Files
Aris-140M/last-checkpoint/trainer_state.json
ModelHub XC 17a7eb0a28 初始化项目,由ModelHub XC社区提供模型
Model: joseph-ai/Aris-140M
Source: Original Platform
2026-10-03 19:24:36 +08:00

5504 lines
143 KiB
JSON

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 7630,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.001310615989515072,
"grad_norm": 4.797830104827881,
"learning_rate": 1.1249999999999999e-05,
"loss": 10.801536560058594,
"step": 10
},
{
"epoch": 0.002621231979030144,
"grad_norm": 2.243802070617676,
"learning_rate": 2.375e-05,
"loss": 10.004041290283203,
"step": 20
},
{
"epoch": 0.003931847968545216,
"grad_norm": 1.9459115266799927,
"learning_rate": 3.625e-05,
"loss": 9.504635620117188,
"step": 30
},
{
"epoch": 0.005242463958060288,
"grad_norm": 1.8689903020858765,
"learning_rate": 4.8750000000000006e-05,
"loss": 9.19006805419922,
"step": 40
},
{
"epoch": 0.00655307994757536,
"grad_norm": 1.6034637689590454,
"learning_rate": 6.125e-05,
"loss": 8.801864624023438,
"step": 50
},
{
"epoch": 0.007863695937090432,
"grad_norm": 1.4121754169464111,
"learning_rate": 7.375e-05,
"loss": 8.340190124511718,
"step": 60
},
{
"epoch": 0.009174311926605505,
"grad_norm": 1.2778944969177246,
"learning_rate": 8.625e-05,
"loss": 7.885487365722656,
"step": 70
},
{
"epoch": 0.010484927916120577,
"grad_norm": 0.8538752794265747,
"learning_rate": 9.875e-05,
"loss": 7.52657470703125,
"step": 80
},
{
"epoch": 0.011795543905635648,
"grad_norm": 0.6117246150970459,
"learning_rate": 0.00011125000000000001,
"loss": 7.2668617248535154,
"step": 90
},
{
"epoch": 0.01310615989515072,
"grad_norm": 0.8442140817642212,
"learning_rate": 0.00012375,
"loss": 7.108723449707031,
"step": 100
},
{
"epoch": 0.014416775884665793,
"grad_norm": 0.849466860294342,
"learning_rate": 0.00013625,
"loss": 6.939354705810547,
"step": 110
},
{
"epoch": 0.015727391874180863,
"grad_norm": 1.1190710067749023,
"learning_rate": 0.00014874999999999998,
"loss": 6.787942504882812,
"step": 120
},
{
"epoch": 0.01703800786369594,
"grad_norm": 1.656615972518921,
"learning_rate": 0.00016125000000000002,
"loss": 6.669708251953125,
"step": 130
},
{
"epoch": 0.01834862385321101,
"grad_norm": 0.9534133672714233,
"learning_rate": 0.00017375,
"loss": 6.547358703613281,
"step": 140
},
{
"epoch": 0.019659239842726082,
"grad_norm": 1.0048075914382935,
"learning_rate": 0.00018625,
"loss": 6.438671875,
"step": 150
},
{
"epoch": 0.020969855832241154,
"grad_norm": 0.915958821773529,
"learning_rate": 0.00019875,
"loss": 6.329353332519531,
"step": 160
},
{
"epoch": 0.022280471821756225,
"grad_norm": 1.0014761686325073,
"learning_rate": 0.00021124999999999998,
"loss": 6.273650360107422,
"step": 170
},
{
"epoch": 0.023591087811271297,
"grad_norm": 1.1724601984024048,
"learning_rate": 0.00022375000000000002,
"loss": 6.156678771972656,
"step": 180
},
{
"epoch": 0.02490170380078637,
"grad_norm": 0.9091385006904602,
"learning_rate": 0.00023625,
"loss": 6.108370208740235,
"step": 190
},
{
"epoch": 0.02621231979030144,
"grad_norm": 1.2376604080200195,
"learning_rate": 0.00024875,
"loss": 6.0070655822753904,
"step": 200
},
{
"epoch": 0.027522935779816515,
"grad_norm": 0.8809821009635925,
"learning_rate": 0.00024999909492070475,
"loss": 5.950724029541016,
"step": 210
},
{
"epoch": 0.028833551769331587,
"grad_norm": 1.0397638082504272,
"learning_rate": 0.0002499959662683632,
"loss": 5.917636871337891,
"step": 220
},
{
"epoch": 0.03014416775884666,
"grad_norm": 1.0171058177947998,
"learning_rate": 0.00024999060292507777,
"loss": 5.841743469238281,
"step": 230
},
{
"epoch": 0.03145478374836173,
"grad_norm": 0.8431490063667297,
"learning_rate": 0.00024998300498673494,
"loss": 5.791047668457031,
"step": 240
},
{
"epoch": 0.0327653997378768,
"grad_norm": 0.7999606132507324,
"learning_rate": 0.00024997317258917143,
"loss": 5.748330307006836,
"step": 250
},
{
"epoch": 0.03407601572739188,
"grad_norm": 0.8790252208709717,
"learning_rate": 0.00024996110590817187,
"loss": 5.715357971191406,
"step": 260
},
{
"epoch": 0.035386631716906945,
"grad_norm": 0.9629982709884644,
"learning_rate": 0.00024994680515946576,
"loss": 5.6543128967285154,
"step": 270
},
{
"epoch": 0.03669724770642202,
"grad_norm": 1.3806921243667603,
"learning_rate": 0.0002499302705987232,
"loss": 5.599337005615235,
"step": 280
},
{
"epoch": 0.03800786369593709,
"grad_norm": 1.2885777950286865,
"learning_rate": 0.0002499115025215509,
"loss": 5.585296630859375,
"step": 290
},
{
"epoch": 0.039318479685452164,
"grad_norm": 1.0204627513885498,
"learning_rate": 0.00024989050126348646,
"loss": 5.526985931396484,
"step": 300
},
{
"epoch": 0.04062909567496723,
"grad_norm": 0.985386312007904,
"learning_rate": 0.0002498672671999926,
"loss": 5.485442352294922,
"step": 310
},
{
"epoch": 0.04193971166448231,
"grad_norm": 0.7440592646598816,
"learning_rate": 0.00024984180074645037,
"loss": 5.49345588684082,
"step": 320
},
{
"epoch": 0.04325032765399738,
"grad_norm": 0.9439603686332703,
"learning_rate": 0.0002498141023581515,
"loss": 5.460093688964844,
"step": 330
},
{
"epoch": 0.04456094364351245,
"grad_norm": 0.8410868644714355,
"learning_rate": 0.00024978417253029096,
"loss": 5.404550170898437,
"step": 340
},
{
"epoch": 0.045871559633027525,
"grad_norm": 1.0195112228393555,
"learning_rate": 0.0002497520117979571,
"loss": 5.325064849853516,
"step": 350
},
{
"epoch": 0.047182175622542594,
"grad_norm": 0.8683975338935852,
"learning_rate": 0.000249717620736123,
"loss": 5.288446044921875,
"step": 360
},
{
"epoch": 0.04849279161205767,
"grad_norm": 0.7637448310852051,
"learning_rate": 0.00024968099995963556,
"loss": 5.294237518310547,
"step": 370
},
{
"epoch": 0.04980340760157274,
"grad_norm": 1.0485244989395142,
"learning_rate": 0.00024964215012320483,
"loss": 5.245368194580078,
"step": 380
},
{
"epoch": 0.05111402359108781,
"grad_norm": 1.0468641519546509,
"learning_rate": 0.00024960107192139236,
"loss": 5.211923599243164,
"step": 390
},
{
"epoch": 0.05242463958060288,
"grad_norm": 0.7256688475608826,
"learning_rate": 0.0002495577660885985,
"loss": 5.204988861083985,
"step": 400
},
{
"epoch": 0.053735255570117955,
"grad_norm": 0.9746586084365845,
"learning_rate": 0.00024951223339904935,
"loss": 5.174029541015625,
"step": 410
},
{
"epoch": 0.05504587155963303,
"grad_norm": 0.94045490026474,
"learning_rate": 0.00024946447466678317,
"loss": 5.139980316162109,
"step": 420
},
{
"epoch": 0.0563564875491481,
"grad_norm": 0.744636058807373,
"learning_rate": 0.0002494144907456356,
"loss": 5.113622665405273,
"step": 430
},
{
"epoch": 0.057667103538663174,
"grad_norm": 1.1388003826141357,
"learning_rate": 0.0002493622825292244,
"loss": 5.0717517852783205,
"step": 440
},
{
"epoch": 0.05897771952817824,
"grad_norm": 0.9612365961074829,
"learning_rate": 0.0002493078509509334,
"loss": 5.072021484375,
"step": 450
},
{
"epoch": 0.06028833551769332,
"grad_norm": 0.7358167171478271,
"learning_rate": 0.00024925119698389613,
"loss": 5.071805953979492,
"step": 460
},
{
"epoch": 0.061598951507208385,
"grad_norm": 0.8576675653457642,
"learning_rate": 0.0002491923216409782,
"loss": 5.0013587951660154,
"step": 470
},
{
"epoch": 0.06290956749672345,
"grad_norm": 0.7767074704170227,
"learning_rate": 0.000249131225974759,
"loss": 4.991025161743164,
"step": 480
},
{
"epoch": 0.06422018348623854,
"grad_norm": 0.8946435451507568,
"learning_rate": 0.00024906791107751335,
"loss": 4.941701889038086,
"step": 490
},
{
"epoch": 0.0655307994757536,
"grad_norm": 0.9179105758666992,
"learning_rate": 0.00024900237808119153,
"loss": 5.002605438232422,
"step": 500
},
{
"epoch": 0.0655307994757536,
"eval_loss": 4.933480262756348,
"eval_runtime": 0.6979,
"eval_samples_per_second": 183.404,
"eval_steps_per_second": 45.851,
"step": 500
},
{
"epoch": 0.06684141546526867,
"grad_norm": 0.7974295020103455,
"learning_rate": 0.00024893462815739933,
"loss": 4.954715728759766,
"step": 510
},
{
"epoch": 0.06815203145478375,
"grad_norm": 0.5864965915679932,
"learning_rate": 0.0002488646625173769,
"loss": 4.9075065612792965,
"step": 520
},
{
"epoch": 0.06946264744429882,
"grad_norm": 1.231337308883667,
"learning_rate": 0.00024879248241197735,
"loss": 4.884672164916992,
"step": 530
},
{
"epoch": 0.07077326343381389,
"grad_norm": 0.9584572911262512,
"learning_rate": 0.0002487180891316441,
"loss": 4.860324096679688,
"step": 540
},
{
"epoch": 0.07208387942332896,
"grad_norm": 0.7450931072235107,
"learning_rate": 0.000248641484006388,
"loss": 4.845925903320312,
"step": 550
},
{
"epoch": 0.07339449541284404,
"grad_norm": 0.9317758083343506,
"learning_rate": 0.0002485626684057635,
"loss": 4.836086273193359,
"step": 560
},
{
"epoch": 0.07470511140235911,
"grad_norm": 0.6534205675125122,
"learning_rate": 0.00024848164373884407,
"loss": 4.798537826538086,
"step": 570
},
{
"epoch": 0.07601572739187418,
"grad_norm": 0.590699315071106,
"learning_rate": 0.0002483984114541972,
"loss": 4.776090240478515,
"step": 580
},
{
"epoch": 0.07732634338138926,
"grad_norm": 0.6772918105125427,
"learning_rate": 0.0002483129730398585,
"loss": 4.76301155090332,
"step": 590
},
{
"epoch": 0.07863695937090433,
"grad_norm": 0.870791494846344,
"learning_rate": 0.00024822533002330467,
"loss": 4.717774200439453,
"step": 600
},
{
"epoch": 0.0799475753604194,
"grad_norm": 0.5128806233406067,
"learning_rate": 0.0002481354839714268,
"loss": 4.6943504333496096,
"step": 610
},
{
"epoch": 0.08125819134993446,
"grad_norm": 0.7424812912940979,
"learning_rate": 0.0002480434364905021,
"loss": 4.6645557403564455,
"step": 620
},
{
"epoch": 0.08256880733944955,
"grad_norm": 0.7780288457870483,
"learning_rate": 0.0002479491892261648,
"loss": 4.689810943603516,
"step": 630
},
{
"epoch": 0.08387942332896461,
"grad_norm": 0.9385794401168823,
"learning_rate": 0.0002478527438633776,
"loss": 4.710404586791992,
"step": 640
},
{
"epoch": 0.08519003931847968,
"grad_norm": 0.8205022215843201,
"learning_rate": 0.00024775410212640064,
"loss": 4.668232727050781,
"step": 650
},
{
"epoch": 0.08650065530799476,
"grad_norm": 0.7820428013801575,
"learning_rate": 0.00024765326577876116,
"loss": 4.626142120361328,
"step": 660
},
{
"epoch": 0.08781127129750983,
"grad_norm": 0.9351951479911804,
"learning_rate": 0.000247550236623222,
"loss": 4.5863594055175785,
"step": 670
},
{
"epoch": 0.0891218872870249,
"grad_norm": 0.7418632507324219,
"learning_rate": 0.0002474450165017492,
"loss": 4.617713928222656,
"step": 680
},
{
"epoch": 0.09043250327653997,
"grad_norm": 0.6272849440574646,
"learning_rate": 0.00024733760729547897,
"loss": 4.553762054443359,
"step": 690
},
{
"epoch": 0.09174311926605505,
"grad_norm": 0.7893446087837219,
"learning_rate": 0.0002472280109246844,
"loss": 4.534370422363281,
"step": 700
},
{
"epoch": 0.09305373525557012,
"grad_norm": 0.7383981943130493,
"learning_rate": 0.0002471162293487408,
"loss": 4.580450439453125,
"step": 710
},
{
"epoch": 0.09436435124508519,
"grad_norm": 0.6358956694602966,
"learning_rate": 0.00024700226456609095,
"loss": 4.499582672119141,
"step": 720
},
{
"epoch": 0.09567496723460026,
"grad_norm": 0.8162468075752258,
"learning_rate": 0.00024688611861420904,
"loss": 4.509400177001953,
"step": 730
},
{
"epoch": 0.09698558322411534,
"grad_norm": 0.7628506422042847,
"learning_rate": 0.00024676779356956456,
"loss": 4.491813659667969,
"step": 740
},
{
"epoch": 0.0982961992136304,
"grad_norm": 0.5897623300552368,
"learning_rate": 0.00024664729154758513,
"loss": 4.439973449707031,
"step": 750
},
{
"epoch": 0.09960681520314547,
"grad_norm": 0.6999703049659729,
"learning_rate": 0.0002465246147026183,
"loss": 4.486153411865234,
"step": 760
},
{
"epoch": 0.10091743119266056,
"grad_norm": 0.5613738894462585,
"learning_rate": 0.0002463997652278937,
"loss": 4.451129913330078,
"step": 770
},
{
"epoch": 0.10222804718217562,
"grad_norm": 0.8398756980895996,
"learning_rate": 0.0002462727453554832,
"loss": 4.443064117431641,
"step": 780
},
{
"epoch": 0.10353866317169069,
"grad_norm": 0.581312358379364,
"learning_rate": 0.0002461435573562613,
"loss": 4.395467376708984,
"step": 790
},
{
"epoch": 0.10484927916120576,
"grad_norm": 0.5368242263793945,
"learning_rate": 0.0002460122035398646,
"loss": 4.407203674316406,
"step": 800
},
{
"epoch": 0.10615989515072084,
"grad_norm": 0.6799978017807007,
"learning_rate": 0.0002458786862546504,
"loss": 4.35401496887207,
"step": 810
},
{
"epoch": 0.10747051114023591,
"grad_norm": 0.800698459148407,
"learning_rate": 0.0002457430078876546,
"loss": 4.374430465698242,
"step": 820
},
{
"epoch": 0.10878112712975098,
"grad_norm": 0.6181208491325378,
"learning_rate": 0.0002456051708645492,
"loss": 4.375891494750976,
"step": 830
},
{
"epoch": 0.11009174311926606,
"grad_norm": 0.5560409426689148,
"learning_rate": 0.00024546517764959897,
"loss": 4.339243316650391,
"step": 840
},
{
"epoch": 0.11140235910878113,
"grad_norm": 0.7018496990203857,
"learning_rate": 0.00024532303074561724,
"loss": 4.345343780517578,
"step": 850
},
{
"epoch": 0.1127129750982962,
"grad_norm": 0.7125679850578308,
"learning_rate": 0.00024517873269392114,
"loss": 4.343055343627929,
"step": 860
},
{
"epoch": 0.11402359108781127,
"grad_norm": 0.7190083861351013,
"learning_rate": 0.0002450322860742864,
"loss": 4.2893321990966795,
"step": 870
},
{
"epoch": 0.11533420707732635,
"grad_norm": 0.5694275498390198,
"learning_rate": 0.0002448836935049009,
"loss": 4.3499504089355465,
"step": 880
},
{
"epoch": 0.11664482306684142,
"grad_norm": 0.598127543926239,
"learning_rate": 0.00024473295764231815,
"loss": 4.2917427062988285,
"step": 890
},
{
"epoch": 0.11795543905635648,
"grad_norm": 0.7066140174865723,
"learning_rate": 0.0002445800811814097,
"loss": 4.278008270263672,
"step": 900
},
{
"epoch": 0.11926605504587157,
"grad_norm": 0.5317610502243042,
"learning_rate": 0.0002444250668553168,
"loss": 4.244252014160156,
"step": 910
},
{
"epoch": 0.12057667103538663,
"grad_norm": 0.5876745581626892,
"learning_rate": 0.0002442679174354018,
"loss": 4.253451538085938,
"step": 920
},
{
"epoch": 0.1218872870249017,
"grad_norm": 0.5453299283981323,
"learning_rate": 0.00024410863573119847,
"loss": 4.218951416015625,
"step": 930
},
{
"epoch": 0.12319790301441677,
"grad_norm": 0.5818265676498413,
"learning_rate": 0.00024394722459036177,
"loss": 4.227011871337891,
"step": 940
},
{
"epoch": 0.12450851900393185,
"grad_norm": 0.9410419464111328,
"learning_rate": 0.00024378368689861692,
"loss": 4.2196708679199215,
"step": 950
},
{
"epoch": 0.1258191349934469,
"grad_norm": 0.5526502132415771,
"learning_rate": 0.00024361802557970784,
"loss": 4.1913597106933596,
"step": 960
},
{
"epoch": 0.127129750982962,
"grad_norm": 0.6215739846229553,
"learning_rate": 0.0002434502435953449,
"loss": 4.180731582641601,
"step": 970
},
{
"epoch": 0.12844036697247707,
"grad_norm": 0.5965920686721802,
"learning_rate": 0.00024328034394515197,
"loss": 4.179488754272461,
"step": 980
},
{
"epoch": 0.12975098296199214,
"grad_norm": 0.748943030834198,
"learning_rate": 0.00024310832966661277,
"loss": 4.194348907470703,
"step": 990
},
{
"epoch": 0.1310615989515072,
"grad_norm": 0.5486030578613281,
"learning_rate": 0.00024293420383501645,
"loss": 4.1511577606201175,
"step": 1000
},
{
"epoch": 0.1310615989515072,
"eval_loss": 4.1450324058532715,
"eval_runtime": 0.697,
"eval_samples_per_second": 183.646,
"eval_steps_per_second": 45.912,
"step": 1000
},
{
"epoch": 0.13237221494102228,
"grad_norm": 0.6322348713874817,
"learning_rate": 0.0002427579695634029,
"loss": 4.164319610595703,
"step": 1010
},
{
"epoch": 0.13368283093053734,
"grad_norm": 0.6631852388381958,
"learning_rate": 0.00024257963000250682,
"loss": 4.142946243286133,
"step": 1020
},
{
"epoch": 0.1349934469200524,
"grad_norm": 0.5122567415237427,
"learning_rate": 0.0002423991883407015,
"loss": 4.127189254760742,
"step": 1030
},
{
"epoch": 0.1363040629095675,
"grad_norm": 0.763272225856781,
"learning_rate": 0.00024221664780394185,
"loss": 4.170484924316407,
"step": 1040
},
{
"epoch": 0.13761467889908258,
"grad_norm": 0.6399340033531189,
"learning_rate": 0.0002420320116557066,
"loss": 4.122929382324219,
"step": 1050
},
{
"epoch": 0.13892529488859764,
"grad_norm": 0.47659626603126526,
"learning_rate": 0.00024184528319694015,
"loss": 4.153424835205078,
"step": 1060
},
{
"epoch": 0.1402359108781127,
"grad_norm": 0.6294122934341431,
"learning_rate": 0.0002416564657659934,
"loss": 4.09582748413086,
"step": 1070
},
{
"epoch": 0.14154652686762778,
"grad_norm": 0.5373665690422058,
"learning_rate": 0.00024146556273856406,
"loss": 4.130755615234375,
"step": 1080
},
{
"epoch": 0.14285714285714285,
"grad_norm": 0.8437304496765137,
"learning_rate": 0.00024127257752763642,
"loss": 4.111982727050782,
"step": 1090
},
{
"epoch": 0.14416775884665792,
"grad_norm": 0.690742015838623,
"learning_rate": 0.00024107751358342028,
"loss": 4.126205825805664,
"step": 1100
},
{
"epoch": 0.145478374836173,
"grad_norm": 0.5229361057281494,
"learning_rate": 0.00024088037439328914,
"loss": 4.135040283203125,
"step": 1110
},
{
"epoch": 0.14678899082568808,
"grad_norm": 0.4880145490169525,
"learning_rate": 0.0002406811634817181,
"loss": 4.091117858886719,
"step": 1120
},
{
"epoch": 0.14809960681520315,
"grad_norm": 0.6884066462516785,
"learning_rate": 0.0002404798844102206,
"loss": 4.066956329345703,
"step": 1130
},
{
"epoch": 0.14941022280471822,
"grad_norm": 0.5559552311897278,
"learning_rate": 0.00024027654077728497,
"loss": 4.062178039550782,
"step": 1140
},
{
"epoch": 0.15072083879423329,
"grad_norm": 0.6642301678657532,
"learning_rate": 0.0002400711362183099,
"loss": 4.045191955566406,
"step": 1150
},
{
"epoch": 0.15203145478374835,
"grad_norm": 0.7238027453422546,
"learning_rate": 0.0002398636744055395,
"loss": 4.065839004516602,
"step": 1160
},
{
"epoch": 0.15334207077326342,
"grad_norm": 0.6288616061210632,
"learning_rate": 0.00023965415904799781,
"loss": 4.068585205078125,
"step": 1170
},
{
"epoch": 0.15465268676277852,
"grad_norm": 0.6159414649009705,
"learning_rate": 0.0002394425938914222,
"loss": 4.057157897949219,
"step": 1180
},
{
"epoch": 0.1559633027522936,
"grad_norm": 0.4982898235321045,
"learning_rate": 0.00023922898271819677,
"loss": 4.025564575195313,
"step": 1190
},
{
"epoch": 0.15727391874180865,
"grad_norm": 0.44572532176971436,
"learning_rate": 0.00023901332934728425,
"loss": 4.023765563964844,
"step": 1200
},
{
"epoch": 0.15858453473132372,
"grad_norm": 0.497937947511673,
"learning_rate": 0.0002387956376341582,
"loss": 4.034201812744141,
"step": 1210
},
{
"epoch": 0.1598951507208388,
"grad_norm": 0.523916482925415,
"learning_rate": 0.0002385759114707338,
"loss": 4.0323326110839846,
"step": 1220
},
{
"epoch": 0.16120576671035386,
"grad_norm": 0.527370274066925,
"learning_rate": 0.00023835415478529843,
"loss": 4.004526519775391,
"step": 1230
},
{
"epoch": 0.16251638269986893,
"grad_norm": 0.5039871335029602,
"learning_rate": 0.00023813037154244115,
"loss": 3.995006561279297,
"step": 1240
},
{
"epoch": 0.16382699868938402,
"grad_norm": 0.6175016164779663,
"learning_rate": 0.00023790456574298224,
"loss": 3.9849922180175783,
"step": 1250
},
{
"epoch": 0.1651376146788991,
"grad_norm": 0.6377896070480347,
"learning_rate": 0.0002376767414239014,
"loss": 3.988655853271484,
"step": 1260
},
{
"epoch": 0.16644823066841416,
"grad_norm": 0.4930308163166046,
"learning_rate": 0.00023744690265826548,
"loss": 4.000828552246094,
"step": 1270
},
{
"epoch": 0.16775884665792923,
"grad_norm": 0.6034945249557495,
"learning_rate": 0.00023721505355515606,
"loss": 3.9951881408691405,
"step": 1280
},
{
"epoch": 0.1690694626474443,
"grad_norm": 0.5141519904136658,
"learning_rate": 0.00023698119825959563,
"loss": 3.975829315185547,
"step": 1290
},
{
"epoch": 0.17038007863695936,
"grad_norm": 0.5124510526657104,
"learning_rate": 0.0002367453409524735,
"loss": 3.946180725097656,
"step": 1300
},
{
"epoch": 0.17169069462647443,
"grad_norm": 0.5529271960258484,
"learning_rate": 0.00023650748585047135,
"loss": 3.967394256591797,
"step": 1310
},
{
"epoch": 0.17300131061598953,
"grad_norm": 0.5892326235771179,
"learning_rate": 0.00023626763720598748,
"loss": 3.9429306030273437,
"step": 1320
},
{
"epoch": 0.1743119266055046,
"grad_norm": 0.6214840412139893,
"learning_rate": 0.00023602579930706103,
"loss": 3.9684001922607424,
"step": 1330
},
{
"epoch": 0.17562254259501967,
"grad_norm": 0.4972020983695984,
"learning_rate": 0.00023578197647729527,
"loss": 3.959344482421875,
"step": 1340
},
{
"epoch": 0.17693315858453473,
"grad_norm": 0.4380028545856476,
"learning_rate": 0.00023553617307578014,
"loss": 3.9436431884765626,
"step": 1350
},
{
"epoch": 0.1782437745740498,
"grad_norm": 0.4352765381336212,
"learning_rate": 0.0002352883934970146,
"loss": 3.933491516113281,
"step": 1360
},
{
"epoch": 0.17955439056356487,
"grad_norm": 0.542396068572998,
"learning_rate": 0.00023503864217082777,
"loss": 3.9454887390136717,
"step": 1370
},
{
"epoch": 0.18086500655307994,
"grad_norm": 0.4181693196296692,
"learning_rate": 0.00023478692356230003,
"loss": 3.90989990234375,
"step": 1380
},
{
"epoch": 0.182175622542595,
"grad_norm": 0.5905563235282898,
"learning_rate": 0.00023453324217168292,
"loss": 3.911590576171875,
"step": 1390
},
{
"epoch": 0.1834862385321101,
"grad_norm": 0.5032172799110413,
"learning_rate": 0.0002342776025343189,
"loss": 3.926470947265625,
"step": 1400
},
{
"epoch": 0.18479685452162517,
"grad_norm": 0.5959187746047974,
"learning_rate": 0.00023402000922056004,
"loss": 3.892247772216797,
"step": 1410
},
{
"epoch": 0.18610747051114024,
"grad_norm": 0.5463594794273376,
"learning_rate": 0.00023376046683568652,
"loss": 3.9080841064453127,
"step": 1420
},
{
"epoch": 0.1874180865006553,
"grad_norm": 0.5404369831085205,
"learning_rate": 0.00023349898001982428,
"loss": 3.9146347045898438,
"step": 1430
},
{
"epoch": 0.18872870249017037,
"grad_norm": 0.5168806910514832,
"learning_rate": 0.0002332355534478619,
"loss": 3.9106430053710937,
"step": 1440
},
{
"epoch": 0.19003931847968544,
"grad_norm": 0.44752195477485657,
"learning_rate": 0.00023297019182936713,
"loss": 3.9024272918701173,
"step": 1450
},
{
"epoch": 0.1913499344692005,
"grad_norm": 0.5084620714187622,
"learning_rate": 0.00023270289990850275,
"loss": 3.8871227264404298,
"step": 1460
},
{
"epoch": 0.1926605504587156,
"grad_norm": 0.489383727312088,
"learning_rate": 0.00023243368246394158,
"loss": 3.894936370849609,
"step": 1470
},
{
"epoch": 0.19397116644823068,
"grad_norm": 0.4871124029159546,
"learning_rate": 0.00023216254430878127,
"loss": 3.891936492919922,
"step": 1480
},
{
"epoch": 0.19528178243774574,
"grad_norm": 0.5670658946037292,
"learning_rate": 0.00023188949029045807,
"loss": 3.9086143493652346,
"step": 1490
},
{
"epoch": 0.1965923984272608,
"grad_norm": 0.4445112645626068,
"learning_rate": 0.0002316145252906603,
"loss": 3.9029220581054687,
"step": 1500
},
{
"epoch": 0.1965923984272608,
"eval_loss": 3.8595924377441406,
"eval_runtime": 0.6992,
"eval_samples_per_second": 183.055,
"eval_steps_per_second": 45.764,
"step": 1500
},
{
"epoch": 0.19790301441677588,
"grad_norm": 0.4595864713191986,
"learning_rate": 0.0002313376542252409,
"loss": 3.8990947723388674,
"step": 1510
},
{
"epoch": 0.19921363040629095,
"grad_norm": 0.44442445039749146,
"learning_rate": 0.00023105888204412968,
"loss": 3.866590118408203,
"step": 1520
},
{
"epoch": 0.20052424639580602,
"grad_norm": 0.46469536423683167,
"learning_rate": 0.00023077821373124484,
"loss": 3.9012496948242186,
"step": 1530
},
{
"epoch": 0.2018348623853211,
"grad_norm": 0.5682144165039062,
"learning_rate": 0.00023049565430440384,
"loss": 3.947462463378906,
"step": 1540
},
{
"epoch": 0.20314547837483618,
"grad_norm": 0.5124285817146301,
"learning_rate": 0.00023021120881523357,
"loss": 3.8624008178710936,
"step": 1550
},
{
"epoch": 0.20445609436435125,
"grad_norm": 0.4463842809200287,
"learning_rate": 0.00022992488234908028,
"loss": 3.872223663330078,
"step": 1560
},
{
"epoch": 0.20576671035386632,
"grad_norm": 0.43720901012420654,
"learning_rate": 0.00022963668002491837,
"loss": 3.851451873779297,
"step": 1570
},
{
"epoch": 0.20707732634338138,
"grad_norm": 0.43668457865715027,
"learning_rate": 0.00022934660699525913,
"loss": 3.869467926025391,
"step": 1580
},
{
"epoch": 0.20838794233289645,
"grad_norm": 0.5460702180862427,
"learning_rate": 0.00022905466844605848,
"loss": 3.8474761962890627,
"step": 1590
},
{
"epoch": 0.20969855832241152,
"grad_norm": 0.5258567929267883,
"learning_rate": 0.00022876086959662434,
"loss": 3.845472717285156,
"step": 1600
},
{
"epoch": 0.21100917431192662,
"grad_norm": 0.47654300928115845,
"learning_rate": 0.00022846521569952323,
"loss": 3.852452850341797,
"step": 1610
},
{
"epoch": 0.21231979030144169,
"grad_norm": 0.4372918903827667,
"learning_rate": 0.00022816771204048637,
"loss": 3.8639129638671874,
"step": 1620
},
{
"epoch": 0.21363040629095675,
"grad_norm": 0.4561574161052704,
"learning_rate": 0.00022786836393831527,
"loss": 3.8518062591552735,
"step": 1630
},
{
"epoch": 0.21494102228047182,
"grad_norm": 0.629420280456543,
"learning_rate": 0.00022756717674478658,
"loss": 3.834371566772461,
"step": 1640
},
{
"epoch": 0.2162516382699869,
"grad_norm": 0.5868667364120483,
"learning_rate": 0.00022726415584455641,
"loss": 3.8643478393554687,
"step": 1650
},
{
"epoch": 0.21756225425950196,
"grad_norm": 0.40777280926704407,
"learning_rate": 0.0002269593066550641,
"loss": 3.8676849365234376,
"step": 1660
},
{
"epoch": 0.21887287024901703,
"grad_norm": 0.4707736372947693,
"learning_rate": 0.00022665263462643526,
"loss": 3.8358665466308595,
"step": 1670
},
{
"epoch": 0.22018348623853212,
"grad_norm": 0.4575025737285614,
"learning_rate": 0.0002263441452413845,
"loss": 3.8249862670898436,
"step": 1680
},
{
"epoch": 0.2214941022280472,
"grad_norm": 0.41110938787460327,
"learning_rate": 0.00022603384401511723,
"loss": 3.8356868743896486,
"step": 1690
},
{
"epoch": 0.22280471821756226,
"grad_norm": 0.47845396399497986,
"learning_rate": 0.00022572173649523125,
"loss": 3.839044189453125,
"step": 1700
},
{
"epoch": 0.22411533420707733,
"grad_norm": 0.5048379898071289,
"learning_rate": 0.00022540782826161736,
"loss": 3.8390274047851562,
"step": 1710
},
{
"epoch": 0.2254259501965924,
"grad_norm": 0.44569337368011475,
"learning_rate": 0.00022509212492635976,
"loss": 3.8324737548828125,
"step": 1720
},
{
"epoch": 0.22673656618610746,
"grad_norm": 0.43920955061912537,
"learning_rate": 0.00022477463213363573,
"loss": 3.8614524841308593,
"step": 1730
},
{
"epoch": 0.22804718217562253,
"grad_norm": 0.5752826929092407,
"learning_rate": 0.0002244553555596145,
"loss": 3.8162994384765625,
"step": 1740
},
{
"epoch": 0.22935779816513763,
"grad_norm": 0.49210062623023987,
"learning_rate": 0.000224134300912356,
"loss": 3.838577651977539,
"step": 1750
},
{
"epoch": 0.2306684141546527,
"grad_norm": 0.386764794588089,
"learning_rate": 0.00022381147393170888,
"loss": 3.812305450439453,
"step": 1760
},
{
"epoch": 0.23197903014416776,
"grad_norm": 0.41887620091438293,
"learning_rate": 0.0002234868803892075,
"loss": 3.807159423828125,
"step": 1770
},
{
"epoch": 0.23328964613368283,
"grad_norm": 0.3812844157218933,
"learning_rate": 0.0002231605260879692,
"loss": 3.8062633514404296,
"step": 1780
},
{
"epoch": 0.2346002621231979,
"grad_norm": 0.392352432012558,
"learning_rate": 0.00022283241686259022,
"loss": 3.8229133605957033,
"step": 1790
},
{
"epoch": 0.23591087811271297,
"grad_norm": 0.4600699841976166,
"learning_rate": 0.0002225025585790416,
"loss": 3.7980079650878906,
"step": 1800
},
{
"epoch": 0.23722149410222804,
"grad_norm": 0.5496444702148438,
"learning_rate": 0.00022217095713456423,
"loss": 3.7862876892089843,
"step": 1810
},
{
"epoch": 0.23853211009174313,
"grad_norm": 0.4653802216053009,
"learning_rate": 0.00022183761845756333,
"loss": 3.7867252349853517,
"step": 1820
},
{
"epoch": 0.2398427260812582,
"grad_norm": 0.4444230794906616,
"learning_rate": 0.0002215025485075026,
"loss": 3.753264617919922,
"step": 1830
},
{
"epoch": 0.24115334207077327,
"grad_norm": 0.4293712079524994,
"learning_rate": 0.0002211657532747976,
"loss": 3.8199562072753905,
"step": 1840
},
{
"epoch": 0.24246395806028834,
"grad_norm": 0.4202822744846344,
"learning_rate": 0.00022082723878070874,
"loss": 3.7781608581542967,
"step": 1850
},
{
"epoch": 0.2437745740498034,
"grad_norm": 0.4034839868545532,
"learning_rate": 0.00022048701107723346,
"loss": 3.783266067504883,
"step": 1860
},
{
"epoch": 0.24508519003931847,
"grad_norm": 0.4446223974227905,
"learning_rate": 0.0002201450762469982,
"loss": 3.798517608642578,
"step": 1870
},
{
"epoch": 0.24639580602883354,
"grad_norm": 0.4842108190059662,
"learning_rate": 0.0002198014404031496,
"loss": 3.769219970703125,
"step": 1880
},
{
"epoch": 0.24770642201834864,
"grad_norm": 0.4581057131290436,
"learning_rate": 0.0002194561096892451,
"loss": 3.7815696716308596,
"step": 1890
},
{
"epoch": 0.2490170380078637,
"grad_norm": 0.45605164766311646,
"learning_rate": 0.0002191090902791434,
"loss": 3.7596714019775392,
"step": 1900
},
{
"epoch": 0.2503276539973788,
"grad_norm": 0.4910837709903717,
"learning_rate": 0.00021876038837689371,
"loss": 3.771947479248047,
"step": 1910
},
{
"epoch": 0.2516382699868938,
"grad_norm": 0.4998149871826172,
"learning_rate": 0.00021841001021662504,
"loss": 3.7736221313476563,
"step": 1920
},
{
"epoch": 0.2529488859764089,
"grad_norm": 0.4697710871696472,
"learning_rate": 0.00021805796206243478,
"loss": 3.7950729370117187,
"step": 1930
},
{
"epoch": 0.254259501965924,
"grad_norm": 0.4391549229621887,
"learning_rate": 0.0002177042502082766,
"loss": 3.772383117675781,
"step": 1940
},
{
"epoch": 0.25557011795543905,
"grad_norm": 0.42760396003723145,
"learning_rate": 0.000217348880977848,
"loss": 3.7615383148193358,
"step": 1950
},
{
"epoch": 0.25688073394495414,
"grad_norm": 0.4630841910839081,
"learning_rate": 0.0002169918607244772,
"loss": 3.7982250213623048,
"step": 1960
},
{
"epoch": 0.2581913499344692,
"grad_norm": 0.44697806239128113,
"learning_rate": 0.00021663319583100957,
"loss": 3.7610572814941405,
"step": 1970
},
{
"epoch": 0.2595019659239843,
"grad_norm": 0.3792683482170105,
"learning_rate": 0.00021627289270969357,
"loss": 3.7640010833740236,
"step": 1980
},
{
"epoch": 0.2608125819134993,
"grad_norm": 0.4451923966407776,
"learning_rate": 0.00021591095780206605,
"loss": 3.721126174926758,
"step": 1990
},
{
"epoch": 0.2621231979030144,
"grad_norm": 0.4384404122829437,
"learning_rate": 0.00021554739757883706,
"loss": 3.755757141113281,
"step": 2000
},
{
"epoch": 0.2621231979030144,
"eval_loss": 3.70749568939209,
"eval_runtime": 0.6984,
"eval_samples_per_second": 183.269,
"eval_steps_per_second": 45.817,
"step": 2000
},
{
"epoch": 0.2634338138925295,
"grad_norm": 0.4921295642852783,
"learning_rate": 0.00021518221853977426,
"loss": 3.7308048248291015,
"step": 2010
},
{
"epoch": 0.26474442988204455,
"grad_norm": 0.41485610604286194,
"learning_rate": 0.0002148154272135867,
"loss": 3.7479896545410156,
"step": 2020
},
{
"epoch": 0.26605504587155965,
"grad_norm": 0.4536568820476532,
"learning_rate": 0.00021444703015780807,
"loss": 3.7584232330322265,
"step": 2030
},
{
"epoch": 0.2673656618610747,
"grad_norm": 0.4791676700115204,
"learning_rate": 0.0002140770339586793,
"loss": 3.7264373779296873,
"step": 2040
},
{
"epoch": 0.2686762778505898,
"grad_norm": 0.41909492015838623,
"learning_rate": 0.0002137054452310312,
"loss": 3.7034957885742186,
"step": 2050
},
{
"epoch": 0.2699868938401048,
"grad_norm": 0.38637474179267883,
"learning_rate": 0.00021333227061816579,
"loss": 3.7362186431884767,
"step": 2060
},
{
"epoch": 0.2712975098296199,
"grad_norm": 0.35627466440200806,
"learning_rate": 0.0002129575167917378,
"loss": 3.7326896667480467,
"step": 2070
},
{
"epoch": 0.272608125819135,
"grad_norm": 0.5081406235694885,
"learning_rate": 0.00021258119045163538,
"loss": 3.710901641845703,
"step": 2080
},
{
"epoch": 0.27391874180865006,
"grad_norm": 0.4458751380443573,
"learning_rate": 0.00021220329832586,
"loss": 3.708026885986328,
"step": 2090
},
{
"epoch": 0.27522935779816515,
"grad_norm": 0.37385478615760803,
"learning_rate": 0.00021182384717040667,
"loss": 3.7077880859375,
"step": 2100
},
{
"epoch": 0.2765399737876802,
"grad_norm": 0.4499031901359558,
"learning_rate": 0.0002114428437691428,
"loss": 3.735845947265625,
"step": 2110
},
{
"epoch": 0.2778505897771953,
"grad_norm": 0.4073517918586731,
"learning_rate": 0.00021106029493368696,
"loss": 3.7483619689941405,
"step": 2120
},
{
"epoch": 0.27916120576671033,
"grad_norm": 0.4546799063682556,
"learning_rate": 0.0002106762075032872,
"loss": 3.7409950256347657,
"step": 2130
},
{
"epoch": 0.2804718217562254,
"grad_norm": 0.4047580361366272,
"learning_rate": 0.00021029058834469872,
"loss": 3.6970542907714843,
"step": 2140
},
{
"epoch": 0.2817824377457405,
"grad_norm": 0.38918593525886536,
"learning_rate": 0.00020990344435206117,
"loss": 3.656460189819336,
"step": 2150
},
{
"epoch": 0.28309305373525556,
"grad_norm": 0.40187233686447144,
"learning_rate": 0.00020951478244677523,
"loss": 3.7148818969726562,
"step": 2160
},
{
"epoch": 0.28440366972477066,
"grad_norm": 0.3731965124607086,
"learning_rate": 0.00020912460957737915,
"loss": 3.6839984893798827,
"step": 2170
},
{
"epoch": 0.2857142857142857,
"grad_norm": 0.4111628830432892,
"learning_rate": 0.00020873293271942415,
"loss": 3.703624725341797,
"step": 2180
},
{
"epoch": 0.2870249017038008,
"grad_norm": 0.3874831795692444,
"learning_rate": 0.00020833975887535005,
"loss": 3.6873214721679686,
"step": 2190
},
{
"epoch": 0.28833551769331583,
"grad_norm": 0.39543426036834717,
"learning_rate": 0.0002079450950743599,
"loss": 3.673131561279297,
"step": 2200
},
{
"epoch": 0.28964613368283093,
"grad_norm": 0.3771519958972931,
"learning_rate": 0.00020754894837229443,
"loss": 3.7100814819335937,
"step": 2210
},
{
"epoch": 0.290956749672346,
"grad_norm": 0.3592368960380554,
"learning_rate": 0.00020715132585150565,
"loss": 3.715135955810547,
"step": 2220
},
{
"epoch": 0.29226736566186107,
"grad_norm": 0.4421813488006592,
"learning_rate": 0.00020675223462073068,
"loss": 3.680974578857422,
"step": 2230
},
{
"epoch": 0.29357798165137616,
"grad_norm": 0.43614616990089417,
"learning_rate": 0.00020635168181496404,
"loss": 3.7040740966796877,
"step": 2240
},
{
"epoch": 0.2948885976408912,
"grad_norm": 0.3772644102573395,
"learning_rate": 0.00020594967459533085,
"loss": 3.6878517150878904,
"step": 2250
},
{
"epoch": 0.2961992136304063,
"grad_norm": 0.41714298725128174,
"learning_rate": 0.00020554622014895806,
"loss": 3.6695098876953125,
"step": 2260
},
{
"epoch": 0.29750982961992134,
"grad_norm": 0.4420553743839264,
"learning_rate": 0.0002051413256888464,
"loss": 3.6577522277832033,
"step": 2270
},
{
"epoch": 0.29882044560943644,
"grad_norm": 0.38797250390052795,
"learning_rate": 0.00020473499845374143,
"loss": 3.6768707275390624,
"step": 2280
},
{
"epoch": 0.30013106159895153,
"grad_norm": 0.39683839678764343,
"learning_rate": 0.00020432724570800383,
"loss": 3.682420349121094,
"step": 2290
},
{
"epoch": 0.30144167758846657,
"grad_norm": 0.3876133859157562,
"learning_rate": 0.00020391807474147978,
"loss": 3.65374755859375,
"step": 2300
},
{
"epoch": 0.30275229357798167,
"grad_norm": 0.48667997121810913,
"learning_rate": 0.0002035074928693706,
"loss": 3.6826313018798826,
"step": 2310
},
{
"epoch": 0.3040629095674967,
"grad_norm": 0.3844733238220215,
"learning_rate": 0.00020309550743210188,
"loss": 3.670343017578125,
"step": 2320
},
{
"epoch": 0.3053735255570118,
"grad_norm": 0.3886657655239105,
"learning_rate": 0.00020268212579519236,
"loss": 3.6630802154541016,
"step": 2330
},
{
"epoch": 0.30668414154652685,
"grad_norm": 0.3683916926383972,
"learning_rate": 0.00020226735534912201,
"loss": 3.6767845153808594,
"step": 2340
},
{
"epoch": 0.30799475753604194,
"grad_norm": 0.39340534806251526,
"learning_rate": 0.00020185120350920034,
"loss": 3.671356201171875,
"step": 2350
},
{
"epoch": 0.30930537352555704,
"grad_norm": 0.4987418055534363,
"learning_rate": 0.00020143367771543333,
"loss": 3.6589839935302733,
"step": 2360
},
{
"epoch": 0.3106159895150721,
"grad_norm": 0.43473708629608154,
"learning_rate": 0.00020101478543239077,
"loss": 3.6682861328125,
"step": 2370
},
{
"epoch": 0.3119266055045872,
"grad_norm": 0.3953477144241333,
"learning_rate": 0.0002005945341490727,
"loss": 3.6322235107421874,
"step": 2380
},
{
"epoch": 0.3132372214941022,
"grad_norm": 0.5532742738723755,
"learning_rate": 0.00020017293137877538,
"loss": 3.663126754760742,
"step": 2390
},
{
"epoch": 0.3145478374836173,
"grad_norm": 0.4194016456604004,
"learning_rate": 0.00019974998465895728,
"loss": 3.635631561279297,
"step": 2400
},
{
"epoch": 0.31585845347313235,
"grad_norm": 0.4373791217803955,
"learning_rate": 0.00019932570155110395,
"loss": 3.6377632141113283,
"step": 2410
},
{
"epoch": 0.31716906946264745,
"grad_norm": 0.4019886553287506,
"learning_rate": 0.00019890008964059316,
"loss": 3.6277240753173827,
"step": 2420
},
{
"epoch": 0.31847968545216254,
"grad_norm": 0.37631332874298096,
"learning_rate": 0.00019847315653655914,
"loss": 3.641008758544922,
"step": 2430
},
{
"epoch": 0.3197903014416776,
"grad_norm": 0.3595082461833954,
"learning_rate": 0.0001980449098717565,
"loss": 3.645294189453125,
"step": 2440
},
{
"epoch": 0.3211009174311927,
"grad_norm": 0.3677518367767334,
"learning_rate": 0.00019761535730242387,
"loss": 3.6396930694580076,
"step": 2450
},
{
"epoch": 0.3224115334207077,
"grad_norm": 0.38670068979263306,
"learning_rate": 0.00019718450650814696,
"loss": 3.623303985595703,
"step": 2460
},
{
"epoch": 0.3237221494102228,
"grad_norm": 0.38947510719299316,
"learning_rate": 0.00019675236519172123,
"loss": 3.649179458618164,
"step": 2470
},
{
"epoch": 0.32503276539973786,
"grad_norm": 0.40161463618278503,
"learning_rate": 0.00019631894107901434,
"loss": 3.6486370086669924,
"step": 2480
},
{
"epoch": 0.32634338138925295,
"grad_norm": 0.47498735785484314,
"learning_rate": 0.00019588424191882786,
"loss": 3.6256385803222657,
"step": 2490
},
{
"epoch": 0.32765399737876805,
"grad_norm": 0.4358643591403961,
"learning_rate": 0.0001954482754827588,
"loss": 3.592013931274414,
"step": 2500
},
{
"epoch": 0.32765399737876805,
"eval_loss": 3.61002254486084,
"eval_runtime": 0.699,
"eval_samples_per_second": 183.109,
"eval_steps_per_second": 45.777,
"step": 2500
},
{
"epoch": 0.3289646133682831,
"grad_norm": 0.35884666442871094,
"learning_rate": 0.00019501104956506056,
"loss": 3.597154998779297,
"step": 2510
},
{
"epoch": 0.3302752293577982,
"grad_norm": 0.6850298643112183,
"learning_rate": 0.00019457257198250385,
"loss": 3.6494251251220704,
"step": 2520
},
{
"epoch": 0.3315858453473132,
"grad_norm": 0.40816670656204224,
"learning_rate": 0.00019413285057423674,
"loss": 3.6235000610351564,
"step": 2530
},
{
"epoch": 0.3328964613368283,
"grad_norm": 0.42288416624069214,
"learning_rate": 0.00019369189320164444,
"loss": 3.619795227050781,
"step": 2540
},
{
"epoch": 0.33420707732634336,
"grad_norm": 0.37256839871406555,
"learning_rate": 0.00019324970774820894,
"loss": 3.600733184814453,
"step": 2550
},
{
"epoch": 0.33551769331585846,
"grad_norm": 0.4219217002391815,
"learning_rate": 0.00019280630211936794,
"loss": 3.62880859375,
"step": 2560
},
{
"epoch": 0.33682830930537355,
"grad_norm": 0.34511440992355347,
"learning_rate": 0.0001923616842423737,
"loss": 3.6224590301513673,
"step": 2570
},
{
"epoch": 0.3381389252948886,
"grad_norm": 0.3655567169189453,
"learning_rate": 0.00019191586206615096,
"loss": 3.614408493041992,
"step": 2580
},
{
"epoch": 0.3394495412844037,
"grad_norm": 0.3706367015838623,
"learning_rate": 0.00019146884356115523,
"loss": 3.602632522583008,
"step": 2590
},
{
"epoch": 0.34076015727391873,
"grad_norm": 0.40935879945755005,
"learning_rate": 0.00019102063671923007,
"loss": 3.58270263671875,
"step": 2600
},
{
"epoch": 0.3420707732634338,
"grad_norm": 0.415536105632782,
"learning_rate": 0.00019057124955346413,
"loss": 3.6320755004882814,
"step": 2610
},
{
"epoch": 0.34338138925294887,
"grad_norm": 0.36466723680496216,
"learning_rate": 0.00019012069009804818,
"loss": 3.624856948852539,
"step": 2620
},
{
"epoch": 0.34469200524246396,
"grad_norm": 0.4434064030647278,
"learning_rate": 0.00018966896640813124,
"loss": 3.6535144805908204,
"step": 2630
},
{
"epoch": 0.34600262123197906,
"grad_norm": 0.39541974663734436,
"learning_rate": 0.00018921608655967658,
"loss": 3.616999053955078,
"step": 2640
},
{
"epoch": 0.3473132372214941,
"grad_norm": 0.3162301480770111,
"learning_rate": 0.0001887620586493175,
"loss": 3.616608810424805,
"step": 2650
},
{
"epoch": 0.3486238532110092,
"grad_norm": 0.39664795994758606,
"learning_rate": 0.00018830689079421238,
"loss": 3.6184730529785156,
"step": 2660
},
{
"epoch": 0.34993446920052423,
"grad_norm": 0.43438875675201416,
"learning_rate": 0.00018785059113189974,
"loss": 3.6125778198242187,
"step": 2670
},
{
"epoch": 0.35124508519003933,
"grad_norm": 0.38295066356658936,
"learning_rate": 0.00018739316782015247,
"loss": 3.6037372589111327,
"step": 2680
},
{
"epoch": 0.35255570117955437,
"grad_norm": 0.3755257725715637,
"learning_rate": 0.0001869346290368325,
"loss": 3.618543243408203,
"step": 2690
},
{
"epoch": 0.35386631716906947,
"grad_norm": 0.37485548853874207,
"learning_rate": 0.00018647498297974405,
"loss": 3.574940872192383,
"step": 2700
},
{
"epoch": 0.35517693315858456,
"grad_norm": 0.4129403233528137,
"learning_rate": 0.0001860142378664873,
"loss": 3.6213829040527346,
"step": 2710
},
{
"epoch": 0.3564875491480996,
"grad_norm": 0.4433867931365967,
"learning_rate": 0.0001855524019343116,
"loss": 3.581695556640625,
"step": 2720
},
{
"epoch": 0.3577981651376147,
"grad_norm": 0.38252827525138855,
"learning_rate": 0.0001850894834399679,
"loss": 3.601634979248047,
"step": 2730
},
{
"epoch": 0.35910878112712974,
"grad_norm": 0.3612919747829437,
"learning_rate": 0.00018462549065956143,
"loss": 3.6092594146728514,
"step": 2740
},
{
"epoch": 0.36041939711664484,
"grad_norm": 0.3477737009525299,
"learning_rate": 0.0001841604318884037,
"loss": 3.5911041259765626,
"step": 2750
},
{
"epoch": 0.3617300131061599,
"grad_norm": 0.43406569957733154,
"learning_rate": 0.00018369431544086384,
"loss": 3.5967559814453125,
"step": 2760
},
{
"epoch": 0.36304062909567497,
"grad_norm": 0.3807731568813324,
"learning_rate": 0.00018322714965022057,
"loss": 3.6038742065429688,
"step": 2770
},
{
"epoch": 0.36435124508519,
"grad_norm": 0.3919323980808258,
"learning_rate": 0.00018275894286851256,
"loss": 3.5964141845703126,
"step": 2780
},
{
"epoch": 0.3656618610747051,
"grad_norm": 0.44667112827301025,
"learning_rate": 0.00018228970346638963,
"loss": 3.5915550231933593,
"step": 2790
},
{
"epoch": 0.3669724770642202,
"grad_norm": 0.3827419579029083,
"learning_rate": 0.0001818194398329628,
"loss": 3.56480712890625,
"step": 2800
},
{
"epoch": 0.36828309305373524,
"grad_norm": 0.3612150549888611,
"learning_rate": 0.00018134816037565453,
"loss": 3.6135284423828127,
"step": 2810
},
{
"epoch": 0.36959370904325034,
"grad_norm": 0.3850660026073456,
"learning_rate": 0.0001808758735200481,
"loss": 3.6066566467285157,
"step": 2820
},
{
"epoch": 0.3709043250327654,
"grad_norm": 0.4873389005661011,
"learning_rate": 0.00018040258770973724,
"loss": 3.5608505249023437,
"step": 2830
},
{
"epoch": 0.3722149410222805,
"grad_norm": 0.380490243434906,
"learning_rate": 0.00017992831140617522,
"loss": 3.587847900390625,
"step": 2840
},
{
"epoch": 0.3735255570117955,
"grad_norm": 0.43044331669807434,
"learning_rate": 0.0001794530530885233,
"loss": 3.604927825927734,
"step": 2850
},
{
"epoch": 0.3748361730013106,
"grad_norm": 0.416038453578949,
"learning_rate": 0.00017897682125349925,
"loss": 3.6181507110595703,
"step": 2860
},
{
"epoch": 0.3761467889908257,
"grad_norm": 0.5963427424430847,
"learning_rate": 0.00017849962441522567,
"loss": 3.629560852050781,
"step": 2870
},
{
"epoch": 0.37745740498034075,
"grad_norm": 0.3649164140224457,
"learning_rate": 0.00017802147110507747,
"loss": 3.5839302062988283,
"step": 2880
},
{
"epoch": 0.37876802096985585,
"grad_norm": 0.35835522413253784,
"learning_rate": 0.00017754236987152944,
"loss": 3.5891387939453123,
"step": 2890
},
{
"epoch": 0.3800786369593709,
"grad_norm": 0.3978164494037628,
"learning_rate": 0.00017706232928000357,
"loss": 3.5973220825195313,
"step": 2900
},
{
"epoch": 0.381389252948886,
"grad_norm": 0.3993713855743408,
"learning_rate": 0.00017658135791271561,
"loss": 3.5841751098632812,
"step": 2910
},
{
"epoch": 0.382699868938401,
"grad_norm": 0.367861807346344,
"learning_rate": 0.00017609946436852202,
"loss": 3.59312744140625,
"step": 2920
},
{
"epoch": 0.3840104849279161,
"grad_norm": 0.34430232644081116,
"learning_rate": 0.00017561665726276583,
"loss": 3.603657531738281,
"step": 2930
},
{
"epoch": 0.3853211009174312,
"grad_norm": 0.42508161067962646,
"learning_rate": 0.00017513294522712296,
"loss": 3.6063438415527345,
"step": 2940
},
{
"epoch": 0.38663171690694625,
"grad_norm": 0.4225197732448578,
"learning_rate": 0.0001746483369094477,
"loss": 3.6155670166015623,
"step": 2950
},
{
"epoch": 0.38794233289646135,
"grad_norm": 0.3732278347015381,
"learning_rate": 0.00017416284097361822,
"loss": 3.6005844116210937,
"step": 2960
},
{
"epoch": 0.3892529488859764,
"grad_norm": 0.35795629024505615,
"learning_rate": 0.00017367646609938155,
"loss": 3.5956985473632814,
"step": 2970
},
{
"epoch": 0.3905635648754915,
"grad_norm": 0.3546122610569,
"learning_rate": 0.00017318922098219852,
"loss": 3.5926021575927733,
"step": 2980
},
{
"epoch": 0.3918741808650065,
"grad_norm": 0.4074459373950958,
"learning_rate": 0.00017270111433308828,
"loss": 3.594530487060547,
"step": 2990
},
{
"epoch": 0.3931847968545216,
"grad_norm": 0.35302865505218506,
"learning_rate": 0.0001722121548784724,
"loss": 3.565687561035156,
"step": 3000
},
{
"epoch": 0.3931847968545216,
"eval_loss": 3.5327887535095215,
"eval_runtime": 0.6989,
"eval_samples_per_second": 183.143,
"eval_steps_per_second": 45.786,
"step": 3000
},
{
"epoch": 0.3944954128440367,
"grad_norm": 0.3555203080177307,
"learning_rate": 0.00017172235136001912,
"loss": 3.58720703125,
"step": 3010
},
{
"epoch": 0.39580602883355176,
"grad_norm": 0.3768612742424011,
"learning_rate": 0.00017123171253448702,
"loss": 3.5762924194335937,
"step": 3020
},
{
"epoch": 0.39711664482306686,
"grad_norm": 0.38559049367904663,
"learning_rate": 0.00017074024717356813,
"loss": 3.6308860778808594,
"step": 3030
},
{
"epoch": 0.3984272608125819,
"grad_norm": 0.34866032004356384,
"learning_rate": 0.00017024796406373168,
"loss": 3.579779052734375,
"step": 3040
},
{
"epoch": 0.399737876802097,
"grad_norm": 0.394422709941864,
"learning_rate": 0.00016975487200606636,
"loss": 3.620661163330078,
"step": 3050
},
{
"epoch": 0.40104849279161203,
"grad_norm": 0.37144729495048523,
"learning_rate": 0.00016926097981612362,
"loss": 3.609404373168945,
"step": 3060
},
{
"epoch": 0.40235910878112713,
"grad_norm": 0.38082075119018555,
"learning_rate": 0.0001687662963237596,
"loss": 3.600936508178711,
"step": 3070
},
{
"epoch": 0.4036697247706422,
"grad_norm": 0.4842313230037689,
"learning_rate": 0.00016827083037297742,
"loss": 3.5481918334960936,
"step": 3080
},
{
"epoch": 0.40498034076015726,
"grad_norm": 0.3545914888381958,
"learning_rate": 0.00016777459082176918,
"loss": 3.592304992675781,
"step": 3090
},
{
"epoch": 0.40629095674967236,
"grad_norm": 0.3619571328163147,
"learning_rate": 0.00016727758654195736,
"loss": 3.561437225341797,
"step": 3100
},
{
"epoch": 0.4076015727391874,
"grad_norm": 0.36174827814102173,
"learning_rate": 0.00016677982641903643,
"loss": 3.5731430053710938,
"step": 3110
},
{
"epoch": 0.4089121887287025,
"grad_norm": 0.3398175537586212,
"learning_rate": 0.00016628131935201388,
"loss": 3.572527313232422,
"step": 3120
},
{
"epoch": 0.41022280471821754,
"grad_norm": 0.37768232822418213,
"learning_rate": 0.0001657820742532511,
"loss": 3.551443862915039,
"step": 3130
},
{
"epoch": 0.41153342070773263,
"grad_norm": 0.3821849524974823,
"learning_rate": 0.00016528210004830414,
"loss": 3.5756664276123047,
"step": 3140
},
{
"epoch": 0.41284403669724773,
"grad_norm": 0.39957883954048157,
"learning_rate": 0.00016478140567576408,
"loss": 3.556570816040039,
"step": 3150
},
{
"epoch": 0.41415465268676277,
"grad_norm": 0.35470208525657654,
"learning_rate": 0.0001642800000870972,
"loss": 3.5717140197753907,
"step": 3160
},
{
"epoch": 0.41546526867627787,
"grad_norm": 0.5372022390365601,
"learning_rate": 0.00016377789224648504,
"loss": 3.553394317626953,
"step": 3170
},
{
"epoch": 0.4167758846657929,
"grad_norm": 0.43058937788009644,
"learning_rate": 0.00016327509113066397,
"loss": 3.5411109924316406,
"step": 3180
},
{
"epoch": 0.418086500655308,
"grad_norm": 0.3757922351360321,
"learning_rate": 0.00016277160572876495,
"loss": 3.5552886962890624,
"step": 3190
},
{
"epoch": 0.41939711664482304,
"grad_norm": 0.4144268333911896,
"learning_rate": 0.0001622674450421525,
"loss": 3.5552597045898438,
"step": 3200
},
{
"epoch": 0.42070773263433814,
"grad_norm": 0.42654407024383545,
"learning_rate": 0.00016176261808426416,
"loss": 3.5842060089111327,
"step": 3210
},
{
"epoch": 0.42201834862385323,
"grad_norm": 0.36649268865585327,
"learning_rate": 0.00016125713388044887,
"loss": 3.562412643432617,
"step": 3220
},
{
"epoch": 0.4233289646133683,
"grad_norm": 0.32557469606399536,
"learning_rate": 0.0001607510014678062,
"loss": 3.5480194091796875,
"step": 3230
},
{
"epoch": 0.42463958060288337,
"grad_norm": 0.3594341576099396,
"learning_rate": 0.0001602442298950242,
"loss": 3.547731399536133,
"step": 3240
},
{
"epoch": 0.4259501965923984,
"grad_norm": 0.38341233134269714,
"learning_rate": 0.00015973682822221802,
"loss": 3.5556121826171876,
"step": 3250
},
{
"epoch": 0.4272608125819135,
"grad_norm": 0.3498956561088562,
"learning_rate": 0.0001592288055207678,
"loss": 3.5513984680175783,
"step": 3260
},
{
"epoch": 0.42857142857142855,
"grad_norm": 0.39473220705986023,
"learning_rate": 0.00015872017087315644,
"loss": 3.507306671142578,
"step": 3270
},
{
"epoch": 0.42988204456094364,
"grad_norm": 0.39170652627944946,
"learning_rate": 0.00015821093337280728,
"loss": 3.5521778106689452,
"step": 3280
},
{
"epoch": 0.43119266055045874,
"grad_norm": 0.46084344387054443,
"learning_rate": 0.00015770110212392162,
"loss": 3.5556732177734376,
"step": 3290
},
{
"epoch": 0.4325032765399738,
"grad_norm": 0.3563990294933319,
"learning_rate": 0.00015719068624131578,
"loss": 3.524671173095703,
"step": 3300
},
{
"epoch": 0.4338138925294889,
"grad_norm": 0.35894671082496643,
"learning_rate": 0.00015667969485025823,
"loss": 3.5170135498046875,
"step": 3310
},
{
"epoch": 0.4351245085190039,
"grad_norm": 0.37164202332496643,
"learning_rate": 0.0001561681370863064,
"loss": 3.5301841735839843,
"step": 3320
},
{
"epoch": 0.436435124508519,
"grad_norm": 0.35135722160339355,
"learning_rate": 0.00015565602209514348,
"loss": 3.539019775390625,
"step": 3330
},
{
"epoch": 0.43774574049803405,
"grad_norm": 0.3510042428970337,
"learning_rate": 0.00015514335903241478,
"loss": 3.5199127197265625,
"step": 3340
},
{
"epoch": 0.43905635648754915,
"grad_norm": 0.36538711190223694,
"learning_rate": 0.0001546301570635641,
"loss": 3.5380203247070314,
"step": 3350
},
{
"epoch": 0.44036697247706424,
"grad_norm": 0.42367586493492126,
"learning_rate": 0.00015411642536366984,
"loss": 3.522888946533203,
"step": 3360
},
{
"epoch": 0.4416775884665793,
"grad_norm": 0.39081627130508423,
"learning_rate": 0.00015360217311728103,
"loss": 3.547618865966797,
"step": 3370
},
{
"epoch": 0.4429882044560944,
"grad_norm": 0.32569605112075806,
"learning_rate": 0.00015308740951825301,
"loss": 3.4833248138427733,
"step": 3380
},
{
"epoch": 0.4442988204456094,
"grad_norm": 0.35152286291122437,
"learning_rate": 0.00015257214376958322,
"loss": 3.4959945678710938,
"step": 3390
},
{
"epoch": 0.4456094364351245,
"grad_norm": 0.4807989001274109,
"learning_rate": 0.00015205638508324654,
"loss": 3.502777862548828,
"step": 3400
},
{
"epoch": 0.44692005242463956,
"grad_norm": 0.3716277480125427,
"learning_rate": 0.00015154014268003063,
"loss": 3.5264617919921877,
"step": 3410
},
{
"epoch": 0.44823066841415465,
"grad_norm": 0.34413373470306396,
"learning_rate": 0.00015102342578937118,
"loss": 3.51181640625,
"step": 3420
},
{
"epoch": 0.44954128440366975,
"grad_norm": 0.3799779713153839,
"learning_rate": 0.0001505062436491866,
"loss": 3.5248443603515627,
"step": 3430
},
{
"epoch": 0.4508519003931848,
"grad_norm": 0.39848634600639343,
"learning_rate": 0.0001499886055057134,
"loss": 3.538591766357422,
"step": 3440
},
{
"epoch": 0.4521625163826999,
"grad_norm": 0.38903528451919556,
"learning_rate": 0.00014947052061334022,
"loss": 3.579244613647461,
"step": 3450
},
{
"epoch": 0.4534731323722149,
"grad_norm": 0.345285564661026,
"learning_rate": 0.00014895199823444303,
"loss": 3.5041595458984376,
"step": 3460
},
{
"epoch": 0.45478374836173,
"grad_norm": 0.379616379737854,
"learning_rate": 0.00014843304763921902,
"loss": 3.5296497344970703,
"step": 3470
},
{
"epoch": 0.45609436435124506,
"grad_norm": 0.39554885029792786,
"learning_rate": 0.00014791367810552118,
"loss": 3.5207672119140625,
"step": 3480
},
{
"epoch": 0.45740498034076016,
"grad_norm": 0.3259015381336212,
"learning_rate": 0.00014739389891869225,
"loss": 3.5236297607421876,
"step": 3490
},
{
"epoch": 0.45871559633027525,
"grad_norm": 0.3829706013202667,
"learning_rate": 0.0001468737193713989,
"loss": 3.499440002441406,
"step": 3500
},
{
"epoch": 0.45871559633027525,
"eval_loss": 3.4763739109039307,
"eval_runtime": 0.6985,
"eval_samples_per_second": 183.241,
"eval_steps_per_second": 45.81,
"step": 3500
},
{
"epoch": 0.4600262123197903,
"grad_norm": 0.34715378284454346,
"learning_rate": 0.00014635314876346536,
"loss": 3.4787666320800783,
"step": 3510
},
{
"epoch": 0.4613368283093054,
"grad_norm": 0.34763696789741516,
"learning_rate": 0.00014583219640170738,
"loss": 3.475682830810547,
"step": 3520
},
{
"epoch": 0.46264744429882043,
"grad_norm": 0.3834385573863983,
"learning_rate": 0.00014531087159976573,
"loss": 3.4582134246826173,
"step": 3530
},
{
"epoch": 0.4639580602883355,
"grad_norm": 0.37635257840156555,
"learning_rate": 0.00014478918367793962,
"loss": 3.5125938415527345,
"step": 3540
},
{
"epoch": 0.46526867627785057,
"grad_norm": 0.38097307085990906,
"learning_rate": 0.00014426714196302023,
"loss": 3.4908729553222657,
"step": 3550
},
{
"epoch": 0.46657929226736566,
"grad_norm": 0.37288933992385864,
"learning_rate": 0.00014374475578812396,
"loss": 3.5178199768066407,
"step": 3560
},
{
"epoch": 0.46788990825688076,
"grad_norm": 0.38614487648010254,
"learning_rate": 0.00014322203449252532,
"loss": 3.493552398681641,
"step": 3570
},
{
"epoch": 0.4692005242463958,
"grad_norm": 0.31520113348960876,
"learning_rate": 0.00014269898742149035,
"loss": 3.4913848876953124,
"step": 3580
},
{
"epoch": 0.4705111402359109,
"grad_norm": 0.358929306268692,
"learning_rate": 0.0001421756239261091,
"loss": 3.509693145751953,
"step": 3590
},
{
"epoch": 0.47182175622542594,
"grad_norm": 0.3876150846481323,
"learning_rate": 0.0001416519533631289,
"loss": 3.5168502807617186,
"step": 3600
},
{
"epoch": 0.47313237221494103,
"grad_norm": 0.3398151099681854,
"learning_rate": 0.0001411279850947868,
"loss": 3.4790908813476564,
"step": 3610
},
{
"epoch": 0.4744429882044561,
"grad_norm": 0.34104910492897034,
"learning_rate": 0.00014060372848864224,
"loss": 3.4779521942138674,
"step": 3620
},
{
"epoch": 0.47575360419397117,
"grad_norm": 0.3506695330142975,
"learning_rate": 0.00014007919291740954,
"loss": 3.495278549194336,
"step": 3630
},
{
"epoch": 0.47706422018348627,
"grad_norm": 0.36414405703544617,
"learning_rate": 0.00013955438775879045,
"loss": 3.4941474914550783,
"step": 3640
},
{
"epoch": 0.4783748361730013,
"grad_norm": 0.3429199755191803,
"learning_rate": 0.00013902932239530646,
"loss": 3.469829559326172,
"step": 3650
},
{
"epoch": 0.4796854521625164,
"grad_norm": 0.34807538986206055,
"learning_rate": 0.00013850400621413103,
"loss": 3.4917461395263674,
"step": 3660
},
{
"epoch": 0.48099606815203144,
"grad_norm": 0.36988309025764465,
"learning_rate": 0.00013797844860692163,
"loss": 3.506679153442383,
"step": 3670
},
{
"epoch": 0.48230668414154654,
"grad_norm": 0.35055309534072876,
"learning_rate": 0.00013745265896965219,
"loss": 3.500745391845703,
"step": 3680
},
{
"epoch": 0.4836173001310616,
"grad_norm": 0.3981817662715912,
"learning_rate": 0.0001369266467024446,
"loss": 3.4869926452636717,
"step": 3690
},
{
"epoch": 0.4849279161205767,
"grad_norm": 0.3786320090293884,
"learning_rate": 0.0001364004212094013,
"loss": 3.488050842285156,
"step": 3700
},
{
"epoch": 0.48623853211009177,
"grad_norm": 0.3471132516860962,
"learning_rate": 0.00013587399189843655,
"loss": 3.492092514038086,
"step": 3710
},
{
"epoch": 0.4875491480996068,
"grad_norm": 0.3803897500038147,
"learning_rate": 0.0001353473681811086,
"loss": 3.478464126586914,
"step": 3720
},
{
"epoch": 0.4888597640891219,
"grad_norm": 0.36248594522476196,
"learning_rate": 0.0001348205594724513,
"loss": 3.4785011291503904,
"step": 3730
},
{
"epoch": 0.49017038007863695,
"grad_norm": 0.35516679286956787,
"learning_rate": 0.0001342935751908057,
"loss": 3.4770259857177734,
"step": 3740
},
{
"epoch": 0.49148099606815204,
"grad_norm": 0.36115631461143494,
"learning_rate": 0.00013376642475765194,
"loss": 3.467881774902344,
"step": 3750
},
{
"epoch": 0.4927916120576671,
"grad_norm": 0.3623756468296051,
"learning_rate": 0.00013323911759744047,
"loss": 3.5005516052246093,
"step": 3760
},
{
"epoch": 0.4941022280471822,
"grad_norm": 0.3679109811782837,
"learning_rate": 0.00013271166313742386,
"loss": 3.458334732055664,
"step": 3770
},
{
"epoch": 0.4954128440366973,
"grad_norm": 0.40925225615501404,
"learning_rate": 0.00013218407080748797,
"loss": 3.4681320190429688,
"step": 3780
},
{
"epoch": 0.4967234600262123,
"grad_norm": 0.368941992521286,
"learning_rate": 0.00013165635003998367,
"loss": 3.4709869384765626,
"step": 3790
},
{
"epoch": 0.4980340760157274,
"grad_norm": 0.3967722952365875,
"learning_rate": 0.00013112851026955786,
"loss": 3.4564468383789064,
"step": 3800
},
{
"epoch": 0.49934469200524245,
"grad_norm": 0.37133631110191345,
"learning_rate": 0.00013060056093298529,
"loss": 3.4897750854492187,
"step": 3810
},
{
"epoch": 0.5006553079947575,
"grad_norm": 0.3902031481266022,
"learning_rate": 0.0001300725114689991,
"loss": 3.504320526123047,
"step": 3820
},
{
"epoch": 0.5019659239842726,
"grad_norm": 0.37370365858078003,
"learning_rate": 0.00012954437131812308,
"loss": 3.5297950744628905,
"step": 3830
},
{
"epoch": 0.5032765399737876,
"grad_norm": 0.3488170802593231,
"learning_rate": 0.00012901614992250177,
"loss": 3.4684242248535155,
"step": 3840
},
{
"epoch": 0.5045871559633027,
"grad_norm": 0.3804292678833008,
"learning_rate": 0.0001284878567257326,
"loss": 3.476953887939453,
"step": 3850
},
{
"epoch": 0.5058977719528178,
"grad_norm": 0.3592618405818939,
"learning_rate": 0.00012795950117269644,
"loss": 3.4947891235351562,
"step": 3860
},
{
"epoch": 0.5072083879423329,
"grad_norm": 0.3395201861858368,
"learning_rate": 0.00012743109270938916,
"loss": 3.4853584289550783,
"step": 3870
},
{
"epoch": 0.508519003931848,
"grad_norm": 0.4130866825580597,
"learning_rate": 0.00012690264078275246,
"loss": 3.514727783203125,
"step": 3880
},
{
"epoch": 0.509829619921363,
"grad_norm": 0.31629058718681335,
"learning_rate": 0.00012637415484050503,
"loss": 3.499773406982422,
"step": 3890
},
{
"epoch": 0.5111402359108781,
"grad_norm": 0.32408541440963745,
"learning_rate": 0.0001258456443309738,
"loss": 3.4784332275390626,
"step": 3900
},
{
"epoch": 0.5124508519003932,
"grad_norm": 0.37189796566963196,
"learning_rate": 0.00012531711870292488,
"loss": 3.5128494262695313,
"step": 3910
},
{
"epoch": 0.5137614678899083,
"grad_norm": 0.3827756345272064,
"learning_rate": 0.0001247885874053947,
"loss": 3.489826202392578,
"step": 3920
},
{
"epoch": 0.5150720838794234,
"grad_norm": 0.3406204879283905,
"learning_rate": 0.00012426005988752092,
"loss": 3.474610137939453,
"step": 3930
},
{
"epoch": 0.5163826998689384,
"grad_norm": 0.3418485224246979,
"learning_rate": 0.00012373154559837377,
"loss": 3.5139701843261717,
"step": 3940
},
{
"epoch": 0.5176933158584535,
"grad_norm": 0.3592812418937683,
"learning_rate": 0.00012320305398678694,
"loss": 3.4879188537597656,
"step": 3950
},
{
"epoch": 0.5190039318479686,
"grad_norm": 0.3936125338077545,
"learning_rate": 0.0001226745945011887,
"loss": 3.4845611572265627,
"step": 3960
},
{
"epoch": 0.5203145478374837,
"grad_norm": 0.3792613744735718,
"learning_rate": 0.00012214617658943294,
"loss": 3.479914093017578,
"step": 3970
},
{
"epoch": 0.5216251638269986,
"grad_norm": 0.34662505984306335,
"learning_rate": 0.00012161780969863019,
"loss": 3.476584243774414,
"step": 3980
},
{
"epoch": 0.5229357798165137,
"grad_norm": 0.35217994451522827,
"learning_rate": 0.00012108950327497903,
"loss": 3.476770782470703,
"step": 3990
},
{
"epoch": 0.5242463958060288,
"grad_norm": 0.4655326306819916,
"learning_rate": 0.00012056126676359688,
"loss": 3.4810821533203127,
"step": 4000
},
{
"epoch": 0.5242463958060288,
"eval_loss": 3.433178186416626,
"eval_runtime": 0.6985,
"eval_samples_per_second": 183.24,
"eval_steps_per_second": 45.81,
"step": 4000
},
{
"epoch": 0.5255570117955439,
"grad_norm": 0.33986395597457886,
"learning_rate": 0.00012003310960835128,
"loss": 3.504052734375,
"step": 4010
},
{
"epoch": 0.526867627785059,
"grad_norm": 0.38196566700935364,
"learning_rate": 0.00011950504125169095,
"loss": 3.477123260498047,
"step": 4020
},
{
"epoch": 0.528178243774574,
"grad_norm": 0.352018266916275,
"learning_rate": 0.00011897707113447724,
"loss": 3.4726978302001954,
"step": 4030
},
{
"epoch": 0.5294888597640891,
"grad_norm": 0.3507325053215027,
"learning_rate": 0.00011844920869581503,
"loss": 3.454428482055664,
"step": 4040
},
{
"epoch": 0.5307994757536042,
"grad_norm": 0.34447765350341797,
"learning_rate": 0.00011792146337288411,
"loss": 3.4298377990722657,
"step": 4050
},
{
"epoch": 0.5321100917431193,
"grad_norm": 0.398247629404068,
"learning_rate": 0.00011739384460077061,
"loss": 3.501710891723633,
"step": 4060
},
{
"epoch": 0.5334207077326344,
"grad_norm": 0.3600616455078125,
"learning_rate": 0.00011686636181229795,
"loss": 3.458553695678711,
"step": 4070
},
{
"epoch": 0.5347313237221494,
"grad_norm": 0.4164774715900421,
"learning_rate": 0.00011633902443785856,
"loss": 3.445058822631836,
"step": 4080
},
{
"epoch": 0.5360419397116645,
"grad_norm": 0.44055747985839844,
"learning_rate": 0.0001158118419052451,
"loss": 3.505634307861328,
"step": 4090
},
{
"epoch": 0.5373525557011796,
"grad_norm": 0.3581311106681824,
"learning_rate": 0.000115284823639482,
"loss": 3.5074691772460938,
"step": 4100
},
{
"epoch": 0.5386631716906947,
"grad_norm": 0.3456391394138336,
"learning_rate": 0.00011475797906265689,
"loss": 3.469898223876953,
"step": 4110
},
{
"epoch": 0.5399737876802096,
"grad_norm": 0.3439826965332031,
"learning_rate": 0.0001142313175937521,
"loss": 3.46533203125,
"step": 4120
},
{
"epoch": 0.5412844036697247,
"grad_norm": 0.5179669260978699,
"learning_rate": 0.0001137048486484764,
"loss": 3.503631591796875,
"step": 4130
},
{
"epoch": 0.5425950196592398,
"grad_norm": 0.3679772615432739,
"learning_rate": 0.00011317858163909662,
"loss": 3.5136138916015627,
"step": 4140
},
{
"epoch": 0.5439056356487549,
"grad_norm": 0.31712621450424194,
"learning_rate": 0.00011265252597426932,
"loss": 3.477741241455078,
"step": 4150
},
{
"epoch": 0.54521625163827,
"grad_norm": 0.33221423625946045,
"learning_rate": 0.00011212669105887274,
"loss": 3.448292922973633,
"step": 4160
},
{
"epoch": 0.546526867627785,
"grad_norm": 0.34244126081466675,
"learning_rate": 0.00011160108629383824,
"loss": 3.4491950988769533,
"step": 4170
},
{
"epoch": 0.5478374836173001,
"grad_norm": 0.3568786084651947,
"learning_rate": 0.00011107572107598285,
"loss": 3.461478424072266,
"step": 4180
},
{
"epoch": 0.5491480996068152,
"grad_norm": 0.40663549304008484,
"learning_rate": 0.00011055060479784076,
"loss": 3.449353790283203,
"step": 4190
},
{
"epoch": 0.5504587155963303,
"grad_norm": 0.3818509578704834,
"learning_rate": 0.00011002574684749572,
"loss": 3.4897056579589845,
"step": 4200
},
{
"epoch": 0.5517693315858454,
"grad_norm": 0.3730104863643646,
"learning_rate": 0.00010950115660841299,
"loss": 3.468395233154297,
"step": 4210
},
{
"epoch": 0.5530799475753604,
"grad_norm": 0.367310494184494,
"learning_rate": 0.00010897684345927163,
"loss": 3.488247299194336,
"step": 4220
},
{
"epoch": 0.5543905635648755,
"grad_norm": 0.450967937707901,
"learning_rate": 0.00010845281677379687,
"loss": 3.433974838256836,
"step": 4230
},
{
"epoch": 0.5557011795543906,
"grad_norm": 0.33636024594306946,
"learning_rate": 0.00010792908592059261,
"loss": 3.512376403808594,
"step": 4240
},
{
"epoch": 0.5570117955439057,
"grad_norm": 0.34425368905067444,
"learning_rate": 0.00010740566026297374,
"loss": 3.4768524169921875,
"step": 4250
},
{
"epoch": 0.5583224115334207,
"grad_norm": 0.35946354269981384,
"learning_rate": 0.00010688254915879888,
"loss": 3.4618865966796877,
"step": 4260
},
{
"epoch": 0.5596330275229358,
"grad_norm": 0.3705649673938751,
"learning_rate": 0.0001063597619603029,
"loss": 3.4575927734375,
"step": 4270
},
{
"epoch": 0.5609436435124509,
"grad_norm": 0.35676929354667664,
"learning_rate": 0.00010583730801393005,
"loss": 3.459075927734375,
"step": 4280
},
{
"epoch": 0.562254259501966,
"grad_norm": 0.4031120538711548,
"learning_rate": 0.00010531519666016654,
"loss": 3.4172103881835936,
"step": 4290
},
{
"epoch": 0.563564875491481,
"grad_norm": 0.37692543864250183,
"learning_rate": 0.00010479343723337365,
"loss": 3.4404541015625,
"step": 4300
},
{
"epoch": 0.564875491480996,
"grad_norm": 0.35645318031311035,
"learning_rate": 0.00010427203906162102,
"loss": 3.452524948120117,
"step": 4310
},
{
"epoch": 0.5661861074705111,
"grad_norm": 0.42808640003204346,
"learning_rate": 0.00010375101146651958,
"loss": 3.48497314453125,
"step": 4320
},
{
"epoch": 0.5674967234600262,
"grad_norm": 0.34046339988708496,
"learning_rate": 0.00010323036376305512,
"loss": 3.426773452758789,
"step": 4330
},
{
"epoch": 0.5688073394495413,
"grad_norm": 0.34263497591018677,
"learning_rate": 0.00010271010525942162,
"loss": 3.453144836425781,
"step": 4340
},
{
"epoch": 0.5701179554390564,
"grad_norm": 0.3416344225406647,
"learning_rate": 0.00010219024525685505,
"loss": 3.4319427490234373,
"step": 4350
},
{
"epoch": 0.5714285714285714,
"grad_norm": 0.3529037833213806,
"learning_rate": 0.00010167079304946684,
"loss": 3.446512985229492,
"step": 4360
},
{
"epoch": 0.5727391874180865,
"grad_norm": 0.39450159668922424,
"learning_rate": 0.00010115175792407777,
"loss": 3.4684822082519533,
"step": 4370
},
{
"epoch": 0.5740498034076016,
"grad_norm": 0.3465752601623535,
"learning_rate": 0.00010063314916005204,
"loss": 3.4847396850585937,
"step": 4380
},
{
"epoch": 0.5753604193971167,
"grad_norm": 0.34021201729774475,
"learning_rate": 0.00010011497602913135,
"loss": 3.4533958435058594,
"step": 4390
},
{
"epoch": 0.5766710353866317,
"grad_norm": 0.33759158849716187,
"learning_rate": 9.959724779526902e-05,
"loss": 3.441851806640625,
"step": 4400
},
{
"epoch": 0.5779816513761468,
"grad_norm": 0.3410870432853699,
"learning_rate": 9.907997371446464e-05,
"loss": 3.455671691894531,
"step": 4410
},
{
"epoch": 0.5792922673656619,
"grad_norm": 0.32743537425994873,
"learning_rate": 9.856316303459808e-05,
"loss": 3.446603012084961,
"step": 4420
},
{
"epoch": 0.580602883355177,
"grad_norm": 0.35926544666290283,
"learning_rate": 9.804682499526479e-05,
"loss": 3.421849822998047,
"step": 4430
},
{
"epoch": 0.581913499344692,
"grad_norm": 0.332614541053772,
"learning_rate": 9.753096882761012e-05,
"loss": 3.4696685791015627,
"step": 4440
},
{
"epoch": 0.583224115334207,
"grad_norm": 0.34293821454048157,
"learning_rate": 9.701560375416457e-05,
"loss": 3.46300048828125,
"step": 4450
},
{
"epoch": 0.5845347313237221,
"grad_norm": 0.33514559268951416,
"learning_rate": 9.650073898867879e-05,
"loss": 3.4373500823974608,
"step": 4460
},
{
"epoch": 0.5858453473132372,
"grad_norm": 0.3574112057685852,
"learning_rate": 9.598638373595879e-05,
"loss": 3.4546417236328124,
"step": 4470
},
{
"epoch": 0.5871559633027523,
"grad_norm": 0.3636513948440552,
"learning_rate": 9.54725471917015e-05,
"loss": 3.4387813568115235,
"step": 4480
},
{
"epoch": 0.5884665792922673,
"grad_norm": 0.35861721634864807,
"learning_rate": 9.49592385423304e-05,
"loss": 3.4477272033691406,
"step": 4490
},
{
"epoch": 0.5897771952817824,
"grad_norm": 0.3176049292087555,
"learning_rate": 9.444646696483114e-05,
"loss": 3.4042896270751952,
"step": 4500
},
{
"epoch": 0.5897771952817824,
"eval_loss": 3.397040367126465,
"eval_runtime": 0.6514,
"eval_samples_per_second": 196.506,
"eval_steps_per_second": 49.127,
"step": 4500
},
{
"epoch": 0.5910878112712975,
"grad_norm": 0.355270117521286,
"learning_rate": 9.393424162658757e-05,
"loss": 3.4663818359375,
"step": 4510
},
{
"epoch": 0.5923984272608126,
"grad_norm": 0.3310716450214386,
"learning_rate": 9.34225716852177e-05,
"loss": 3.4160083770751952,
"step": 4520
},
{
"epoch": 0.5937090432503277,
"grad_norm": 0.338791161775589,
"learning_rate": 9.291146628841033e-05,
"loss": 3.445051574707031,
"step": 4530
},
{
"epoch": 0.5950196592398427,
"grad_norm": 0.36331963539123535,
"learning_rate": 9.240093457376105e-05,
"loss": 3.461710739135742,
"step": 4540
},
{
"epoch": 0.5963302752293578,
"grad_norm": 0.347826212644577,
"learning_rate": 9.189098566860921e-05,
"loss": 3.420879364013672,
"step": 4550
},
{
"epoch": 0.5976408912188729,
"grad_norm": 0.3780054748058319,
"learning_rate": 9.138162868987468e-05,
"loss": 3.442432403564453,
"step": 4560
},
{
"epoch": 0.598951507208388,
"grad_norm": 0.39614027738571167,
"learning_rate": 9.087287274389466e-05,
"loss": 3.4345935821533202,
"step": 4570
},
{
"epoch": 0.6002621231979031,
"grad_norm": 0.33810099959373474,
"learning_rate": 9.036472692626113e-05,
"loss": 3.41512451171875,
"step": 4580
},
{
"epoch": 0.601572739187418,
"grad_norm": 0.34188082814216614,
"learning_rate": 8.985720032165808e-05,
"loss": 3.4464752197265627,
"step": 4590
},
{
"epoch": 0.6028833551769331,
"grad_norm": 0.3292509913444519,
"learning_rate": 8.935030200369926e-05,
"loss": 3.440515899658203,
"step": 4600
},
{
"epoch": 0.6041939711664482,
"grad_norm": 0.3394361138343811,
"learning_rate": 8.884404103476566e-05,
"loss": 3.419635772705078,
"step": 4610
},
{
"epoch": 0.6055045871559633,
"grad_norm": 0.3512326180934906,
"learning_rate": 8.83384264658438e-05,
"loss": 3.4148048400878905,
"step": 4620
},
{
"epoch": 0.6068152031454783,
"grad_norm": 0.3446638882160187,
"learning_rate": 8.783346733636376e-05,
"loss": 3.4195236206054687,
"step": 4630
},
{
"epoch": 0.6081258191349934,
"grad_norm": 0.33051419258117676,
"learning_rate": 8.732917267403767e-05,
"loss": 3.427733612060547,
"step": 4640
},
{
"epoch": 0.6094364351245085,
"grad_norm": 0.3206946849822998,
"learning_rate": 8.682555149469814e-05,
"loss": 3.4442039489746095,
"step": 4650
},
{
"epoch": 0.6107470511140236,
"grad_norm": 0.32173067331314087,
"learning_rate": 8.632261280213725e-05,
"loss": 3.4339874267578123,
"step": 4660
},
{
"epoch": 0.6120576671035387,
"grad_norm": 0.3285246789455414,
"learning_rate": 8.582036558794543e-05,
"loss": 3.428564453125,
"step": 4670
},
{
"epoch": 0.6133682830930537,
"grad_norm": 0.3776794672012329,
"learning_rate": 8.53188188313509e-05,
"loss": 3.4429168701171875,
"step": 4680
},
{
"epoch": 0.6146788990825688,
"grad_norm": 0.3147306442260742,
"learning_rate": 8.481798149905895e-05,
"loss": 3.4195117950439453,
"step": 4690
},
{
"epoch": 0.6159895150720839,
"grad_norm": 0.3394879102706909,
"learning_rate": 8.431786254509183e-05,
"loss": 3.41383056640625,
"step": 4700
},
{
"epoch": 0.617300131061599,
"grad_norm": 0.347411185503006,
"learning_rate": 8.38184709106283e-05,
"loss": 3.4398727416992188,
"step": 4710
},
{
"epoch": 0.6186107470511141,
"grad_norm": 0.38795945048332214,
"learning_rate": 8.331981552384425e-05,
"loss": 3.422657775878906,
"step": 4720
},
{
"epoch": 0.6199213630406291,
"grad_norm": 0.33243945240974426,
"learning_rate": 8.282190529975275e-05,
"loss": 3.4115135192871096,
"step": 4730
},
{
"epoch": 0.6212319790301442,
"grad_norm": 0.3807089924812317,
"learning_rate": 8.232474914004486e-05,
"loss": 3.4477569580078127,
"step": 4740
},
{
"epoch": 0.6225425950196593,
"grad_norm": 0.3600168824195862,
"learning_rate": 8.182835593293034e-05,
"loss": 3.4017192840576174,
"step": 4750
},
{
"epoch": 0.6238532110091743,
"grad_norm": 0.33293479681015015,
"learning_rate": 8.133273455297873e-05,
"loss": 3.4247257232666017,
"step": 4760
},
{
"epoch": 0.6251638269986893,
"grad_norm": 0.34662172198295593,
"learning_rate": 8.083789386096089e-05,
"loss": 3.426392364501953,
"step": 4770
},
{
"epoch": 0.6264744429882044,
"grad_norm": 0.3521794080734253,
"learning_rate": 8.034384270369041e-05,
"loss": 3.38433723449707,
"step": 4780
},
{
"epoch": 0.6277850589777195,
"grad_norm": 0.4303857386112213,
"learning_rate": 7.98505899138655e-05,
"loss": 3.4620620727539064,
"step": 4790
},
{
"epoch": 0.6290956749672346,
"grad_norm": 0.335654079914093,
"learning_rate": 7.935814430991109e-05,
"loss": 3.3878963470458983,
"step": 4800
},
{
"epoch": 0.6304062909567497,
"grad_norm": 0.3300749361515045,
"learning_rate": 7.886651469582107e-05,
"loss": 3.4436004638671873,
"step": 4810
},
{
"epoch": 0.6317169069462647,
"grad_norm": 0.3882623612880707,
"learning_rate": 7.83757098610011e-05,
"loss": 3.4116146087646486,
"step": 4820
},
{
"epoch": 0.6330275229357798,
"grad_norm": 0.35634687542915344,
"learning_rate": 7.788573858011125e-05,
"loss": 3.4195274353027343,
"step": 4830
},
{
"epoch": 0.6343381389252949,
"grad_norm": 0.3425414562225342,
"learning_rate": 7.739660961290928e-05,
"loss": 3.4286224365234377,
"step": 4840
},
{
"epoch": 0.63564875491481,
"grad_norm": 0.3372134566307068,
"learning_rate": 7.690833170409405e-05,
"loss": 3.406309127807617,
"step": 4850
},
{
"epoch": 0.6369593709043251,
"grad_norm": 0.357086181640625,
"learning_rate": 7.642091358314893e-05,
"loss": 3.3772621154785156,
"step": 4860
},
{
"epoch": 0.6382699868938401,
"grad_norm": 0.3391537368297577,
"learning_rate": 7.593436396418601e-05,
"loss": 3.3964859008789063,
"step": 4870
},
{
"epoch": 0.6395806028833552,
"grad_norm": 0.43100473284721375,
"learning_rate": 7.544869154579024e-05,
"loss": 3.4649642944335937,
"step": 4880
},
{
"epoch": 0.6408912188728703,
"grad_norm": 0.32557618618011475,
"learning_rate": 7.496390501086385e-05,
"loss": 3.39740104675293,
"step": 4890
},
{
"epoch": 0.6422018348623854,
"grad_norm": 0.35657286643981934,
"learning_rate": 7.44800130264712e-05,
"loss": 3.4110061645507814,
"step": 4900
},
{
"epoch": 0.6435124508519003,
"grad_norm": 0.3218068480491638,
"learning_rate": 7.399702424368368e-05,
"loss": 3.3983528137207033,
"step": 4910
},
{
"epoch": 0.6448230668414154,
"grad_norm": 0.32021915912628174,
"learning_rate": 7.35149472974252e-05,
"loss": 3.4089466094970704,
"step": 4920
},
{
"epoch": 0.6461336828309305,
"grad_norm": 0.3458782732486725,
"learning_rate": 7.303379080631784e-05,
"loss": 3.4090194702148438,
"step": 4930
},
{
"epoch": 0.6474442988204456,
"grad_norm": 0.3527551293373108,
"learning_rate": 7.255356337252752e-05,
"loss": 3.387525177001953,
"step": 4940
},
{
"epoch": 0.6487549148099607,
"grad_norm": 0.3285190761089325,
"learning_rate": 7.207427358161063e-05,
"loss": 3.408121109008789,
"step": 4950
},
{
"epoch": 0.6500655307994757,
"grad_norm": 0.3367261588573456,
"learning_rate": 7.159593000235997e-05,
"loss": 3.400335693359375,
"step": 4960
},
{
"epoch": 0.6513761467889908,
"grad_norm": 0.3399582505226135,
"learning_rate": 7.111854118665203e-05,
"loss": 3.4165496826171875,
"step": 4970
},
{
"epoch": 0.6526867627785059,
"grad_norm": 0.35921865701675415,
"learning_rate": 7.064211566929403e-05,
"loss": 3.389024353027344,
"step": 4980
},
{
"epoch": 0.653997378768021,
"grad_norm": 0.3397617042064667,
"learning_rate": 7.016666196787111e-05,
"loss": 3.394414520263672,
"step": 4990
},
{
"epoch": 0.6553079947575361,
"grad_norm": 0.3500848114490509,
"learning_rate": 6.969218858259423e-05,
"loss": 3.3963043212890627,
"step": 5000
},
{
"epoch": 0.6553079947575361,
"eval_loss": 3.369736671447754,
"eval_runtime": 0.6527,
"eval_samples_per_second": 196.119,
"eval_steps_per_second": 49.03,
"step": 5000
},
{
"epoch": 0.6566186107470511,
"grad_norm": 0.3623511493206024,
"learning_rate": 6.921870399614803e-05,
"loss": 3.4134674072265625,
"step": 5010
},
{
"epoch": 0.6579292267365662,
"grad_norm": 0.35100898146629333,
"learning_rate": 6.874621667353953e-05,
"loss": 3.416904830932617,
"step": 5020
},
{
"epoch": 0.6592398427260813,
"grad_norm": 0.35833802819252014,
"learning_rate": 6.827473506194643e-05,
"loss": 3.409159851074219,
"step": 5030
},
{
"epoch": 0.6605504587155964,
"grad_norm": 0.35042181611061096,
"learning_rate": 6.780426759056622e-05,
"loss": 3.4056846618652346,
"step": 5040
},
{
"epoch": 0.6618610747051114,
"grad_norm": 0.33665770292282104,
"learning_rate": 6.733482267046551e-05,
"loss": 3.414917755126953,
"step": 5050
},
{
"epoch": 0.6631716906946264,
"grad_norm": 0.35702770948410034,
"learning_rate": 6.686640869442962e-05,
"loss": 3.4223007202148437,
"step": 5060
},
{
"epoch": 0.6644823066841415,
"grad_norm": 0.3478568196296692,
"learning_rate": 6.639903403681258e-05,
"loss": 3.4018138885498046,
"step": 5070
},
{
"epoch": 0.6657929226736566,
"grad_norm": 0.37595197558403015,
"learning_rate": 6.59327070533873e-05,
"loss": 3.415419006347656,
"step": 5080
},
{
"epoch": 0.6671035386631717,
"grad_norm": 0.3227323591709137,
"learning_rate": 6.546743608119633e-05,
"loss": 3.3868553161621096,
"step": 5090
},
{
"epoch": 0.6684141546526867,
"grad_norm": 0.4119921624660492,
"learning_rate": 6.500322943840273e-05,
"loss": 3.4066440582275392,
"step": 5100
},
{
"epoch": 0.6697247706422018,
"grad_norm": 0.3280180096626282,
"learning_rate": 6.454009542414136e-05,
"loss": 3.4018421173095703,
"step": 5110
},
{
"epoch": 0.6710353866317169,
"grad_norm": 0.34080690145492554,
"learning_rate": 6.407804231837048e-05,
"loss": 3.4216339111328127,
"step": 5120
},
{
"epoch": 0.672346002621232,
"grad_norm": 0.3441828787326813,
"learning_rate": 6.361707838172382e-05,
"loss": 3.4043300628662108,
"step": 5130
},
{
"epoch": 0.6736566186107471,
"grad_norm": 0.3287142813205719,
"learning_rate": 6.315721185536274e-05,
"loss": 3.410292053222656,
"step": 5140
},
{
"epoch": 0.6749672346002621,
"grad_norm": 0.3822394013404846,
"learning_rate": 6.26984509608292e-05,
"loss": 3.417401885986328,
"step": 5150
},
{
"epoch": 0.6762778505897772,
"grad_norm": 0.3449343740940094,
"learning_rate": 6.224080389989824e-05,
"loss": 3.3693260192871093,
"step": 5160
},
{
"epoch": 0.6775884665792923,
"grad_norm": 0.49308013916015625,
"learning_rate": 6.178427885443187e-05,
"loss": 3.375142288208008,
"step": 5170
},
{
"epoch": 0.6788990825688074,
"grad_norm": 0.3545759618282318,
"learning_rate": 6.132888398623251e-05,
"loss": 3.386444091796875,
"step": 5180
},
{
"epoch": 0.6802096985583224,
"grad_norm": 0.36718955636024475,
"learning_rate": 6.08746274368973e-05,
"loss": 3.389984130859375,
"step": 5190
},
{
"epoch": 0.6815203145478375,
"grad_norm": 0.36378800868988037,
"learning_rate": 6.042151732767226e-05,
"loss": 3.41534423828125,
"step": 5200
},
{
"epoch": 0.6828309305373526,
"grad_norm": 0.354679673910141,
"learning_rate": 5.996956175930718e-05,
"loss": 3.4176490783691404,
"step": 5210
},
{
"epoch": 0.6841415465268676,
"grad_norm": 0.3615143299102783,
"learning_rate": 5.9518768811910925e-05,
"loss": 3.3957984924316404,
"step": 5220
},
{
"epoch": 0.6854521625163827,
"grad_norm": 0.31297728419303894,
"learning_rate": 5.906914654480701e-05,
"loss": 3.39869384765625,
"step": 5230
},
{
"epoch": 0.6867627785058977,
"grad_norm": 0.31141722202301025,
"learning_rate": 5.86207029963893e-05,
"loss": 3.3952251434326173,
"step": 5240
},
{
"epoch": 0.6880733944954128,
"grad_norm": 0.3368931710720062,
"learning_rate": 5.8173446183978294e-05,
"loss": 3.392802429199219,
"step": 5250
},
{
"epoch": 0.6893840104849279,
"grad_norm": 0.3885488510131836,
"learning_rate": 5.772738410367802e-05,
"loss": 3.4020896911621095,
"step": 5260
},
{
"epoch": 0.690694626474443,
"grad_norm": 0.3572815954685211,
"learning_rate": 5.7282524730233045e-05,
"loss": 3.4177703857421875,
"step": 5270
},
{
"epoch": 0.6920052424639581,
"grad_norm": 0.31040751934051514,
"learning_rate": 5.6838876016885686e-05,
"loss": 3.3765377044677733,
"step": 5280
},
{
"epoch": 0.6933158584534731,
"grad_norm": 0.34986966848373413,
"learning_rate": 5.639644589523402e-05,
"loss": 3.405791091918945,
"step": 5290
},
{
"epoch": 0.6946264744429882,
"grad_norm": 0.3639046549797058,
"learning_rate": 5.595524227508986e-05,
"loss": 3.399429702758789,
"step": 5300
},
{
"epoch": 0.6959370904325033,
"grad_norm": 0.4057208001613617,
"learning_rate": 5.551527304433775e-05,
"loss": 3.4495635986328126,
"step": 5310
},
{
"epoch": 0.6972477064220184,
"grad_norm": 0.3493952453136444,
"learning_rate": 5.50765460687935e-05,
"loss": 3.393666076660156,
"step": 5320
},
{
"epoch": 0.6985583224115334,
"grad_norm": 0.43060457706451416,
"learning_rate": 5.4639069192063866e-05,
"loss": 3.406134033203125,
"step": 5330
},
{
"epoch": 0.6998689384010485,
"grad_norm": 0.34414297342300415,
"learning_rate": 5.420285023540611e-05,
"loss": 3.398945617675781,
"step": 5340
},
{
"epoch": 0.7011795543905636,
"grad_norm": 0.38662222027778625,
"learning_rate": 5.376789699758833e-05,
"loss": 3.39422492980957,
"step": 5350
},
{
"epoch": 0.7024901703800787,
"grad_norm": 0.3366408944129944,
"learning_rate": 5.3334217254749945e-05,
"loss": 3.3609878540039064,
"step": 5360
},
{
"epoch": 0.7038007863695938,
"grad_norm": 0.32966917753219604,
"learning_rate": 5.290181876026272e-05,
"loss": 3.4071975708007813,
"step": 5370
},
{
"epoch": 0.7051114023591087,
"grad_norm": 0.3106803894042969,
"learning_rate": 5.2470709244592104e-05,
"loss": 3.3670631408691407,
"step": 5380
},
{
"epoch": 0.7064220183486238,
"grad_norm": 0.3144097328186035,
"learning_rate": 5.204089641515905e-05,
"loss": 3.383616256713867,
"step": 5390
},
{
"epoch": 0.7077326343381389,
"grad_norm": 0.3441525101661682,
"learning_rate": 5.161238795620225e-05,
"loss": 3.3808822631835938,
"step": 5400
},
{
"epoch": 0.709043250327654,
"grad_norm": 0.32820984721183777,
"learning_rate": 5.118519152864068e-05,
"loss": 3.3869140625,
"step": 5410
},
{
"epoch": 0.7103538663171691,
"grad_norm": 0.36829277873039246,
"learning_rate": 5.075931476993671e-05,
"loss": 3.442747116088867,
"step": 5420
},
{
"epoch": 0.7116644823066841,
"grad_norm": 0.33520880341529846,
"learning_rate": 5.033476529395947e-05,
"loss": 3.4088211059570312,
"step": 5430
},
{
"epoch": 0.7129750982961992,
"grad_norm": 0.3686036169528961,
"learning_rate": 4.9911550690849004e-05,
"loss": 3.389745330810547,
"step": 5440
},
{
"epoch": 0.7142857142857143,
"grad_norm": 0.3428293466567993,
"learning_rate": 4.948967852688006e-05,
"loss": 3.390821838378906,
"step": 5450
},
{
"epoch": 0.7155963302752294,
"grad_norm": 0.31483086943626404,
"learning_rate": 4.906915634432729e-05,
"loss": 3.4046356201171877,
"step": 5460
},
{
"epoch": 0.7169069462647444,
"grad_norm": 0.3391151428222656,
"learning_rate": 4.8649991661330175e-05,
"loss": 3.370349884033203,
"step": 5470
},
{
"epoch": 0.7182175622542595,
"grad_norm": 0.35527873039245605,
"learning_rate": 4.823219197175879e-05,
"loss": 3.3873008728027343,
"step": 5480
},
{
"epoch": 0.7195281782437746,
"grad_norm": 0.33960920572280884,
"learning_rate": 4.781576474507965e-05,
"loss": 3.357405090332031,
"step": 5490
},
{
"epoch": 0.7208387942332897,
"grad_norm": 0.37227174639701843,
"learning_rate": 4.740071742622211e-05,
"loss": 3.3784297943115233,
"step": 5500
},
{
"epoch": 0.7208387942332897,
"eval_loss": 3.3471105098724365,
"eval_runtime": 0.653,
"eval_samples_per_second": 196.023,
"eval_steps_per_second": 49.006,
"step": 5500
},
{
"epoch": 0.7221494102228048,
"grad_norm": 0.3386659622192383,
"learning_rate": 4.6987057435445515e-05,
"loss": 3.3939659118652346,
"step": 5510
},
{
"epoch": 0.7234600262123198,
"grad_norm": 0.32990965247154236,
"learning_rate": 4.6574792168206484e-05,
"loss": 3.4213134765625,
"step": 5520
},
{
"epoch": 0.7247706422018348,
"grad_norm": 0.35072508454322815,
"learning_rate": 4.6163928995026486e-05,
"loss": 3.379620361328125,
"step": 5530
},
{
"epoch": 0.7260812581913499,
"grad_norm": 0.33484020829200745,
"learning_rate": 4.5754475261360285e-05,
"loss": 3.384080505371094,
"step": 5540
},
{
"epoch": 0.727391874180865,
"grad_norm": 0.34197279810905457,
"learning_rate": 4.5346438287464416e-05,
"loss": 3.4106285095214846,
"step": 5550
},
{
"epoch": 0.72870249017038,
"grad_norm": 0.32156234979629517,
"learning_rate": 4.493982536826663e-05,
"loss": 3.367784881591797,
"step": 5560
},
{
"epoch": 0.7300131061598951,
"grad_norm": 0.33544281125068665,
"learning_rate": 4.453464377323513e-05,
"loss": 3.393235778808594,
"step": 5570
},
{
"epoch": 0.7313237221494102,
"grad_norm": 0.3419542610645294,
"learning_rate": 4.4130900746248815e-05,
"loss": 3.4026866912841798,
"step": 5580
},
{
"epoch": 0.7326343381389253,
"grad_norm": 0.325393408536911,
"learning_rate": 4.37286035054677e-05,
"loss": 3.3925567626953126,
"step": 5590
},
{
"epoch": 0.7339449541284404,
"grad_norm": 0.33149510622024536,
"learning_rate": 4.33277592432039e-05,
"loss": 3.4121665954589844,
"step": 5600
},
{
"epoch": 0.7352555701179554,
"grad_norm": 0.3202405571937561,
"learning_rate": 4.292837512579298e-05,
"loss": 3.387676239013672,
"step": 5610
},
{
"epoch": 0.7365661861074705,
"grad_norm": 0.3338731527328491,
"learning_rate": 4.253045829346594e-05,
"loss": 3.3993797302246094,
"step": 5620
},
{
"epoch": 0.7378768020969856,
"grad_norm": 0.319058895111084,
"learning_rate": 4.213401586022149e-05,
"loss": 3.3972732543945314,
"step": 5630
},
{
"epoch": 0.7391874180865007,
"grad_norm": 0.3269672095775604,
"learning_rate": 4.173905491369884e-05,
"loss": 3.4002712249755858,
"step": 5640
},
{
"epoch": 0.7404980340760158,
"grad_norm": 0.32364776730537415,
"learning_rate": 4.134558251505108e-05,
"loss": 3.373546600341797,
"step": 5650
},
{
"epoch": 0.7418086500655308,
"grad_norm": 0.31596776843070984,
"learning_rate": 4.095360569881884e-05,
"loss": 3.395953369140625,
"step": 5660
},
{
"epoch": 0.7431192660550459,
"grad_norm": 0.44278842210769653,
"learning_rate": 4.0563131472804605e-05,
"loss": 3.414118194580078,
"step": 5670
},
{
"epoch": 0.744429882044561,
"grad_norm": 0.32000258564949036,
"learning_rate": 4.0174166817947346e-05,
"loss": 3.3579486846923827,
"step": 5680
},
{
"epoch": 0.745740498034076,
"grad_norm": 0.39397454261779785,
"learning_rate": 3.978671868819792e-05,
"loss": 3.3779945373535156,
"step": 5690
},
{
"epoch": 0.747051114023591,
"grad_norm": 0.3453938364982605,
"learning_rate": 3.940079401039434e-05,
"loss": 3.3643959045410154,
"step": 5700
},
{
"epoch": 0.7483617300131061,
"grad_norm": 0.332226425409317,
"learning_rate": 3.901639968413835e-05,
"loss": 3.412553405761719,
"step": 5710
},
{
"epoch": 0.7496723460026212,
"grad_norm": 0.33490025997161865,
"learning_rate": 3.863354258167184e-05,
"loss": 3.394591522216797,
"step": 5720
},
{
"epoch": 0.7509829619921363,
"grad_norm": 0.33534327149391174,
"learning_rate": 3.825222954775419e-05,
"loss": 3.4038040161132814,
"step": 5730
},
{
"epoch": 0.7522935779816514,
"grad_norm": 0.36024338006973267,
"learning_rate": 3.787246739953962e-05,
"loss": 3.385595703125,
"step": 5740
},
{
"epoch": 0.7536041939711664,
"grad_norm": 0.31551989912986755,
"learning_rate": 3.7494262926455426e-05,
"loss": 3.363161087036133,
"step": 5750
},
{
"epoch": 0.7549148099606815,
"grad_norm": 0.33746281266212463,
"learning_rate": 3.711762289008069e-05,
"loss": 3.3952056884765627,
"step": 5760
},
{
"epoch": 0.7562254259501966,
"grad_norm": 0.3276306390762329,
"learning_rate": 3.6742554024025416e-05,
"loss": 3.3604564666748047,
"step": 5770
},
{
"epoch": 0.7575360419397117,
"grad_norm": 0.37417009472846985,
"learning_rate": 3.6369063033809946e-05,
"loss": 3.394389343261719,
"step": 5780
},
{
"epoch": 0.7588466579292268,
"grad_norm": 0.3108018636703491,
"learning_rate": 3.599715659674524e-05,
"loss": 3.3967559814453123,
"step": 5790
},
{
"epoch": 0.7601572739187418,
"grad_norm": 0.36986300349235535,
"learning_rate": 3.562684136181334e-05,
"loss": 3.3757484436035154,
"step": 5800
},
{
"epoch": 0.7614678899082569,
"grad_norm": 0.32950860261917114,
"learning_rate": 3.525812394954879e-05,
"loss": 3.372197723388672,
"step": 5810
},
{
"epoch": 0.762778505897772,
"grad_norm": 0.34042319655418396,
"learning_rate": 3.489101095192e-05,
"loss": 3.363986587524414,
"step": 5820
},
{
"epoch": 0.7640891218872871,
"grad_norm": 0.35090264678001404,
"learning_rate": 3.4525508932211484e-05,
"loss": 3.346660614013672,
"step": 5830
},
{
"epoch": 0.765399737876802,
"grad_norm": 0.33715710043907166,
"learning_rate": 3.416162442490653e-05,
"loss": 3.3397201538085937,
"step": 5840
},
{
"epoch": 0.7667103538663171,
"grad_norm": 0.33030956983566284,
"learning_rate": 3.3799363935570387e-05,
"loss": 3.366960144042969,
"step": 5850
},
{
"epoch": 0.7680209698558322,
"grad_norm": 0.322314977645874,
"learning_rate": 3.343873394073392e-05,
"loss": 3.393926239013672,
"step": 5860
},
{
"epoch": 0.7693315858453473,
"grad_norm": 0.3647744357585907,
"learning_rate": 3.3079740887777865e-05,
"loss": 3.3585338592529297,
"step": 5870
},
{
"epoch": 0.7706422018348624,
"grad_norm": 0.3519931137561798,
"learning_rate": 3.2722391194817516e-05,
"loss": 3.3352012634277344,
"step": 5880
},
{
"epoch": 0.7719528178243774,
"grad_norm": 0.33610445261001587,
"learning_rate": 3.236669125058803e-05,
"loss": 3.3586959838867188,
"step": 5890
},
{
"epoch": 0.7732634338138925,
"grad_norm": 0.3407491445541382,
"learning_rate": 3.201264741433016e-05,
"loss": 3.355939483642578,
"step": 5900
},
{
"epoch": 0.7745740498034076,
"grad_norm": 0.3252614140510559,
"learning_rate": 3.166026601567661e-05,
"loss": 3.3707916259765627,
"step": 5910
},
{
"epoch": 0.7758846657929227,
"grad_norm": 0.3055562973022461,
"learning_rate": 3.130955335453882e-05,
"loss": 3.3595699310302733,
"step": 5920
},
{
"epoch": 0.7771952817824378,
"grad_norm": 0.34379708766937256,
"learning_rate": 3.096051570099441e-05,
"loss": 3.3652412414550783,
"step": 5930
},
{
"epoch": 0.7785058977719528,
"grad_norm": 0.3022652566432953,
"learning_rate": 3.0613159295175e-05,
"loss": 3.3455970764160154,
"step": 5940
},
{
"epoch": 0.7798165137614679,
"grad_norm": 0.3257646858692169,
"learning_rate": 3.0267490347154702e-05,
"loss": 3.3622432708740235,
"step": 5950
},
{
"epoch": 0.781127129750983,
"grad_norm": 0.3029178977012634,
"learning_rate": 2.99235150368391e-05,
"loss": 3.3372623443603517,
"step": 5960
},
{
"epoch": 0.7824377457404981,
"grad_norm": 0.3360319435596466,
"learning_rate": 2.958123951385469e-05,
"loss": 3.3521957397460938,
"step": 5970
},
{
"epoch": 0.783748361730013,
"grad_norm": 0.3412819504737854,
"learning_rate": 2.924066989743912e-05,
"loss": 3.3656593322753907,
"step": 5980
},
{
"epoch": 0.7850589777195282,
"grad_norm": 0.30636218190193176,
"learning_rate": 2.890181227633147e-05,
"loss": 3.3589092254638673,
"step": 5990
},
{
"epoch": 0.7863695937090432,
"grad_norm": 0.3335491120815277,
"learning_rate": 2.856467270866374e-05,
"loss": 3.3817459106445313,
"step": 6000
},
{
"epoch": 0.7863695937090432,
"eval_loss": 3.3303537368774414,
"eval_runtime": 0.6506,
"eval_samples_per_second": 196.738,
"eval_steps_per_second": 49.185,
"step": 6000
},
{
"epoch": 0.7876802096985583,
"grad_norm": 0.35363516211509705,
"learning_rate": 2.8229257221852295e-05,
"loss": 3.3436180114746095,
"step": 6010
},
{
"epoch": 0.7889908256880734,
"grad_norm": 0.37370988726615906,
"learning_rate": 2.7895571812490332e-05,
"loss": 3.3338638305664063,
"step": 6020
},
{
"epoch": 0.7903014416775884,
"grad_norm": 0.34174343943595886,
"learning_rate": 2.756362244624046e-05,
"loss": 3.373862457275391,
"step": 6030
},
{
"epoch": 0.7916120576671035,
"grad_norm": 0.3042144179344177,
"learning_rate": 2.7233415057728043e-05,
"loss": 3.375458526611328,
"step": 6040
},
{
"epoch": 0.7929226736566186,
"grad_norm": 0.3233564794063568,
"learning_rate": 2.690495555043529e-05,
"loss": 3.35299072265625,
"step": 6050
},
{
"epoch": 0.7942332896461337,
"grad_norm": 0.34451305866241455,
"learning_rate": 2.657824979659565e-05,
"loss": 3.3339828491210937,
"step": 6060
},
{
"epoch": 0.7955439056356488,
"grad_norm": 0.3269437253475189,
"learning_rate": 2.6253303637088692e-05,
"loss": 3.3266883850097657,
"step": 6070
},
{
"epoch": 0.7968545216251638,
"grad_norm": 0.3344271183013916,
"learning_rate": 2.5930122881335838e-05,
"loss": 3.3458206176757814,
"step": 6080
},
{
"epoch": 0.7981651376146789,
"grad_norm": 0.3378041982650757,
"learning_rate": 2.5608713307196288e-05,
"loss": 3.376486587524414,
"step": 6090
},
{
"epoch": 0.799475753604194,
"grad_norm": 0.33639028668403625,
"learning_rate": 2.5289080660864104e-05,
"loss": 3.3825927734375,
"step": 6100
},
{
"epoch": 0.8007863695937091,
"grad_norm": 0.33317652344703674,
"learning_rate": 2.49712306567651e-05,
"loss": 3.3571189880371093,
"step": 6110
},
{
"epoch": 0.8020969855832241,
"grad_norm": 0.41228970885276794,
"learning_rate": 2.465516897745483e-05,
"loss": 3.365692138671875,
"step": 6120
},
{
"epoch": 0.8034076015727392,
"grad_norm": 0.3285669684410095,
"learning_rate": 2.4340901273517054e-05,
"loss": 3.34945068359375,
"step": 6130
},
{
"epoch": 0.8047182175622543,
"grad_norm": 0.33901575207710266,
"learning_rate": 2.4028433163462592e-05,
"loss": 3.3318550109863283,
"step": 6140
},
{
"epoch": 0.8060288335517694,
"grad_norm": 0.3647589385509491,
"learning_rate": 2.3717770233628963e-05,
"loss": 3.3123458862304687,
"step": 6150
},
{
"epoch": 0.8073394495412844,
"grad_norm": 0.32430407404899597,
"learning_rate": 2.3408918038080485e-05,
"loss": 3.32352294921875,
"step": 6160
},
{
"epoch": 0.8086500655307994,
"grad_norm": 0.3205670714378357,
"learning_rate": 2.3101882098508984e-05,
"loss": 3.3723487854003906,
"step": 6170
},
{
"epoch": 0.8099606815203145,
"grad_norm": 0.3097008466720581,
"learning_rate": 2.279666790413505e-05,
"loss": 3.345772552490234,
"step": 6180
},
{
"epoch": 0.8112712975098296,
"grad_norm": 0.33213186264038086,
"learning_rate": 2.2493280911609945e-05,
"loss": 3.361161804199219,
"step": 6190
},
{
"epoch": 0.8125819134993447,
"grad_norm": 0.33103683590888977,
"learning_rate": 2.2191726544918024e-05,
"loss": 3.3440555572509765,
"step": 6200
},
{
"epoch": 0.8138925294888598,
"grad_norm": 0.3270765542984009,
"learning_rate": 2.1892010195279753e-05,
"loss": 3.314558410644531,
"step": 6210
},
{
"epoch": 0.8152031454783748,
"grad_norm": 0.8729628324508667,
"learning_rate": 2.1594137221055313e-05,
"loss": 3.3897079467773437,
"step": 6220
},
{
"epoch": 0.8165137614678899,
"grad_norm": 0.327976793050766,
"learning_rate": 2.1298112947648953e-05,
"loss": 3.3648033142089844,
"step": 6230
},
{
"epoch": 0.817824377457405,
"grad_norm": 0.333496630191803,
"learning_rate": 2.1003942667413457e-05,
"loss": 3.347941589355469,
"step": 6240
},
{
"epoch": 0.8191349934469201,
"grad_norm": 0.3156273066997528,
"learning_rate": 2.0711631639555874e-05,
"loss": 3.3855525970458986,
"step": 6250
},
{
"epoch": 0.8204456094364351,
"grad_norm": 0.32962316274642944,
"learning_rate": 2.042118509004326e-05,
"loss": 3.3365623474121096,
"step": 6260
},
{
"epoch": 0.8217562254259502,
"grad_norm": 0.3112349510192871,
"learning_rate": 2.0132608211509432e-05,
"loss": 3.3350597381591798,
"step": 6270
},
{
"epoch": 0.8230668414154653,
"grad_norm": 0.331908643245697,
"learning_rate": 1.984590616316198e-05,
"loss": 3.335915374755859,
"step": 6280
},
{
"epoch": 0.8243774574049804,
"grad_norm": 0.3129262328147888,
"learning_rate": 1.9561084070690003e-05,
"loss": 3.376775360107422,
"step": 6290
},
{
"epoch": 0.8256880733944955,
"grad_norm": 0.35328081250190735,
"learning_rate": 1.9278147026172643e-05,
"loss": 3.337753677368164,
"step": 6300
},
{
"epoch": 0.8269986893840104,
"grad_norm": 0.3121175169944763,
"learning_rate": 1.8997100087987995e-05,
"loss": 3.3632957458496096,
"step": 6310
},
{
"epoch": 0.8283093053735255,
"grad_norm": 0.3124881684780121,
"learning_rate": 1.8717948280722542e-05,
"loss": 3.35325927734375,
"step": 6320
},
{
"epoch": 0.8296199213630406,
"grad_norm": 0.3167072832584381,
"learning_rate": 1.8440696595081503e-05,
"loss": 3.356714630126953,
"step": 6330
},
{
"epoch": 0.8309305373525557,
"grad_norm": 0.33579739928245544,
"learning_rate": 1.8165349987799382e-05,
"loss": 3.362018585205078,
"step": 6340
},
{
"epoch": 0.8322411533420708,
"grad_norm": 0.32899734377861023,
"learning_rate": 1.789191338155169e-05,
"loss": 3.3694324493408203,
"step": 6350
},
{
"epoch": 0.8335517693315858,
"grad_norm": 0.32633092999458313,
"learning_rate": 1.7620391664866593e-05,
"loss": 3.349412536621094,
"step": 6360
},
{
"epoch": 0.8348623853211009,
"grad_norm": 0.38317421078681946,
"learning_rate": 1.7350789692037725e-05,
"loss": 3.3652263641357423,
"step": 6370
},
{
"epoch": 0.836173001310616,
"grad_norm": 0.3290563225746155,
"learning_rate": 1.7083112283037315e-05,
"loss": 3.351622772216797,
"step": 6380
},
{
"epoch": 0.8374836173001311,
"grad_norm": 0.32409361004829407,
"learning_rate": 1.6817364223430053e-05,
"loss": 3.369770812988281,
"step": 6390
},
{
"epoch": 0.8387942332896461,
"grad_norm": 0.4066806435585022,
"learning_rate": 1.655355026428751e-05,
"loss": 3.348828125,
"step": 6400
},
{
"epoch": 0.8401048492791612,
"grad_norm": 0.43215715885162354,
"learning_rate": 1.6291675122103194e-05,
"loss": 3.3478652954101564,
"step": 6410
},
{
"epoch": 0.8414154652686763,
"grad_norm": 0.32412537932395935,
"learning_rate": 1.6031743478708282e-05,
"loss": 3.3778121948242186,
"step": 6420
},
{
"epoch": 0.8427260812581914,
"grad_norm": 0.32044336199760437,
"learning_rate": 1.5773759981187837e-05,
"loss": 3.358892059326172,
"step": 6430
},
{
"epoch": 0.8440366972477065,
"grad_norm": 0.31787800788879395,
"learning_rate": 1.5517729241797784e-05,
"loss": 3.379691314697266,
"step": 6440
},
{
"epoch": 0.8453473132372215,
"grad_norm": 0.3083689212799072,
"learning_rate": 1.5263655837882418e-05,
"loss": 3.372517776489258,
"step": 6450
},
{
"epoch": 0.8466579292267365,
"grad_norm": 0.3075832426548004,
"learning_rate": 1.5011544311792608e-05,
"loss": 3.3828834533691405,
"step": 6460
},
{
"epoch": 0.8479685452162516,
"grad_norm": 0.31305742263793945,
"learning_rate": 1.4761399170804573e-05,
"loss": 3.3757080078125,
"step": 6470
},
{
"epoch": 0.8492791612057667,
"grad_norm": 0.30489581823349,
"learning_rate": 1.4513224887039253e-05,
"loss": 3.380139923095703,
"step": 6480
},
{
"epoch": 0.8505897771952818,
"grad_norm": 0.30987152457237244,
"learning_rate": 1.4267025897382421e-05,
"loss": 3.340727996826172,
"step": 6490
},
{
"epoch": 0.8519003931847968,
"grad_norm": 0.36789146065711975,
"learning_rate": 1.4022806603405352e-05,
"loss": 3.369355010986328,
"step": 6500
},
{
"epoch": 0.8519003931847968,
"eval_loss": 3.3188538551330566,
"eval_runtime": 0.6546,
"eval_samples_per_second": 195.543,
"eval_steps_per_second": 48.886,
"step": 6500
},
{
"epoch": 0.8532110091743119,
"grad_norm": 0.310918927192688,
"learning_rate": 1.3780571371286063e-05,
"loss": 3.3753097534179686,
"step": 6510
},
{
"epoch": 0.854521625163827,
"grad_norm": 0.3079811632633209,
"learning_rate": 1.3540324531731416e-05,
"loss": 3.376170349121094,
"step": 6520
},
{
"epoch": 0.8558322411533421,
"grad_norm": 0.3083153963088989,
"learning_rate": 1.3302070379899444e-05,
"loss": 3.386785125732422,
"step": 6530
},
{
"epoch": 0.8571428571428571,
"grad_norm": 0.3154764771461487,
"learning_rate": 1.3065813175322766e-05,
"loss": 3.4144176483154296,
"step": 6540
},
{
"epoch": 0.8584534731323722,
"grad_norm": 0.3203883469104767,
"learning_rate": 1.2831557141832374e-05,
"loss": 3.4260269165039063,
"step": 6550
},
{
"epoch": 0.8597640891218873,
"grad_norm": 0.3174280822277069,
"learning_rate": 1.2599306467482142e-05,
"loss": 3.354481887817383,
"step": 6560
},
{
"epoch": 0.8610747051114024,
"grad_norm": 0.29426464438438416,
"learning_rate": 1.236906530447389e-05,
"loss": 3.370252990722656,
"step": 6570
},
{
"epoch": 0.8623853211009175,
"grad_norm": 0.3100093603134155,
"learning_rate": 1.2140837769083133e-05,
"loss": 3.38878173828125,
"step": 6580
},
{
"epoch": 0.8636959370904325,
"grad_norm": 0.32260891795158386,
"learning_rate": 1.1914627941585599e-05,
"loss": 3.3707855224609373,
"step": 6590
},
{
"epoch": 0.8650065530799476,
"grad_norm": 0.3071698248386383,
"learning_rate": 1.169043986618426e-05,
"loss": 3.371693420410156,
"step": 6600
},
{
"epoch": 0.8663171690694627,
"grad_norm": 0.31154584884643555,
"learning_rate": 1.1468277550936921e-05,
"loss": 3.3806121826171873,
"step": 6610
},
{
"epoch": 0.8676277850589778,
"grad_norm": 0.29435452818870544,
"learning_rate": 1.1248144967684712e-05,
"loss": 3.396368408203125,
"step": 6620
},
{
"epoch": 0.8689384010484927,
"grad_norm": 0.30826202034950256,
"learning_rate": 1.1030046051980852e-05,
"loss": 3.382862091064453,
"step": 6630
},
{
"epoch": 0.8702490170380078,
"grad_norm": 0.3180260956287384,
"learning_rate": 1.081398470302064e-05,
"loss": 3.379864501953125,
"step": 6640
},
{
"epoch": 0.8715596330275229,
"grad_norm": 0.3043128550052643,
"learning_rate": 1.0599964783571406e-05,
"loss": 3.368876647949219,
"step": 6650
},
{
"epoch": 0.872870249017038,
"grad_norm": 0.3154434263706207,
"learning_rate": 1.0387990119903628e-05,
"loss": 3.3803356170654295,
"step": 6660
},
{
"epoch": 0.8741808650065531,
"grad_norm": 0.3053065538406372,
"learning_rate": 1.0178064501722497e-05,
"loss": 3.367156219482422,
"step": 6670
},
{
"epoch": 0.8754914809960681,
"grad_norm": 0.30004337430000305,
"learning_rate": 9.970191682100144e-06,
"loss": 3.322949981689453,
"step": 6680
},
{
"epoch": 0.8768020969855832,
"grad_norm": 0.3109953999519348,
"learning_rate": 9.764375377408554e-06,
"loss": 3.4001792907714843,
"step": 6690
},
{
"epoch": 0.8781127129750983,
"grad_norm": 0.30131492018699646,
"learning_rate": 9.560619267253128e-06,
"loss": 3.368103790283203,
"step": 6700
},
{
"epoch": 0.8794233289646134,
"grad_norm": 0.2979066073894501,
"learning_rate": 9.358926994406894e-06,
"loss": 3.3624393463134767,
"step": 6710
},
{
"epoch": 0.8807339449541285,
"grad_norm": 0.30490824580192566,
"learning_rate": 9.159302164745365e-06,
"loss": 3.356029510498047,
"step": 6720
},
{
"epoch": 0.8820445609436435,
"grad_norm": 0.3191581964492798,
"learning_rate": 8.961748347182122e-06,
"loss": 3.3503108978271485,
"step": 6730
},
{
"epoch": 0.8833551769331586,
"grad_norm": 0.30742236971855164,
"learning_rate": 8.766269073604921e-06,
"loss": 3.3578086853027345,
"step": 6740
},
{
"epoch": 0.8846657929226737,
"grad_norm": 0.3141056001186371,
"learning_rate": 8.57286783881267e-06,
"loss": 3.3904998779296873,
"step": 6750
},
{
"epoch": 0.8859764089121888,
"grad_norm": 0.30159813165664673,
"learning_rate": 8.381548100452816e-06,
"loss": 3.3533233642578124,
"step": 6760
},
{
"epoch": 0.8872870249017037,
"grad_norm": 0.3002426028251648,
"learning_rate": 8.192313278959695e-06,
"loss": 3.379141616821289,
"step": 6770
},
{
"epoch": 0.8885976408912188,
"grad_norm": 0.32040804624557495,
"learning_rate": 8.005166757493146e-06,
"loss": 3.3639598846435548,
"step": 6780
},
{
"epoch": 0.8899082568807339,
"grad_norm": 0.29884403944015503,
"learning_rate": 7.820111881878256e-06,
"loss": 3.338395690917969,
"step": 6790
},
{
"epoch": 0.891218872870249,
"grad_norm": 0.3116324543952942,
"learning_rate": 7.637151960545416e-06,
"loss": 3.3901512145996096,
"step": 6800
},
{
"epoch": 0.8925294888597641,
"grad_norm": 0.31241050362586975,
"learning_rate": 7.456290264471241e-06,
"loss": 3.36478271484375,
"step": 6810
},
{
"epoch": 0.8938401048492791,
"grad_norm": 0.3349895477294922,
"learning_rate": 7.277530027120033e-06,
"loss": 3.431895446777344,
"step": 6820
},
{
"epoch": 0.8951507208387942,
"grad_norm": 0.32072752714157104,
"learning_rate": 7.1008744443859515e-06,
"loss": 3.3821666717529295,
"step": 6830
},
{
"epoch": 0.8964613368283093,
"grad_norm": 0.3245388865470886,
"learning_rate": 6.926326674535965e-06,
"loss": 3.4020755767822264,
"step": 6840
},
{
"epoch": 0.8977719528178244,
"grad_norm": 0.28425949811935425,
"learning_rate": 6.753889838153365e-06,
"loss": 3.3578716278076173,
"step": 6850
},
{
"epoch": 0.8990825688073395,
"grad_norm": 0.29869386553764343,
"learning_rate": 6.583567018081924e-06,
"loss": 3.347605895996094,
"step": 6860
},
{
"epoch": 0.9003931847968545,
"grad_norm": 0.3033890426158905,
"learning_rate": 6.415361259370811e-06,
"loss": 3.323773193359375,
"step": 6870
},
{
"epoch": 0.9017038007863696,
"grad_norm": 0.2927691340446472,
"learning_rate": 6.249275569220128e-06,
"loss": 3.3410198211669924,
"step": 6880
},
{
"epoch": 0.9030144167758847,
"grad_norm": 0.3325936198234558,
"learning_rate": 6.085312916927224e-06,
"loss": 3.3386634826660155,
"step": 6890
},
{
"epoch": 0.9043250327653998,
"grad_norm": 0.2980666160583496,
"learning_rate": 5.923476233833519e-06,
"loss": 3.3201866149902344,
"step": 6900
},
{
"epoch": 0.9056356487549148,
"grad_norm": 0.39060717821121216,
"learning_rate": 5.763768413272141e-06,
"loss": 3.3882377624511717,
"step": 6910
},
{
"epoch": 0.9069462647444299,
"grad_norm": 0.31787121295928955,
"learning_rate": 5.606192310516192e-06,
"loss": 3.3214431762695313,
"step": 6920
},
{
"epoch": 0.908256880733945,
"grad_norm": 0.3000994622707367,
"learning_rate": 5.450750742727703e-06,
"loss": 3.307904052734375,
"step": 6930
},
{
"epoch": 0.90956749672346,
"grad_norm": 0.3255162239074707,
"learning_rate": 5.2974464889072455e-06,
"loss": 3.3502056121826174,
"step": 6940
},
{
"epoch": 0.9108781127129751,
"grad_norm": 0.29743945598602295,
"learning_rate": 5.146282289844304e-06,
"loss": 3.378376770019531,
"step": 6950
},
{
"epoch": 0.9121887287024901,
"grad_norm": 0.2946659028530121,
"learning_rate": 4.9972608480682165e-06,
"loss": 3.321432113647461,
"step": 6960
},
{
"epoch": 0.9134993446920052,
"grad_norm": 0.2932695150375366,
"learning_rate": 4.8503848277998976e-06,
"loss": 3.345353698730469,
"step": 6970
},
{
"epoch": 0.9148099606815203,
"grad_norm": 0.30997851490974426,
"learning_rate": 4.705656854904164e-06,
"loss": 3.3472625732421877,
"step": 6980
},
{
"epoch": 0.9161205766710354,
"grad_norm": 0.29063260555267334,
"learning_rate": 4.56307951684283e-06,
"loss": 3.339314270019531,
"step": 6990
},
{
"epoch": 0.9174311926605505,
"grad_norm": 0.2816169261932373,
"learning_rate": 4.42265536262848e-06,
"loss": 3.333665466308594,
"step": 7000
},
{
"epoch": 0.9174311926605505,
"eval_loss": 3.3129336833953857,
"eval_runtime": 0.6522,
"eval_samples_per_second": 196.258,
"eval_steps_per_second": 49.064,
"step": 7000
},
{
"epoch": 0.9187418086500655,
"grad_norm": 0.3051961362361908,
"learning_rate": 4.2843869027787555e-06,
"loss": 3.3713245391845703,
"step": 7010
},
{
"epoch": 0.9200524246395806,
"grad_norm": 0.2888372540473938,
"learning_rate": 4.148276609271695e-06,
"loss": 3.3540958404541015,
"step": 7020
},
{
"epoch": 0.9213630406290957,
"grad_norm": 0.29332298040390015,
"learning_rate": 4.014326915501299e-06,
"loss": 3.3700469970703124,
"step": 7030
},
{
"epoch": 0.9226736566186108,
"grad_norm": 0.3304659426212311,
"learning_rate": 3.882540216234201e-06,
"loss": 3.3370128631591798,
"step": 7040
},
{
"epoch": 0.9239842726081258,
"grad_norm": 0.3269523084163666,
"learning_rate": 3.7529188675667625e-06,
"loss": 3.338623809814453,
"step": 7050
},
{
"epoch": 0.9252948885976409,
"grad_norm": 0.3190535008907318,
"learning_rate": 3.6254651868830174e-06,
"loss": 3.3183475494384767,
"step": 7060
},
{
"epoch": 0.926605504587156,
"grad_norm": 0.2958862781524658,
"learning_rate": 3.500181452813209e-06,
"loss": 3.330657196044922,
"step": 7070
},
{
"epoch": 0.927916120576671,
"grad_norm": 0.3097800612449646,
"learning_rate": 3.3770699051929586e-06,
"loss": 3.321266937255859,
"step": 7080
},
{
"epoch": 0.9292267365661862,
"grad_norm": 0.30177006125450134,
"learning_rate": 3.256132745023413e-06,
"loss": 3.3275909423828125,
"step": 7090
},
{
"epoch": 0.9305373525557011,
"grad_norm": 0.30706194043159485,
"learning_rate": 3.1373721344317705e-06,
"loss": 3.3286991119384766,
"step": 7100
},
{
"epoch": 0.9318479685452162,
"grad_norm": 0.3056046664714813,
"learning_rate": 3.020790196632622e-06,
"loss": 3.3269805908203125,
"step": 7110
},
{
"epoch": 0.9331585845347313,
"grad_norm": 0.292073518037796,
"learning_rate": 2.9063890158900206e-06,
"loss": 3.351909637451172,
"step": 7120
},
{
"epoch": 0.9344692005242464,
"grad_norm": 0.2931078374385834,
"learning_rate": 2.7941706374802338e-06,
"loss": 3.336248016357422,
"step": 7130
},
{
"epoch": 0.9357798165137615,
"grad_norm": 0.29219377040863037,
"learning_rate": 2.6841370676551624e-06,
"loss": 3.35142822265625,
"step": 7140
},
{
"epoch": 0.9370904325032765,
"grad_norm": 0.289546936750412,
"learning_rate": 2.576290273606452e-06,
"loss": 3.3211265563964845,
"step": 7150
},
{
"epoch": 0.9384010484927916,
"grad_norm": 0.36794570088386536,
"learning_rate": 2.4706321834303686e-06,
"loss": 3.3085655212402343,
"step": 7160
},
{
"epoch": 0.9397116644823067,
"grad_norm": 0.30128714442253113,
"learning_rate": 2.3671646860932706e-06,
"loss": 3.332135009765625,
"step": 7170
},
{
"epoch": 0.9410222804718218,
"grad_norm": 0.3192578852176666,
"learning_rate": 2.2658896313978856e-06,
"loss": 3.3328075408935547,
"step": 7180
},
{
"epoch": 0.9423328964613368,
"grad_norm": 0.2986798584461212,
"learning_rate": 2.1668088299502396e-06,
"loss": 3.3342510223388673,
"step": 7190
},
{
"epoch": 0.9436435124508519,
"grad_norm": 0.3024228811264038,
"learning_rate": 2.0699240531272394e-06,
"loss": 3.333052062988281,
"step": 7200
},
{
"epoch": 0.944954128440367,
"grad_norm": 0.35249948501586914,
"learning_rate": 1.9752370330450576e-06,
"loss": 3.3223030090332033,
"step": 7210
},
{
"epoch": 0.9462647444298821,
"grad_norm": 0.29251888394355774,
"learning_rate": 1.882749462528116e-06,
"loss": 3.3603363037109375,
"step": 7220
},
{
"epoch": 0.9475753604193972,
"grad_norm": 0.3287796676158905,
"learning_rate": 1.7924629950788612e-06,
"loss": 3.338475799560547,
"step": 7230
},
{
"epoch": 0.9488859764089121,
"grad_norm": 0.32185614109039307,
"learning_rate": 1.7043792448482032e-06,
"loss": 3.3265007019042967,
"step": 7240
},
{
"epoch": 0.9501965923984272,
"grad_norm": 0.3013335168361664,
"learning_rate": 1.6184997866066364e-06,
"loss": 3.3488208770751955,
"step": 7250
},
{
"epoch": 0.9515072083879423,
"grad_norm": 0.28101852536201477,
"learning_rate": 1.5348261557160819e-06,
"loss": 3.323994445800781,
"step": 7260
},
{
"epoch": 0.9528178243774574,
"grad_norm": 0.3079557716846466,
"learning_rate": 1.4533598481024367e-06,
"loss": 3.323910140991211,
"step": 7270
},
{
"epoch": 0.9541284403669725,
"grad_norm": 0.29948699474334717,
"learning_rate": 1.3741023202288872e-06,
"loss": 3.32139892578125,
"step": 7280
},
{
"epoch": 0.9554390563564875,
"grad_norm": 0.2919541001319885,
"learning_rate": 1.2970549890697764e-06,
"loss": 3.320056915283203,
"step": 7290
},
{
"epoch": 0.9567496723460026,
"grad_norm": 0.2965683341026306,
"learning_rate": 1.2222192320853337e-06,
"loss": 3.320098114013672,
"step": 7300
},
{
"epoch": 0.9580602883355177,
"grad_norm": 0.3311969041824341,
"learning_rate": 1.1495963871970826e-06,
"loss": 3.3380081176757814,
"step": 7310
},
{
"epoch": 0.9593709043250328,
"grad_norm": 0.3010317087173462,
"learning_rate": 1.079187752763791e-06,
"loss": 3.343079376220703,
"step": 7320
},
{
"epoch": 0.9606815203145478,
"grad_norm": 0.2998013198375702,
"learning_rate": 1.010994587558392e-06,
"loss": 3.3433082580566404,
"step": 7330
},
{
"epoch": 0.9619921363040629,
"grad_norm": 0.3191259801387787,
"learning_rate": 9.450181107454186e-07,
"loss": 3.340503692626953,
"step": 7340
},
{
"epoch": 0.963302752293578,
"grad_norm": 0.3275178074836731,
"learning_rate": 8.812595018592162e-07,
"loss": 3.407551956176758,
"step": 7350
},
{
"epoch": 0.9646133682830931,
"grad_norm": 0.28972694277763367,
"learning_rate": 8.197199007828616e-07,
"loss": 3.3389026641845705,
"step": 7360
},
{
"epoch": 0.9659239842726082,
"grad_norm": 0.29658886790275574,
"learning_rate": 7.604004077277765e-07,
"loss": 3.3539688110351564,
"step": 7370
},
{
"epoch": 0.9672346002621232,
"grad_norm": 0.2887858748435974,
"learning_rate": 7.033020832140219e-07,
"loss": 3.356988525390625,
"step": 7380
},
{
"epoch": 0.9685452162516383,
"grad_norm": 0.2883237898349762,
"learning_rate": 6.48425948051437e-07,
"loss": 3.3269508361816404,
"step": 7390
},
{
"epoch": 0.9698558322411533,
"grad_norm": 0.3045375645160675,
"learning_rate": 5.957729833213077e-07,
"loss": 3.361216735839844,
"step": 7400
},
{
"epoch": 0.9711664482306684,
"grad_norm": 0.2914453148841858,
"learning_rate": 5.453441303588241e-07,
"loss": 3.3756134033203127,
"step": 7410
},
{
"epoch": 0.9724770642201835,
"grad_norm": 0.3032694160938263,
"learning_rate": 4.971402907362755e-07,
"loss": 3.3739913940429687,
"step": 7420
},
{
"epoch": 0.9737876802096985,
"grad_norm": 0.3123086988925934,
"learning_rate": 4.5116232624696494e-07,
"loss": 3.3590274810791017,
"step": 7430
},
{
"epoch": 0.9750982961992136,
"grad_norm": 0.295000821352005,
"learning_rate": 4.0741105888973674e-07,
"loss": 3.358423614501953,
"step": 7440
},
{
"epoch": 0.9764089121887287,
"grad_norm": 0.2938995361328125,
"learning_rate": 3.658872708543204e-07,
"loss": 3.356646728515625,
"step": 7450
},
{
"epoch": 0.9777195281782438,
"grad_norm": 0.37629279494285583,
"learning_rate": 3.2659170450734246e-07,
"loss": 3.390125274658203,
"step": 7460
},
{
"epoch": 0.9790301441677588,
"grad_norm": 0.2896755039691925,
"learning_rate": 2.895250623790313e-07,
"loss": 3.3407524108886717,
"step": 7470
},
{
"epoch": 0.9803407601572739,
"grad_norm": 0.2916204035282135,
"learning_rate": 2.546880071506857e-07,
"loss": 3.3886882781982424,
"step": 7480
},
{
"epoch": 0.981651376146789,
"grad_norm": 0.3027748167514801,
"learning_rate": 2.2208116164282311e-07,
"loss": 3.345502471923828,
"step": 7490
},
{
"epoch": 0.9829619921363041,
"grad_norm": 0.294237345457077,
"learning_rate": 1.9170510880404944e-07,
"loss": 3.3678871154785157,
"step": 7500
},
{
"epoch": 0.9829619921363041,
"eval_loss": 3.311072587966919,
"eval_runtime": 0.652,
"eval_samples_per_second": 196.324,
"eval_steps_per_second": 49.081,
"step": 7500
},
{
"epoch": 0.9842726081258192,
"grad_norm": 0.3106505274772644,
"learning_rate": 1.6356039170059567e-07,
"loss": 3.3929988861083986,
"step": 7510
},
{
"epoch": 0.9855832241153342,
"grad_norm": 0.2878120541572571,
"learning_rate": 1.3764751350664463e-07,
"loss": 3.3702327728271486,
"step": 7520
},
{
"epoch": 0.9868938401048493,
"grad_norm": 0.30748748779296875,
"learning_rate": 1.1396693749536612e-07,
"loss": 3.327832794189453,
"step": 7530
},
{
"epoch": 0.9882044560943644,
"grad_norm": 0.293449342250824,
"learning_rate": 9.251908703054867e-08,
"loss": 3.353495407104492,
"step": 7540
},
{
"epoch": 0.9895150720838795,
"grad_norm": 0.2871973216533661,
"learning_rate": 7.330434555911925e-08,
"loss": 3.366381072998047,
"step": 7550
},
{
"epoch": 0.9908256880733946,
"grad_norm": 0.2895137071609497,
"learning_rate": 5.6323056604218326e-08,
"loss": 3.3602920532226563,
"step": 7560
},
{
"epoch": 0.9921363040629095,
"grad_norm": 0.3256204128265381,
"learning_rate": 4.1575523759121434e-08,
"loss": 3.3837661743164062,
"step": 7570
},
{
"epoch": 0.9934469200524246,
"grad_norm": 0.28854256868362427,
"learning_rate": 2.90620106817574e-08,
"loss": 3.3838157653808594,
"step": 7580
},
{
"epoch": 0.9947575360419397,
"grad_norm": 0.31502512097358704,
"learning_rate": 1.8782741090003776e-08,
"loss": 3.3796337127685545,
"step": 7590
},
{
"epoch": 0.9960681520314548,
"grad_norm": 0.2949548661708832,
"learning_rate": 1.0737898757731679e-08,
"loss": 3.3461822509765624,
"step": 7600
},
{
"epoch": 0.9973787680209698,
"grad_norm": 0.2914263606071472,
"learning_rate": 4.92762751147513e-09,
"loss": 3.3411937713623048,
"step": 7610
},
{
"epoch": 0.9986893840104849,
"grad_norm": 0.2979587912559509,
"learning_rate": 1.3520312278636481e-09,
"loss": 3.339801788330078,
"step": 7620
},
{
"epoch": 1.0,
"grad_norm": 0.29464030265808105,
"learning_rate": 1.1173831776511989e-11,
"loss": 3.35032958984375,
"step": 7630
},
{
"epoch": 1.0,
"eval_loss": 3.3110201358795166,
"eval_runtime": 0.6512,
"eval_samples_per_second": 196.565,
"eval_steps_per_second": 49.141,
"step": 7630
}
],
"logging_steps": 10,
"max_steps": 7630,
"num_input_tokens_seen": 0,
"num_train_epochs": 9223372036854775807,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.2318806662918963e+18,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}