Skip to page content
SAVRN
Search Plan a project

Dataset

maxrl_dppo_qwen3_4B_base_knk_rollouts

by Fahim Tajwar ftajwar/maxrl_dppo_qwen3_4B_base_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.

Rows1,318,912
Configurations1
Size12.5 GB
License—
AccessPublicly accessible
Monthly Downloads458

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 1,318,912 rows

SplitRowsSize
train1,318,91233.9 GB
inputstringoutputstringgtsstringscorefloat64stepint64accfloat64prompt_token_idslistresponse_token_idslistrollout_log_probslistuidstring

Details

Repository
ftajwar/maxrl_dppo_qwen3_4B_base_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
021dc87bd7c54b38f5be7caf25407d7fd7fe6758
Last updated
2026-10-08

Files

324 files, 12.5 GB in total.

Data322 files · 12.5 GB
Documentation1 file · 597 B
Repository1 file · 2.5 KB
Every file
FileTypeSizeSHA-256
data/step-0001.parquetData41.1 MB44074b330435
data/step-0002.parquetData42.4 MB0edb15648a1e
data/step-0003.parquetData40.4 MB2551ff68b8e2
data/step-0004.parquetData40.9 MB38425b3aa40c
data/step-0005.parquetData41.0 MBa7f4571f80e4
data/step-0006.parquetData38.9 MB0eaf2d5eb2a2
data/step-0007.parquetData38.1 MBb964b37af362
data/step-0008.parquetData37.1 MB1752bf22ffad
data/step-0009.parquetData38.1 MB8c424ac04588
data/step-0010.parquetData37.0 MBd239d7dbc955
data/step-0011.parquetData34.8 MBabd89a375904
data/step-0012.parquetData31.3 MBb9f8c4398daf
data/step-0013.parquetData33.8 MBc1c44ae2075b
data/step-0014.parquetData33.9 MB285c9f11ee5d
data/step-0015.parquetData33.4 MB9f0cb4e9738c
data/step-0016.parquetData33.4 MBc45718760eb9
data/step-0017.parquetData33.2 MB5607a2e20399
data/step-0018.parquetData32.4 MBdc85a5fdc231
data/step-0019.parquetData32.4 MB806ac153580b
data/step-0020.parquetData33.8 MB14e7bee5a4c5
data/step-0021.parquetData34.2 MB5af2fb131d1b
data/step-0022.parquetData33.2 MBf2eee28816e7
data/step-0023.parquetData33.7 MB03932208b03e
data/step-0024.parquetData33.6 MBb8ca7d3dfac5
data/step-0025.parquetData35.0 MB1a375f28822f
data/step-0026.parquetData36.9 MB9c53254619b8
data/step-0027.parquetData36.2 MBb855d6a6cc2e
data/step-0028.parquetData34.5 MBb21cb526b799
data/step-0029.parquetData37.5 MBfc033bc2ac8c
data/step-0030.parquetData37.3 MBe3587dd5a5af
data/step-0031.parquetData36.4 MB90399404b7a6
data/step-0032.parquetData39.7 MB139b15ae61dd
data/step-0033.parquetData38.1 MBf8f6fd498ef6
data/step-0034.parquetData38.9 MB704163f76e9d
data/step-0035.parquetData39.9 MBe0bf12380534
data/step-0036.parquetData40.2 MB654166afbbf5
data/step-0037.parquetData40.8 MB122c85fca6ab
data/step-0038.parquetData42.8 MBb9f2e38cb7ae
data/step-0039.parquetData43.3 MB86f1ba91bc07
data/step-0040.parquetData42.0 MBe7a9339c919e
data/step-0041.parquetData45.2 MB759dc099f064
data/step-0042.parquetData46.5 MB314893e5f826
data/step-0043.parquetData46.8 MBeb9646ae2580
data/step-0044.parquetData47.2 MB638691acfc3a
data/step-0045.parquetData48.0 MB3f5ac61e7b56
data/step-0046.parquetData46.8 MB821b69e0e5f4
data/step-0047.parquetData47.8 MB8d506056330e
data/step-0048.parquetData49.4 MB2a02a2fdcf65
data/step-0049.parquetData50.3 MBe57c1888fe13
data/step-0050.parquetData52.2 MB6f7f4fb3eb0d
data/step-0051.parquetData51.3 MB6ecb935525cc
data/step-0052.parquetData52.5 MBaff5e9f2cd1f
data/step-0053.parquetData56.2 MB9500fce713f4
data/step-0054.parquetData54.5 MBcc9d699eaf29
data/step-0055.parquetData54.1 MB66ec70712ded
data/step-0056.parquetData53.4 MB1fe43fa0b811
data/step-0057.parquetData54.0 MB03bea55b823d
data/step-0058.parquetData51.3 MBd850b05c4935
data/step-0059.parquetData52.6 MB83487cadb94c
data/step-0060.parquetData54.2 MB5fbb79aa1ffe
data/step-0061.parquetData54.7 MB7db49aa10d0d
data/step-0062.parquetData51.4 MBdd80d0cc3079
data/step-0063.parquetData54.8 MB8850607f09b1
data/step-0064.parquetData56.7 MB95178c2f16b0
data/step-0065.parquetData54.5 MBf3094f595db3
data/step-0066.parquetData53.2 MB745d3946707c
data/step-0067.parquetData55.3 MBd743af6888b0
data/step-0068.parquetData54.0 MBc21f14e2ad92
data/step-0069.parquetData51.4 MB5a00b32bb387
data/step-0070.parquetData54.1 MBe0991e4bbd43
data/step-0071.parquetData54.3 MB269d1c2875bd
data/step-0072.parquetData52.5 MB3215f3fa3ee4
data/step-0073.parquetData52.0 MB66814cc76f4d
data/step-0074.parquetData53.6 MB2445efb6b14f
data/step-0075.parquetData52.0 MBdf58a4c185c5
data/step-0076.parquetData50.6 MB01f689f80bbf
data/step-0077.parquetData49.3 MB83470c421dae
data/step-0078.parquetData50.4 MBc3bcbd848ad5
data/step-0079.parquetData51.4 MB5276bcfff7a0
data/step-0080.parquetData52.0 MB9045677b8613
data/step-0081.parquetData52.3 MB4f290503a261
data/step-0082.parquetData51.4 MBc16f0788fb55
data/step-0083.parquetData53.3 MB3ce5c81ce389
data/step-0084.parquetData52.6 MB19ae1c4d2ab1
data/step-0085.parquetData51.0 MBd4ca955ac041
data/step-0086.parquetData52.1 MBc3800d0c419a
data/step-0087.parquetData50.8 MBc3565ad24faf
data/step-0088.parquetData49.7 MB03d77a9fb8f6
data/step-0089.parquetData52.5 MB2e179b9f9c6f
data/step-0090.parquetData53.2 MB5cdd353a89cd
data/step-0091.parquetData50.7 MB5cc301c181f0
data/step-0092.parquetData51.3 MB0e79532a1276
data/step-0093.parquetData50.8 MB3d31ae5c01c0
data/step-0094.parquetData49.5 MBf792892b772c
data/step-0095.parquetData51.9 MBa7be8541c79a
data/step-0096.parquetData51.0 MB0e07ef2ff01b
data/step-0097.parquetData51.1 MB3d55c19f0235
data/step-0098.parquetData48.2 MB7e9cbf5b3f7f
data/step-0099.parquetData47.6 MB3205d63b6585
data/step-0100.parquetData47.1 MB5dfcc58ac9b3
data/step-0101.parquetData48.5 MB5b5900bc4e12
data/step-0102.parquetData47.3 MB55c7b0fa9161
data/step-0103.parquetData47.3 MBa76d45956f2e
data/step-0104.parquetData48.6 MB18c1087c6a44
data/step-0105.parquetData47.4 MBbed302715b91
data/step-0106.parquetData46.6 MB377a6267a8b1
data/step-0107.parquetData46.4 MB5b9f3dbe3d6b
data/step-0108.parquetData46.1 MBabce7c4b5aed
data/step-0109.parquetData48.4 MB6420d41e88c0
data/step-0110.parquetData44.6 MB362961c8761e
data/step-0111.parquetData46.1 MB0bb245563234
data/step-0112.parquetData48.8 MBcb3339cdbcef
data/step-0113.parquetData46.5 MB6044baec46c3
data/step-0114.parquetData46.3 MB6596f7bdc8be
data/step-0115.parquetData46.9 MB92b1c9eed974
data/step-0116.parquetData47.8 MB915209ccd5bc
data/step-0117.parquetData47.9 MB0c7974542d6e
data/step-0118.parquetData45.9 MBabde7b15bb42
data/step-0119.parquetData46.2 MBf41189a2b5ea
data/step-0120.parquetData44.1 MBfbaf99c46461
data/step-0121.parquetData48.5 MB64802397b0cf
data/step-0122.parquetData46.8 MBb978625d4c4f
data/step-0123.parquetData46.6 MB7093fb1e9062
data/step-0124.parquetData46.1 MB7b8fe0c71647
data/step-0125.parquetData45.2 MBf2e51b70e083
data/step-0126.parquetData43.6 MBb0d2da9808b8
data/step-0127.parquetData45.0 MB132c0d4aa2cc
data/step-0128.parquetData44.3 MB899c26e26e73
data/step-0129.parquetData45.3 MB279b39364809
data/step-0130.parquetData43.0 MBd7ef2fbe885e
data/step-0131.parquetData43.8 MB23a7dfbd2b0a
data/step-0132.parquetData45.7 MB5fc0a9abf627
data/step-0133.parquetData42.7 MBdb4296339ddf
data/step-0134.parquetData40.8 MB2cfd2d0cd2a5
data/step-0135.parquetData44.2 MB0be4c35fe982
data/step-0136.parquetData41.9 MB0cc12735d9b0
data/step-0137.parquetData41.5 MB63da46a13542
data/step-0138.parquetData43.0 MBc92ca7192543
data/step-0139.parquetData43.0 MB6bc29234ee42
data/step-0140.parquetData41.2 MBb5bb42f6cc3d
data/step-0141.parquetData43.2 MBcbcebb1cd92c
data/step-0142.parquetData41.6 MBbcfef7107f0b
data/step-0143.parquetData40.6 MB29b81971a90b
data/step-0144.parquetData40.0 MB3db0998718fa
data/step-0145.parquetData40.2 MB92e8a9767aae
data/step-0146.parquetData41.0 MB0e80f05da5e0
data/step-0147.parquetData39.9 MB5d685b2031bc
data/step-0148.parquetData41.0 MBeaa7597ae28e
data/step-0149.parquetData43.1 MB17d830431b58
data/step-0150.parquetData42.3 MBf26a9473d680
data/step-0151.parquetData41.8 MB11a64f35b718
data/step-0152.parquetData41.9 MBd1e493c7a217
data/step-0153.parquetData40.4 MBccaa7c18ee58
data/step-0154.parquetData42.6 MBf5c63d30e829
data/step-0155.parquetData40.0 MBa6ad8ef42ece
data/step-0156.parquetData40.6 MB03d0b272c059
data/step-0157.parquetData42.2 MB90d944cee7b5
data/step-0158.parquetData42.3 MBd6c612af726d
data/step-0159.parquetData43.7 MB8bc874e5b241
data/step-0160.parquetData39.5 MB2c6c4211a86d
data/step-0161.parquetData44.6 MB96c76ae0fdcc
data/step-0162.parquetData42.2 MB037e3bd52b13
data/step-0163.parquetData42.3 MB02e001c34c79
data/step-0164.parquetData44.0 MB46f0f88f0fc1
data/step-0165.parquetData42.2 MB51b3104e1e5d
data/step-0166.parquetData43.9 MBfee8cc01b06a
data/step-0167.parquetData44.4 MB7ca53d02d466
data/step-0168.parquetData43.4 MBc0eac0d1f276
data/step-0169.parquetData42.3 MB7c0b2ceda221
data/step-0170.parquetData41.6 MB51a12722051a
data/step-0171.parquetData43.5 MB176f98315484
data/step-0172.parquetData42.6 MBce59074bc691
data/step-0173.parquetData40.4 MBea263cf3af0c
data/step-0174.parquetData43.0 MBb3c1ef3fea69
data/step-0175.parquetData41.1 MB0083891cd572
data/step-0176.parquetData41.3 MB3e532217e80e
data/step-0177.parquetData41.1 MB595b905c43f0
data/step-0178.parquetData40.9 MBb3dc2561ef1d
data/step-0179.parquetData40.2 MBe8ee85ab3e7d
data/step-0180.parquetData37.8 MB79e7d26b815e
data/step-0181.parquetData40.3 MB87a1cc49710b
data/step-0182.parquetData41.2 MB475dcc052edc
data/step-0183.parquetData38.6 MB39ef6753bfbc
data/step-0184.parquetData37.9 MB2aaa433f6058
data/step-0185.parquetData36.2 MB062c8ad45737
data/step-0186.parquetData37.7 MBf0db35f5dfa4
data/step-0187.parquetData37.6 MB2c67fb3abd9c
data/step-0188.parquetData36.6 MB5ef90025847a
data/step-0189.parquetData39.0 MB2da569684b10
data/step-0190.parquetData35.8 MB691c37b91ac2
data/step-0191.parquetData37.3 MBfd3edbfcaa8a
data/step-0192.parquetData36.1 MB110e22153627
data/step-0193.parquetData36.0 MB74e4c0c9473d
data/step-0194.parquetData36.4 MB000949dd3d34
data/step-0195.parquetData37.9 MB72d0abe0e3ec
data/step-0196.parquetData37.1 MB5613ee36f47c
data/step-0197.parquetData37.6 MBc9e9003b98d3
data/step-0198.parquetData35.1 MB42af00674c35
data/step-0199.parquetData36.9 MB2ae4cc901be3
data/step-0200.parquetData35.3 MB2fb170b2ed9b
data/step-0201.parquetData35.6 MBe3cd961c42c7
data/step-0202.parquetData36.0 MBc7254a829dcd
data/step-0203.parquetData34.1 MBf00b1031aac9
data/step-0204.parquetData35.8 MB9346d8505d9c
data/step-0205.parquetData34.2 MB767cea5d5eef
data/step-0206.parquetData34.9 MB465391fb6c33
data/step-0207.parquetData34.9 MBfd1d2ae4d763
data/step-0208.parquetData33.2 MB5f0806193517
data/step-0209.parquetData33.7 MB1d978ffd9e4f
data/step-0210.parquetData33.9 MB4d8b62218029
data/step-0211.parquetData33.3 MBdef86c1311ed
data/step-0212.parquetData32.3 MB3ef13dd9e5a9
data/step-0213.parquetData34.4 MB3d02bcae8e91
data/step-0214.parquetData33.5 MBd8d14985dc5b
data/step-0215.parquetData33.7 MB8e4e0b5d619d
data/step-0216.parquetData32.7 MB8103014d6f8b
data/step-0217.parquetData33.3 MBf59fed20ebbf
data/step-0218.parquetData32.4 MB99a352dadbfc
data/step-0219.parquetData31.9 MBbb2cadaf9a13
data/step-0220.parquetData32.4 MB1a211205e762
data/step-0221.parquetData32.2 MB48874d77407f
data/step-0222.parquetData32.7 MB920a69dc0497
data/step-0223.parquetData32.0 MBd48482cb394d
data/step-0224.parquetData32.2 MBd65a6e9a49a5
data/step-0225.parquetData32.9 MB5652d217fb67
data/step-0226.parquetData33.3 MBf5e3357dd05c
data/step-0227.parquetData32.3 MB39f08720f3ed
data/step-0228.parquetData30.5 MBfb374e85b3a9
data/step-0229.parquetData31.4 MB9ed8e3d905e4
data/step-0230.parquetData32.5 MBb7523c51ad2f
data/step-0231.parquetData30.3 MB0708ee69eb7b
data/step-0232.parquetData31.0 MB76f61bff476a
data/step-0233.parquetData32.9 MBc0dba98a21c8
data/step-0234.parquetData33.5 MB6688c556f5d3
data/step-0235.parquetData32.2 MB254986025945
data/step-0236.parquetData32.0 MBcbea699160f6
data/step-0237.parquetData31.8 MB2780ec7b4f58
data/step-0238.parquetData31.7 MB2d87221b62de
data/step-0239.parquetData31.8 MB1ba28cc6d5c3
data/step-0240.parquetData31.6 MB03bbb5f7ff4c
data/step-0241.parquetData31.3 MB283d4bd89ff2
data/step-0242.parquetData32.2 MBe04a5c491a2c
data/step-0243.parquetData30.8 MBb6a93a112d94
data/step-0244.parquetData31.7 MBbbd2cf2a2489
data/step-0245.parquetData33.0 MBd76defe4d7b6
data/step-0246.parquetData32.7 MB5f3aff2d56db
data/step-0247.parquetData31.1 MB407aad52f3bb
data/step-0248.parquetData31.5 MBb0895cecf246
data/step-0249.parquetData32.1 MB58af7bc3bad4
data/step-0250.parquetData31.6 MB856179731acc
data/step-0251.parquetData31.9 MB5b439db1d54c
data/step-0252.parquetData32.4 MBe9b7eb2f9353
data/step-0253.parquetData29.4 MB620de15f0aba
data/step-0254.parquetData30.8 MBdec52a691cb6
data/step-0255.parquetData31.7 MB1addd3d6e860
data/step-0256.parquetData30.5 MBd61a688d568a
data/step-0257.parquetData30.8 MBe161beac48e5
data/step-0258.parquetData30.8 MBf7a5ab2a7de2
data/step-0259.parquetData30.4 MBccb86537bd0c
data/step-0260.parquetData30.9 MB2edbdb3d7b2b
data/step-0261.parquetData30.9 MBbded8e2778d3
data/step-0262.parquetData29.8 MB4fbf16f7f14a
data/step-0263.parquetData30.3 MB623755e9fc89
data/step-0264.parquetData28.9 MB6ddafca7a557
data/step-0265.parquetData30.6 MB9183ac177232
data/step-0266.parquetData30.4 MB84cfef11a1ff
data/step-0267.parquetData29.5 MB4958482bdc35
data/step-0268.parquetData28.5 MB94603830bb47
data/step-0269.parquetData28.9 MB03bca42aa4b9
data/step-0270.parquetData29.7 MB1299e68f1e0e
data/step-0271.parquetData29.3 MB34ceb57e762a
data/step-0272.parquetData31.3 MB105b57b8c5c9
data/step-0273.parquetData30.8 MB6aae2750aa31
data/step-0274.parquetData30.4 MB200d957d1bd4
data/step-0275.parquetData32.4 MBa9d704edddf3
data/step-0276.parquetData29.2 MB519b7fdefedf
data/step-0277.parquetData31.0 MBb8dab5dd0020
data/step-0278.parquetData30.1 MB3dfb6595bbb6
data/step-0279.parquetData28.5 MB8e1064b9b020
data/step-0280.parquetData30.9 MBa9613d1085a6
data/step-0281.parquetData28.6 MB6287a9245a36
data/step-0282.parquetData29.7 MB11378d7e0f57
data/step-0283.parquetData30.1 MBbabde6eaa79c
data/step-0284.parquetData29.1 MB29658e29eec7
data/step-0285.parquetData30.4 MBa44b38e9fc50
data/step-0286.parquetData30.0 MBcbdfe9ac29bc
data/step-0287.parquetData29.8 MB7264c797ca64
data/step-0288.parquetData29.4 MB99c65cb3a066
data/step-0289.parquetData30.0 MBe3f0802185d1
data/step-0290.parquetData29.8 MB80d14dd936b1
data/step-0291.parquetData29.1 MBbf58dd803370
data/step-0292.parquetData30.6 MBdd7bcda0efb1
data/step-0293.parquetData31.3 MBd61aa419fc0e
data/step-0294.parquetData29.7 MBca707e4ea310
data/step-0295.parquetData29.1 MB1efac4e756cd
data/step-0296.parquetData29.2 MB44fe66f64bac
data/step-0297.parquetData29.8 MB70a10ccd3283
data/step-0298.parquetData29.6 MB248ccceaf49a
data/step-0299.parquetData28.4 MB3bb26214381b
data/step-0300.parquetData28.4 MB0134cac2fdb3
data/step-0301.parquetData28.8 MB75c5fa080cc4
data/step-0302.parquetData25.8 MB4fe9a553392d
data/step-0303.parquetData28.3 MB5355217f7821
data/step-0304.parquetData27.8 MB842e9cfddc9a
data/step-0305.parquetData27.9 MBc69ab7b47544
data/step-0306.parquetData28.1 MB3e882eeaf405
data/step-0307.parquetData27.2 MB3142265af691
data/step-0308.parquetData27.0 MB0fae1c9f3d7d
data/step-0309.parquetData27.6 MB23a1897b5cb8
data/step-0310.parquetData27.2 MB74a445a08d63
data/step-0311.parquetData26.9 MBb51cd50fa6e3
data/step-0312.parquetData26.9 MBa3c2e982e479
data/step-0313.parquetData26.3 MB0c3078cb4d22
data/step-0314.parquetData25.4 MB965177724f54
data/step-0315.parquetData25.6 MBabf3890e0fce
data/step-0316.parquetData25.6 MBf2f47968138d
data/step-0317.parquetData25.9 MB273db11158d3
data/step-0318.parquetData26.2 MBe6e8072ee77d
data/step-0319.parquetData24.4 MBef645eb7de82
data/step-0320.parquetData26.0 MBac628d2635bf
data/step-0321.parquetData25.4 MB396e7102ba44
data/step-0322.parquetData27.1 MBa738b6e5c68c
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.