opt-peter-1.3B / trainer_state.json
pszemraj's picture
update model with approx 1.6 epochs training
687ed54
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.6457290244272493,
"global_step": 1016,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 2.4193548387096776e-06,
"loss": 3.4647,
"step": 5
},
{
"epoch": 0.02,
"learning_rate": 4.838709677419355e-06,
"loss": 2.4108,
"step": 10
},
{
"epoch": 0.02,
"learning_rate": 7.258064516129032e-06,
"loss": 1.5747,
"step": 15
},
{
"epoch": 0.03,
"learning_rate": 9.67741935483871e-06,
"loss": 1.5005,
"step": 20
},
{
"epoch": 0.04,
"learning_rate": 1.2096774193548387e-05,
"loss": 1.4694,
"step": 25
},
{
"epoch": 0.05,
"learning_rate": 1.4516129032258065e-05,
"loss": 1.4399,
"step": 30
},
{
"epoch": 0.06,
"learning_rate": 1.6935483870967744e-05,
"loss": 1.4215,
"step": 35
},
{
"epoch": 0.06,
"learning_rate": 1.935483870967742e-05,
"loss": 1.4526,
"step": 40
},
{
"epoch": 0.07,
"learning_rate": 2.1774193548387097e-05,
"loss": 1.4458,
"step": 45
},
{
"epoch": 0.08,
"learning_rate": 2.4193548387096773e-05,
"loss": 1.4237,
"step": 50
},
{
"epoch": 0.09,
"learning_rate": 2.6612903225806453e-05,
"loss": 1.4339,
"step": 55
},
{
"epoch": 0.1,
"learning_rate": 2.903225806451613e-05,
"loss": 1.4294,
"step": 60
},
{
"epoch": 0.11,
"learning_rate": 2.999951499529191e-05,
"loss": 1.4968,
"step": 65
},
{
"epoch": 0.11,
"learning_rate": 2.9996551191211948e-05,
"loss": 1.4261,
"step": 70
},
{
"epoch": 0.12,
"learning_rate": 2.9990893561853812e-05,
"loss": 1.4371,
"step": 75
},
{
"epoch": 0.13,
"learning_rate": 2.9982543123495507e-05,
"loss": 1.4412,
"step": 80
},
{
"epoch": 0.14,
"learning_rate": 2.9971501376123366e-05,
"loss": 1.4638,
"step": 85
},
{
"epoch": 0.15,
"learning_rate": 2.9957770303162634e-05,
"loss": 1.4498,
"step": 90
},
{
"epoch": 0.15,
"learning_rate": 2.9941352371121173e-05,
"loss": 1.4393,
"step": 95
},
{
"epoch": 0.16,
"learning_rate": 2.992225052914641e-05,
"loss": 1.4291,
"step": 100
},
{
"epoch": 0.17,
"learning_rate": 2.990046820849558e-05,
"loss": 1.4587,
"step": 105
},
{
"epoch": 0.18,
"learning_rate": 2.9876009321919372e-05,
"loss": 1.4272,
"step": 110
},
{
"epoch": 0.19,
"learning_rate": 2.9848878262959076e-05,
"loss": 1.4622,
"step": 115
},
{
"epoch": 0.19,
"learning_rate": 2.981907990515739e-05,
"loss": 1.4863,
"step": 120
},
{
"epoch": 0.2,
"learning_rate": 2.9786619601182965e-05,
"loss": 1.4274,
"step": 125
},
{
"epoch": 0.21,
"learning_rate": 2.975150318186892e-05,
"loss": 1.4382,
"step": 130
},
{
"epoch": 0.22,
"learning_rate": 2.9713736955165456e-05,
"loss": 1.4413,
"step": 135
},
{
"epoch": 0.23,
"learning_rate": 2.9673327705006745e-05,
"loss": 1.431,
"step": 140
},
{
"epoch": 0.23,
"learning_rate": 2.963028269009235e-05,
"loss": 1.4659,
"step": 145
},
{
"epoch": 0.24,
"learning_rate": 2.9584609642583337e-05,
"loss": 1.4426,
"step": 150
},
{
"epoch": 0.25,
"learning_rate": 2.9536316766713357e-05,
"loss": 1.4733,
"step": 155
},
{
"epoch": 0.26,
"learning_rate": 2.9485412737314923e-05,
"loss": 1.4595,
"step": 160
},
{
"epoch": 0.27,
"learning_rate": 2.9431906698261136e-05,
"loss": 1.4845,
"step": 165
},
{
"epoch": 0.28,
"learning_rate": 2.9375808260823192e-05,
"loss": 1.5219,
"step": 170
},
{
"epoch": 0.28,
"learning_rate": 2.931712750194392e-05,
"loss": 1.532,
"step": 175
},
{
"epoch": 0.29,
"learning_rate": 2.9255874962427638e-05,
"loss": 1.4632,
"step": 180
},
{
"epoch": 0.3,
"learning_rate": 2.9192061645046724e-05,
"loss": 1.5057,
"step": 185
},
{
"epoch": 0.31,
"learning_rate": 2.9125699012565204e-05,
"loss": 1.5023,
"step": 190
},
{
"epoch": 0.32,
"learning_rate": 2.9056798985679688e-05,
"loss": 1.4423,
"step": 195
},
{
"epoch": 0.32,
"learning_rate": 2.8985373940878053e-05,
"loss": 1.4968,
"step": 200
},
{
"epoch": 0.33,
"learning_rate": 2.8911436708216276e-05,
"loss": 1.4767,
"step": 205
},
{
"epoch": 0.34,
"learning_rate": 2.883500056901376e-05,
"loss": 1.4968,
"step": 210
},
{
"epoch": 0.35,
"learning_rate": 2.875607925346762e-05,
"loss": 1.4842,
"step": 215
},
{
"epoch": 0.36,
"learning_rate": 2.867468693818634e-05,
"loss": 1.4921,
"step": 220
},
{
"epoch": 0.36,
"learning_rate": 2.859083824364323e-05,
"loss": 1.4969,
"step": 225
},
{
"epoch": 0.37,
"learning_rate": 2.8504548231550143e-05,
"loss": 1.4423,
"step": 230
},
{
"epoch": 0.38,
"learning_rate": 2.8415832402151956e-05,
"loss": 1.5165,
"step": 235
},
{
"epoch": 0.39,
"learning_rate": 2.832470669144227e-05,
"loss": 1.4715,
"step": 240
},
{
"epoch": 0.4,
"learning_rate": 2.8231187468300836e-05,
"loss": 1.4747,
"step": 245
},
{
"epoch": 0.4,
"learning_rate": 2.8135291531553192e-05,
"loss": 1.466,
"step": 250
},
{
"epoch": 0.41,
"learning_rate": 2.8037036106953134e-05,
"loss": 1.4609,
"step": 255
},
{
"epoch": 0.42,
"learning_rate": 2.793643884408843e-05,
"loss": 1.4649,
"step": 260
},
{
"epoch": 0.43,
"learning_rate": 2.7833517813210437e-05,
"loss": 1.4923,
"step": 265
},
{
"epoch": 0.44,
"learning_rate": 2.7728291501988173e-05,
"loss": 1.4968,
"step": 270
},
{
"epoch": 0.45,
"learning_rate": 2.7620778812187338e-05,
"loss": 1.4545,
"step": 275
},
{
"epoch": 0.45,
"learning_rate": 2.7510999056275038e-05,
"loss": 1.4791,
"step": 280
},
{
"epoch": 0.46,
"learning_rate": 2.739897195395067e-05,
"loss": 1.4849,
"step": 285
},
{
"epoch": 0.47,
"learning_rate": 2.728471762860369e-05,
"loss": 1.4492,
"step": 290
},
{
"epoch": 0.48,
"learning_rate": 2.716825660369885e-05,
"loss": 1.5052,
"step": 295
},
{
"epoch": 0.49,
"learning_rate": 2.704960979908957e-05,
"loss": 1.4701,
"step": 300
},
{
"epoch": 0.49,
"learning_rate": 2.6928798527260127e-05,
"loss": 1.48,
"step": 305
},
{
"epoch": 0.5,
"learning_rate": 2.680584448949729e-05,
"loss": 1.5158,
"step": 310
},
{
"epoch": 0.51,
"learning_rate": 2.6680769771992136e-05,
"loss": 1.5047,
"step": 315
},
{
"epoch": 0.52,
"learning_rate": 2.6553596841872682e-05,
"loss": 1.5023,
"step": 320
},
{
"epoch": 0.53,
"learning_rate": 2.6424348543168177e-05,
"loss": 1.4932,
"step": 325
},
{
"epoch": 0.53,
"learning_rate": 2.6293048092705586e-05,
"loss": 1.5241,
"step": 330
},
{
"epoch": 0.54,
"learning_rate": 2.6159719075939196e-05,
"loss": 1.4969,
"step": 335
},
{
"epoch": 0.55,
"learning_rate": 2.602438544271395e-05,
"loss": 1.4983,
"step": 340
},
{
"epoch": 0.56,
"learning_rate": 2.5887071502963338e-05,
"loss": 1.4801,
"step": 345
},
{
"epoch": 0.57,
"learning_rate": 2.574780192234264e-05,
"loss": 1.4595,
"step": 350
},
{
"epoch": 0.57,
"learning_rate": 2.5606601717798212e-05,
"loss": 1.4901,
"step": 355
},
{
"epoch": 0.58,
"learning_rate": 2.5463496253073726e-05,
"loss": 1.4939,
"step": 360
},
{
"epoch": 0.59,
"learning_rate": 2.531851123415406e-05,
"loss": 1.4791,
"step": 365
},
{
"epoch": 0.6,
"learning_rate": 2.5171672704647785e-05,
"loss": 1.448,
"step": 370
},
{
"epoch": 0.61,
"learning_rate": 2.502300704110891e-05,
"loss": 1.4857,
"step": 375
},
{
"epoch": 0.61,
"learning_rate": 2.4872540948298913e-05,
"loss": 1.4829,
"step": 380
},
{
"epoch": 0.62,
"learning_rate": 2.472030145438974e-05,
"loss": 1.4786,
"step": 385
},
{
"epoch": 0.63,
"learning_rate": 2.4566315906108772e-05,
"loss": 1.4118,
"step": 390
},
{
"epoch": 0.64,
"learning_rate": 2.4410611963826522e-05,
"loss": 1.429,
"step": 395
},
{
"epoch": 0.65,
"learning_rate": 2.4253217596588036e-05,
"loss": 1.4719,
"step": 400
},
{
"epoch": 0.66,
"learning_rate": 2.4094161077088784e-05,
"loss": 1.5054,
"step": 405
},
{
"epoch": 0.66,
"learning_rate": 2.3933470976596088e-05,
"loss": 1.4493,
"step": 410
},
{
"epoch": 0.67,
"learning_rate": 2.3771176159816846e-05,
"loss": 1.4957,
"step": 415
},
{
"epoch": 0.68,
"learning_rate": 2.360730577971259e-05,
"loss": 1.4764,
"step": 420
},
{
"epoch": 0.69,
"learning_rate": 2.3441889272262742e-05,
"loss": 1.5056,
"step": 425
},
{
"epoch": 0.7,
"learning_rate": 2.3274956351177037e-05,
"loss": 1.4732,
"step": 430
},
{
"epoch": 0.7,
"learning_rate": 2.3106537002558074e-05,
"loss": 1.5047,
"step": 435
},
{
"epoch": 0.71,
"learning_rate": 2.293666147951491e-05,
"loss": 1.5098,
"step": 440
},
{
"epoch": 0.72,
"learning_rate": 2.2765360296728697e-05,
"loss": 1.504,
"step": 445
},
{
"epoch": 0.73,
"learning_rate": 2.259266422497137e-05,
"loss": 1.4742,
"step": 450
},
{
"epoch": 0.74,
"learning_rate": 2.2418604285578273e-05,
"loss": 1.482,
"step": 455
},
{
"epoch": 0.74,
"learning_rate": 2.2243211744875818e-05,
"loss": 1.4702,
"step": 460
},
{
"epoch": 0.75,
"learning_rate": 2.2066518108565137e-05,
"loss": 1.4839,
"step": 465
},
{
"epoch": 0.76,
"learning_rate": 2.18885551160627e-05,
"loss": 1.4942,
"step": 470
},
{
"epoch": 0.77,
"learning_rate": 2.1709354734798998e-05,
"loss": 1.4497,
"step": 475
},
{
"epoch": 0.78,
"learning_rate": 2.152894915447624e-05,
"loss": 1.4927,
"step": 480
},
{
"epoch": 0.78,
"learning_rate": 2.134737078128611e-05,
"loss": 1.4794,
"step": 485
},
{
"epoch": 0.79,
"learning_rate": 2.1164652232088674e-05,
"loss": 1.5034,
"step": 490
},
{
"epoch": 0.8,
"learning_rate": 2.0980826328553416e-05,
"loss": 1.5157,
"step": 495
},
{
"epoch": 0.81,
"learning_rate": 2.0795926091263504e-05,
"loss": 1.4681,
"step": 500
},
{
"epoch": 0.82,
"learning_rate": 2.0609984733784287e-05,
"loss": 1.4732,
"step": 505
},
{
"epoch": 0.83,
"learning_rate": 2.042303565669719e-05,
"loss": 1.5047,
"step": 510
},
{
"epoch": 0.83,
"learning_rate": 2.0235112441599948e-05,
"loss": 1.5093,
"step": 515
},
{
"epoch": 0.84,
"learning_rate": 2.0046248845074373e-05,
"loss": 1.4515,
"step": 520
},
{
"epoch": 0.85,
"learning_rate": 1.9856478792622666e-05,
"loss": 1.5051,
"step": 525
},
{
"epoch": 0.86,
"learning_rate": 1.9665836372573397e-05,
"loss": 1.5073,
"step": 530
},
{
"epoch": 0.87,
"learning_rate": 1.947435582995821e-05,
"loss": 1.4952,
"step": 535
},
{
"epoch": 0.87,
"learning_rate": 1.928207156036043e-05,
"loss": 1.4308,
"step": 540
},
{
"epoch": 0.88,
"learning_rate": 1.9089018103736568e-05,
"loss": 1.4588,
"step": 545
},
{
"epoch": 0.89,
"learning_rate": 1.8895230138211942e-05,
"loss": 1.5477,
"step": 550
},
{
"epoch": 0.9,
"learning_rate": 1.870074247385144e-05,
"loss": 1.4979,
"step": 555
},
{
"epoch": 0.91,
"learning_rate": 1.8505590046406615e-05,
"loss": 1.4487,
"step": 560
},
{
"epoch": 0.91,
"learning_rate": 1.8309807911040186e-05,
"loss": 1.4671,
"step": 565
},
{
"epoch": 0.92,
"learning_rate": 1.8113431236029078e-05,
"loss": 1.4486,
"step": 570
},
{
"epoch": 0.93,
"learning_rate": 1.7916495296447162e-05,
"loss": 1.483,
"step": 575
},
{
"epoch": 0.94,
"learning_rate": 1.771903546782883e-05,
"loss": 1.4896,
"step": 580
},
{
"epoch": 0.95,
"learning_rate": 1.7521087219814454e-05,
"loss": 1.5259,
"step": 585
},
{
"epoch": 0.95,
"learning_rate": 1.7322686109779032e-05,
"loss": 1.4845,
"step": 590
},
{
"epoch": 0.96,
"learning_rate": 1.7123867776445e-05,
"loss": 1.4866,
"step": 595
},
{
"epoch": 0.97,
"learning_rate": 1.692466793348047e-05,
"loss": 1.4968,
"step": 600
},
{
"epoch": 0.98,
"learning_rate": 1.6725122363084004e-05,
"loss": 1.4582,
"step": 605
},
{
"epoch": 0.99,
"learning_rate": 1.6525266909557046e-05,
"loss": 1.4605,
"step": 610
},
{
"epoch": 1.0,
"learning_rate": 1.6325137472865262e-05,
"loss": 1.4391,
"step": 615
},
{
"epoch": 1.0,
"eval_loss": 3.6259799003601074,
"eval_runtime": 967.3528,
"eval_samples_per_second": 8.998,
"eval_steps_per_second": 2.249,
"step": 617
},
{
"epoch": 1.0,
"learning_rate": 1.6124770002189804e-05,
"loss": 1.5992,
"step": 620
},
{
"epoch": 1.01,
"learning_rate": 1.5924200489469782e-05,
"loss": 1.1688,
"step": 625
},
{
"epoch": 1.02,
"learning_rate": 1.572346496293706e-05,
"loss": 1.1778,
"step": 630
},
{
"epoch": 1.03,
"learning_rate": 1.5522599480644496e-05,
"loss": 1.1652,
"step": 635
},
{
"epoch": 1.04,
"learning_rate": 1.532164012398886e-05,
"loss": 1.1344,
"step": 640
},
{
"epoch": 1.05,
"learning_rate": 1.5120622991229545e-05,
"loss": 1.1474,
"step": 645
},
{
"epoch": 1.05,
"learning_rate": 1.4919584191004244e-05,
"loss": 1.1457,
"step": 650
},
{
"epoch": 1.06,
"learning_rate": 1.471855983584276e-05,
"loss": 1.1441,
"step": 655
},
{
"epoch": 1.07,
"learning_rate": 1.4517586035680145e-05,
"loss": 1.1546,
"step": 660
},
{
"epoch": 1.08,
"learning_rate": 1.431669889137027e-05,
"loss": 1.1526,
"step": 665
},
{
"epoch": 1.09,
"learning_rate": 1.4115934488201047e-05,
"loss": 1.1778,
"step": 670
},
{
"epoch": 1.09,
"learning_rate": 1.3915328889412434e-05,
"loss": 1.1468,
"step": 675
},
{
"epoch": 1.1,
"learning_rate": 1.3714918129718418e-05,
"loss": 1.1367,
"step": 680
},
{
"epoch": 1.11,
"learning_rate": 1.3514738208834112e-05,
"loss": 1.1972,
"step": 685
},
{
"epoch": 1.12,
"learning_rate": 1.331482508500912e-05,
"loss": 1.1701,
"step": 690
},
{
"epoch": 1.13,
"learning_rate": 1.31152146685684e-05,
"loss": 1.1911,
"step": 695
},
{
"epoch": 1.13,
"learning_rate": 1.2915942815461677e-05,
"loss": 1.1758,
"step": 700
},
{
"epoch": 1.14,
"learning_rate": 1.2717045320822658e-05,
"loss": 1.1486,
"step": 705
},
{
"epoch": 1.15,
"learning_rate": 1.2518557912539185e-05,
"loss": 1.1502,
"step": 710
},
{
"epoch": 1.16,
"learning_rate": 1.232051624483541e-05,
"loss": 1.1459,
"step": 715
},
{
"epoch": 1.17,
"learning_rate": 1.2122955891867278e-05,
"loss": 1.1546,
"step": 720
},
{
"epoch": 1.17,
"learning_rate": 1.1925912341332324e-05,
"loss": 1.165,
"step": 725
},
{
"epoch": 1.18,
"learning_rate": 1.1729420988095042e-05,
"loss": 1.1548,
"step": 730
},
{
"epoch": 1.19,
"learning_rate": 1.1533517127828926e-05,
"loss": 1.1454,
"step": 735
},
{
"epoch": 1.2,
"learning_rate": 1.1338235950676305e-05,
"loss": 1.19,
"step": 740
},
{
"epoch": 1.21,
"learning_rate": 1.1143612534927153e-05,
"loss": 1.1475,
"step": 745
},
{
"epoch": 1.22,
"learning_rate": 1.0949681840717997e-05,
"loss": 1.1754,
"step": 750
},
{
"epoch": 1.22,
"learning_rate": 1.0756478703752036e-05,
"loss": 1.1041,
"step": 755
},
{
"epoch": 1.23,
"learning_rate": 1.0564037829041609e-05,
"loss": 1.1465,
"step": 760
},
{
"epoch": 1.24,
"learning_rate": 1.037239378467416e-05,
"loss": 1.1704,
"step": 765
},
{
"epoch": 1.25,
"learning_rate": 1.0181580995602766e-05,
"loss": 1.1716,
"step": 770
},
{
"epoch": 1.26,
"learning_rate": 9.991633737462405e-06,
"loss": 1.1902,
"step": 775
},
{
"epoch": 1.26,
"learning_rate": 9.802586130413045e-06,
"loss": 1.1342,
"step": 780
},
{
"epoch": 1.27,
"learning_rate": 9.614472133010623e-06,
"loss": 1.1848,
"step": 785
},
{
"epoch": 1.28,
"learning_rate": 9.42732553610712e-06,
"loss": 1.1652,
"step": 790
},
{
"epoch": 1.29,
"learning_rate": 9.241179956780689e-06,
"loss": 1.1832,
"step": 795
},
{
"epoch": 1.3,
"learning_rate": 9.056068832297041e-06,
"loss": 1.1935,
"step": 800
},
{
"epoch": 1.3,
"learning_rate": 8.872025414103135e-06,
"loss": 1.1714,
"step": 805
},
{
"epoch": 1.31,
"learning_rate": 8.689082761854213e-06,
"loss": 1.1525,
"step": 810
},
{
"epoch": 1.32,
"learning_rate": 8.507273737475307e-06,
"loss": 1.1701,
"step": 815
},
{
"epoch": 1.33,
"learning_rate": 8.326630999258286e-06,
"loss": 1.1761,
"step": 820
},
{
"epoch": 1.34,
"learning_rate": 8.14718699599542e-06,
"loss": 1.1323,
"step": 825
},
{
"epoch": 1.34,
"learning_rate": 7.968973961150653e-06,
"loss": 1.1768,
"step": 830
},
{
"epoch": 1.35,
"learning_rate": 7.792023907069486e-06,
"loss": 1.1487,
"step": 835
},
{
"epoch": 1.36,
"learning_rate": 7.616368619228645e-06,
"loss": 1.1445,
"step": 840
},
{
"epoch": 1.37,
"learning_rate": 7.442039650526419e-06,
"loss": 1.1348,
"step": 845
},
{
"epoch": 1.38,
"learning_rate": 7.2690683156148705e-06,
"loss": 1.1478,
"step": 850
},
{
"epoch": 1.39,
"learning_rate": 7.097485685274776e-06,
"loss": 1.1625,
"step": 855
},
{
"epoch": 1.39,
"learning_rate": 6.927322580834376e-06,
"loss": 1.1331,
"step": 860
},
{
"epoch": 1.4,
"learning_rate": 6.758609568632982e-06,
"loss": 1.1608,
"step": 865
},
{
"epoch": 1.41,
"learning_rate": 6.591376954530345e-06,
"loss": 1.1349,
"step": 870
},
{
"epoch": 1.42,
"learning_rate": 6.4256547784628e-06,
"loss": 1.1569,
"step": 875
},
{
"epoch": 1.43,
"learning_rate": 6.261472809047244e-06,
"loss": 1.152,
"step": 880
},
{
"epoch": 1.43,
"learning_rate": 6.098860538233769e-06,
"loss": 1.1498,
"step": 885
},
{
"epoch": 1.44,
"learning_rate": 5.937847176008072e-06,
"loss": 1.1802,
"step": 890
},
{
"epoch": 1.45,
"learning_rate": 5.778461645144438e-06,
"loss": 1.1717,
"step": 895
},
{
"epoch": 1.46,
"learning_rate": 5.6207325760103845e-06,
"loss": 1.152,
"step": 900
},
{
"epoch": 1.47,
"learning_rate": 5.464688301423782e-06,
"loss": 1.1414,
"step": 905
},
{
"epoch": 1.47,
"learning_rate": 5.310356851563427e-06,
"loss": 1.154,
"step": 910
},
{
"epoch": 1.48,
"learning_rate": 5.1577659489340255e-06,
"loss": 1.1353,
"step": 915
},
{
"epoch": 1.49,
"learning_rate": 5.00694300338638e-06,
"loss": 1.1783,
"step": 920
},
{
"epoch": 1.5,
"learning_rate": 4.857915107193783e-06,
"loss": 1.14,
"step": 925
},
{
"epoch": 1.51,
"learning_rate": 4.710709030185422e-06,
"loss": 1.1755,
"step": 930
},
{
"epoch": 1.51,
"learning_rate": 4.565351214937748e-06,
"loss": 1.1121,
"step": 935
},
{
"epoch": 1.52,
"learning_rate": 4.421867772024601e-06,
"loss": 1.171,
"step": 940
},
{
"epoch": 1.53,
"learning_rate": 4.280284475326948e-06,
"loss": 1.1517,
"step": 945
},
{
"epoch": 1.54,
"learning_rate": 4.140626757403176e-06,
"loss": 1.1694,
"step": 950
},
{
"epoch": 1.55,
"learning_rate": 4.002919704920607e-06,
"loss": 1.1465,
"step": 955
},
{
"epoch": 1.56,
"learning_rate": 3.8671880541492236e-06,
"loss": 1.179,
"step": 960
},
{
"epoch": 1.56,
"learning_rate": 3.7334561865182694e-06,
"loss": 1.1498,
"step": 965
},
{
"epoch": 1.57,
"learning_rate": 3.6017481242366503e-06,
"loss": 1.1438,
"step": 970
},
{
"epoch": 1.58,
"learning_rate": 3.472087525977823e-06,
"loss": 1.1647,
"step": 975
},
{
"epoch": 1.59,
"learning_rate": 3.3444976826299754e-06,
"loss": 1.1475,
"step": 980
},
{
"epoch": 1.6,
"learning_rate": 3.219001513112329e-06,
"loss": 1.15,
"step": 985
},
{
"epoch": 1.6,
"learning_rate": 3.0956215602581933e-06,
"loss": 1.1613,
"step": 990
},
{
"epoch": 1.61,
"learning_rate": 2.974379986765622e-06,
"loss": 1.1672,
"step": 995
},
{
"epoch": 1.62,
"learning_rate": 2.855298571216316e-06,
"loss": 1.1702,
"step": 1000
},
{
"epoch": 1.63,
"learning_rate": 2.738398704163561e-06,
"loss": 1.1634,
"step": 1005
},
{
"epoch": 1.64,
"learning_rate": 2.6237013842898533e-06,
"loss": 1.1756,
"step": 1010
},
{
"epoch": 1.64,
"learning_rate": 2.511227214634887e-06,
"loss": 1.1075,
"step": 1015
}
],
"max_steps": 1234,
"num_train_epochs": 2,
"total_flos": 4.827958566725878e+17,
"trial_name": null,
"trial_params": null
}