Qwen2.5-7B-SFT-LoRA / trainer_state.json
mssongit's picture
Upload folder using huggingface_hub
5a15cc5 verified
{
"best_metric": 0.7716161012649536,
"best_model_checkpoint": "./outputs/lora-out/checkpoint-11000",
"epoch": 1.3056379821958457,
"eval_steps": 500,
"global_step": 11000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0011869436201780415,
"grad_norm": 0.5230351686477661,
"learning_rate": 5.938242280285035e-07,
"loss": 1.0767,
"step": 10
},
{
"epoch": 0.002373887240356083,
"grad_norm": 0.37288129329681396,
"learning_rate": 1.187648456057007e-06,
"loss": 1.1895,
"step": 20
},
{
"epoch": 0.0035608308605341245,
"grad_norm": 0.42279818654060364,
"learning_rate": 1.7814726840855108e-06,
"loss": 1.1203,
"step": 30
},
{
"epoch": 0.004747774480712166,
"grad_norm": 0.47712886333465576,
"learning_rate": 2.375296912114014e-06,
"loss": 1.1429,
"step": 40
},
{
"epoch": 0.005934718100890208,
"grad_norm": 0.3551290035247803,
"learning_rate": 2.9691211401425176e-06,
"loss": 1.1629,
"step": 50
},
{
"epoch": 0.007121661721068249,
"grad_norm": 0.4963815212249756,
"learning_rate": 3.5629453681710215e-06,
"loss": 1.0595,
"step": 60
},
{
"epoch": 0.00830860534124629,
"grad_norm": 0.3718396723270416,
"learning_rate": 4.156769596199525e-06,
"loss": 1.0844,
"step": 70
},
{
"epoch": 0.009495548961424332,
"grad_norm": 0.4744749069213867,
"learning_rate": 4.750593824228028e-06,
"loss": 1.0476,
"step": 80
},
{
"epoch": 0.010682492581602374,
"grad_norm": 0.4049204885959625,
"learning_rate": 5.344418052256532e-06,
"loss": 1.0769,
"step": 90
},
{
"epoch": 0.011869436201780416,
"grad_norm": 0.5128685235977173,
"learning_rate": 5.938242280285035e-06,
"loss": 1.0685,
"step": 100
},
{
"epoch": 0.013056379821958458,
"grad_norm": 0.4374200403690338,
"learning_rate": 6.532066508313539e-06,
"loss": 1.1161,
"step": 110
},
{
"epoch": 0.014243323442136498,
"grad_norm": 0.3933285176753998,
"learning_rate": 7.125890736342043e-06,
"loss": 1.0004,
"step": 120
},
{
"epoch": 0.01543026706231454,
"grad_norm": 0.41642361879348755,
"learning_rate": 7.719714964370546e-06,
"loss": 1.0188,
"step": 130
},
{
"epoch": 0.01661721068249258,
"grad_norm": 0.6790082454681396,
"learning_rate": 8.31353919239905e-06,
"loss": 1.0037,
"step": 140
},
{
"epoch": 0.017804154302670624,
"grad_norm": 0.46090933680534363,
"learning_rate": 8.907363420427554e-06,
"loss": 1.1503,
"step": 150
},
{
"epoch": 0.018991097922848664,
"grad_norm": 0.46357613801956177,
"learning_rate": 9.501187648456057e-06,
"loss": 0.937,
"step": 160
},
{
"epoch": 0.020178041543026708,
"grad_norm": 0.5010367035865784,
"learning_rate": 1.0095011876484562e-05,
"loss": 1.1009,
"step": 170
},
{
"epoch": 0.021364985163204748,
"grad_norm": 0.53301602602005,
"learning_rate": 1.0688836104513065e-05,
"loss": 0.983,
"step": 180
},
{
"epoch": 0.022551928783382788,
"grad_norm": 0.5423634052276611,
"learning_rate": 1.1282660332541568e-05,
"loss": 1.0063,
"step": 190
},
{
"epoch": 0.02373887240356083,
"grad_norm": 0.7087329626083374,
"learning_rate": 1.187648456057007e-05,
"loss": 0.9804,
"step": 200
},
{
"epoch": 0.024925816023738872,
"grad_norm": 0.5914109945297241,
"learning_rate": 1.2470308788598575e-05,
"loss": 1.0208,
"step": 210
},
{
"epoch": 0.026112759643916916,
"grad_norm": 0.4900115430355072,
"learning_rate": 1.3064133016627078e-05,
"loss": 0.9603,
"step": 220
},
{
"epoch": 0.027299703264094956,
"grad_norm": 0.5469143986701965,
"learning_rate": 1.3657957244655583e-05,
"loss": 1.0107,
"step": 230
},
{
"epoch": 0.028486646884272996,
"grad_norm": 0.6561472415924072,
"learning_rate": 1.4251781472684086e-05,
"loss": 1.0653,
"step": 240
},
{
"epoch": 0.02967359050445104,
"grad_norm": 0.6878321766853333,
"learning_rate": 1.4845605700712589e-05,
"loss": 0.9832,
"step": 250
},
{
"epoch": 0.03086053412462908,
"grad_norm": 0.4900878965854645,
"learning_rate": 1.5439429928741092e-05,
"loss": 0.941,
"step": 260
},
{
"epoch": 0.032047477744807124,
"grad_norm": 0.7876664996147156,
"learning_rate": 1.60332541567696e-05,
"loss": 0.9555,
"step": 270
},
{
"epoch": 0.03323442136498516,
"grad_norm": 1.0006091594696045,
"learning_rate": 1.66270783847981e-05,
"loss": 0.9885,
"step": 280
},
{
"epoch": 0.034421364985163204,
"grad_norm": 0.8071354627609253,
"learning_rate": 1.7220902612826605e-05,
"loss": 0.8773,
"step": 290
},
{
"epoch": 0.03560830860534125,
"grad_norm": 0.9984568953514099,
"learning_rate": 1.7814726840855108e-05,
"loss": 0.9989,
"step": 300
},
{
"epoch": 0.03679525222551929,
"grad_norm": 0.8211751580238342,
"learning_rate": 1.840855106888361e-05,
"loss": 0.8656,
"step": 310
},
{
"epoch": 0.03798219584569733,
"grad_norm": 1.264052152633667,
"learning_rate": 1.9002375296912114e-05,
"loss": 0.9641,
"step": 320
},
{
"epoch": 0.03916913946587537,
"grad_norm": 0.8444392085075378,
"learning_rate": 1.9596199524940617e-05,
"loss": 0.8856,
"step": 330
},
{
"epoch": 0.040356083086053415,
"grad_norm": 0.7926741242408752,
"learning_rate": 2.0190023752969123e-05,
"loss": 0.8962,
"step": 340
},
{
"epoch": 0.04154302670623145,
"grad_norm": 0.8968790173530579,
"learning_rate": 2.0783847980997626e-05,
"loss": 0.9583,
"step": 350
},
{
"epoch": 0.042729970326409496,
"grad_norm": 0.9435778260231018,
"learning_rate": 2.137767220902613e-05,
"loss": 0.9534,
"step": 360
},
{
"epoch": 0.04391691394658754,
"grad_norm": 0.9596304893493652,
"learning_rate": 2.1971496437054635e-05,
"loss": 0.9235,
"step": 370
},
{
"epoch": 0.045103857566765576,
"grad_norm": 0.9254686236381531,
"learning_rate": 2.2565320665083135e-05,
"loss": 0.9345,
"step": 380
},
{
"epoch": 0.04629080118694362,
"grad_norm": 0.7298787832260132,
"learning_rate": 2.3159144893111638e-05,
"loss": 0.9894,
"step": 390
},
{
"epoch": 0.04747774480712166,
"grad_norm": 0.8593952655792236,
"learning_rate": 2.375296912114014e-05,
"loss": 0.9318,
"step": 400
},
{
"epoch": 0.0486646884272997,
"grad_norm": 0.9398151636123657,
"learning_rate": 2.4346793349168648e-05,
"loss": 0.9175,
"step": 410
},
{
"epoch": 0.049851632047477744,
"grad_norm": 0.9733231663703918,
"learning_rate": 2.494061757719715e-05,
"loss": 1.0808,
"step": 420
},
{
"epoch": 0.05103857566765579,
"grad_norm": 1.7867991924285889,
"learning_rate": 2.5534441805225657e-05,
"loss": 0.9826,
"step": 430
},
{
"epoch": 0.05222551928783383,
"grad_norm": 0.951397716999054,
"learning_rate": 2.6128266033254157e-05,
"loss": 0.9248,
"step": 440
},
{
"epoch": 0.05341246290801187,
"grad_norm": 0.7252931594848633,
"learning_rate": 2.6722090261282663e-05,
"loss": 0.9616,
"step": 450
},
{
"epoch": 0.05459940652818991,
"grad_norm": 1.4507744312286377,
"learning_rate": 2.7315914489311166e-05,
"loss": 0.9346,
"step": 460
},
{
"epoch": 0.055786350148367955,
"grad_norm": 0.9934329390525818,
"learning_rate": 2.790973871733967e-05,
"loss": 0.9847,
"step": 470
},
{
"epoch": 0.05697329376854599,
"grad_norm": 1.0395673513412476,
"learning_rate": 2.8503562945368172e-05,
"loss": 0.9946,
"step": 480
},
{
"epoch": 0.058160237388724036,
"grad_norm": 1.156590461730957,
"learning_rate": 2.909738717339668e-05,
"loss": 1.0282,
"step": 490
},
{
"epoch": 0.05934718100890208,
"grad_norm": 0.8681560754776001,
"learning_rate": 2.9691211401425178e-05,
"loss": 0.9639,
"step": 500
},
{
"epoch": 0.05934718100890208,
"eval_loss": 0.8606800436973572,
"eval_runtime": 772.8877,
"eval_samples_per_second": 2.295,
"eval_steps_per_second": 0.574,
"step": 500
},
{
"epoch": 0.060534124629080116,
"grad_norm": 0.9961158037185669,
"learning_rate": 3.0285035629453685e-05,
"loss": 0.9167,
"step": 510
},
{
"epoch": 0.06172106824925816,
"grad_norm": 1.1789659261703491,
"learning_rate": 3.0878859857482184e-05,
"loss": 0.9911,
"step": 520
},
{
"epoch": 0.0629080118694362,
"grad_norm": 0.9042471647262573,
"learning_rate": 3.147268408551069e-05,
"loss": 0.9563,
"step": 530
},
{
"epoch": 0.06409495548961425,
"grad_norm": 0.8174954652786255,
"learning_rate": 3.20665083135392e-05,
"loss": 0.899,
"step": 540
},
{
"epoch": 0.06528189910979229,
"grad_norm": 1.0244214534759521,
"learning_rate": 3.2660332541567697e-05,
"loss": 1.0177,
"step": 550
},
{
"epoch": 0.06646884272997032,
"grad_norm": 1.0030802488327026,
"learning_rate": 3.32541567695962e-05,
"loss": 0.8967,
"step": 560
},
{
"epoch": 0.06765578635014836,
"grad_norm": 1.0239039659500122,
"learning_rate": 3.384798099762471e-05,
"loss": 1.003,
"step": 570
},
{
"epoch": 0.06884272997032641,
"grad_norm": 0.8476904034614563,
"learning_rate": 3.444180522565321e-05,
"loss": 0.8915,
"step": 580
},
{
"epoch": 0.07002967359050445,
"grad_norm": 0.8580217957496643,
"learning_rate": 3.503562945368171e-05,
"loss": 0.9793,
"step": 590
},
{
"epoch": 0.0712166172106825,
"grad_norm": 1.0635051727294922,
"learning_rate": 3.5629453681710215e-05,
"loss": 0.939,
"step": 600
},
{
"epoch": 0.07240356083086054,
"grad_norm": 1.0473207235336304,
"learning_rate": 3.6223277909738715e-05,
"loss": 0.9026,
"step": 610
},
{
"epoch": 0.07359050445103858,
"grad_norm": 0.9743014574050903,
"learning_rate": 3.681710213776722e-05,
"loss": 0.9039,
"step": 620
},
{
"epoch": 0.07477744807121661,
"grad_norm": 0.9305230379104614,
"learning_rate": 3.741092636579573e-05,
"loss": 0.9828,
"step": 630
},
{
"epoch": 0.07596439169139466,
"grad_norm": 0.8575608134269714,
"learning_rate": 3.800475059382423e-05,
"loss": 0.9613,
"step": 640
},
{
"epoch": 0.0771513353115727,
"grad_norm": 1.1435967683792114,
"learning_rate": 3.8598574821852734e-05,
"loss": 0.9162,
"step": 650
},
{
"epoch": 0.07833827893175074,
"grad_norm": 1.2059640884399414,
"learning_rate": 3.919239904988123e-05,
"loss": 0.9601,
"step": 660
},
{
"epoch": 0.07952522255192879,
"grad_norm": 1.2867904901504517,
"learning_rate": 3.978622327790974e-05,
"loss": 0.8977,
"step": 670
},
{
"epoch": 0.08071216617210683,
"grad_norm": 1.225451111793518,
"learning_rate": 4.0380047505938246e-05,
"loss": 0.9324,
"step": 680
},
{
"epoch": 0.08189910979228486,
"grad_norm": 0.9785982966423035,
"learning_rate": 4.0973871733966746e-05,
"loss": 0.9825,
"step": 690
},
{
"epoch": 0.0830860534124629,
"grad_norm": 1.006413221359253,
"learning_rate": 4.156769596199525e-05,
"loss": 0.992,
"step": 700
},
{
"epoch": 0.08427299703264095,
"grad_norm": 0.9456585049629211,
"learning_rate": 4.216152019002376e-05,
"loss": 0.9368,
"step": 710
},
{
"epoch": 0.08545994065281899,
"grad_norm": 0.8581728339195251,
"learning_rate": 4.275534441805226e-05,
"loss": 0.888,
"step": 720
},
{
"epoch": 0.08664688427299704,
"grad_norm": 0.8342024683952332,
"learning_rate": 4.3349168646080765e-05,
"loss": 0.8947,
"step": 730
},
{
"epoch": 0.08783382789317508,
"grad_norm": 1.0038254261016846,
"learning_rate": 4.394299287410927e-05,
"loss": 0.8982,
"step": 740
},
{
"epoch": 0.08902077151335312,
"grad_norm": 1.085750699043274,
"learning_rate": 4.453681710213777e-05,
"loss": 0.8623,
"step": 750
},
{
"epoch": 0.09020771513353115,
"grad_norm": 1.2366139888763428,
"learning_rate": 4.513064133016627e-05,
"loss": 0.8971,
"step": 760
},
{
"epoch": 0.0913946587537092,
"grad_norm": 1.0508228540420532,
"learning_rate": 4.5724465558194777e-05,
"loss": 0.9076,
"step": 770
},
{
"epoch": 0.09258160237388724,
"grad_norm": 1.1448291540145874,
"learning_rate": 4.6318289786223276e-05,
"loss": 1.0223,
"step": 780
},
{
"epoch": 0.09376854599406528,
"grad_norm": 0.9256898760795593,
"learning_rate": 4.691211401425178e-05,
"loss": 0.9233,
"step": 790
},
{
"epoch": 0.09495548961424333,
"grad_norm": 0.7629861235618591,
"learning_rate": 4.750593824228028e-05,
"loss": 0.8658,
"step": 800
},
{
"epoch": 0.09614243323442137,
"grad_norm": 1.2851282358169556,
"learning_rate": 4.809976247030879e-05,
"loss": 0.9022,
"step": 810
},
{
"epoch": 0.0973293768545994,
"grad_norm": 1.0901813507080078,
"learning_rate": 4.8693586698337295e-05,
"loss": 0.9875,
"step": 820
},
{
"epoch": 0.09851632047477744,
"grad_norm": 0.9551505446434021,
"learning_rate": 4.9287410926365795e-05,
"loss": 0.9256,
"step": 830
},
{
"epoch": 0.09970326409495549,
"grad_norm": 1.760140299797058,
"learning_rate": 4.98812351543943e-05,
"loss": 0.9199,
"step": 840
},
{
"epoch": 0.10089020771513353,
"grad_norm": 1.2546443939208984,
"learning_rate": 4.9999969188311976e-05,
"loss": 0.9354,
"step": 850
},
{
"epoch": 0.10207715133531158,
"grad_norm": 1.1511316299438477,
"learning_rate": 4.9999844015959526e-05,
"loss": 0.9547,
"step": 860
},
{
"epoch": 0.10326409495548962,
"grad_norm": 1.0263421535491943,
"learning_rate": 4.9999622557693894e-05,
"loss": 0.9864,
"step": 870
},
{
"epoch": 0.10445103857566766,
"grad_norm": 1.0085169076919556,
"learning_rate": 4.9999304814368e-05,
"loss": 0.9439,
"step": 880
},
{
"epoch": 0.10563798219584569,
"grad_norm": 0.8821007609367371,
"learning_rate": 4.9998890787205634e-05,
"loss": 0.9689,
"step": 890
},
{
"epoch": 0.10682492581602374,
"grad_norm": 1.2020398378372192,
"learning_rate": 4.999838047780139e-05,
"loss": 0.8825,
"step": 900
},
{
"epoch": 0.10801186943620178,
"grad_norm": 1.1312310695648193,
"learning_rate": 4.999777388812072e-05,
"loss": 0.954,
"step": 910
},
{
"epoch": 0.10919881305637982,
"grad_norm": 0.8461742997169495,
"learning_rate": 4.999707102049986e-05,
"loss": 0.9593,
"step": 920
},
{
"epoch": 0.11038575667655787,
"grad_norm": 0.9337742328643799,
"learning_rate": 4.9996271877645906e-05,
"loss": 0.8842,
"step": 930
},
{
"epoch": 0.11157270029673591,
"grad_norm": 1.5750162601470947,
"learning_rate": 4.99953764626367e-05,
"loss": 0.9315,
"step": 940
},
{
"epoch": 0.11275964391691394,
"grad_norm": 0.8479421734809875,
"learning_rate": 4.9994384778920914e-05,
"loss": 0.9483,
"step": 950
},
{
"epoch": 0.11394658753709198,
"grad_norm": 1.4970643520355225,
"learning_rate": 4.999329683031796e-05,
"loss": 0.8933,
"step": 960
},
{
"epoch": 0.11513353115727003,
"grad_norm": 1.1047669649124146,
"learning_rate": 4.9992112621018046e-05,
"loss": 1.0143,
"step": 970
},
{
"epoch": 0.11632047477744807,
"grad_norm": 1.0329428911209106,
"learning_rate": 4.99908321555821e-05,
"loss": 0.9286,
"step": 980
},
{
"epoch": 0.11750741839762611,
"grad_norm": 0.7338259816169739,
"learning_rate": 4.998945543894179e-05,
"loss": 0.9012,
"step": 990
},
{
"epoch": 0.11869436201780416,
"grad_norm": 0.9167189598083496,
"learning_rate": 4.998798247639947e-05,
"loss": 0.8714,
"step": 1000
},
{
"epoch": 0.11869436201780416,
"eval_loss": 0.8358961939811707,
"eval_runtime": 772.8376,
"eval_samples_per_second": 2.295,
"eval_steps_per_second": 0.575,
"step": 1000
},
{
"epoch": 0.1198813056379822,
"grad_norm": 0.7216008305549622,
"learning_rate": 4.998641327362823e-05,
"loss": 0.797,
"step": 1010
},
{
"epoch": 0.12106824925816023,
"grad_norm": 0.8900012373924255,
"learning_rate": 4.9984747836671754e-05,
"loss": 0.9256,
"step": 1020
},
{
"epoch": 0.12225519287833828,
"grad_norm": 1.069724202156067,
"learning_rate": 4.9982986171944436e-05,
"loss": 0.8876,
"step": 1030
},
{
"epoch": 0.12344213649851632,
"grad_norm": 1.2220836877822876,
"learning_rate": 4.9981128286231235e-05,
"loss": 0.9672,
"step": 1040
},
{
"epoch": 0.12462908011869436,
"grad_norm": 1.055732011795044,
"learning_rate": 4.997917418668774e-05,
"loss": 0.8998,
"step": 1050
},
{
"epoch": 0.1258160237388724,
"grad_norm": 1.0066126585006714,
"learning_rate": 4.997712388084009e-05,
"loss": 0.965,
"step": 1060
},
{
"epoch": 0.12700296735905045,
"grad_norm": 0.9691961407661438,
"learning_rate": 4.9974977376584945e-05,
"loss": 0.8449,
"step": 1070
},
{
"epoch": 0.1281899109792285,
"grad_norm": 0.9969437122344971,
"learning_rate": 4.997273468218949e-05,
"loss": 0.9465,
"step": 1080
},
{
"epoch": 0.12937685459940654,
"grad_norm": 0.9269385933876038,
"learning_rate": 4.997039580629137e-05,
"loss": 0.9694,
"step": 1090
},
{
"epoch": 0.13056379821958458,
"grad_norm": 1.0190030336380005,
"learning_rate": 4.996796075789868e-05,
"loss": 0.8879,
"step": 1100
},
{
"epoch": 0.13175074183976263,
"grad_norm": 0.9268912076950073,
"learning_rate": 4.9965429546389906e-05,
"loss": 0.9089,
"step": 1110
},
{
"epoch": 0.13293768545994064,
"grad_norm": 0.8694944977760315,
"learning_rate": 4.996280218151391e-05,
"loss": 0.8724,
"step": 1120
},
{
"epoch": 0.13412462908011868,
"grad_norm": 0.862114667892456,
"learning_rate": 4.996007867338989e-05,
"loss": 0.93,
"step": 1130
},
{
"epoch": 0.13531157270029673,
"grad_norm": 1.0992276668548584,
"learning_rate": 4.995725903250732e-05,
"loss": 0.9419,
"step": 1140
},
{
"epoch": 0.13649851632047477,
"grad_norm": 1.1718074083328247,
"learning_rate": 4.995434326972595e-05,
"loss": 0.9153,
"step": 1150
},
{
"epoch": 0.13768545994065282,
"grad_norm": 1.34727942943573,
"learning_rate": 4.99513313962757e-05,
"loss": 0.8732,
"step": 1160
},
{
"epoch": 0.13887240356083086,
"grad_norm": 0.717542290687561,
"learning_rate": 4.994822342375671e-05,
"loss": 0.8217,
"step": 1170
},
{
"epoch": 0.1400593471810089,
"grad_norm": 1.2748587131500244,
"learning_rate": 4.99450193641392e-05,
"loss": 0.8905,
"step": 1180
},
{
"epoch": 0.14124629080118695,
"grad_norm": 1.3889360427856445,
"learning_rate": 4.994171922976348e-05,
"loss": 0.8903,
"step": 1190
},
{
"epoch": 0.142433234421365,
"grad_norm": 0.7903980016708374,
"learning_rate": 4.993832303333989e-05,
"loss": 0.9225,
"step": 1200
},
{
"epoch": 0.14362017804154303,
"grad_norm": 1.0810192823410034,
"learning_rate": 4.993483078794875e-05,
"loss": 0.8787,
"step": 1210
},
{
"epoch": 0.14480712166172108,
"grad_norm": 1.3086780309677124,
"learning_rate": 4.9931242507040294e-05,
"loss": 0.8956,
"step": 1220
},
{
"epoch": 0.14599406528189912,
"grad_norm": 1.00796639919281,
"learning_rate": 4.992755820443465e-05,
"loss": 0.8845,
"step": 1230
},
{
"epoch": 0.14718100890207717,
"grad_norm": 1.0078282356262207,
"learning_rate": 4.992377789432177e-05,
"loss": 0.9332,
"step": 1240
},
{
"epoch": 0.14836795252225518,
"grad_norm": 1.0940144062042236,
"learning_rate": 4.9919901591261366e-05,
"loss": 0.9136,
"step": 1250
},
{
"epoch": 0.14955489614243322,
"grad_norm": 0.9257196187973022,
"learning_rate": 4.991592931018286e-05,
"loss": 0.8778,
"step": 1260
},
{
"epoch": 0.15074183976261127,
"grad_norm": 0.6467018723487854,
"learning_rate": 4.991186106638534e-05,
"loss": 0.9091,
"step": 1270
},
{
"epoch": 0.1519287833827893,
"grad_norm": 0.9637243747711182,
"learning_rate": 4.990769687553749e-05,
"loss": 0.9091,
"step": 1280
},
{
"epoch": 0.15311572700296736,
"grad_norm": 1.1351808309555054,
"learning_rate": 4.990343675367752e-05,
"loss": 0.9378,
"step": 1290
},
{
"epoch": 0.1543026706231454,
"grad_norm": 1.2510889768600464,
"learning_rate": 4.989908071721313e-05,
"loss": 0.8714,
"step": 1300
},
{
"epoch": 0.15548961424332344,
"grad_norm": 1.0249766111373901,
"learning_rate": 4.989462878292142e-05,
"loss": 0.9535,
"step": 1310
},
{
"epoch": 0.1566765578635015,
"grad_norm": 1.5715773105621338,
"learning_rate": 4.9890080967948836e-05,
"loss": 0.8567,
"step": 1320
},
{
"epoch": 0.15786350148367953,
"grad_norm": 1.0343701839447021,
"learning_rate": 4.9885437289811106e-05,
"loss": 0.9312,
"step": 1330
},
{
"epoch": 0.15905044510385757,
"grad_norm": 0.952111542224884,
"learning_rate": 4.988069776639318e-05,
"loss": 0.8942,
"step": 1340
},
{
"epoch": 0.16023738872403562,
"grad_norm": 1.1799042224884033,
"learning_rate": 4.987586241594914e-05,
"loss": 0.9557,
"step": 1350
},
{
"epoch": 0.16142433234421366,
"grad_norm": 0.9772776365280151,
"learning_rate": 4.987093125710215e-05,
"loss": 0.9164,
"step": 1360
},
{
"epoch": 0.1626112759643917,
"grad_norm": 0.9001238346099854,
"learning_rate": 4.986590430884437e-05,
"loss": 0.8403,
"step": 1370
},
{
"epoch": 0.16379821958456972,
"grad_norm": 1.0725136995315552,
"learning_rate": 4.98607815905369e-05,
"loss": 0.9615,
"step": 1380
},
{
"epoch": 0.16498516320474776,
"grad_norm": 1.0103825330734253,
"learning_rate": 4.985556312190968e-05,
"loss": 0.904,
"step": 1390
},
{
"epoch": 0.1661721068249258,
"grad_norm": 1.019081473350525,
"learning_rate": 4.985024892306145e-05,
"loss": 0.8856,
"step": 1400
},
{
"epoch": 0.16735905044510385,
"grad_norm": 0.9215166568756104,
"learning_rate": 4.984483901445961e-05,
"loss": 0.9263,
"step": 1410
},
{
"epoch": 0.1685459940652819,
"grad_norm": 0.9046373963356018,
"learning_rate": 4.9839333416940245e-05,
"loss": 0.8989,
"step": 1420
},
{
"epoch": 0.16973293768545994,
"grad_norm": 0.683239758014679,
"learning_rate": 4.9833732151707915e-05,
"loss": 0.8229,
"step": 1430
},
{
"epoch": 0.17091988130563798,
"grad_norm": 0.9845369458198547,
"learning_rate": 4.982803524033569e-05,
"loss": 0.8801,
"step": 1440
},
{
"epoch": 0.17210682492581603,
"grad_norm": 0.92441326379776,
"learning_rate": 4.9822242704764996e-05,
"loss": 0.9166,
"step": 1450
},
{
"epoch": 0.17329376854599407,
"grad_norm": 1.1213726997375488,
"learning_rate": 4.9816354567305555e-05,
"loss": 0.855,
"step": 1460
},
{
"epoch": 0.1744807121661721,
"grad_norm": 0.9685346484184265,
"learning_rate": 4.98103708506353e-05,
"loss": 0.9458,
"step": 1470
},
{
"epoch": 0.17566765578635016,
"grad_norm": 0.827942967414856,
"learning_rate": 4.980429157780027e-05,
"loss": 0.9032,
"step": 1480
},
{
"epoch": 0.1768545994065282,
"grad_norm": 0.951607882976532,
"learning_rate": 4.979811677221455e-05,
"loss": 0.8916,
"step": 1490
},
{
"epoch": 0.17804154302670624,
"grad_norm": 0.8278565406799316,
"learning_rate": 4.979184645766016e-05,
"loss": 0.9589,
"step": 1500
},
{
"epoch": 0.17804154302670624,
"eval_loss": 0.8247362971305847,
"eval_runtime": 773.006,
"eval_samples_per_second": 2.295,
"eval_steps_per_second": 0.574,
"step": 1500
},
{
"epoch": 0.17922848664688426,
"grad_norm": 1.4448524713516235,
"learning_rate": 4.978548065828697e-05,
"loss": 0.9034,
"step": 1510
},
{
"epoch": 0.1804154302670623,
"grad_norm": 0.9414309859275818,
"learning_rate": 4.9779019398612605e-05,
"loss": 0.8396,
"step": 1520
},
{
"epoch": 0.18160237388724035,
"grad_norm": 1.0360982418060303,
"learning_rate": 4.977246270352235e-05,
"loss": 0.8701,
"step": 1530
},
{
"epoch": 0.1827893175074184,
"grad_norm": 0.9132982492446899,
"learning_rate": 4.976581059826906e-05,
"loss": 0.9002,
"step": 1540
},
{
"epoch": 0.18397626112759644,
"grad_norm": 1.1046500205993652,
"learning_rate": 4.9759063108473046e-05,
"loss": 0.9659,
"step": 1550
},
{
"epoch": 0.18516320474777448,
"grad_norm": 0.7856910228729248,
"learning_rate": 4.975222026012202e-05,
"loss": 0.8569,
"step": 1560
},
{
"epoch": 0.18635014836795252,
"grad_norm": 1.0288382768630981,
"learning_rate": 4.974528207957092e-05,
"loss": 0.8933,
"step": 1570
},
{
"epoch": 0.18753709198813057,
"grad_norm": 1.1528115272521973,
"learning_rate": 4.9738248593541894e-05,
"loss": 0.9673,
"step": 1580
},
{
"epoch": 0.1887240356083086,
"grad_norm": 1.0612051486968994,
"learning_rate": 4.973111982912412e-05,
"loss": 0.8482,
"step": 1590
},
{
"epoch": 0.18991097922848665,
"grad_norm": 0.6942639350891113,
"learning_rate": 4.972389581377376e-05,
"loss": 0.87,
"step": 1600
},
{
"epoch": 0.1910979228486647,
"grad_norm": 1.0134094953536987,
"learning_rate": 4.971657657531383e-05,
"loss": 0.9021,
"step": 1610
},
{
"epoch": 0.19228486646884274,
"grad_norm": 1.2450144290924072,
"learning_rate": 4.9709162141934076e-05,
"loss": 0.9054,
"step": 1620
},
{
"epoch": 0.19347181008902078,
"grad_norm": 0.9599930644035339,
"learning_rate": 4.97016525421909e-05,
"loss": 0.9167,
"step": 1630
},
{
"epoch": 0.1946587537091988,
"grad_norm": 1.022996425628662,
"learning_rate": 4.9694047805007236e-05,
"loss": 0.8035,
"step": 1640
},
{
"epoch": 0.19584569732937684,
"grad_norm": 0.9903687834739685,
"learning_rate": 4.968634795967243e-05,
"loss": 0.8117,
"step": 1650
},
{
"epoch": 0.1970326409495549,
"grad_norm": 0.7823516726493835,
"learning_rate": 4.9678553035842125e-05,
"loss": 0.836,
"step": 1660
},
{
"epoch": 0.19821958456973293,
"grad_norm": 0.896584689617157,
"learning_rate": 4.967066306353816e-05,
"loss": 0.9844,
"step": 1670
},
{
"epoch": 0.19940652818991098,
"grad_norm": 1.0771842002868652,
"learning_rate": 4.966267807314845e-05,
"loss": 0.8248,
"step": 1680
},
{
"epoch": 0.20059347181008902,
"grad_norm": 0.7897894978523254,
"learning_rate": 4.965459809542689e-05,
"loss": 0.801,
"step": 1690
},
{
"epoch": 0.20178041543026706,
"grad_norm": 0.8380696773529053,
"learning_rate": 4.964642316149318e-05,
"loss": 0.9112,
"step": 1700
},
{
"epoch": 0.2029673590504451,
"grad_norm": 0.9411647915840149,
"learning_rate": 4.963815330283276e-05,
"loss": 0.9455,
"step": 1710
},
{
"epoch": 0.20415430267062315,
"grad_norm": 0.9571262001991272,
"learning_rate": 4.9629788551296656e-05,
"loss": 0.897,
"step": 1720
},
{
"epoch": 0.2053412462908012,
"grad_norm": 0.8774480819702148,
"learning_rate": 4.962132893910139e-05,
"loss": 0.9364,
"step": 1730
},
{
"epoch": 0.20652818991097924,
"grad_norm": 0.8009739518165588,
"learning_rate": 4.961277449882882e-05,
"loss": 0.8938,
"step": 1740
},
{
"epoch": 0.20771513353115728,
"grad_norm": 0.8146045207977295,
"learning_rate": 4.960412526342604e-05,
"loss": 0.8371,
"step": 1750
},
{
"epoch": 0.20890207715133532,
"grad_norm": 0.9336399435997009,
"learning_rate": 4.9595381266205234e-05,
"loss": 0.8568,
"step": 1760
},
{
"epoch": 0.21008902077151334,
"grad_norm": 1.1002113819122314,
"learning_rate": 4.958654254084355e-05,
"loss": 0.8137,
"step": 1770
},
{
"epoch": 0.21127596439169138,
"grad_norm": 1.308406949043274,
"learning_rate": 4.9577609121383015e-05,
"loss": 0.96,
"step": 1780
},
{
"epoch": 0.21246290801186943,
"grad_norm": 0.9280428290367126,
"learning_rate": 4.956858104223033e-05,
"loss": 0.8977,
"step": 1790
},
{
"epoch": 0.21364985163204747,
"grad_norm": 0.9551159739494324,
"learning_rate": 4.9559458338156787e-05,
"loss": 0.8581,
"step": 1800
},
{
"epoch": 0.21483679525222552,
"grad_norm": 0.8361184597015381,
"learning_rate": 4.955024104429812e-05,
"loss": 0.9148,
"step": 1810
},
{
"epoch": 0.21602373887240356,
"grad_norm": 0.9807586669921875,
"learning_rate": 4.954092919615436e-05,
"loss": 0.8276,
"step": 1820
},
{
"epoch": 0.2172106824925816,
"grad_norm": 0.8617917895317078,
"learning_rate": 4.953152282958975e-05,
"loss": 0.965,
"step": 1830
},
{
"epoch": 0.21839762611275965,
"grad_norm": 0.8824058175086975,
"learning_rate": 4.952202198083252e-05,
"loss": 0.9464,
"step": 1840
},
{
"epoch": 0.2195845697329377,
"grad_norm": 1.0937623977661133,
"learning_rate": 4.9512426686474825e-05,
"loss": 0.9038,
"step": 1850
},
{
"epoch": 0.22077151335311573,
"grad_norm": 0.8538326025009155,
"learning_rate": 4.950273698347254e-05,
"loss": 0.9066,
"step": 1860
},
{
"epoch": 0.22195845697329378,
"grad_norm": 0.882277250289917,
"learning_rate": 4.9492952909145206e-05,
"loss": 0.8483,
"step": 1870
},
{
"epoch": 0.22314540059347182,
"grad_norm": 1.1620628833770752,
"learning_rate": 4.948307450117578e-05,
"loss": 0.9323,
"step": 1880
},
{
"epoch": 0.22433234421364986,
"grad_norm": 0.9091768860816956,
"learning_rate": 4.947310179761057e-05,
"loss": 0.9345,
"step": 1890
},
{
"epoch": 0.22551928783382788,
"grad_norm": 0.9526820182800293,
"learning_rate": 4.9463034836859035e-05,
"loss": 0.8783,
"step": 1900
},
{
"epoch": 0.22670623145400592,
"grad_norm": 1.2667665481567383,
"learning_rate": 4.94528736576937e-05,
"loss": 0.8743,
"step": 1910
},
{
"epoch": 0.22789317507418397,
"grad_norm": 0.9735057950019836,
"learning_rate": 4.944261829924992e-05,
"loss": 0.8674,
"step": 1920
},
{
"epoch": 0.229080118694362,
"grad_norm": 1.0499471426010132,
"learning_rate": 4.943226880102583e-05,
"loss": 0.984,
"step": 1930
},
{
"epoch": 0.23026706231454006,
"grad_norm": 1.031782865524292,
"learning_rate": 4.94218252028821e-05,
"loss": 0.8371,
"step": 1940
},
{
"epoch": 0.2314540059347181,
"grad_norm": 0.6854504346847534,
"learning_rate": 4.9411287545041826e-05,
"loss": 0.9478,
"step": 1950
},
{
"epoch": 0.23264094955489614,
"grad_norm": 0.7416971325874329,
"learning_rate": 4.940065586809041e-05,
"loss": 0.8734,
"step": 1960
},
{
"epoch": 0.2338278931750742,
"grad_norm": 0.9267434477806091,
"learning_rate": 4.938993021297531e-05,
"loss": 0.9185,
"step": 1970
},
{
"epoch": 0.23501483679525223,
"grad_norm": 1.0826581716537476,
"learning_rate": 4.937911062100597e-05,
"loss": 0.8529,
"step": 1980
},
{
"epoch": 0.23620178041543027,
"grad_norm": 1.1237525939941406,
"learning_rate": 4.936819713385363e-05,
"loss": 0.884,
"step": 1990
},
{
"epoch": 0.23738872403560832,
"grad_norm": 0.8531584739685059,
"learning_rate": 4.935718979355115e-05,
"loss": 0.8282,
"step": 2000
},
{
"epoch": 0.23738872403560832,
"eval_loss": 0.8174661993980408,
"eval_runtime": 770.587,
"eval_samples_per_second": 2.302,
"eval_steps_per_second": 0.576,
"step": 2000
},
{
"epoch": 0.23857566765578636,
"grad_norm": 1.2661949396133423,
"learning_rate": 4.9346088642492867e-05,
"loss": 0.903,
"step": 2010
},
{
"epoch": 0.2397626112759644,
"grad_norm": 0.8154357671737671,
"learning_rate": 4.933489372343443e-05,
"loss": 0.8108,
"step": 2020
},
{
"epoch": 0.24094955489614242,
"grad_norm": 1.1874347925186157,
"learning_rate": 4.9323605079492616e-05,
"loss": 0.8955,
"step": 2030
},
{
"epoch": 0.24213649851632046,
"grad_norm": 0.9091614484786987,
"learning_rate": 4.931222275414521e-05,
"loss": 0.8428,
"step": 2040
},
{
"epoch": 0.2433234421364985,
"grad_norm": 0.9067183136940002,
"learning_rate": 4.9300746791230775e-05,
"loss": 0.9133,
"step": 2050
},
{
"epoch": 0.24451038575667655,
"grad_norm": 0.9364748597145081,
"learning_rate": 4.9289177234948535e-05,
"loss": 0.8841,
"step": 2060
},
{
"epoch": 0.2456973293768546,
"grad_norm": 1.1770869493484497,
"learning_rate": 4.927751412985818e-05,
"loss": 0.9274,
"step": 2070
},
{
"epoch": 0.24688427299703264,
"grad_norm": 0.9044451117515564,
"learning_rate": 4.926575752087971e-05,
"loss": 0.9128,
"step": 2080
},
{
"epoch": 0.24807121661721068,
"grad_norm": 0.9170666933059692,
"learning_rate": 4.925390745329321e-05,
"loss": 0.9526,
"step": 2090
},
{
"epoch": 0.24925816023738873,
"grad_norm": 0.7166856527328491,
"learning_rate": 4.924196397273878e-05,
"loss": 0.8833,
"step": 2100
},
{
"epoch": 0.25044510385756674,
"grad_norm": 1.122950792312622,
"learning_rate": 4.9229927125216245e-05,
"loss": 0.9077,
"step": 2110
},
{
"epoch": 0.2516320474777448,
"grad_norm": 0.7830738425254822,
"learning_rate": 4.921779695708506e-05,
"loss": 0.9394,
"step": 2120
},
{
"epoch": 0.25281899109792283,
"grad_norm": 0.8085533380508423,
"learning_rate": 4.920557351506409e-05,
"loss": 0.9048,
"step": 2130
},
{
"epoch": 0.2540059347181009,
"grad_norm": 0.8500973582267761,
"learning_rate": 4.9193256846231425e-05,
"loss": 0.9618,
"step": 2140
},
{
"epoch": 0.2551928783382789,
"grad_norm": 1.318861961364746,
"learning_rate": 4.9180846998024235e-05,
"loss": 0.9177,
"step": 2150
},
{
"epoch": 0.256379821958457,
"grad_norm": 0.7787696123123169,
"learning_rate": 4.916834401823857e-05,
"loss": 0.9182,
"step": 2160
},
{
"epoch": 0.257566765578635,
"grad_norm": 1.0064891576766968,
"learning_rate": 4.9155747955029156e-05,
"loss": 0.8871,
"step": 2170
},
{
"epoch": 0.2587537091988131,
"grad_norm": 0.8516352772712708,
"learning_rate": 4.9143058856909235e-05,
"loss": 0.8684,
"step": 2180
},
{
"epoch": 0.2599406528189911,
"grad_norm": 0.846522867679596,
"learning_rate": 4.9130276772750374e-05,
"loss": 0.8566,
"step": 2190
},
{
"epoch": 0.26112759643916916,
"grad_norm": 0.8422502875328064,
"learning_rate": 4.911740175178226e-05,
"loss": 0.9234,
"step": 2200
},
{
"epoch": 0.2623145400593472,
"grad_norm": 1.1003506183624268,
"learning_rate": 4.910443384359251e-05,
"loss": 0.9092,
"step": 2210
},
{
"epoch": 0.26350148367952525,
"grad_norm": 1.1265966892242432,
"learning_rate": 4.9091373098126546e-05,
"loss": 0.856,
"step": 2220
},
{
"epoch": 0.26468842729970327,
"grad_norm": 0.9657439589500427,
"learning_rate": 4.90782195656873e-05,
"loss": 0.8534,
"step": 2230
},
{
"epoch": 0.2658753709198813,
"grad_norm": 0.6582869291305542,
"learning_rate": 4.9064973296935075e-05,
"loss": 0.9634,
"step": 2240
},
{
"epoch": 0.26706231454005935,
"grad_norm": 1.1937134265899658,
"learning_rate": 4.9051634342887356e-05,
"loss": 0.911,
"step": 2250
},
{
"epoch": 0.26824925816023737,
"grad_norm": 0.9606165289878845,
"learning_rate": 4.90382027549186e-05,
"loss": 0.8507,
"step": 2260
},
{
"epoch": 0.26943620178041544,
"grad_norm": 0.9722040295600891,
"learning_rate": 4.9024678584760056e-05,
"loss": 0.9747,
"step": 2270
},
{
"epoch": 0.27062314540059346,
"grad_norm": 0.8512585163116455,
"learning_rate": 4.90110618844995e-05,
"loss": 0.8801,
"step": 2280
},
{
"epoch": 0.27181008902077153,
"grad_norm": 0.8925254940986633,
"learning_rate": 4.899735270658114e-05,
"loss": 0.8713,
"step": 2290
},
{
"epoch": 0.27299703264094954,
"grad_norm": 1.0183494091033936,
"learning_rate": 4.898355110380531e-05,
"loss": 0.8947,
"step": 2300
},
{
"epoch": 0.2741839762611276,
"grad_norm": 0.7539317011833191,
"learning_rate": 4.896965712932837e-05,
"loss": 0.9179,
"step": 2310
},
{
"epoch": 0.27537091988130563,
"grad_norm": 1.0620583295822144,
"learning_rate": 4.895567083666239e-05,
"loss": 0.9275,
"step": 2320
},
{
"epoch": 0.2765578635014837,
"grad_norm": 1.908141851425171,
"learning_rate": 4.8941592279675037e-05,
"loss": 0.9131,
"step": 2330
},
{
"epoch": 0.2777448071216617,
"grad_norm": 0.7129389643669128,
"learning_rate": 4.892742151258931e-05,
"loss": 0.8657,
"step": 2340
},
{
"epoch": 0.2789317507418398,
"grad_norm": 0.8836341500282288,
"learning_rate": 4.8913158589983374e-05,
"loss": 0.8815,
"step": 2350
},
{
"epoch": 0.2801186943620178,
"grad_norm": 0.9302510619163513,
"learning_rate": 4.8898803566790296e-05,
"loss": 0.86,
"step": 2360
},
{
"epoch": 0.2813056379821958,
"grad_norm": 0.717185378074646,
"learning_rate": 4.8884356498297884e-05,
"loss": 0.8734,
"step": 2370
},
{
"epoch": 0.2824925816023739,
"grad_norm": 1.0630261898040771,
"learning_rate": 4.886981744014846e-05,
"loss": 0.8946,
"step": 2380
},
{
"epoch": 0.2836795252225519,
"grad_norm": 0.7842198014259338,
"learning_rate": 4.885518644833863e-05,
"loss": 0.877,
"step": 2390
},
{
"epoch": 0.28486646884273,
"grad_norm": 0.9807887077331543,
"learning_rate": 4.8840463579219086e-05,
"loss": 0.8142,
"step": 2400
},
{
"epoch": 0.286053412462908,
"grad_norm": 0.9767608642578125,
"learning_rate": 4.882564888949437e-05,
"loss": 0.9082,
"step": 2410
},
{
"epoch": 0.28724035608308607,
"grad_norm": 1.0220454931259155,
"learning_rate": 4.881074243622268e-05,
"loss": 0.9359,
"step": 2420
},
{
"epoch": 0.2884272997032641,
"grad_norm": 0.791378378868103,
"learning_rate": 4.8795744276815624e-05,
"loss": 0.8607,
"step": 2430
},
{
"epoch": 0.28961424332344216,
"grad_norm": 0.9800243973731995,
"learning_rate": 4.8780654469038036e-05,
"loss": 0.9268,
"step": 2440
},
{
"epoch": 0.29080118694362017,
"grad_norm": 0.9791480302810669,
"learning_rate": 4.8765473071007706e-05,
"loss": 0.8905,
"step": 2450
},
{
"epoch": 0.29198813056379824,
"grad_norm": 0.9130463004112244,
"learning_rate": 4.8750200141195194e-05,
"loss": 0.8537,
"step": 2460
},
{
"epoch": 0.29317507418397626,
"grad_norm": 0.7925013899803162,
"learning_rate": 4.8734835738423586e-05,
"loss": 0.9325,
"step": 2470
},
{
"epoch": 0.29436201780415433,
"grad_norm": 1.0440537929534912,
"learning_rate": 4.871937992186827e-05,
"loss": 0.9535,
"step": 2480
},
{
"epoch": 0.29554896142433235,
"grad_norm": 0.9307257533073425,
"learning_rate": 4.870383275105671e-05,
"loss": 0.8196,
"step": 2490
},
{
"epoch": 0.29673590504451036,
"grad_norm": 0.9311190843582153,
"learning_rate": 4.8688194285868235e-05,
"loss": 0.9039,
"step": 2500
},
{
"epoch": 0.29673590504451036,
"eval_loss": 0.8113005757331848,
"eval_runtime": 769.3974,
"eval_samples_per_second": 2.306,
"eval_steps_per_second": 0.577,
"step": 2500
},
{
"epoch": 0.29792284866468843,
"grad_norm": 0.8999786376953125,
"learning_rate": 4.867246458653377e-05,
"loss": 0.8452,
"step": 2510
},
{
"epoch": 0.29910979228486645,
"grad_norm": 0.8556974530220032,
"learning_rate": 4.865664371363563e-05,
"loss": 0.8501,
"step": 2520
},
{
"epoch": 0.3002967359050445,
"grad_norm": 0.8129183650016785,
"learning_rate": 4.864073172810728e-05,
"loss": 0.8773,
"step": 2530
},
{
"epoch": 0.30148367952522254,
"grad_norm": 1.002024531364441,
"learning_rate": 4.862472869123311e-05,
"loss": 0.897,
"step": 2540
},
{
"epoch": 0.3026706231454006,
"grad_norm": 0.8188593983650208,
"learning_rate": 4.860863466464818e-05,
"loss": 0.88,
"step": 2550
},
{
"epoch": 0.3038575667655786,
"grad_norm": 1.4040788412094116,
"learning_rate": 4.8592449710338004e-05,
"loss": 0.9121,
"step": 2560
},
{
"epoch": 0.3050445103857567,
"grad_norm": 1.0157005786895752,
"learning_rate": 4.8576173890638295e-05,
"loss": 0.8512,
"step": 2570
},
{
"epoch": 0.3062314540059347,
"grad_norm": 0.9177596569061279,
"learning_rate": 4.855980726823472e-05,
"loss": 0.9312,
"step": 2580
},
{
"epoch": 0.3074183976261128,
"grad_norm": 1.0521137714385986,
"learning_rate": 4.8543349906162705e-05,
"loss": 0.8539,
"step": 2590
},
{
"epoch": 0.3086053412462908,
"grad_norm": 0.7588869333267212,
"learning_rate": 4.85268018678071e-05,
"loss": 0.8604,
"step": 2600
},
{
"epoch": 0.30979228486646887,
"grad_norm": 0.6621432900428772,
"learning_rate": 4.851016321690205e-05,
"loss": 0.8125,
"step": 2610
},
{
"epoch": 0.3109792284866469,
"grad_norm": 1.1285654306411743,
"learning_rate": 4.849343401753064e-05,
"loss": 0.9558,
"step": 2620
},
{
"epoch": 0.3121661721068249,
"grad_norm": 0.9147692918777466,
"learning_rate": 4.847661433412474e-05,
"loss": 0.8431,
"step": 2630
},
{
"epoch": 0.313353115727003,
"grad_norm": 0.9889045357704163,
"learning_rate": 4.845970423146467e-05,
"loss": 0.9009,
"step": 2640
},
{
"epoch": 0.314540059347181,
"grad_norm": 1.1277235746383667,
"learning_rate": 4.844270377467905e-05,
"loss": 0.8435,
"step": 2650
},
{
"epoch": 0.31572700296735906,
"grad_norm": 1.0129153728485107,
"learning_rate": 4.842561302924447e-05,
"loss": 0.8972,
"step": 2660
},
{
"epoch": 0.3169139465875371,
"grad_norm": 0.8509721159934998,
"learning_rate": 4.8408432060985244e-05,
"loss": 0.8769,
"step": 2670
},
{
"epoch": 0.31810089020771515,
"grad_norm": 0.8103829622268677,
"learning_rate": 4.83911609360732e-05,
"loss": 0.8339,
"step": 2680
},
{
"epoch": 0.31928783382789316,
"grad_norm": 0.6785222887992859,
"learning_rate": 4.83737997210274e-05,
"loss": 0.8878,
"step": 2690
},
{
"epoch": 0.32047477744807124,
"grad_norm": 0.9442030191421509,
"learning_rate": 4.835634848271387e-05,
"loss": 0.8441,
"step": 2700
},
{
"epoch": 0.32166172106824925,
"grad_norm": 0.7956482768058777,
"learning_rate": 4.833880728834538e-05,
"loss": 0.823,
"step": 2710
},
{
"epoch": 0.3228486646884273,
"grad_norm": 0.8883194327354431,
"learning_rate": 4.832117620548114e-05,
"loss": 0.8942,
"step": 2720
},
{
"epoch": 0.32403560830860534,
"grad_norm": 0.713594377040863,
"learning_rate": 4.8303455302026565e-05,
"loss": 0.8289,
"step": 2730
},
{
"epoch": 0.3252225519287834,
"grad_norm": 0.9183544516563416,
"learning_rate": 4.828564464623302e-05,
"loss": 0.8776,
"step": 2740
},
{
"epoch": 0.3264094955489614,
"grad_norm": 1.0352176427841187,
"learning_rate": 4.826774430669755e-05,
"loss": 0.8295,
"step": 2750
},
{
"epoch": 0.32759643916913944,
"grad_norm": 0.8310732245445251,
"learning_rate": 4.824975435236258e-05,
"loss": 0.9217,
"step": 2760
},
{
"epoch": 0.3287833827893175,
"grad_norm": 0.8640980124473572,
"learning_rate": 4.8231674852515754e-05,
"loss": 0.7977,
"step": 2770
},
{
"epoch": 0.32997032640949553,
"grad_norm": 1.2364047765731812,
"learning_rate": 4.821350587678952e-05,
"loss": 0.8313,
"step": 2780
},
{
"epoch": 0.3311572700296736,
"grad_norm": 0.9419341683387756,
"learning_rate": 4.819524749516098e-05,
"loss": 0.8888,
"step": 2790
},
{
"epoch": 0.3323442136498516,
"grad_norm": 1.1738229990005493,
"learning_rate": 4.817689977795157e-05,
"loss": 0.9192,
"step": 2800
},
{
"epoch": 0.3335311572700297,
"grad_norm": 0.9781436920166016,
"learning_rate": 4.81584627958268e-05,
"loss": 0.9115,
"step": 2810
},
{
"epoch": 0.3347181008902077,
"grad_norm": 1.0958514213562012,
"learning_rate": 4.813993661979598e-05,
"loss": 0.9746,
"step": 2820
},
{
"epoch": 0.3359050445103858,
"grad_norm": 0.9003396034240723,
"learning_rate": 4.812132132121195e-05,
"loss": 0.9283,
"step": 2830
},
{
"epoch": 0.3370919881305638,
"grad_norm": 1.0439910888671875,
"learning_rate": 4.8102616971770786e-05,
"loss": 0.8966,
"step": 2840
},
{
"epoch": 0.33827893175074186,
"grad_norm": 0.9176671504974365,
"learning_rate": 4.808382364351157e-05,
"loss": 0.9132,
"step": 2850
},
{
"epoch": 0.3394658753709199,
"grad_norm": 1.3821845054626465,
"learning_rate": 4.806494140881605e-05,
"loss": 0.807,
"step": 2860
},
{
"epoch": 0.34065281899109795,
"grad_norm": 0.6699957847595215,
"learning_rate": 4.804597034040841e-05,
"loss": 0.8337,
"step": 2870
},
{
"epoch": 0.34183976261127597,
"grad_norm": 0.8552253842353821,
"learning_rate": 4.802691051135497e-05,
"loss": 0.8939,
"step": 2880
},
{
"epoch": 0.343026706231454,
"grad_norm": 0.8702000379562378,
"learning_rate": 4.800776199506392e-05,
"loss": 0.8901,
"step": 2890
},
{
"epoch": 0.34421364985163205,
"grad_norm": 0.9271148443222046,
"learning_rate": 4.798852486528501e-05,
"loss": 0.9171,
"step": 2900
},
{
"epoch": 0.34540059347181007,
"grad_norm": 0.7314199805259705,
"learning_rate": 4.796919919610929e-05,
"loss": 0.9328,
"step": 2910
},
{
"epoch": 0.34658753709198814,
"grad_norm": 0.7465083599090576,
"learning_rate": 4.79497850619688e-05,
"loss": 0.8635,
"step": 2920
},
{
"epoch": 0.34777448071216616,
"grad_norm": 0.9858341217041016,
"learning_rate": 4.793028253763633e-05,
"loss": 0.9312,
"step": 2930
},
{
"epoch": 0.3489614243323442,
"grad_norm": 0.8937369585037231,
"learning_rate": 4.791069169822507e-05,
"loss": 0.8388,
"step": 2940
},
{
"epoch": 0.35014836795252224,
"grad_norm": 1.0096062421798706,
"learning_rate": 4.789101261918837e-05,
"loss": 0.8698,
"step": 2950
},
{
"epoch": 0.3513353115727003,
"grad_norm": 0.7233800888061523,
"learning_rate": 4.787124537631942e-05,
"loss": 0.8332,
"step": 2960
},
{
"epoch": 0.35252225519287833,
"grad_norm": 0.7524568438529968,
"learning_rate": 4.785139004575099e-05,
"loss": 0.9154,
"step": 2970
},
{
"epoch": 0.3537091988130564,
"grad_norm": 1.2735203504562378,
"learning_rate": 4.783144670395509e-05,
"loss": 0.8637,
"step": 2980
},
{
"epoch": 0.3548961424332344,
"grad_norm": 1.1231369972229004,
"learning_rate": 4.781141542774272e-05,
"loss": 0.8696,
"step": 2990
},
{
"epoch": 0.3560830860534125,
"grad_norm": 0.8862468004226685,
"learning_rate": 4.7791296294263556e-05,
"loss": 0.9512,
"step": 3000
},
{
"epoch": 0.3560830860534125,
"eval_loss": 0.8069484829902649,
"eval_runtime": 774.9961,
"eval_samples_per_second": 2.289,
"eval_steps_per_second": 0.573,
"step": 3000
},
{
"epoch": 0.3572700296735905,
"grad_norm": 0.8185063600540161,
"learning_rate": 4.7771089381005636e-05,
"loss": 0.9453,
"step": 3010
},
{
"epoch": 0.3584569732937685,
"grad_norm": 0.8352025747299194,
"learning_rate": 4.7750794765795095e-05,
"loss": 0.8775,
"step": 3020
},
{
"epoch": 0.3596439169139466,
"grad_norm": 0.7337610125541687,
"learning_rate": 4.773041252679584e-05,
"loss": 0.7979,
"step": 3030
},
{
"epoch": 0.3608308605341246,
"grad_norm": 0.980402946472168,
"learning_rate": 4.770994274250926e-05,
"loss": 0.923,
"step": 3040
},
{
"epoch": 0.3620178041543027,
"grad_norm": 0.970067024230957,
"learning_rate": 4.768938549177393e-05,
"loss": 0.9579,
"step": 3050
},
{
"epoch": 0.3632047477744807,
"grad_norm": 0.7478693127632141,
"learning_rate": 4.766874085376528e-05,
"loss": 0.9448,
"step": 3060
},
{
"epoch": 0.36439169139465877,
"grad_norm": 0.8010281920433044,
"learning_rate": 4.764800890799532e-05,
"loss": 0.7986,
"step": 3070
},
{
"epoch": 0.3655786350148368,
"grad_norm": 0.8199505805969238,
"learning_rate": 4.762718973431234e-05,
"loss": 0.8876,
"step": 3080
},
{
"epoch": 0.36676557863501486,
"grad_norm": 0.9049654006958008,
"learning_rate": 4.760628341290054e-05,
"loss": 0.8882,
"step": 3090
},
{
"epoch": 0.36795252225519287,
"grad_norm": 0.9176567196846008,
"learning_rate": 4.7585290024279806e-05,
"loss": 0.9358,
"step": 3100
},
{
"epoch": 0.36913946587537094,
"grad_norm": 0.9601384997367859,
"learning_rate": 4.756420964930535e-05,
"loss": 0.9252,
"step": 3110
},
{
"epoch": 0.37032640949554896,
"grad_norm": 1.0265899896621704,
"learning_rate": 4.754304236916739e-05,
"loss": 0.9221,
"step": 3120
},
{
"epoch": 0.37151335311572703,
"grad_norm": 1.52367103099823,
"learning_rate": 4.752178826539089e-05,
"loss": 0.8951,
"step": 3130
},
{
"epoch": 0.37270029673590505,
"grad_norm": 0.9129610657691956,
"learning_rate": 4.750044741983517e-05,
"loss": 0.9562,
"step": 3140
},
{
"epoch": 0.37388724035608306,
"grad_norm": 0.8349341154098511,
"learning_rate": 4.747901991469367e-05,
"loss": 0.8603,
"step": 3150
},
{
"epoch": 0.37507418397626113,
"grad_norm": 1.1370227336883545,
"learning_rate": 4.7457505832493584e-05,
"loss": 0.8294,
"step": 3160
},
{
"epoch": 0.37626112759643915,
"grad_norm": 0.9587147235870361,
"learning_rate": 4.743590525609554e-05,
"loss": 0.8465,
"step": 3170
},
{
"epoch": 0.3774480712166172,
"grad_norm": 0.8471585512161255,
"learning_rate": 4.741421826869331e-05,
"loss": 0.8936,
"step": 3180
},
{
"epoch": 0.37863501483679524,
"grad_norm": 0.9739141464233398,
"learning_rate": 4.7392444953813476e-05,
"loss": 0.8797,
"step": 3190
},
{
"epoch": 0.3798219584569733,
"grad_norm": 0.9010668992996216,
"learning_rate": 4.737058539531509e-05,
"loss": 0.801,
"step": 3200
},
{
"epoch": 0.3810089020771513,
"grad_norm": 1.0589792728424072,
"learning_rate": 4.7348639677389395e-05,
"loss": 0.8597,
"step": 3210
},
{
"epoch": 0.3821958456973294,
"grad_norm": 1.1330857276916504,
"learning_rate": 4.732660788455944e-05,
"loss": 0.9204,
"step": 3220
},
{
"epoch": 0.3833827893175074,
"grad_norm": 1.1525068283081055,
"learning_rate": 4.730449010167982e-05,
"loss": 0.8399,
"step": 3230
},
{
"epoch": 0.3845697329376855,
"grad_norm": 1.045083999633789,
"learning_rate": 4.72822864139363e-05,
"loss": 0.8264,
"step": 3240
},
{
"epoch": 0.3857566765578635,
"grad_norm": 0.9954350590705872,
"learning_rate": 4.72599969068455e-05,
"loss": 0.8966,
"step": 3250
},
{
"epoch": 0.38694362017804157,
"grad_norm": 0.9580268859863281,
"learning_rate": 4.7237621666254596e-05,
"loss": 0.8649,
"step": 3260
},
{
"epoch": 0.3881305637982196,
"grad_norm": 0.945152997970581,
"learning_rate": 4.7215160778340925e-05,
"loss": 0.8898,
"step": 3270
},
{
"epoch": 0.3893175074183976,
"grad_norm": 0.8537838459014893,
"learning_rate": 4.7192614329611715e-05,
"loss": 0.8578,
"step": 3280
},
{
"epoch": 0.3905044510385757,
"grad_norm": 0.8144123554229736,
"learning_rate": 4.716998240690373e-05,
"loss": 0.8975,
"step": 3290
},
{
"epoch": 0.3916913946587537,
"grad_norm": 1.0924110412597656,
"learning_rate": 4.714726509738293e-05,
"loss": 0.9747,
"step": 3300
},
{
"epoch": 0.39287833827893176,
"grad_norm": 0.9143339395523071,
"learning_rate": 4.712446248854413e-05,
"loss": 0.8604,
"step": 3310
},
{
"epoch": 0.3940652818991098,
"grad_norm": 0.9802985191345215,
"learning_rate": 4.710157466821068e-05,
"loss": 0.9112,
"step": 3320
},
{
"epoch": 0.39525222551928785,
"grad_norm": 0.8650179505348206,
"learning_rate": 4.7078601724534124e-05,
"loss": 0.9519,
"step": 3330
},
{
"epoch": 0.39643916913946586,
"grad_norm": 0.9761397242546082,
"learning_rate": 4.7055543745993824e-05,
"loss": 0.8376,
"step": 3340
},
{
"epoch": 0.39762611275964393,
"grad_norm": 1.0427495241165161,
"learning_rate": 4.70324008213967e-05,
"loss": 0.8693,
"step": 3350
},
{
"epoch": 0.39881305637982195,
"grad_norm": 0.8753753304481506,
"learning_rate": 4.700917303987682e-05,
"loss": 0.8574,
"step": 3360
},
{
"epoch": 0.4,
"grad_norm": 0.9678823947906494,
"learning_rate": 4.698586049089504e-05,
"loss": 0.8288,
"step": 3370
},
{
"epoch": 0.40118694362017804,
"grad_norm": 0.7577453851699829,
"learning_rate": 4.696246326423874e-05,
"loss": 0.9234,
"step": 3380
},
{
"epoch": 0.4023738872403561,
"grad_norm": 0.9350581169128418,
"learning_rate": 4.6938981450021416e-05,
"loss": 0.8608,
"step": 3390
},
{
"epoch": 0.4035608308605341,
"grad_norm": 0.9648170471191406,
"learning_rate": 4.691541513868234e-05,
"loss": 1.0063,
"step": 3400
},
{
"epoch": 0.40474777448071214,
"grad_norm": 0.9693950414657593,
"learning_rate": 4.689176442098624e-05,
"loss": 0.9051,
"step": 3410
},
{
"epoch": 0.4059347181008902,
"grad_norm": 1.0306636095046997,
"learning_rate": 4.6868029388022925e-05,
"loss": 0.8788,
"step": 3420
},
{
"epoch": 0.40712166172106823,
"grad_norm": 0.9469864368438721,
"learning_rate": 4.684421013120694e-05,
"loss": 0.8339,
"step": 3430
},
{
"epoch": 0.4083086053412463,
"grad_norm": 1.083781123161316,
"learning_rate": 4.6820306742277184e-05,
"loss": 0.816,
"step": 3440
},
{
"epoch": 0.4094955489614243,
"grad_norm": 0.7388644218444824,
"learning_rate": 4.6796319313296653e-05,
"loss": 0.9355,
"step": 3450
},
{
"epoch": 0.4106824925816024,
"grad_norm": 0.7487624287605286,
"learning_rate": 4.677224793665197e-05,
"loss": 0.8786,
"step": 3460
},
{
"epoch": 0.4118694362017804,
"grad_norm": 1.0239750146865845,
"learning_rate": 4.674809270505308e-05,
"loss": 0.8933,
"step": 3470
},
{
"epoch": 0.4130563798219585,
"grad_norm": 0.9962897896766663,
"learning_rate": 4.672385371153293e-05,
"loss": 0.9012,
"step": 3480
},
{
"epoch": 0.4142433234421365,
"grad_norm": 1.0697799921035767,
"learning_rate": 4.669953104944703e-05,
"loss": 0.8838,
"step": 3490
},
{
"epoch": 0.41543026706231456,
"grad_norm": 0.9961535930633545,
"learning_rate": 4.667512481247316e-05,
"loss": 0.8607,
"step": 3500
},
{
"epoch": 0.41543026706231456,
"eval_loss": 0.8028744459152222,
"eval_runtime": 777.2663,
"eval_samples_per_second": 2.282,
"eval_steps_per_second": 0.571,
"step": 3500
},
{
"epoch": 0.4166172106824926,
"grad_norm": 0.8088350892066956,
"learning_rate": 4.665063509461097e-05,
"loss": 0.7517,
"step": 3510
},
{
"epoch": 0.41780415430267065,
"grad_norm": 0.9155455827713013,
"learning_rate": 4.6626061990181655e-05,
"loss": 0.8841,
"step": 3520
},
{
"epoch": 0.41899109792284867,
"grad_norm": 1.1847515106201172,
"learning_rate": 4.660140559382754e-05,
"loss": 0.8914,
"step": 3530
},
{
"epoch": 0.4201780415430267,
"grad_norm": 1.216593623161316,
"learning_rate": 4.6576666000511795e-05,
"loss": 0.8503,
"step": 3540
},
{
"epoch": 0.42136498516320475,
"grad_norm": 0.8132240772247314,
"learning_rate": 4.6551843305517955e-05,
"loss": 0.8313,
"step": 3550
},
{
"epoch": 0.42255192878338277,
"grad_norm": 0.7744690775871277,
"learning_rate": 4.652693760444967e-05,
"loss": 0.8644,
"step": 3560
},
{
"epoch": 0.42373887240356084,
"grad_norm": 1.0237466096878052,
"learning_rate": 4.650194899323025e-05,
"loss": 0.7594,
"step": 3570
},
{
"epoch": 0.42492581602373886,
"grad_norm": 1.1930066347122192,
"learning_rate": 4.647687756810236e-05,
"loss": 0.8841,
"step": 3580
},
{
"epoch": 0.4261127596439169,
"grad_norm": 1.263022780418396,
"learning_rate": 4.64517234256276e-05,
"loss": 0.8631,
"step": 3590
},
{
"epoch": 0.42729970326409494,
"grad_norm": 0.7010476589202881,
"learning_rate": 4.642648666268616e-05,
"loss": 0.929,
"step": 3600
},
{
"epoch": 0.428486646884273,
"grad_norm": 0.9364275336265564,
"learning_rate": 4.6401167376476426e-05,
"loss": 0.8954,
"step": 3610
},
{
"epoch": 0.42967359050445103,
"grad_norm": 0.752667248249054,
"learning_rate": 4.637576566451465e-05,
"loss": 0.8819,
"step": 3620
},
{
"epoch": 0.4308605341246291,
"grad_norm": 0.9347110986709595,
"learning_rate": 4.63502816246345e-05,
"loss": 0.9257,
"step": 3630
},
{
"epoch": 0.4320474777448071,
"grad_norm": 0.7974644303321838,
"learning_rate": 4.632471535498677e-05,
"loss": 0.9112,
"step": 3640
},
{
"epoch": 0.4332344213649852,
"grad_norm": 0.8897979855537415,
"learning_rate": 4.629906695403894e-05,
"loss": 0.8222,
"step": 3650
},
{
"epoch": 0.4344213649851632,
"grad_norm": 0.7003577351570129,
"learning_rate": 4.6273336520574796e-05,
"loss": 0.8635,
"step": 3660
},
{
"epoch": 0.4356083086053412,
"grad_norm": 0.741892397403717,
"learning_rate": 4.62475241536941e-05,
"loss": 0.8561,
"step": 3670
},
{
"epoch": 0.4367952522255193,
"grad_norm": 0.816657304763794,
"learning_rate": 4.622162995281216e-05,
"loss": 0.9304,
"step": 3680
},
{
"epoch": 0.4379821958456973,
"grad_norm": 0.8286069631576538,
"learning_rate": 4.6195654017659475e-05,
"loss": 0.9195,
"step": 3690
},
{
"epoch": 0.4391691394658754,
"grad_norm": 1.2624543905258179,
"learning_rate": 4.616959644828133e-05,
"loss": 0.8274,
"step": 3700
},
{
"epoch": 0.4403560830860534,
"grad_norm": 0.7968791127204895,
"learning_rate": 4.614345734503743e-05,
"loss": 0.8574,
"step": 3710
},
{
"epoch": 0.44154302670623147,
"grad_norm": 0.7390718460083008,
"learning_rate": 4.6117236808601495e-05,
"loss": 0.8595,
"step": 3720
},
{
"epoch": 0.4427299703264095,
"grad_norm": 0.8652002811431885,
"learning_rate": 4.6090934939960894e-05,
"loss": 0.9118,
"step": 3730
},
{
"epoch": 0.44391691394658755,
"grad_norm": 0.7718791365623474,
"learning_rate": 4.606455184041622e-05,
"loss": 0.9111,
"step": 3740
},
{
"epoch": 0.44510385756676557,
"grad_norm": 0.7578750848770142,
"learning_rate": 4.603808761158097e-05,
"loss": 0.831,
"step": 3750
},
{
"epoch": 0.44629080118694364,
"grad_norm": 0.7663601636886597,
"learning_rate": 4.6011542355381055e-05,
"loss": 0.8584,
"step": 3760
},
{
"epoch": 0.44747774480712166,
"grad_norm": 0.9673477411270142,
"learning_rate": 4.5984916174054494e-05,
"loss": 0.8244,
"step": 3770
},
{
"epoch": 0.44866468842729973,
"grad_norm": 0.8995377421379089,
"learning_rate": 4.5958209170150956e-05,
"loss": 0.8669,
"step": 3780
},
{
"epoch": 0.44985163204747775,
"grad_norm": 0.7621747851371765,
"learning_rate": 4.593142144653143e-05,
"loss": 0.7874,
"step": 3790
},
{
"epoch": 0.45103857566765576,
"grad_norm": 0.9863945841789246,
"learning_rate": 4.5904553106367774e-05,
"loss": 0.8802,
"step": 3800
},
{
"epoch": 0.45222551928783383,
"grad_norm": 0.9811915159225464,
"learning_rate": 4.5877604253142336e-05,
"loss": 0.8655,
"step": 3810
},
{
"epoch": 0.45341246290801185,
"grad_norm": 0.9134243130683899,
"learning_rate": 4.5850574990647566e-05,
"loss": 0.843,
"step": 3820
},
{
"epoch": 0.4545994065281899,
"grad_norm": 0.788000226020813,
"learning_rate": 4.58234654229856e-05,
"loss": 0.9024,
"step": 3830
},
{
"epoch": 0.45578635014836794,
"grad_norm": 0.9368435740470886,
"learning_rate": 4.579627565456786e-05,
"loss": 0.8806,
"step": 3840
},
{
"epoch": 0.456973293768546,
"grad_norm": 0.9639891982078552,
"learning_rate": 4.576900579011469e-05,
"loss": 0.8829,
"step": 3850
},
{
"epoch": 0.458160237388724,
"grad_norm": 0.9119428992271423,
"learning_rate": 4.574165593465487e-05,
"loss": 0.8648,
"step": 3860
},
{
"epoch": 0.4593471810089021,
"grad_norm": 0.9542068243026733,
"learning_rate": 4.571422619352531e-05,
"loss": 0.9334,
"step": 3870
},
{
"epoch": 0.4605341246290801,
"grad_norm": 0.9423951506614685,
"learning_rate": 4.5686716672370574e-05,
"loss": 0.8339,
"step": 3880
},
{
"epoch": 0.4617210682492582,
"grad_norm": 0.8227980732917786,
"learning_rate": 4.565912747714248e-05,
"loss": 0.9334,
"step": 3890
},
{
"epoch": 0.4629080118694362,
"grad_norm": 0.5783571004867554,
"learning_rate": 4.563145871409974e-05,
"loss": 0.818,
"step": 3900
},
{
"epoch": 0.46409495548961427,
"grad_norm": 0.8463440537452698,
"learning_rate": 4.560371048980751e-05,
"loss": 0.864,
"step": 3910
},
{
"epoch": 0.4652818991097923,
"grad_norm": 0.6289271116256714,
"learning_rate": 4.5575882911136966e-05,
"loss": 0.9059,
"step": 3920
},
{
"epoch": 0.4664688427299703,
"grad_norm": 0.9256350994110107,
"learning_rate": 4.554797608526493e-05,
"loss": 0.8776,
"step": 3930
},
{
"epoch": 0.4676557863501484,
"grad_norm": 0.9969144463539124,
"learning_rate": 4.5519990119673454e-05,
"loss": 0.8477,
"step": 3940
},
{
"epoch": 0.4688427299703264,
"grad_norm": 0.7296236753463745,
"learning_rate": 4.5491925122149373e-05,
"loss": 0.9374,
"step": 3950
},
{
"epoch": 0.47002967359050446,
"grad_norm": 0.9291085004806519,
"learning_rate": 4.546378120078392e-05,
"loss": 0.9612,
"step": 3960
},
{
"epoch": 0.4712166172106825,
"grad_norm": 0.977351188659668,
"learning_rate": 4.543555846397229e-05,
"loss": 0.8603,
"step": 3970
},
{
"epoch": 0.47240356083086055,
"grad_norm": 0.6831493377685547,
"learning_rate": 4.540725702041326e-05,
"loss": 0.8115,
"step": 3980
},
{
"epoch": 0.47359050445103856,
"grad_norm": 1.238186001777649,
"learning_rate": 4.537887697910871e-05,
"loss": 0.7757,
"step": 3990
},
{
"epoch": 0.47477744807121663,
"grad_norm": 0.8050864338874817,
"learning_rate": 4.535041844936324e-05,
"loss": 0.8233,
"step": 4000
},
{
"epoch": 0.47477744807121663,
"eval_loss": 0.7993733286857605,
"eval_runtime": 773.0065,
"eval_samples_per_second": 2.295,
"eval_steps_per_second": 0.574,
"step": 4000
},
{
"epoch": 0.47596439169139465,
"grad_norm": 0.8855427503585815,
"learning_rate": 4.532188154078377e-05,
"loss": 0.82,
"step": 4010
},
{
"epoch": 0.4771513353115727,
"grad_norm": 0.9453259706497192,
"learning_rate": 4.5293266363279084e-05,
"loss": 0.9316,
"step": 4020
},
{
"epoch": 0.47833827893175074,
"grad_norm": 0.8510525226593018,
"learning_rate": 4.526457302705939e-05,
"loss": 0.8287,
"step": 4030
},
{
"epoch": 0.4795252225519288,
"grad_norm": 0.757831871509552,
"learning_rate": 4.5235801642635956e-05,
"loss": 0.8773,
"step": 4040
},
{
"epoch": 0.4807121661721068,
"grad_norm": 0.649082362651825,
"learning_rate": 4.520695232082062e-05,
"loss": 0.8056,
"step": 4050
},
{
"epoch": 0.48189910979228484,
"grad_norm": 0.7211092710494995,
"learning_rate": 4.517802517272542e-05,
"loss": 0.8503,
"step": 4060
},
{
"epoch": 0.4830860534124629,
"grad_norm": 1.1383498907089233,
"learning_rate": 4.514902030976212e-05,
"loss": 0.9434,
"step": 4070
},
{
"epoch": 0.48427299703264093,
"grad_norm": 0.8772423267364502,
"learning_rate": 4.5119937843641816e-05,
"loss": 0.8708,
"step": 4080
},
{
"epoch": 0.485459940652819,
"grad_norm": 0.8966817259788513,
"learning_rate": 4.509077788637446e-05,
"loss": 0.881,
"step": 4090
},
{
"epoch": 0.486646884272997,
"grad_norm": 0.7666980624198914,
"learning_rate": 4.506154055026849e-05,
"loss": 0.8542,
"step": 4100
},
{
"epoch": 0.4878338278931751,
"grad_norm": 0.8315165042877197,
"learning_rate": 4.5032225947930354e-05,
"loss": 0.8535,
"step": 4110
},
{
"epoch": 0.4890207715133531,
"grad_norm": 0.9881699681282043,
"learning_rate": 4.5002834192264074e-05,
"loss": 0.9076,
"step": 4120
},
{
"epoch": 0.4902077151335312,
"grad_norm": 0.7229058146476746,
"learning_rate": 4.497336539647086e-05,
"loss": 0.8915,
"step": 4130
},
{
"epoch": 0.4913946587537092,
"grad_norm": 0.8700774312019348,
"learning_rate": 4.494381967404859e-05,
"loss": 0.8986,
"step": 4140
},
{
"epoch": 0.49258160237388726,
"grad_norm": 0.8586352467536926,
"learning_rate": 4.491419713879147e-05,
"loss": 0.886,
"step": 4150
},
{
"epoch": 0.4937685459940653,
"grad_norm": 0.9397504329681396,
"learning_rate": 4.488449790478952e-05,
"loss": 0.8476,
"step": 4160
},
{
"epoch": 0.49495548961424335,
"grad_norm": 0.9243069887161255,
"learning_rate": 4.4854722086428155e-05,
"loss": 0.8302,
"step": 4170
},
{
"epoch": 0.49614243323442137,
"grad_norm": 0.9511041641235352,
"learning_rate": 4.482486979838779e-05,
"loss": 0.8928,
"step": 4180
},
{
"epoch": 0.4973293768545994,
"grad_norm": 0.7937717437744141,
"learning_rate": 4.479494115564332e-05,
"loss": 0.8834,
"step": 4190
},
{
"epoch": 0.49851632047477745,
"grad_norm": 0.7729777693748474,
"learning_rate": 4.4764936273463734e-05,
"loss": 0.8346,
"step": 4200
},
{
"epoch": 0.49970326409495547,
"grad_norm": 0.8554561734199524,
"learning_rate": 4.473485526741164e-05,
"loss": 0.9021,
"step": 4210
},
{
"epoch": 0.5008902077151335,
"grad_norm": 1.142722725868225,
"learning_rate": 4.470469825334287e-05,
"loss": 0.8927,
"step": 4220
},
{
"epoch": 0.5020771513353116,
"grad_norm": 0.8637948632240295,
"learning_rate": 4.467446534740596e-05,
"loss": 0.9063,
"step": 4230
},
{
"epoch": 0.5032640949554896,
"grad_norm": 0.7488009929656982,
"learning_rate": 4.464415666604176e-05,
"loss": 0.8494,
"step": 4240
},
{
"epoch": 0.5044510385756676,
"grad_norm": 1.1592786312103271,
"learning_rate": 4.4613772325982954e-05,
"loss": 0.8737,
"step": 4250
},
{
"epoch": 0.5056379821958457,
"grad_norm": 0.8699279427528381,
"learning_rate": 4.458331244425364e-05,
"loss": 0.8825,
"step": 4260
},
{
"epoch": 0.5068249258160238,
"grad_norm": 0.8104872703552246,
"learning_rate": 4.4552777138168836e-05,
"loss": 0.8296,
"step": 4270
},
{
"epoch": 0.5080118694362018,
"grad_norm": 0.9313340187072754,
"learning_rate": 4.452216652533409e-05,
"loss": 0.92,
"step": 4280
},
{
"epoch": 0.5091988130563798,
"grad_norm": 1.2348231077194214,
"learning_rate": 4.4491480723644944e-05,
"loss": 0.9548,
"step": 4290
},
{
"epoch": 0.5103857566765578,
"grad_norm": 0.7835610508918762,
"learning_rate": 4.446071985128657e-05,
"loss": 0.8719,
"step": 4300
},
{
"epoch": 0.511572700296736,
"grad_norm": 0.7401361465454102,
"learning_rate": 4.442988402673326e-05,
"loss": 0.9493,
"step": 4310
},
{
"epoch": 0.512759643916914,
"grad_norm": 0.9936283826828003,
"learning_rate": 4.439897336874796e-05,
"loss": 0.849,
"step": 4320
},
{
"epoch": 0.513946587537092,
"grad_norm": 0.7580452561378479,
"learning_rate": 4.436798799638186e-05,
"loss": 0.8755,
"step": 4330
},
{
"epoch": 0.51513353115727,
"grad_norm": 0.6950749754905701,
"learning_rate": 4.433692802897389e-05,
"loss": 0.8511,
"step": 4340
},
{
"epoch": 0.516320474777448,
"grad_norm": 0.9436466693878174,
"learning_rate": 4.43057935861503e-05,
"loss": 0.8442,
"step": 4350
},
{
"epoch": 0.5175074183976262,
"grad_norm": 0.6432386636734009,
"learning_rate": 4.4274584787824166e-05,
"loss": 0.8074,
"step": 4360
},
{
"epoch": 0.5186943620178042,
"grad_norm": 0.6607010364532471,
"learning_rate": 4.424330175419495e-05,
"loss": 0.8124,
"step": 4370
},
{
"epoch": 0.5198813056379822,
"grad_norm": 0.7476988434791565,
"learning_rate": 4.421194460574801e-05,
"loss": 0.9036,
"step": 4380
},
{
"epoch": 0.5210682492581602,
"grad_norm": 0.6674066781997681,
"learning_rate": 4.418051346325417e-05,
"loss": 0.8132,
"step": 4390
},
{
"epoch": 0.5222551928783383,
"grad_norm": 1.0428473949432373,
"learning_rate": 4.414900844776924e-05,
"loss": 0.897,
"step": 4400
},
{
"epoch": 0.5234421364985163,
"grad_norm": 1.014649510383606,
"learning_rate": 4.411742968063354e-05,
"loss": 0.8464,
"step": 4410
},
{
"epoch": 0.5246290801186944,
"grad_norm": 0.7525945901870728,
"learning_rate": 4.408577728347144e-05,
"loss": 0.8195,
"step": 4420
},
{
"epoch": 0.5258160237388724,
"grad_norm": 0.8915858268737793,
"learning_rate": 4.4054051378190915e-05,
"loss": 0.8397,
"step": 4430
},
{
"epoch": 0.5270029673590505,
"grad_norm": 0.9980171918869019,
"learning_rate": 4.402225208698303e-05,
"loss": 0.8515,
"step": 4440
},
{
"epoch": 0.5281899109792285,
"grad_norm": 0.862018883228302,
"learning_rate": 4.3990379532321495e-05,
"loss": 0.9129,
"step": 4450
},
{
"epoch": 0.5293768545994065,
"grad_norm": 0.6026172041893005,
"learning_rate": 4.395843383696221e-05,
"loss": 0.9185,
"step": 4460
},
{
"epoch": 0.5305637982195845,
"grad_norm": 0.8191978335380554,
"learning_rate": 4.392641512394274e-05,
"loss": 0.8652,
"step": 4470
},
{
"epoch": 0.5317507418397626,
"grad_norm": 0.7325008511543274,
"learning_rate": 4.389432351658193e-05,
"loss": 0.8111,
"step": 4480
},
{
"epoch": 0.5329376854599407,
"grad_norm": 0.7030380368232727,
"learning_rate": 4.3862159138479305e-05,
"loss": 0.8795,
"step": 4490
},
{
"epoch": 0.5341246290801187,
"grad_norm": 1.045393466949463,
"learning_rate": 4.382992211351471e-05,
"loss": 0.9116,
"step": 4500
},
{
"epoch": 0.5341246290801187,
"eval_loss": 0.7951666116714478,
"eval_runtime": 772.1963,
"eval_samples_per_second": 2.297,
"eval_steps_per_second": 0.575,
"step": 4500
},
{
"epoch": 0.5353115727002967,
"grad_norm": 0.9076895713806152,
"learning_rate": 4.3797612565847785e-05,
"loss": 0.8539,
"step": 4510
},
{
"epoch": 0.5364985163204747,
"grad_norm": 0.7308244109153748,
"learning_rate": 4.376523061991747e-05,
"loss": 0.8028,
"step": 4520
},
{
"epoch": 0.5376854599406529,
"grad_norm": 0.8944220542907715,
"learning_rate": 4.373277640044156e-05,
"loss": 0.8532,
"step": 4530
},
{
"epoch": 0.5388724035608309,
"grad_norm": 0.7320210337638855,
"learning_rate": 4.370025003241618e-05,
"loss": 0.8444,
"step": 4540
},
{
"epoch": 0.5400593471810089,
"grad_norm": 0.7950882315635681,
"learning_rate": 4.366765164111538e-05,
"loss": 0.8683,
"step": 4550
},
{
"epoch": 0.5412462908011869,
"grad_norm": 0.9182088375091553,
"learning_rate": 4.3634981352090555e-05,
"loss": 0.8671,
"step": 4560
},
{
"epoch": 0.542433234421365,
"grad_norm": 0.943240761756897,
"learning_rate": 4.3602239291170056e-05,
"loss": 0.8846,
"step": 4570
},
{
"epoch": 0.5436201780415431,
"grad_norm": 0.8208020925521851,
"learning_rate": 4.356942558445862e-05,
"loss": 0.8565,
"step": 4580
},
{
"epoch": 0.5448071216617211,
"grad_norm": 1.0908632278442383,
"learning_rate": 4.353654035833697e-05,
"loss": 0.9188,
"step": 4590
},
{
"epoch": 0.5459940652818991,
"grad_norm": 0.796347439289093,
"learning_rate": 4.3503583739461235e-05,
"loss": 0.8693,
"step": 4600
},
{
"epoch": 0.5471810089020771,
"grad_norm": 0.9802513718605042,
"learning_rate": 4.347055585476254e-05,
"loss": 0.8915,
"step": 4610
},
{
"epoch": 0.5483679525222552,
"grad_norm": 0.9925609827041626,
"learning_rate": 4.343745683144647e-05,
"loss": 0.9135,
"step": 4620
},
{
"epoch": 0.5495548961424332,
"grad_norm": 0.9672980904579163,
"learning_rate": 4.340428679699262e-05,
"loss": 0.8645,
"step": 4630
},
{
"epoch": 0.5507418397626113,
"grad_norm": 0.7970063090324402,
"learning_rate": 4.337104587915407e-05,
"loss": 0.8438,
"step": 4640
},
{
"epoch": 0.5519287833827893,
"grad_norm": 0.8159870505332947,
"learning_rate": 4.3337734205956905e-05,
"loss": 0.8973,
"step": 4650
},
{
"epoch": 0.5531157270029674,
"grad_norm": 0.7052998542785645,
"learning_rate": 4.3304351905699714e-05,
"loss": 0.9022,
"step": 4660
},
{
"epoch": 0.5543026706231454,
"grad_norm": 1.0930620431900024,
"learning_rate": 4.3270899106953105e-05,
"loss": 0.8904,
"step": 4670
},
{
"epoch": 0.5554896142433234,
"grad_norm": 0.8277913928031921,
"learning_rate": 4.323737593855922e-05,
"loss": 0.9505,
"step": 4680
},
{
"epoch": 0.5566765578635015,
"grad_norm": 1.1288316249847412,
"learning_rate": 4.320378252963123e-05,
"loss": 0.8257,
"step": 4690
},
{
"epoch": 0.5578635014836796,
"grad_norm": 0.7659019231796265,
"learning_rate": 4.31701190095528e-05,
"loss": 0.8508,
"step": 4700
},
{
"epoch": 0.5590504451038576,
"grad_norm": 0.7803889513015747,
"learning_rate": 4.3136385507977674e-05,
"loss": 0.9115,
"step": 4710
},
{
"epoch": 0.5602373887240356,
"grad_norm": 1.101162075996399,
"learning_rate": 4.310258215482909e-05,
"loss": 0.8446,
"step": 4720
},
{
"epoch": 0.5614243323442136,
"grad_norm": 0.9544945955276489,
"learning_rate": 4.306870908029933e-05,
"loss": 0.9256,
"step": 4730
},
{
"epoch": 0.5626112759643916,
"grad_norm": 0.6556288003921509,
"learning_rate": 4.303476641484921e-05,
"loss": 0.8348,
"step": 4740
},
{
"epoch": 0.5637982195845698,
"grad_norm": 0.7247066497802734,
"learning_rate": 4.300075428920756e-05,
"loss": 0.8539,
"step": 4750
},
{
"epoch": 0.5649851632047478,
"grad_norm": 0.8974981904029846,
"learning_rate": 4.2966672834370746e-05,
"loss": 0.8447,
"step": 4760
},
{
"epoch": 0.5661721068249258,
"grad_norm": 0.7167539596557617,
"learning_rate": 4.2932522181602146e-05,
"loss": 0.881,
"step": 4770
},
{
"epoch": 0.5673590504451038,
"grad_norm": 0.8274498581886292,
"learning_rate": 4.289830246243165e-05,
"loss": 0.834,
"step": 4780
},
{
"epoch": 0.568545994065282,
"grad_norm": 0.9641168117523193,
"learning_rate": 4.2864013808655155e-05,
"loss": 0.7953,
"step": 4790
},
{
"epoch": 0.56973293768546,
"grad_norm": 1.0160974264144897,
"learning_rate": 4.282965635233407e-05,
"loss": 0.9052,
"step": 4800
},
{
"epoch": 0.570919881305638,
"grad_norm": 0.7646157145500183,
"learning_rate": 4.279523022579477e-05,
"loss": 0.8616,
"step": 4810
},
{
"epoch": 0.572106824925816,
"grad_norm": 0.6680961847305298,
"learning_rate": 4.276073556162813e-05,
"loss": 0.8366,
"step": 4820
},
{
"epoch": 0.5732937685459941,
"grad_norm": 0.5522508025169373,
"learning_rate": 4.272617249268899e-05,
"loss": 0.7855,
"step": 4830
},
{
"epoch": 0.5744807121661721,
"grad_norm": 0.989126443862915,
"learning_rate": 4.2691541152095636e-05,
"loss": 0.8128,
"step": 4840
},
{
"epoch": 0.5756676557863502,
"grad_norm": 0.8510265946388245,
"learning_rate": 4.265684167322932e-05,
"loss": 0.8525,
"step": 4850
},
{
"epoch": 0.5768545994065282,
"grad_norm": 0.9399095177650452,
"learning_rate": 4.2622074189733715e-05,
"loss": 0.801,
"step": 4860
},
{
"epoch": 0.5780415430267062,
"grad_norm": 0.8262656331062317,
"learning_rate": 4.258723883551441e-05,
"loss": 0.8555,
"step": 4870
},
{
"epoch": 0.5792284866468843,
"grad_norm": 0.9158207178115845,
"learning_rate": 4.25523357447384e-05,
"loss": 0.8475,
"step": 4880
},
{
"epoch": 0.5804154302670623,
"grad_norm": 0.8734104633331299,
"learning_rate": 4.251736505183356e-05,
"loss": 0.8769,
"step": 4890
},
{
"epoch": 0.5816023738872403,
"grad_norm": 0.6260775923728943,
"learning_rate": 4.2482326891488146e-05,
"loss": 0.9077,
"step": 4900
},
{
"epoch": 0.5827893175074184,
"grad_norm": 0.7941166758537292,
"learning_rate": 4.244722139865023e-05,
"loss": 0.8298,
"step": 4910
},
{
"epoch": 0.5839762611275965,
"grad_norm": 1.005319356918335,
"learning_rate": 4.241204870852725e-05,
"loss": 0.8613,
"step": 4920
},
{
"epoch": 0.5851632047477745,
"grad_norm": 0.7308863401412964,
"learning_rate": 4.237680895658543e-05,
"loss": 0.8962,
"step": 4930
},
{
"epoch": 0.5863501483679525,
"grad_norm": 0.7635941505432129,
"learning_rate": 4.2341502278549285e-05,
"loss": 0.7996,
"step": 4940
},
{
"epoch": 0.5875370919881305,
"grad_norm": 0.8385136723518372,
"learning_rate": 4.230612881040108e-05,
"loss": 0.8976,
"step": 4950
},
{
"epoch": 0.5887240356083087,
"grad_norm": 1.0231964588165283,
"learning_rate": 4.227068868838035e-05,
"loss": 0.8559,
"step": 4960
},
{
"epoch": 0.5899109792284867,
"grad_norm": 1.0560635328292847,
"learning_rate": 4.223518204898332e-05,
"loss": 0.8956,
"step": 4970
},
{
"epoch": 0.5910979228486647,
"grad_norm": 0.8823438286781311,
"learning_rate": 4.2199609028962414e-05,
"loss": 0.8867,
"step": 4980
},
{
"epoch": 0.5922848664688427,
"grad_norm": 0.8164929747581482,
"learning_rate": 4.216396976532571e-05,
"loss": 0.8373,
"step": 4990
},
{
"epoch": 0.5934718100890207,
"grad_norm": 1.1652288436889648,
"learning_rate": 4.2128264395336436e-05,
"loss": 0.8498,
"step": 5000
},
{
"epoch": 0.5934718100890207,
"eval_loss": 0.7921267747879028,
"eval_runtime": 774.3302,
"eval_samples_per_second": 2.291,
"eval_steps_per_second": 0.573,
"step": 5000
},
{
"epoch": 0.5946587537091989,
"grad_norm": 0.7231582999229431,
"learning_rate": 4.20924930565124e-05,
"loss": 0.8121,
"step": 5010
},
{
"epoch": 0.5958456973293769,
"grad_norm": 0.8947294354438782,
"learning_rate": 4.205665588662553e-05,
"loss": 0.8989,
"step": 5020
},
{
"epoch": 0.5970326409495549,
"grad_norm": 0.9510688781738281,
"learning_rate": 4.202075302370124e-05,
"loss": 0.8759,
"step": 5030
},
{
"epoch": 0.5982195845697329,
"grad_norm": 0.892690896987915,
"learning_rate": 4.198478460601801e-05,
"loss": 0.8538,
"step": 5040
},
{
"epoch": 0.599406528189911,
"grad_norm": 0.8187076449394226,
"learning_rate": 4.194875077210677e-05,
"loss": 0.852,
"step": 5050
},
{
"epoch": 0.600593471810089,
"grad_norm": 0.9023217558860779,
"learning_rate": 4.191265166075043e-05,
"loss": 0.8713,
"step": 5060
},
{
"epoch": 0.6017804154302671,
"grad_norm": 0.797012209892273,
"learning_rate": 4.1876487410983275e-05,
"loss": 0.8255,
"step": 5070
},
{
"epoch": 0.6029673590504451,
"grad_norm": 0.7908094525337219,
"learning_rate": 4.184025816209049e-05,
"loss": 0.8534,
"step": 5080
},
{
"epoch": 0.6041543026706232,
"grad_norm": 0.8755646347999573,
"learning_rate": 4.180396405360761e-05,
"loss": 0.8395,
"step": 5090
},
{
"epoch": 0.6053412462908012,
"grad_norm": 1.0383625030517578,
"learning_rate": 4.176760522531995e-05,
"loss": 0.8838,
"step": 5100
},
{
"epoch": 0.6065281899109792,
"grad_norm": 0.937275767326355,
"learning_rate": 4.17311818172621e-05,
"loss": 0.8722,
"step": 5110
},
{
"epoch": 0.6077151335311572,
"grad_norm": 0.9110156297683716,
"learning_rate": 4.169469396971739e-05,
"loss": 0.8799,
"step": 5120
},
{
"epoch": 0.6089020771513353,
"grad_norm": 1.0142481327056885,
"learning_rate": 4.165814182321732e-05,
"loss": 0.8696,
"step": 5130
},
{
"epoch": 0.6100890207715134,
"grad_norm": 1.2037073373794556,
"learning_rate": 4.162152551854105e-05,
"loss": 0.8787,
"step": 5140
},
{
"epoch": 0.6112759643916914,
"grad_norm": 1.0719590187072754,
"learning_rate": 4.1584845196714815e-05,
"loss": 0.9276,
"step": 5150
},
{
"epoch": 0.6124629080118694,
"grad_norm": 0.6941603422164917,
"learning_rate": 4.1548100999011444e-05,
"loss": 0.8573,
"step": 5160
},
{
"epoch": 0.6136498516320474,
"grad_norm": 1.162614107131958,
"learning_rate": 4.1511293066949766e-05,
"loss": 0.852,
"step": 5170
},
{
"epoch": 0.6148367952522256,
"grad_norm": 0.8986325860023499,
"learning_rate": 4.147442154229406e-05,
"loss": 0.8469,
"step": 5180
},
{
"epoch": 0.6160237388724036,
"grad_norm": 0.7604609131813049,
"learning_rate": 4.1437486567053565e-05,
"loss": 0.7998,
"step": 5190
},
{
"epoch": 0.6172106824925816,
"grad_norm": 0.7332816123962402,
"learning_rate": 4.140048828348188e-05,
"loss": 0.8393,
"step": 5200
},
{
"epoch": 0.6183976261127596,
"grad_norm": 1.0732022523880005,
"learning_rate": 4.1363426834076435e-05,
"loss": 0.8923,
"step": 5210
},
{
"epoch": 0.6195845697329377,
"grad_norm": 0.7702149748802185,
"learning_rate": 4.132630236157793e-05,
"loss": 0.8341,
"step": 5220
},
{
"epoch": 0.6207715133531158,
"grad_norm": 0.7818199992179871,
"learning_rate": 4.128911500896983e-05,
"loss": 0.8694,
"step": 5230
},
{
"epoch": 0.6219584569732938,
"grad_norm": 1.0020356178283691,
"learning_rate": 4.1251864919477736e-05,
"loss": 0.7892,
"step": 5240
},
{
"epoch": 0.6231454005934718,
"grad_norm": 0.595758855342865,
"learning_rate": 4.12145522365689e-05,
"loss": 0.7445,
"step": 5250
},
{
"epoch": 0.6243323442136498,
"grad_norm": 0.8680073618888855,
"learning_rate": 4.117717710395166e-05,
"loss": 0.8066,
"step": 5260
},
{
"epoch": 0.6255192878338279,
"grad_norm": 1.1112210750579834,
"learning_rate": 4.1139739665574856e-05,
"loss": 0.8548,
"step": 5270
},
{
"epoch": 0.626706231454006,
"grad_norm": 0.8211454153060913,
"learning_rate": 4.11022400656273e-05,
"loss": 0.8481,
"step": 5280
},
{
"epoch": 0.627893175074184,
"grad_norm": 0.9181323051452637,
"learning_rate": 4.1064678448537244e-05,
"loss": 0.8785,
"step": 5290
},
{
"epoch": 0.629080118694362,
"grad_norm": 1.1044420003890991,
"learning_rate": 4.1027054958971746e-05,
"loss": 0.8726,
"step": 5300
},
{
"epoch": 0.6302670623145401,
"grad_norm": 0.8787449598312378,
"learning_rate": 4.098936974183621e-05,
"loss": 0.8111,
"step": 5310
},
{
"epoch": 0.6314540059347181,
"grad_norm": 0.7336760759353638,
"learning_rate": 4.095162294227377e-05,
"loss": 0.8404,
"step": 5320
},
{
"epoch": 0.6326409495548961,
"grad_norm": 0.8776949048042297,
"learning_rate": 4.091381470566472e-05,
"loss": 0.8317,
"step": 5330
},
{
"epoch": 0.6338278931750742,
"grad_norm": 1.099990963935852,
"learning_rate": 4.0875945177625996e-05,
"loss": 0.8799,
"step": 5340
},
{
"epoch": 0.6350148367952523,
"grad_norm": 0.7875162959098816,
"learning_rate": 4.08380145040106e-05,
"loss": 0.8337,
"step": 5350
},
{
"epoch": 0.6362017804154303,
"grad_norm": 0.8717107772827148,
"learning_rate": 4.0800022830907015e-05,
"loss": 0.8433,
"step": 5360
},
{
"epoch": 0.6373887240356083,
"grad_norm": 0.9052039980888367,
"learning_rate": 4.076197030463868e-05,
"loss": 0.9926,
"step": 5370
},
{
"epoch": 0.6385756676557863,
"grad_norm": 0.9107869863510132,
"learning_rate": 4.072385707176338e-05,
"loss": 0.8785,
"step": 5380
},
{
"epoch": 0.6397626112759643,
"grad_norm": 0.8352899551391602,
"learning_rate": 4.068568327907275e-05,
"loss": 0.8804,
"step": 5390
},
{
"epoch": 0.6409495548961425,
"grad_norm": 1.1323997974395752,
"learning_rate": 4.064744907359164e-05,
"loss": 0.8881,
"step": 5400
},
{
"epoch": 0.6421364985163205,
"grad_norm": 0.8301699161529541,
"learning_rate": 4.060915460257759e-05,
"loss": 0.8865,
"step": 5410
},
{
"epoch": 0.6433234421364985,
"grad_norm": 0.9869147539138794,
"learning_rate": 4.057080001352023e-05,
"loss": 0.8698,
"step": 5420
},
{
"epoch": 0.6445103857566765,
"grad_norm": 0.8932684063911438,
"learning_rate": 4.0532385454140764e-05,
"loss": 0.8295,
"step": 5430
},
{
"epoch": 0.6456973293768546,
"grad_norm": 0.9118059277534485,
"learning_rate": 4.0493911072391344e-05,
"loss": 0.8094,
"step": 5440
},
{
"epoch": 0.6468842729970327,
"grad_norm": 0.9855166673660278,
"learning_rate": 4.0455377016454526e-05,
"loss": 0.864,
"step": 5450
},
{
"epoch": 0.6480712166172107,
"grad_norm": 0.7110412120819092,
"learning_rate": 4.041678343474271e-05,
"loss": 0.887,
"step": 5460
},
{
"epoch": 0.6492581602373887,
"grad_norm": 0.886139988899231,
"learning_rate": 4.037813047589757e-05,
"loss": 0.8107,
"step": 5470
},
{
"epoch": 0.6504451038575668,
"grad_norm": 0.9182664155960083,
"learning_rate": 4.0339418288789444e-05,
"loss": 0.8721,
"step": 5480
},
{
"epoch": 0.6516320474777448,
"grad_norm": 0.9091856479644775,
"learning_rate": 4.030064702251678e-05,
"loss": 0.8063,
"step": 5490
},
{
"epoch": 0.6528189910979229,
"grad_norm": 0.971271812915802,
"learning_rate": 4.02618168264056e-05,
"loss": 0.934,
"step": 5500
},
{
"epoch": 0.6528189910979229,
"eval_loss": 0.7894634008407593,
"eval_runtime": 773.6717,
"eval_samples_per_second": 2.293,
"eval_steps_per_second": 0.574,
"step": 5500
},
{
"epoch": 0.6540059347181009,
"grad_norm": 1.1243486404418945,
"learning_rate": 4.022292785000885e-05,
"loss": 0.8822,
"step": 5510
},
{
"epoch": 0.6551928783382789,
"grad_norm": 0.9398186206817627,
"learning_rate": 4.0183980243105915e-05,
"loss": 0.8621,
"step": 5520
},
{
"epoch": 0.656379821958457,
"grad_norm": 0.7023178935050964,
"learning_rate": 4.0144974155701954e-05,
"loss": 0.8111,
"step": 5530
},
{
"epoch": 0.657566765578635,
"grad_norm": 0.939687192440033,
"learning_rate": 4.0105909738027365e-05,
"loss": 0.8812,
"step": 5540
},
{
"epoch": 0.658753709198813,
"grad_norm": 1.0677947998046875,
"learning_rate": 4.006678714053723e-05,
"loss": 0.8459,
"step": 5550
},
{
"epoch": 0.6599406528189911,
"grad_norm": 0.9122028350830078,
"learning_rate": 4.002760651391068e-05,
"loss": 0.8306,
"step": 5560
},
{
"epoch": 0.6611275964391692,
"grad_norm": 0.8712452054023743,
"learning_rate": 3.9988368009050355e-05,
"loss": 0.8285,
"step": 5570
},
{
"epoch": 0.6623145400593472,
"grad_norm": 0.7736614942550659,
"learning_rate": 3.994907177708181e-05,
"loss": 0.8848,
"step": 5580
},
{
"epoch": 0.6635014836795252,
"grad_norm": 0.9067423343658447,
"learning_rate": 3.990971796935293e-05,
"loss": 0.9422,
"step": 5590
},
{
"epoch": 0.6646884272997032,
"grad_norm": 0.8553429841995239,
"learning_rate": 3.987030673743335e-05,
"loss": 0.8574,
"step": 5600
},
{
"epoch": 0.6658753709198814,
"grad_norm": 0.8687973618507385,
"learning_rate": 3.9830838233113884e-05,
"loss": 0.8979,
"step": 5610
},
{
"epoch": 0.6670623145400594,
"grad_norm": 0.9838027954101562,
"learning_rate": 3.9791312608405916e-05,
"loss": 0.8126,
"step": 5620
},
{
"epoch": 0.6682492581602374,
"grad_norm": 0.9024927020072937,
"learning_rate": 3.9751730015540835e-05,
"loss": 0.8863,
"step": 5630
},
{
"epoch": 0.6694362017804154,
"grad_norm": 0.7636504769325256,
"learning_rate": 3.9712090606969425e-05,
"loss": 0.881,
"step": 5640
},
{
"epoch": 0.6706231454005934,
"grad_norm": 1.080702543258667,
"learning_rate": 3.967239453536131e-05,
"loss": 0.8608,
"step": 5650
},
{
"epoch": 0.6718100890207716,
"grad_norm": 0.9500882029533386,
"learning_rate": 3.963264195360435e-05,
"loss": 0.8783,
"step": 5660
},
{
"epoch": 0.6729970326409496,
"grad_norm": 0.8997413516044617,
"learning_rate": 3.9592833014804044e-05,
"loss": 0.7662,
"step": 5670
},
{
"epoch": 0.6741839762611276,
"grad_norm": 1.0616391897201538,
"learning_rate": 3.955296787228294e-05,
"loss": 0.9328,
"step": 5680
},
{
"epoch": 0.6753709198813056,
"grad_norm": 0.9198963642120361,
"learning_rate": 3.9513046679580075e-05,
"loss": 0.8603,
"step": 5690
},
{
"epoch": 0.6765578635014837,
"grad_norm": 0.8361203670501709,
"learning_rate": 3.947306959045034e-05,
"loss": 0.8299,
"step": 5700
},
{
"epoch": 0.6777448071216617,
"grad_norm": 0.8562357425689697,
"learning_rate": 3.943303675886391e-05,
"loss": 0.8098,
"step": 5710
},
{
"epoch": 0.6789317507418398,
"grad_norm": 0.8105937242507935,
"learning_rate": 3.939294833900568e-05,
"loss": 0.8345,
"step": 5720
},
{
"epoch": 0.6801186943620178,
"grad_norm": 1.072594165802002,
"learning_rate": 3.9352804485274586e-05,
"loss": 0.8251,
"step": 5730
},
{
"epoch": 0.6813056379821959,
"grad_norm": 0.7309070229530334,
"learning_rate": 3.9312605352283124e-05,
"loss": 0.8493,
"step": 5740
},
{
"epoch": 0.6824925816023739,
"grad_norm": 0.9707894921302795,
"learning_rate": 3.9272351094856665e-05,
"loss": 0.9283,
"step": 5750
},
{
"epoch": 0.6836795252225519,
"grad_norm": 0.742894172668457,
"learning_rate": 3.9232041868032885e-05,
"loss": 0.8594,
"step": 5760
},
{
"epoch": 0.68486646884273,
"grad_norm": 0.8216899633407593,
"learning_rate": 3.919167782706119e-05,
"loss": 0.8136,
"step": 5770
},
{
"epoch": 0.686053412462908,
"grad_norm": 0.9572916626930237,
"learning_rate": 3.915125912740208e-05,
"loss": 0.8865,
"step": 5780
},
{
"epoch": 0.6872403560830861,
"grad_norm": 0.8739311695098877,
"learning_rate": 3.911078592472659e-05,
"loss": 0.9605,
"step": 5790
},
{
"epoch": 0.6884272997032641,
"grad_norm": 0.8961655497550964,
"learning_rate": 3.907025837491566e-05,
"loss": 0.7693,
"step": 5800
},
{
"epoch": 0.6896142433234421,
"grad_norm": 0.6542685627937317,
"learning_rate": 3.902967663405956e-05,
"loss": 0.7998,
"step": 5810
},
{
"epoch": 0.6908011869436201,
"grad_norm": 0.7300246953964233,
"learning_rate": 3.898904085845726e-05,
"loss": 0.8037,
"step": 5820
},
{
"epoch": 0.6919881305637983,
"grad_norm": 0.778400719165802,
"learning_rate": 3.894835120461584e-05,
"loss": 0.8407,
"step": 5830
},
{
"epoch": 0.6931750741839763,
"grad_norm": 0.7665027379989624,
"learning_rate": 3.8907607829249896e-05,
"loss": 0.8662,
"step": 5840
},
{
"epoch": 0.6943620178041543,
"grad_norm": 0.7500550150871277,
"learning_rate": 3.8866810889280944e-05,
"loss": 0.7828,
"step": 5850
},
{
"epoch": 0.6955489614243323,
"grad_norm": 1.1609930992126465,
"learning_rate": 3.882596054183678e-05,
"loss": 0.8885,
"step": 5860
},
{
"epoch": 0.6967359050445104,
"grad_norm": 0.8244234323501587,
"learning_rate": 3.8785056944250905e-05,
"loss": 0.8757,
"step": 5870
},
{
"epoch": 0.6979228486646885,
"grad_norm": 0.9945170879364014,
"learning_rate": 3.874410025406191e-05,
"loss": 0.8839,
"step": 5880
},
{
"epoch": 0.6991097922848665,
"grad_norm": 0.9832311868667603,
"learning_rate": 3.8703090629012885e-05,
"loss": 0.906,
"step": 5890
},
{
"epoch": 0.7002967359050445,
"grad_norm": 0.8854712843894958,
"learning_rate": 3.866202822705077e-05,
"loss": 0.8103,
"step": 5900
},
{
"epoch": 0.7014836795252225,
"grad_norm": 0.8822460770606995,
"learning_rate": 3.862091320632579e-05,
"loss": 0.8785,
"step": 5910
},
{
"epoch": 0.7026706231454006,
"grad_norm": 0.9770189523696899,
"learning_rate": 3.857974572519083e-05,
"loss": 0.8647,
"step": 5920
},
{
"epoch": 0.7038575667655786,
"grad_norm": 1.002086877822876,
"learning_rate": 3.853852594220082e-05,
"loss": 0.8382,
"step": 5930
},
{
"epoch": 0.7050445103857567,
"grad_norm": 1.0117483139038086,
"learning_rate": 3.849725401611212e-05,
"loss": 0.8456,
"step": 5940
},
{
"epoch": 0.7062314540059347,
"grad_norm": 0.8960864543914795,
"learning_rate": 3.845593010588193e-05,
"loss": 0.8474,
"step": 5950
},
{
"epoch": 0.7074183976261128,
"grad_norm": 0.9335463047027588,
"learning_rate": 3.8414554370667666e-05,
"loss": 0.7904,
"step": 5960
},
{
"epoch": 0.7086053412462908,
"grad_norm": 0.8023741841316223,
"learning_rate": 3.837312696982633e-05,
"loss": 0.9063,
"step": 5970
},
{
"epoch": 0.7097922848664688,
"grad_norm": 0.9262803792953491,
"learning_rate": 3.833164806291393e-05,
"loss": 0.8648,
"step": 5980
},
{
"epoch": 0.7109792284866469,
"grad_norm": 0.8695958852767944,
"learning_rate": 3.829011780968483e-05,
"loss": 0.7609,
"step": 5990
},
{
"epoch": 0.712166172106825,
"grad_norm": 0.7884494662284851,
"learning_rate": 3.824853637009118e-05,
"loss": 0.8551,
"step": 6000
},
{
"epoch": 0.712166172106825,
"eval_loss": 0.7863911390304565,
"eval_runtime": 771.9015,
"eval_samples_per_second": 2.298,
"eval_steps_per_second": 0.575,
"step": 6000
},
{
"epoch": 0.713353115727003,
"grad_norm": 0.6585323810577393,
"learning_rate": 3.820690390428224e-05,
"loss": 0.8449,
"step": 6010
},
{
"epoch": 0.714540059347181,
"grad_norm": 1.00165855884552,
"learning_rate": 3.816522057260382e-05,
"loss": 0.8509,
"step": 6020
},
{
"epoch": 0.715727002967359,
"grad_norm": 1.0901848077774048,
"learning_rate": 3.812348653559762e-05,
"loss": 0.9183,
"step": 6030
},
{
"epoch": 0.716913946587537,
"grad_norm": 1.2764759063720703,
"learning_rate": 3.808170195400064e-05,
"loss": 0.8629,
"step": 6040
},
{
"epoch": 0.7181008902077152,
"grad_norm": 1.0480196475982666,
"learning_rate": 3.803986698874454e-05,
"loss": 0.8116,
"step": 6050
},
{
"epoch": 0.7192878338278932,
"grad_norm": 0.7966551184654236,
"learning_rate": 3.7997981800955066e-05,
"loss": 0.7801,
"step": 6060
},
{
"epoch": 0.7204747774480712,
"grad_norm": 1.2469865083694458,
"learning_rate": 3.795604655195133e-05,
"loss": 0.8411,
"step": 6070
},
{
"epoch": 0.7216617210682492,
"grad_norm": 0.8085722923278809,
"learning_rate": 3.7914061403245305e-05,
"loss": 0.8729,
"step": 6080
},
{
"epoch": 0.7228486646884273,
"grad_norm": 0.9539546370506287,
"learning_rate": 3.7872026516541146e-05,
"loss": 0.8266,
"step": 6090
},
{
"epoch": 0.7240356083086054,
"grad_norm": 0.8947449922561646,
"learning_rate": 3.782994205373455e-05,
"loss": 0.8971,
"step": 6100
},
{
"epoch": 0.7252225519287834,
"grad_norm": 0.654983401298523,
"learning_rate": 3.778780817691217e-05,
"loss": 0.8984,
"step": 6110
},
{
"epoch": 0.7264094955489614,
"grad_norm": 1.3490341901779175,
"learning_rate": 3.7745625048350967e-05,
"loss": 0.9364,
"step": 6120
},
{
"epoch": 0.7275964391691395,
"grad_norm": 0.7951292395591736,
"learning_rate": 3.7703392830517594e-05,
"loss": 0.8804,
"step": 6130
},
{
"epoch": 0.7287833827893175,
"grad_norm": 0.8905290961265564,
"learning_rate": 3.766111168606778e-05,
"loss": 0.7922,
"step": 6140
},
{
"epoch": 0.7299703264094956,
"grad_norm": 0.976385235786438,
"learning_rate": 3.761878177784568e-05,
"loss": 0.7535,
"step": 6150
},
{
"epoch": 0.7311572700296736,
"grad_norm": 0.865348219871521,
"learning_rate": 3.757640326888327e-05,
"loss": 0.7965,
"step": 6160
},
{
"epoch": 0.7323442136498516,
"grad_norm": 1.9425514936447144,
"learning_rate": 3.7533976322399704e-05,
"loss": 0.9106,
"step": 6170
},
{
"epoch": 0.7335311572700297,
"grad_norm": 0.8154645562171936,
"learning_rate": 3.749150110180069e-05,
"loss": 0.8822,
"step": 6180
},
{
"epoch": 0.7347181008902077,
"grad_norm": 1.1397662162780762,
"learning_rate": 3.744897777067787e-05,
"loss": 0.9469,
"step": 6190
},
{
"epoch": 0.7359050445103857,
"grad_norm": 0.8680721521377563,
"learning_rate": 3.740640649280817e-05,
"loss": 0.8945,
"step": 6200
},
{
"epoch": 0.7370919881305638,
"grad_norm": 0.7108725309371948,
"learning_rate": 3.7363787432153195e-05,
"loss": 0.788,
"step": 6210
},
{
"epoch": 0.7382789317507419,
"grad_norm": 0.7222614884376526,
"learning_rate": 3.7321120752858576e-05,
"loss": 0.8395,
"step": 6220
},
{
"epoch": 0.7394658753709199,
"grad_norm": 0.888178288936615,
"learning_rate": 3.727840661925334e-05,
"loss": 0.7956,
"step": 6230
},
{
"epoch": 0.7406528189910979,
"grad_norm": 0.8803139328956604,
"learning_rate": 3.72356451958493e-05,
"loss": 0.865,
"step": 6240
},
{
"epoch": 0.7418397626112759,
"grad_norm": 1.3589036464691162,
"learning_rate": 3.719283664734037e-05,
"loss": 0.8589,
"step": 6250
},
{
"epoch": 0.7430267062314541,
"grad_norm": 0.8697090148925781,
"learning_rate": 3.7149981138602e-05,
"loss": 0.8484,
"step": 6260
},
{
"epoch": 0.7442136498516321,
"grad_norm": 0.9363373517990112,
"learning_rate": 3.710707883469049e-05,
"loss": 0.8558,
"step": 6270
},
{
"epoch": 0.7454005934718101,
"grad_norm": 0.9355902075767517,
"learning_rate": 3.706412990084237e-05,
"loss": 0.7856,
"step": 6280
},
{
"epoch": 0.7465875370919881,
"grad_norm": 0.9179161190986633,
"learning_rate": 3.7021134502473765e-05,
"loss": 0.8404,
"step": 6290
},
{
"epoch": 0.7477744807121661,
"grad_norm": 1.0483746528625488,
"learning_rate": 3.6978092805179764e-05,
"loss": 0.9132,
"step": 6300
},
{
"epoch": 0.7489614243323442,
"grad_norm": 0.6334797739982605,
"learning_rate": 3.693500497473376e-05,
"loss": 0.8527,
"step": 6310
},
{
"epoch": 0.7501483679525223,
"grad_norm": 0.9985843300819397,
"learning_rate": 3.6891871177086826e-05,
"loss": 0.869,
"step": 6320
},
{
"epoch": 0.7513353115727003,
"grad_norm": 0.7376993298530579,
"learning_rate": 3.684869157836709e-05,
"loss": 0.8631,
"step": 6330
},
{
"epoch": 0.7525222551928783,
"grad_norm": 1.340173363685608,
"learning_rate": 3.6805466344879065e-05,
"loss": 0.9274,
"step": 6340
},
{
"epoch": 0.7537091988130564,
"grad_norm": 0.8143832087516785,
"learning_rate": 3.676219564310305e-05,
"loss": 0.8822,
"step": 6350
},
{
"epoch": 0.7548961424332344,
"grad_norm": 0.7084901928901672,
"learning_rate": 3.671887963969443e-05,
"loss": 0.7689,
"step": 6360
},
{
"epoch": 0.7560830860534125,
"grad_norm": 0.8772728443145752,
"learning_rate": 3.6675518501483086e-05,
"loss": 0.8648,
"step": 6370
},
{
"epoch": 0.7572700296735905,
"grad_norm": 0.8810312747955322,
"learning_rate": 3.663211239547274e-05,
"loss": 0.864,
"step": 6380
},
{
"epoch": 0.7584569732937686,
"grad_norm": 0.7641283869743347,
"learning_rate": 3.658866148884029e-05,
"loss": 0.8561,
"step": 6390
},
{
"epoch": 0.7596439169139466,
"grad_norm": 0.7614161372184753,
"learning_rate": 3.654516594893519e-05,
"loss": 0.8465,
"step": 6400
},
{
"epoch": 0.7608308605341246,
"grad_norm": 0.8121643662452698,
"learning_rate": 3.6501625943278805e-05,
"loss": 0.8615,
"step": 6410
},
{
"epoch": 0.7620178041543026,
"grad_norm": 0.7303301095962524,
"learning_rate": 3.645804163956374e-05,
"loss": 0.8287,
"step": 6420
},
{
"epoch": 0.7632047477744807,
"grad_norm": 0.8093650937080383,
"learning_rate": 3.641441320565322e-05,
"loss": 0.7774,
"step": 6430
},
{
"epoch": 0.7643916913946588,
"grad_norm": 0.9081480503082275,
"learning_rate": 3.637074080958044e-05,
"loss": 0.8813,
"step": 6440
},
{
"epoch": 0.7655786350148368,
"grad_norm": 0.9593875408172607,
"learning_rate": 3.632702461954792e-05,
"loss": 0.8371,
"step": 6450
},
{
"epoch": 0.7667655786350148,
"grad_norm": 0.862206757068634,
"learning_rate": 3.628326480392683e-05,
"loss": 0.8444,
"step": 6460
},
{
"epoch": 0.7679525222551928,
"grad_norm": 0.7164747714996338,
"learning_rate": 3.623946153125638e-05,
"loss": 0.9414,
"step": 6470
},
{
"epoch": 0.769139465875371,
"grad_norm": 0.964017391204834,
"learning_rate": 3.619561497024315e-05,
"loss": 0.8678,
"step": 6480
},
{
"epoch": 0.770326409495549,
"grad_norm": 0.7740942239761353,
"learning_rate": 3.615172528976043e-05,
"loss": 0.8334,
"step": 6490
},
{
"epoch": 0.771513353115727,
"grad_norm": 0.9031283855438232,
"learning_rate": 3.6107792658847595e-05,
"loss": 0.8742,
"step": 6500
},
{
"epoch": 0.771513353115727,
"eval_loss": 0.7851033806800842,
"eval_runtime": 773.4625,
"eval_samples_per_second": 2.294,
"eval_steps_per_second": 0.574,
"step": 6500
},
{
"epoch": 0.772700296735905,
"grad_norm": 0.9258557558059692,
"learning_rate": 3.6063817246709444e-05,
"loss": 0.8296,
"step": 6510
},
{
"epoch": 0.7738872403560831,
"grad_norm": 0.8308103084564209,
"learning_rate": 3.6019799222715544e-05,
"loss": 0.8445,
"step": 6520
},
{
"epoch": 0.7750741839762612,
"grad_norm": 0.8182960748672485,
"learning_rate": 3.5975738756399574e-05,
"loss": 0.8563,
"step": 6530
},
{
"epoch": 0.7762611275964392,
"grad_norm": 0.9615590572357178,
"learning_rate": 3.593163601745869e-05,
"loss": 0.9132,
"step": 6540
},
{
"epoch": 0.7774480712166172,
"grad_norm": 1.027933955192566,
"learning_rate": 3.588749117575284e-05,
"loss": 0.8199,
"step": 6550
},
{
"epoch": 0.7786350148367952,
"grad_norm": 0.9510583877563477,
"learning_rate": 3.584330440130415e-05,
"loss": 0.8841,
"step": 6560
},
{
"epoch": 0.7798219584569733,
"grad_norm": 1.3776129484176636,
"learning_rate": 3.5799075864296244e-05,
"loss": 0.8644,
"step": 6570
},
{
"epoch": 0.7810089020771513,
"grad_norm": 0.9228107333183289,
"learning_rate": 3.5754805735073575e-05,
"loss": 0.8239,
"step": 6580
},
{
"epoch": 0.7821958456973294,
"grad_norm": 0.9446859359741211,
"learning_rate": 3.571049418414081e-05,
"loss": 0.8789,
"step": 6590
},
{
"epoch": 0.7833827893175074,
"grad_norm": 0.8383123278617859,
"learning_rate": 3.566614138216212e-05,
"loss": 0.8505,
"step": 6600
},
{
"epoch": 0.7845697329376855,
"grad_norm": 0.9476184248924255,
"learning_rate": 3.562174749996059e-05,
"loss": 0.8603,
"step": 6610
},
{
"epoch": 0.7857566765578635,
"grad_norm": 0.8598759770393372,
"learning_rate": 3.557731270851751e-05,
"loss": 0.8384,
"step": 6620
},
{
"epoch": 0.7869436201780415,
"grad_norm": 0.7761598825454712,
"learning_rate": 3.553283717897171e-05,
"loss": 0.8635,
"step": 6630
},
{
"epoch": 0.7881305637982196,
"grad_norm": 0.7737132906913757,
"learning_rate": 3.548832108261896e-05,
"loss": 0.8172,
"step": 6640
},
{
"epoch": 0.7893175074183977,
"grad_norm": 0.8764254450798035,
"learning_rate": 3.5443764590911234e-05,
"loss": 0.8453,
"step": 6650
},
{
"epoch": 0.7905044510385757,
"grad_norm": 0.9228744506835938,
"learning_rate": 3.5399167875456094e-05,
"loss": 0.8332,
"step": 6660
},
{
"epoch": 0.7916913946587537,
"grad_norm": 0.9316686987876892,
"learning_rate": 3.5354531108016054e-05,
"loss": 0.8833,
"step": 6670
},
{
"epoch": 0.7928783382789317,
"grad_norm": 0.9275509119033813,
"learning_rate": 3.5309854460507845e-05,
"loss": 0.8096,
"step": 6680
},
{
"epoch": 0.7940652818991097,
"grad_norm": 0.7147977352142334,
"learning_rate": 3.526513810500182e-05,
"loss": 0.8531,
"step": 6690
},
{
"epoch": 0.7952522255192879,
"grad_norm": 0.8468539118766785,
"learning_rate": 3.5220382213721256e-05,
"loss": 0.8305,
"step": 6700
},
{
"epoch": 0.7964391691394659,
"grad_norm": 0.9143972992897034,
"learning_rate": 3.517558695904171e-05,
"loss": 0.8462,
"step": 6710
},
{
"epoch": 0.7976261127596439,
"grad_norm": 0.8958293199539185,
"learning_rate": 3.513075251349033e-05,
"loss": 0.8374,
"step": 6720
},
{
"epoch": 0.7988130563798219,
"grad_norm": 0.9110985398292542,
"learning_rate": 3.508587904974522e-05,
"loss": 0.8081,
"step": 6730
},
{
"epoch": 0.8,
"grad_norm": 1.0743153095245361,
"learning_rate": 3.5040966740634746e-05,
"loss": 0.8702,
"step": 6740
},
{
"epoch": 0.8011869436201781,
"grad_norm": 0.9236382842063904,
"learning_rate": 3.499601575913691e-05,
"loss": 0.8654,
"step": 6750
},
{
"epoch": 0.8023738872403561,
"grad_norm": 0.7747888565063477,
"learning_rate": 3.495102627837863e-05,
"loss": 0.8405,
"step": 6760
},
{
"epoch": 0.8035608308605341,
"grad_norm": 1.18403959274292,
"learning_rate": 3.490599847163512e-05,
"loss": 0.9657,
"step": 6770
},
{
"epoch": 0.8047477744807122,
"grad_norm": 0.8413063287734985,
"learning_rate": 3.486093251232921e-05,
"loss": 0.816,
"step": 6780
},
{
"epoch": 0.8059347181008902,
"grad_norm": 1.244125485420227,
"learning_rate": 3.4815828574030674e-05,
"loss": 0.8854,
"step": 6790
},
{
"epoch": 0.8071216617210683,
"grad_norm": 0.6361657977104187,
"learning_rate": 3.477068683045552e-05,
"loss": 0.8278,
"step": 6800
},
{
"epoch": 0.8083086053412463,
"grad_norm": 0.6628708243370056,
"learning_rate": 3.472550745546542e-05,
"loss": 0.8873,
"step": 6810
},
{
"epoch": 0.8094955489614243,
"grad_norm": 1.2260278463363647,
"learning_rate": 3.468029062306694e-05,
"loss": 0.8603,
"step": 6820
},
{
"epoch": 0.8106824925816024,
"grad_norm": 0.9426025748252869,
"learning_rate": 3.4635036507410935e-05,
"loss": 0.8679,
"step": 6830
},
{
"epoch": 0.8118694362017804,
"grad_norm": 0.6960638761520386,
"learning_rate": 3.4589745282791844e-05,
"loss": 0.7996,
"step": 6840
},
{
"epoch": 0.8130563798219584,
"grad_norm": 0.8688523173332214,
"learning_rate": 3.4544417123647024e-05,
"loss": 0.8351,
"step": 6850
},
{
"epoch": 0.8142433234421365,
"grad_norm": 0.9678128957748413,
"learning_rate": 3.4499052204556104e-05,
"loss": 0.9089,
"step": 6860
},
{
"epoch": 0.8154302670623146,
"grad_norm": 0.9642847180366516,
"learning_rate": 3.4453650700240276e-05,
"loss": 0.8045,
"step": 6870
},
{
"epoch": 0.8166172106824926,
"grad_norm": 0.7829703688621521,
"learning_rate": 3.4408212785561645e-05,
"loss": 0.8584,
"step": 6880
},
{
"epoch": 0.8178041543026706,
"grad_norm": 1.0218199491500854,
"learning_rate": 3.436273863552254e-05,
"loss": 0.8441,
"step": 6890
},
{
"epoch": 0.8189910979228486,
"grad_norm": 0.9257336854934692,
"learning_rate": 3.431722842526487e-05,
"loss": 0.8729,
"step": 6900
},
{
"epoch": 0.8201780415430268,
"grad_norm": 0.9871886372566223,
"learning_rate": 3.427168233006941e-05,
"loss": 0.8093,
"step": 6910
},
{
"epoch": 0.8213649851632048,
"grad_norm": 1.0205531120300293,
"learning_rate": 3.422610052535514e-05,
"loss": 0.9016,
"step": 6920
},
{
"epoch": 0.8225519287833828,
"grad_norm": 0.848477303981781,
"learning_rate": 3.418048318667858e-05,
"loss": 0.8633,
"step": 6930
},
{
"epoch": 0.8237388724035608,
"grad_norm": 1.0038056373596191,
"learning_rate": 3.4134830489733124e-05,
"loss": 0.9181,
"step": 6940
},
{
"epoch": 0.8249258160237388,
"grad_norm": 0.6654286980628967,
"learning_rate": 3.408914261034834e-05,
"loss": 0.819,
"step": 6950
},
{
"epoch": 0.826112759643917,
"grad_norm": 0.9518380165100098,
"learning_rate": 3.404341972448928e-05,
"loss": 0.8409,
"step": 6960
},
{
"epoch": 0.827299703264095,
"grad_norm": 0.8517202734947205,
"learning_rate": 3.399766200825583e-05,
"loss": 0.8002,
"step": 6970
},
{
"epoch": 0.828486646884273,
"grad_norm": 0.872999906539917,
"learning_rate": 3.3951869637882045e-05,
"loss": 0.8235,
"step": 6980
},
{
"epoch": 0.829673590504451,
"grad_norm": 0.9970819354057312,
"learning_rate": 3.390604278973543e-05,
"loss": 0.8936,
"step": 6990
},
{
"epoch": 0.8308605341246291,
"grad_norm": 0.9367091059684753,
"learning_rate": 3.386018164031627e-05,
"loss": 0.8477,
"step": 7000
},
{
"epoch": 0.8308605341246291,
"eval_loss": 0.7821624875068665,
"eval_runtime": 773.75,
"eval_samples_per_second": 2.293,
"eval_steps_per_second": 0.574,
"step": 7000
},
{
"epoch": 0.8320474777448071,
"grad_norm": 1.0959240198135376,
"learning_rate": 3.381428636625698e-05,
"loss": 0.8012,
"step": 7010
},
{
"epoch": 0.8332344213649852,
"grad_norm": 0.7259630560874939,
"learning_rate": 3.3768357144321406e-05,
"loss": 0.823,
"step": 7020
},
{
"epoch": 0.8344213649851632,
"grad_norm": 0.8429070711135864,
"learning_rate": 3.372239415140413e-05,
"loss": 0.9024,
"step": 7030
},
{
"epoch": 0.8356083086053413,
"grad_norm": 0.859829306602478,
"learning_rate": 3.367639756452981e-05,
"loss": 0.8471,
"step": 7040
},
{
"epoch": 0.8367952522255193,
"grad_norm": 0.8430136442184448,
"learning_rate": 3.363036756085247e-05,
"loss": 0.7838,
"step": 7050
},
{
"epoch": 0.8379821958456973,
"grad_norm": 0.7361006140708923,
"learning_rate": 3.3584304317654866e-05,
"loss": 0.8315,
"step": 7060
},
{
"epoch": 0.8391691394658753,
"grad_norm": 0.9041669368743896,
"learning_rate": 3.3538208012347774e-05,
"loss": 0.8307,
"step": 7070
},
{
"epoch": 0.8403560830860534,
"grad_norm": 0.8590983152389526,
"learning_rate": 3.349207882246927e-05,
"loss": 0.8197,
"step": 7080
},
{
"epoch": 0.8415430267062315,
"grad_norm": 0.8333597779273987,
"learning_rate": 3.3445916925684125e-05,
"loss": 0.8802,
"step": 7090
},
{
"epoch": 0.8427299703264095,
"grad_norm": 0.8488454818725586,
"learning_rate": 3.339972249978306e-05,
"loss": 0.8412,
"step": 7100
},
{
"epoch": 0.8439169139465875,
"grad_norm": 0.990714430809021,
"learning_rate": 3.335349572268209e-05,
"loss": 0.9064,
"step": 7110
},
{
"epoch": 0.8451038575667655,
"grad_norm": 0.9207063317298889,
"learning_rate": 3.3307236772421823e-05,
"loss": 0.8309,
"step": 7120
},
{
"epoch": 0.8462908011869437,
"grad_norm": 0.9114146828651428,
"learning_rate": 3.326094582716678e-05,
"loss": 0.8841,
"step": 7130
},
{
"epoch": 0.8474777448071217,
"grad_norm": 1.2737470865249634,
"learning_rate": 3.32146230652047e-05,
"loss": 0.8373,
"step": 7140
},
{
"epoch": 0.8486646884272997,
"grad_norm": 0.9316775798797607,
"learning_rate": 3.3168268664945886e-05,
"loss": 0.8177,
"step": 7150
},
{
"epoch": 0.8498516320474777,
"grad_norm": 0.9175634980201721,
"learning_rate": 3.3121882804922484e-05,
"loss": 0.8933,
"step": 7160
},
{
"epoch": 0.8510385756676558,
"grad_norm": 0.9439007639884949,
"learning_rate": 3.3075465663787794e-05,
"loss": 0.7867,
"step": 7170
},
{
"epoch": 0.8522255192878339,
"grad_norm": 0.893551230430603,
"learning_rate": 3.30290174203156e-05,
"loss": 0.9118,
"step": 7180
},
{
"epoch": 0.8534124629080119,
"grad_norm": 1.0411717891693115,
"learning_rate": 3.29825382533995e-05,
"loss": 0.8291,
"step": 7190
},
{
"epoch": 0.8545994065281899,
"grad_norm": 0.8265506029129028,
"learning_rate": 3.293602834205216e-05,
"loss": 0.8228,
"step": 7200
},
{
"epoch": 0.8557863501483679,
"grad_norm": 0.9281073808670044,
"learning_rate": 3.2889487865404674e-05,
"loss": 0.8694,
"step": 7210
},
{
"epoch": 0.856973293768546,
"grad_norm": 0.7977082133293152,
"learning_rate": 3.284291700270584e-05,
"loss": 0.8242,
"step": 7220
},
{
"epoch": 0.858160237388724,
"grad_norm": 0.8856098055839539,
"learning_rate": 3.279631593332151e-05,
"loss": 0.8507,
"step": 7230
},
{
"epoch": 0.8593471810089021,
"grad_norm": 1.0277818441390991,
"learning_rate": 3.2749684836733866e-05,
"loss": 0.8775,
"step": 7240
},
{
"epoch": 0.8605341246290801,
"grad_norm": 1.0487391948699951,
"learning_rate": 3.270302389254074e-05,
"loss": 0.8404,
"step": 7250
},
{
"epoch": 0.8617210682492582,
"grad_norm": 0.8497503995895386,
"learning_rate": 3.26563332804549e-05,
"loss": 0.9356,
"step": 7260
},
{
"epoch": 0.8629080118694362,
"grad_norm": 0.7366178631782532,
"learning_rate": 3.2609613180303413e-05,
"loss": 0.8132,
"step": 7270
},
{
"epoch": 0.8640949554896142,
"grad_norm": 0.8012394905090332,
"learning_rate": 3.25628637720269e-05,
"loss": 0.7763,
"step": 7280
},
{
"epoch": 0.8652818991097923,
"grad_norm": 1.029145359992981,
"learning_rate": 3.251608523567886e-05,
"loss": 0.8116,
"step": 7290
},
{
"epoch": 0.8664688427299704,
"grad_norm": 0.9467004537582397,
"learning_rate": 3.246927775142498e-05,
"loss": 0.8444,
"step": 7300
},
{
"epoch": 0.8676557863501484,
"grad_norm": 0.8625933527946472,
"learning_rate": 3.242244149954243e-05,
"loss": 0.9053,
"step": 7310
},
{
"epoch": 0.8688427299703264,
"grad_norm": 1.0368592739105225,
"learning_rate": 3.237557666041922e-05,
"loss": 0.9067,
"step": 7320
},
{
"epoch": 0.8700296735905044,
"grad_norm": 0.760718822479248,
"learning_rate": 3.232868341455339e-05,
"loss": 0.8853,
"step": 7330
},
{
"epoch": 0.8712166172106824,
"grad_norm": 0.7736045718193054,
"learning_rate": 3.2281761942552446e-05,
"loss": 0.9119,
"step": 7340
},
{
"epoch": 0.8724035608308606,
"grad_norm": 0.9275757670402527,
"learning_rate": 3.223481242513259e-05,
"loss": 0.7932,
"step": 7350
},
{
"epoch": 0.8735905044510386,
"grad_norm": 0.8196761012077332,
"learning_rate": 3.218783504311803e-05,
"loss": 0.8893,
"step": 7360
},
{
"epoch": 0.8747774480712166,
"grad_norm": 0.8558308482170105,
"learning_rate": 3.21408299774403e-05,
"loss": 0.8948,
"step": 7370
},
{
"epoch": 0.8759643916913946,
"grad_norm": 0.7309697270393372,
"learning_rate": 3.209379740913756e-05,
"loss": 0.8142,
"step": 7380
},
{
"epoch": 0.8771513353115727,
"grad_norm": 0.7917837500572205,
"learning_rate": 3.204673751935389e-05,
"loss": 0.8557,
"step": 7390
},
{
"epoch": 0.8783382789317508,
"grad_norm": 1.0087717771530151,
"learning_rate": 3.199965048933859e-05,
"loss": 0.9215,
"step": 7400
},
{
"epoch": 0.8795252225519288,
"grad_norm": 0.9754852056503296,
"learning_rate": 3.195253650044552e-05,
"loss": 0.8229,
"step": 7410
},
{
"epoch": 0.8807121661721068,
"grad_norm": 0.8485432863235474,
"learning_rate": 3.1905395734132326e-05,
"loss": 0.8653,
"step": 7420
},
{
"epoch": 0.8818991097922849,
"grad_norm": 0.8773938417434692,
"learning_rate": 3.185822837195983e-05,
"loss": 0.9774,
"step": 7430
},
{
"epoch": 0.8830860534124629,
"grad_norm": 0.7707668542861938,
"learning_rate": 3.181103459559126e-05,
"loss": 0.8606,
"step": 7440
},
{
"epoch": 0.884272997032641,
"grad_norm": 0.7578542828559875,
"learning_rate": 3.176381458679158e-05,
"loss": 0.8229,
"step": 7450
},
{
"epoch": 0.885459940652819,
"grad_norm": 0.9707187414169312,
"learning_rate": 3.17165685274268e-05,
"loss": 0.8259,
"step": 7460
},
{
"epoch": 0.886646884272997,
"grad_norm": 1.3090546131134033,
"learning_rate": 3.166929659946325e-05,
"loss": 0.8843,
"step": 7470
},
{
"epoch": 0.8878338278931751,
"grad_norm": 0.7714403867721558,
"learning_rate": 3.1621998984966894e-05,
"loss": 0.8204,
"step": 7480
},
{
"epoch": 0.8890207715133531,
"grad_norm": 1.0389307737350464,
"learning_rate": 3.157467586610262e-05,
"loss": 0.8493,
"step": 7490
},
{
"epoch": 0.8902077151335311,
"grad_norm": 0.8754733204841614,
"learning_rate": 3.152732742513357e-05,
"loss": 0.8492,
"step": 7500
},
{
"epoch": 0.8902077151335311,
"eval_loss": 0.7797255516052246,
"eval_runtime": 769.2673,
"eval_samples_per_second": 2.306,
"eval_steps_per_second": 0.577,
"step": 7500
},
{
"epoch": 0.8913946587537092,
"grad_norm": 0.5790079832077026,
"learning_rate": 3.147995384442039e-05,
"loss": 0.7781,
"step": 7510
},
{
"epoch": 0.8925816023738873,
"grad_norm": 0.8539468050003052,
"learning_rate": 3.143255530642056e-05,
"loss": 0.8033,
"step": 7520
},
{
"epoch": 0.8937685459940653,
"grad_norm": 0.7788002490997314,
"learning_rate": 3.138513199368768e-05,
"loss": 0.8488,
"step": 7530
},
{
"epoch": 0.8949554896142433,
"grad_norm": 0.6952248215675354,
"learning_rate": 3.133768408887076e-05,
"loss": 0.8358,
"step": 7540
},
{
"epoch": 0.8961424332344213,
"grad_norm": 0.7432988882064819,
"learning_rate": 3.129021177471354e-05,
"loss": 0.8785,
"step": 7550
},
{
"epoch": 0.8973293768545995,
"grad_norm": 0.8068860173225403,
"learning_rate": 3.124271523405377e-05,
"loss": 0.8018,
"step": 7560
},
{
"epoch": 0.8985163204747775,
"grad_norm": 0.8224064707756042,
"learning_rate": 3.11951946498225e-05,
"loss": 0.7825,
"step": 7570
},
{
"epoch": 0.8997032640949555,
"grad_norm": 1.1241661310195923,
"learning_rate": 3.114765020504339e-05,
"loss": 0.9344,
"step": 7580
},
{
"epoch": 0.9008902077151335,
"grad_norm": 1.0378000736236572,
"learning_rate": 3.110008208283197e-05,
"loss": 0.8613,
"step": 7590
},
{
"epoch": 0.9020771513353115,
"grad_norm": 0.9253892302513123,
"learning_rate": 3.1052490466395023e-05,
"loss": 0.8725,
"step": 7600
},
{
"epoch": 0.9032640949554896,
"grad_norm": 0.9954454898834229,
"learning_rate": 3.1004875539029755e-05,
"loss": 0.7652,
"step": 7610
},
{
"epoch": 0.9044510385756677,
"grad_norm": 0.9441783428192139,
"learning_rate": 3.0957237484123196e-05,
"loss": 0.9363,
"step": 7620
},
{
"epoch": 0.9056379821958457,
"grad_norm": 0.8821224570274353,
"learning_rate": 3.090957648515142e-05,
"loss": 0.9007,
"step": 7630
},
{
"epoch": 0.9068249258160237,
"grad_norm": 0.8489567041397095,
"learning_rate": 3.086189272567891e-05,
"loss": 0.9078,
"step": 7640
},
{
"epoch": 0.9080118694362018,
"grad_norm": 0.894049882888794,
"learning_rate": 3.0814186389357765e-05,
"loss": 0.8681,
"step": 7650
},
{
"epoch": 0.9091988130563798,
"grad_norm": 1.0218708515167236,
"learning_rate": 3.076645765992706e-05,
"loss": 0.8021,
"step": 7660
},
{
"epoch": 0.9103857566765579,
"grad_norm": 1.236533761024475,
"learning_rate": 3.0718706721212116e-05,
"loss": 0.8619,
"step": 7670
},
{
"epoch": 0.9115727002967359,
"grad_norm": 0.8600239157676697,
"learning_rate": 3.067093375712379e-05,
"loss": 0.7805,
"step": 7680
},
{
"epoch": 0.912759643916914,
"grad_norm": 0.8407261371612549,
"learning_rate": 3.062313895165777e-05,
"loss": 0.834,
"step": 7690
},
{
"epoch": 0.913946587537092,
"grad_norm": 1.0202083587646484,
"learning_rate": 3.057532248889386e-05,
"loss": 0.8123,
"step": 7700
},
{
"epoch": 0.91513353115727,
"grad_norm": 1.0518484115600586,
"learning_rate": 3.052748455299529e-05,
"loss": 0.7799,
"step": 7710
},
{
"epoch": 0.916320474777448,
"grad_norm": 0.6867684721946716,
"learning_rate": 3.047962532820797e-05,
"loss": 0.7433,
"step": 7720
},
{
"epoch": 0.9175074183976261,
"grad_norm": 0.8702245354652405,
"learning_rate": 3.0431744998859834e-05,
"loss": 0.9001,
"step": 7730
},
{
"epoch": 0.9186943620178042,
"grad_norm": 0.7238080501556396,
"learning_rate": 3.0383843749360063e-05,
"loss": 0.8513,
"step": 7740
},
{
"epoch": 0.9198813056379822,
"grad_norm": 0.8329240083694458,
"learning_rate": 3.0335921764198437e-05,
"loss": 0.8731,
"step": 7750
},
{
"epoch": 0.9210682492581602,
"grad_norm": 0.6718214750289917,
"learning_rate": 3.02879792279446e-05,
"loss": 0.796,
"step": 7760
},
{
"epoch": 0.9222551928783382,
"grad_norm": 0.8724163770675659,
"learning_rate": 3.0240016325247332e-05,
"loss": 0.812,
"step": 7770
},
{
"epoch": 0.9234421364985164,
"grad_norm": 0.820673942565918,
"learning_rate": 3.019203324083386e-05,
"loss": 0.8487,
"step": 7780
},
{
"epoch": 0.9246290801186944,
"grad_norm": 0.8288638591766357,
"learning_rate": 3.0144030159509145e-05,
"loss": 0.9421,
"step": 7790
},
{
"epoch": 0.9258160237388724,
"grad_norm": 0.8184947371482849,
"learning_rate": 3.0096007266155167e-05,
"loss": 0.8175,
"step": 7800
},
{
"epoch": 0.9270029673590504,
"grad_norm": 0.7271890044212341,
"learning_rate": 3.0047964745730188e-05,
"loss": 0.7866,
"step": 7810
},
{
"epoch": 0.9281899109792285,
"grad_norm": 0.8679256439208984,
"learning_rate": 2.9999902783268096e-05,
"loss": 0.9335,
"step": 7820
},
{
"epoch": 0.9293768545994066,
"grad_norm": 0.9273269772529602,
"learning_rate": 2.995182156387763e-05,
"loss": 0.8438,
"step": 7830
},
{
"epoch": 0.9305637982195846,
"grad_norm": 1.0963612794876099,
"learning_rate": 2.9903721272741726e-05,
"loss": 0.8234,
"step": 7840
},
{
"epoch": 0.9317507418397626,
"grad_norm": 0.8465256094932556,
"learning_rate": 2.9855602095116746e-05,
"loss": 0.9186,
"step": 7850
},
{
"epoch": 0.9329376854599406,
"grad_norm": 0.6978071331977844,
"learning_rate": 2.9807464216331815e-05,
"loss": 0.7885,
"step": 7860
},
{
"epoch": 0.9341246290801187,
"grad_norm": 0.7609380483627319,
"learning_rate": 2.975930782178807e-05,
"loss": 0.8534,
"step": 7870
},
{
"epoch": 0.9353115727002967,
"grad_norm": 1.0055066347122192,
"learning_rate": 2.9711133096957962e-05,
"loss": 0.886,
"step": 7880
},
{
"epoch": 0.9364985163204748,
"grad_norm": 0.7272253036499023,
"learning_rate": 2.9662940227384544e-05,
"loss": 0.8306,
"step": 7890
},
{
"epoch": 0.9376854599406528,
"grad_norm": 0.877644419670105,
"learning_rate": 2.961472939868075e-05,
"loss": 0.8807,
"step": 7900
},
{
"epoch": 0.9388724035608309,
"grad_norm": 0.8707640171051025,
"learning_rate": 2.95665007965287e-05,
"loss": 0.9264,
"step": 7910
},
{
"epoch": 0.9400593471810089,
"grad_norm": 0.9951992034912109,
"learning_rate": 2.9518254606678935e-05,
"loss": 0.8345,
"step": 7920
},
{
"epoch": 0.9412462908011869,
"grad_norm": 0.9738534092903137,
"learning_rate": 2.946999101494976e-05,
"loss": 0.8475,
"step": 7930
},
{
"epoch": 0.942433234421365,
"grad_norm": 0.8127856254577637,
"learning_rate": 2.94217102072265e-05,
"loss": 0.7886,
"step": 7940
},
{
"epoch": 0.9436201780415431,
"grad_norm": 0.8734619617462158,
"learning_rate": 2.9373412369460774e-05,
"loss": 0.8797,
"step": 7950
},
{
"epoch": 0.9448071216617211,
"grad_norm": 0.7642955780029297,
"learning_rate": 2.93250976876698e-05,
"loss": 0.782,
"step": 7960
},
{
"epoch": 0.9459940652818991,
"grad_norm": 1.211449146270752,
"learning_rate": 2.927676634793567e-05,
"loss": 0.8433,
"step": 7970
},
{
"epoch": 0.9471810089020771,
"grad_norm": 1.0548808574676514,
"learning_rate": 2.9228418536404633e-05,
"loss": 0.8425,
"step": 7980
},
{
"epoch": 0.9483679525222551,
"grad_norm": 0.7786169052124023,
"learning_rate": 2.9180054439286377e-05,
"loss": 0.7831,
"step": 7990
},
{
"epoch": 0.9495548961424333,
"grad_norm": 0.7001704573631287,
"learning_rate": 2.9131674242853318e-05,
"loss": 0.8113,
"step": 8000
},
{
"epoch": 0.9495548961424333,
"eval_loss": 0.7776389718055725,
"eval_runtime": 768.6108,
"eval_samples_per_second": 2.308,
"eval_steps_per_second": 0.578,
"step": 8000
},
{
"epoch": 0.9507418397626113,
"grad_norm": 0.7324917316436768,
"learning_rate": 2.9083278133439883e-05,
"loss": 0.8663,
"step": 8010
},
{
"epoch": 0.9519287833827893,
"grad_norm": 1.0254888534545898,
"learning_rate": 2.903486629744176e-05,
"loss": 0.8569,
"step": 8020
},
{
"epoch": 0.9531157270029673,
"grad_norm": 1.0356833934783936,
"learning_rate": 2.8986438921315234e-05,
"loss": 0.8032,
"step": 8030
},
{
"epoch": 0.9543026706231454,
"grad_norm": 0.8793463706970215,
"learning_rate": 2.8937996191576432e-05,
"loss": 0.841,
"step": 8040
},
{
"epoch": 0.9554896142433235,
"grad_norm": 0.891936182975769,
"learning_rate": 2.888953829480062e-05,
"loss": 0.7662,
"step": 8050
},
{
"epoch": 0.9566765578635015,
"grad_norm": 1.5397156476974487,
"learning_rate": 2.884106541762148e-05,
"loss": 0.8108,
"step": 8060
},
{
"epoch": 0.9578635014836795,
"grad_norm": 0.9605924487113953,
"learning_rate": 2.8792577746730375e-05,
"loss": 0.8661,
"step": 8070
},
{
"epoch": 0.9590504451038576,
"grad_norm": 0.9698401093482971,
"learning_rate": 2.8744075468875664e-05,
"loss": 0.8318,
"step": 8080
},
{
"epoch": 0.9602373887240356,
"grad_norm": 0.7636862993240356,
"learning_rate": 2.8695558770861956e-05,
"loss": 0.8516,
"step": 8090
},
{
"epoch": 0.9614243323442137,
"grad_norm": 0.8955727815628052,
"learning_rate": 2.8647027839549407e-05,
"loss": 0.755,
"step": 8100
},
{
"epoch": 0.9626112759643917,
"grad_norm": 1.0949968099594116,
"learning_rate": 2.8598482861852976e-05,
"loss": 0.8672,
"step": 8110
},
{
"epoch": 0.9637982195845697,
"grad_norm": 0.9382548928260803,
"learning_rate": 2.8549924024741736e-05,
"loss": 0.811,
"step": 8120
},
{
"epoch": 0.9649851632047478,
"grad_norm": 0.6319847106933594,
"learning_rate": 2.8501351515238135e-05,
"loss": 0.826,
"step": 8130
},
{
"epoch": 0.9661721068249258,
"grad_norm": 0.8856794834136963,
"learning_rate": 2.8452765520417264e-05,
"loss": 0.8287,
"step": 8140
},
{
"epoch": 0.9673590504451038,
"grad_norm": 0.9224793314933777,
"learning_rate": 2.840416622740617e-05,
"loss": 0.834,
"step": 8150
},
{
"epoch": 0.9685459940652819,
"grad_norm": 0.8026999831199646,
"learning_rate": 2.8355553823383118e-05,
"loss": 0.9407,
"step": 8160
},
{
"epoch": 0.96973293768546,
"grad_norm": 0.9832804203033447,
"learning_rate": 2.8306928495576868e-05,
"loss": 0.8656,
"step": 8170
},
{
"epoch": 0.970919881305638,
"grad_norm": 0.9716234803199768,
"learning_rate": 2.8258290431265944e-05,
"loss": 0.9096,
"step": 8180
},
{
"epoch": 0.972106824925816,
"grad_norm": 0.8760083317756653,
"learning_rate": 2.820963981777794e-05,
"loss": 0.8766,
"step": 8190
},
{
"epoch": 0.973293768545994,
"grad_norm": 1.0046391487121582,
"learning_rate": 2.816097684248877e-05,
"loss": 0.895,
"step": 8200
},
{
"epoch": 0.9744807121661722,
"grad_norm": 0.6522348523139954,
"learning_rate": 2.8112301692821964e-05,
"loss": 0.7833,
"step": 8210
},
{
"epoch": 0.9756676557863502,
"grad_norm": 1.0381343364715576,
"learning_rate": 2.8063614556247948e-05,
"loss": 0.8459,
"step": 8220
},
{
"epoch": 0.9768545994065282,
"grad_norm": 0.6645428538322449,
"learning_rate": 2.8014915620283294e-05,
"loss": 0.9006,
"step": 8230
},
{
"epoch": 0.9780415430267062,
"grad_norm": 0.9998354315757751,
"learning_rate": 2.796620507249005e-05,
"loss": 0.7815,
"step": 8240
},
{
"epoch": 0.9792284866468842,
"grad_norm": 0.9861043095588684,
"learning_rate": 2.7917483100474955e-05,
"loss": 0.9115,
"step": 8250
},
{
"epoch": 0.9804154302670623,
"grad_norm": 1.03464937210083,
"learning_rate": 2.7868749891888767e-05,
"loss": 0.7882,
"step": 8260
},
{
"epoch": 0.9816023738872404,
"grad_norm": 0.8017896413803101,
"learning_rate": 2.7820005634425528e-05,
"loss": 0.8348,
"step": 8270
},
{
"epoch": 0.9827893175074184,
"grad_norm": 1.063607931137085,
"learning_rate": 2.7771250515821807e-05,
"loss": 0.9464,
"step": 8280
},
{
"epoch": 0.9839762611275964,
"grad_norm": 1.0467710494995117,
"learning_rate": 2.772248472385603e-05,
"loss": 0.8656,
"step": 8290
},
{
"epoch": 0.9851632047477745,
"grad_norm": 0.7622478604316711,
"learning_rate": 2.7673708446347712e-05,
"loss": 0.8404,
"step": 8300
},
{
"epoch": 0.9863501483679525,
"grad_norm": 0.7604171633720398,
"learning_rate": 2.7624921871156772e-05,
"loss": 0.8404,
"step": 8310
},
{
"epoch": 0.9875370919881306,
"grad_norm": 1.0093992948532104,
"learning_rate": 2.7576125186182772e-05,
"loss": 0.8871,
"step": 8320
},
{
"epoch": 0.9887240356083086,
"grad_norm": 0.5812623500823975,
"learning_rate": 2.752731857936423e-05,
"loss": 0.8223,
"step": 8330
},
{
"epoch": 0.9899109792284867,
"grad_norm": 0.5302821397781372,
"learning_rate": 2.7478502238677862e-05,
"loss": 0.8357,
"step": 8340
},
{
"epoch": 0.9910979228486647,
"grad_norm": 0.8963286876678467,
"learning_rate": 2.7429676352137885e-05,
"loss": 0.9103,
"step": 8350
},
{
"epoch": 0.9922848664688427,
"grad_norm": 0.8985591530799866,
"learning_rate": 2.738084110779526e-05,
"loss": 0.9101,
"step": 8360
},
{
"epoch": 0.9934718100890207,
"grad_norm": 1.0268195867538452,
"learning_rate": 2.733199669373701e-05,
"loss": 0.8353,
"step": 8370
},
{
"epoch": 0.9946587537091988,
"grad_norm": 0.897013783454895,
"learning_rate": 2.7283143298085474e-05,
"loss": 0.8526,
"step": 8380
},
{
"epoch": 0.9958456973293769,
"grad_norm": 1.0083746910095215,
"learning_rate": 2.723428110899757e-05,
"loss": 0.8349,
"step": 8390
},
{
"epoch": 0.9970326409495549,
"grad_norm": 0.9338353276252747,
"learning_rate": 2.7185410314664107e-05,
"loss": 0.8304,
"step": 8400
},
{
"epoch": 0.9982195845697329,
"grad_norm": 1.0432991981506348,
"learning_rate": 2.713653110330901e-05,
"loss": 0.8171,
"step": 8410
},
{
"epoch": 0.9994065281899109,
"grad_norm": 1.0495964288711548,
"learning_rate": 2.7087643663188633e-05,
"loss": 0.8842,
"step": 8420
},
{
"epoch": 1.000593471810089,
"grad_norm": 0.9517408609390259,
"learning_rate": 2.7038748182591034e-05,
"loss": 0.9287,
"step": 8430
},
{
"epoch": 1.001780415430267,
"grad_norm": 0.7760940790176392,
"learning_rate": 2.698984484983522e-05,
"loss": 0.783,
"step": 8440
},
{
"epoch": 1.002967359050445,
"grad_norm": 0.9729003310203552,
"learning_rate": 2.694093385327045e-05,
"loss": 0.8654,
"step": 8450
},
{
"epoch": 1.0041543026706232,
"grad_norm": 1.066876769065857,
"learning_rate": 2.6892015381275493e-05,
"loss": 0.8803,
"step": 8460
},
{
"epoch": 1.0053412462908011,
"grad_norm": 0.8508303165435791,
"learning_rate": 2.684308962225793e-05,
"loss": 0.8473,
"step": 8470
},
{
"epoch": 1.0065281899109793,
"grad_norm": 0.95997154712677,
"learning_rate": 2.6794156764653367e-05,
"loss": 0.8053,
"step": 8480
},
{
"epoch": 1.0077151335311574,
"grad_norm": 0.8591273427009583,
"learning_rate": 2.67452169969248e-05,
"loss": 0.8858,
"step": 8490
},
{
"epoch": 1.0089020771513353,
"grad_norm": 1.0032804012298584,
"learning_rate": 2.66962705075618e-05,
"loss": 0.7683,
"step": 8500
},
{
"epoch": 1.0089020771513353,
"eval_loss": 0.7766696810722351,
"eval_runtime": 769.9177,
"eval_samples_per_second": 2.304,
"eval_steps_per_second": 0.577,
"step": 8500
},
{
"epoch": 1.0100890207715134,
"grad_norm": 1.2956509590148926,
"learning_rate": 2.6647317485079843e-05,
"loss": 0.8622,
"step": 8510
},
{
"epoch": 1.0112759643916913,
"grad_norm": 0.9024373292922974,
"learning_rate": 2.659835811801956e-05,
"loss": 0.8628,
"step": 8520
},
{
"epoch": 1.0124629080118694,
"grad_norm": 1.0521224737167358,
"learning_rate": 2.6549392594946037e-05,
"loss": 0.8375,
"step": 8530
},
{
"epoch": 1.0136498516320476,
"grad_norm": 0.8524640202522278,
"learning_rate": 2.650042110444804e-05,
"loss": 0.8161,
"step": 8540
},
{
"epoch": 1.0148367952522255,
"grad_norm": 0.8786371350288391,
"learning_rate": 2.6451443835137342e-05,
"loss": 0.8241,
"step": 8550
},
{
"epoch": 1.0160237388724036,
"grad_norm": 0.993325412273407,
"learning_rate": 2.640246097564796e-05,
"loss": 0.827,
"step": 8560
},
{
"epoch": 1.0172106824925815,
"grad_norm": 0.9485884308815002,
"learning_rate": 2.635347271463544e-05,
"loss": 0.8,
"step": 8570
},
{
"epoch": 1.0183976261127596,
"grad_norm": 0.879743218421936,
"learning_rate": 2.6304479240776147e-05,
"loss": 0.7966,
"step": 8580
},
{
"epoch": 1.0195845697329378,
"grad_norm": 0.9597845673561096,
"learning_rate": 2.625548074276651e-05,
"loss": 0.9103,
"step": 8590
},
{
"epoch": 1.0207715133531157,
"grad_norm": 0.7718109488487244,
"learning_rate": 2.6206477409322306e-05,
"loss": 0.8311,
"step": 8600
},
{
"epoch": 1.0219584569732938,
"grad_norm": 0.8141760230064392,
"learning_rate": 2.615746942917795e-05,
"loss": 0.8162,
"step": 8610
},
{
"epoch": 1.0231454005934717,
"grad_norm": 1.0272105932235718,
"learning_rate": 2.610845699108574e-05,
"loss": 0.8109,
"step": 8620
},
{
"epoch": 1.0243323442136498,
"grad_norm": 1.010335922241211,
"learning_rate": 2.6059440283815146e-05,
"loss": 0.786,
"step": 8630
},
{
"epoch": 1.025519287833828,
"grad_norm": 0.9021561741828918,
"learning_rate": 2.6010419496152083e-05,
"loss": 0.8407,
"step": 8640
},
{
"epoch": 1.0267062314540059,
"grad_norm": 0.8104283809661865,
"learning_rate": 2.5961394816898184e-05,
"loss": 0.8011,
"step": 8650
},
{
"epoch": 1.027893175074184,
"grad_norm": 0.7947595715522766,
"learning_rate": 2.5912366434870066e-05,
"loss": 0.7901,
"step": 8660
},
{
"epoch": 1.0290801186943621,
"grad_norm": 0.8429394960403442,
"learning_rate": 2.586333453889861e-05,
"loss": 0.8268,
"step": 8670
},
{
"epoch": 1.03026706231454,
"grad_norm": 1.1205742359161377,
"learning_rate": 2.5814299317828234e-05,
"loss": 0.8265,
"step": 8680
},
{
"epoch": 1.0314540059347181,
"grad_norm": 0.9771138429641724,
"learning_rate": 2.576526096051615e-05,
"loss": 0.8365,
"step": 8690
},
{
"epoch": 1.032640949554896,
"grad_norm": 1.1407651901245117,
"learning_rate": 2.571621965583166e-05,
"loss": 0.8202,
"step": 8700
},
{
"epoch": 1.0338278931750742,
"grad_norm": 0.9523341059684753,
"learning_rate": 2.5667175592655413e-05,
"loss": 0.8774,
"step": 8710
},
{
"epoch": 1.0350148367952523,
"grad_norm": 1.2493401765823364,
"learning_rate": 2.5618128959878686e-05,
"loss": 0.8825,
"step": 8720
},
{
"epoch": 1.0362017804154302,
"grad_norm": 1.1929856538772583,
"learning_rate": 2.556907994640264e-05,
"loss": 0.8126,
"step": 8730
},
{
"epoch": 1.0373887240356083,
"grad_norm": 0.6640522480010986,
"learning_rate": 2.5520028741137636e-05,
"loss": 0.7949,
"step": 8740
},
{
"epoch": 1.0385756676557865,
"grad_norm": 0.9167020320892334,
"learning_rate": 2.547097553300245e-05,
"loss": 0.7853,
"step": 8750
},
{
"epoch": 1.0397626112759644,
"grad_norm": 1.055112361907959,
"learning_rate": 2.5421920510923568e-05,
"loss": 0.8087,
"step": 8760
},
{
"epoch": 1.0409495548961425,
"grad_norm": 0.9777671098709106,
"learning_rate": 2.5372863863834484e-05,
"loss": 0.7782,
"step": 8770
},
{
"epoch": 1.0421364985163204,
"grad_norm": 0.8514682054519653,
"learning_rate": 2.532380578067493e-05,
"loss": 0.8282,
"step": 8780
},
{
"epoch": 1.0433234421364985,
"grad_norm": 1.0609797239303589,
"learning_rate": 2.5274746450390186e-05,
"loss": 0.8202,
"step": 8790
},
{
"epoch": 1.0445103857566767,
"grad_norm": 0.8998866081237793,
"learning_rate": 2.5225686061930326e-05,
"loss": 0.8358,
"step": 8800
},
{
"epoch": 1.0456973293768546,
"grad_norm": 0.9781126379966736,
"learning_rate": 2.51766248042495e-05,
"loss": 0.8162,
"step": 8810
},
{
"epoch": 1.0468842729970327,
"grad_norm": 0.8800467252731323,
"learning_rate": 2.5127562866305206e-05,
"loss": 0.8072,
"step": 8820
},
{
"epoch": 1.0480712166172106,
"grad_norm": 1.0159127712249756,
"learning_rate": 2.5078500437057556e-05,
"loss": 0.7879,
"step": 8830
},
{
"epoch": 1.0492581602373887,
"grad_norm": 0.8794515132904053,
"learning_rate": 2.5029437705468577e-05,
"loss": 0.7577,
"step": 8840
},
{
"epoch": 1.0504451038575668,
"grad_norm": 1.0653094053268433,
"learning_rate": 2.4980374860501436e-05,
"loss": 0.8209,
"step": 8850
},
{
"epoch": 1.0516320474777447,
"grad_norm": 0.8193674683570862,
"learning_rate": 2.493131209111974e-05,
"loss": 0.796,
"step": 8860
},
{
"epoch": 1.0528189910979229,
"grad_norm": 0.8971123099327087,
"learning_rate": 2.4882249586286825e-05,
"loss": 0.8305,
"step": 8870
},
{
"epoch": 1.0540059347181008,
"grad_norm": 1.074666976928711,
"learning_rate": 2.4833187534964975e-05,
"loss": 0.8023,
"step": 8880
},
{
"epoch": 1.055192878338279,
"grad_norm": 0.8462083339691162,
"learning_rate": 2.4784126126114766e-05,
"loss": 0.6995,
"step": 8890
},
{
"epoch": 1.056379821958457,
"grad_norm": 0.9592199921607971,
"learning_rate": 2.4735065548694263e-05,
"loss": 0.8354,
"step": 8900
},
{
"epoch": 1.057566765578635,
"grad_norm": 1.0060994625091553,
"learning_rate": 2.4686005991658345e-05,
"loss": 0.8547,
"step": 8910
},
{
"epoch": 1.058753709198813,
"grad_norm": 1.0199803113937378,
"learning_rate": 2.463694764395797e-05,
"loss": 0.7854,
"step": 8920
},
{
"epoch": 1.0599406528189912,
"grad_norm": 1.029325008392334,
"learning_rate": 2.458789069453942e-05,
"loss": 0.8594,
"step": 8930
},
{
"epoch": 1.061127596439169,
"grad_norm": 0.9073032140731812,
"learning_rate": 2.453883533234361e-05,
"loss": 0.8842,
"step": 8940
},
{
"epoch": 1.0623145400593472,
"grad_norm": 0.9671692252159119,
"learning_rate": 2.4489781746305326e-05,
"loss": 0.7664,
"step": 8950
},
{
"epoch": 1.0635014836795251,
"grad_norm": 0.9562916159629822,
"learning_rate": 2.4440730125352528e-05,
"loss": 0.7997,
"step": 8960
},
{
"epoch": 1.0646884272997033,
"grad_norm": 1.353378176689148,
"learning_rate": 2.4391680658405585e-05,
"loss": 0.7817,
"step": 8970
},
{
"epoch": 1.0658753709198814,
"grad_norm": 1.0125224590301514,
"learning_rate": 2.4342633534376606e-05,
"loss": 0.7787,
"step": 8980
},
{
"epoch": 1.0670623145400593,
"grad_norm": 0.8546326160430908,
"learning_rate": 2.4293588942168626e-05,
"loss": 0.7713,
"step": 8990
},
{
"epoch": 1.0682492581602374,
"grad_norm": 1.0358268022537231,
"learning_rate": 2.424454707067499e-05,
"loss": 0.7992,
"step": 9000
},
{
"epoch": 1.0682492581602374,
"eval_loss": 0.7761093974113464,
"eval_runtime": 766.9914,
"eval_samples_per_second": 2.313,
"eval_steps_per_second": 0.579,
"step": 9000
},
{
"epoch": 1.0694362017804155,
"grad_norm": 1.4168592691421509,
"learning_rate": 2.4195508108778496e-05,
"loss": 0.8385,
"step": 9010
},
{
"epoch": 1.0706231454005934,
"grad_norm": 0.9119871854782104,
"learning_rate": 2.4146472245350805e-05,
"loss": 0.7911,
"step": 9020
},
{
"epoch": 1.0718100890207716,
"grad_norm": 1.2158715724945068,
"learning_rate": 2.4097439669251595e-05,
"loss": 0.7497,
"step": 9030
},
{
"epoch": 1.0729970326409495,
"grad_norm": 0.9663563966751099,
"learning_rate": 2.4048410569327892e-05,
"loss": 0.7418,
"step": 9040
},
{
"epoch": 1.0741839762611276,
"grad_norm": 0.8806156516075134,
"learning_rate": 2.399938513441336e-05,
"loss": 0.7516,
"step": 9050
},
{
"epoch": 1.0753709198813057,
"grad_norm": 1.1792296171188354,
"learning_rate": 2.3950363553327507e-05,
"loss": 0.8281,
"step": 9060
},
{
"epoch": 1.0765578635014836,
"grad_norm": 0.8951212763786316,
"learning_rate": 2.3901346014875037e-05,
"loss": 0.8098,
"step": 9070
},
{
"epoch": 1.0777448071216618,
"grad_norm": 0.8044575452804565,
"learning_rate": 2.3852332707845047e-05,
"loss": 0.779,
"step": 9080
},
{
"epoch": 1.0789317507418397,
"grad_norm": 1.0501245260238647,
"learning_rate": 2.380332382101038e-05,
"loss": 0.8397,
"step": 9090
},
{
"epoch": 1.0801186943620178,
"grad_norm": 1.2533349990844727,
"learning_rate": 2.375431954312681e-05,
"loss": 0.8961,
"step": 9100
},
{
"epoch": 1.081305637982196,
"grad_norm": 1.1930317878723145,
"learning_rate": 2.370532006293239e-05,
"loss": 0.8086,
"step": 9110
},
{
"epoch": 1.0824925816023738,
"grad_norm": 0.9481077790260315,
"learning_rate": 2.365632556914668e-05,
"loss": 0.7704,
"step": 9120
},
{
"epoch": 1.083679525222552,
"grad_norm": 0.9799014925956726,
"learning_rate": 2.360733625047005e-05,
"loss": 0.8051,
"step": 9130
},
{
"epoch": 1.0848664688427299,
"grad_norm": 0.669889509677887,
"learning_rate": 2.355835229558292e-05,
"loss": 0.829,
"step": 9140
},
{
"epoch": 1.086053412462908,
"grad_norm": 0.7075235843658447,
"learning_rate": 2.350937389314506e-05,
"loss": 0.7304,
"step": 9150
},
{
"epoch": 1.0872403560830861,
"grad_norm": 0.8748484253883362,
"learning_rate": 2.3460401231794865e-05,
"loss": 0.7795,
"step": 9160
},
{
"epoch": 1.088427299703264,
"grad_norm": 1.0108648538589478,
"learning_rate": 2.3411434500148592e-05,
"loss": 0.8265,
"step": 9170
},
{
"epoch": 1.0896142433234421,
"grad_norm": 1.1676548719406128,
"learning_rate": 2.3362473886799682e-05,
"loss": 0.8811,
"step": 9180
},
{
"epoch": 1.0908011869436203,
"grad_norm": 0.9329530000686646,
"learning_rate": 2.3313519580318025e-05,
"loss": 0.795,
"step": 9190
},
{
"epoch": 1.0919881305637982,
"grad_norm": 1.159879207611084,
"learning_rate": 2.3264571769249175e-05,
"loss": 0.7857,
"step": 9200
},
{
"epoch": 1.0931750741839763,
"grad_norm": 0.8164434432983398,
"learning_rate": 2.3215630642113716e-05,
"loss": 0.7752,
"step": 9210
},
{
"epoch": 1.0943620178041542,
"grad_norm": 0.9076749682426453,
"learning_rate": 2.316669638740645e-05,
"loss": 0.827,
"step": 9220
},
{
"epoch": 1.0955489614243323,
"grad_norm": 0.9610769748687744,
"learning_rate": 2.3117769193595756e-05,
"loss": 0.8259,
"step": 9230
},
{
"epoch": 1.0967359050445105,
"grad_norm": 1.0237845182418823,
"learning_rate": 2.3068849249122763e-05,
"loss": 0.7899,
"step": 9240
},
{
"epoch": 1.0979228486646884,
"grad_norm": 0.9556646347045898,
"learning_rate": 2.3019936742400724e-05,
"loss": 0.8264,
"step": 9250
},
{
"epoch": 1.0991097922848665,
"grad_norm": 0.9002740383148193,
"learning_rate": 2.2971031861814223e-05,
"loss": 0.8421,
"step": 9260
},
{
"epoch": 1.1002967359050446,
"grad_norm": 1.3685026168823242,
"learning_rate": 2.2922134795718477e-05,
"loss": 0.8126,
"step": 9270
},
{
"epoch": 1.1014836795252225,
"grad_norm": 0.9254816174507141,
"learning_rate": 2.287324573243862e-05,
"loss": 0.8157,
"step": 9280
},
{
"epoch": 1.1026706231454007,
"grad_norm": 1.0241316556930542,
"learning_rate": 2.2824364860268927e-05,
"loss": 0.8039,
"step": 9290
},
{
"epoch": 1.1038575667655786,
"grad_norm": 0.941983699798584,
"learning_rate": 2.277549236747218e-05,
"loss": 0.8532,
"step": 9300
},
{
"epoch": 1.1050445103857567,
"grad_norm": 0.837225615978241,
"learning_rate": 2.2726628442278826e-05,
"loss": 0.7953,
"step": 9310
},
{
"epoch": 1.1062314540059348,
"grad_norm": 1.406443476676941,
"learning_rate": 2.2677773272886373e-05,
"loss": 0.8247,
"step": 9320
},
{
"epoch": 1.1074183976261127,
"grad_norm": 0.8839634656906128,
"learning_rate": 2.2628927047458558e-05,
"loss": 0.8739,
"step": 9330
},
{
"epoch": 1.1086053412462908,
"grad_norm": 1.0190342664718628,
"learning_rate": 2.258008995412471e-05,
"loss": 0.7675,
"step": 9340
},
{
"epoch": 1.1097922848664687,
"grad_norm": 1.0243916511535645,
"learning_rate": 2.2531262180978952e-05,
"loss": 0.7931,
"step": 9350
},
{
"epoch": 1.1109792284866469,
"grad_norm": 1.0709168910980225,
"learning_rate": 2.2482443916079534e-05,
"loss": 0.8604,
"step": 9360
},
{
"epoch": 1.112166172106825,
"grad_norm": 0.733767569065094,
"learning_rate": 2.2433635347448084e-05,
"loss": 0.7005,
"step": 9370
},
{
"epoch": 1.113353115727003,
"grad_norm": 1.1727817058563232,
"learning_rate": 2.238483666306886e-05,
"loss": 0.7749,
"step": 9380
},
{
"epoch": 1.114540059347181,
"grad_norm": 1.2899572849273682,
"learning_rate": 2.2336048050888095e-05,
"loss": 0.7893,
"step": 9390
},
{
"epoch": 1.115727002967359,
"grad_norm": 1.053328037261963,
"learning_rate": 2.228726969881318e-05,
"loss": 0.7456,
"step": 9400
},
{
"epoch": 1.116913946587537,
"grad_norm": 1.1463031768798828,
"learning_rate": 2.2238501794712034e-05,
"loss": 0.8456,
"step": 9410
},
{
"epoch": 1.1181008902077152,
"grad_norm": 1.030958652496338,
"learning_rate": 2.2189744526412297e-05,
"loss": 0.8649,
"step": 9420
},
{
"epoch": 1.119287833827893,
"grad_norm": 1.2256028652191162,
"learning_rate": 2.2140998081700676e-05,
"loss": 0.8161,
"step": 9430
},
{
"epoch": 1.1204747774480712,
"grad_norm": 1.2485847473144531,
"learning_rate": 2.209226264832219e-05,
"loss": 0.8105,
"step": 9440
},
{
"epoch": 1.1216617210682494,
"grad_norm": 1.0359870195388794,
"learning_rate": 2.2043538413979425e-05,
"loss": 0.8313,
"step": 9450
},
{
"epoch": 1.1228486646884273,
"grad_norm": 1.0015512704849243,
"learning_rate": 2.199482556633185e-05,
"loss": 0.8401,
"step": 9460
},
{
"epoch": 1.1240356083086054,
"grad_norm": 1.3324792385101318,
"learning_rate": 2.1946124292995075e-05,
"loss": 0.8473,
"step": 9470
},
{
"epoch": 1.1252225519287833,
"grad_norm": 0.9533900618553162,
"learning_rate": 2.1897434781540144e-05,
"loss": 0.7981,
"step": 9480
},
{
"epoch": 1.1264094955489614,
"grad_norm": 1.0634273290634155,
"learning_rate": 2.1848757219492768e-05,
"loss": 0.7351,
"step": 9490
},
{
"epoch": 1.1275964391691395,
"grad_norm": 1.1996541023254395,
"learning_rate": 2.1800091794332672e-05,
"loss": 0.8202,
"step": 9500
},
{
"epoch": 1.1275964391691395,
"eval_loss": 0.7751370072364807,
"eval_runtime": 772.0681,
"eval_samples_per_second": 2.298,
"eval_steps_per_second": 0.575,
"step": 9500
},
{
"epoch": 1.1287833827893174,
"grad_norm": 0.9720987677574158,
"learning_rate": 2.1751438693492834e-05,
"loss": 0.7615,
"step": 9510
},
{
"epoch": 1.1299703264094956,
"grad_norm": 1.2057288885116577,
"learning_rate": 2.1702798104358728e-05,
"loss": 0.8152,
"step": 9520
},
{
"epoch": 1.1311572700296737,
"grad_norm": 1.2114310264587402,
"learning_rate": 2.1654170214267682e-05,
"loss": 0.7721,
"step": 9530
},
{
"epoch": 1.1323442136498516,
"grad_norm": 0.9816735982894897,
"learning_rate": 2.1605555210508086e-05,
"loss": 0.8552,
"step": 9540
},
{
"epoch": 1.1335311572700297,
"grad_norm": 1.4814951419830322,
"learning_rate": 2.155695328031872e-05,
"loss": 0.9111,
"step": 9550
},
{
"epoch": 1.1347181008902076,
"grad_norm": 1.1997096538543701,
"learning_rate": 2.150836461088799e-05,
"loss": 0.8304,
"step": 9560
},
{
"epoch": 1.1359050445103858,
"grad_norm": 0.9619604349136353,
"learning_rate": 2.1459789389353254e-05,
"loss": 0.8256,
"step": 9570
},
{
"epoch": 1.137091988130564,
"grad_norm": 0.8761364221572876,
"learning_rate": 2.1411227802800043e-05,
"loss": 0.832,
"step": 9580
},
{
"epoch": 1.1382789317507418,
"grad_norm": 1.1116739511489868,
"learning_rate": 2.13626800382614e-05,
"loss": 0.81,
"step": 9590
},
{
"epoch": 1.13946587537092,
"grad_norm": 1.316489577293396,
"learning_rate": 2.1314146282717144e-05,
"loss": 0.7807,
"step": 9600
},
{
"epoch": 1.1406528189910978,
"grad_norm": 0.8193405866622925,
"learning_rate": 2.1265626723093087e-05,
"loss": 0.7839,
"step": 9610
},
{
"epoch": 1.141839762611276,
"grad_norm": 1.4620572328567505,
"learning_rate": 2.121712154626043e-05,
"loss": 0.7752,
"step": 9620
},
{
"epoch": 1.143026706231454,
"grad_norm": 1.1185797452926636,
"learning_rate": 2.1168630939034923e-05,
"loss": 0.8407,
"step": 9630
},
{
"epoch": 1.144213649851632,
"grad_norm": 1.177851676940918,
"learning_rate": 2.112015508817625e-05,
"loss": 0.8171,
"step": 9640
},
{
"epoch": 1.1454005934718101,
"grad_norm": 1.0961945056915283,
"learning_rate": 2.1071694180387217e-05,
"loss": 0.7837,
"step": 9650
},
{
"epoch": 1.146587537091988,
"grad_norm": 0.984928548336029,
"learning_rate": 2.1023248402313118e-05,
"loss": 0.7953,
"step": 9660
},
{
"epoch": 1.1477744807121661,
"grad_norm": 1.216512680053711,
"learning_rate": 2.097481794054094e-05,
"loss": 0.8274,
"step": 9670
},
{
"epoch": 1.1489614243323443,
"grad_norm": 1.20542573928833,
"learning_rate": 2.092640298159871e-05,
"loss": 0.8014,
"step": 9680
},
{
"epoch": 1.1501483679525222,
"grad_norm": 1.1971958875656128,
"learning_rate": 2.0878003711954727e-05,
"loss": 0.8743,
"step": 9690
},
{
"epoch": 1.1513353115727003,
"grad_norm": 0.7235596776008606,
"learning_rate": 2.0829620318016863e-05,
"loss": 0.7358,
"step": 9700
},
{
"epoch": 1.1525222551928784,
"grad_norm": 0.8866942524909973,
"learning_rate": 2.0781252986131867e-05,
"loss": 0.7841,
"step": 9710
},
{
"epoch": 1.1537091988130563,
"grad_norm": 0.8466529846191406,
"learning_rate": 2.073290190258459e-05,
"loss": 0.7781,
"step": 9720
},
{
"epoch": 1.1548961424332345,
"grad_norm": 1.2863926887512207,
"learning_rate": 2.068456725359735e-05,
"loss": 0.7886,
"step": 9730
},
{
"epoch": 1.1560830860534124,
"grad_norm": 1.4690605401992798,
"learning_rate": 2.0636249225329105e-05,
"loss": 0.8238,
"step": 9740
},
{
"epoch": 1.1572700296735905,
"grad_norm": 1.0647927522659302,
"learning_rate": 2.058794800387486e-05,
"loss": 0.8475,
"step": 9750
},
{
"epoch": 1.1584569732937686,
"grad_norm": 0.860795795917511,
"learning_rate": 2.053966377526487e-05,
"loss": 0.7295,
"step": 9760
},
{
"epoch": 1.1596439169139465,
"grad_norm": 1.1758381128311157,
"learning_rate": 2.0491396725463916e-05,
"loss": 0.7597,
"step": 9770
},
{
"epoch": 1.1608308605341247,
"grad_norm": 1.068555235862732,
"learning_rate": 2.0443147040370657e-05,
"loss": 0.7811,
"step": 9780
},
{
"epoch": 1.1620178041543028,
"grad_norm": 0.9324743151664734,
"learning_rate": 2.0394914905816836e-05,
"loss": 0.8155,
"step": 9790
},
{
"epoch": 1.1632047477744807,
"grad_norm": 1.0542433261871338,
"learning_rate": 2.0346700507566628e-05,
"loss": 0.8496,
"step": 9800
},
{
"epoch": 1.1643916913946588,
"grad_norm": 1.0209101438522339,
"learning_rate": 2.0298504031315875e-05,
"loss": 0.8186,
"step": 9810
},
{
"epoch": 1.1655786350148367,
"grad_norm": 1.0995596647262573,
"learning_rate": 2.0250325662691422e-05,
"loss": 0.8151,
"step": 9820
},
{
"epoch": 1.1667655786350148,
"grad_norm": 1.0881950855255127,
"learning_rate": 2.020216558725033e-05,
"loss": 0.7281,
"step": 9830
},
{
"epoch": 1.167952522255193,
"grad_norm": 0.8314598202705383,
"learning_rate": 2.0154023990479244e-05,
"loss": 0.7935,
"step": 9840
},
{
"epoch": 1.1691394658753709,
"grad_norm": 0.8414393067359924,
"learning_rate": 2.010590105779364e-05,
"loss": 0.7565,
"step": 9850
},
{
"epoch": 1.170326409495549,
"grad_norm": 1.2225565910339355,
"learning_rate": 2.0057796974537066e-05,
"loss": 0.8241,
"step": 9860
},
{
"epoch": 1.171513353115727,
"grad_norm": 1.2027302980422974,
"learning_rate": 2.000971192598053e-05,
"loss": 0.8656,
"step": 9870
},
{
"epoch": 1.172700296735905,
"grad_norm": 0.9175437688827515,
"learning_rate": 1.996164609732167e-05,
"loss": 0.7888,
"step": 9880
},
{
"epoch": 1.1738872403560832,
"grad_norm": 1.1863752603530884,
"learning_rate": 1.991359967368416e-05,
"loss": 0.8331,
"step": 9890
},
{
"epoch": 1.175074183976261,
"grad_norm": 0.907460629940033,
"learning_rate": 1.9865572840116888e-05,
"loss": 0.8613,
"step": 9900
},
{
"epoch": 1.1762611275964392,
"grad_norm": 1.053999423980713,
"learning_rate": 1.9817565781593302e-05,
"loss": 0.806,
"step": 9910
},
{
"epoch": 1.177448071216617,
"grad_norm": 0.7389574646949768,
"learning_rate": 1.9769578683010714e-05,
"loss": 0.8112,
"step": 9920
},
{
"epoch": 1.1786350148367952,
"grad_norm": 1.1020927429199219,
"learning_rate": 1.9721611729189527e-05,
"loss": 0.8336,
"step": 9930
},
{
"epoch": 1.1798219584569734,
"grad_norm": 1.0626370906829834,
"learning_rate": 1.967366510487258e-05,
"loss": 0.8218,
"step": 9940
},
{
"epoch": 1.1810089020771513,
"grad_norm": 1.0998564958572388,
"learning_rate": 1.96257389947244e-05,
"loss": 0.7585,
"step": 9950
},
{
"epoch": 1.1821958456973294,
"grad_norm": 1.141060471534729,
"learning_rate": 1.9577833583330518e-05,
"loss": 0.8047,
"step": 9960
},
{
"epoch": 1.1833827893175075,
"grad_norm": 1.1182451248168945,
"learning_rate": 1.9529949055196724e-05,
"loss": 0.7609,
"step": 9970
},
{
"epoch": 1.1845697329376854,
"grad_norm": 0.8619179129600525,
"learning_rate": 1.94820855947484e-05,
"loss": 0.7636,
"step": 9980
},
{
"epoch": 1.1857566765578635,
"grad_norm": 1.414411187171936,
"learning_rate": 1.9434243386329764e-05,
"loss": 0.7444,
"step": 9990
},
{
"epoch": 1.1869436201780414,
"grad_norm": 1.1006312370300293,
"learning_rate": 1.9386422614203212e-05,
"loss": 0.8047,
"step": 10000
},
{
"epoch": 1.1869436201780414,
"eval_loss": 0.7749348282814026,
"eval_runtime": 772.3216,
"eval_samples_per_second": 2.297,
"eval_steps_per_second": 0.575,
"step": 10000
},
{
"epoch": 1.1881305637982196,
"grad_norm": 0.8358052968978882,
"learning_rate": 1.933862346254855e-05,
"loss": 0.793,
"step": 10010
},
{
"epoch": 1.1893175074183977,
"grad_norm": 0.9573304653167725,
"learning_rate": 1.929084611546233e-05,
"loss": 0.8799,
"step": 10020
},
{
"epoch": 1.1905044510385756,
"grad_norm": 1.1877682209014893,
"learning_rate": 1.924309075695712e-05,
"loss": 0.8131,
"step": 10030
},
{
"epoch": 1.1916913946587537,
"grad_norm": 1.4036526679992676,
"learning_rate": 1.919535757096079e-05,
"loss": 0.8705,
"step": 10040
},
{
"epoch": 1.1928783382789319,
"grad_norm": 1.054739236831665,
"learning_rate": 1.914764674131584e-05,
"loss": 0.8792,
"step": 10050
},
{
"epoch": 1.1940652818991098,
"grad_norm": 1.0724046230316162,
"learning_rate": 1.909995845177863e-05,
"loss": 0.7757,
"step": 10060
},
{
"epoch": 1.195252225519288,
"grad_norm": 1.3081406354904175,
"learning_rate": 1.9052292886018745e-05,
"loss": 0.7989,
"step": 10070
},
{
"epoch": 1.1964391691394658,
"grad_norm": 1.3429408073425293,
"learning_rate": 1.9004650227618215e-05,
"loss": 0.8631,
"step": 10080
},
{
"epoch": 1.197626112759644,
"grad_norm": 1.1177902221679688,
"learning_rate": 1.8957030660070862e-05,
"loss": 0.7897,
"step": 10090
},
{
"epoch": 1.198813056379822,
"grad_norm": 1.1529994010925293,
"learning_rate": 1.8909434366781592e-05,
"loss": 0.8375,
"step": 10100
},
{
"epoch": 1.2,
"grad_norm": 1.2478564977645874,
"learning_rate": 1.8861861531065622e-05,
"loss": 0.7458,
"step": 10110
},
{
"epoch": 1.201186943620178,
"grad_norm": 0.9045754671096802,
"learning_rate": 1.881431233614788e-05,
"loss": 0.8016,
"step": 10120
},
{
"epoch": 1.202373887240356,
"grad_norm": 1.0951228141784668,
"learning_rate": 1.8766786965162198e-05,
"loss": 0.7926,
"step": 10130
},
{
"epoch": 1.2035608308605341,
"grad_norm": 1.0310611724853516,
"learning_rate": 1.8719285601150665e-05,
"loss": 0.859,
"step": 10140
},
{
"epoch": 1.2047477744807122,
"grad_norm": 1.3347513675689697,
"learning_rate": 1.8671808427062924e-05,
"loss": 0.8695,
"step": 10150
},
{
"epoch": 1.2059347181008901,
"grad_norm": 1.2841068506240845,
"learning_rate": 1.8624355625755423e-05,
"loss": 0.7943,
"step": 10160
},
{
"epoch": 1.2071216617210683,
"grad_norm": 1.022180199623108,
"learning_rate": 1.8576927379990776e-05,
"loss": 0.8996,
"step": 10170
},
{
"epoch": 1.2083086053412462,
"grad_norm": 0.9655975103378296,
"learning_rate": 1.852952387243698e-05,
"loss": 0.7883,
"step": 10180
},
{
"epoch": 1.2094955489614243,
"grad_norm": 1.4115021228790283,
"learning_rate": 1.848214528566679e-05,
"loss": 0.7879,
"step": 10190
},
{
"epoch": 1.2106824925816024,
"grad_norm": 1.3277826309204102,
"learning_rate": 1.843479180215695e-05,
"loss": 0.8263,
"step": 10200
},
{
"epoch": 1.2118694362017803,
"grad_norm": 1.1939499378204346,
"learning_rate": 1.8387463604287554e-05,
"loss": 0.8068,
"step": 10210
},
{
"epoch": 1.2130563798219585,
"grad_norm": 1.1342905759811401,
"learning_rate": 1.834016087434126e-05,
"loss": 0.7618,
"step": 10220
},
{
"epoch": 1.2142433234421366,
"grad_norm": 1.5079452991485596,
"learning_rate": 1.8292883794502697e-05,
"loss": 0.7817,
"step": 10230
},
{
"epoch": 1.2154302670623145,
"grad_norm": 0.8599739074707031,
"learning_rate": 1.8245632546857655e-05,
"loss": 0.7645,
"step": 10240
},
{
"epoch": 1.2166172106824926,
"grad_norm": 1.26143217086792,
"learning_rate": 1.8198407313392447e-05,
"loss": 0.8963,
"step": 10250
},
{
"epoch": 1.2178041543026705,
"grad_norm": 1.018254041671753,
"learning_rate": 1.815120827599322e-05,
"loss": 0.8858,
"step": 10260
},
{
"epoch": 1.2189910979228487,
"grad_norm": 1.004217505455017,
"learning_rate": 1.8104035616445182e-05,
"loss": 0.8171,
"step": 10270
},
{
"epoch": 1.2201780415430268,
"grad_norm": 1.0886305570602417,
"learning_rate": 1.8056889516431986e-05,
"loss": 0.7771,
"step": 10280
},
{
"epoch": 1.2213649851632047,
"grad_norm": 1.4587886333465576,
"learning_rate": 1.8009770157534955e-05,
"loss": 0.8639,
"step": 10290
},
{
"epoch": 1.2225519287833828,
"grad_norm": 1.419061303138733,
"learning_rate": 1.7962677721232475e-05,
"loss": 0.7838,
"step": 10300
},
{
"epoch": 1.223738872403561,
"grad_norm": 0.8688335418701172,
"learning_rate": 1.7915612388899173e-05,
"loss": 0.7259,
"step": 10310
},
{
"epoch": 1.2249258160237388,
"grad_norm": 1.0464247465133667,
"learning_rate": 1.786857434180534e-05,
"loss": 0.7191,
"step": 10320
},
{
"epoch": 1.226112759643917,
"grad_norm": 1.0211148262023926,
"learning_rate": 1.7821563761116157e-05,
"loss": 0.7761,
"step": 10330
},
{
"epoch": 1.2272997032640949,
"grad_norm": 1.3132156133651733,
"learning_rate": 1.777458082789102e-05,
"loss": 0.8556,
"step": 10340
},
{
"epoch": 1.228486646884273,
"grad_norm": 1.148878812789917,
"learning_rate": 1.7727625723082843e-05,
"loss": 0.8134,
"step": 10350
},
{
"epoch": 1.2296735905044511,
"grad_norm": 1.038454294204712,
"learning_rate": 1.7680698627537355e-05,
"loss": 0.8661,
"step": 10360
},
{
"epoch": 1.230860534124629,
"grad_norm": 1.1032830476760864,
"learning_rate": 1.7633799721992423e-05,
"loss": 0.8575,
"step": 10370
},
{
"epoch": 1.2320474777448072,
"grad_norm": 0.923642635345459,
"learning_rate": 1.7586929187077318e-05,
"loss": 0.8379,
"step": 10380
},
{
"epoch": 1.233234421364985,
"grad_norm": 1.3686367273330688,
"learning_rate": 1.7540087203312074e-05,
"loss": 0.8902,
"step": 10390
},
{
"epoch": 1.2344213649851632,
"grad_norm": 1.2714051008224487,
"learning_rate": 1.7493273951106727e-05,
"loss": 0.9215,
"step": 10400
},
{
"epoch": 1.2356083086053413,
"grad_norm": 1.0131839513778687,
"learning_rate": 1.744648961076068e-05,
"loss": 0.8196,
"step": 10410
},
{
"epoch": 1.2367952522255192,
"grad_norm": 1.1317576169967651,
"learning_rate": 1.7399734362461985e-05,
"loss": 0.816,
"step": 10420
},
{
"epoch": 1.2379821958456974,
"grad_norm": 0.8517586588859558,
"learning_rate": 1.7353008386286614e-05,
"loss": 0.8241,
"step": 10430
},
{
"epoch": 1.2391691394658753,
"grad_norm": 1.0312011241912842,
"learning_rate": 1.7306311862197843e-05,
"loss": 0.8058,
"step": 10440
},
{
"epoch": 1.2403560830860534,
"grad_norm": 0.7310718894004822,
"learning_rate": 1.725964497004548e-05,
"loss": 0.7813,
"step": 10450
},
{
"epoch": 1.2415430267062315,
"grad_norm": 0.9712517857551575,
"learning_rate": 1.7213007889565225e-05,
"loss": 0.8308,
"step": 10460
},
{
"epoch": 1.2427299703264094,
"grad_norm": 1.0188968181610107,
"learning_rate": 1.7166400800377948e-05,
"loss": 0.7791,
"step": 10470
},
{
"epoch": 1.2439169139465875,
"grad_norm": 1.0058274269104004,
"learning_rate": 1.7119823881989012e-05,
"loss": 0.8117,
"step": 10480
},
{
"epoch": 1.2451038575667657,
"grad_norm": 1.0918079614639282,
"learning_rate": 1.7073277313787602e-05,
"loss": 0.7714,
"step": 10490
},
{
"epoch": 1.2462908011869436,
"grad_norm": 0.7621090412139893,
"learning_rate": 1.7026761275045965e-05,
"loss": 0.7538,
"step": 10500
},
{
"epoch": 1.2462908011869436,
"eval_loss": 0.7732818722724915,
"eval_runtime": 804.3434,
"eval_samples_per_second": 2.206,
"eval_steps_per_second": 0.552,
"step": 10500
},
{
"epoch": 1.2474777448071217,
"grad_norm": 1.076611876487732,
"learning_rate": 1.6980275944918804e-05,
"loss": 0.7582,
"step": 10510
},
{
"epoch": 1.2486646884272996,
"grad_norm": 0.965393602848053,
"learning_rate": 1.6933821502442527e-05,
"loss": 0.7812,
"step": 10520
},
{
"epoch": 1.2498516320474777,
"grad_norm": 1.0871965885162354,
"learning_rate": 1.6887398126534594e-05,
"loss": 0.7601,
"step": 10530
},
{
"epoch": 1.2510385756676559,
"grad_norm": 0.9238999485969543,
"learning_rate": 1.68410059959928e-05,
"loss": 0.8194,
"step": 10540
},
{
"epoch": 1.2522255192878338,
"grad_norm": 1.1522690057754517,
"learning_rate": 1.679464528949461e-05,
"loss": 0.8596,
"step": 10550
},
{
"epoch": 1.253412462908012,
"grad_norm": 1.2266411781311035,
"learning_rate": 1.674831618559645e-05,
"loss": 0.7715,
"step": 10560
},
{
"epoch": 1.25459940652819,
"grad_norm": 1.0644117593765259,
"learning_rate": 1.6702018862733045e-05,
"loss": 0.7796,
"step": 10570
},
{
"epoch": 1.255786350148368,
"grad_norm": 1.1088664531707764,
"learning_rate": 1.6655753499216708e-05,
"loss": 0.8424,
"step": 10580
},
{
"epoch": 1.256973293768546,
"grad_norm": 1.0059804916381836,
"learning_rate": 1.660952027323665e-05,
"loss": 0.8276,
"step": 10590
},
{
"epoch": 1.258160237388724,
"grad_norm": 0.947927713394165,
"learning_rate": 1.6563319362858333e-05,
"loss": 0.8039,
"step": 10600
},
{
"epoch": 1.259347181008902,
"grad_norm": 1.1396689414978027,
"learning_rate": 1.6517150946022725e-05,
"loss": 0.8359,
"step": 10610
},
{
"epoch": 1.2605341246290802,
"grad_norm": 1.1726751327514648,
"learning_rate": 1.6471015200545682e-05,
"loss": 0.7481,
"step": 10620
},
{
"epoch": 1.2617210682492581,
"grad_norm": 0.9562637805938721,
"learning_rate": 1.6424912304117183e-05,
"loss": 0.7441,
"step": 10630
},
{
"epoch": 1.2629080118694362,
"grad_norm": 1.0816010236740112,
"learning_rate": 1.6378842434300746e-05,
"loss": 0.8136,
"step": 10640
},
{
"epoch": 1.2640949554896141,
"grad_norm": 1.0869776010513306,
"learning_rate": 1.6332805768532628e-05,
"loss": 0.8231,
"step": 10650
},
{
"epoch": 1.2652818991097923,
"grad_norm": 0.6635779142379761,
"learning_rate": 1.6286802484121244e-05,
"loss": 0.7143,
"step": 10660
},
{
"epoch": 1.2664688427299704,
"grad_norm": 1.0061339139938354,
"learning_rate": 1.624083275824644e-05,
"loss": 0.7553,
"step": 10670
},
{
"epoch": 1.2676557863501483,
"grad_norm": 1.0020318031311035,
"learning_rate": 1.6194896767958785e-05,
"loss": 0.7943,
"step": 10680
},
{
"epoch": 1.2688427299703264,
"grad_norm": 0.9754176735877991,
"learning_rate": 1.614899469017894e-05,
"loss": 0.7514,
"step": 10690
},
{
"epoch": 1.2700296735905043,
"grad_norm": 0.9844280481338501,
"learning_rate": 1.610312670169695e-05,
"loss": 0.8493,
"step": 10700
},
{
"epoch": 1.2712166172106825,
"grad_norm": 1.2735682725906372,
"learning_rate": 1.6057292979171565e-05,
"loss": 0.8771,
"step": 10710
},
{
"epoch": 1.2724035608308606,
"grad_norm": 1.0701887607574463,
"learning_rate": 1.6011493699129544e-05,
"loss": 0.8163,
"step": 10720
},
{
"epoch": 1.2735905044510385,
"grad_norm": 1.1788859367370605,
"learning_rate": 1.596572903796502e-05,
"loss": 0.8017,
"step": 10730
},
{
"epoch": 1.2747774480712166,
"grad_norm": 1.3015997409820557,
"learning_rate": 1.5919999171938792e-05,
"loss": 0.8182,
"step": 10740
},
{
"epoch": 1.2759643916913945,
"grad_norm": 0.8140105605125427,
"learning_rate": 1.5874304277177616e-05,
"loss": 0.8411,
"step": 10750
},
{
"epoch": 1.2771513353115727,
"grad_norm": 1.0661457777023315,
"learning_rate": 1.582864452967359e-05,
"loss": 0.7887,
"step": 10760
},
{
"epoch": 1.2783382789317508,
"grad_norm": 1.1822696924209595,
"learning_rate": 1.5783020105283416e-05,
"loss": 0.878,
"step": 10770
},
{
"epoch": 1.279525222551929,
"grad_norm": 1.4426151514053345,
"learning_rate": 1.5737431179727785e-05,
"loss": 0.7839,
"step": 10780
},
{
"epoch": 1.2807121661721068,
"grad_norm": 1.173456072807312,
"learning_rate": 1.5691877928590624e-05,
"loss": 0.7866,
"step": 10790
},
{
"epoch": 1.281899109792285,
"grad_norm": 1.166433334350586,
"learning_rate": 1.5646360527318498e-05,
"loss": 0.7891,
"step": 10800
},
{
"epoch": 1.2830860534124628,
"grad_norm": 0.8934957385063171,
"learning_rate": 1.5600879151219878e-05,
"loss": 0.8032,
"step": 10810
},
{
"epoch": 1.284272997032641,
"grad_norm": 0.9699043035507202,
"learning_rate": 1.5555433975464488e-05,
"loss": 0.8116,
"step": 10820
},
{
"epoch": 1.285459940652819,
"grad_norm": 1.186395287513733,
"learning_rate": 1.551002517508264e-05,
"loss": 0.7575,
"step": 10830
},
{
"epoch": 1.286646884272997,
"grad_norm": 1.1024322509765625,
"learning_rate": 1.546465292496452e-05,
"loss": 0.8077,
"step": 10840
},
{
"epoch": 1.2878338278931751,
"grad_norm": 1.213882565498352,
"learning_rate": 1.541931739985958e-05,
"loss": 0.7592,
"step": 10850
},
{
"epoch": 1.289020771513353,
"grad_norm": 1.1226390600204468,
"learning_rate": 1.537401877437578e-05,
"loss": 0.7934,
"step": 10860
},
{
"epoch": 1.2902077151335312,
"grad_norm": 0.8497867584228516,
"learning_rate": 1.5328757222979025e-05,
"loss": 0.787,
"step": 10870
},
{
"epoch": 1.2913946587537093,
"grad_norm": 1.041912317276001,
"learning_rate": 1.5283532919992362e-05,
"loss": 0.8114,
"step": 10880
},
{
"epoch": 1.2925816023738872,
"grad_norm": 1.0151110887527466,
"learning_rate": 1.5238346039595433e-05,
"loss": 0.8071,
"step": 10890
},
{
"epoch": 1.2937685459940653,
"grad_norm": 1.183481216430664,
"learning_rate": 1.5193196755823714e-05,
"loss": 0.7622,
"step": 10900
},
{
"epoch": 1.2949554896142432,
"grad_norm": 1.1755239963531494,
"learning_rate": 1.5148085242567894e-05,
"loss": 0.8257,
"step": 10910
},
{
"epoch": 1.2961424332344214,
"grad_norm": 1.057356834411621,
"learning_rate": 1.5103011673573193e-05,
"loss": 0.8093,
"step": 10920
},
{
"epoch": 1.2973293768545995,
"grad_norm": 0.8958760499954224,
"learning_rate": 1.5057976222438667e-05,
"loss": 0.722,
"step": 10930
},
{
"epoch": 1.2985163204747774,
"grad_norm": 1.42214834690094,
"learning_rate": 1.50129790626166e-05,
"loss": 0.8296,
"step": 10940
},
{
"epoch": 1.2997032640949555,
"grad_norm": 1.1652711629867554,
"learning_rate": 1.4968020367411758e-05,
"loss": 0.8211,
"step": 10950
},
{
"epoch": 1.3008902077151334,
"grad_norm": 0.9645259380340576,
"learning_rate": 1.4923100309980805e-05,
"loss": 0.7363,
"step": 10960
},
{
"epoch": 1.3020771513353115,
"grad_norm": 1.1167161464691162,
"learning_rate": 1.487821906333155e-05,
"loss": 0.8433,
"step": 10970
},
{
"epoch": 1.3032640949554897,
"grad_norm": 0.841799795627594,
"learning_rate": 1.4833376800322352e-05,
"loss": 0.8135,
"step": 10980
},
{
"epoch": 1.3044510385756676,
"grad_norm": 1.0974364280700684,
"learning_rate": 1.4788573693661429e-05,
"loss": 0.8523,
"step": 10990
},
{
"epoch": 1.3056379821958457,
"grad_norm": 1.2325725555419922,
"learning_rate": 1.4743809915906162e-05,
"loss": 0.8301,
"step": 11000
},
{
"epoch": 1.3056379821958457,
"eval_loss": 0.7716161012649536,
"eval_runtime": 773.4944,
"eval_samples_per_second": 2.293,
"eval_steps_per_second": 0.574,
"step": 11000
}
],
"logging_steps": 10,
"max_steps": 16850,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 500,
"stateful_callbacks": {
"EarlyStoppingCallback": {
"args": {
"early_stopping_patience": 4,
"early_stopping_threshold": 0.0
},
"attributes": {
"early_stopping_patience_counter": 0
}
},
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 7.667075811983229e+18,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}