svjack's picture
Upload 66 files
9c174fc verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.4624871531346352,
"eval_steps": 500,
"global_step": 900,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 4.9999979150098865e-05,
"loss": 4.0418,
"step": 5
},
{
"epoch": 0.01,
"learning_rate": 4.999989444743508e-05,
"loss": 3.8361,
"step": 10
},
{
"epoch": 0.01,
"learning_rate": 4.999974458911041e-05,
"loss": 3.8737,
"step": 15
},
{
"epoch": 0.01,
"learning_rate": 4.9999529575515425e-05,
"loss": 3.806,
"step": 20
},
{
"epoch": 0.01,
"learning_rate": 4.9999249407210515e-05,
"loss": 3.4125,
"step": 25
},
{
"epoch": 0.02,
"learning_rate": 4.9998904084925845e-05,
"loss": 3.6489,
"step": 30
},
{
"epoch": 0.02,
"learning_rate": 4.999849360956141e-05,
"loss": 3.4719,
"step": 35
},
{
"epoch": 0.02,
"learning_rate": 4.999811831976564e-05,
"loss": 3.5265,
"step": 40
},
{
"epoch": 0.02,
"learning_rate": 4.999759057166764e-05,
"loss": 3.6889,
"step": 45
},
{
"epoch": 0.03,
"learning_rate": 4.9996997673913204e-05,
"loss": 3.5104,
"step": 50
},
{
"epoch": 0.03,
"learning_rate": 4.999633962804756e-05,
"loss": 3.5778,
"step": 55
},
{
"epoch": 0.03,
"learning_rate": 4.999561643578575e-05,
"loss": 3.3353,
"step": 60
},
{
"epoch": 0.03,
"learning_rate": 4.999482809901257e-05,
"loss": 3.3511,
"step": 65
},
{
"epoch": 0.04,
"learning_rate": 4.9993974619782625e-05,
"loss": 3.6355,
"step": 70
},
{
"epoch": 0.04,
"learning_rate": 4.999305600032027e-05,
"loss": 3.3915,
"step": 75
},
{
"epoch": 0.04,
"learning_rate": 4.999207224301965e-05,
"loss": 3.4926,
"step": 80
},
{
"epoch": 0.04,
"learning_rate": 4.999102335044467e-05,
"loss": 3.4918,
"step": 85
},
{
"epoch": 0.05,
"learning_rate": 4.9989909325328996e-05,
"loss": 3.3984,
"step": 90
},
{
"epoch": 0.05,
"learning_rate": 4.998873017057605e-05,
"loss": 3.4989,
"step": 95
},
{
"epoch": 0.05,
"learning_rate": 4.998748588925897e-05,
"loss": 3.6368,
"step": 100
},
{
"epoch": 0.05,
"learning_rate": 4.9986176484620665e-05,
"loss": 3.4412,
"step": 105
},
{
"epoch": 0.06,
"learning_rate": 4.998480196007375e-05,
"loss": 3.4708,
"step": 110
},
{
"epoch": 0.06,
"learning_rate": 4.9983362319200554e-05,
"loss": 3.37,
"step": 115
},
{
"epoch": 0.06,
"learning_rate": 4.998185756575313e-05,
"loss": 3.4524,
"step": 120
},
{
"epoch": 0.06,
"learning_rate": 4.9980287703653225e-05,
"loss": 3.3863,
"step": 125
},
{
"epoch": 0.07,
"learning_rate": 4.997865273699227e-05,
"loss": 3.4414,
"step": 130
},
{
"epoch": 0.07,
"learning_rate": 4.997729789123732e-05,
"loss": 3.3335,
"step": 135
},
{
"epoch": 0.07,
"learning_rate": 4.997554574721849e-05,
"loss": 3.231,
"step": 140
},
{
"epoch": 0.07,
"learning_rate": 4.997372851099727e-05,
"loss": 3.5962,
"step": 145
},
{
"epoch": 0.08,
"learning_rate": 4.997184618730983e-05,
"loss": 3.2795,
"step": 150
},
{
"epoch": 0.08,
"learning_rate": 4.996989878106192e-05,
"loss": 3.5384,
"step": 155
},
{
"epoch": 0.08,
"learning_rate": 4.996788629732897e-05,
"loss": 3.3274,
"step": 160
},
{
"epoch": 0.08,
"learning_rate": 4.996580874135597e-05,
"loss": 3.2292,
"step": 165
},
{
"epoch": 0.09,
"learning_rate": 4.996366611855753e-05,
"loss": 3.4829,
"step": 170
},
{
"epoch": 0.09,
"learning_rate": 4.996145843451785e-05,
"loss": 3.1998,
"step": 175
},
{
"epoch": 0.09,
"learning_rate": 4.9959185694990655e-05,
"loss": 3.3757,
"step": 180
},
{
"epoch": 0.1,
"learning_rate": 4.995684790589926e-05,
"loss": 3.1001,
"step": 185
},
{
"epoch": 0.1,
"learning_rate": 4.99544450733365e-05,
"loss": 3.2301,
"step": 190
},
{
"epoch": 0.1,
"learning_rate": 4.995197720356471e-05,
"loss": 3.3187,
"step": 195
},
{
"epoch": 0.1,
"learning_rate": 4.994944430301576e-05,
"loss": 3.527,
"step": 200
},
{
"epoch": 0.11,
"learning_rate": 4.994684637829098e-05,
"loss": 3.3808,
"step": 205
},
{
"epoch": 0.11,
"learning_rate": 4.994418343616119e-05,
"loss": 3.2406,
"step": 210
},
{
"epoch": 0.11,
"learning_rate": 4.994145548356664e-05,
"loss": 3.3463,
"step": 215
},
{
"epoch": 0.11,
"learning_rate": 4.993866252761702e-05,
"loss": 3.2972,
"step": 220
},
{
"epoch": 0.12,
"learning_rate": 4.993580457559145e-05,
"loss": 3.4135,
"step": 225
},
{
"epoch": 0.12,
"learning_rate": 4.9932881634938424e-05,
"loss": 3.4025,
"step": 230
},
{
"epoch": 0.12,
"learning_rate": 4.992989371327581e-05,
"loss": 3.662,
"step": 235
},
{
"epoch": 0.12,
"learning_rate": 4.992684081839086e-05,
"loss": 3.4279,
"step": 240
},
{
"epoch": 0.13,
"learning_rate": 4.9923722958240124e-05,
"loss": 3.3075,
"step": 245
},
{
"epoch": 0.13,
"learning_rate": 4.992054014094951e-05,
"loss": 3.2826,
"step": 250
},
{
"epoch": 0.13,
"learning_rate": 4.991729237481417e-05,
"loss": 3.4992,
"step": 255
},
{
"epoch": 0.13,
"learning_rate": 4.991397966829856e-05,
"loss": 3.2421,
"step": 260
},
{
"epoch": 0.14,
"learning_rate": 4.991060203003639e-05,
"loss": 3.2615,
"step": 265
},
{
"epoch": 0.14,
"learning_rate": 4.9907159468830585e-05,
"loss": 3.3299,
"step": 270
},
{
"epoch": 0.14,
"learning_rate": 4.990365199365327e-05,
"loss": 3.3117,
"step": 275
},
{
"epoch": 0.14,
"learning_rate": 4.9900079613645754e-05,
"loss": 3.4729,
"step": 280
},
{
"epoch": 0.15,
"learning_rate": 4.9896442338118524e-05,
"loss": 3.2521,
"step": 285
},
{
"epoch": 0.15,
"learning_rate": 4.989274017655117e-05,
"loss": 3.261,
"step": 290
},
{
"epoch": 0.15,
"learning_rate": 4.9888973138592414e-05,
"loss": 3.4403,
"step": 295
},
{
"epoch": 0.15,
"learning_rate": 4.988514123406004e-05,
"loss": 3.3408,
"step": 300
},
{
"epoch": 0.16,
"learning_rate": 4.9881244472940914e-05,
"loss": 3.3208,
"step": 305
},
{
"epoch": 0.16,
"learning_rate": 4.9877282865390904e-05,
"loss": 3.6889,
"step": 310
},
{
"epoch": 0.16,
"learning_rate": 4.9873256421734915e-05,
"loss": 3.1537,
"step": 315
},
{
"epoch": 0.16,
"learning_rate": 4.986916515246681e-05,
"loss": 3.1837,
"step": 320
},
{
"epoch": 0.17,
"learning_rate": 4.986500906824941e-05,
"loss": 3.23,
"step": 325
},
{
"epoch": 0.17,
"learning_rate": 4.9860788179914456e-05,
"loss": 3.4343,
"step": 330
},
{
"epoch": 0.17,
"learning_rate": 4.985650249846258e-05,
"loss": 3.3317,
"step": 335
},
{
"epoch": 0.17,
"learning_rate": 4.9852152035063294e-05,
"loss": 3.155,
"step": 340
},
{
"epoch": 0.18,
"learning_rate": 4.984773680105493e-05,
"loss": 3.3527,
"step": 345
},
{
"epoch": 0.18,
"learning_rate": 4.984415798673684e-05,
"loss": 3.3783,
"step": 350
},
{
"epoch": 0.18,
"learning_rate": 4.9839626194743575e-05,
"loss": 3.2567,
"step": 355
},
{
"epoch": 0.18,
"learning_rate": 4.983502966478656e-05,
"loss": 3.1278,
"step": 360
},
{
"epoch": 0.19,
"learning_rate": 4.9830368408845425e-05,
"loss": 3.2344,
"step": 365
},
{
"epoch": 0.19,
"learning_rate": 4.982564243906852e-05,
"loss": 3.3458,
"step": 370
},
{
"epoch": 0.19,
"learning_rate": 4.9820851767772844e-05,
"loss": 3.1467,
"step": 375
},
{
"epoch": 0.2,
"learning_rate": 4.981599640744402e-05,
"loss": 3.217,
"step": 380
},
{
"epoch": 0.2,
"learning_rate": 4.981107637073628e-05,
"loss": 3.4011,
"step": 385
},
{
"epoch": 0.2,
"learning_rate": 4.98060916704724e-05,
"loss": 3.4098,
"step": 390
},
{
"epoch": 0.2,
"learning_rate": 4.9801042319643686e-05,
"loss": 3.3332,
"step": 395
},
{
"epoch": 0.21,
"learning_rate": 4.979592833140996e-05,
"loss": 3.4042,
"step": 400
},
{
"epoch": 0.21,
"learning_rate": 4.979074971909948e-05,
"loss": 3.3127,
"step": 405
},
{
"epoch": 0.21,
"learning_rate": 4.978550649620894e-05,
"loss": 3.1726,
"step": 410
},
{
"epoch": 0.21,
"learning_rate": 4.978019867640342e-05,
"loss": 3.3029,
"step": 415
},
{
"epoch": 0.22,
"learning_rate": 4.977482627351637e-05,
"loss": 3.3519,
"step": 420
},
{
"epoch": 0.22,
"learning_rate": 4.976938930154954e-05,
"loss": 3.2507,
"step": 425
},
{
"epoch": 0.22,
"learning_rate": 4.976388777467297e-05,
"loss": 3.1322,
"step": 430
},
{
"epoch": 0.22,
"learning_rate": 4.9758321707224954e-05,
"loss": 3.1306,
"step": 435
},
{
"epoch": 0.23,
"learning_rate": 4.975269111371197e-05,
"loss": 3.189,
"step": 440
},
{
"epoch": 0.23,
"learning_rate": 4.974699600880869e-05,
"loss": 3.2598,
"step": 445
},
{
"epoch": 0.23,
"learning_rate": 4.974123640735791e-05,
"loss": 3.3483,
"step": 450
},
{
"epoch": 0.23,
"learning_rate": 4.973541232437052e-05,
"loss": 3.4678,
"step": 455
},
{
"epoch": 0.24,
"learning_rate": 4.972952377502545e-05,
"loss": 2.9668,
"step": 460
},
{
"epoch": 0.24,
"learning_rate": 4.9723570774669687e-05,
"loss": 3.372,
"step": 465
},
{
"epoch": 0.24,
"learning_rate": 4.971755333881814e-05,
"loss": 3.2808,
"step": 470
},
{
"epoch": 0.24,
"learning_rate": 4.971147148315367e-05,
"loss": 3.2235,
"step": 475
},
{
"epoch": 0.25,
"learning_rate": 4.9705325223527057e-05,
"loss": 3.0441,
"step": 480
},
{
"epoch": 0.25,
"learning_rate": 4.96991145759569e-05,
"loss": 3.4724,
"step": 485
},
{
"epoch": 0.25,
"learning_rate": 4.969283955662963e-05,
"loss": 3.389,
"step": 490
},
{
"epoch": 0.25,
"learning_rate": 4.968650018189943e-05,
"loss": 3.3817,
"step": 495
},
{
"epoch": 0.26,
"learning_rate": 4.968009646828822e-05,
"loss": 3.2597,
"step": 500
},
{
"epoch": 0.26,
"learning_rate": 4.967362843248559e-05,
"loss": 3.253,
"step": 505
},
{
"epoch": 0.26,
"learning_rate": 4.966709609134878e-05,
"loss": 3.1903,
"step": 510
},
{
"epoch": 0.26,
"learning_rate": 4.9660499461902645e-05,
"loss": 3.1286,
"step": 515
},
{
"epoch": 0.27,
"learning_rate": 4.965383856133953e-05,
"loss": 3.292,
"step": 520
},
{
"epoch": 0.27,
"learning_rate": 4.964711340701935e-05,
"loss": 3.4217,
"step": 525
},
{
"epoch": 0.27,
"learning_rate": 4.964032401646944e-05,
"loss": 3.2587,
"step": 530
},
{
"epoch": 0.27,
"learning_rate": 4.963347040738457e-05,
"loss": 3.3325,
"step": 535
},
{
"epoch": 0.28,
"learning_rate": 4.962655259762687e-05,
"loss": 3.3475,
"step": 540
},
{
"epoch": 0.28,
"learning_rate": 4.961957060522581e-05,
"loss": 3.2882,
"step": 545
},
{
"epoch": 0.28,
"learning_rate": 4.961252444837809e-05,
"loss": 3.3631,
"step": 550
},
{
"epoch": 0.29,
"learning_rate": 4.9605414145447696e-05,
"loss": 3.2589,
"step": 555
},
{
"epoch": 0.29,
"learning_rate": 4.959823971496574e-05,
"loss": 3.4855,
"step": 560
},
{
"epoch": 0.29,
"learning_rate": 4.959100117563052e-05,
"loss": 3.1354,
"step": 565
},
{
"epoch": 0.29,
"learning_rate": 4.958369854630737e-05,
"loss": 3.3037,
"step": 570
},
{
"epoch": 0.3,
"learning_rate": 4.957633184602867e-05,
"loss": 3.1023,
"step": 575
},
{
"epoch": 0.3,
"learning_rate": 4.956890109399381e-05,
"loss": 3.1933,
"step": 580
},
{
"epoch": 0.3,
"learning_rate": 4.956140630956908e-05,
"loss": 3.3994,
"step": 585
},
{
"epoch": 0.3,
"learning_rate": 4.955384751228767e-05,
"loss": 3.1178,
"step": 590
},
{
"epoch": 0.31,
"learning_rate": 4.954622472184961e-05,
"loss": 3.2513,
"step": 595
},
{
"epoch": 0.31,
"learning_rate": 4.9538537958121686e-05,
"loss": 3.5002,
"step": 600
},
{
"epoch": 0.31,
"learning_rate": 4.953078724113744e-05,
"loss": 3.1867,
"step": 605
},
{
"epoch": 0.31,
"learning_rate": 4.9522972591097074e-05,
"loss": 3.3883,
"step": 610
},
{
"epoch": 0.32,
"learning_rate": 4.9515094028367425e-05,
"loss": 3.3708,
"step": 615
},
{
"epoch": 0.32,
"learning_rate": 4.950715157348191e-05,
"loss": 3.2425,
"step": 620
},
{
"epoch": 0.32,
"learning_rate": 4.9499145247140434e-05,
"loss": 3.1046,
"step": 625
},
{
"epoch": 0.32,
"learning_rate": 4.9491075070209385e-05,
"loss": 3.3236,
"step": 630
},
{
"epoch": 0.33,
"learning_rate": 4.9482941063721576e-05,
"loss": 3.4957,
"step": 635
},
{
"epoch": 0.33,
"learning_rate": 4.947474324887614e-05,
"loss": 3.2225,
"step": 640
},
{
"epoch": 0.33,
"learning_rate": 4.9466481647038554e-05,
"loss": 3.5314,
"step": 645
},
{
"epoch": 0.33,
"learning_rate": 4.9458156279740484e-05,
"loss": 3.1028,
"step": 650
},
{
"epoch": 0.34,
"learning_rate": 4.944976716867984e-05,
"loss": 3.2322,
"step": 655
},
{
"epoch": 0.34,
"learning_rate": 4.9441314335720616e-05,
"loss": 3.1332,
"step": 660
},
{
"epoch": 0.34,
"learning_rate": 4.943279780289292e-05,
"loss": 3.2028,
"step": 665
},
{
"epoch": 0.34,
"learning_rate": 4.9424217592392854e-05,
"loss": 3.2408,
"step": 670
},
{
"epoch": 0.35,
"learning_rate": 4.9415573726582465e-05,
"loss": 3.0121,
"step": 675
},
{
"epoch": 0.35,
"learning_rate": 4.940686622798974e-05,
"loss": 3.2182,
"step": 680
},
{
"epoch": 0.35,
"learning_rate": 4.93980951193085e-05,
"loss": 3.4794,
"step": 685
},
{
"epoch": 0.35,
"learning_rate": 4.93892604233983e-05,
"loss": 3.3226,
"step": 690
},
{
"epoch": 0.36,
"learning_rate": 4.938036216328448e-05,
"loss": 3.1917,
"step": 695
},
{
"epoch": 0.36,
"learning_rate": 4.9371400362158016e-05,
"loss": 3.2176,
"step": 700
},
{
"epoch": 0.36,
"learning_rate": 4.936237504337549e-05,
"loss": 2.933,
"step": 705
},
{
"epoch": 0.36,
"learning_rate": 4.935328623045902e-05,
"loss": 3.134,
"step": 710
},
{
"epoch": 0.37,
"learning_rate": 4.934413394709621e-05,
"loss": 2.983,
"step": 715
},
{
"epoch": 0.37,
"learning_rate": 4.933491821714009e-05,
"loss": 3.4337,
"step": 720
},
{
"epoch": 0.37,
"learning_rate": 4.9325639064609045e-05,
"loss": 3.4699,
"step": 725
},
{
"epoch": 0.38,
"learning_rate": 4.931629651368673e-05,
"loss": 3.4081,
"step": 730
},
{
"epoch": 0.38,
"learning_rate": 4.930689058872208e-05,
"loss": 3.2174,
"step": 735
},
{
"epoch": 0.38,
"learning_rate": 4.929742131422913e-05,
"loss": 3.4953,
"step": 740
},
{
"epoch": 0.38,
"learning_rate": 4.92878887148871e-05,
"loss": 3.217,
"step": 745
},
{
"epoch": 0.39,
"learning_rate": 4.927829281554017e-05,
"loss": 3.2689,
"step": 750
},
{
"epoch": 0.39,
"learning_rate": 4.926863364119756e-05,
"loss": 3.2994,
"step": 755
},
{
"epoch": 0.39,
"learning_rate": 4.9258911217033355e-05,
"loss": 3.3868,
"step": 760
},
{
"epoch": 0.39,
"learning_rate": 4.924912556838651e-05,
"loss": 3.3169,
"step": 765
},
{
"epoch": 0.4,
"learning_rate": 4.923927672076075e-05,
"loss": 3.0466,
"step": 770
},
{
"epoch": 0.4,
"learning_rate": 4.92293646998245e-05,
"loss": 3.0857,
"step": 775
},
{
"epoch": 0.4,
"learning_rate": 4.921938953141084e-05,
"loss": 3.5001,
"step": 780
},
{
"epoch": 0.4,
"learning_rate": 4.920935124151744e-05,
"loss": 3.1633,
"step": 785
},
{
"epoch": 0.41,
"learning_rate": 4.919924985630647e-05,
"loss": 3.4816,
"step": 790
},
{
"epoch": 0.41,
"learning_rate": 4.9189085402104516e-05,
"loss": 3.2032,
"step": 795
},
{
"epoch": 0.41,
"learning_rate": 4.917885790540258e-05,
"loss": 3.164,
"step": 800
},
{
"epoch": 0.41,
"learning_rate": 4.916856739285593e-05,
"loss": 3.3047,
"step": 805
},
{
"epoch": 0.42,
"learning_rate": 4.91582138912841e-05,
"loss": 3.2366,
"step": 810
},
{
"epoch": 0.42,
"learning_rate": 4.9147797427670766e-05,
"loss": 3.3105,
"step": 815
},
{
"epoch": 0.42,
"learning_rate": 4.91373180291637e-05,
"loss": 3.4871,
"step": 820
},
{
"epoch": 0.42,
"learning_rate": 4.912677572307473e-05,
"loss": 3.0442,
"step": 825
},
{
"epoch": 0.43,
"learning_rate": 4.911617053687957e-05,
"loss": 3.1934,
"step": 830
},
{
"epoch": 0.43,
"learning_rate": 4.910550249821787e-05,
"loss": 3.27,
"step": 835
},
{
"epoch": 0.43,
"learning_rate": 4.9094771634893075e-05,
"loss": 3.3252,
"step": 840
},
{
"epoch": 0.43,
"learning_rate": 4.908397797487237e-05,
"loss": 2.9852,
"step": 845
},
{
"epoch": 0.44,
"learning_rate": 4.907312154628658e-05,
"loss": 3.1331,
"step": 850
},
{
"epoch": 0.44,
"learning_rate": 4.906220237743014e-05,
"loss": 3.1444,
"step": 855
},
{
"epoch": 0.44,
"learning_rate": 4.905122049676101e-05,
"loss": 3.3702,
"step": 860
},
{
"epoch": 0.44,
"learning_rate": 4.904017593290056e-05,
"loss": 3.0531,
"step": 865
},
{
"epoch": 0.45,
"learning_rate": 4.902906871463355e-05,
"loss": 3.2007,
"step": 870
},
{
"epoch": 0.45,
"learning_rate": 4.901789887090804e-05,
"loss": 3.0837,
"step": 875
},
{
"epoch": 0.45,
"learning_rate": 4.900666643083528e-05,
"loss": 3.2268,
"step": 880
},
{
"epoch": 0.45,
"learning_rate": 4.8995371423689684e-05,
"loss": 2.9249,
"step": 885
},
{
"epoch": 0.46,
"learning_rate": 4.898401387890873e-05,
"loss": 3.2058,
"step": 890
},
{
"epoch": 0.46,
"learning_rate": 4.897259382609285e-05,
"loss": 3.0971,
"step": 895
},
{
"epoch": 0.46,
"learning_rate": 4.8961111295005446e-05,
"loss": 3.2227,
"step": 900
}
],
"logging_steps": 5,
"max_steps": 9730,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 300,
"total_flos": 1.0630979953041408e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}