OpenKurdish_V2_Checkpoint / trainer_state.json
Azadio's picture
Upload folder using huggingface_hub
dec9afe
Raw History Blame Contribute Delete
167 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.004685567203528545,
"eval_steps": 500,
"global_step": 15000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 2.4989691752152238e-05,
"grad_norm": 9.5625,
"learning_rate": 1.9998e-05,
"loss": 6.0983,
"step": 10
},
{
"epoch": 4.9979383504304475e-05,
"grad_norm": 11.125,
"learning_rate": 1.9996000000000003e-05,
"loss": 5.5915,
"step": 20
},
{
"epoch": 7.496907525645671e-05,
"grad_norm": 11.0625,
"learning_rate": 1.9994000000000002e-05,
"loss": 5.6105,
"step": 30
},
{
"epoch": 9.995876700860895e-05,
"grad_norm": 7.53125,
"learning_rate": 1.9992e-05,
"loss": 5.593,
"step": 40
},
{
"epoch": 0.00012494845876076117,
"grad_norm": 9.125,
"learning_rate": 1.9990000000000003e-05,
"loss": 5.6984,
"step": 50
},
{
"epoch": 0.00014993815051291343,
"grad_norm": 9.75,
"learning_rate": 1.9988000000000002e-05,
"loss": 5.6445,
"step": 60
},
{
"epoch": 0.00017492784226506565,
"grad_norm": 9.3125,
"learning_rate": 1.9986e-05,
"loss": 5.6131,
"step": 70
},
{
"epoch": 0.0001999175340172179,
"grad_norm": 12.9375,
"learning_rate": 1.9984e-05,
"loss": 5.5871,
"step": 80
},
{
"epoch": 0.00022490722576937012,
"grad_norm": 7.375,
"learning_rate": 1.9982000000000003e-05,
"loss": 5.6247,
"step": 90
},
{
"epoch": 0.00024989691752152235,
"grad_norm": 6.6875,
"learning_rate": 1.9980000000000002e-05,
"loss": 5.6734,
"step": 100
},
{
"epoch": 0.0002748866092736746,
"grad_norm": 7.9375,
"learning_rate": 1.9978e-05,
"loss": 5.5597,
"step": 110
},
{
"epoch": 0.00029987630102582685,
"grad_norm": 9.125,
"learning_rate": 1.9976000000000003e-05,
"loss": 5.5558,
"step": 120
},
{
"epoch": 0.0003248659927779791,
"grad_norm": 7.40625,
"learning_rate": 1.9974000000000002e-05,
"loss": 5.6126,
"step": 130
},
{
"epoch": 0.0003498556845301313,
"grad_norm": 12.8125,
"learning_rate": 1.9972e-05,
"loss": 5.4829,
"step": 140
},
{
"epoch": 0.00037484537628228355,
"grad_norm": 7.75,
"learning_rate": 1.9970000000000004e-05,
"loss": 5.5903,
"step": 150
},
{
"epoch": 0.0003998350680344358,
"grad_norm": 8.3125,
"learning_rate": 1.9968e-05,
"loss": 5.6543,
"step": 160
},
{
"epoch": 0.00042482475978658805,
"grad_norm": 7.53125,
"learning_rate": 1.9966e-05,
"loss": 5.6299,
"step": 170
},
{
"epoch": 0.00044981445153874025,
"grad_norm": 8.875,
"learning_rate": 1.9964e-05,
"loss": 5.6187,
"step": 180
},
{
"epoch": 0.0004748041432908925,
"grad_norm": 7.0625,
"learning_rate": 1.9962000000000003e-05,
"loss": 5.5635,
"step": 190
},
{
"epoch": 0.0004997938350430447,
"grad_norm": 7.8125,
"learning_rate": 1.9960000000000002e-05,
"loss": 5.4657,
"step": 200
},
{
"epoch": 0.000524783526795197,
"grad_norm": 8.3125,
"learning_rate": 1.9958e-05,
"loss": 5.5554,
"step": 210
},
{
"epoch": 0.0005497732185473492,
"grad_norm": 10.625,
"learning_rate": 1.9956000000000003e-05,
"loss": 5.6229,
"step": 220
},
{
"epoch": 0.0005747629102995015,
"grad_norm": 8.4375,
"learning_rate": 1.9954000000000002e-05,
"loss": 5.4991,
"step": 230
},
{
"epoch": 0.0005997526020516537,
"grad_norm": 7.15625,
"learning_rate": 1.9952e-05,
"loss": 5.5635,
"step": 240
},
{
"epoch": 0.0006247422938038059,
"grad_norm": 7.125,
"learning_rate": 1.9950000000000004e-05,
"loss": 5.6043,
"step": 250
},
{
"epoch": 0.0006497319855559582,
"grad_norm": 11.0625,
"learning_rate": 1.9948e-05,
"loss": 5.6521,
"step": 260
},
{
"epoch": 0.0006747216773081104,
"grad_norm": 7.59375,
"learning_rate": 1.9946000000000002e-05,
"loss": 5.5403,
"step": 270
},
{
"epoch": 0.0006997113690602626,
"grad_norm": 8.8125,
"learning_rate": 1.9944e-05,
"loss": 5.5962,
"step": 280
},
{
"epoch": 0.0007247010608124149,
"grad_norm": 8.25,
"learning_rate": 1.9942e-05,
"loss": 5.6138,
"step": 290
},
{
"epoch": 0.0007496907525645671,
"grad_norm": 7.625,
"learning_rate": 1.9940000000000002e-05,
"loss": 5.6668,
"step": 300
},
{
"epoch": 0.0007746804443167194,
"grad_norm": 9.75,
"learning_rate": 1.9938e-05,
"loss": 5.5858,
"step": 310
},
{
"epoch": 0.0007996701360688716,
"grad_norm": 8.875,
"learning_rate": 1.9936000000000004e-05,
"loss": 5.6348,
"step": 320
},
{
"epoch": 0.0008246598278210238,
"grad_norm": 8.5,
"learning_rate": 1.9934000000000003e-05,
"loss": 5.48,
"step": 330
},
{
"epoch": 0.0008496495195731761,
"grad_norm": 7.25,
"learning_rate": 1.9932e-05,
"loss": 5.5633,
"step": 340
},
{
"epoch": 0.0008746392113253283,
"grad_norm": 10.75,
"learning_rate": 1.9930000000000004e-05,
"loss": 5.5039,
"step": 350
},
{
"epoch": 0.0008996289030774805,
"grad_norm": 6.40625,
"learning_rate": 1.9928e-05,
"loss": 5.6185,
"step": 360
},
{
"epoch": 0.0009246185948296328,
"grad_norm": 7.15625,
"learning_rate": 1.9926000000000002e-05,
"loss": 5.6497,
"step": 370
},
{
"epoch": 0.000949608286581785,
"grad_norm": 8.5,
"learning_rate": 1.9924e-05,
"loss": 5.6076,
"step": 380
},
{
"epoch": 0.0009745979783339372,
"grad_norm": 7.6875,
"learning_rate": 1.9922e-05,
"loss": 5.549,
"step": 390
},
{
"epoch": 0.0009995876700860894,
"grad_norm": 7.03125,
"learning_rate": 1.9920000000000002e-05,
"loss": 5.537,
"step": 400
},
{
"epoch": 0.0010245773618382417,
"grad_norm": 8.25,
"learning_rate": 1.9918e-05,
"loss": 5.5424,
"step": 410
},
{
"epoch": 0.001049567053590394,
"grad_norm": 9.3125,
"learning_rate": 1.9916e-05,
"loss": 5.6549,
"step": 420
},
{
"epoch": 0.001074556745342546,
"grad_norm": 6.15625,
"learning_rate": 1.9914000000000003e-05,
"loss": 5.4906,
"step": 430
},
{
"epoch": 0.0010995464370946984,
"grad_norm": 8.875,
"learning_rate": 1.9912000000000002e-05,
"loss": 5.6258,
"step": 440
},
{
"epoch": 0.0011245361288468507,
"grad_norm": 7.78125,
"learning_rate": 1.9910000000000004e-05,
"loss": 5.5201,
"step": 450
},
{
"epoch": 0.001149525820599003,
"grad_norm": 7.0,
"learning_rate": 1.9908e-05,
"loss": 5.6204,
"step": 460
},
{
"epoch": 0.001174515512351155,
"grad_norm": 9.125,
"learning_rate": 1.9906000000000002e-05,
"loss": 5.5345,
"step": 470
},
{
"epoch": 0.0011995052041033074,
"grad_norm": 6.25,
"learning_rate": 1.9904e-05,
"loss": 5.5461,
"step": 480
},
{
"epoch": 0.0012244948958554597,
"grad_norm": 6.84375,
"learning_rate": 1.9902e-05,
"loss": 5.6241,
"step": 490
},
{
"epoch": 0.0012494845876076118,
"grad_norm": 7.34375,
"learning_rate": 1.9900000000000003e-05,
"loss": 5.5851,
"step": 500
},
{
"epoch": 0.001274474279359764,
"grad_norm": 6.25,
"learning_rate": 1.9898e-05,
"loss": 5.415,
"step": 510
},
{
"epoch": 0.0012994639711119164,
"grad_norm": 6.46875,
"learning_rate": 1.9896e-05,
"loss": 5.4773,
"step": 520
},
{
"epoch": 0.0013244536628640685,
"grad_norm": 7.15625,
"learning_rate": 1.9894000000000003e-05,
"loss": 5.6274,
"step": 530
},
{
"epoch": 0.0013494433546162208,
"grad_norm": 6.875,
"learning_rate": 1.9892000000000002e-05,
"loss": 5.5787,
"step": 540
},
{
"epoch": 0.001374433046368373,
"grad_norm": 7.15625,
"learning_rate": 1.989e-05,
"loss": 5.5155,
"step": 550
},
{
"epoch": 0.0013994227381205252,
"grad_norm": 7.875,
"learning_rate": 1.9888e-05,
"loss": 5.6302,
"step": 560
},
{
"epoch": 0.0014244124298726775,
"grad_norm": 6.90625,
"learning_rate": 1.9886000000000002e-05,
"loss": 5.6188,
"step": 570
},
{
"epoch": 0.0014494021216248298,
"grad_norm": 7.0625,
"learning_rate": 1.9884e-05,
"loss": 5.6807,
"step": 580
},
{
"epoch": 0.001474391813376982,
"grad_norm": 6.625,
"learning_rate": 1.9882e-05,
"loss": 5.5243,
"step": 590
},
{
"epoch": 0.0014993815051291342,
"grad_norm": 7.4375,
"learning_rate": 1.9880000000000003e-05,
"loss": 5.6199,
"step": 600
},
{
"epoch": 0.0015243711968812865,
"grad_norm": 7.96875,
"learning_rate": 1.9878000000000002e-05,
"loss": 5.5931,
"step": 610
},
{
"epoch": 0.0015493608886334388,
"grad_norm": 7.75,
"learning_rate": 1.9876e-05,
"loss": 5.4432,
"step": 620
},
{
"epoch": 0.001574350580385591,
"grad_norm": 7.5625,
"learning_rate": 1.9874000000000003e-05,
"loss": 5.5082,
"step": 630
},
{
"epoch": 0.0015993402721377432,
"grad_norm": 9.0625,
"learning_rate": 1.9872000000000002e-05,
"loss": 5.5388,
"step": 640
},
{
"epoch": 0.0016243299638898955,
"grad_norm": 6.40625,
"learning_rate": 1.987e-05,
"loss": 5.6473,
"step": 650
},
{
"epoch": 0.0016493196556420476,
"grad_norm": 7.375,
"learning_rate": 1.9868e-05,
"loss": 5.5754,
"step": 660
},
{
"epoch": 0.0016743093473942,
"grad_norm": 6.71875,
"learning_rate": 1.9866e-05,
"loss": 5.6752,
"step": 670
},
{
"epoch": 0.0016992990391463522,
"grad_norm": 7.53125,
"learning_rate": 1.9864e-05,
"loss": 5.6575,
"step": 680
},
{
"epoch": 0.0017242887308985043,
"grad_norm": 7.5,
"learning_rate": 1.9862e-05,
"loss": 5.4459,
"step": 690
},
{
"epoch": 0.0017492784226506566,
"grad_norm": 8.8125,
"learning_rate": 1.9860000000000003e-05,
"loss": 5.5611,
"step": 700
},
{
"epoch": 0.001774268114402809,
"grad_norm": 7.46875,
"learning_rate": 1.9858000000000002e-05,
"loss": 5.532,
"step": 710
},
{
"epoch": 0.001799257806154961,
"grad_norm": 7.90625,
"learning_rate": 1.9856e-05,
"loss": 5.5263,
"step": 720
},
{
"epoch": 0.0018242474979071133,
"grad_norm": 7.375,
"learning_rate": 1.9854000000000003e-05,
"loss": 5.6173,
"step": 730
},
{
"epoch": 0.0018492371896592656,
"grad_norm": 5.34375,
"learning_rate": 1.9852000000000002e-05,
"loss": 5.4807,
"step": 740
},
{
"epoch": 0.0018742268814114177,
"grad_norm": 6.5625,
"learning_rate": 1.985e-05,
"loss": 5.5658,
"step": 750
},
{
"epoch": 0.00189921657316357,
"grad_norm": 7.09375,
"learning_rate": 1.9848e-05,
"loss": 5.4922,
"step": 760
},
{
"epoch": 0.0019242062649157223,
"grad_norm": 6.875,
"learning_rate": 1.9846e-05,
"loss": 5.5456,
"step": 770
},
{
"epoch": 0.0019491959566678744,
"grad_norm": 6.5625,
"learning_rate": 1.9844000000000002e-05,
"loss": 5.5476,
"step": 780
},
{
"epoch": 0.0019741856484200267,
"grad_norm": 6.09375,
"learning_rate": 1.9842e-05,
"loss": 5.5522,
"step": 790
},
{
"epoch": 0.001999175340172179,
"grad_norm": 7.375,
"learning_rate": 1.9840000000000003e-05,
"loss": 5.6347,
"step": 800
},
{
"epoch": 0.0020241650319243313,
"grad_norm": 8.5,
"learning_rate": 1.9838000000000002e-05,
"loss": 5.5714,
"step": 810
},
{
"epoch": 0.0020491547236764834,
"grad_norm": 7.125,
"learning_rate": 1.9836e-05,
"loss": 5.5227,
"step": 820
},
{
"epoch": 0.0020741444154286355,
"grad_norm": 8.125,
"learning_rate": 1.9834000000000004e-05,
"loss": 5.5476,
"step": 830
},
{
"epoch": 0.002099134107180788,
"grad_norm": 5.96875,
"learning_rate": 1.9832000000000003e-05,
"loss": 5.5645,
"step": 840
},
{
"epoch": 0.00212412379893294,
"grad_norm": 5.84375,
"learning_rate": 1.983e-05,
"loss": 5.5257,
"step": 850
},
{
"epoch": 0.002149113490685092,
"grad_norm": 7.5,
"learning_rate": 1.9828e-05,
"loss": 5.5988,
"step": 860
},
{
"epoch": 0.0021741031824372447,
"grad_norm": 7.625,
"learning_rate": 1.9826e-05,
"loss": 5.466,
"step": 870
},
{
"epoch": 0.002199092874189397,
"grad_norm": 6.0625,
"learning_rate": 1.9824000000000002e-05,
"loss": 5.5425,
"step": 880
},
{
"epoch": 0.0022240825659415493,
"grad_norm": 5.71875,
"learning_rate": 1.9822e-05,
"loss": 5.5708,
"step": 890
},
{
"epoch": 0.0022490722576937014,
"grad_norm": 7.21875,
"learning_rate": 1.982e-05,
"loss": 5.6318,
"step": 900
},
{
"epoch": 0.0022740619494458535,
"grad_norm": 7.25,
"learning_rate": 1.9818000000000002e-05,
"loss": 5.5574,
"step": 910
},
{
"epoch": 0.002299051641198006,
"grad_norm": 7.3125,
"learning_rate": 1.9816e-05,
"loss": 5.4384,
"step": 920
},
{
"epoch": 0.002324041332950158,
"grad_norm": 6.4375,
"learning_rate": 1.9814000000000004e-05,
"loss": 5.5454,
"step": 930
},
{
"epoch": 0.00234903102470231,
"grad_norm": 7.0625,
"learning_rate": 1.9812000000000003e-05,
"loss": 5.6111,
"step": 940
},
{
"epoch": 0.0023740207164544627,
"grad_norm": 6.84375,
"learning_rate": 1.9810000000000002e-05,
"loss": 5.4391,
"step": 950
},
{
"epoch": 0.002399010408206615,
"grad_norm": 7.875,
"learning_rate": 1.9808e-05,
"loss": 5.5132,
"step": 960
},
{
"epoch": 0.002424000099958767,
"grad_norm": 6.53125,
"learning_rate": 1.9806e-05,
"loss": 5.5817,
"step": 970
},
{
"epoch": 0.0024489897917109194,
"grad_norm": 6.90625,
"learning_rate": 1.9804000000000002e-05,
"loss": 5.5528,
"step": 980
},
{
"epoch": 0.0024739794834630715,
"grad_norm": 5.6875,
"learning_rate": 1.9802e-05,
"loss": 5.6482,
"step": 990
},
{
"epoch": 0.0024989691752152236,
"grad_norm": 6.84375,
"learning_rate": 1.98e-05,
"loss": 5.6296,
"step": 1000
},
{
"epoch": 0.000315494858370922,
"grad_norm": 14.75,
"learning_rate": 1.9798000000000003e-05,
"loss": 5.4533,
"step": 1010
},
{
"epoch": 0.000318618569839941,
"grad_norm": 11.5,
"learning_rate": 1.9796e-05,
"loss": 5.5744,
"step": 1020
},
{
"epoch": 0.00032174228130896004,
"grad_norm": 19.625,
"learning_rate": 1.9794e-05,
"loss": 5.3765,
"step": 1030
},
{
"epoch": 0.0003248659927779791,
"grad_norm": 15.375,
"learning_rate": 1.9792000000000003e-05,
"loss": 5.7717,
"step": 1040
},
{
"epoch": 0.0003279897042469981,
"grad_norm": 14.6875,
"learning_rate": 1.9790000000000002e-05,
"loss": 5.418,
"step": 1050
},
{
"epoch": 0.0003311134157160171,
"grad_norm": 19.625,
"learning_rate": 1.9788e-05,
"loss": 5.3012,
"step": 1060
},
{
"epoch": 0.0003342371271850362,
"grad_norm": 18.0,
"learning_rate": 1.9786e-05,
"loss": 5.665,
"step": 1070
},
{
"epoch": 0.0003373608386540552,
"grad_norm": 15.8125,
"learning_rate": 1.9784000000000002e-05,
"loss": 5.6366,
"step": 1080
},
{
"epoch": 0.0003404845501230742,
"grad_norm": 18.375,
"learning_rate": 1.9782e-05,
"loss": 5.2612,
"step": 1090
},
{
"epoch": 0.0003436082615920933,
"grad_norm": 13.125,
"learning_rate": 1.978e-05,
"loss": 5.3365,
"step": 1100
},
{
"epoch": 0.0003467319730611123,
"grad_norm": 11.75,
"learning_rate": 1.9778000000000003e-05,
"loss": 4.9134,
"step": 1110
},
{
"epoch": 0.0003498556845301313,
"grad_norm": 20.875,
"learning_rate": 1.9776000000000002e-05,
"loss": 5.5621,
"step": 1120
},
{
"epoch": 0.00035297939599915036,
"grad_norm": 14.5625,
"learning_rate": 1.9774e-05,
"loss": 5.4201,
"step": 1130
},
{
"epoch": 0.0003561031074681694,
"grad_norm": 26.25,
"learning_rate": 1.9772000000000003e-05,
"loss": 5.4606,
"step": 1140
},
{
"epoch": 0.0003592268189371884,
"grad_norm": 21.75,
"learning_rate": 1.9770000000000002e-05,
"loss": 5.6511,
"step": 1150
},
{
"epoch": 0.00036235053040620745,
"grad_norm": 15.8125,
"learning_rate": 1.9768e-05,
"loss": 5.1869,
"step": 1160
},
{
"epoch": 0.00036547424187522646,
"grad_norm": 15.875,
"learning_rate": 1.9766e-05,
"loss": 5.5171,
"step": 1170
},
{
"epoch": 0.0003685979533442455,
"grad_norm": 14.25,
"learning_rate": 1.9764000000000003e-05,
"loss": 5.5247,
"step": 1180
},
{
"epoch": 0.00037172166481326454,
"grad_norm": 14.0625,
"learning_rate": 1.9762e-05,
"loss": 5.86,
"step": 1190
},
{
"epoch": 0.00037484537628228355,
"grad_norm": 17.0,
"learning_rate": 1.976e-05,
"loss": 5.6677,
"step": 1200
},
{
"epoch": 0.00037796908775130256,
"grad_norm": 14.125,
"learning_rate": 1.9758000000000003e-05,
"loss": 5.7599,
"step": 1210
},
{
"epoch": 0.0003810927992203216,
"grad_norm": 18.875,
"learning_rate": 1.9756000000000002e-05,
"loss": 5.6479,
"step": 1220
},
{
"epoch": 0.00038421651068934064,
"grad_norm": 17.25,
"learning_rate": 1.9754e-05,
"loss": 5.7051,
"step": 1230
},
{
"epoch": 0.0003873402221583597,
"grad_norm": 13.0625,
"learning_rate": 1.9752000000000003e-05,
"loss": 5.413,
"step": 1240
},
{
"epoch": 0.0003904639336273787,
"grad_norm": 15.75,
"learning_rate": 1.9750000000000002e-05,
"loss": 5.4345,
"step": 1250
},
{
"epoch": 0.0003935876450963977,
"grad_norm": 20.25,
"learning_rate": 1.9748e-05,
"loss": 5.7715,
"step": 1260
},
{
"epoch": 0.0003967113565654168,
"grad_norm": 22.875,
"learning_rate": 1.9746e-05,
"loss": 5.6046,
"step": 1270
},
{
"epoch": 0.0003998350680344358,
"grad_norm": 19.375,
"learning_rate": 1.9744e-05,
"loss": 5.6705,
"step": 1280
},
{
"epoch": 0.0004029587795034548,
"grad_norm": 13.1875,
"learning_rate": 1.9742000000000002e-05,
"loss": 5.7126,
"step": 1290
},
{
"epoch": 0.0004060824909724739,
"grad_norm": 11.25,
"learning_rate": 1.974e-05,
"loss": 5.2286,
"step": 1300
},
{
"epoch": 0.0004092062024414929,
"grad_norm": 17.125,
"learning_rate": 1.9738000000000003e-05,
"loss": 5.4317,
"step": 1310
},
{
"epoch": 0.0004123299139105119,
"grad_norm": 13.4375,
"learning_rate": 1.9736000000000002e-05,
"loss": 5.7747,
"step": 1320
},
{
"epoch": 0.00041545362537953097,
"grad_norm": 17.25,
"learning_rate": 1.9734e-05,
"loss": 5.6014,
"step": 1330
},
{
"epoch": 0.00041857733684855,
"grad_norm": 23.25,
"learning_rate": 1.9732000000000004e-05,
"loss": 5.6342,
"step": 1340
},
{
"epoch": 0.000421701048317569,
"grad_norm": 17.75,
"learning_rate": 1.9730000000000003e-05,
"loss": 5.714,
"step": 1350
},
{
"epoch": 0.00042482475978658805,
"grad_norm": 18.75,
"learning_rate": 1.9728e-05,
"loss": 5.7005,
"step": 1360
},
{
"epoch": 0.00042794847125560706,
"grad_norm": 14.0625,
"learning_rate": 1.9726e-05,
"loss": 5.773,
"step": 1370
},
{
"epoch": 0.0004310721827246261,
"grad_norm": 21.125,
"learning_rate": 1.9724e-05,
"loss": 5.6311,
"step": 1380
},
{
"epoch": 0.00043419589419364514,
"grad_norm": 17.375,
"learning_rate": 1.9722000000000002e-05,
"loss": 5.6152,
"step": 1390
},
{
"epoch": 0.00043731960566266415,
"grad_norm": 27.25,
"learning_rate": 1.972e-05,
"loss": 5.442,
"step": 1400
},
{
"epoch": 0.00044044331713168316,
"grad_norm": 14.375,
"learning_rate": 1.9718e-05,
"loss": 5.4459,
"step": 1410
},
{
"epoch": 0.00044356702860070223,
"grad_norm": 15.6875,
"learning_rate": 1.9716000000000002e-05,
"loss": 5.4062,
"step": 1420
},
{
"epoch": 0.00044669074006972124,
"grad_norm": 21.25,
"learning_rate": 1.9714e-05,
"loss": 5.7603,
"step": 1430
},
{
"epoch": 0.00044981445153874025,
"grad_norm": 18.125,
"learning_rate": 1.9712000000000004e-05,
"loss": 5.5305,
"step": 1440
},
{
"epoch": 0.0004529381630077593,
"grad_norm": 21.375,
"learning_rate": 1.9710000000000003e-05,
"loss": 5.5689,
"step": 1450
},
{
"epoch": 0.0004560618744767783,
"grad_norm": 20.625,
"learning_rate": 1.9708000000000002e-05,
"loss": 5.2929,
"step": 1460
},
{
"epoch": 0.00045918558594579734,
"grad_norm": 14.75,
"learning_rate": 1.9706e-05,
"loss": 5.7608,
"step": 1470
},
{
"epoch": 0.0004623092974148164,
"grad_norm": 14.25,
"learning_rate": 1.9704e-05,
"loss": 5.5301,
"step": 1480
},
{
"epoch": 0.0004654330088838354,
"grad_norm": 15.4375,
"learning_rate": 1.9702000000000002e-05,
"loss": 5.5514,
"step": 1490
},
{
"epoch": 0.0004685567203528544,
"grad_norm": 16.75,
"learning_rate": 1.97e-05,
"loss": 5.4936,
"step": 1500
},
{
"epoch": 0.0004716804318218735,
"grad_norm": 18.0,
"learning_rate": 1.9698e-05,
"loss": 5.5664,
"step": 1510
},
{
"epoch": 0.0004748041432908925,
"grad_norm": 24.125,
"learning_rate": 1.9696000000000003e-05,
"loss": 5.2128,
"step": 1520
},
{
"epoch": 0.0004779278547599115,
"grad_norm": 19.375,
"learning_rate": 1.9694e-05,
"loss": 5.6116,
"step": 1530
},
{
"epoch": 0.0004810515662289306,
"grad_norm": 13.375,
"learning_rate": 1.9692000000000004e-05,
"loss": 5.65,
"step": 1540
},
{
"epoch": 0.0004841752776979496,
"grad_norm": 20.0,
"learning_rate": 1.9690000000000003e-05,
"loss": 5.6937,
"step": 1550
},
{
"epoch": 0.0004872989891669686,
"grad_norm": 23.375,
"learning_rate": 1.9688000000000002e-05,
"loss": 5.3234,
"step": 1560
},
{
"epoch": 0.0004904227006359876,
"grad_norm": 17.375,
"learning_rate": 1.9686e-05,
"loss": 5.5981,
"step": 1570
},
{
"epoch": 0.0004935464121050067,
"grad_norm": 19.5,
"learning_rate": 1.9684e-05,
"loss": 4.9602,
"step": 1580
},
{
"epoch": 0.0004966701235740257,
"grad_norm": 17.25,
"learning_rate": 1.9682000000000002e-05,
"loss": 5.3535,
"step": 1590
},
{
"epoch": 0.0004997938350430447,
"grad_norm": 17.25,
"learning_rate": 1.968e-05,
"loss": 5.1287,
"step": 1600
},
{
"epoch": 0.0005029175465120638,
"grad_norm": 16.75,
"learning_rate": 1.9678e-05,
"loss": 5.6272,
"step": 1610
},
{
"epoch": 0.0005060412579810828,
"grad_norm": 17.75,
"learning_rate": 1.9676000000000003e-05,
"loss": 5.432,
"step": 1620
},
{
"epoch": 0.0005091649694501018,
"grad_norm": 11.8125,
"learning_rate": 1.9674000000000002e-05,
"loss": 5.3767,
"step": 1630
},
{
"epoch": 0.0005122886809191209,
"grad_norm": 19.25,
"learning_rate": 1.9672e-05,
"loss": 5.3967,
"step": 1640
},
{
"epoch": 0.0005154123923881399,
"grad_norm": 14.9375,
"learning_rate": 1.9670000000000003e-05,
"loss": 5.2857,
"step": 1650
},
{
"epoch": 0.0005185361038571589,
"grad_norm": 16.0,
"learning_rate": 1.9668000000000002e-05,
"loss": 5.6774,
"step": 1660
},
{
"epoch": 0.0005216598153261779,
"grad_norm": 21.5,
"learning_rate": 1.9666e-05,
"loss": 5.6296,
"step": 1670
},
{
"epoch": 0.000524783526795197,
"grad_norm": 17.625,
"learning_rate": 1.9664e-05,
"loss": 5.4919,
"step": 1680
},
{
"epoch": 0.000527907238264216,
"grad_norm": 15.5625,
"learning_rate": 1.9662000000000003e-05,
"loss": 5.7115,
"step": 1690
},
{
"epoch": 0.000531030949733235,
"grad_norm": 15.25,
"learning_rate": 1.966e-05,
"loss": 5.7642,
"step": 1700
},
{
"epoch": 0.0005341546612022541,
"grad_norm": 19.0,
"learning_rate": 1.9658e-05,
"loss": 5.444,
"step": 1710
},
{
"epoch": 0.000537278372671273,
"grad_norm": 16.375,
"learning_rate": 1.9656000000000003e-05,
"loss": 5.4625,
"step": 1720
},
{
"epoch": 0.0005404020841402921,
"grad_norm": 15.0,
"learning_rate": 1.9654000000000002e-05,
"loss": 5.7821,
"step": 1730
},
{
"epoch": 0.0005435257956093112,
"grad_norm": 13.25,
"learning_rate": 1.9652e-05,
"loss": 5.6565,
"step": 1740
},
{
"epoch": 0.0005466495070783301,
"grad_norm": 17.25,
"learning_rate": 1.9650000000000003e-05,
"loss": 5.4938,
"step": 1750
},
{
"epoch": 0.0005497732185473492,
"grad_norm": 17.25,
"learning_rate": 1.9648000000000002e-05,
"loss": 5.4783,
"step": 1760
},
{
"epoch": 0.0005528969300163683,
"grad_norm": 12.875,
"learning_rate": 1.9646e-05,
"loss": 5.5006,
"step": 1770
},
{
"epoch": 0.0005560206414853873,
"grad_norm": 20.375,
"learning_rate": 1.9644e-05,
"loss": 5.3737,
"step": 1780
},
{
"epoch": 0.0005591443529544063,
"grad_norm": 18.875,
"learning_rate": 1.9642000000000003e-05,
"loss": 5.5817,
"step": 1790
},
{
"epoch": 0.0005622680644234254,
"grad_norm": 19.375,
"learning_rate": 1.9640000000000002e-05,
"loss": 5.2545,
"step": 1800
},
{
"epoch": 0.0005653917758924444,
"grad_norm": 12.125,
"learning_rate": 1.9638e-05,
"loss": 5.5962,
"step": 1810
},
{
"epoch": 0.0005685154873614634,
"grad_norm": 18.25,
"learning_rate": 1.9636000000000003e-05,
"loss": 5.3886,
"step": 1820
},
{
"epoch": 0.0005716391988304824,
"grad_norm": 20.125,
"learning_rate": 1.9634000000000002e-05,
"loss": 5.4536,
"step": 1830
},
{
"epoch": 0.0005747629102995015,
"grad_norm": 15.1875,
"learning_rate": 1.9632e-05,
"loss": 5.539,
"step": 1840
},
{
"epoch": 0.0005778866217685205,
"grad_norm": 15.8125,
"learning_rate": 1.9630000000000003e-05,
"loss": 5.5931,
"step": 1850
},
{
"epoch": 0.0005810103332375395,
"grad_norm": 13.6875,
"learning_rate": 1.9628000000000002e-05,
"loss": 5.544,
"step": 1860
},
{
"epoch": 0.0005841340447065586,
"grad_norm": 15.875,
"learning_rate": 1.9626e-05,
"loss": 5.344,
"step": 1870
},
{
"epoch": 0.0005872577561755775,
"grad_norm": 12.75,
"learning_rate": 1.9624e-05,
"loss": 5.4029,
"step": 1880
},
{
"epoch": 0.0005903814676445966,
"grad_norm": 12.9375,
"learning_rate": 1.9622e-05,
"loss": 5.7129,
"step": 1890
},
{
"epoch": 0.0005935051791136157,
"grad_norm": 17.625,
"learning_rate": 1.9620000000000002e-05,
"loss": 5.5131,
"step": 1900
},
{
"epoch": 0.0005966288905826346,
"grad_norm": 15.3125,
"learning_rate": 1.9618e-05,
"loss": 5.4311,
"step": 1910
},
{
"epoch": 0.0005997526020516537,
"grad_norm": 18.5,
"learning_rate": 1.9616000000000003e-05,
"loss": 5.6053,
"step": 1920
},
{
"epoch": 0.0006028763135206728,
"grad_norm": 15.8125,
"learning_rate": 1.9614000000000002e-05,
"loss": 5.6931,
"step": 1930
},
{
"epoch": 0.0006060000249896917,
"grad_norm": 20.0,
"learning_rate": 1.9612e-05,
"loss": 5.5993,
"step": 1940
},
{
"epoch": 0.0006091237364587108,
"grad_norm": 17.625,
"learning_rate": 1.9610000000000004e-05,
"loss": 5.5418,
"step": 1950
},
{
"epoch": 0.0006122474479277299,
"grad_norm": 14.4375,
"learning_rate": 1.9608000000000003e-05,
"loss": 5.4671,
"step": 1960
},
{
"epoch": 0.0006153711593967488,
"grad_norm": 15.1875,
"learning_rate": 1.9606e-05,
"loss": 5.5768,
"step": 1970
},
{
"epoch": 0.0006184948708657679,
"grad_norm": 13.0625,
"learning_rate": 1.9604e-05,
"loss": 5.6612,
"step": 1980
},
{
"epoch": 0.0006216185823347869,
"grad_norm": 11.625,
"learning_rate": 1.9602e-05,
"loss": 5.4864,
"step": 1990
},
{
"epoch": 0.0006247422938038059,
"grad_norm": 20.875,
"learning_rate": 1.9600000000000002e-05,
"loss": 5.6614,
"step": 2000
},
{
"epoch": 0.000627866005272825,
"grad_norm": 17.25,
"learning_rate": 1.9598e-05,
"loss": 5.6901,
"step": 2010
},
{
"epoch": 0.000630989716741844,
"grad_norm": 17.375,
"learning_rate": 1.9596e-05,
"loss": 5.7166,
"step": 2020
},
{
"epoch": 0.000634113428210863,
"grad_norm": 19.25,
"learning_rate": 1.9594000000000002e-05,
"loss": 5.7282,
"step": 2030
},
{
"epoch": 0.000637237139679882,
"grad_norm": 18.5,
"learning_rate": 1.9592e-05,
"loss": 5.6836,
"step": 2040
},
{
"epoch": 0.0006403608511489011,
"grad_norm": 19.625,
"learning_rate": 1.9590000000000004e-05,
"loss": 5.6814,
"step": 2050
},
{
"epoch": 0.0006434845626179201,
"grad_norm": 17.625,
"learning_rate": 1.9588000000000003e-05,
"loss": 5.4854,
"step": 2060
},
{
"epoch": 0.0006466082740869391,
"grad_norm": 13.8125,
"learning_rate": 1.9586000000000002e-05,
"loss": 5.7422,
"step": 2070
},
{
"epoch": 0.0006497319855559582,
"grad_norm": 19.125,
"learning_rate": 1.9584e-05,
"loss": 5.2697,
"step": 2080
},
{
"epoch": 0.0006528556970249772,
"grad_norm": 17.125,
"learning_rate": 1.9582e-05,
"loss": 5.4215,
"step": 2090
},
{
"epoch": 0.0006559794084939962,
"grad_norm": 17.375,
"learning_rate": 1.9580000000000002e-05,
"loss": 5.4331,
"step": 2100
},
{
"epoch": 0.0006591031199630153,
"grad_norm": 15.5625,
"learning_rate": 1.9578e-05,
"loss": 5.6154,
"step": 2110
},
{
"epoch": 0.0006622268314320342,
"grad_norm": 22.625,
"learning_rate": 1.9576e-05,
"loss": 5.1004,
"step": 2120
},
{
"epoch": 0.0006653505429010533,
"grad_norm": 16.75,
"learning_rate": 1.9574000000000003e-05,
"loss": 5.5134,
"step": 2130
},
{
"epoch": 0.0006684742543700724,
"grad_norm": 14.875,
"learning_rate": 1.9572e-05,
"loss": 5.5428,
"step": 2140
},
{
"epoch": 0.0006715979658390913,
"grad_norm": 23.0,
"learning_rate": 1.957e-05,
"loss": 5.4856,
"step": 2150
},
{
"epoch": 0.0006747216773081104,
"grad_norm": 14.5,
"learning_rate": 1.9568000000000003e-05,
"loss": 5.7353,
"step": 2160
},
{
"epoch": 0.0006778453887771295,
"grad_norm": 13.875,
"learning_rate": 1.9566000000000002e-05,
"loss": 5.4402,
"step": 2170
},
{
"epoch": 0.0006809691002461484,
"grad_norm": 21.75,
"learning_rate": 1.9564e-05,
"loss": 5.842,
"step": 2180
},
{
"epoch": 0.0006840928117151675,
"grad_norm": 15.9375,
"learning_rate": 1.9562e-05,
"loss": 5.4636,
"step": 2190
},
{
"epoch": 0.0006872165231841866,
"grad_norm": 16.125,
"learning_rate": 1.9560000000000002e-05,
"loss": 5.3003,
"step": 2200
},
{
"epoch": 0.0006903402346532055,
"grad_norm": 18.75,
"learning_rate": 1.9558e-05,
"loss": 5.6459,
"step": 2210
},
{
"epoch": 0.0006934639461222246,
"grad_norm": 11.875,
"learning_rate": 1.9556e-05,
"loss": 5.5871,
"step": 2220
},
{
"epoch": 0.0006965876575912436,
"grad_norm": 20.375,
"learning_rate": 1.9554000000000003e-05,
"loss": 5.5089,
"step": 2230
},
{
"epoch": 0.0006997113690602626,
"grad_norm": 16.375,
"learning_rate": 1.9552000000000002e-05,
"loss": 5.7504,
"step": 2240
},
{
"epoch": 0.0007028350805292817,
"grad_norm": 28.375,
"learning_rate": 1.955e-05,
"loss": 5.5179,
"step": 2250
},
{
"epoch": 0.0007059587919983007,
"grad_norm": 13.6875,
"learning_rate": 1.9548000000000003e-05,
"loss": 5.4784,
"step": 2260
},
{
"epoch": 0.0007090825034673197,
"grad_norm": 14.375,
"learning_rate": 1.9546e-05,
"loss": 5.6661,
"step": 2270
},
{
"epoch": 0.0007122062149363388,
"grad_norm": 13.75,
"learning_rate": 1.9544e-05,
"loss": 5.4985,
"step": 2280
},
{
"epoch": 0.0007153299264053578,
"grad_norm": 16.0,
"learning_rate": 1.9542e-05,
"loss": 5.8197,
"step": 2290
},
{
"epoch": 0.0007184536378743768,
"grad_norm": 19.125,
"learning_rate": 1.9540000000000003e-05,
"loss": 5.3348,
"step": 2300
},
{
"epoch": 0.0007215773493433958,
"grad_norm": 13.5,
"learning_rate": 1.9538e-05,
"loss": 5.6816,
"step": 2310
},
{
"epoch": 0.0007247010608124149,
"grad_norm": 15.4375,
"learning_rate": 1.9536e-05,
"loss": 5.7536,
"step": 2320
},
{
"epoch": 0.0007278247722814339,
"grad_norm": 19.625,
"learning_rate": 1.9534000000000003e-05,
"loss": 5.7437,
"step": 2330
},
{
"epoch": 0.0007309484837504529,
"grad_norm": 16.875,
"learning_rate": 1.9532000000000002e-05,
"loss": 5.5242,
"step": 2340
},
{
"epoch": 0.000734072195219472,
"grad_norm": 15.6875,
"learning_rate": 1.953e-05,
"loss": 5.7513,
"step": 2350
},
{
"epoch": 0.000737195906688491,
"grad_norm": 12.5625,
"learning_rate": 1.9528000000000003e-05,
"loss": 5.7211,
"step": 2360
},
{
"epoch": 0.00074031961815751,
"grad_norm": 16.875,
"learning_rate": 1.9526e-05,
"loss": 5.7414,
"step": 2370
},
{
"epoch": 0.0007434433296265291,
"grad_norm": 23.25,
"learning_rate": 1.9524e-05,
"loss": 5.4911,
"step": 2380
},
{
"epoch": 0.000746567041095548,
"grad_norm": 14.5625,
"learning_rate": 1.9522e-05,
"loss": 5.7355,
"step": 2390
},
{
"epoch": 0.0007496907525645671,
"grad_norm": 20.75,
"learning_rate": 1.9520000000000003e-05,
"loss": 5.5423,
"step": 2400
},
{
"epoch": 0.0007528144640335862,
"grad_norm": 16.125,
"learning_rate": 1.9518000000000002e-05,
"loss": 5.4726,
"step": 2410
},
{
"epoch": 0.0007559381755026051,
"grad_norm": 19.0,
"learning_rate": 1.9516e-05,
"loss": 5.7674,
"step": 2420
},
{
"epoch": 0.0007590618869716242,
"grad_norm": 13.25,
"learning_rate": 1.9514000000000003e-05,
"loss": 5.65,
"step": 2430
},
{
"epoch": 0.0007621855984406433,
"grad_norm": 14.0,
"learning_rate": 1.9512000000000002e-05,
"loss": 5.777,
"step": 2440
},
{
"epoch": 0.0007653093099096622,
"grad_norm": 17.125,
"learning_rate": 1.951e-05,
"loss": 5.2448,
"step": 2450
},
{
"epoch": 0.0007684330213786813,
"grad_norm": 11.1875,
"learning_rate": 1.9508000000000004e-05,
"loss": 5.6465,
"step": 2460
},
{
"epoch": 0.0007715567328477003,
"grad_norm": 15.3125,
"learning_rate": 1.9506e-05,
"loss": 5.6983,
"step": 2470
},
{
"epoch": 0.0007746804443167194,
"grad_norm": 12.1875,
"learning_rate": 1.9504e-05,
"loss": 5.2056,
"step": 2480
},
{
"epoch": 0.0007778041557857384,
"grad_norm": 22.125,
"learning_rate": 1.9502e-05,
"loss": 5.675,
"step": 2490
},
{
"epoch": 0.0007809278672547574,
"grad_norm": 20.25,
"learning_rate": 1.95e-05,
"loss": 5.6855,
"step": 2500
},
{
"epoch": 0.0007840515787237765,
"grad_norm": 12.25,
"learning_rate": 1.9498000000000002e-05,
"loss": 5.7423,
"step": 2510
},
{
"epoch": 0.0007871752901927955,
"grad_norm": 15.5,
"learning_rate": 1.9496e-05,
"loss": 5.4792,
"step": 2520
},
{
"epoch": 0.0007902990016618145,
"grad_norm": 17.375,
"learning_rate": 1.9494000000000003e-05,
"loss": 5.562,
"step": 2530
},
{
"epoch": 0.0007934227131308336,
"grad_norm": 14.1875,
"learning_rate": 1.9492000000000002e-05,
"loss": 5.6019,
"step": 2540
},
{
"epoch": 0.0007965464245998525,
"grad_norm": 17.625,
"learning_rate": 1.949e-05,
"loss": 5.6609,
"step": 2550
},
{
"epoch": 0.0007996701360688716,
"grad_norm": 18.75,
"learning_rate": 1.9488000000000004e-05,
"loss": 5.374,
"step": 2560
},
{
"epoch": 0.0008027938475378907,
"grad_norm": 22.25,
"learning_rate": 1.9486000000000003e-05,
"loss": 5.4583,
"step": 2570
},
{
"epoch": 0.0008059175590069096,
"grad_norm": 13.8125,
"learning_rate": 1.9484000000000002e-05,
"loss": 5.2723,
"step": 2580
},
{
"epoch": 0.0008090412704759287,
"grad_norm": 12.5625,
"learning_rate": 1.9482e-05,
"loss": 5.328,
"step": 2590
},
{
"epoch": 0.0008121649819449478,
"grad_norm": 12.375,
"learning_rate": 1.948e-05,
"loss": 5.4062,
"step": 2600
},
{
"epoch": 0.0008152886934139667,
"grad_norm": 13.125,
"learning_rate": 1.9478000000000002e-05,
"loss": 5.7013,
"step": 2610
},
{
"epoch": 0.0008184124048829858,
"grad_norm": 17.625,
"learning_rate": 1.9476e-05,
"loss": 5.5071,
"step": 2620
},
{
"epoch": 0.0008215361163520048,
"grad_norm": 14.4375,
"learning_rate": 1.9474e-05,
"loss": 5.6346,
"step": 2630
},
{
"epoch": 0.0008246598278210238,
"grad_norm": 17.0,
"learning_rate": 1.9472000000000003e-05,
"loss": 5.1499,
"step": 2640
},
{
"epoch": 0.0008277835392900429,
"grad_norm": 13.5,
"learning_rate": 1.947e-05,
"loss": 5.547,
"step": 2650
},
{
"epoch": 0.0008309072507590619,
"grad_norm": 16.0,
"learning_rate": 1.9468000000000004e-05,
"loss": 5.7553,
"step": 2660
},
{
"epoch": 0.0008340309622280809,
"grad_norm": 19.875,
"learning_rate": 1.9466000000000003e-05,
"loss": 5.4614,
"step": 2670
},
{
"epoch": 0.0008371546736971,
"grad_norm": 14.5625,
"learning_rate": 1.9464000000000002e-05,
"loss": 5.4983,
"step": 2680
},
{
"epoch": 0.000840278385166119,
"grad_norm": 17.75,
"learning_rate": 1.9462e-05,
"loss": 5.1406,
"step": 2690
},
{
"epoch": 0.000843402096635138,
"grad_norm": 16.625,
"learning_rate": 1.946e-05,
"loss": 5.4609,
"step": 2700
},
{
"epoch": 0.000846525808104157,
"grad_norm": 21.5,
"learning_rate": 1.9458000000000002e-05,
"loss": 5.7063,
"step": 2710
},
{
"epoch": 0.0008496495195731761,
"grad_norm": 15.625,
"learning_rate": 1.9456e-05,
"loss": 5.6036,
"step": 2720
},
{
"epoch": 0.0008527732310421951,
"grad_norm": 13.9375,
"learning_rate": 1.9454e-05,
"loss": 5.5167,
"step": 2730
},
{
"epoch": 0.0008558969425112141,
"grad_norm": 10.5,
"learning_rate": 1.9452000000000003e-05,
"loss": 5.3084,
"step": 2740
},
{
"epoch": 0.0008590206539802332,
"grad_norm": 12.5625,
"learning_rate": 1.9450000000000002e-05,
"loss": 5.2286,
"step": 2750
},
{
"epoch": 0.0008621443654492521,
"grad_norm": 17.125,
"learning_rate": 1.9448e-05,
"loss": 5.4303,
"step": 2760
},
{
"epoch": 0.0008652680769182712,
"grad_norm": 14.75,
"learning_rate": 1.9446000000000003e-05,
"loss": 5.476,
"step": 2770
},
{
"epoch": 0.0008683917883872903,
"grad_norm": 16.75,
"learning_rate": 1.9444000000000002e-05,
"loss": 5.7369,
"step": 2780
},
{
"epoch": 0.0008715154998563092,
"grad_norm": 12.3125,
"learning_rate": 1.9442e-05,
"loss": 5.7263,
"step": 2790
},
{
"epoch": 0.0008746392113253283,
"grad_norm": 20.25,
"learning_rate": 1.944e-05,
"loss": 5.1745,
"step": 2800
},
{
"epoch": 0.0008777629227943474,
"grad_norm": 11.25,
"learning_rate": 1.9438000000000003e-05,
"loss": 5.5658,
"step": 2810
},
{
"epoch": 0.0008808866342633663,
"grad_norm": 16.625,
"learning_rate": 1.9436e-05,
"loss": 5.4065,
"step": 2820
},
{
"epoch": 0.0008840103457323854,
"grad_norm": 15.5,
"learning_rate": 1.9434e-05,
"loss": 5.6371,
"step": 2830
},
{
"epoch": 0.0008871340572014045,
"grad_norm": 13.3125,
"learning_rate": 1.9432000000000003e-05,
"loss": 5.771,
"step": 2840
},
{
"epoch": 0.0008902577686704234,
"grad_norm": 16.375,
"learning_rate": 1.9430000000000002e-05,
"loss": 5.5594,
"step": 2850
},
{
"epoch": 0.0008933814801394425,
"grad_norm": 14.5625,
"learning_rate": 1.9428e-05,
"loss": 5.5727,
"step": 2860
},
{
"epoch": 0.0008965051916084615,
"grad_norm": 21.625,
"learning_rate": 1.9426000000000003e-05,
"loss": 5.6611,
"step": 2870
},
{
"epoch": 0.0008996289030774805,
"grad_norm": 21.375,
"learning_rate": 1.9424e-05,
"loss": 5.5716,
"step": 2880
},
{
"epoch": 0.0009027526145464996,
"grad_norm": 21.375,
"learning_rate": 1.9422e-05,
"loss": 5.4191,
"step": 2890
},
{
"epoch": 0.0009058763260155186,
"grad_norm": 14.375,
"learning_rate": 1.942e-05,
"loss": 5.5003,
"step": 2900
},
{
"epoch": 0.0009090000374845376,
"grad_norm": 18.375,
"learning_rate": 1.9418000000000003e-05,
"loss": 5.6882,
"step": 2910
},
{
"epoch": 0.0009121237489535567,
"grad_norm": 16.125,
"learning_rate": 1.9416000000000002e-05,
"loss": 5.7119,
"step": 2920
},
{
"epoch": 0.0009152474604225757,
"grad_norm": 15.75,
"learning_rate": 1.9414e-05,
"loss": 5.6209,
"step": 2930
},
{
"epoch": 0.0009183711718915947,
"grad_norm": 15.9375,
"learning_rate": 1.9412000000000003e-05,
"loss": 5.7167,
"step": 2940
},
{
"epoch": 0.0009214948833606137,
"grad_norm": 11.5625,
"learning_rate": 1.9410000000000002e-05,
"loss": 5.6873,
"step": 2950
},
{
"epoch": 0.0009246185948296328,
"grad_norm": 14.4375,
"learning_rate": 1.9408e-05,
"loss": 5.66,
"step": 2960
},
{
"epoch": 0.0009277423062986518,
"grad_norm": 16.375,
"learning_rate": 1.9406000000000004e-05,
"loss": 5.6787,
"step": 2970
},
{
"epoch": 0.0009308660177676708,
"grad_norm": 16.0,
"learning_rate": 1.9404e-05,
"loss": 5.014,
"step": 2980
},
{
"epoch": 0.0009339897292366899,
"grad_norm": 15.6875,
"learning_rate": 1.9402e-05,
"loss": 5.6579,
"step": 2990
},
{
"epoch": 0.0009371134407057088,
"grad_norm": 18.125,
"learning_rate": 1.94e-05,
"loss": 5.4841,
"step": 3000
},
{
"epoch": 0.0009402371521747279,
"grad_norm": 20.125,
"learning_rate": 1.9398e-05,
"loss": 5.7642,
"step": 3010
},
{
"epoch": 0.000943360863643747,
"grad_norm": 14.3125,
"learning_rate": 1.9396000000000002e-05,
"loss": 5.7511,
"step": 3020
},
{
"epoch": 0.0009464845751127659,
"grad_norm": 19.75,
"learning_rate": 1.9394e-05,
"loss": 5.6252,
"step": 3030
},
{
"epoch": 0.000949608286581785,
"grad_norm": 14.8125,
"learning_rate": 1.9392000000000003e-05,
"loss": 5.6126,
"step": 3040
},
{
"epoch": 0.0009527319980508041,
"grad_norm": 15.75,
"learning_rate": 1.9390000000000002e-05,
"loss": 5.2285,
"step": 3050
},
{
"epoch": 0.000955855709519823,
"grad_norm": 21.0,
"learning_rate": 1.9388e-05,
"loss": 5.712,
"step": 3060
},
{
"epoch": 0.0009589794209888421,
"grad_norm": 12.1875,
"learning_rate": 1.9386000000000004e-05,
"loss": 5.4932,
"step": 3070
},
{
"epoch": 0.0009621031324578612,
"grad_norm": 16.25,
"learning_rate": 1.9384e-05,
"loss": 5.6017,
"step": 3080
},
{
"epoch": 0.0009652268439268801,
"grad_norm": 14.3125,
"learning_rate": 1.9382000000000002e-05,
"loss": 5.6923,
"step": 3090
},
{
"epoch": 0.0009683505553958992,
"grad_norm": 22.0,
"learning_rate": 1.938e-05,
"loss": 5.6061,
"step": 3100
},
{
"epoch": 0.0009714742668649182,
"grad_norm": 12.375,
"learning_rate": 1.9378e-05,
"loss": 5.1761,
"step": 3110
},
{
"epoch": 0.0009745979783339372,
"grad_norm": 10.875,
"learning_rate": 1.9376000000000002e-05,
"loss": 5.6632,
"step": 3120
},
{
"epoch": 0.0009777216898029563,
"grad_norm": 18.875,
"learning_rate": 1.9374e-05,
"loss": 5.7265,
"step": 3130
},
{
"epoch": 0.0009808454012719752,
"grad_norm": 13.0,
"learning_rate": 1.9372000000000004e-05,
"loss": 5.726,
"step": 3140
},
{
"epoch": 0.0009839691127409944,
"grad_norm": 17.0,
"learning_rate": 1.9370000000000003e-05,
"loss": 5.1336,
"step": 3150
},
{
"epoch": 0.0009870928242100134,
"grad_norm": 17.25,
"learning_rate": 1.9368e-05,
"loss": 5.5581,
"step": 3160
},
{
"epoch": 0.0009902165356790323,
"grad_norm": 17.0,
"learning_rate": 1.9366000000000004e-05,
"loss": 5.3809,
"step": 3170
},
{
"epoch": 0.0009933402471480515,
"grad_norm": 13.75,
"learning_rate": 1.9364e-05,
"loss": 5.6005,
"step": 3180
},
{
"epoch": 0.0009964639586170704,
"grad_norm": 16.75,
"learning_rate": 1.9362000000000002e-05,
"loss": 5.304,
"step": 3190
},
{
"epoch": 0.0009995876700860894,
"grad_norm": 13.125,
"learning_rate": 1.936e-05,
"loss": 5.4925,
"step": 3200
},
{
"epoch": 0.0010027113815551086,
"grad_norm": 12.1875,
"learning_rate": 1.9358e-05,
"loss": 5.6312,
"step": 3210
},
{
"epoch": 0.0010058350930241275,
"grad_norm": 15.625,
"learning_rate": 1.9356000000000002e-05,
"loss": 5.5569,
"step": 3220
},
{
"epoch": 0.0010089588044931465,
"grad_norm": 16.5,
"learning_rate": 1.9354e-05,
"loss": 5.7318,
"step": 3230
},
{
"epoch": 0.0010120825159621657,
"grad_norm": 20.75,
"learning_rate": 1.9352e-05,
"loss": 5.3372,
"step": 3240
},
{
"epoch": 0.0010152062274311846,
"grad_norm": 15.8125,
"learning_rate": 1.9350000000000003e-05,
"loss": 5.3677,
"step": 3250
},
{
"epoch": 0.0010183299389002036,
"grad_norm": 12.75,
"learning_rate": 1.9348000000000002e-05,
"loss": 5.6175,
"step": 3260
},
{
"epoch": 0.0010214536503692227,
"grad_norm": 14.25,
"learning_rate": 1.9346000000000004e-05,
"loss": 5.3534,
"step": 3270
},
{
"epoch": 0.0010245773618382417,
"grad_norm": 12.4375,
"learning_rate": 1.9344e-05,
"loss": 5.1925,
"step": 3280
},
{
"epoch": 0.0010277010733072607,
"grad_norm": 13.5625,
"learning_rate": 1.9342000000000002e-05,
"loss": 5.8256,
"step": 3290
},
{
"epoch": 0.0010308247847762798,
"grad_norm": 14.5625,
"learning_rate": 1.934e-05,
"loss": 5.7592,
"step": 3300
},
{
"epoch": 0.0010339484962452988,
"grad_norm": 11.75,
"learning_rate": 1.9338e-05,
"loss": 5.6782,
"step": 3310
},
{
"epoch": 0.0010370722077143177,
"grad_norm": 23.25,
"learning_rate": 1.9336000000000003e-05,
"loss": 5.4134,
"step": 3320
},
{
"epoch": 0.001040195919183337,
"grad_norm": 14.625,
"learning_rate": 1.9334e-05,
"loss": 5.6013,
"step": 3330
},
{
"epoch": 0.0010433196306523559,
"grad_norm": 17.625,
"learning_rate": 1.9332e-05,
"loss": 5.4116,
"step": 3340
},
{
"epoch": 0.0010464433421213748,
"grad_norm": 10.5625,
"learning_rate": 1.9330000000000003e-05,
"loss": 5.5778,
"step": 3350
},
{
"epoch": 0.001049567053590394,
"grad_norm": 16.75,
"learning_rate": 1.9328000000000002e-05,
"loss": 5.7124,
"step": 3360
},
{
"epoch": 0.001052690765059413,
"grad_norm": 15.8125,
"learning_rate": 1.9326e-05,
"loss": 5.5873,
"step": 3370
},
{
"epoch": 0.001055814476528432,
"grad_norm": 15.4375,
"learning_rate": 1.9324e-05,
"loss": 5.706,
"step": 3380
},
{
"epoch": 0.001058938187997451,
"grad_norm": 14.375,
"learning_rate": 1.9322000000000002e-05,
"loss": 5.4256,
"step": 3390
},
{
"epoch": 0.00106206189946647,
"grad_norm": 19.0,
"learning_rate": 1.932e-05,
"loss": 5.3917,
"step": 3400
},
{
"epoch": 0.001065185610935489,
"grad_norm": 13.1875,
"learning_rate": 1.9318e-05,
"loss": 5.6195,
"step": 3410
},
{
"epoch": 0.0010683093224045082,
"grad_norm": 14.8125,
"learning_rate": 1.9316000000000003e-05,
"loss": 5.3445,
"step": 3420
},
{
"epoch": 0.0010714330338735271,
"grad_norm": 11.75,
"learning_rate": 1.9314000000000002e-05,
"loss": 5.155,
"step": 3430
},
{
"epoch": 0.001074556745342546,
"grad_norm": 13.0625,
"learning_rate": 1.9312e-05,
"loss": 5.3025,
"step": 3440
},
{
"epoch": 0.0010776804568115653,
"grad_norm": 14.5,
"learning_rate": 1.9310000000000003e-05,
"loss": 5.4074,
"step": 3450
},
{
"epoch": 0.0010808041682805842,
"grad_norm": 16.375,
"learning_rate": 1.9308000000000002e-05,
"loss": 5.6609,
"step": 3460
},
{
"epoch": 0.0010839278797496032,
"grad_norm": 19.25,
"learning_rate": 1.9306e-05,
"loss": 5.4313,
"step": 3470
},
{
"epoch": 0.0010870515912186224,
"grad_norm": 15.6875,
"learning_rate": 1.9304e-05,
"loss": 5.5033,
"step": 3480
},
{
"epoch": 0.0010901753026876413,
"grad_norm": 18.625,
"learning_rate": 1.9302e-05,
"loss": 5.7241,
"step": 3490
},
{
"epoch": 0.0010932990141566603,
"grad_norm": 13.0625,
"learning_rate": 1.93e-05,
"loss": 5.6599,
"step": 3500
},
{
"epoch": 0.0010964227256256794,
"grad_norm": 14.6875,
"learning_rate": 1.9298e-05,
"loss": 5.7546,
"step": 3510
},
{
"epoch": 0.0010995464370946984,
"grad_norm": 13.1875,
"learning_rate": 1.9296000000000003e-05,
"loss": 5.6877,
"step": 3520
},
{
"epoch": 0.0011026701485637176,
"grad_norm": 12.1875,
"learning_rate": 1.9294000000000002e-05,
"loss": 5.5043,
"step": 3530
},
{
"epoch": 0.0011057938600327365,
"grad_norm": 23.25,
"learning_rate": 1.9292e-05,
"loss": 5.3722,
"step": 3540
},
{
"epoch": 0.0011089175715017555,
"grad_norm": 16.625,
"learning_rate": 1.9290000000000003e-05,
"loss": 5.1707,
"step": 3550
},
{
"epoch": 0.0011120412829707747,
"grad_norm": 19.25,
"learning_rate": 1.9288000000000002e-05,
"loss": 5.6722,
"step": 3560
},
{
"epoch": 0.0011151649944397936,
"grad_norm": 12.75,
"learning_rate": 1.9286e-05,
"loss": 5.3163,
"step": 3570
},
{
"epoch": 0.0011182887059088126,
"grad_norm": 17.625,
"learning_rate": 1.9284e-05,
"loss": 5.4662,
"step": 3580
},
{
"epoch": 0.0011214124173778317,
"grad_norm": 15.75,
"learning_rate": 1.9282e-05,
"loss": 5.4683,
"step": 3590
},
{
"epoch": 0.0011245361288468507,
"grad_norm": 12.9375,
"learning_rate": 1.9280000000000002e-05,
"loss": 5.8348,
"step": 3600
},
{
"epoch": 0.0011276598403158697,
"grad_norm": 20.625,
"learning_rate": 1.9278e-05,
"loss": 5.5024,
"step": 3610
},
{
"epoch": 0.0011307835517848888,
"grad_norm": 19.125,
"learning_rate": 1.9276e-05,
"loss": 5.8194,
"step": 3620
},
{
"epoch": 0.0011339072632539078,
"grad_norm": 15.3125,
"learning_rate": 1.9274000000000002e-05,
"loss": 5.662,
"step": 3630
},
{
"epoch": 0.0011370309747229267,
"grad_norm": 15.3125,
"learning_rate": 1.9272e-05,
"loss": 5.6406,
"step": 3640
},
{
"epoch": 0.001140154686191946,
"grad_norm": 18.0,
"learning_rate": 1.9270000000000004e-05,
"loss": 5.5363,
"step": 3650
},
{
"epoch": 0.0011432783976609649,
"grad_norm": 13.4375,
"learning_rate": 1.9268000000000003e-05,
"loss": 5.4715,
"step": 3660
},
{
"epoch": 0.0011464021091299838,
"grad_norm": 16.25,
"learning_rate": 1.9266e-05,
"loss": 5.6616,
"step": 3670
},
{
"epoch": 0.001149525820599003,
"grad_norm": 11.375,
"learning_rate": 1.9264e-05,
"loss": 5.3432,
"step": 3680
},
{
"epoch": 0.001152649532068022,
"grad_norm": 20.0,
"learning_rate": 1.9262e-05,
"loss": 5.6658,
"step": 3690
},
{
"epoch": 0.001155773243537041,
"grad_norm": 10.375,
"learning_rate": 1.9260000000000002e-05,
"loss": 5.6279,
"step": 3700
},
{
"epoch": 0.00115889695500606,
"grad_norm": 14.4375,
"learning_rate": 1.9258e-05,
"loss": 5.7396,
"step": 3710
},
{
"epoch": 0.001162020666475079,
"grad_norm": 19.0,
"learning_rate": 1.9256e-05,
"loss": 5.3711,
"step": 3720
},
{
"epoch": 0.001165144377944098,
"grad_norm": 13.75,
"learning_rate": 1.9254000000000002e-05,
"loss": 5.0692,
"step": 3730
},
{
"epoch": 0.0011682680894131172,
"grad_norm": 14.625,
"learning_rate": 1.9252e-05,
"loss": 5.6739,
"step": 3740
},
{
"epoch": 0.0011713918008821361,
"grad_norm": 16.875,
"learning_rate": 1.925e-05,
"loss": 5.3631,
"step": 3750
},
{
"epoch": 0.001174515512351155,
"grad_norm": 12.375,
"learning_rate": 1.9248000000000003e-05,
"loss": 5.3772,
"step": 3760
},
{
"epoch": 0.0011776392238201743,
"grad_norm": 17.5,
"learning_rate": 1.9246000000000002e-05,
"loss": 5.5263,
"step": 3770
},
{
"epoch": 0.0011807629352891932,
"grad_norm": 9.25,
"learning_rate": 1.9244000000000004e-05,
"loss": 5.0929,
"step": 3780
},
{
"epoch": 0.0011838866467582122,
"grad_norm": 23.875,
"learning_rate": 1.9242e-05,
"loss": 5.7394,
"step": 3790
},
{
"epoch": 0.0011870103582272314,
"grad_norm": 11.0,
"learning_rate": 1.9240000000000002e-05,
"loss": 5.7045,
"step": 3800
},
{
"epoch": 0.0011901340696962503,
"grad_norm": 14.0,
"learning_rate": 1.9238e-05,
"loss": 5.6738,
"step": 3810
},
{
"epoch": 0.0011932577811652693,
"grad_norm": 15.375,
"learning_rate": 1.9236e-05,
"loss": 5.3815,
"step": 3820
},
{
"epoch": 0.0011963814926342884,
"grad_norm": 19.125,
"learning_rate": 1.9234000000000003e-05,
"loss": 5.0305,
"step": 3830
},
{
"epoch": 0.0011995052041033074,
"grad_norm": 13.4375,
"learning_rate": 1.9232e-05,
"loss": 5.7874,
"step": 3840
},
{
"epoch": 0.0012026289155723264,
"grad_norm": 17.125,
"learning_rate": 1.923e-05,
"loss": 5.4694,
"step": 3850
},
{
"epoch": 0.0012057526270413455,
"grad_norm": 14.625,
"learning_rate": 1.9228000000000003e-05,
"loss": 5.5002,
"step": 3860
},
{
"epoch": 0.0012088763385103645,
"grad_norm": 12.75,
"learning_rate": 1.9226000000000002e-05,
"loss": 5.7154,
"step": 3870
},
{
"epoch": 0.0012120000499793834,
"grad_norm": 17.625,
"learning_rate": 1.9224000000000004e-05,
"loss": 5.4635,
"step": 3880
},
{
"epoch": 0.0012151237614484026,
"grad_norm": 18.5,
"learning_rate": 1.9222e-05,
"loss": 5.7866,
"step": 3890
},
{
"epoch": 0.0012182474729174216,
"grad_norm": 12.625,
"learning_rate": 1.9220000000000002e-05,
"loss": 5.7408,
"step": 3900
},
{
"epoch": 0.0012213711843864405,
"grad_norm": 17.25,
"learning_rate": 1.9218e-05,
"loss": 5.5827,
"step": 3910
},
{
"epoch": 0.0012244948958554597,
"grad_norm": 18.625,
"learning_rate": 1.9216e-05,
"loss": 5.5495,
"step": 3920
},
{
"epoch": 0.0012276186073244787,
"grad_norm": 23.0,
"learning_rate": 1.9214000000000003e-05,
"loss": 5.7164,
"step": 3930
},
{
"epoch": 0.0012307423187934976,
"grad_norm": 16.625,
"learning_rate": 1.9212000000000002e-05,
"loss": 5.5238,
"step": 3940
},
{
"epoch": 0.0012338660302625168,
"grad_norm": 23.625,
"learning_rate": 1.921e-05,
"loss": 5.1463,
"step": 3950
},
{
"epoch": 0.0012369897417315358,
"grad_norm": 14.875,
"learning_rate": 1.9208000000000003e-05,
"loss": 5.4224,
"step": 3960
},
{
"epoch": 0.0012401134532005547,
"grad_norm": 12.3125,
"learning_rate": 1.9206000000000002e-05,
"loss": 5.6267,
"step": 3970
},
{
"epoch": 0.0012432371646695739,
"grad_norm": 12.9375,
"learning_rate": 1.9204e-05,
"loss": 5.3007,
"step": 3980
},
{
"epoch": 0.0012463608761385928,
"grad_norm": 26.75,
"learning_rate": 1.9202e-05,
"loss": 5.5985,
"step": 3990
},
{
"epoch": 0.0012494845876076118,
"grad_norm": 13.25,
"learning_rate": 1.9200000000000003e-05,
"loss": 5.8027,
"step": 4000
},
{
"epoch": 0.001252608299076631,
"grad_norm": 15.5625,
"learning_rate": 1.9198e-05,
"loss": 5.4355,
"step": 4010
},
{
"epoch": 0.00125573201054565,
"grad_norm": 15.125,
"learning_rate": 1.9196e-05,
"loss": 5.2763,
"step": 4020
},
{
"epoch": 0.0012588557220146689,
"grad_norm": 11.8125,
"learning_rate": 1.9194000000000003e-05,
"loss": 5.3245,
"step": 4030
},
{
"epoch": 0.001261979433483688,
"grad_norm": 14.6875,
"learning_rate": 1.9192000000000002e-05,
"loss": 4.9815,
"step": 4040
},
{
"epoch": 0.001265103144952707,
"grad_norm": 13.875,
"learning_rate": 1.919e-05,
"loss": 5.5233,
"step": 4050
},
{
"epoch": 0.001268226856421726,
"grad_norm": 17.625,
"learning_rate": 1.9188000000000003e-05,
"loss": 5.6294,
"step": 4060
},
{
"epoch": 0.0012713505678907451,
"grad_norm": 20.5,
"learning_rate": 1.9186000000000002e-05,
"loss": 5.2967,
"step": 4070
},
{
"epoch": 0.001274474279359764,
"grad_norm": 14.75,
"learning_rate": 1.9184e-05,
"loss": 5.0256,
"step": 4080
},
{
"epoch": 0.001277597990828783,
"grad_norm": 20.875,
"learning_rate": 1.9182e-05,
"loss": 5.0885,
"step": 4090
},
{
"epoch": 0.0012807217022978022,
"grad_norm": 11.75,
"learning_rate": 1.918e-05,
"loss": 5.4286,
"step": 4100
},
{
"epoch": 0.0012838454137668212,
"grad_norm": 14.5625,
"learning_rate": 1.9178000000000002e-05,
"loss": 5.344,
"step": 4110
},
{
"epoch": 0.0012869691252358401,
"grad_norm": 24.125,
"learning_rate": 1.9176e-05,
"loss": 5.362,
"step": 4120
},
{
"epoch": 0.0012900928367048593,
"grad_norm": 24.125,
"learning_rate": 1.9174000000000003e-05,
"loss": 5.6709,
"step": 4130
},
{
"epoch": 0.0012932165481738783,
"grad_norm": 15.6875,
"learning_rate": 1.9172000000000002e-05,
"loss": 5.5542,
"step": 4140
},
{
"epoch": 0.0012963402596428972,
"grad_norm": 16.125,
"learning_rate": 1.917e-05,
"loss": 5.2315,
"step": 4150
},
{
"epoch": 0.0012994639711119164,
"grad_norm": 17.5,
"learning_rate": 1.9168000000000004e-05,
"loss": 5.538,
"step": 4160
},
{
"epoch": 0.0013025876825809354,
"grad_norm": 15.9375,
"learning_rate": 1.9166000000000003e-05,
"loss": 5.5096,
"step": 4170
},
{
"epoch": 0.0013057113940499543,
"grad_norm": 11.125,
"learning_rate": 1.9164e-05,
"loss": 5.7454,
"step": 4180
},
{
"epoch": 0.0013088351055189735,
"grad_norm": 11.6875,
"learning_rate": 1.9162e-05,
"loss": 5.5636,
"step": 4190
},
{
"epoch": 0.0013119588169879925,
"grad_norm": 12.4375,
"learning_rate": 1.916e-05,
"loss": 5.6127,
"step": 4200
},
{
"epoch": 0.0013150825284570114,
"grad_norm": 11.875,
"learning_rate": 1.9158000000000002e-05,
"loss": 5.6977,
"step": 4210
},
{
"epoch": 0.0013182062399260306,
"grad_norm": 16.75,
"learning_rate": 1.9156e-05,
"loss": 5.7171,
"step": 4220
},
{
"epoch": 0.0013213299513950495,
"grad_norm": 17.125,
"learning_rate": 1.9154e-05,
"loss": 5.4361,
"step": 4230
},
{
"epoch": 0.0013244536628640685,
"grad_norm": 14.625,
"learning_rate": 1.9152000000000002e-05,
"loss": 5.3326,
"step": 4240
},
{
"epoch": 0.0013275773743330877,
"grad_norm": 15.4375,
"learning_rate": 1.915e-05,
"loss": 5.438,
"step": 4250
},
{
"epoch": 0.0013307010858021066,
"grad_norm": 17.0,
"learning_rate": 1.9148000000000004e-05,
"loss": 5.5329,
"step": 4260
},
{
"epoch": 0.0013338247972711256,
"grad_norm": 21.125,
"learning_rate": 1.9146000000000003e-05,
"loss": 5.791,
"step": 4270
},
{
"epoch": 0.0013369485087401448,
"grad_norm": 20.125,
"learning_rate": 1.9144000000000002e-05,
"loss": 5.7484,
"step": 4280
},
{
"epoch": 0.0013400722202091637,
"grad_norm": 15.3125,
"learning_rate": 1.9142e-05,
"loss": 5.0317,
"step": 4290
},
{
"epoch": 0.0013431959316781827,
"grad_norm": 25.0,
"learning_rate": 1.914e-05,
"loss": 5.6427,
"step": 4300
},
{
"epoch": 0.0013463196431472018,
"grad_norm": 21.375,
"learning_rate": 1.9138000000000002e-05,
"loss": 5.4972,
"step": 4310
},
{
"epoch": 0.0013494433546162208,
"grad_norm": 11.1875,
"learning_rate": 1.9136e-05,
"loss": 5.5393,
"step": 4320
},
{
"epoch": 0.0013525670660852398,
"grad_norm": 27.625,
"learning_rate": 1.9134e-05,
"loss": 5.3816,
"step": 4330
},
{
"epoch": 0.001355690777554259,
"grad_norm": 24.75,
"learning_rate": 1.9132000000000002e-05,
"loss": 5.5932,
"step": 4340
},
{
"epoch": 0.0013588144890232779,
"grad_norm": 19.25,
"learning_rate": 1.913e-05,
"loss": 5.7437,
"step": 4350
},
{
"epoch": 0.0013619382004922968,
"grad_norm": 22.625,
"learning_rate": 1.9128e-05,
"loss": 5.453,
"step": 4360
},
{
"epoch": 0.001365061911961316,
"grad_norm": 13.5625,
"learning_rate": 1.9126000000000003e-05,
"loss": 5.4275,
"step": 4370
},
{
"epoch": 0.001368185623430335,
"grad_norm": 18.375,
"learning_rate": 1.9124000000000002e-05,
"loss": 5.1498,
"step": 4380
},
{
"epoch": 0.001371309334899354,
"grad_norm": 13.25,
"learning_rate": 1.9122e-05,
"loss": 5.72,
"step": 4390
},
{
"epoch": 0.001374433046368373,
"grad_norm": 23.0,
"learning_rate": 1.912e-05,
"loss": 5.2518,
"step": 4400
},
{
"epoch": 0.001377556757837392,
"grad_norm": 12.8125,
"learning_rate": 1.9118000000000002e-05,
"loss": 5.5847,
"step": 4410
},
{
"epoch": 0.001380680469306411,
"grad_norm": 21.375,
"learning_rate": 1.9116e-05,
"loss": 5.6463,
"step": 4420
},
{
"epoch": 0.0013838041807754302,
"grad_norm": 19.0,
"learning_rate": 1.9114e-05,
"loss": 5.5771,
"step": 4430
},
{
"epoch": 0.0013869278922444492,
"grad_norm": 19.625,
"learning_rate": 1.9112000000000003e-05,
"loss": 5.5444,
"step": 4440
},
{
"epoch": 0.0013900516037134681,
"grad_norm": 11.875,
"learning_rate": 1.911e-05,
"loss": 5.5599,
"step": 4450
},
{
"epoch": 0.0013931753151824873,
"grad_norm": 14.1875,
"learning_rate": 1.9108e-05,
"loss": 5.6097,
"step": 4460
},
{
"epoch": 0.0013962990266515062,
"grad_norm": 15.625,
"learning_rate": 1.9106000000000003e-05,
"loss": 5.4403,
"step": 4470
},
{
"epoch": 0.0013994227381205252,
"grad_norm": 15.5,
"learning_rate": 1.9104000000000002e-05,
"loss": 5.6646,
"step": 4480
},
{
"epoch": 0.0014025464495895444,
"grad_norm": 16.375,
"learning_rate": 1.9102e-05,
"loss": 5.5024,
"step": 4490
},
{
"epoch": 0.0014056701610585633,
"grad_norm": 22.0,
"learning_rate": 1.91e-05,
"loss": 5.5846,
"step": 4500
},
{
"epoch": 0.0014087938725275823,
"grad_norm": 12.5625,
"learning_rate": 1.9098000000000002e-05,
"loss": 5.7252,
"step": 4510
},
{
"epoch": 0.0014119175839966015,
"grad_norm": 16.25,
"learning_rate": 1.9096e-05,
"loss": 5.7331,
"step": 4520
},
{
"epoch": 0.0014150412954656204,
"grad_norm": 17.5,
"learning_rate": 1.9094e-05,
"loss": 5.6089,
"step": 4530
},
{
"epoch": 0.0014181650069346394,
"grad_norm": 15.625,
"learning_rate": 1.9092000000000003e-05,
"loss": 5.5612,
"step": 4540
},
{
"epoch": 0.0014212887184036585,
"grad_norm": 10.3125,
"learning_rate": 1.9090000000000002e-05,
"loss": 5.4987,
"step": 4550
},
{
"epoch": 0.0014244124298726775,
"grad_norm": 18.625,
"learning_rate": 1.9088e-05,
"loss": 5.4008,
"step": 4560
},
{
"epoch": 0.0014275361413416965,
"grad_norm": 15.0625,
"learning_rate": 1.9086000000000003e-05,
"loss": 5.8175,
"step": 4570
},
{
"epoch": 0.0014306598528107156,
"grad_norm": 17.875,
"learning_rate": 1.9084000000000002e-05,
"loss": 5.7524,
"step": 4580
},
{
"epoch": 0.0014337835642797346,
"grad_norm": 14.0625,
"learning_rate": 1.9082e-05,
"loss": 5.6834,
"step": 4590
},
{
"epoch": 0.0014369072757487535,
"grad_norm": 17.375,
"learning_rate": 1.908e-05,
"loss": 5.7438,
"step": 4600
},
{
"epoch": 0.0014400309872177727,
"grad_norm": 14.625,
"learning_rate": 1.9078e-05,
"loss": 5.6002,
"step": 4610
},
{
"epoch": 0.0014431546986867917,
"grad_norm": 17.0,
"learning_rate": 1.9076e-05,
"loss": 5.5971,
"step": 4620
},
{
"epoch": 0.0014462784101558106,
"grad_norm": 12.5,
"learning_rate": 1.9074e-05,
"loss": 5.6446,
"step": 4630
},
{
"epoch": 0.0014494021216248298,
"grad_norm": 22.75,
"learning_rate": 1.9072000000000003e-05,
"loss": 5.2944,
"step": 4640
},
{
"epoch": 0.0014525258330938488,
"grad_norm": 13.25,
"learning_rate": 1.9070000000000002e-05,
"loss": 5.6228,
"step": 4650
},
{
"epoch": 0.0014556495445628677,
"grad_norm": 13.25,
"learning_rate": 1.9068e-05,
"loss": 5.1766,
"step": 4660
},
{
"epoch": 0.001458773256031887,
"grad_norm": 20.875,
"learning_rate": 1.9066000000000003e-05,
"loss": 5.1814,
"step": 4670
},
{
"epoch": 0.0014618969675009059,
"grad_norm": 16.625,
"learning_rate": 1.9064000000000002e-05,
"loss": 5.2023,
"step": 4680
},
{
"epoch": 0.0014650206789699248,
"grad_norm": 18.75,
"learning_rate": 1.9062e-05,
"loss": 5.5188,
"step": 4690
},
{
"epoch": 0.001468144390438944,
"grad_norm": 16.875,
"learning_rate": 1.906e-05,
"loss": 5.5913,
"step": 4700
},
{
"epoch": 0.001471268101907963,
"grad_norm": 12.3125,
"learning_rate": 1.9058e-05,
"loss": 5.5999,
"step": 4710
},
{
"epoch": 0.001474391813376982,
"grad_norm": 16.0,
"learning_rate": 1.9056000000000002e-05,
"loss": 5.7381,
"step": 4720
},
{
"epoch": 0.001477515524846001,
"grad_norm": 12.625,
"learning_rate": 1.9054e-05,
"loss": 5.7001,
"step": 4730
},
{
"epoch": 0.00148063923631502,
"grad_norm": 13.875,
"learning_rate": 1.9052000000000003e-05,
"loss": 5.5039,
"step": 4740
},
{
"epoch": 0.001483762947784039,
"grad_norm": 16.75,
"learning_rate": 1.9050000000000002e-05,
"loss": 5.7352,
"step": 4750
},
{
"epoch": 0.0014868866592530582,
"grad_norm": 15.1875,
"learning_rate": 1.9048e-05,
"loss": 5.5764,
"step": 4760
},
{
"epoch": 0.0014900103707220771,
"grad_norm": 12.75,
"learning_rate": 1.9046000000000004e-05,
"loss": 5.6157,
"step": 4770
},
{
"epoch": 0.001493134082191096,
"grad_norm": 13.625,
"learning_rate": 1.9044000000000003e-05,
"loss": 5.5936,
"step": 4780
},
{
"epoch": 0.0014962577936601152,
"grad_norm": 16.75,
"learning_rate": 1.9042e-05,
"loss": 5.5932,
"step": 4790
},
{
"epoch": 0.0014993815051291342,
"grad_norm": 22.75,
"learning_rate": 1.904e-05,
"loss": 5.1952,
"step": 4800
},
{
"epoch": 0.0015025052165981532,
"grad_norm": 10.625,
"learning_rate": 1.9038e-05,
"loss": 5.2644,
"step": 4810
},
{
"epoch": 0.0015056289280671723,
"grad_norm": 12.1875,
"learning_rate": 1.9036000000000002e-05,
"loss": 5.6933,
"step": 4820
},
{
"epoch": 0.0015087526395361913,
"grad_norm": 14.875,
"learning_rate": 1.9034e-05,
"loss": 5.7847,
"step": 4830
},
{
"epoch": 0.0015118763510052102,
"grad_norm": 23.625,
"learning_rate": 1.9032e-05,
"loss": 5.3767,
"step": 4840
},
{
"epoch": 0.0015150000624742294,
"grad_norm": 17.875,
"learning_rate": 1.9030000000000002e-05,
"loss": 5.6129,
"step": 4850
},
{
"epoch": 0.0015181237739432484,
"grad_norm": 15.4375,
"learning_rate": 1.9028e-05,
"loss": 5.6761,
"step": 4860
},
{
"epoch": 0.0015212474854122673,
"grad_norm": 13.4375,
"learning_rate": 1.9026000000000004e-05,
"loss": 5.5118,
"step": 4870
},
{
"epoch": 0.0015243711968812865,
"grad_norm": 14.625,
"learning_rate": 1.9024000000000003e-05,
"loss": 5.5173,
"step": 4880
},
{
"epoch": 0.0015274949083503055,
"grad_norm": 15.3125,
"learning_rate": 1.9022000000000002e-05,
"loss": 5.5349,
"step": 4890
},
{
"epoch": 0.0015306186198193244,
"grad_norm": 17.25,
"learning_rate": 1.902e-05,
"loss": 5.1842,
"step": 4900
},
{
"epoch": 0.0015337423312883436,
"grad_norm": 28.5,
"learning_rate": 1.9018e-05,
"loss": 5.2254,
"step": 4910
},
{
"epoch": 0.0015368660427573626,
"grad_norm": 15.6875,
"learning_rate": 1.9016000000000002e-05,
"loss": 5.4484,
"step": 4920
},
{
"epoch": 0.0015399897542263817,
"grad_norm": 13.0,
"learning_rate": 1.9014e-05,
"loss": 5.1899,
"step": 4930
},
{
"epoch": 0.0015431134656954007,
"grad_norm": 15.5,
"learning_rate": 1.9012e-05,
"loss": 5.7682,
"step": 4940
},
{
"epoch": 0.0015462371771644196,
"grad_norm": 16.125,
"learning_rate": 1.9010000000000003e-05,
"loss": 5.564,
"step": 4950
},
{
"epoch": 0.0015493608886334388,
"grad_norm": 15.3125,
"learning_rate": 1.9008e-05,
"loss": 5.0649,
"step": 4960
},
{
"epoch": 0.0015524846001024578,
"grad_norm": 14.375,
"learning_rate": 1.9006e-05,
"loss": 5.1236,
"step": 4970
},
{
"epoch": 0.0015556083115714767,
"grad_norm": 13.0,
"learning_rate": 1.9004000000000003e-05,
"loss": 5.6358,
"step": 4980
},
{
"epoch": 0.001558732023040496,
"grad_norm": 13.0625,
"learning_rate": 1.9002000000000002e-05,
"loss": 5.4334,
"step": 4990
},
{
"epoch": 0.0015618557345095149,
"grad_norm": 10.8125,
"learning_rate": 1.9e-05,
"loss": 5.5315,
"step": 5000
},
{
"epoch": 0.0015649794459785338,
"grad_norm": 17.125,
"learning_rate": 1.8998e-05,
"loss": 5.4787,
"step": 5010
},
{
"epoch": 0.001568103157447553,
"grad_norm": 12.0625,
"learning_rate": 1.8996000000000002e-05,
"loss": 5.6937,
"step": 5020
},
{
"epoch": 0.001571226868916572,
"grad_norm": 14.8125,
"learning_rate": 1.8994e-05,
"loss": 5.29,
"step": 5030
},
{
"epoch": 0.001574350580385591,
"grad_norm": 22.375,
"learning_rate": 1.8992e-05,
"loss": 5.4227,
"step": 5040
},
{
"epoch": 0.00157747429185461,
"grad_norm": 14.1875,
"learning_rate": 1.8990000000000003e-05,
"loss": 5.4948,
"step": 5050
},
{
"epoch": 0.001580598003323629,
"grad_norm": 18.625,
"learning_rate": 1.8988000000000002e-05,
"loss": 5.6641,
"step": 5060
},
{
"epoch": 0.001583721714792648,
"grad_norm": 21.375,
"learning_rate": 1.8986e-05,
"loss": 5.6905,
"step": 5070
},
{
"epoch": 0.0015868454262616672,
"grad_norm": 21.75,
"learning_rate": 1.8984000000000003e-05,
"loss": 5.2701,
"step": 5080
},
{
"epoch": 0.0015899691377306861,
"grad_norm": 18.625,
"learning_rate": 1.8982000000000002e-05,
"loss": 5.697,
"step": 5090
},
{
"epoch": 0.001593092849199705,
"grad_norm": 20.0,
"learning_rate": 1.898e-05,
"loss": 5.455,
"step": 5100
},
{
"epoch": 0.0015962165606687242,
"grad_norm": 16.125,
"learning_rate": 1.8978e-05,
"loss": 5.4842,
"step": 5110
},
{
"epoch": 0.0015993402721377432,
"grad_norm": 15.5,
"learning_rate": 1.8976000000000003e-05,
"loss": 5.2105,
"step": 5120
},
{
"epoch": 0.0016024639836067622,
"grad_norm": 19.125,
"learning_rate": 1.8974e-05,
"loss": 5.7378,
"step": 5130
},
{
"epoch": 0.0016055876950757813,
"grad_norm": 15.625,
"learning_rate": 1.8972e-05,
"loss": 5.7174,
"step": 5140
},
{
"epoch": 0.0016087114065448003,
"grad_norm": 14.4375,
"learning_rate": 1.8970000000000003e-05,
"loss": 5.6989,
"step": 5150
},
{
"epoch": 0.0016118351180138193,
"grad_norm": 16.125,
"learning_rate": 1.8968000000000002e-05,
"loss": 5.7208,
"step": 5160
},
{
"epoch": 0.0016149588294828384,
"grad_norm": 14.375,
"learning_rate": 1.8966e-05,
"loss": 5.5058,
"step": 5170
},
{
"epoch": 0.0016180825409518574,
"grad_norm": 16.5,
"learning_rate": 1.8964000000000003e-05,
"loss": 5.2135,
"step": 5180
},
{
"epoch": 0.0016212062524208763,
"grad_norm": 18.125,
"learning_rate": 1.8962000000000002e-05,
"loss": 5.6446,
"step": 5190
},
{
"epoch": 0.0016243299638898955,
"grad_norm": 17.375,
"learning_rate": 1.896e-05,
"loss": 5.7546,
"step": 5200
},
{
"epoch": 0.0016274536753589145,
"grad_norm": 13.125,
"learning_rate": 1.8958e-05,
"loss": 5.4918,
"step": 5210
},
{
"epoch": 0.0016305773868279334,
"grad_norm": 21.375,
"learning_rate": 1.8956e-05,
"loss": 5.4425,
"step": 5220
},
{
"epoch": 0.0016337010982969526,
"grad_norm": 13.5625,
"learning_rate": 1.8954000000000002e-05,
"loss": 5.5864,
"step": 5230
},
{
"epoch": 0.0016368248097659716,
"grad_norm": 18.5,
"learning_rate": 1.8952e-05,
"loss": 5.7043,
"step": 5240
},
{
"epoch": 0.0016399485212349905,
"grad_norm": 18.625,
"learning_rate": 1.8950000000000003e-05,
"loss": 5.5245,
"step": 5250
},
{
"epoch": 0.0016430722327040097,
"grad_norm": 13.375,
"learning_rate": 1.8948000000000002e-05,
"loss": 5.5455,
"step": 5260
},
{
"epoch": 0.0016461959441730286,
"grad_norm": 17.125,
"learning_rate": 1.8946e-05,
"loss": 5.5937,
"step": 5270
},
{
"epoch": 0.0016493196556420476,
"grad_norm": 22.0,
"learning_rate": 1.8944000000000004e-05,
"loss": 5.4067,
"step": 5280
},
{
"epoch": 0.0016524433671110668,
"grad_norm": 18.125,
"learning_rate": 1.8942000000000003e-05,
"loss": 5.7337,
"step": 5290
},
{
"epoch": 0.0016555670785800857,
"grad_norm": 14.25,
"learning_rate": 1.894e-05,
"loss": 5.5917,
"step": 5300
},
{
"epoch": 0.0016586907900491047,
"grad_norm": 17.25,
"learning_rate": 1.8938e-05,
"loss": 5.4728,
"step": 5310
},
{
"epoch": 0.0016618145015181239,
"grad_norm": 15.4375,
"learning_rate": 1.8936e-05,
"loss": 5.5482,
"step": 5320
},
{
"epoch": 0.0016649382129871428,
"grad_norm": 20.25,
"learning_rate": 1.8934000000000002e-05,
"loss": 5.6185,
"step": 5330
},
{
"epoch": 0.0016680619244561618,
"grad_norm": 14.6875,
"learning_rate": 1.8932e-05,
"loss": 5.7582,
"step": 5340
},
{
"epoch": 0.001671185635925181,
"grad_norm": 20.875,
"learning_rate": 1.893e-05,
"loss": 5.733,
"step": 5350
},
{
"epoch": 0.0016743093473942,
"grad_norm": 11.625,
"learning_rate": 1.8928000000000002e-05,
"loss": 5.754,
"step": 5360
},
{
"epoch": 0.0016774330588632189,
"grad_norm": 14.6875,
"learning_rate": 1.8926e-05,
"loss": 5.5967,
"step": 5370
},
{
"epoch": 0.001680556770332238,
"grad_norm": 15.375,
"learning_rate": 1.8924000000000004e-05,
"loss": 5.5708,
"step": 5380
},
{
"epoch": 0.001683680481801257,
"grad_norm": 19.625,
"learning_rate": 1.8922000000000003e-05,
"loss": 5.4091,
"step": 5390
},
{
"epoch": 0.001686804193270276,
"grad_norm": 14.125,
"learning_rate": 1.8920000000000002e-05,
"loss": 5.7056,
"step": 5400
},
{
"epoch": 0.0016899279047392951,
"grad_norm": 14.1875,
"learning_rate": 1.8918e-05,
"loss": 5.7095,
"step": 5410
},
{
"epoch": 0.001693051616208314,
"grad_norm": 17.625,
"learning_rate": 1.8916e-05,
"loss": 5.632,
"step": 5420
},
{
"epoch": 0.001696175327677333,
"grad_norm": 17.0,
"learning_rate": 1.8914000000000002e-05,
"loss": 5.5599,
"step": 5430
},
{
"epoch": 0.0016992990391463522,
"grad_norm": 14.5625,
"learning_rate": 1.8912e-05,
"loss": 5.6591,
"step": 5440
},
{
"epoch": 0.0017024227506153712,
"grad_norm": 14.9375,
"learning_rate": 1.891e-05,
"loss": 5.5385,
"step": 5450
},
{
"epoch": 0.0017055464620843901,
"grad_norm": 19.625,
"learning_rate": 1.8908000000000003e-05,
"loss": 5.4329,
"step": 5460
},
{
"epoch": 0.0017086701735534093,
"grad_norm": 22.375,
"learning_rate": 1.8906e-05,
"loss": 5.2307,
"step": 5470
},
{
"epoch": 0.0017117938850224283,
"grad_norm": 19.125,
"learning_rate": 1.8904000000000004e-05,
"loss": 5.2522,
"step": 5480
},
{
"epoch": 0.0017149175964914472,
"grad_norm": 11.625,
"learning_rate": 1.8902000000000003e-05,
"loss": 4.9953,
"step": 5490
},
{
"epoch": 0.0017180413079604664,
"grad_norm": 19.125,
"learning_rate": 1.8900000000000002e-05,
"loss": 5.6609,
"step": 5500
},
{
"epoch": 0.0017211650194294853,
"grad_norm": 18.75,
"learning_rate": 1.8898e-05,
"loss": 5.4622,
"step": 5510
},
{
"epoch": 0.0017242887308985043,
"grad_norm": 12.8125,
"learning_rate": 1.8896e-05,
"loss": 5.2735,
"step": 5520
},
{
"epoch": 0.0017274124423675235,
"grad_norm": 13.5625,
"learning_rate": 1.8894000000000002e-05,
"loss": 5.6178,
"step": 5530
},
{
"epoch": 0.0017305361538365424,
"grad_norm": 19.125,
"learning_rate": 1.8892e-05,
"loss": 5.0805,
"step": 5540
},
{
"epoch": 0.0017336598653055614,
"grad_norm": 20.125,
"learning_rate": 1.889e-05,
"loss": 5.3146,
"step": 5550
},
{
"epoch": 0.0017367835767745806,
"grad_norm": 11.375,
"learning_rate": 1.8888000000000003e-05,
"loss": 5.741,
"step": 5560
},
{
"epoch": 0.0017399072882435995,
"grad_norm": 15.5625,
"learning_rate": 1.8886000000000002e-05,
"loss": 5.7295,
"step": 5570
},
{
"epoch": 0.0017430309997126185,
"grad_norm": 12.1875,
"learning_rate": 1.8884e-05,
"loss": 5.4365,
"step": 5580
},
{
"epoch": 0.0017461547111816376,
"grad_norm": 18.125,
"learning_rate": 1.8882000000000003e-05,
"loss": 5.6188,
"step": 5590
},
{
"epoch": 0.0017492784226506566,
"grad_norm": 18.75,
"learning_rate": 1.8880000000000002e-05,
"loss": 5.4696,
"step": 5600
},
{
"epoch": 0.0017524021341196756,
"grad_norm": 12.25,
"learning_rate": 1.8878e-05,
"loss": 5.5221,
"step": 5610
},
{
"epoch": 0.0017555258455886947,
"grad_norm": 21.75,
"learning_rate": 1.8876e-05,
"loss": 5.1922,
"step": 5620
},
{
"epoch": 0.0017586495570577137,
"grad_norm": 17.125,
"learning_rate": 1.8874000000000003e-05,
"loss": 5.4216,
"step": 5630
},
{
"epoch": 0.0017617732685267326,
"grad_norm": 17.75,
"learning_rate": 1.8872e-05,
"loss": 5.4876,
"step": 5640
},
{
"epoch": 0.0017648969799957518,
"grad_norm": 17.0,
"learning_rate": 1.887e-05,
"loss": 5.595,
"step": 5650
},
{
"epoch": 0.0017680206914647708,
"grad_norm": 20.25,
"learning_rate": 1.8868000000000003e-05,
"loss": 5.6682,
"step": 5660
},
{
"epoch": 0.0017711444029337897,
"grad_norm": 17.375,
"learning_rate": 1.8866000000000002e-05,
"loss": 5.2636,
"step": 5670
},
{
"epoch": 0.001774268114402809,
"grad_norm": 14.875,
"learning_rate": 1.8864e-05,
"loss": 5.4282,
"step": 5680
},
{
"epoch": 0.0017773918258718279,
"grad_norm": 18.375,
"learning_rate": 1.8862000000000003e-05,
"loss": 5.0176,
"step": 5690
},
{
"epoch": 0.0017805155373408468,
"grad_norm": 16.375,
"learning_rate": 1.886e-05,
"loss": 5.6029,
"step": 5700
},
{
"epoch": 0.001783639248809866,
"grad_norm": 9.875,
"learning_rate": 1.8858e-05,
"loss": 5.7118,
"step": 5710
},
{
"epoch": 0.001786762960278885,
"grad_norm": 21.25,
"learning_rate": 1.8856e-05,
"loss": 5.0426,
"step": 5720
},
{
"epoch": 0.001789886671747904,
"grad_norm": 20.5,
"learning_rate": 1.8854000000000003e-05,
"loss": 5.5434,
"step": 5730
},
{
"epoch": 0.001793010383216923,
"grad_norm": 12.75,
"learning_rate": 1.8852000000000002e-05,
"loss": 5.665,
"step": 5740
},
{
"epoch": 0.001796134094685942,
"grad_norm": 16.375,
"learning_rate": 1.885e-05,
"loss": 5.6768,
"step": 5750
},
{
"epoch": 0.001799257806154961,
"grad_norm": 19.25,
"learning_rate": 1.8848000000000003e-05,
"loss": 5.3965,
"step": 5760
},
{
"epoch": 0.0018023815176239802,
"grad_norm": 13.0625,
"learning_rate": 1.8846000000000002e-05,
"loss": 5.6074,
"step": 5770
},
{
"epoch": 0.0018055052290929991,
"grad_norm": 20.875,
"learning_rate": 1.8844e-05,
"loss": 5.6579,
"step": 5780
},
{
"epoch": 0.001808628940562018,
"grad_norm": 15.75,
"learning_rate": 1.8842000000000004e-05,
"loss": 5.5465,
"step": 5790
},
{
"epoch": 0.0018117526520310373,
"grad_norm": 14.8125,
"learning_rate": 1.884e-05,
"loss": 5.542,
"step": 5800
},
{
"epoch": 0.0018148763635000562,
"grad_norm": 14.3125,
"learning_rate": 1.8838e-05,
"loss": 5.7126,
"step": 5810
},
{
"epoch": 0.0018180000749690752,
"grad_norm": 14.125,
"learning_rate": 1.8836e-05,
"loss": 5.6531,
"step": 5820
},
{
"epoch": 0.0018211237864380943,
"grad_norm": 11.9375,
"learning_rate": 1.8834e-05,
"loss": 5.1662,
"step": 5830
},
{
"epoch": 0.0018242474979071133,
"grad_norm": 18.625,
"learning_rate": 1.8832000000000002e-05,
"loss": 5.7536,
"step": 5840
},
{
"epoch": 0.0018273712093761323,
"grad_norm": 13.625,
"learning_rate": 1.883e-05,
"loss": 5.7136,
"step": 5850
},
{
"epoch": 0.0018304949208451514,
"grad_norm": 13.75,
"learning_rate": 1.8828000000000003e-05,
"loss": 5.4302,
"step": 5860
},
{
"epoch": 0.0018336186323141704,
"grad_norm": 15.75,
"learning_rate": 1.8826000000000002e-05,
"loss": 5.3569,
"step": 5870
},
{
"epoch": 0.0018367423437831893,
"grad_norm": 15.25,
"learning_rate": 1.8824e-05,
"loss": 4.9267,
"step": 5880
},
{
"epoch": 0.0018398660552522085,
"grad_norm": 13.0,
"learning_rate": 1.8822000000000004e-05,
"loss": 5.2913,
"step": 5890
},
{
"epoch": 0.0018429897667212275,
"grad_norm": 22.875,
"learning_rate": 1.882e-05,
"loss": 5.4008,
"step": 5900
},
{
"epoch": 0.0018461134781902464,
"grad_norm": 15.1875,
"learning_rate": 1.8818000000000002e-05,
"loss": 5.378,
"step": 5910
},
{
"epoch": 0.0018492371896592656,
"grad_norm": 11.375,
"learning_rate": 1.8816e-05,
"loss": 5.8182,
"step": 5920
},
{
"epoch": 0.0018523609011282846,
"grad_norm": 18.0,
"learning_rate": 1.8814e-05,
"loss": 4.9918,
"step": 5930
},
{
"epoch": 0.0018554846125973035,
"grad_norm": 12.4375,
"learning_rate": 1.8812000000000002e-05,
"loss": 5.6478,
"step": 5940
},
{
"epoch": 0.0018586083240663227,
"grad_norm": 11.875,
"learning_rate": 1.881e-05,
"loss": 5.4287,
"step": 5950
},
{
"epoch": 0.0018617320355353417,
"grad_norm": 15.3125,
"learning_rate": 1.8808e-05,
"loss": 5.6799,
"step": 5960
},
{
"epoch": 0.0018648557470043606,
"grad_norm": 14.625,
"learning_rate": 1.8806000000000003e-05,
"loss": 5.4295,
"step": 5970
},
{
"epoch": 0.0018679794584733798,
"grad_norm": 13.5625,
"learning_rate": 1.8804e-05,
"loss": 5.7169,
"step": 5980
},
{
"epoch": 0.0018711031699423987,
"grad_norm": 14.625,
"learning_rate": 1.8802000000000004e-05,
"loss": 5.6801,
"step": 5990
},
{
"epoch": 0.0018742268814114177,
"grad_norm": 17.875,
"learning_rate": 1.88e-05,
"loss": 5.5941,
"step": 6000
},
{
"epoch": 0.0018773505928804369,
"grad_norm": 13.1875,
"learning_rate": 1.8798000000000002e-05,
"loss": 5.4782,
"step": 6010
},
{
"epoch": 0.0018804743043494558,
"grad_norm": 19.5,
"learning_rate": 1.8796e-05,
"loss": 5.371,
"step": 6020
},
{
"epoch": 0.0018835980158184748,
"grad_norm": 14.0,
"learning_rate": 1.8794e-05,
"loss": 5.0534,
"step": 6030
},
{
"epoch": 0.001886721727287494,
"grad_norm": 18.125,
"learning_rate": 1.8792000000000002e-05,
"loss": 5.5754,
"step": 6040
},
{
"epoch": 0.001889845438756513,
"grad_norm": 14.875,
"learning_rate": 1.879e-05,
"loss": 5.4178,
"step": 6050
},
{
"epoch": 0.0018929691502255319,
"grad_norm": 10.875,
"learning_rate": 1.8788e-05,
"loss": 5.6989,
"step": 6060
},
{
"epoch": 0.001896092861694551,
"grad_norm": 14.125,
"learning_rate": 1.8786000000000003e-05,
"loss": 5.1613,
"step": 6070
},
{
"epoch": 0.00189921657316357,
"grad_norm": 13.3125,
"learning_rate": 1.8784000000000002e-05,
"loss": 5.7284,
"step": 6080
},
{
"epoch": 0.001902340284632589,
"grad_norm": 17.375,
"learning_rate": 1.8782e-05,
"loss": 5.4167,
"step": 6090
},
{
"epoch": 0.0019054639961016081,
"grad_norm": 21.375,
"learning_rate": 1.878e-05,
"loss": 5.3408,
"step": 6100
},
{
"epoch": 0.001908587707570627,
"grad_norm": 16.5,
"learning_rate": 1.8778000000000002e-05,
"loss": 5.6557,
"step": 6110
},
{
"epoch": 0.001911711419039646,
"grad_norm": 13.0,
"learning_rate": 1.8776e-05,
"loss": 5.4712,
"step": 6120
},
{
"epoch": 0.0019148351305086652,
"grad_norm": 16.0,
"learning_rate": 1.8774e-05,
"loss": 5.7737,
"step": 6130
},
{
"epoch": 0.0019179588419776842,
"grad_norm": 13.0,
"learning_rate": 1.8772000000000003e-05,
"loss": 5.6721,
"step": 6140
},
{
"epoch": 0.0019210825534467031,
"grad_norm": 13.0625,
"learning_rate": 1.877e-05,
"loss": 5.729,
"step": 6150
},
{
"epoch": 0.0019242062649157223,
"grad_norm": 13.8125,
"learning_rate": 1.8768e-05,
"loss": 4.9197,
"step": 6160
},
{
"epoch": 0.0019273299763847413,
"grad_norm": 18.75,
"learning_rate": 1.8766000000000003e-05,
"loss": 5.704,
"step": 6170
},
{
"epoch": 0.0019304536878537602,
"grad_norm": 13.3125,
"learning_rate": 1.8764000000000002e-05,
"loss": 5.7139,
"step": 6180
},
{
"epoch": 0.0019335773993227794,
"grad_norm": 15.875,
"learning_rate": 1.8762e-05,
"loss": 5.1654,
"step": 6190
},
{
"epoch": 0.0019367011107917984,
"grad_norm": 16.625,
"learning_rate": 1.876e-05,
"loss": 5.3066,
"step": 6200
},
{
"epoch": 0.0019398248222608173,
"grad_norm": 11.125,
"learning_rate": 1.8758000000000002e-05,
"loss": 5.585,
"step": 6210
},
{
"epoch": 0.0019429485337298365,
"grad_norm": 12.75,
"learning_rate": 1.8756e-05,
"loss": 5.4716,
"step": 6220
},
{
"epoch": 0.0019460722451988554,
"grad_norm": 11.375,
"learning_rate": 1.8754e-05,
"loss": 5.4859,
"step": 6230
},
{
"epoch": 0.0019491959566678744,
"grad_norm": 12.5,
"learning_rate": 1.8752000000000003e-05,
"loss": 5.6636,
"step": 6240
},
{
"epoch": 0.0019523196681368936,
"grad_norm": 11.1875,
"learning_rate": 1.8750000000000002e-05,
"loss": 5.4264,
"step": 6250
},
{
"epoch": 0.0019554433796059125,
"grad_norm": 19.375,
"learning_rate": 1.8748e-05,
"loss": 5.3095,
"step": 6260
},
{
"epoch": 0.0019585670910749315,
"grad_norm": 12.75,
"learning_rate": 1.8746000000000003e-05,
"loss": 5.4934,
"step": 6270
},
{
"epoch": 0.0019616908025439504,
"grad_norm": 15.9375,
"learning_rate": 1.8744000000000002e-05,
"loss": 5.6084,
"step": 6280
},
{
"epoch": 0.00196481451401297,
"grad_norm": 11.5,
"learning_rate": 1.8742e-05,
"loss": 5.695,
"step": 6290
},
{
"epoch": 0.001967938225481989,
"grad_norm": 17.5,
"learning_rate": 1.8740000000000004e-05,
"loss": 5.456,
"step": 6300
},
{
"epoch": 0.0019710619369510077,
"grad_norm": 21.375,
"learning_rate": 1.8738e-05,
"loss": 5.2945,
"step": 6310
},
{
"epoch": 0.0019741856484200267,
"grad_norm": 15.875,
"learning_rate": 1.8736e-05,
"loss": 5.7415,
"step": 6320
},
{
"epoch": 0.0019773093598890457,
"grad_norm": 15.6875,
"learning_rate": 1.8734e-05,
"loss": 5.7126,
"step": 6330
},
{
"epoch": 0.0019804330713580646,
"grad_norm": 19.5,
"learning_rate": 1.8732000000000003e-05,
"loss": 5.5993,
"step": 6340
},
{
"epoch": 0.001983556782827084,
"grad_norm": 15.9375,
"learning_rate": 1.8730000000000002e-05,
"loss": 5.6778,
"step": 6350
},
{
"epoch": 0.001986680494296103,
"grad_norm": 15.8125,
"learning_rate": 1.8728e-05,
"loss": 5.5652,
"step": 6360
},
{
"epoch": 0.001989804205765122,
"grad_norm": 11.25,
"learning_rate": 1.8726000000000003e-05,
"loss": 5.5042,
"step": 6370
},
{
"epoch": 0.001992927917234141,
"grad_norm": 19.125,
"learning_rate": 1.8724000000000002e-05,
"loss": 5.7456,
"step": 6380
},
{
"epoch": 0.00199605162870316,
"grad_norm": 18.75,
"learning_rate": 1.8722e-05,
"loss": 5.7884,
"step": 6390
},
{
"epoch": 0.001999175340172179,
"grad_norm": 16.875,
"learning_rate": 1.8720000000000004e-05,
"loss": 5.3617,
"step": 6400
},
{
"epoch": 0.002002299051641198,
"grad_norm": 17.125,
"learning_rate": 1.8718e-05,
"loss": 5.0735,
"step": 6410
},
{
"epoch": 0.002005422763110217,
"grad_norm": 12.5625,
"learning_rate": 1.8716000000000002e-05,
"loss": 5.5877,
"step": 6420
},
{
"epoch": 0.002008546474579236,
"grad_norm": 15.0625,
"learning_rate": 1.8714e-05,
"loss": 5.5967,
"step": 6430
},
{
"epoch": 0.002011670186048255,
"grad_norm": 18.875,
"learning_rate": 1.8712e-05,
"loss": 5.4915,
"step": 6440
},
{
"epoch": 0.002014793897517274,
"grad_norm": 11.0625,
"learning_rate": 1.8710000000000002e-05,
"loss": 5.4465,
"step": 6450
},
{
"epoch": 0.002017917608986293,
"grad_norm": 18.125,
"learning_rate": 1.8708e-05,
"loss": 5.7355,
"step": 6460
},
{
"epoch": 0.0020210413204553124,
"grad_norm": 11.8125,
"learning_rate": 1.8706000000000004e-05,
"loss": 5.7497,
"step": 6470
},
{
"epoch": 0.0020241650319243313,
"grad_norm": 21.375,
"learning_rate": 1.8704000000000003e-05,
"loss": 5.501,
"step": 6480
},
{
"epoch": 0.0020272887433933503,
"grad_norm": 14.25,
"learning_rate": 1.8702e-05,
"loss": 5.3654,
"step": 6490
},
{
"epoch": 0.0020304124548623692,
"grad_norm": 14.25,
"learning_rate": 1.8700000000000004e-05,
"loss": 5.4165,
"step": 6500
},
{
"epoch": 0.002033536166331388,
"grad_norm": 14.75,
"learning_rate": 1.8698e-05,
"loss": 5.0379,
"step": 6510
},
{
"epoch": 0.002036659877800407,
"grad_norm": 18.25,
"learning_rate": 1.8696000000000002e-05,
"loss": 5.718,
"step": 6520
},
{
"epoch": 0.0020397835892694265,
"grad_norm": 21.0,
"learning_rate": 1.8694e-05,
"loss": 5.3976,
"step": 6530
},
{
"epoch": 0.0020429073007384455,
"grad_norm": 12.5,
"learning_rate": 1.8692e-05,
"loss": 5.745,
"step": 6540
},
{
"epoch": 0.0020460310122074644,
"grad_norm": 27.5,
"learning_rate": 1.8690000000000002e-05,
"loss": 5.4708,
"step": 6550
},
{
"epoch": 0.0020491547236764834,
"grad_norm": 15.75,
"learning_rate": 1.8688e-05,
"loss": 5.5559,
"step": 6560
},
{
"epoch": 0.0020522784351455024,
"grad_norm": 14.25,
"learning_rate": 1.8686e-05,
"loss": 5.4009,
"step": 6570
},
{
"epoch": 0.0020554021466145213,
"grad_norm": 14.375,
"learning_rate": 1.8684000000000003e-05,
"loss": 5.4685,
"step": 6580
},
{
"epoch": 0.0020585258580835407,
"grad_norm": 22.25,
"learning_rate": 1.8682000000000002e-05,
"loss": 5.4608,
"step": 6590
},
{
"epoch": 0.0020616495695525597,
"grad_norm": 12.8125,
"learning_rate": 1.8680000000000004e-05,
"loss": 5.6396,
"step": 6600
},
{
"epoch": 0.0020647732810215786,
"grad_norm": 17.375,
"learning_rate": 1.8678e-05,
"loss": 5.4356,
"step": 6610
},
{
"epoch": 0.0020678969924905976,
"grad_norm": 19.375,
"learning_rate": 1.8676000000000002e-05,
"loss": 5.4307,
"step": 6620
},
{
"epoch": 0.0020710207039596165,
"grad_norm": 15.3125,
"learning_rate": 1.8674e-05,
"loss": 5.7456,
"step": 6630
},
{
"epoch": 0.0020741444154286355,
"grad_norm": 18.625,
"learning_rate": 1.8672e-05,
"loss": 5.5859,
"step": 6640
},
{
"epoch": 0.002077268126897655,
"grad_norm": 16.5,
"learning_rate": 1.8670000000000003e-05,
"loss": 5.5676,
"step": 6650
},
{
"epoch": 0.002080391838366674,
"grad_norm": 17.875,
"learning_rate": 1.8668e-05,
"loss": 5.6596,
"step": 6660
},
{
"epoch": 0.002083515549835693,
"grad_norm": 16.375,
"learning_rate": 1.8666e-05,
"loss": 5.1641,
"step": 6670
},
{
"epoch": 0.0020866392613047118,
"grad_norm": 17.125,
"learning_rate": 1.8664000000000003e-05,
"loss": 5.5482,
"step": 6680
},
{
"epoch": 0.0020897629727737307,
"grad_norm": 17.375,
"learning_rate": 1.8662000000000002e-05,
"loss": 5.5011,
"step": 6690
},
{
"epoch": 0.0020928866842427497,
"grad_norm": 15.4375,
"learning_rate": 1.866e-05,
"loss": 5.7944,
"step": 6700
},
{
"epoch": 0.002096010395711769,
"grad_norm": 18.625,
"learning_rate": 1.8658e-05,
"loss": 5.5027,
"step": 6710
},
{
"epoch": 0.002099134107180788,
"grad_norm": 14.0625,
"learning_rate": 1.8656000000000002e-05,
"loss": 5.1993,
"step": 6720
},
{
"epoch": 0.002102257818649807,
"grad_norm": 38.25,
"learning_rate": 1.8654e-05,
"loss": 5.6462,
"step": 6730
},
{
"epoch": 0.002105381530118826,
"grad_norm": 12.8125,
"learning_rate": 1.8652e-05,
"loss": 5.1891,
"step": 6740
},
{
"epoch": 0.002108505241587845,
"grad_norm": 11.875,
"learning_rate": 1.8650000000000003e-05,
"loss": 5.4365,
"step": 6750
},
{
"epoch": 0.002111628953056864,
"grad_norm": 16.125,
"learning_rate": 1.8648000000000002e-05,
"loss": 5.3265,
"step": 6760
},
{
"epoch": 0.0021147526645258832,
"grad_norm": 18.25,
"learning_rate": 1.8646e-05,
"loss": 5.4414,
"step": 6770
},
{
"epoch": 0.002117876375994902,
"grad_norm": 13.8125,
"learning_rate": 1.8644000000000003e-05,
"loss": 5.447,
"step": 6780
},
{
"epoch": 0.002121000087463921,
"grad_norm": 13.5625,
"learning_rate": 1.8642000000000002e-05,
"loss": 5.6893,
"step": 6790
},
{
"epoch": 0.00212412379893294,
"grad_norm": 16.75,
"learning_rate": 1.864e-05,
"loss": 5.7147,
"step": 6800
},
{
"epoch": 0.002127247510401959,
"grad_norm": 13.125,
"learning_rate": 1.8638e-05,
"loss": 4.9141,
"step": 6810
},
{
"epoch": 0.002130371221870978,
"grad_norm": 17.625,
"learning_rate": 1.8636e-05,
"loss": 5.6681,
"step": 6820
},
{
"epoch": 0.0021334949333399974,
"grad_norm": 13.4375,
"learning_rate": 1.8634e-05,
"loss": 5.753,
"step": 6830
},
{
"epoch": 0.0021366186448090164,
"grad_norm": 14.5,
"learning_rate": 1.8632e-05,
"loss": 5.7403,
"step": 6840
},
{
"epoch": 0.0021397423562780353,
"grad_norm": 15.5,
"learning_rate": 1.8630000000000003e-05,
"loss": 5.6975,
"step": 6850
},
{
"epoch": 0.0021428660677470543,
"grad_norm": 19.125,
"learning_rate": 1.8628000000000002e-05,
"loss": 5.2514,
"step": 6860
},
{
"epoch": 0.0021459897792160732,
"grad_norm": 12.4375,
"learning_rate": 1.8626e-05,
"loss": 5.7343,
"step": 6870
},
{
"epoch": 0.002149113490685092,
"grad_norm": 12.75,
"learning_rate": 1.8624000000000003e-05,
"loss": 5.7297,
"step": 6880
},
{
"epoch": 0.0021522372021541116,
"grad_norm": 19.0,
"learning_rate": 1.8622000000000002e-05,
"loss": 5.4692,
"step": 6890
},
{
"epoch": 0.0021553609136231305,
"grad_norm": 18.75,
"learning_rate": 1.862e-05,
"loss": 5.602,
"step": 6900
},
{
"epoch": 0.0021584846250921495,
"grad_norm": 11.8125,
"learning_rate": 1.8618e-05,
"loss": 5.4007,
"step": 6910
},
{
"epoch": 0.0021616083365611685,
"grad_norm": 16.125,
"learning_rate": 1.8616e-05,
"loss": 5.4039,
"step": 6920
},
{
"epoch": 0.0021647320480301874,
"grad_norm": 18.625,
"learning_rate": 1.8614e-05,
"loss": 4.9848,
"step": 6930
},
{
"epoch": 0.0021678557594992064,
"grad_norm": 19.875,
"learning_rate": 1.8612e-05,
"loss": 5.3236,
"step": 6940
},
{
"epoch": 0.0021709794709682258,
"grad_norm": 15.3125,
"learning_rate": 1.8610000000000003e-05,
"loss": 5.5269,
"step": 6950
},
{
"epoch": 0.0021741031824372447,
"grad_norm": 18.125,
"learning_rate": 1.8608000000000002e-05,
"loss": 5.515,
"step": 6960
},
{
"epoch": 0.0021772268939062637,
"grad_norm": 19.125,
"learning_rate": 1.8606e-05,
"loss": 5.392,
"step": 6970
},
{
"epoch": 0.0021803506053752826,
"grad_norm": 15.6875,
"learning_rate": 1.8604000000000003e-05,
"loss": 5.2582,
"step": 6980
},
{
"epoch": 0.0021834743168443016,
"grad_norm": 14.25,
"learning_rate": 1.8602000000000002e-05,
"loss": 5.7263,
"step": 6990
},
{
"epoch": 0.0021865980283133205,
"grad_norm": 15.0625,
"learning_rate": 1.86e-05,
"loss": 5.7057,
"step": 7000
},
{
"epoch": 0.00218972173978234,
"grad_norm": 14.9375,
"learning_rate": 1.8598e-05,
"loss": 5.0865,
"step": 7010
},
{
"epoch": 0.002192845451251359,
"grad_norm": 13.125,
"learning_rate": 1.8596e-05,
"loss": 5.5477,
"step": 7020
},
{
"epoch": 0.002195969162720378,
"grad_norm": 20.25,
"learning_rate": 1.8594000000000002e-05,
"loss": 5.3479,
"step": 7030
},
{
"epoch": 0.002199092874189397,
"grad_norm": 16.25,
"learning_rate": 1.8592e-05,
"loss": 5.7431,
"step": 7040
},
{
"epoch": 0.0022022165856584158,
"grad_norm": 14.375,
"learning_rate": 1.859e-05,
"loss": 5.6489,
"step": 7050
},
{
"epoch": 0.002205340297127435,
"grad_norm": 13.0,
"learning_rate": 1.8588000000000002e-05,
"loss": 5.7156,
"step": 7060
},
{
"epoch": 0.002208464008596454,
"grad_norm": 11.5,
"learning_rate": 1.8586e-05,
"loss": 5.2813,
"step": 7070
},
{
"epoch": 0.002211587720065473,
"grad_norm": 13.5625,
"learning_rate": 1.8584000000000004e-05,
"loss": 5.5629,
"step": 7080
},
{
"epoch": 0.002214711431534492,
"grad_norm": 16.125,
"learning_rate": 1.8582000000000003e-05,
"loss": 5.2464,
"step": 7090
},
{
"epoch": 0.002217835143003511,
"grad_norm": 12.6875,
"learning_rate": 1.858e-05,
"loss": 5.327,
"step": 7100
},
{
"epoch": 0.00222095885447253,
"grad_norm": 15.3125,
"learning_rate": 1.8578e-05,
"loss": 5.5996,
"step": 7110
},
{
"epoch": 0.0022240825659415493,
"grad_norm": 21.25,
"learning_rate": 1.8576e-05,
"loss": 5.7259,
"step": 7120
},
{
"epoch": 0.0022272062774105683,
"grad_norm": 10.75,
"learning_rate": 1.8574000000000002e-05,
"loss": 5.6215,
"step": 7130
},
{
"epoch": 0.0022303299888795872,
"grad_norm": 11.625,
"learning_rate": 1.8572e-05,
"loss": 5.7345,
"step": 7140
},
{
"epoch": 0.002233453700348606,
"grad_norm": 11.4375,
"learning_rate": 1.857e-05,
"loss": 5.5052,
"step": 7150
},
{
"epoch": 0.002236577411817625,
"grad_norm": 15.375,
"learning_rate": 1.8568000000000002e-05,
"loss": 5.2982,
"step": 7160
},
{
"epoch": 0.002239701123286644,
"grad_norm": 16.125,
"learning_rate": 1.8566e-05,
"loss": 5.6919,
"step": 7170
},
{
"epoch": 0.0022428248347556635,
"grad_norm": 12.375,
"learning_rate": 1.8564e-05,
"loss": 5.7517,
"step": 7180
},
{
"epoch": 0.0022459485462246825,
"grad_norm": 20.125,
"learning_rate": 1.8562000000000003e-05,
"loss": 5.4861,
"step": 7190
},
{
"epoch": 0.0022490722576937014,
"grad_norm": 20.875,
"learning_rate": 1.8560000000000002e-05,
"loss": 5.6446,
"step": 7200
},
{
"epoch": 0.0022521959691627204,
"grad_norm": 12.0,
"learning_rate": 1.8558e-05,
"loss": 5.4446,
"step": 7210
},
{
"epoch": 0.0022553196806317393,
"grad_norm": 15.0625,
"learning_rate": 1.8556e-05,
"loss": 5.4171,
"step": 7220
},
{
"epoch": 0.0022584433921007583,
"grad_norm": 14.4375,
"learning_rate": 1.8554000000000002e-05,
"loss": 5.2308,
"step": 7230
},
{
"epoch": 0.0022615671035697777,
"grad_norm": 18.75,
"learning_rate": 1.8552e-05,
"loss": 5.7176,
"step": 7240
},
{
"epoch": 0.0022646908150387966,
"grad_norm": 18.375,
"learning_rate": 1.855e-05,
"loss": 5.3464,
"step": 7250
},
{
"epoch": 0.0022678145265078156,
"grad_norm": 15.375,
"learning_rate": 1.8548000000000003e-05,
"loss": 5.7412,
"step": 7260
},
{
"epoch": 0.0022709382379768345,
"grad_norm": 21.0,
"learning_rate": 1.8546e-05,
"loss": 5.434,
"step": 7270
},
{
"epoch": 0.0022740619494458535,
"grad_norm": 12.375,
"learning_rate": 1.8544e-05,
"loss": 5.572,
"step": 7280
},
{
"epoch": 0.0022771856609148725,
"grad_norm": 13.125,
"learning_rate": 1.8542000000000003e-05,
"loss": 5.4166,
"step": 7290
},
{
"epoch": 0.002280309372383892,
"grad_norm": 13.25,
"learning_rate": 1.8540000000000002e-05,
"loss": 5.7315,
"step": 7300
},
{
"epoch": 0.002283433083852911,
"grad_norm": 12.25,
"learning_rate": 1.8538e-05,
"loss": 5.0891,
"step": 7310
},
{
"epoch": 0.0022865567953219298,
"grad_norm": 18.125,
"learning_rate": 1.8536e-05,
"loss": 5.3659,
"step": 7320
},
{
"epoch": 0.0022896805067909487,
"grad_norm": 19.125,
"learning_rate": 1.8534000000000002e-05,
"loss": 5.2582,
"step": 7330
},
{
"epoch": 0.0022928042182599677,
"grad_norm": 15.8125,
"learning_rate": 1.8532e-05,
"loss": 5.6617,
"step": 7340
},
{
"epoch": 0.0022959279297289866,
"grad_norm": 15.1875,
"learning_rate": 1.853e-05,
"loss": 5.2852,
"step": 7350
},
{
"epoch": 0.002299051641198006,
"grad_norm": 17.625,
"learning_rate": 1.8528000000000003e-05,
"loss": 5.1184,
"step": 7360
},
{
"epoch": 0.002302175352667025,
"grad_norm": 14.6875,
"learning_rate": 1.8526000000000002e-05,
"loss": 5.4026,
"step": 7370
},
{
"epoch": 0.002305299064136044,
"grad_norm": 14.4375,
"learning_rate": 1.8524e-05,
"loss": 5.5041,
"step": 7380
},
{
"epoch": 0.002308422775605063,
"grad_norm": 19.5,
"learning_rate": 1.8522000000000003e-05,
"loss": 5.3296,
"step": 7390
},
{
"epoch": 0.002311546487074082,
"grad_norm": 16.75,
"learning_rate": 1.8520000000000002e-05,
"loss": 5.39,
"step": 7400
},
{
"epoch": 0.002314670198543101,
"grad_norm": 18.25,
"learning_rate": 1.8518e-05,
"loss": 5.6362,
"step": 7410
},
{
"epoch": 0.00231779391001212,
"grad_norm": 13.625,
"learning_rate": 1.8516e-05,
"loss": 5.3206,
"step": 7420
},
{
"epoch": 0.002320917621481139,
"grad_norm": 18.875,
"learning_rate": 1.8514e-05,
"loss": 5.7378,
"step": 7430
},
{
"epoch": 0.002324041332950158,
"grad_norm": 18.0,
"learning_rate": 1.8512e-05,
"loss": 5.7009,
"step": 7440
},
{
"epoch": 0.002327165044419177,
"grad_norm": 14.75,
"learning_rate": 1.851e-05,
"loss": 5.6616,
"step": 7450
},
{
"epoch": 0.002330288755888196,
"grad_norm": 17.125,
"learning_rate": 1.8508000000000003e-05,
"loss": 5.4864,
"step": 7460
},
{
"epoch": 0.002333412467357215,
"grad_norm": 15.0,
"learning_rate": 1.8506000000000002e-05,
"loss": 5.725,
"step": 7470
},
{
"epoch": 0.0023365361788262344,
"grad_norm": 17.5,
"learning_rate": 1.8504e-05,
"loss": 5.4064,
"step": 7480
},
{
"epoch": 0.0023396598902952533,
"grad_norm": 13.6875,
"learning_rate": 1.8502000000000003e-05,
"loss": 5.4624,
"step": 7490
},
{
"epoch": 0.0023427836017642723,
"grad_norm": 18.625,
"learning_rate": 1.8500000000000002e-05,
"loss": 5.4293,
"step": 7500
},
{
"epoch": 0.0023459073132332912,
"grad_norm": 11.1875,
"learning_rate": 1.8498e-05,
"loss": 5.7245,
"step": 7510
},
{
"epoch": 0.00234903102470231,
"grad_norm": 15.75,
"learning_rate": 1.8496e-05,
"loss": 5.6408,
"step": 7520
},
{
"epoch": 0.002352154736171329,
"grad_norm": 18.875,
"learning_rate": 1.8494e-05,
"loss": 5.6747,
"step": 7530
},
{
"epoch": 0.0023552784476403485,
"grad_norm": 13.875,
"learning_rate": 1.8492000000000002e-05,
"loss": 5.0168,
"step": 7540
},
{
"epoch": 0.0023584021591093675,
"grad_norm": 12.125,
"learning_rate": 1.849e-05,
"loss": 5.0578,
"step": 7550
},
{
"epoch": 0.0023615258705783865,
"grad_norm": 22.875,
"learning_rate": 1.8488e-05,
"loss": 5.3784,
"step": 7560
},
{
"epoch": 0.0023646495820474054,
"grad_norm": 15.5,
"learning_rate": 1.8486000000000002e-05,
"loss": 5.3212,
"step": 7570
},
{
"epoch": 0.0023677732935164244,
"grad_norm": 11.9375,
"learning_rate": 1.8484e-05,
"loss": 5.4473,
"step": 7580
},
{
"epoch": 0.0023708970049854433,
"grad_norm": 13.9375,
"learning_rate": 1.8482000000000004e-05,
"loss": 5.2285,
"step": 7590
},
{
"epoch": 0.0023740207164544627,
"grad_norm": 21.5,
"learning_rate": 1.8480000000000003e-05,
"loss": 5.6203,
"step": 7600
},
{
"epoch": 0.0023771444279234817,
"grad_norm": 21.875,
"learning_rate": 1.8478e-05,
"loss": 5.5058,
"step": 7610
},
{
"epoch": 0.0023802681393925006,
"grad_norm": 15.8125,
"learning_rate": 1.8476e-05,
"loss": 5.4249,
"step": 7620
},
{
"epoch": 0.0023833918508615196,
"grad_norm": 13.0625,
"learning_rate": 1.8474e-05,
"loss": 5.3753,
"step": 7630
},
{
"epoch": 0.0023865155623305385,
"grad_norm": 18.0,
"learning_rate": 1.8472000000000002e-05,
"loss": 5.4768,
"step": 7640
},
{
"epoch": 0.0023896392737995575,
"grad_norm": 16.0,
"learning_rate": 1.847e-05,
"loss": 5.4579,
"step": 7650
},
{
"epoch": 0.002392762985268577,
"grad_norm": 14.8125,
"learning_rate": 1.8468e-05,
"loss": 5.6076,
"step": 7660
},
{
"epoch": 0.002395886696737596,
"grad_norm": 15.4375,
"learning_rate": 1.8466000000000002e-05,
"loss": 5.3755,
"step": 7670
},
{
"epoch": 0.002399010408206615,
"grad_norm": 15.9375,
"learning_rate": 1.8464e-05,
"loss": 5.3464,
"step": 7680
},
{
"epoch": 0.0024021341196756338,
"grad_norm": 16.25,
"learning_rate": 1.8462000000000004e-05,
"loss": 5.4139,
"step": 7690
},
{
"epoch": 0.0024052578311446527,
"grad_norm": 54.25,
"learning_rate": 1.8460000000000003e-05,
"loss": 5.1928,
"step": 7700
},
{
"epoch": 0.0024083815426136717,
"grad_norm": 11.75,
"learning_rate": 1.8458000000000002e-05,
"loss": 5.8011,
"step": 7710
},
{
"epoch": 0.002411505254082691,
"grad_norm": 11.9375,
"learning_rate": 1.8456e-05,
"loss": 5.7197,
"step": 7720
},
{
"epoch": 0.00241462896555171,
"grad_norm": 15.8125,
"learning_rate": 1.8454e-05,
"loss": 5.7809,
"step": 7730
},
{
"epoch": 0.002417752677020729,
"grad_norm": 18.75,
"learning_rate": 1.8452000000000002e-05,
"loss": 5.5839,
"step": 7740
},
{
"epoch": 0.002420876388489748,
"grad_norm": 14.5625,
"learning_rate": 1.845e-05,
"loss": 5.7785,
"step": 7750
},
{
"epoch": 0.002424000099958767,
"grad_norm": 20.5,
"learning_rate": 1.8448e-05,
"loss": 5.124,
"step": 7760
},
{
"epoch": 0.002427123811427786,
"grad_norm": 21.75,
"learning_rate": 1.8446000000000003e-05,
"loss": 5.7176,
"step": 7770
},
{
"epoch": 0.0024302475228968052,
"grad_norm": 16.25,
"learning_rate": 1.8444e-05,
"loss": 5.5727,
"step": 7780
},
{
"epoch": 0.002433371234365824,
"grad_norm": 19.0,
"learning_rate": 1.8442e-05,
"loss": 5.5922,
"step": 7790
},
{
"epoch": 0.002436494945834843,
"grad_norm": 15.9375,
"learning_rate": 1.8440000000000003e-05,
"loss": 5.4033,
"step": 7800
},
{
"epoch": 0.002439618657303862,
"grad_norm": 11.875,
"learning_rate": 1.8438000000000002e-05,
"loss": 5.3833,
"step": 7810
},
{
"epoch": 0.002442742368772881,
"grad_norm": 14.875,
"learning_rate": 1.8436e-05,
"loss": 5.5731,
"step": 7820
},
{
"epoch": 0.0024458660802419,
"grad_norm": 13.875,
"learning_rate": 1.8434e-05,
"loss": 5.2196,
"step": 7830
},
{
"epoch": 0.0024489897917109194,
"grad_norm": 17.0,
"learning_rate": 1.8432000000000002e-05,
"loss": 5.5524,
"step": 7840
},
{
"epoch": 0.0024521135031799384,
"grad_norm": 19.75,
"learning_rate": 1.843e-05,
"loss": 5.3533,
"step": 7850
},
{
"epoch": 0.0024552372146489573,
"grad_norm": 18.375,
"learning_rate": 1.8428e-05,
"loss": 5.2578,
"step": 7860
},
{
"epoch": 0.0024583609261179763,
"grad_norm": 13.0,
"learning_rate": 1.8426000000000003e-05,
"loss": 5.6817,
"step": 7870
},
{
"epoch": 0.0024614846375869952,
"grad_norm": 20.375,
"learning_rate": 1.8424000000000002e-05,
"loss": 5.6943,
"step": 7880
},
{
"epoch": 0.002464608349056014,
"grad_norm": 17.875,
"learning_rate": 1.8422e-05,
"loss": 5.7388,
"step": 7890
},
{
"epoch": 0.0024677320605250336,
"grad_norm": 14.9375,
"learning_rate": 1.8420000000000003e-05,
"loss": 5.6932,
"step": 7900
},
{
"epoch": 0.0024708557719940526,
"grad_norm": 15.75,
"learning_rate": 1.8418000000000002e-05,
"loss": 5.6659,
"step": 7910
},
{
"epoch": 0.0024739794834630715,
"grad_norm": 15.8125,
"learning_rate": 1.8416e-05,
"loss": 5.7062,
"step": 7920
},
{
"epoch": 0.0024771031949320905,
"grad_norm": 17.5,
"learning_rate": 1.8414e-05,
"loss": 5.4439,
"step": 7930
},
{
"epoch": 0.0024802269064011094,
"grad_norm": 17.875,
"learning_rate": 1.8412000000000003e-05,
"loss": 5.5437,
"step": 7940
},
{
"epoch": 0.0024833506178701284,
"grad_norm": 19.0,
"learning_rate": 1.841e-05,
"loss": 5.6529,
"step": 7950
},
{
"epoch": 0.0024864743293391478,
"grad_norm": 11.75,
"learning_rate": 1.8408e-05,
"loss": 5.5425,
"step": 7960
},
{
"epoch": 0.0024895980408081667,
"grad_norm": 19.375,
"learning_rate": 1.8406000000000003e-05,
"loss": 5.6127,
"step": 7970
},
{
"epoch": 0.0024927217522771857,
"grad_norm": 14.25,
"learning_rate": 1.8404000000000002e-05,
"loss": 5.5463,
"step": 7980
},
{
"epoch": 0.0024958454637462046,
"grad_norm": 11.1875,
"learning_rate": 1.8402e-05,
"loss": 5.6528,
"step": 7990
},
{
"epoch": 0.0024989691752152236,
"grad_norm": 16.125,
"learning_rate": 1.8400000000000003e-05,
"loss": 5.6768,
"step": 8000
},
{
"epoch": 0.0025020928866842426,
"grad_norm": 12.9375,
"learning_rate": 1.8398000000000002e-05,
"loss": 5.529,
"step": 8010
},
{
"epoch": 0.002505216598153262,
"grad_norm": 19.25,
"learning_rate": 1.8396e-05,
"loss": 5.6349,
"step": 8020
},
{
"epoch": 0.002508340309622281,
"grad_norm": 15.8125,
"learning_rate": 1.8394e-05,
"loss": 5.7431,
"step": 8030
},
{
"epoch": 0.0025114640210913,
"grad_norm": 18.125,
"learning_rate": 1.8392e-05,
"loss": 5.6143,
"step": 8040
},
{
"epoch": 0.002514587732560319,
"grad_norm": 25.75,
"learning_rate": 1.8390000000000002e-05,
"loss": 5.5138,
"step": 8050
},
{
"epoch": 0.0025177114440293378,
"grad_norm": 12.375,
"learning_rate": 1.8388e-05,
"loss": 5.5673,
"step": 8060
},
{
"epoch": 0.0025208351554983567,
"grad_norm": 12.875,
"learning_rate": 1.8386000000000003e-05,
"loss": 5.7416,
"step": 8070
},
{
"epoch": 0.002523958866967376,
"grad_norm": 13.9375,
"learning_rate": 1.8384000000000002e-05,
"loss": 5.6055,
"step": 8080
},
{
"epoch": 0.002527082578436395,
"grad_norm": 15.4375,
"learning_rate": 1.8382e-05,
"loss": 5.5338,
"step": 8090
},
{
"epoch": 0.002530206289905414,
"grad_norm": 24.75,
"learning_rate": 1.8380000000000004e-05,
"loss": 5.5553,
"step": 8100
},
{
"epoch": 0.002533330001374433,
"grad_norm": 16.625,
"learning_rate": 1.8378000000000003e-05,
"loss": 5.4739,
"step": 8110
},
{
"epoch": 0.002536453712843452,
"grad_norm": 20.25,
"learning_rate": 1.8376e-05,
"loss": 5.4292,
"step": 8120
},
{
"epoch": 0.002539577424312471,
"grad_norm": 22.5,
"learning_rate": 1.8374e-05,
"loss": 5.6956,
"step": 8130
},
{
"epoch": 0.0025427011357814903,
"grad_norm": 12.8125,
"learning_rate": 1.8372e-05,
"loss": 5.4388,
"step": 8140
},
{
"epoch": 0.0025458248472505093,
"grad_norm": 13.6875,
"learning_rate": 1.8370000000000002e-05,
"loss": 5.3834,
"step": 8150
},
{
"epoch": 0.002548948558719528,
"grad_norm": 17.625,
"learning_rate": 1.8368e-05,
"loss": 5.6013,
"step": 8160
},
{
"epoch": 0.002552072270188547,
"grad_norm": 13.1875,
"learning_rate": 1.8366e-05,
"loss": 5.6469,
"step": 8170
},
{
"epoch": 0.002555195981657566,
"grad_norm": 13.4375,
"learning_rate": 1.8364000000000002e-05,
"loss": 5.7029,
"step": 8180
},
{
"epoch": 0.002558319693126585,
"grad_norm": 16.75,
"learning_rate": 1.8362e-05,
"loss": 5.6941,
"step": 8190
},
{
"epoch": 0.0025614434045956045,
"grad_norm": 14.375,
"learning_rate": 1.8360000000000004e-05,
"loss": 5.6255,
"step": 8200
},
{
"epoch": 0.0025645671160646234,
"grad_norm": 12.8125,
"learning_rate": 1.8358000000000003e-05,
"loss": 5.5778,
"step": 8210
},
{
"epoch": 0.0025676908275336424,
"grad_norm": 13.0625,
"learning_rate": 1.8356000000000002e-05,
"loss": 5.7121,
"step": 8220
},
{
"epoch": 0.0025708145390026613,
"grad_norm": 18.5,
"learning_rate": 1.8354e-05,
"loss": 5.6337,
"step": 8230
},
{
"epoch": 0.0025739382504716803,
"grad_norm": 14.75,
"learning_rate": 1.8352e-05,
"loss": 5.781,
"step": 8240
},
{
"epoch": 0.0025770619619406993,
"grad_norm": 16.375,
"learning_rate": 1.8350000000000002e-05,
"loss": 5.7368,
"step": 8250
},
{
"epoch": 0.0025801856734097186,
"grad_norm": 14.4375,
"learning_rate": 1.8348e-05,
"loss": 5.614,
"step": 8260
},
{
"epoch": 0.0025833093848787376,
"grad_norm": 12.9375,
"learning_rate": 1.8346e-05,
"loss": 5.2018,
"step": 8270
},
{
"epoch": 0.0025864330963477566,
"grad_norm": 16.25,
"learning_rate": 1.8344000000000003e-05,
"loss": 5.5791,
"step": 8280
},
{
"epoch": 0.0025895568078167755,
"grad_norm": 13.3125,
"learning_rate": 1.8342e-05,
"loss": 5.7152,
"step": 8290
},
{
"epoch": 0.0025926805192857945,
"grad_norm": 20.5,
"learning_rate": 1.834e-05,
"loss": 5.6225,
"step": 8300
},
{
"epoch": 0.0025958042307548134,
"grad_norm": 18.625,
"learning_rate": 1.8338000000000003e-05,
"loss": 5.5868,
"step": 8310
},
{
"epoch": 0.002598927942223833,
"grad_norm": 15.625,
"learning_rate": 1.8336000000000002e-05,
"loss": 5.1618,
"step": 8320
},
{
"epoch": 0.0026020516536928518,
"grad_norm": 15.25,
"learning_rate": 1.8334e-05,
"loss": 5.6222,
"step": 8330
},
{
"epoch": 0.0026051753651618707,
"grad_norm": 14.5625,
"learning_rate": 1.8332e-05,
"loss": 5.441,
"step": 8340
},
{
"epoch": 0.0026082990766308897,
"grad_norm": 20.625,
"learning_rate": 1.8330000000000002e-05,
"loss": 5.8026,
"step": 8350
},
{
"epoch": 0.0026114227880999086,
"grad_norm": 16.0,
"learning_rate": 1.8328e-05,
"loss": 5.643,
"step": 8360
},
{
"epoch": 0.0026145464995689276,
"grad_norm": 18.5,
"learning_rate": 1.8326e-05,
"loss": 5.744,
"step": 8370
},
{
"epoch": 0.002617670211037947,
"grad_norm": 12.3125,
"learning_rate": 1.8324000000000003e-05,
"loss": 5.8405,
"step": 8380
},
{
"epoch": 0.002620793922506966,
"grad_norm": 14.125,
"learning_rate": 1.8322000000000002e-05,
"loss": 5.6224,
"step": 8390
},
{
"epoch": 0.002623917633975985,
"grad_norm": 14.3125,
"learning_rate": 1.832e-05,
"loss": 5.5866,
"step": 8400
},
{
"epoch": 0.002627041345445004,
"grad_norm": 14.875,
"learning_rate": 1.8318000000000003e-05,
"loss": 5.6274,
"step": 8410
},
{
"epoch": 0.002630165056914023,
"grad_norm": 18.375,
"learning_rate": 1.8316e-05,
"loss": 5.71,
"step": 8420
},
{
"epoch": 0.002633288768383042,
"grad_norm": 14.625,
"learning_rate": 1.8314e-05,
"loss": 5.8016,
"step": 8430
},
{
"epoch": 0.002636412479852061,
"grad_norm": 13.6875,
"learning_rate": 1.8312e-05,
"loss": 5.7552,
"step": 8440
},
{
"epoch": 0.00263953619132108,
"grad_norm": 15.8125,
"learning_rate": 1.8310000000000003e-05,
"loss": 5.7443,
"step": 8450
},
{
"epoch": 0.002642659902790099,
"grad_norm": 17.875,
"learning_rate": 1.8308e-05,
"loss": 5.7018,
"step": 8460
},
{
"epoch": 0.002645783614259118,
"grad_norm": 14.0,
"learning_rate": 1.8306e-05,
"loss": 5.7637,
"step": 8470
},
{
"epoch": 0.002648907325728137,
"grad_norm": 19.375,
"learning_rate": 1.8304000000000003e-05,
"loss": 5.6508,
"step": 8480
},
{
"epoch": 0.0026520310371971564,
"grad_norm": 11.5625,
"learning_rate": 1.8302000000000002e-05,
"loss": 5.3358,
"step": 8490
},
{
"epoch": 0.0026551547486661753,
"grad_norm": 13.3125,
"learning_rate": 1.83e-05,
"loss": 5.5337,
"step": 8500
},
{
"epoch": 0.0026582784601351943,
"grad_norm": 14.875,
"learning_rate": 1.8298000000000003e-05,
"loss": 5.5015,
"step": 8510
},
{
"epoch": 0.0026614021716042133,
"grad_norm": 15.3125,
"learning_rate": 1.8296e-05,
"loss": 5.3137,
"step": 8520
},
{
"epoch": 0.002664525883073232,
"grad_norm": 14.25,
"learning_rate": 1.8294e-05,
"loss": 5.6978,
"step": 8530
},
{
"epoch": 0.002667649594542251,
"grad_norm": 12.5625,
"learning_rate": 1.8292e-05,
"loss": 5.6982,
"step": 8540
},
{
"epoch": 0.0026707733060112706,
"grad_norm": 15.5,
"learning_rate": 1.8290000000000003e-05,
"loss": 5.6831,
"step": 8550
},
{
"epoch": 0.0026738970174802895,
"grad_norm": 16.875,
"learning_rate": 1.8288000000000002e-05,
"loss": 5.607,
"step": 8560
},
{
"epoch": 0.0026770207289493085,
"grad_norm": 15.3125,
"learning_rate": 1.8286e-05,
"loss": 5.4306,
"step": 8570
},
{
"epoch": 0.0026801444404183274,
"grad_norm": 15.8125,
"learning_rate": 1.8284000000000003e-05,
"loss": 5.5547,
"step": 8580
},
{
"epoch": 0.0026832681518873464,
"grad_norm": 15.25,
"learning_rate": 1.8282000000000002e-05,
"loss": 5.6788,
"step": 8590
},
{
"epoch": 0.0026863918633563653,
"grad_norm": 14.4375,
"learning_rate": 1.828e-05,
"loss": 5.7335,
"step": 8600
},
{
"epoch": 0.0026895155748253847,
"grad_norm": 17.375,
"learning_rate": 1.8278000000000004e-05,
"loss": 5.6141,
"step": 8610
},
{
"epoch": 0.0026926392862944037,
"grad_norm": 15.25,
"learning_rate": 1.8276e-05,
"loss": 5.1071,
"step": 8620
},
{
"epoch": 0.0026957629977634226,
"grad_norm": 13.875,
"learning_rate": 1.8274e-05,
"loss": 5.3496,
"step": 8630
},
{
"epoch": 0.0026988867092324416,
"grad_norm": 13.25,
"learning_rate": 1.8272e-05,
"loss": 5.6896,
"step": 8640
},
{
"epoch": 0.0027020104207014606,
"grad_norm": 14.8125,
"learning_rate": 1.827e-05,
"loss": 5.6356,
"step": 8650
},
{
"epoch": 0.0027051341321704795,
"grad_norm": 14.0,
"learning_rate": 1.8268000000000002e-05,
"loss": 5.3134,
"step": 8660
},
{
"epoch": 0.002708257843639499,
"grad_norm": 13.8125,
"learning_rate": 1.8266e-05,
"loss": 5.7228,
"step": 8670
},
{
"epoch": 0.002711381555108518,
"grad_norm": 12.0,
"learning_rate": 1.8264000000000003e-05,
"loss": 5.7139,
"step": 8680
},
{
"epoch": 0.002714505266577537,
"grad_norm": 16.625,
"learning_rate": 1.8262000000000002e-05,
"loss": 5.5275,
"step": 8690
},
{
"epoch": 0.0027176289780465558,
"grad_norm": 17.0,
"learning_rate": 1.826e-05,
"loss": 5.4343,
"step": 8700
},
{
"epoch": 0.0027207526895155747,
"grad_norm": 17.5,
"learning_rate": 1.8258000000000004e-05,
"loss": 5.7115,
"step": 8710
},
{
"epoch": 0.0027238764009845937,
"grad_norm": 18.875,
"learning_rate": 1.8256e-05,
"loss": 5.7312,
"step": 8720
},
{
"epoch": 0.002727000112453613,
"grad_norm": 13.4375,
"learning_rate": 1.8254000000000002e-05,
"loss": 5.783,
"step": 8730
},
{
"epoch": 0.002730123823922632,
"grad_norm": 19.875,
"learning_rate": 1.8252e-05,
"loss": 5.6312,
"step": 8740
},
{
"epoch": 0.002733247535391651,
"grad_norm": 18.625,
"learning_rate": 1.825e-05,
"loss": 5.285,
"step": 8750
},
{
"epoch": 0.00273637124686067,
"grad_norm": 17.125,
"learning_rate": 1.8248000000000002e-05,
"loss": 5.5014,
"step": 8760
},
{
"epoch": 0.002739494958329689,
"grad_norm": 14.9375,
"learning_rate": 1.8246e-05,
"loss": 5.6598,
"step": 8770
},
{
"epoch": 0.002742618669798708,
"grad_norm": 15.75,
"learning_rate": 1.8244e-05,
"loss": 5.323,
"step": 8780
},
{
"epoch": 0.0027457423812677273,
"grad_norm": 17.25,
"learning_rate": 1.8242000000000003e-05,
"loss": 5.7593,
"step": 8790
},
{
"epoch": 0.002748866092736746,
"grad_norm": 16.875,
"learning_rate": 1.824e-05,
"loss": 5.2561,
"step": 8800
},
{
"epoch": 0.002751989804205765,
"grad_norm": 16.625,
"learning_rate": 1.8238000000000004e-05,
"loss": 5.7641,
"step": 8810
},
{
"epoch": 0.002755113515674784,
"grad_norm": 12.5,
"learning_rate": 1.8236000000000003e-05,
"loss": 5.2643,
"step": 8820
},
{
"epoch": 0.002758237227143803,
"grad_norm": 18.75,
"learning_rate": 1.8234000000000002e-05,
"loss": 5.7293,
"step": 8830
},
{
"epoch": 0.002761360938612822,
"grad_norm": 16.5,
"learning_rate": 1.8232e-05,
"loss": 5.5185,
"step": 8840
},
{
"epoch": 0.0027644846500818414,
"grad_norm": 14.1875,
"learning_rate": 1.823e-05,
"loss": 5.2863,
"step": 8850
},
{
"epoch": 0.0027676083615508604,
"grad_norm": 13.5625,
"learning_rate": 1.8228000000000002e-05,
"loss": 5.6614,
"step": 8860
},
{
"epoch": 0.0027707320730198793,
"grad_norm": 13.3125,
"learning_rate": 1.8226e-05,
"loss": 5.6269,
"step": 8870
},
{
"epoch": 0.0027738557844888983,
"grad_norm": 17.375,
"learning_rate": 1.8224e-05,
"loss": 5.6308,
"step": 8880
},
{
"epoch": 0.0027769794959579173,
"grad_norm": 12.5,
"learning_rate": 1.8222000000000003e-05,
"loss": 5.5156,
"step": 8890
},
{
"epoch": 0.0027801032074269362,
"grad_norm": 11.5,
"learning_rate": 1.8220000000000002e-05,
"loss": 5.7255,
"step": 8900
},
{
"epoch": 0.0027832269188959556,
"grad_norm": 14.625,
"learning_rate": 1.8218e-05,
"loss": 5.7292,
"step": 8910
},
{
"epoch": 0.0027863506303649746,
"grad_norm": 17.625,
"learning_rate": 1.8216000000000003e-05,
"loss": 5.709,
"step": 8920
},
{
"epoch": 0.0027894743418339935,
"grad_norm": 15.1875,
"learning_rate": 1.8214000000000002e-05,
"loss": 5.5746,
"step": 8930
},
{
"epoch": 0.0027925980533030125,
"grad_norm": 14.5,
"learning_rate": 1.8212e-05,
"loss": 5.7584,
"step": 8940
},
{
"epoch": 0.0027957217647720314,
"grad_norm": 12.3125,
"learning_rate": 1.821e-05,
"loss": 5.6617,
"step": 8950
},
{
"epoch": 0.0027988454762410504,
"grad_norm": 18.5,
"learning_rate": 1.8208000000000003e-05,
"loss": 5.4972,
"step": 8960
},
{
"epoch": 0.00280196918771007,
"grad_norm": 13.625,
"learning_rate": 1.8206e-05,
"loss": 5.7059,
"step": 8970
},
{
"epoch": 0.0028050928991790887,
"grad_norm": 13.125,
"learning_rate": 1.8204e-05,
"loss": 5.6355,
"step": 8980
},
{
"epoch": 0.0028082166106481077,
"grad_norm": 18.875,
"learning_rate": 1.8202000000000003e-05,
"loss": 5.3638,
"step": 8990
},
{
"epoch": 0.0028113403221171267,
"grad_norm": 24.375,
"learning_rate": 1.8200000000000002e-05,
"loss": 5.4894,
"step": 9000
},
{
"epoch": 0.0028144640335861456,
"grad_norm": 30.75,
"learning_rate": 1.8198e-05,
"loss": 5.4093,
"step": 9010
},
{
"epoch": 0.0028175877450551646,
"grad_norm": 14.6875,
"learning_rate": 1.8196000000000003e-05,
"loss": 5.5249,
"step": 9020
},
{
"epoch": 0.002820711456524184,
"grad_norm": 14.5625,
"learning_rate": 1.8194e-05,
"loss": 5.5573,
"step": 9030
},
{
"epoch": 0.002823835167993203,
"grad_norm": 14.8125,
"learning_rate": 1.8192e-05,
"loss": 5.7256,
"step": 9040
},
{
"epoch": 0.002826958879462222,
"grad_norm": 13.1875,
"learning_rate": 1.819e-05,
"loss": 5.5939,
"step": 9050
},
{
"epoch": 0.002830082590931241,
"grad_norm": 16.75,
"learning_rate": 1.8188000000000003e-05,
"loss": 5.6293,
"step": 9060
},
{
"epoch": 0.00283320630240026,
"grad_norm": 16.75,
"learning_rate": 1.8186000000000002e-05,
"loss": 5.7227,
"step": 9070
},
{
"epoch": 0.0028363300138692787,
"grad_norm": 14.6875,
"learning_rate": 1.8184e-05,
"loss": 5.6747,
"step": 9080
},
{
"epoch": 0.002839453725338298,
"grad_norm": 17.875,
"learning_rate": 1.8182000000000003e-05,
"loss": 5.7045,
"step": 9090
},
{
"epoch": 0.002842577436807317,
"grad_norm": 15.125,
"learning_rate": 1.8180000000000002e-05,
"loss": 5.7102,
"step": 9100
},
{
"epoch": 0.002845701148276336,
"grad_norm": 11.8125,
"learning_rate": 1.8178e-05,
"loss": 5.5916,
"step": 9110
},
{
"epoch": 0.002848824859745355,
"grad_norm": 16.625,
"learning_rate": 1.8176000000000004e-05,
"loss": 5.6392,
"step": 9120
},
{
"epoch": 0.002851948571214374,
"grad_norm": 17.25,
"learning_rate": 1.8174e-05,
"loss": 5.4425,
"step": 9130
},
{
"epoch": 0.002855072282683393,
"grad_norm": 13.8125,
"learning_rate": 1.8172e-05,
"loss": 5.6837,
"step": 9140
},
{
"epoch": 0.0028581959941524123,
"grad_norm": 14.375,
"learning_rate": 1.817e-05,
"loss": 5.6339,
"step": 9150
},
{
"epoch": 0.0028613197056214313,
"grad_norm": 22.75,
"learning_rate": 1.8168e-05,
"loss": 5.3867,
"step": 9160
},
{
"epoch": 0.0028644434170904502,
"grad_norm": 13.75,
"learning_rate": 1.8166000000000002e-05,
"loss": 5.5268,
"step": 9170
},
{
"epoch": 0.002867567128559469,
"grad_norm": 29.375,
"learning_rate": 1.8164e-05,
"loss": 5.5336,
"step": 9180
},
{
"epoch": 0.002870690840028488,
"grad_norm": 18.125,
"learning_rate": 1.8162000000000003e-05,
"loss": 5.3273,
"step": 9190
},
{
"epoch": 0.002873814551497507,
"grad_norm": 13.75,
"learning_rate": 1.8160000000000002e-05,
"loss": 5.7044,
"step": 9200
},
{
"epoch": 0.0028769382629665265,
"grad_norm": 16.5,
"learning_rate": 1.8158e-05,
"loss": 5.5071,
"step": 9210
},
{
"epoch": 0.0028800619744355454,
"grad_norm": 21.625,
"learning_rate": 1.8156000000000004e-05,
"loss": 5.5489,
"step": 9220
},
{
"epoch": 0.0028831856859045644,
"grad_norm": 18.125,
"learning_rate": 1.8154e-05,
"loss": 5.7425,
"step": 9230
},
{
"epoch": 0.0028863093973735834,
"grad_norm": 19.625,
"learning_rate": 1.8152000000000002e-05,
"loss": 5.4189,
"step": 9240
},
{
"epoch": 0.0028894331088426023,
"grad_norm": 15.6875,
"learning_rate": 1.815e-05,
"loss": 5.607,
"step": 9250
},
{
"epoch": 0.0028925568203116213,
"grad_norm": 14.0625,
"learning_rate": 1.8148e-05,
"loss": 5.441,
"step": 9260
},
{
"epoch": 0.0028956805317806407,
"grad_norm": 16.375,
"learning_rate": 1.8146000000000002e-05,
"loss": 5.7174,
"step": 9270
},
{
"epoch": 0.0028988042432496596,
"grad_norm": 14.0,
"learning_rate": 1.8144e-05,
"loss": 5.736,
"step": 9280
},
{
"epoch": 0.0029019279547186786,
"grad_norm": 16.0,
"learning_rate": 1.8142000000000004e-05,
"loss": 5.5643,
"step": 9290
},
{
"epoch": 0.0029050516661876975,
"grad_norm": 14.875,
"learning_rate": 1.8140000000000003e-05,
"loss": 5.6589,
"step": 9300
},
{
"epoch": 0.0029081753776567165,
"grad_norm": 16.25,
"learning_rate": 1.8138e-05,
"loss": 5.5534,
"step": 9310
},
{
"epoch": 0.0029112990891257354,
"grad_norm": 15.375,
"learning_rate": 1.8136000000000004e-05,
"loss": 5.5841,
"step": 9320
},
{
"epoch": 0.002914422800594755,
"grad_norm": 14.875,
"learning_rate": 1.8134e-05,
"loss": 5.6218,
"step": 9330
},
{
"epoch": 0.002917546512063774,
"grad_norm": 12.8125,
"learning_rate": 1.8132000000000002e-05,
"loss": 5.5351,
"step": 9340
},
{
"epoch": 0.0029206702235327927,
"grad_norm": 14.3125,
"learning_rate": 1.813e-05,
"loss": 5.5595,
"step": 9350
},
{
"epoch": 0.0029237939350018117,
"grad_norm": 12.375,
"learning_rate": 1.8128e-05,
"loss": 5.6869,
"step": 9360
},
{
"epoch": 0.0029269176464708307,
"grad_norm": 18.125,
"learning_rate": 1.8126000000000002e-05,
"loss": 5.4235,
"step": 9370
},
{
"epoch": 0.0029300413579398496,
"grad_norm": 18.125,
"learning_rate": 1.8124e-05,
"loss": 5.583,
"step": 9380
},
{
"epoch": 0.002933165069408869,
"grad_norm": 13.375,
"learning_rate": 1.8122e-05,
"loss": 5.6377,
"step": 9390
},
{
"epoch": 0.002936288780877888,
"grad_norm": 14.9375,
"learning_rate": 1.8120000000000003e-05,
"loss": 5.5843,
"step": 9400
},
{
"epoch": 0.002939412492346907,
"grad_norm": 18.875,
"learning_rate": 1.8118000000000002e-05,
"loss": 5.4062,
"step": 9410
},
{
"epoch": 0.002942536203815926,
"grad_norm": 13.375,
"learning_rate": 1.8116000000000004e-05,
"loss": 5.6681,
"step": 9420
},
{
"epoch": 0.002945659915284945,
"grad_norm": 11.6875,
"learning_rate": 1.8114e-05,
"loss": 5.5529,
"step": 9430
},
{
"epoch": 0.002948783626753964,
"grad_norm": 20.875,
"learning_rate": 1.8112000000000002e-05,
"loss": 5.6737,
"step": 9440
},
{
"epoch": 0.002951907338222983,
"grad_norm": 13.5625,
"learning_rate": 1.811e-05,
"loss": 5.6382,
"step": 9450
},
{
"epoch": 0.002955031049692002,
"grad_norm": 17.75,
"learning_rate": 1.8108e-05,
"loss": 5.5691,
"step": 9460
},
{
"epoch": 0.002958154761161021,
"grad_norm": 13.8125,
"learning_rate": 1.8106000000000002e-05,
"loss": 5.5743,
"step": 9470
},
{
"epoch": 0.00296127847263004,
"grad_norm": 14.3125,
"learning_rate": 1.8104e-05,
"loss": 5.3204,
"step": 9480
},
{
"epoch": 0.002964402184099059,
"grad_norm": 14.625,
"learning_rate": 1.8102e-05,
"loss": 5.6013,
"step": 9490
},
{
"epoch": 0.002967525895568078,
"grad_norm": 14.5625,
"learning_rate": 1.8100000000000003e-05,
"loss": 5.619,
"step": 9500
},
{
"epoch": 0.0029706496070370974,
"grad_norm": 20.75,
"learning_rate": 1.8098000000000002e-05,
"loss": 5.2921,
"step": 9510
},
{
"epoch": 0.0029737733185061163,
"grad_norm": 15.6875,
"learning_rate": 1.8096e-05,
"loss": 5.5717,
"step": 9520
},
{
"epoch": 0.0029768970299751353,
"grad_norm": 12.0,
"learning_rate": 1.8094e-05,
"loss": 5.2209,
"step": 9530
},
{
"epoch": 0.0029800207414441542,
"grad_norm": 16.5,
"learning_rate": 1.8092000000000002e-05,
"loss": 5.6648,
"step": 9540
},
{
"epoch": 0.002983144452913173,
"grad_norm": 14.125,
"learning_rate": 1.809e-05,
"loss": 5.5017,
"step": 9550
},
{
"epoch": 0.002986268164382192,
"grad_norm": 14.625,
"learning_rate": 1.8088e-05,
"loss": 5.6437,
"step": 9560
},
{
"epoch": 0.0029893918758512115,
"grad_norm": 12.875,
"learning_rate": 1.8086000000000003e-05,
"loss": 5.4502,
"step": 9570
},
{
"epoch": 0.0029925155873202305,
"grad_norm": 17.75,
"learning_rate": 1.8084e-05,
"loss": 5.4951,
"step": 9580
},
{
"epoch": 0.0029956392987892494,
"grad_norm": 14.5625,
"learning_rate": 1.8082e-05,
"loss": 5.5379,
"step": 9590
},
{
"epoch": 0.0029987630102582684,
"grad_norm": 12.0625,
"learning_rate": 1.8080000000000003e-05,
"loss": 5.3454,
"step": 9600
},
{
"epoch": 0.0030018867217272874,
"grad_norm": 12.25,
"learning_rate": 1.8078000000000002e-05,
"loss": 5.4042,
"step": 9610
},
{
"epoch": 0.0030050104331963063,
"grad_norm": 21.25,
"learning_rate": 1.8076e-05,
"loss": 5.1163,
"step": 9620
},
{
"epoch": 0.0030081341446653257,
"grad_norm": 13.4375,
"learning_rate": 1.8074e-05,
"loss": 5.622,
"step": 9630
},
{
"epoch": 0.0030112578561343447,
"grad_norm": 17.5,
"learning_rate": 1.8072e-05,
"loss": 5.6973,
"step": 9640
},
{
"epoch": 0.0030143815676033636,
"grad_norm": 18.5,
"learning_rate": 1.807e-05,
"loss": 5.6359,
"step": 9650
},
{
"epoch": 0.0030175052790723826,
"grad_norm": 17.625,
"learning_rate": 1.8068e-05,
"loss": 5.3276,
"step": 9660
},
{
"epoch": 0.0030206289905414015,
"grad_norm": 14.5625,
"learning_rate": 1.8066000000000003e-05,
"loss": 5.3333,
"step": 9670
},
{
"epoch": 0.0030237527020104205,
"grad_norm": 13.375,
"learning_rate": 1.8064000000000002e-05,
"loss": 5.6219,
"step": 9680
},
{
"epoch": 0.00302687641347944,
"grad_norm": 12.8125,
"learning_rate": 1.8062e-05,
"loss": 5.6096,
"step": 9690
},
{
"epoch": 0.003030000124948459,
"grad_norm": 17.5,
"learning_rate": 1.8060000000000003e-05,
"loss": 5.4421,
"step": 9700
},
{
"epoch": 0.003033123836417478,
"grad_norm": 18.25,
"learning_rate": 1.8058000000000002e-05,
"loss": 5.6122,
"step": 9710
},
{
"epoch": 0.0030362475478864968,
"grad_norm": 14.1875,
"learning_rate": 1.8056e-05,
"loss": 5.5429,
"step": 9720
},
{
"epoch": 0.0030393712593555157,
"grad_norm": 16.875,
"learning_rate": 1.8054e-05,
"loss": 5.7268,
"step": 9730
},
{
"epoch": 0.0030424949708245347,
"grad_norm": 13.8125,
"learning_rate": 1.8052e-05,
"loss": 5.5956,
"step": 9740
},
{
"epoch": 0.003045618682293554,
"grad_norm": 11.1875,
"learning_rate": 1.805e-05,
"loss": 5.8065,
"step": 9750
},
{
"epoch": 0.003048742393762573,
"grad_norm": 12.875,
"learning_rate": 1.8048e-05,
"loss": 5.445,
"step": 9760
},
{
"epoch": 0.003051866105231592,
"grad_norm": 18.0,
"learning_rate": 1.8046e-05,
"loss": 5.5073,
"step": 9770
},
{
"epoch": 0.003054989816700611,
"grad_norm": 12.25,
"learning_rate": 1.8044000000000002e-05,
"loss": 5.6303,
"step": 9780
},
{
"epoch": 0.00305811352816963,
"grad_norm": 12.375,
"learning_rate": 1.8042e-05,
"loss": 5.4904,
"step": 9790
},
{
"epoch": 0.003061237239638649,
"grad_norm": 14.375,
"learning_rate": 1.8040000000000003e-05,
"loss": 5.455,
"step": 9800
},
{
"epoch": 0.0030643609511076682,
"grad_norm": 14.1875,
"learning_rate": 1.8038000000000002e-05,
"loss": 5.5094,
"step": 9810
},
{
"epoch": 0.003067484662576687,
"grad_norm": 18.25,
"learning_rate": 1.8036e-05,
"loss": 5.7042,
"step": 9820
},
{
"epoch": 0.003070608374045706,
"grad_norm": 15.25,
"learning_rate": 1.8034e-05,
"loss": 5.6813,
"step": 9830
},
{
"epoch": 0.003073732085514725,
"grad_norm": 11.75,
"learning_rate": 1.8032e-05,
"loss": 5.5455,
"step": 9840
},
{
"epoch": 0.003076855796983744,
"grad_norm": 12.6875,
"learning_rate": 1.8030000000000002e-05,
"loss": 5.7319,
"step": 9850
},
{
"epoch": 0.0030799795084527634,
"grad_norm": 11.4375,
"learning_rate": 1.8028e-05,
"loss": 5.7984,
"step": 9860
},
{
"epoch": 0.0030831032199217824,
"grad_norm": 12.0625,
"learning_rate": 1.8026e-05,
"loss": 5.6482,
"step": 9870
},
{
"epoch": 0.0030862269313908014,
"grad_norm": 12.75,
"learning_rate": 1.8024000000000002e-05,
"loss": 5.8138,
"step": 9880
},
{
"epoch": 0.0030893506428598203,
"grad_norm": 15.1875,
"learning_rate": 1.8022e-05,
"loss": 5.542,
"step": 9890
},
{
"epoch": 0.0030924743543288393,
"grad_norm": 15.4375,
"learning_rate": 1.802e-05,
"loss": 5.3519,
"step": 9900
},
{
"epoch": 0.0030955980657978582,
"grad_norm": 12.875,
"learning_rate": 1.8018000000000003e-05,
"loss": 5.6849,
"step": 9910
},
{
"epoch": 0.0030987217772668776,
"grad_norm": 16.25,
"learning_rate": 1.8016e-05,
"loss": 5.5522,
"step": 9920
},
{
"epoch": 0.0031018454887358966,
"grad_norm": 17.875,
"learning_rate": 1.8014e-05,
"loss": 5.3861,
"step": 9930
},
{
"epoch": 0.0031049692002049155,
"grad_norm": 18.125,
"learning_rate": 1.8012e-05,
"loss": 5.7198,
"step": 9940
},
{
"epoch": 0.0031080929116739345,
"grad_norm": 18.375,
"learning_rate": 1.8010000000000002e-05,
"loss": 5.722,
"step": 9950
},
{
"epoch": 0.0031112166231429535,
"grad_norm": 9.5625,
"learning_rate": 1.8008e-05,
"loss": 5.4564,
"step": 9960
},
{
"epoch": 0.0031143403346119724,
"grad_norm": 12.75,
"learning_rate": 1.8006e-05,
"loss": 5.6358,
"step": 9970
},
{
"epoch": 0.003117464046080992,
"grad_norm": 13.0,
"learning_rate": 1.8004000000000002e-05,
"loss": 5.6605,
"step": 9980
},
{
"epoch": 0.0031205877575500108,
"grad_norm": 14.375,
"learning_rate": 1.8002e-05,
"loss": 5.6693,
"step": 9990
},
{
"epoch": 0.0031237114690190297,
"grad_norm": 17.875,
"learning_rate": 1.8e-05,
"loss": 5.1117,
"step": 10000
},
{
"epoch": 0.00315494858370922,
"grad_norm": 12.6875,
"learning_rate": 1.798e-05,
"loss": 5.5697,
"step": 10100
},
{
"epoch": 0.00318618569839941,
"grad_norm": 17.75,
"learning_rate": 1.796e-05,
"loss": 5.6353,
"step": 10200
},
{
"epoch": 0.0032174228130896006,
"grad_norm": 14.1875,
"learning_rate": 1.794e-05,
"loss": 5.5588,
"step": 10300
},
{
"epoch": 0.003248659927779791,
"grad_norm": 21.0,
"learning_rate": 1.792e-05,
"loss": 5.5504,
"step": 10400
},
{
"epoch": 0.003279897042469981,
"grad_norm": 20.5,
"learning_rate": 1.79e-05,
"loss": 5.5269,
"step": 10500
},
{
"epoch": 0.0033111341571601715,
"grad_norm": 20.625,
"learning_rate": 1.788e-05,
"loss": 5.4473,
"step": 10600
},
{
"epoch": 0.003342371271850362,
"grad_norm": 13.4375,
"learning_rate": 1.7860000000000002e-05,
"loss": 5.5347,
"step": 10700
},
{
"epoch": 0.003373608386540552,
"grad_norm": 17.625,
"learning_rate": 1.7840000000000002e-05,
"loss": 5.4918,
"step": 10800
},
{
"epoch": 0.0034048455012307423,
"grad_norm": 17.75,
"learning_rate": 1.7820000000000002e-05,
"loss": 5.5585,
"step": 10900
},
{
"epoch": 0.0034360826159209328,
"grad_norm": 20.125,
"learning_rate": 1.7800000000000002e-05,
"loss": 5.578,
"step": 11000
},
{
"epoch": 0.0034673197306111228,
"grad_norm": 17.25,
"learning_rate": 1.7780000000000003e-05,
"loss": 5.5814,
"step": 11100
},
{
"epoch": 0.003498556845301313,
"grad_norm": 18.25,
"learning_rate": 1.7760000000000003e-05,
"loss": 5.5033,
"step": 11200
},
{
"epoch": 0.0035297939599915036,
"grad_norm": 12.8125,
"learning_rate": 1.7740000000000003e-05,
"loss": 5.5739,
"step": 11300
},
{
"epoch": 0.0035610310746816936,
"grad_norm": 13.75,
"learning_rate": 1.7720000000000003e-05,
"loss": 5.5183,
"step": 11400
},
{
"epoch": 0.003592268189371884,
"grad_norm": 17.375,
"learning_rate": 1.77e-05,
"loss": 5.58,
"step": 11500
},
{
"epoch": 0.0036235053040620745,
"grad_norm": 15.875,
"learning_rate": 1.768e-05,
"loss": 5.6626,
"step": 11600
},
{
"epoch": 0.0036547424187522645,
"grad_norm": 15.5,
"learning_rate": 1.766e-05,
"loss": 5.4935,
"step": 11700
},
{
"epoch": 0.003685979533442455,
"grad_norm": 17.625,
"learning_rate": 1.764e-05,
"loss": 5.516,
"step": 11800
},
{
"epoch": 0.0037172166481326454,
"grad_norm": 14.75,
"learning_rate": 1.762e-05,
"loss": 5.564,
"step": 11900
},
{
"epoch": 0.0037484537628228354,
"grad_norm": 16.125,
"learning_rate": 1.76e-05,
"loss": 5.5759,
"step": 12000
},
{
"epoch": 0.003779690877513026,
"grad_norm": 18.25,
"learning_rate": 1.758e-05,
"loss": 5.5458,
"step": 12100
},
{
"epoch": 0.0038109279922032163,
"grad_norm": 20.75,
"learning_rate": 1.756e-05,
"loss": 5.6255,
"step": 12200
},
{
"epoch": 0.0038421651068934063,
"grad_norm": 15.5,
"learning_rate": 1.754e-05,
"loss": 5.5217,
"step": 12300
},
{
"epoch": 0.0038734022215835967,
"grad_norm": 15.3125,
"learning_rate": 1.752e-05,
"loss": 5.5526,
"step": 12400
},
{
"epoch": 0.003904639336273787,
"grad_norm": 18.625,
"learning_rate": 1.7500000000000002e-05,
"loss": 5.5548,
"step": 12500
},
{
"epoch": 0.003935876450963978,
"grad_norm": 15.4375,
"learning_rate": 1.7480000000000002e-05,
"loss": 5.5379,
"step": 12600
},
{
"epoch": 0.003967113565654168,
"grad_norm": 19.0,
"learning_rate": 1.7460000000000002e-05,
"loss": 5.5492,
"step": 12700
},
{
"epoch": 0.003998350680344358,
"grad_norm": 16.875,
"learning_rate": 1.7440000000000002e-05,
"loss": 5.5787,
"step": 12800
},
{
"epoch": 0.004029587795034548,
"grad_norm": 10.3125,
"learning_rate": 1.7420000000000003e-05,
"loss": 5.4924,
"step": 12900
},
{
"epoch": 0.0040608249097247385,
"grad_norm": 14.375,
"learning_rate": 1.7400000000000003e-05,
"loss": 5.4351,
"step": 13000
},
{
"epoch": 0.004092062024414929,
"grad_norm": 17.25,
"learning_rate": 1.7380000000000003e-05,
"loss": 5.5044,
"step": 13100
},
{
"epoch": 0.004123299139105119,
"grad_norm": 13.9375,
"learning_rate": 1.736e-05,
"loss": 5.5039,
"step": 13200
},
{
"epoch": 0.00415453625379531,
"grad_norm": 15.4375,
"learning_rate": 1.734e-05,
"loss": 5.5401,
"step": 13300
},
{
"epoch": 0.004185773368485499,
"grad_norm": 15.375,
"learning_rate": 1.732e-05,
"loss": 5.6314,
"step": 13400
},
{
"epoch": 0.00421701048317569,
"grad_norm": 13.875,
"learning_rate": 1.73e-05,
"loss": 5.6265,
"step": 13500
},
{
"epoch": 0.00424824759786588,
"grad_norm": 14.375,
"learning_rate": 1.728e-05,
"loss": 5.6575,
"step": 13600
},
{
"epoch": 0.004279484712556071,
"grad_norm": 13.125,
"learning_rate": 1.726e-05,
"loss": 5.5269,
"step": 13700
},
{
"epoch": 0.004310721827246261,
"grad_norm": 16.375,
"learning_rate": 1.724e-05,
"loss": 5.5832,
"step": 13800
},
{
"epoch": 0.0043419589419364515,
"grad_norm": 15.25,
"learning_rate": 1.722e-05,
"loss": 5.5344,
"step": 13900
},
{
"epoch": 0.004373196056626641,
"grad_norm": 13.875,
"learning_rate": 1.72e-05,
"loss": 5.6058,
"step": 14000
},
{
"epoch": 0.0044044331713168315,
"grad_norm": 13.25,
"learning_rate": 1.718e-05,
"loss": 5.5807,
"step": 14100
},
{
"epoch": 0.004435670286007022,
"grad_norm": 15.4375,
"learning_rate": 1.7160000000000002e-05,
"loss": 5.569,
"step": 14200
},
{
"epoch": 0.004466907400697212,
"grad_norm": 15.625,
"learning_rate": 1.7140000000000002e-05,
"loss": 5.5629,
"step": 14300
},
{
"epoch": 0.004498144515387403,
"grad_norm": 14.3125,
"learning_rate": 1.7120000000000002e-05,
"loss": 5.4952,
"step": 14400
},
{
"epoch": 0.004529381630077593,
"grad_norm": 12.25,
"learning_rate": 1.7100000000000002e-05,
"loss": 5.567,
"step": 14500
},
{
"epoch": 0.004560618744767784,
"grad_norm": 15.5625,
"learning_rate": 1.7080000000000002e-05,
"loss": 5.6412,
"step": 14600
},
{
"epoch": 0.004591855859457973,
"grad_norm": 12.8125,
"learning_rate": 1.7060000000000003e-05,
"loss": 5.548,
"step": 14700
},
{
"epoch": 0.004623092974148164,
"grad_norm": 18.875,
"learning_rate": 1.704e-05,
"loss": 5.5444,
"step": 14800
},
{
"epoch": 0.004654330088838354,
"grad_norm": 14.1875,
"learning_rate": 1.702e-05,
"loss": 5.5596,
"step": 14900
},
{
"epoch": 0.004685567203528545,
"grad_norm": 11.125,
"learning_rate": 1.7e-05,
"loss": 5.6135,
"step": 15000
}
],
"logging_steps": 100,
"max_steps": 100000,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 5000,
"total_flos": 2.028851316129792e+18,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}