.parquetData43.4 MBef48b325087b
data/step-0136.parquetData41.8 MB057f95c94038
data/step-0137.parquetData41.4 MBf12eb1bc5816
data/step-0138.parquetData43.4 MBc4e2f6fbeacd
data/step-0139.parquetData43.2 MBc99fbc9d68b8
data/step-0140.parquetData41.5 MB8e8172c1fd0c
data/step-0141.parquetData42.1 MB1475640edcac
data/step-0142.parquetData41.2 MB22175092f716
data/step-0143.parquetData40.1 MB2d17f0bb01f5
data/step-0144.parquetData39.6 MBc2fd2a010161
data/step-0145.parquetData40.1 MB31fd43dc257c
data/step-0146.parquetData41.1 MB49e6327fbfcd
data/step-0147.parquetData39.0 MB094b4a9301c3
data/step-0148.parquetData40.4 MBa109c71eabaa
data/step-0149.parquetData41.6 MB0d7ccbee1de9
data/step-0150.parquetData41.3 MBfbf6c367123f
data/step-0151.parquetData40.5 MB2f3356dabc37
data/step-0152.parquetData40.8 MB239bf3bc8c86
data/step-0153.parquetData40.1 MB293f692906a4
data/step-0154.parquetData41.6 MBd9df992961e5
data/step-0155.parquetData38.7 MB973bdd792696
data/step-0156.parquetData38.4 MB1a305b542ecc
data/step-0157.parquetData40.1 MB64026377ec4e
data/step-0158.parquetData40.8 MBddc627708f17
data/step-0159.parquetData41.0 MB1cc5a7448724
data/step-0160.parquetData37.3 MBc52f65fa8e7d
data/step-0161.parquetData40.8 MBefeafa783274
data/step-0162.parquetData40.5 MB4c80e09f02b1
data/step-0163.parquetData39.2 MB26aef6e6573c
data/step-0164.parquetData40.5 MB1000e359300a
data/step-0165.parquetData40.1 MB476d764b4070
data/step-0166.parquetData40.7 MBfa626258e6cd
data/step-0167.parquetData41.6 MB47fb1a06fa67
data/step-0168.parquetData40.8 MBce1c7ee835b4
data/step-0169.parquetData40.7 MB518a3e3aa0c3
data/step-0170.parquetData39.6 MB0c8b95a94840
data/step-0171.parquetData41.9 MB9c49d1a5368d
data/step-0172.parquetData40.8 MB793170189b79
data/step-0173.parquetData39.3 MB5c387baff6b5
data/step-0174.parquetData41.1 MBd9943881ddf1
data/step-0175.parquetData40.2 MBdf653132a62a
data/step-0176.parquetData40.5 MBcb30da6b86d4
data/step-0177.parquetData40.4 MBfea173d7bba9
data/step-0178.parquetData39.9 MB0bee808eabd7
data/step-0179.parquetData39.0 MBb020cd4de4f4
data/step-0180.parquetData37.9 MB544f67d0a4e0
data/step-0181.parquetData39.6 MB00cddc1df746
data/step-0182.parquetData40.4 MB2e39fb9e5248
data/step-0183.parquetData38.7 MBa19fad536857
data/step-0184.parquetData38.6 MBfe8b7a339818
data/step-0185.parquetData36.7 MBd92a87755971
data/step-0186.parquetData38.9 MB258c3709f1c0
data/step-0187.parquetData38.9 MB8bda3fc24130
data/step-0188.parquetData38.2 MBf49d846873e9
data/step-0189.parquetData40.0 MB650b4de1d973
data/step-0190.parquetData38.7 MBb2bb6e89c1b5
data/step-0191.parquetData38.4 MBf995a2f70176
data/step-0192.parquetData38.3 MB69e4e0262a8c
data/step-0193.parquetData38.6 MBbfe3cea7c21d
data/step-0194.parquetData39.1 MB4748f8b8a7d3
data/step-0195.parquetData39.9 MBdd808f5563ac
data/step-0196.parquetData39.2 MB2daf3d6fb508
data/step-0197.parquetData40.1 MB4aa5c1d8f23f
data/step-0198.parquetData38.1 MB8d8050b1db28
data/step-0199.parquetData40.4 MB0b1045c63d6c
data/step-0200.parquetData38.5 MB4612463d469d
data/step-0201.parquetData39.3 MBcc53e459cb3b
data/step-0202.parquetData39.2 MB190db890a15b
data/step-0203.parquetData36.9 MBcf07372a03c1
data/step-0204.parquetData39.3 MBd6d783b5c22a
data/step-0205.parquetData37.9 MBd29f421b94b6
data/step-0206.parquetData38.1 MB60aa8946b784
data/step-0207.parquetData37.1 MBc254dd996815
data/step-0208.parquetData35.9 MB3edf25afaf3f
data/step-0209.parquetData36.2 MBa336bc5b450f
data/step-0210.parquetData35.7 MBa4244c45f890
data/step-0211.parquetData35.5 MBa7c85284b79a
data/step-0212.parquetData34.0 MB85ab330c8a5f
data/step-0213.parquetData35.9 MB357e37b0966d
data/step-0214.parquetData34.3 MB982df1be916f
data/step-0215.parquetData35.1 MB5a2baee721be
data/step-0216.parquetData33.4 MBc8d78aa3e138
data/step-0217.parquetData33.9 MBa60f8449eaa0
data/step-0218.parquetData33.0 MB2d32fc16b70d
data/step-0219.parquetData32.4 MB8574b9d253e5
data/step-0220.parquetData33.0 MBeaf36ca26829
data/step-0221.parquetData32.8 MBacf199012129
data/step-0222.parquetData33.7 MB26964a528b8e
data/step-0223.parquetData33.0 MB03b7c7c17ec6
data/step-0224.parquetData33.3 MBeabf05e080b1
data/step-0225.parquetData33.9 MBc16bd9d4c14f
data/step-0226.parquetData34.6 MB149e072e0770
data/step-0227.parquetData33.1 MBce989465af9d
data/step-0228.parquetData32.1 MB7ead2ff9d4e8
data/step-0229.parquetData32.3 MBe8afbcccd34a
data/step-0230.parquetData33.4 MBb20946670291
data/step-0231.parquetData31.2 MB7c8e40481768
data/step-0232.parquetData32.0 MB33246d780920
data/step-0233.parquetData33.6 MB3fef30d16f3f
data/step-0234.parquetData34.4 MB550005197ea3
data/step-0235.parquetData32.9 MBd0d6dead9708
data/step-0236.parquetData32.6 MB837b6f2a5557
data/step-0237.parquetData32.4 MBde24f1fc7c95
data/step-0238.parquetData32.7 MBecf427946a73
data/step-0239.parquetData32.8 MBce22b27bd5d2
data/step-0240.parquetData31.8 MB58706a449069
data/step-0241.parquetData31.8 MB984386584e6d
data/step-0242.parquetData32.2 MBf48c8d8a9d7a
data/step-0243.parquetData30.6 MB10f4b97bed43
data/step-0244.parquetData31.5 MBccceb33847b3
data/step-0245.parquetData33.0 MBe12e568559dc
data/step-0246.parquetData32.6 MB2f8dc9646ab0
data/step-0247.parquetData31.3 MB05d8e39c83e7
data/step-0248.parquetData31.7 MB7c8b113a8bc6
data/step-0249.parquetData32.2 MBd791996d7fd6
data/step-0250.parquetData32.2 MB39d2ef38897f
data/step-0251.parquetData32.7 MBae31477e329b
data/step-0252.parquetData33.1 MBa5ce7b46f063
data/step-0253.parquetData31.1 MBa06d98b15a52
data/step-0254.parquetData32.3 MB4cc8da976a19
data/step-0255.parquetData33.2 MBdc7089f2f571
data/step-0256.parquetData32.3 MBa887722056c8
data/step-0257.parquetData32.7 MB092a1c10d309
README.mdDocumentation597 B—
.gitattributesRepository2.5 KB—

License and Download

License
Not stated by the source
Access
No access gate
Download from Fahim Tajwar

Released by Fahim Tajwar through its official repository on Hugging Face.