qwen3-emo / trainer_state.json
jasong03's picture
Upload folder using huggingface_hub
6996e37 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 500,
"global_step": 5942,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0033664366268305,
"grad_norm": 61.75,
"learning_rate": 7.563025210084034e-08,
"loss": 4.5144,
"step": 10
},
{
"epoch": 0.006732873253661,
"grad_norm": 62.0,
"learning_rate": 1.596638655462185e-07,
"loss": 4.5034,
"step": 20
},
{
"epoch": 0.0100993098804915,
"grad_norm": 62.75,
"learning_rate": 2.4369747899159664e-07,
"loss": 4.5244,
"step": 30
},
{
"epoch": 0.013465746507322,
"grad_norm": 62.75,
"learning_rate": 3.2773109243697485e-07,
"loss": 4.4946,
"step": 40
},
{
"epoch": 0.0168321831341525,
"grad_norm": 62.25,
"learning_rate": 4.1176470588235295e-07,
"loss": 4.4875,
"step": 50
},
{
"epoch": 0.020198619760983,
"grad_norm": 61.0,
"learning_rate": 4.957983193277312e-07,
"loss": 4.4683,
"step": 60
},
{
"epoch": 0.0235650563878135,
"grad_norm": 61.25,
"learning_rate": 5.798319327731094e-07,
"loss": 4.4164,
"step": 70
},
{
"epoch": 0.026931493014644,
"grad_norm": 57.5,
"learning_rate": 6.638655462184875e-07,
"loss": 4.3726,
"step": 80
},
{
"epoch": 0.0302979296414745,
"grad_norm": 55.0,
"learning_rate": 7.478991596638656e-07,
"loss": 4.2941,
"step": 90
},
{
"epoch": 0.033664366268305,
"grad_norm": 55.0,
"learning_rate": 8.319327731092438e-07,
"loss": 4.2208,
"step": 100
},
{
"epoch": 0.0370308028951355,
"grad_norm": 54.0,
"learning_rate": 9.159663865546219e-07,
"loss": 4.1239,
"step": 110
},
{
"epoch": 0.040397239521966,
"grad_norm": 51.5,
"learning_rate": 1.0000000000000002e-06,
"loss": 4.0198,
"step": 120
},
{
"epoch": 0.0437636761487965,
"grad_norm": 49.0,
"learning_rate": 1.0840336134453784e-06,
"loss": 3.8576,
"step": 130
},
{
"epoch": 0.047130112775627,
"grad_norm": 50.25,
"learning_rate": 1.1680672268907564e-06,
"loss": 3.6773,
"step": 140
},
{
"epoch": 0.0504965494024575,
"grad_norm": 43.75,
"learning_rate": 1.2521008403361346e-06,
"loss": 3.3933,
"step": 150
},
{
"epoch": 0.053862986029288,
"grad_norm": 52.25,
"learning_rate": 1.3361344537815128e-06,
"loss": 3.0487,
"step": 160
},
{
"epoch": 0.0572294226561185,
"grad_norm": 40.0,
"learning_rate": 1.4201680672268908e-06,
"loss": 2.6898,
"step": 170
},
{
"epoch": 0.060595859282949,
"grad_norm": 39.75,
"learning_rate": 1.504201680672269e-06,
"loss": 2.452,
"step": 180
},
{
"epoch": 0.0639622959097795,
"grad_norm": 39.75,
"learning_rate": 1.5882352941176472e-06,
"loss": 2.3426,
"step": 190
},
{
"epoch": 0.06732873253661,
"grad_norm": 40.75,
"learning_rate": 1.6722689075630252e-06,
"loss": 2.2247,
"step": 200
},
{
"epoch": 0.0706951691634405,
"grad_norm": 40.25,
"learning_rate": 1.7563025210084034e-06,
"loss": 2.147,
"step": 210
},
{
"epoch": 0.074061605790271,
"grad_norm": 37.0,
"learning_rate": 1.8403361344537818e-06,
"loss": 2.083,
"step": 220
},
{
"epoch": 0.0774280424171015,
"grad_norm": 22.0,
"learning_rate": 1.92436974789916e-06,
"loss": 2.0281,
"step": 230
},
{
"epoch": 0.080794479043932,
"grad_norm": 10.8125,
"learning_rate": 2.008403361344538e-06,
"loss": 1.9858,
"step": 240
},
{
"epoch": 0.0841609156707625,
"grad_norm": 9.75,
"learning_rate": 2.0924369747899164e-06,
"loss": 1.9059,
"step": 250
},
{
"epoch": 0.087527352297593,
"grad_norm": 9.375,
"learning_rate": 2.176470588235294e-06,
"loss": 1.8522,
"step": 260
},
{
"epoch": 0.0908937889244235,
"grad_norm": 9.25,
"learning_rate": 2.2605042016806724e-06,
"loss": 1.8367,
"step": 270
},
{
"epoch": 0.094260225551254,
"grad_norm": 9.625,
"learning_rate": 2.3445378151260506e-06,
"loss": 1.7908,
"step": 280
},
{
"epoch": 0.0976266621780845,
"grad_norm": 9.0625,
"learning_rate": 2.428571428571429e-06,
"loss": 1.7631,
"step": 290
},
{
"epoch": 0.100993098804915,
"grad_norm": 8.875,
"learning_rate": 2.512605042016807e-06,
"loss": 1.7427,
"step": 300
},
{
"epoch": 0.1043595354317455,
"grad_norm": 10.125,
"learning_rate": 2.596638655462185e-06,
"loss": 1.7185,
"step": 310
},
{
"epoch": 0.107725972058576,
"grad_norm": 9.625,
"learning_rate": 2.680672268907563e-06,
"loss": 1.6808,
"step": 320
},
{
"epoch": 0.1110924086854065,
"grad_norm": 9.5,
"learning_rate": 2.7647058823529416e-06,
"loss": 1.6309,
"step": 330
},
{
"epoch": 0.114458845312237,
"grad_norm": 9.875,
"learning_rate": 2.8487394957983194e-06,
"loss": 1.6266,
"step": 340
},
{
"epoch": 0.1178252819390675,
"grad_norm": 9.4375,
"learning_rate": 2.932773109243698e-06,
"loss": 1.5933,
"step": 350
},
{
"epoch": 0.121191718565898,
"grad_norm": 10.1875,
"learning_rate": 3.016806722689076e-06,
"loss": 1.5802,
"step": 360
},
{
"epoch": 0.1245581551927285,
"grad_norm": 11.25,
"learning_rate": 3.100840336134454e-06,
"loss": 1.5324,
"step": 370
},
{
"epoch": 0.127924591819559,
"grad_norm": 10.4375,
"learning_rate": 3.1848739495798318e-06,
"loss": 1.4969,
"step": 380
},
{
"epoch": 0.13129102844638948,
"grad_norm": 10.125,
"learning_rate": 3.2689075630252104e-06,
"loss": 1.4568,
"step": 390
},
{
"epoch": 0.13465746507322,
"grad_norm": 11.0,
"learning_rate": 3.352941176470588e-06,
"loss": 1.4319,
"step": 400
},
{
"epoch": 0.1380239017000505,
"grad_norm": 10.25,
"learning_rate": 3.436974789915967e-06,
"loss": 1.3945,
"step": 410
},
{
"epoch": 0.141390338326881,
"grad_norm": 10.375,
"learning_rate": 3.521008403361345e-06,
"loss": 1.3808,
"step": 420
},
{
"epoch": 0.1447567749537115,
"grad_norm": 10.5,
"learning_rate": 3.605042016806723e-06,
"loss": 1.3198,
"step": 430
},
{
"epoch": 0.148123211580542,
"grad_norm": 11.25,
"learning_rate": 3.6890756302521014e-06,
"loss": 1.3422,
"step": 440
},
{
"epoch": 0.15148964820737248,
"grad_norm": 9.625,
"learning_rate": 3.7731092436974792e-06,
"loss": 1.3092,
"step": 450
},
{
"epoch": 0.154856084834203,
"grad_norm": 9.3125,
"learning_rate": 3.857142857142858e-06,
"loss": 1.288,
"step": 460
},
{
"epoch": 0.1582225214610335,
"grad_norm": 9.875,
"learning_rate": 3.941176470588236e-06,
"loss": 1.2572,
"step": 470
},
{
"epoch": 0.161588958087864,
"grad_norm": 10.375,
"learning_rate": 4.025210084033614e-06,
"loss": 1.2318,
"step": 480
},
{
"epoch": 0.1649553947146945,
"grad_norm": 10.625,
"learning_rate": 4.109243697478992e-06,
"loss": 1.2322,
"step": 490
},
{
"epoch": 0.168321831341525,
"grad_norm": 9.5625,
"learning_rate": 4.19327731092437e-06,
"loss": 1.1904,
"step": 500
},
{
"epoch": 0.17168826796835548,
"grad_norm": 11.0,
"learning_rate": 4.2773109243697484e-06,
"loss": 1.1914,
"step": 510
},
{
"epoch": 0.175054704595186,
"grad_norm": 7.71875,
"learning_rate": 4.361344537815126e-06,
"loss": 1.2056,
"step": 520
},
{
"epoch": 0.1784211412220165,
"grad_norm": 6.875,
"learning_rate": 4.445378151260504e-06,
"loss": 1.1938,
"step": 530
},
{
"epoch": 0.181787577848847,
"grad_norm": 7.53125,
"learning_rate": 4.529411764705883e-06,
"loss": 1.166,
"step": 540
},
{
"epoch": 0.1851540144756775,
"grad_norm": 8.125,
"learning_rate": 4.613445378151261e-06,
"loss": 1.151,
"step": 550
},
{
"epoch": 0.188520451102508,
"grad_norm": 7.0625,
"learning_rate": 4.697478991596639e-06,
"loss": 1.2043,
"step": 560
},
{
"epoch": 0.19188688772933848,
"grad_norm": 6.625,
"learning_rate": 4.781512605042018e-06,
"loss": 1.1583,
"step": 570
},
{
"epoch": 0.195253324356169,
"grad_norm": 9.375,
"learning_rate": 4.8655462184873955e-06,
"loss": 1.1437,
"step": 580
},
{
"epoch": 0.1986197609829995,
"grad_norm": 8.3125,
"learning_rate": 4.949579831932774e-06,
"loss": 1.1319,
"step": 590
},
{
"epoch": 0.20198619760983,
"grad_norm": 6.625,
"learning_rate": 4.9999930958666185e-06,
"loss": 1.1393,
"step": 600
},
{
"epoch": 0.2053526342366605,
"grad_norm": 6.4375,
"learning_rate": 4.9999154248040175e-06,
"loss": 1.123,
"step": 610
},
{
"epoch": 0.208719070863491,
"grad_norm": 6.5,
"learning_rate": 4.999751455202276e-06,
"loss": 1.1206,
"step": 620
},
{
"epoch": 0.21208550749032148,
"grad_norm": 6.9375,
"learning_rate": 4.9995011927217214e-06,
"loss": 1.101,
"step": 630
},
{
"epoch": 0.215451944117152,
"grad_norm": 5.625,
"learning_rate": 4.99916464600156e-06,
"loss": 1.1162,
"step": 640
},
{
"epoch": 0.2188183807439825,
"grad_norm": 7.3125,
"learning_rate": 4.99874182665958e-06,
"loss": 1.1267,
"step": 650
},
{
"epoch": 0.222184817370813,
"grad_norm": 6.21875,
"learning_rate": 4.998232749291753e-06,
"loss": 1.1111,
"step": 660
},
{
"epoch": 0.2255512539976435,
"grad_norm": 7.84375,
"learning_rate": 4.997637431471726e-06,
"loss": 1.1055,
"step": 670
},
{
"epoch": 0.228917690624474,
"grad_norm": 5.875,
"learning_rate": 4.9969558937502185e-06,
"loss": 1.1354,
"step": 680
},
{
"epoch": 0.23228412725130448,
"grad_norm": 6.28125,
"learning_rate": 4.99618815965431e-06,
"loss": 1.0621,
"step": 690
},
{
"epoch": 0.235650563878135,
"grad_norm": 6.96875,
"learning_rate": 4.995334255686629e-06,
"loss": 1.0846,
"step": 700
},
{
"epoch": 0.2390170005049655,
"grad_norm": 6.8125,
"learning_rate": 4.9943942113244385e-06,
"loss": 1.0556,
"step": 710
},
{
"epoch": 0.242383437131796,
"grad_norm": 5.1875,
"learning_rate": 4.993368059018619e-06,
"loss": 1.0803,
"step": 720
},
{
"epoch": 0.2457498737586265,
"grad_norm": 5.53125,
"learning_rate": 4.992255834192548e-06,
"loss": 1.0619,
"step": 730
},
{
"epoch": 0.249116310385457,
"grad_norm": 6.6875,
"learning_rate": 4.991057575240872e-06,
"loss": 1.0449,
"step": 740
},
{
"epoch": 0.2524827470122875,
"grad_norm": 5.375,
"learning_rate": 4.989773323528192e-06,
"loss": 1.0383,
"step": 750
},
{
"epoch": 0.255849183639118,
"grad_norm": 6.21875,
"learning_rate": 4.988403123387626e-06,
"loss": 1.0606,
"step": 760
},
{
"epoch": 0.2592156202659485,
"grad_norm": 6.0,
"learning_rate": 4.986947022119283e-06,
"loss": 1.0262,
"step": 770
},
{
"epoch": 0.26258205689277897,
"grad_norm": 5.46875,
"learning_rate": 4.985405069988627e-06,
"loss": 1.0482,
"step": 780
},
{
"epoch": 0.2659484935196095,
"grad_norm": 5.71875,
"learning_rate": 4.983777320224745e-06,
"loss": 1.0149,
"step": 790
},
{
"epoch": 0.26931493014644,
"grad_norm": 6.3125,
"learning_rate": 4.9820638290185095e-06,
"loss": 1.0236,
"step": 800
},
{
"epoch": 0.2726813667732705,
"grad_norm": 5.25,
"learning_rate": 4.980264655520637e-06,
"loss": 1.0226,
"step": 810
},
{
"epoch": 0.276047803400101,
"grad_norm": 5.75,
"learning_rate": 4.978379861839647e-06,
"loss": 1.0322,
"step": 820
},
{
"epoch": 0.2794142400269315,
"grad_norm": 5.53125,
"learning_rate": 4.976409513039715e-06,
"loss": 1.0513,
"step": 830
},
{
"epoch": 0.282780676653762,
"grad_norm": 8.4375,
"learning_rate": 4.974353677138433e-06,
"loss": 1.0236,
"step": 840
},
{
"epoch": 0.2861471132805925,
"grad_norm": 5.40625,
"learning_rate": 4.972212425104455e-06,
"loss": 1.0078,
"step": 850
},
{
"epoch": 0.289513549907423,
"grad_norm": 6.5,
"learning_rate": 4.969985830855052e-06,
"loss": 1.0204,
"step": 860
},
{
"epoch": 0.2928799865342535,
"grad_norm": 6.375,
"learning_rate": 4.967673971253555e-06,
"loss": 1.0125,
"step": 870
},
{
"epoch": 0.296246423161084,
"grad_norm": 5.125,
"learning_rate": 4.965276926106707e-06,
"loss": 0.9777,
"step": 880
},
{
"epoch": 0.29961285978791447,
"grad_norm": 5.71875,
"learning_rate": 4.962794778161906e-06,
"loss": 1.0176,
"step": 890
},
{
"epoch": 0.30297929641474497,
"grad_norm": 4.46875,
"learning_rate": 4.960227613104346e-06,
"loss": 1.0243,
"step": 900
},
{
"epoch": 0.3063457330415755,
"grad_norm": 6.125,
"learning_rate": 4.957575519554062e-06,
"loss": 0.9973,
"step": 910
},
{
"epoch": 0.309712169668406,
"grad_norm": 6.875,
"learning_rate": 4.954838589062873e-06,
"loss": 1.0084,
"step": 920
},
{
"epoch": 0.3130786062952365,
"grad_norm": 5.34375,
"learning_rate": 4.952016916111216e-06,
"loss": 1.0132,
"step": 930
},
{
"epoch": 0.316445042922067,
"grad_norm": 6.40625,
"learning_rate": 4.949110598104887e-06,
"loss": 1.0178,
"step": 940
},
{
"epoch": 0.3198114795488975,
"grad_norm": 4.40625,
"learning_rate": 4.94611973537168e-06,
"loss": 0.9945,
"step": 950
},
{
"epoch": 0.323177916175728,
"grad_norm": 4.71875,
"learning_rate": 4.9430444311579215e-06,
"loss": 0.9948,
"step": 960
},
{
"epoch": 0.3265443528025585,
"grad_norm": 4.75,
"learning_rate": 4.9398847916249065e-06,
"loss": 0.9929,
"step": 970
},
{
"epoch": 0.329910789429389,
"grad_norm": 4.4375,
"learning_rate": 4.936640925845236e-06,
"loss": 1.004,
"step": 980
},
{
"epoch": 0.3332772260562195,
"grad_norm": 4.90625,
"learning_rate": 4.93331294579905e-06,
"loss": 1.001,
"step": 990
},
{
"epoch": 0.33664366268305,
"grad_norm": 4.71875,
"learning_rate": 4.929900966370161e-06,
"loss": 1.0096,
"step": 1000
},
{
"epoch": 0.34001009930988046,
"grad_norm": 5.59375,
"learning_rate": 4.9264051053420905e-06,
"loss": 0.9993,
"step": 1010
},
{
"epoch": 0.34337653593671097,
"grad_norm": 4.9375,
"learning_rate": 4.922825483393999e-06,
"loss": 1.0077,
"step": 1020
},
{
"epoch": 0.3467429725635415,
"grad_norm": 4.78125,
"learning_rate": 4.9191622240965275e-06,
"loss": 1.022,
"step": 1030
},
{
"epoch": 0.350109409190372,
"grad_norm": 4.5625,
"learning_rate": 4.9154154539075226e-06,
"loss": 1.0082,
"step": 1040
},
{
"epoch": 0.3534758458172025,
"grad_norm": 7.625,
"learning_rate": 4.911585302167678e-06,
"loss": 1.0103,
"step": 1050
},
{
"epoch": 0.356842282444033,
"grad_norm": 4.25,
"learning_rate": 4.907671901096065e-06,
"loss": 1.0075,
"step": 1060
},
{
"epoch": 0.3602087190708635,
"grad_norm": 4.09375,
"learning_rate": 4.903675385785575e-06,
"loss": 0.9907,
"step": 1070
},
{
"epoch": 0.363575155697694,
"grad_norm": 5.125,
"learning_rate": 4.8995958941982466e-06,
"loss": 0.9833,
"step": 1080
},
{
"epoch": 0.3669415923245245,
"grad_norm": 4.4375,
"learning_rate": 4.8954335671605094e-06,
"loss": 0.9823,
"step": 1090
},
{
"epoch": 0.370308028951355,
"grad_norm": 5.53125,
"learning_rate": 4.891188548358323e-06,
"loss": 0.9998,
"step": 1100
},
{
"epoch": 0.3736744655781855,
"grad_norm": 4.3125,
"learning_rate": 4.886860984332212e-06,
"loss": 0.9878,
"step": 1110
},
{
"epoch": 0.377040902205016,
"grad_norm": 3.875,
"learning_rate": 4.882451024472212e-06,
"loss": 0.9986,
"step": 1120
},
{
"epoch": 0.38040733883184646,
"grad_norm": 4.25,
"learning_rate": 4.877958821012712e-06,
"loss": 0.9927,
"step": 1130
},
{
"epoch": 0.38377377545867697,
"grad_norm": 4.65625,
"learning_rate": 4.873384529027194e-06,
"loss": 1.0192,
"step": 1140
},
{
"epoch": 0.3871402120855075,
"grad_norm": 5.34375,
"learning_rate": 4.8687283064228884e-06,
"loss": 0.9797,
"step": 1150
},
{
"epoch": 0.390506648712338,
"grad_norm": 4.5,
"learning_rate": 4.863990313935315e-06,
"loss": 0.9949,
"step": 1160
},
{
"epoch": 0.3938730853391685,
"grad_norm": 4.21875,
"learning_rate": 4.85917071512274e-06,
"loss": 0.9737,
"step": 1170
},
{
"epoch": 0.397239521965999,
"grad_norm": 4.03125,
"learning_rate": 4.854269676360526e-06,
"loss": 0.9788,
"step": 1180
},
{
"epoch": 0.4006059585928295,
"grad_norm": 5.03125,
"learning_rate": 4.849287366835391e-06,
"loss": 0.9911,
"step": 1190
},
{
"epoch": 0.40397239521966,
"grad_norm": 5.40625,
"learning_rate": 4.844223958539568e-06,
"loss": 0.9896,
"step": 1200
},
{
"epoch": 0.4073388318464905,
"grad_norm": 4.75,
"learning_rate": 4.839079626264863e-06,
"loss": 0.992,
"step": 1210
},
{
"epoch": 0.410705268473321,
"grad_norm": 4.5,
"learning_rate": 4.833854547596631e-06,
"loss": 0.9852,
"step": 1220
},
{
"epoch": 0.4140717051001515,
"grad_norm": 4.0625,
"learning_rate": 4.828548902907635e-06,
"loss": 0.9633,
"step": 1230
},
{
"epoch": 0.417438141726982,
"grad_norm": 4.25,
"learning_rate": 4.823162875351825e-06,
"loss": 0.9925,
"step": 1240
},
{
"epoch": 0.42080457835381246,
"grad_norm": 5.125,
"learning_rate": 4.8176966508580155e-06,
"loss": 0.9948,
"step": 1250
},
{
"epoch": 0.42417101498064297,
"grad_norm": 4.46875,
"learning_rate": 4.812150418123466e-06,
"loss": 0.9828,
"step": 1260
},
{
"epoch": 0.4275374516074735,
"grad_norm": 4.46875,
"learning_rate": 4.806524368607365e-06,
"loss": 0.9902,
"step": 1270
},
{
"epoch": 0.430903888234304,
"grad_norm": 4.03125,
"learning_rate": 4.800818696524226e-06,
"loss": 0.994,
"step": 1280
},
{
"epoch": 0.4342703248611345,
"grad_norm": 4.65625,
"learning_rate": 4.795033598837178e-06,
"loss": 1.0027,
"step": 1290
},
{
"epoch": 0.437636761487965,
"grad_norm": 6.34375,
"learning_rate": 4.789169275251168e-06,
"loss": 0.9776,
"step": 1300
},
{
"epoch": 0.4410031981147955,
"grad_norm": 4.59375,
"learning_rate": 4.783225928206071e-06,
"loss": 0.9684,
"step": 1310
},
{
"epoch": 0.444369634741626,
"grad_norm": 4.125,
"learning_rate": 4.777203762869691e-06,
"loss": 0.982,
"step": 1320
},
{
"epoch": 0.4477360713684565,
"grad_norm": 5.125,
"learning_rate": 4.771102987130691e-06,
"loss": 0.9965,
"step": 1330
},
{
"epoch": 0.451102507995287,
"grad_norm": 3.890625,
"learning_rate": 4.76492381159141e-06,
"loss": 0.9899,
"step": 1340
},
{
"epoch": 0.4544689446221175,
"grad_norm": 5.8125,
"learning_rate": 4.758666449560591e-06,
"loss": 0.9668,
"step": 1350
},
{
"epoch": 0.457835381248948,
"grad_norm": 4.4375,
"learning_rate": 4.752331117046025e-06,
"loss": 0.9829,
"step": 1360
},
{
"epoch": 0.46120181787577846,
"grad_norm": 3.96875,
"learning_rate": 4.7459180327470855e-06,
"loss": 0.9858,
"step": 1370
},
{
"epoch": 0.46456825450260897,
"grad_norm": 3.875,
"learning_rate": 4.739427418047184e-06,
"loss": 0.9645,
"step": 1380
},
{
"epoch": 0.4679346911294395,
"grad_norm": 3.765625,
"learning_rate": 4.732859497006129e-06,
"loss": 0.9701,
"step": 1390
},
{
"epoch": 0.47130112775627,
"grad_norm": 4.8125,
"learning_rate": 4.726214496352385e-06,
"loss": 0.9698,
"step": 1400
},
{
"epoch": 0.4746675643831005,
"grad_norm": 4.59375,
"learning_rate": 4.719492645475253e-06,
"loss": 0.9875,
"step": 1410
},
{
"epoch": 0.478034001009931,
"grad_norm": 3.78125,
"learning_rate": 4.7126941764169476e-06,
"loss": 0.9672,
"step": 1420
},
{
"epoch": 0.4814004376367615,
"grad_norm": 3.953125,
"learning_rate": 4.7058193238645875e-06,
"loss": 0.9866,
"step": 1430
},
{
"epoch": 0.484766874263592,
"grad_norm": 3.8125,
"learning_rate": 4.698868325142093e-06,
"loss": 0.9685,
"step": 1440
},
{
"epoch": 0.4881333108904225,
"grad_norm": 5.125,
"learning_rate": 4.6918414202019975e-06,
"loss": 0.9816,
"step": 1450
},
{
"epoch": 0.491499747517253,
"grad_norm": 4.625,
"learning_rate": 4.684738851617159e-06,
"loss": 0.9774,
"step": 1460
},
{
"epoch": 0.4948661841440835,
"grad_norm": 4.125,
"learning_rate": 4.677560864572389e-06,
"loss": 0.9955,
"step": 1470
},
{
"epoch": 0.498232620770914,
"grad_norm": 4.1875,
"learning_rate": 4.670307706855988e-06,
"loss": 0.9703,
"step": 1480
},
{
"epoch": 0.5015990573977445,
"grad_norm": 4.1875,
"learning_rate": 4.662979628851194e-06,
"loss": 0.9905,
"step": 1490
},
{
"epoch": 0.504965494024575,
"grad_norm": 4.6875,
"learning_rate": 4.655576883527535e-06,
"loss": 0.9872,
"step": 1500
},
{
"epoch": 0.5083319306514055,
"grad_norm": 3.96875,
"learning_rate": 4.648099726432099e-06,
"loss": 0.9651,
"step": 1510
},
{
"epoch": 0.511698367278236,
"grad_norm": 4.4375,
"learning_rate": 4.640548415680711e-06,
"loss": 0.993,
"step": 1520
},
{
"epoch": 0.5150648039050665,
"grad_norm": 4.75,
"learning_rate": 4.632923211949026e-06,
"loss": 0.9399,
"step": 1530
},
{
"epoch": 0.518431240531897,
"grad_norm": 4.15625,
"learning_rate": 4.625224378463524e-06,
"loss": 0.9764,
"step": 1540
},
{
"epoch": 0.5217976771587275,
"grad_norm": 3.75,
"learning_rate": 4.617452180992431e-06,
"loss": 0.9816,
"step": 1550
},
{
"epoch": 0.5251641137855579,
"grad_norm": 3.96875,
"learning_rate": 4.609606887836537e-06,
"loss": 0.9728,
"step": 1560
},
{
"epoch": 0.5285305504123885,
"grad_norm": 3.640625,
"learning_rate": 4.601688769819941e-06,
"loss": 0.9785,
"step": 1570
},
{
"epoch": 0.531896987039219,
"grad_norm": 3.796875,
"learning_rate": 4.593698100280696e-06,
"loss": 0.9698,
"step": 1580
},
{
"epoch": 0.5352634236660495,
"grad_norm": 4.875,
"learning_rate": 4.585635155061377e-06,
"loss": 0.9799,
"step": 1590
},
{
"epoch": 0.53862986029288,
"grad_norm": 3.984375,
"learning_rate": 4.577500212499558e-06,
"loss": 0.9846,
"step": 1600
},
{
"epoch": 0.5419962969197105,
"grad_norm": 4.65625,
"learning_rate": 4.569293553418203e-06,
"loss": 0.9895,
"step": 1610
},
{
"epoch": 0.545362733546541,
"grad_norm": 4.34375,
"learning_rate": 4.561015461115972e-06,
"loss": 0.9968,
"step": 1620
},
{
"epoch": 0.5487291701733715,
"grad_norm": 3.953125,
"learning_rate": 4.55266622135744e-06,
"loss": 0.9789,
"step": 1630
},
{
"epoch": 0.552095606800202,
"grad_norm": 4.40625,
"learning_rate": 4.544246122363237e-06,
"loss": 0.9759,
"step": 1640
},
{
"epoch": 0.5554620434270325,
"grad_norm": 4.53125,
"learning_rate": 4.535755454800093e-06,
"loss": 0.9587,
"step": 1650
},
{
"epoch": 0.558828480053863,
"grad_norm": 5.375,
"learning_rate": 4.527194511770806e-06,
"loss": 0.996,
"step": 1660
},
{
"epoch": 0.5621949166806934,
"grad_norm": 5.0,
"learning_rate": 4.518563588804125e-06,
"loss": 0.9935,
"step": 1670
},
{
"epoch": 0.565561353307524,
"grad_norm": 3.703125,
"learning_rate": 4.509862983844546e-06,
"loss": 0.9802,
"step": 1680
},
{
"epoch": 0.5689277899343544,
"grad_norm": 4.6875,
"learning_rate": 4.50109299724203e-06,
"loss": 0.9656,
"step": 1690
},
{
"epoch": 0.572294226561185,
"grad_norm": 4.46875,
"learning_rate": 4.492253931741632e-06,
"loss": 1.0113,
"step": 1700
},
{
"epoch": 0.5756606631880155,
"grad_norm": 6.65625,
"learning_rate": 4.483346092473049e-06,
"loss": 0.9868,
"step": 1710
},
{
"epoch": 0.579027099814846,
"grad_norm": 4.34375,
"learning_rate": 4.4743697869400916e-06,
"loss": 0.9677,
"step": 1720
},
{
"epoch": 0.5823935364416765,
"grad_norm": 4.65625,
"learning_rate": 4.465325325010063e-06,
"loss": 0.9912,
"step": 1730
},
{
"epoch": 0.585759973068507,
"grad_norm": 4.875,
"learning_rate": 4.456213018903069e-06,
"loss": 0.9797,
"step": 1740
},
{
"epoch": 0.5891264096953375,
"grad_norm": 4.1875,
"learning_rate": 4.44703318318123e-06,
"loss": 0.9665,
"step": 1750
},
{
"epoch": 0.592492846322168,
"grad_norm": 4.96875,
"learning_rate": 4.4377861347378334e-06,
"loss": 0.9932,
"step": 1760
},
{
"epoch": 0.5958592829489985,
"grad_norm": 4.15625,
"learning_rate": 4.428472192786387e-06,
"loss": 0.9954,
"step": 1770
},
{
"epoch": 0.5992257195758289,
"grad_norm": 4.21875,
"learning_rate": 4.4190916788496006e-06,
"loss": 0.961,
"step": 1780
},
{
"epoch": 0.6025921562026595,
"grad_norm": 5.0,
"learning_rate": 4.40964491674829e-06,
"loss": 0.972,
"step": 1790
},
{
"epoch": 0.6059585928294899,
"grad_norm": 4.65625,
"learning_rate": 4.400132232590194e-06,
"loss": 0.9745,
"step": 1800
},
{
"epoch": 0.6093250294563205,
"grad_norm": 4.375,
"learning_rate": 4.390553954758721e-06,
"loss": 0.973,
"step": 1810
},
{
"epoch": 0.612691466083151,
"grad_norm": 4.03125,
"learning_rate": 4.38091041390161e-06,
"loss": 0.9629,
"step": 1820
},
{
"epoch": 0.6160579027099815,
"grad_norm": 3.96875,
"learning_rate": 4.371201942919519e-06,
"loss": 0.9929,
"step": 1830
},
{
"epoch": 0.619424339336812,
"grad_norm": 3.96875,
"learning_rate": 4.3614288769545295e-06,
"loss": 0.9644,
"step": 1840
},
{
"epoch": 0.6227907759636425,
"grad_norm": 4.03125,
"learning_rate": 4.351591553378583e-06,
"loss": 0.9694,
"step": 1850
},
{
"epoch": 0.626157212590473,
"grad_norm": 4.125,
"learning_rate": 4.34169031178183e-06,
"loss": 0.9779,
"step": 1860
},
{
"epoch": 0.6295236492173035,
"grad_norm": 4.125,
"learning_rate": 4.331725493960904e-06,
"loss": 0.9779,
"step": 1870
},
{
"epoch": 0.632890085844134,
"grad_norm": 4.21875,
"learning_rate": 4.321697443907133e-06,
"loss": 0.9634,
"step": 1880
},
{
"epoch": 0.6362565224709645,
"grad_norm": 4.71875,
"learning_rate": 4.311606507794658e-06,
"loss": 0.967,
"step": 1890
},
{
"epoch": 0.639622959097795,
"grad_norm": 4.6875,
"learning_rate": 4.301453033968478e-06,
"loss": 0.9589,
"step": 1900
},
{
"epoch": 0.6429893957246254,
"grad_norm": 3.828125,
"learning_rate": 4.291237372932435e-06,
"loss": 0.9805,
"step": 1910
},
{
"epoch": 0.646355832351456,
"grad_norm": 4.90625,
"learning_rate": 4.28095987733711e-06,
"loss": 0.9919,
"step": 1920
},
{
"epoch": 0.6497222689782864,
"grad_norm": 6.5,
"learning_rate": 4.270620901967647e-06,
"loss": 0.977,
"step": 1930
},
{
"epoch": 0.653088705605117,
"grad_norm": 3.90625,
"learning_rate": 4.260220803731508e-06,
"loss": 0.9591,
"step": 1940
},
{
"epoch": 0.6564551422319475,
"grad_norm": 3.96875,
"learning_rate": 4.249759941646153e-06,
"loss": 0.9746,
"step": 1950
},
{
"epoch": 0.659821578858778,
"grad_norm": 4.90625,
"learning_rate": 4.239238676826645e-06,
"loss": 0.9604,
"step": 1960
},
{
"epoch": 0.6631880154856085,
"grad_norm": 4.15625,
"learning_rate": 4.228657372473184e-06,
"loss": 0.9929,
"step": 1970
},
{
"epoch": 0.666554452112439,
"grad_norm": 4.09375,
"learning_rate": 4.21801639385857e-06,
"loss": 1.0086,
"step": 1980
},
{
"epoch": 0.6699208887392695,
"grad_norm": 3.640625,
"learning_rate": 4.207316108315597e-06,
"loss": 0.9904,
"step": 1990
},
{
"epoch": 0.6732873253661,
"grad_norm": 4.78125,
"learning_rate": 4.196556885224361e-06,
"loss": 0.9817,
"step": 2000
},
{
"epoch": 0.6766537619929305,
"grad_norm": 4.125,
"learning_rate": 4.185739095999524e-06,
"loss": 0.9523,
"step": 2010
},
{
"epoch": 0.6800201986197609,
"grad_norm": 4.71875,
"learning_rate": 4.17486311407748e-06,
"loss": 0.9561,
"step": 2020
},
{
"epoch": 0.6833866352465915,
"grad_norm": 4.53125,
"learning_rate": 4.163929314903472e-06,
"loss": 0.9719,
"step": 2030
},
{
"epoch": 0.6867530718734219,
"grad_norm": 4.375,
"learning_rate": 4.152938075918627e-06,
"loss": 0.9917,
"step": 2040
},
{
"epoch": 0.6901195085002525,
"grad_norm": 3.703125,
"learning_rate": 4.14188977654693e-06,
"loss": 0.9765,
"step": 2050
},
{
"epoch": 0.693485945127083,
"grad_norm": 3.8125,
"learning_rate": 4.130784798182119e-06,
"loss": 0.9562,
"step": 2060
},
{
"epoch": 0.6968523817539135,
"grad_norm": 3.640625,
"learning_rate": 4.119623524174528e-06,
"loss": 0.976,
"step": 2070
},
{
"epoch": 0.700218818380744,
"grad_norm": 4.125,
"learning_rate": 4.108406339817849e-06,
"loss": 0.9649,
"step": 2080
},
{
"epoch": 0.7035852550075745,
"grad_norm": 4.21875,
"learning_rate": 4.09713363233583e-06,
"loss": 0.9865,
"step": 2090
},
{
"epoch": 0.706951691634405,
"grad_norm": 3.765625,
"learning_rate": 4.085805790868911e-06,
"loss": 0.9691,
"step": 2100
},
{
"epoch": 0.7103181282612355,
"grad_norm": 4.28125,
"learning_rate": 4.074423206460789e-06,
"loss": 0.9687,
"step": 2110
},
{
"epoch": 0.713684564888066,
"grad_norm": 4.1875,
"learning_rate": 4.062986272044921e-06,
"loss": 0.9435,
"step": 2120
},
{
"epoch": 0.7170510015148965,
"grad_norm": 4.15625,
"learning_rate": 4.051495382430953e-06,
"loss": 0.9752,
"step": 2130
},
{
"epoch": 0.720417438141727,
"grad_norm": 3.953125,
"learning_rate": 4.039950934291103e-06,
"loss": 0.9793,
"step": 2140
},
{
"epoch": 0.7237838747685574,
"grad_norm": 3.9375,
"learning_rate": 4.028353326146455e-06,
"loss": 0.9708,
"step": 2150
},
{
"epoch": 0.727150311395388,
"grad_norm": 3.515625,
"learning_rate": 4.0167029583532084e-06,
"loss": 0.9598,
"step": 2160
},
{
"epoch": 0.7305167480222184,
"grad_norm": 3.921875,
"learning_rate": 4.005000233088859e-06,
"loss": 0.9772,
"step": 2170
},
{
"epoch": 0.733883184649049,
"grad_norm": 3.921875,
"learning_rate": 3.99324555433831e-06,
"loss": 0.9775,
"step": 2180
},
{
"epoch": 0.7372496212758795,
"grad_norm": 3.875,
"learning_rate": 3.981439327879932e-06,
"loss": 0.9757,
"step": 2190
},
{
"epoch": 0.74061605790271,
"grad_norm": 3.734375,
"learning_rate": 3.9695819612715505e-06,
"loss": 0.9623,
"step": 2200
},
{
"epoch": 0.7439824945295405,
"grad_norm": 4.8125,
"learning_rate": 3.95767386383638e-06,
"loss": 0.9504,
"step": 2210
},
{
"epoch": 0.747348931156371,
"grad_norm": 3.96875,
"learning_rate": 3.9457154466488904e-06,
"loss": 0.9476,
"step": 2220
},
{
"epoch": 0.7507153677832015,
"grad_norm": 4.84375,
"learning_rate": 3.933707122520622e-06,
"loss": 0.9698,
"step": 2230
},
{
"epoch": 0.754081804410032,
"grad_norm": 4.5625,
"learning_rate": 3.921649305985931e-06,
"loss": 0.9757,
"step": 2240
},
{
"epoch": 0.7574482410368625,
"grad_norm": 3.6875,
"learning_rate": 3.909542413287679e-06,
"loss": 0.9837,
"step": 2250
},
{
"epoch": 0.7608146776636929,
"grad_norm": 4.09375,
"learning_rate": 3.897386862362865e-06,
"loss": 0.9797,
"step": 2260
},
{
"epoch": 0.7641811142905235,
"grad_norm": 4.4375,
"learning_rate": 3.885183072828203e-06,
"loss": 0.9609,
"step": 2270
},
{
"epoch": 0.7675475509173539,
"grad_norm": 3.8125,
"learning_rate": 3.872931465965625e-06,
"loss": 0.9863,
"step": 2280
},
{
"epoch": 0.7709139875441845,
"grad_norm": 4.59375,
"learning_rate": 3.860632464707751e-06,
"loss": 0.9637,
"step": 2290
},
{
"epoch": 0.774280424171015,
"grad_norm": 5.0625,
"learning_rate": 3.848286493623278e-06,
"loss": 0.9839,
"step": 2300
},
{
"epoch": 0.7776468607978455,
"grad_norm": 3.953125,
"learning_rate": 3.835893978902334e-06,
"loss": 0.9834,
"step": 2310
},
{
"epoch": 0.781013297424676,
"grad_norm": 5.84375,
"learning_rate": 3.8234553483417584e-06,
"loss": 0.9808,
"step": 2320
},
{
"epoch": 0.7843797340515065,
"grad_norm": 3.671875,
"learning_rate": 3.8109710313303345e-06,
"loss": 0.9727,
"step": 2330
},
{
"epoch": 0.787746170678337,
"grad_norm": 4.25,
"learning_rate": 3.79844145883397e-06,
"loss": 0.9788,
"step": 2340
},
{
"epoch": 0.7911126073051675,
"grad_norm": 3.953125,
"learning_rate": 3.7858670633808198e-06,
"loss": 0.9741,
"step": 2350
},
{
"epoch": 0.794479043931998,
"grad_norm": 4.59375,
"learning_rate": 3.7732482790463497e-06,
"loss": 0.9853,
"step": 2360
},
{
"epoch": 0.7978454805588285,
"grad_norm": 4.4375,
"learning_rate": 3.7605855414383607e-06,
"loss": 0.9777,
"step": 2370
},
{
"epoch": 0.801211917185659,
"grad_norm": 3.875,
"learning_rate": 3.7478792876819412e-06,
"loss": 0.9789,
"step": 2380
},
{
"epoch": 0.8045783538124894,
"grad_norm": 4.1875,
"learning_rate": 3.7351299564043865e-06,
"loss": 0.9638,
"step": 2390
},
{
"epoch": 0.80794479043932,
"grad_norm": 4.9375,
"learning_rate": 3.7223379877200493e-06,
"loss": 1.0005,
"step": 2400
},
{
"epoch": 0.8113112270661504,
"grad_norm": 3.84375,
"learning_rate": 3.7095038232151546e-06,
"loss": 0.9783,
"step": 2410
},
{
"epoch": 0.814677663692981,
"grad_norm": 4.0,
"learning_rate": 3.696627905932549e-06,
"loss": 0.9929,
"step": 2420
},
{
"epoch": 0.8180441003198115,
"grad_norm": 4.28125,
"learning_rate": 3.6837106803564105e-06,
"loss": 0.9667,
"step": 2430
},
{
"epoch": 0.821410536946642,
"grad_norm": 3.90625,
"learning_rate": 3.6707525923969027e-06,
"loss": 0.9667,
"step": 2440
},
{
"epoch": 0.8247769735734725,
"grad_norm": 4.1875,
"learning_rate": 3.6577540893747826e-06,
"loss": 0.9589,
"step": 2450
},
{
"epoch": 0.828143410200303,
"grad_norm": 4.65625,
"learning_rate": 3.6447156200059584e-06,
"loss": 0.9728,
"step": 2460
},
{
"epoch": 0.8315098468271335,
"grad_norm": 3.859375,
"learning_rate": 3.6316376343860027e-06,
"loss": 0.9746,
"step": 2470
},
{
"epoch": 0.834876283453964,
"grad_norm": 4.09375,
"learning_rate": 3.6185205839746084e-06,
"loss": 0.9659,
"step": 2480
},
{
"epoch": 0.8382427200807945,
"grad_norm": 4.0,
"learning_rate": 3.6053649215800096e-06,
"loss": 0.9521,
"step": 2490
},
{
"epoch": 0.8416091567076249,
"grad_norm": 3.828125,
"learning_rate": 3.592171101343348e-06,
"loss": 0.9676,
"step": 2500
},
{
"epoch": 0.8449755933344555,
"grad_norm": 4.15625,
"learning_rate": 3.5789395787229968e-06,
"loss": 0.9616,
"step": 2510
},
{
"epoch": 0.8483420299612859,
"grad_norm": 4.9375,
"learning_rate": 3.5656708104788385e-06,
"loss": 0.9576,
"step": 2520
},
{
"epoch": 0.8517084665881165,
"grad_norm": 4.53125,
"learning_rate": 3.5523652546564933e-06,
"loss": 0.9675,
"step": 2530
},
{
"epoch": 0.855074903214947,
"grad_norm": 4.125,
"learning_rate": 3.5390233705715132e-06,
"loss": 0.9736,
"step": 2540
},
{
"epoch": 0.8584413398417775,
"grad_norm": 3.921875,
"learning_rate": 3.525645618793521e-06,
"loss": 0.9823,
"step": 2550
},
{
"epoch": 0.861807776468608,
"grad_norm": 4.0625,
"learning_rate": 3.512232461130314e-06,
"loss": 0.9906,
"step": 2560
},
{
"epoch": 0.8651742130954385,
"grad_norm": 4.3125,
"learning_rate": 3.498784360611921e-06,
"loss": 0.9635,
"step": 2570
},
{
"epoch": 0.868540649722269,
"grad_norm": 3.90625,
"learning_rate": 3.485301781474619e-06,
"loss": 0.9551,
"step": 2580
},
{
"epoch": 0.8719070863490995,
"grad_norm": 3.921875,
"learning_rate": 3.4717851891449073e-06,
"loss": 0.9744,
"step": 2590
},
{
"epoch": 0.87527352297593,
"grad_norm": 4.4375,
"learning_rate": 3.4582350502234385e-06,
"loss": 0.9581,
"step": 2600
},
{
"epoch": 0.8786399596027604,
"grad_norm": 4.375,
"learning_rate": 3.4446518324689153e-06,
"loss": 0.9651,
"step": 2610
},
{
"epoch": 0.882006396229591,
"grad_norm": 4.8125,
"learning_rate": 3.4310360047819406e-06,
"loss": 0.974,
"step": 2620
},
{
"epoch": 0.8853728328564214,
"grad_norm": 4.40625,
"learning_rate": 3.4173880371888305e-06,
"loss": 0.9568,
"step": 2630
},
{
"epoch": 0.888739269483252,
"grad_norm": 3.6875,
"learning_rate": 3.4037084008253897e-06,
"loss": 0.9665,
"step": 2640
},
{
"epoch": 0.8921057061100824,
"grad_norm": 4.125,
"learning_rate": 3.389997567920648e-06,
"loss": 0.9566,
"step": 2650
},
{
"epoch": 0.895472142736913,
"grad_norm": 4.90625,
"learning_rate": 3.3762560117805565e-06,
"loss": 0.9781,
"step": 2660
},
{
"epoch": 0.8988385793637435,
"grad_norm": 4.375,
"learning_rate": 3.3624842067716518e-06,
"loss": 0.9573,
"step": 2670
},
{
"epoch": 0.902205015990574,
"grad_norm": 4.0,
"learning_rate": 3.3486826283046785e-06,
"loss": 0.9496,
"step": 2680
},
{
"epoch": 0.9055714526174045,
"grad_norm": 3.859375,
"learning_rate": 3.334851752818178e-06,
"loss": 0.9796,
"step": 2690
},
{
"epoch": 0.908937889244235,
"grad_norm": 4.09375,
"learning_rate": 3.3209920577620424e-06,
"loss": 0.956,
"step": 2700
},
{
"epoch": 0.9123043258710655,
"grad_norm": 3.75,
"learning_rate": 3.3071040215810326e-06,
"loss": 0.9597,
"step": 2710
},
{
"epoch": 0.915670762497896,
"grad_norm": 4.59375,
"learning_rate": 3.29318812369826e-06,
"loss": 0.9659,
"step": 2720
},
{
"epoch": 0.9190371991247265,
"grad_norm": 3.453125,
"learning_rate": 3.27924484449864e-06,
"loss": 0.9571,
"step": 2730
},
{
"epoch": 0.9224036357515569,
"grad_norm": 4.375,
"learning_rate": 3.2652746653123068e-06,
"loss": 0.9696,
"step": 2740
},
{
"epoch": 0.9257700723783875,
"grad_norm": 4.75,
"learning_rate": 3.2512780683979972e-06,
"loss": 0.9868,
"step": 2750
},
{
"epoch": 0.9291365090052179,
"grad_norm": 5.125,
"learning_rate": 3.2372555369264024e-06,
"loss": 0.9727,
"step": 2760
},
{
"epoch": 0.9325029456320485,
"grad_norm": 4.0625,
"learning_rate": 3.2232075549634923e-06,
"loss": 0.9707,
"step": 2770
},
{
"epoch": 0.935869382258879,
"grad_norm": 3.828125,
"learning_rate": 3.2091346074537994e-06,
"loss": 0.9599,
"step": 2780
},
{
"epoch": 0.9392358188857095,
"grad_norm": 3.796875,
"learning_rate": 3.1950371802036844e-06,
"loss": 0.9805,
"step": 2790
},
{
"epoch": 0.94260225551254,
"grad_norm": 4.53125,
"learning_rate": 3.1809157598645623e-06,
"loss": 0.9875,
"step": 2800
},
{
"epoch": 0.9459686921393705,
"grad_norm": 4.1875,
"learning_rate": 3.166770833916102e-06,
"loss": 0.9808,
"step": 2810
},
{
"epoch": 0.949335128766201,
"grad_norm": 3.84375,
"learning_rate": 3.1526028906494005e-06,
"loss": 0.9712,
"step": 2820
},
{
"epoch": 0.9527015653930315,
"grad_norm": 3.78125,
"learning_rate": 3.1384124191501257e-06,
"loss": 0.9666,
"step": 2830
},
{
"epoch": 0.956068002019862,
"grad_norm": 3.78125,
"learning_rate": 3.1241999092816356e-06,
"loss": 0.9755,
"step": 2840
},
{
"epoch": 0.9594344386466924,
"grad_norm": 3.703125,
"learning_rate": 3.109965851668062e-06,
"loss": 0.9536,
"step": 2850
},
{
"epoch": 0.962800875273523,
"grad_norm": 4.5,
"learning_rate": 3.09571073767738e-06,
"loss": 0.9707,
"step": 2860
},
{
"epoch": 0.9661673119003534,
"grad_norm": 3.90625,
"learning_rate": 3.081435059404443e-06,
"loss": 0.983,
"step": 2870
},
{
"epoch": 0.969533748527184,
"grad_norm": 4.8125,
"learning_rate": 3.0671393096539925e-06,
"loss": 0.9676,
"step": 2880
},
{
"epoch": 0.9729001851540144,
"grad_norm": 4.65625,
"learning_rate": 3.052823981923653e-06,
"loss": 0.9628,
"step": 2890
},
{
"epoch": 0.976266621780845,
"grad_norm": 4.5,
"learning_rate": 3.0384895703868898e-06,
"loss": 0.978,
"step": 2900
},
{
"epoch": 0.9796330584076754,
"grad_norm": 3.65625,
"learning_rate": 3.024136569875954e-06,
"loss": 0.958,
"step": 2910
},
{
"epoch": 0.982999495034506,
"grad_norm": 3.96875,
"learning_rate": 3.0097654758647975e-06,
"loss": 0.9789,
"step": 2920
},
{
"epoch": 0.9863659316613365,
"grad_norm": 4.625,
"learning_rate": 2.995376784451971e-06,
"loss": 0.9757,
"step": 2930
},
{
"epoch": 0.989732368288167,
"grad_norm": 4.1875,
"learning_rate": 2.9809709923434974e-06,
"loss": 0.9962,
"step": 2940
},
{
"epoch": 0.9930988049149975,
"grad_norm": 3.8125,
"learning_rate": 2.9665485968357276e-06,
"loss": 0.9429,
"step": 2950
},
{
"epoch": 0.996465241541828,
"grad_norm": 4.3125,
"learning_rate": 2.9521100957981676e-06,
"loss": 0.9679,
"step": 2960
},
{
"epoch": 0.9998316781686585,
"grad_norm": 3.6875,
"learning_rate": 2.9376559876562993e-06,
"loss": 0.9672,
"step": 2970
},
{
"epoch": 1.0030297929641474,
"grad_norm": 3.765625,
"learning_rate": 2.9231867713743684e-06,
"loss": 0.9376,
"step": 2980
},
{
"epoch": 1.0063962295909779,
"grad_norm": 4.28125,
"learning_rate": 2.9087029464381643e-06,
"loss": 0.9602,
"step": 2990
},
{
"epoch": 1.0097626662178085,
"grad_norm": 3.890625,
"learning_rate": 2.8942050128377763e-06,
"loss": 0.9546,
"step": 3000
},
{
"epoch": 1.013129102844639,
"grad_norm": 3.78125,
"learning_rate": 2.8796934710503293e-06,
"loss": 0.9648,
"step": 3010
},
{
"epoch": 1.0164955394714694,
"grad_norm": 4.90625,
"learning_rate": 2.8651688220227148e-06,
"loss": 0.9514,
"step": 3020
},
{
"epoch": 1.0198619760982999,
"grad_norm": 3.6875,
"learning_rate": 2.850631567154292e-06,
"loss": 0.9942,
"step": 3030
},
{
"epoch": 1.0232284127251305,
"grad_norm": 3.90625,
"learning_rate": 2.8360822082795803e-06,
"loss": 0.9635,
"step": 3040
},
{
"epoch": 1.026594849351961,
"grad_norm": 5.4375,
"learning_rate": 2.8215212476509386e-06,
"loss": 0.9577,
"step": 3050
},
{
"epoch": 1.0299612859787914,
"grad_norm": 4.4375,
"learning_rate": 2.806949187921223e-06,
"loss": 0.9789,
"step": 3060
},
{
"epoch": 1.033327722605622,
"grad_norm": 4.09375,
"learning_rate": 2.7923665321264386e-06,
"loss": 0.971,
"step": 3070
},
{
"epoch": 1.0366941592324523,
"grad_norm": 3.671875,
"learning_rate": 2.777773783668373e-06,
"loss": 0.9493,
"step": 3080
},
{
"epoch": 1.040060595859283,
"grad_norm": 4.1875,
"learning_rate": 2.7631714462972163e-06,
"loss": 0.9706,
"step": 3090
},
{
"epoch": 1.0434270324861135,
"grad_norm": 3.859375,
"learning_rate": 2.748560024094176e-06,
"loss": 0.9724,
"step": 3100
},
{
"epoch": 1.046793469112944,
"grad_norm": 4.1875,
"learning_rate": 2.733940021454071e-06,
"loss": 0.9782,
"step": 3110
},
{
"epoch": 1.0501599057397744,
"grad_norm": 4.0,
"learning_rate": 2.7193119430679238e-06,
"loss": 0.9649,
"step": 3120
},
{
"epoch": 1.053526342366605,
"grad_norm": 3.421875,
"learning_rate": 2.7046762939055367e-06,
"loss": 0.9688,
"step": 3130
},
{
"epoch": 1.0568927789934355,
"grad_norm": 3.875,
"learning_rate": 2.690033579198056e-06,
"loss": 0.9845,
"step": 3140
},
{
"epoch": 1.060259215620266,
"grad_norm": 3.515625,
"learning_rate": 2.6753843044205392e-06,
"loss": 0.9565,
"step": 3150
},
{
"epoch": 1.0636256522470964,
"grad_norm": 4.625,
"learning_rate": 2.660728975274498e-06,
"loss": 0.9666,
"step": 3160
},
{
"epoch": 1.066992088873927,
"grad_norm": 3.515625,
"learning_rate": 2.646068097670448e-06,
"loss": 0.9667,
"step": 3170
},
{
"epoch": 1.0703585255007575,
"grad_norm": 3.875,
"learning_rate": 2.631402177710436e-06,
"loss": 0.9433,
"step": 3180
},
{
"epoch": 1.073724962127588,
"grad_norm": 3.609375,
"learning_rate": 2.616731721670578e-06,
"loss": 0.9685,
"step": 3190
},
{
"epoch": 1.0770913987544184,
"grad_norm": 3.59375,
"learning_rate": 2.602057235983577e-06,
"loss": 0.9702,
"step": 3200
},
{
"epoch": 1.080457835381249,
"grad_norm": 4.34375,
"learning_rate": 2.5873792272212388e-06,
"loss": 0.985,
"step": 3210
},
{
"epoch": 1.0838242720080795,
"grad_norm": 3.90625,
"learning_rate": 2.5726982020769923e-06,
"loss": 0.9588,
"step": 3220
},
{
"epoch": 1.08719070863491,
"grad_norm": 3.8125,
"learning_rate": 2.558014667348393e-06,
"loss": 0.9583,
"step": 3230
},
{
"epoch": 1.0905571452617404,
"grad_norm": 4.1875,
"learning_rate": 2.5433291299196245e-06,
"loss": 0.9537,
"step": 3240
},
{
"epoch": 1.0939235818885709,
"grad_norm": 4.09375,
"learning_rate": 2.5286420967440096e-06,
"loss": 0.983,
"step": 3250
},
{
"epoch": 1.0972900185154015,
"grad_norm": 3.96875,
"learning_rate": 2.513954074826502e-06,
"loss": 0.9666,
"step": 3260
},
{
"epoch": 1.100656455142232,
"grad_norm": 5.25,
"learning_rate": 2.4992655712061904e-06,
"loss": 0.9484,
"step": 3270
},
{
"epoch": 1.1040228917690624,
"grad_norm": 4.4375,
"learning_rate": 2.484577092938788e-06,
"loss": 0.9369,
"step": 3280
},
{
"epoch": 1.1073893283958929,
"grad_norm": 4.3125,
"learning_rate": 2.4698891470791374e-06,
"loss": 0.9804,
"step": 3290
},
{
"epoch": 1.1107557650227236,
"grad_norm": 4.125,
"learning_rate": 2.455202240663699e-06,
"loss": 0.957,
"step": 3300
},
{
"epoch": 1.114122201649554,
"grad_norm": 3.78125,
"learning_rate": 2.4405168806930513e-06,
"loss": 0.9653,
"step": 3310
},
{
"epoch": 1.1174886382763844,
"grad_norm": 3.671875,
"learning_rate": 2.4258335741143903e-06,
"loss": 0.9987,
"step": 3320
},
{
"epoch": 1.120855074903215,
"grad_norm": 3.4375,
"learning_rate": 2.411152827804027e-06,
"loss": 0.9774,
"step": 3330
},
{
"epoch": 1.1242215115300453,
"grad_norm": 5.34375,
"learning_rate": 2.396475148549889e-06,
"loss": 0.9769,
"step": 3340
},
{
"epoch": 1.127587948156876,
"grad_norm": 3.671875,
"learning_rate": 2.3818010430340273e-06,
"loss": 0.9723,
"step": 3350
},
{
"epoch": 1.1309543847837065,
"grad_norm": 4.125,
"learning_rate": 2.3671310178151294e-06,
"loss": 0.9552,
"step": 3360
},
{
"epoch": 1.134320821410537,
"grad_norm": 4.15625,
"learning_rate": 2.3524655793110242e-06,
"loss": 0.9511,
"step": 3370
},
{
"epoch": 1.1376872580373674,
"grad_norm": 3.84375,
"learning_rate": 2.337805233781205e-06,
"loss": 0.967,
"step": 3380
},
{
"epoch": 1.141053694664198,
"grad_norm": 4.5,
"learning_rate": 2.323150487309356e-06,
"loss": 0.9644,
"step": 3390
},
{
"epoch": 1.1444201312910285,
"grad_norm": 4.1875,
"learning_rate": 2.308501845785875e-06,
"loss": 0.9708,
"step": 3400
},
{
"epoch": 1.147786567917859,
"grad_norm": 4.5,
"learning_rate": 2.2938598148904136e-06,
"loss": 0.9779,
"step": 3410
},
{
"epoch": 1.1511530045446894,
"grad_norm": 3.859375,
"learning_rate": 2.279224900074424e-06,
"loss": 0.9689,
"step": 3420
},
{
"epoch": 1.1545194411715198,
"grad_norm": 4.125,
"learning_rate": 2.2645976065437036e-06,
"loss": 0.9836,
"step": 3430
},
{
"epoch": 1.1578858777983505,
"grad_norm": 4.0,
"learning_rate": 2.2499784392409604e-06,
"loss": 0.945,
"step": 3440
},
{
"epoch": 1.161252314425181,
"grad_norm": 4.03125,
"learning_rate": 2.235367902828379e-06,
"loss": 0.9917,
"step": 3450
},
{
"epoch": 1.1646187510520114,
"grad_norm": 4.09375,
"learning_rate": 2.2207665016702024e-06,
"loss": 0.9361,
"step": 3460
},
{
"epoch": 1.1679851876788419,
"grad_norm": 4.53125,
"learning_rate": 2.206174739815319e-06,
"loss": 0.9677,
"step": 3470
},
{
"epoch": 1.1713516243056725,
"grad_norm": 3.75,
"learning_rate": 2.1915931209798595e-06,
"loss": 0.9439,
"step": 3480
},
{
"epoch": 1.174718060932503,
"grad_norm": 3.515625,
"learning_rate": 2.177022148529818e-06,
"loss": 0.9628,
"step": 3490
},
{
"epoch": 1.1780844975593334,
"grad_norm": 3.890625,
"learning_rate": 2.162462325463663e-06,
"loss": 0.9512,
"step": 3500
},
{
"epoch": 1.1814509341861639,
"grad_norm": 3.890625,
"learning_rate": 2.147914154394982e-06,
"loss": 0.9449,
"step": 3510
},
{
"epoch": 1.1848173708129945,
"grad_norm": 4.0625,
"learning_rate": 2.1333781375351313e-06,
"loss": 0.97,
"step": 3520
},
{
"epoch": 1.188183807439825,
"grad_norm": 3.984375,
"learning_rate": 2.1188547766758925e-06,
"loss": 0.9236,
"step": 3530
},
{
"epoch": 1.1915502440666554,
"grad_norm": 3.546875,
"learning_rate": 2.104344573172157e-06,
"loss": 0.9869,
"step": 3540
},
{
"epoch": 1.1949166806934859,
"grad_norm": 4.1875,
"learning_rate": 2.089848027924617e-06,
"loss": 0.9794,
"step": 3550
},
{
"epoch": 1.1982831173203166,
"grad_norm": 3.6875,
"learning_rate": 2.0753656413624728e-06,
"loss": 0.9689,
"step": 3560
},
{
"epoch": 1.201649553947147,
"grad_norm": 3.71875,
"learning_rate": 2.060897913426158e-06,
"loss": 0.9586,
"step": 3570
},
{
"epoch": 1.2050159905739775,
"grad_norm": 3.71875,
"learning_rate": 2.046445343550084e-06,
"loss": 0.9813,
"step": 3580
},
{
"epoch": 1.208382427200808,
"grad_norm": 4.1875,
"learning_rate": 2.032008430645395e-06,
"loss": 0.9952,
"step": 3590
},
{
"epoch": 1.2117488638276384,
"grad_norm": 4.53125,
"learning_rate": 2.017587673082749e-06,
"loss": 0.9674,
"step": 3600
},
{
"epoch": 1.215115300454469,
"grad_norm": 4.40625,
"learning_rate": 2.00318356867511e-06,
"loss": 0.9539,
"step": 3610
},
{
"epoch": 1.2184817370812995,
"grad_norm": 4.3125,
"learning_rate": 1.988796614660569e-06,
"loss": 0.9704,
"step": 3620
},
{
"epoch": 1.22184817370813,
"grad_norm": 3.890625,
"learning_rate": 1.974427307685171e-06,
"loss": 0.9761,
"step": 3630
},
{
"epoch": 1.2252146103349604,
"grad_norm": 4.625,
"learning_rate": 1.960076143785779e-06,
"loss": 0.9897,
"step": 3640
},
{
"epoch": 1.228581046961791,
"grad_norm": 3.890625,
"learning_rate": 1.9457436183729434e-06,
"loss": 0.9734,
"step": 3650
},
{
"epoch": 1.2319474835886215,
"grad_norm": 3.734375,
"learning_rate": 1.9314302262138047e-06,
"loss": 0.963,
"step": 3660
},
{
"epoch": 1.235313920215452,
"grad_norm": 3.78125,
"learning_rate": 1.917136461415012e-06,
"loss": 0.9531,
"step": 3670
},
{
"epoch": 1.2386803568422824,
"grad_norm": 3.921875,
"learning_rate": 1.9028628174056654e-06,
"loss": 0.9729,
"step": 3680
},
{
"epoch": 1.2420467934691128,
"grad_norm": 5.03125,
"learning_rate": 1.8886097869202868e-06,
"loss": 0.9499,
"step": 3690
},
{
"epoch": 1.2454132300959435,
"grad_norm": 4.1875,
"learning_rate": 1.8743778619818038e-06,
"loss": 0.9714,
"step": 3700
},
{
"epoch": 1.248779666722774,
"grad_norm": 3.8125,
"learning_rate": 1.8601675338845682e-06,
"loss": 0.9851,
"step": 3710
},
{
"epoch": 1.2521461033496044,
"grad_norm": 4.4375,
"learning_rate": 1.8459792931774002e-06,
"loss": 0.9557,
"step": 3720
},
{
"epoch": 1.2555125399764349,
"grad_norm": 4.40625,
"learning_rate": 1.831813629646646e-06,
"loss": 0.9471,
"step": 3730
},
{
"epoch": 1.2588789766032655,
"grad_norm": 4.0625,
"learning_rate": 1.8176710322992757e-06,
"loss": 0.9585,
"step": 3740
},
{
"epoch": 1.262245413230096,
"grad_norm": 3.875,
"learning_rate": 1.8035519893460034e-06,
"loss": 0.9732,
"step": 3750
},
{
"epoch": 1.2656118498569264,
"grad_norm": 4.09375,
"learning_rate": 1.7894569881844296e-06,
"loss": 0.9821,
"step": 3760
},
{
"epoch": 1.2689782864837569,
"grad_norm": 4.3125,
"learning_rate": 1.7753865153822184e-06,
"loss": 0.9436,
"step": 3770
},
{
"epoch": 1.2723447231105873,
"grad_norm": 3.65625,
"learning_rate": 1.7613410566603001e-06,
"loss": 0.955,
"step": 3780
},
{
"epoch": 1.275711159737418,
"grad_norm": 4.09375,
"learning_rate": 1.7473210968761068e-06,
"loss": 0.9546,
"step": 3790
},
{
"epoch": 1.2790775963642484,
"grad_norm": 3.875,
"learning_rate": 1.7333271200068302e-06,
"loss": 0.9614,
"step": 3800
},
{
"epoch": 1.282444032991079,
"grad_norm": 4.625,
"learning_rate": 1.7193596091327158e-06,
"loss": 0.9765,
"step": 3810
},
{
"epoch": 1.2858104696179096,
"grad_norm": 3.859375,
"learning_rate": 1.7054190464203912e-06,
"loss": 0.9712,
"step": 3820
},
{
"epoch": 1.28917690624474,
"grad_norm": 3.796875,
"learning_rate": 1.6915059131062139e-06,
"loss": 0.9542,
"step": 3830
},
{
"epoch": 1.2925433428715705,
"grad_norm": 4.1875,
"learning_rate": 1.677620689479664e-06,
"loss": 0.9569,
"step": 3840
},
{
"epoch": 1.295909779498401,
"grad_norm": 4.53125,
"learning_rate": 1.6637638548667642e-06,
"loss": 0.9832,
"step": 3850
},
{
"epoch": 1.2992762161252314,
"grad_norm": 4.625,
"learning_rate": 1.6499358876135298e-06,
"loss": 0.9692,
"step": 3860
},
{
"epoch": 1.3026426527520618,
"grad_norm": 3.78125,
"learning_rate": 1.636137265069459e-06,
"loss": 0.9697,
"step": 3870
},
{
"epoch": 1.3060090893788925,
"grad_norm": 4.09375,
"learning_rate": 1.6223684635710524e-06,
"loss": 0.9637,
"step": 3880
},
{
"epoch": 1.309375526005723,
"grad_norm": 3.859375,
"learning_rate": 1.6086299584253736e-06,
"loss": 0.9681,
"step": 3890
},
{
"epoch": 1.3127419626325534,
"grad_norm": 5.3125,
"learning_rate": 1.5949222238936356e-06,
"loss": 0.9778,
"step": 3900
},
{
"epoch": 1.316108399259384,
"grad_norm": 3.953125,
"learning_rate": 1.5812457331748316e-06,
"loss": 0.9925,
"step": 3910
},
{
"epoch": 1.3194748358862145,
"grad_norm": 3.875,
"learning_rate": 1.5676009583894034e-06,
"loss": 0.9804,
"step": 3920
},
{
"epoch": 1.322841272513045,
"grad_norm": 3.65625,
"learning_rate": 1.553988370562937e-06,
"loss": 0.9427,
"step": 3930
},
{
"epoch": 1.3262077091398754,
"grad_norm": 4.34375,
"learning_rate": 1.5404084396099064e-06,
"loss": 0.9494,
"step": 3940
},
{
"epoch": 1.3295741457667059,
"grad_norm": 4.28125,
"learning_rate": 1.5268616343174537e-06,
"loss": 0.9535,
"step": 3950
},
{
"epoch": 1.3329405823935365,
"grad_norm": 3.75,
"learning_rate": 1.513348422329201e-06,
"loss": 0.9739,
"step": 3960
},
{
"epoch": 1.336307019020367,
"grad_norm": 3.890625,
"learning_rate": 1.4998692701291106e-06,
"loss": 0.9661,
"step": 3970
},
{
"epoch": 1.3396734556471974,
"grad_norm": 3.765625,
"learning_rate": 1.4864246430253807e-06,
"loss": 0.9583,
"step": 3980
},
{
"epoch": 1.3430398922740279,
"grad_norm": 3.640625,
"learning_rate": 1.4730150051343856e-06,
"loss": 0.9493,
"step": 3990
},
{
"epoch": 1.3464063289008585,
"grad_norm": 3.546875,
"learning_rate": 1.459640819364649e-06,
"loss": 0.9825,
"step": 4000
},
{
"epoch": 1.349772765527689,
"grad_norm": 3.828125,
"learning_rate": 1.4463025474008662e-06,
"loss": 0.9634,
"step": 4010
},
{
"epoch": 1.3531392021545194,
"grad_norm": 3.90625,
"learning_rate": 1.4330006496879701e-06,
"loss": 0.9483,
"step": 4020
},
{
"epoch": 1.3565056387813499,
"grad_norm": 4.25,
"learning_rate": 1.4197355854152305e-06,
"loss": 0.968,
"step": 4030
},
{
"epoch": 1.3598720754081803,
"grad_norm": 4.09375,
"learning_rate": 1.4065078125004044e-06,
"loss": 0.9814,
"step": 4040
},
{
"epoch": 1.363238512035011,
"grad_norm": 3.578125,
"learning_rate": 1.3933177875739346e-06,
"loss": 0.9836,
"step": 4050
},
{
"epoch": 1.3666049486618415,
"grad_norm": 4.125,
"learning_rate": 1.3801659659631743e-06,
"loss": 0.9598,
"step": 4060
},
{
"epoch": 1.369971385288672,
"grad_norm": 3.734375,
"learning_rate": 1.3670528016766815e-06,
"loss": 0.966,
"step": 4070
},
{
"epoch": 1.3733378219155026,
"grad_norm": 3.984375,
"learning_rate": 1.3539787473885384e-06,
"loss": 1.0049,
"step": 4080
},
{
"epoch": 1.376704258542333,
"grad_norm": 3.78125,
"learning_rate": 1.3409442544227283e-06,
"loss": 0.9765,
"step": 4090
},
{
"epoch": 1.3800706951691635,
"grad_norm": 4.25,
"learning_rate": 1.3279497727375518e-06,
"loss": 0.9703,
"step": 4100
},
{
"epoch": 1.383437131795994,
"grad_norm": 3.9375,
"learning_rate": 1.3149957509101002e-06,
"loss": 0.9644,
"step": 4110
},
{
"epoch": 1.3868035684228244,
"grad_norm": 3.90625,
"learning_rate": 1.3020826361207656e-06,
"loss": 0.9869,
"step": 4120
},
{
"epoch": 1.3901700050496548,
"grad_norm": 3.71875,
"learning_rate": 1.2892108741378058e-06,
"loss": 0.9627,
"step": 4130
},
{
"epoch": 1.3935364416764855,
"grad_norm": 4.03125,
"learning_rate": 1.2763809093019525e-06,
"loss": 0.9463,
"step": 4140
},
{
"epoch": 1.396902878303316,
"grad_norm": 4.0625,
"learning_rate": 1.2635931845110816e-06,
"loss": 0.9786,
"step": 4150
},
{
"epoch": 1.4002693149301464,
"grad_norm": 3.921875,
"learning_rate": 1.2508481412049112e-06,
"loss": 0.9593,
"step": 4160
},
{
"epoch": 1.403635751556977,
"grad_norm": 3.5625,
"learning_rate": 1.2381462193497743e-06,
"loss": 0.9663,
"step": 4170
},
{
"epoch": 1.4070021881838075,
"grad_norm": 4.71875,
"learning_rate": 1.2254878574234244e-06,
"loss": 0.9698,
"step": 4180
},
{
"epoch": 1.410368624810638,
"grad_norm": 3.734375,
"learning_rate": 1.2128734923999027e-06,
"loss": 0.9545,
"step": 4190
},
{
"epoch": 1.4137350614374684,
"grad_norm": 3.90625,
"learning_rate": 1.200303559734448e-06,
"loss": 0.9726,
"step": 4200
},
{
"epoch": 1.4171014980642989,
"grad_norm": 3.703125,
"learning_rate": 1.187778493348471e-06,
"loss": 0.9524,
"step": 4210
},
{
"epoch": 1.4204679346911293,
"grad_norm": 3.890625,
"learning_rate": 1.1752987256145717e-06,
"loss": 0.9618,
"step": 4220
},
{
"epoch": 1.42383437131796,
"grad_norm": 4.1875,
"learning_rate": 1.1628646873416152e-06,
"loss": 0.9589,
"step": 4230
},
{
"epoch": 1.4272008079447904,
"grad_norm": 3.515625,
"learning_rate": 1.1504768077598551e-06,
"loss": 0.9604,
"step": 4240
},
{
"epoch": 1.4305672445716209,
"grad_norm": 3.859375,
"learning_rate": 1.1381355145061266e-06,
"loss": 0.9677,
"step": 4250
},
{
"epoch": 1.4339336811984515,
"grad_norm": 4.46875,
"learning_rate": 1.1258412336090707e-06,
"loss": 0.9535,
"step": 4260
},
{
"epoch": 1.437300117825282,
"grad_norm": 4.03125,
"learning_rate": 1.1135943894744383e-06,
"loss": 0.9667,
"step": 4270
},
{
"epoch": 1.4406665544521124,
"grad_norm": 3.625,
"learning_rate": 1.1013954048704356e-06,
"loss": 0.9569,
"step": 4280
},
{
"epoch": 1.444032991078943,
"grad_norm": 4.0625,
"learning_rate": 1.0892447009131284e-06,
"loss": 0.9625,
"step": 4290
},
{
"epoch": 1.4473994277057733,
"grad_norm": 4.40625,
"learning_rate": 1.0771426970519083e-06,
"loss": 0.9634,
"step": 4300
},
{
"epoch": 1.450765864332604,
"grad_norm": 4.5,
"learning_rate": 1.0650898110550084e-06,
"loss": 0.9689,
"step": 4310
},
{
"epoch": 1.4541323009594345,
"grad_norm": 4.0625,
"learning_rate": 1.0530864589950864e-06,
"loss": 0.928,
"step": 4320
},
{
"epoch": 1.457498737586265,
"grad_norm": 3.859375,
"learning_rate": 1.0411330552348598e-06,
"loss": 0.9496,
"step": 4330
},
{
"epoch": 1.4608651742130954,
"grad_norm": 4.0625,
"learning_rate": 1.0292300124128007e-06,
"loss": 0.9851,
"step": 4340
},
{
"epoch": 1.464231610839926,
"grad_norm": 3.984375,
"learning_rate": 1.0173777414288933e-06,
"loss": 0.9642,
"step": 4350
},
{
"epoch": 1.4675980474667565,
"grad_norm": 3.75,
"learning_rate": 1.0055766514304474e-06,
"loss": 0.9733,
"step": 4360
},
{
"epoch": 1.470964484093587,
"grad_norm": 4.28125,
"learning_rate": 9.938271497979747e-07,
"loss": 0.9632,
"step": 4370
},
{
"epoch": 1.4743309207204174,
"grad_norm": 3.8125,
"learning_rate": 9.82129642131128e-07,
"loss": 0.9554,
"step": 4380
},
{
"epoch": 1.4776973573472478,
"grad_norm": 4.03125,
"learning_rate": 9.70484532234697e-07,
"loss": 0.9524,
"step": 4390
},
{
"epoch": 1.4810637939740785,
"grad_norm": 4.4375,
"learning_rate": 9.588922221046704e-07,
"loss": 0.9738,
"step": 4400
},
{
"epoch": 1.484430230600909,
"grad_norm": 3.859375,
"learning_rate": 9.473531119143578e-07,
"loss": 0.936,
"step": 4410
},
{
"epoch": 1.4877966672277394,
"grad_norm": 4.3125,
"learning_rate": 9.35867600000577e-07,
"loss": 0.9358,
"step": 4420
},
{
"epoch": 1.49116310385457,
"grad_norm": 4.21875,
"learning_rate": 9.244360828498988e-07,
"loss": 0.9789,
"step": 4430
},
{
"epoch": 1.4945295404814005,
"grad_norm": 4.65625,
"learning_rate": 9.130589550849669e-07,
"loss": 0.9841,
"step": 4440
},
{
"epoch": 1.497895977108231,
"grad_norm": 4.21875,
"learning_rate": 9.017366094508695e-07,
"loss": 0.9653,
"step": 4450
},
{
"epoch": 1.5012624137350614,
"grad_norm": 3.71875,
"learning_rate": 8.904694368015857e-07,
"loss": 0.9615,
"step": 4460
},
{
"epoch": 1.5046288503618919,
"grad_norm": 3.703125,
"learning_rate": 8.792578260864867e-07,
"loss": 0.9541,
"step": 4470
},
{
"epoch": 1.5079952869887223,
"grad_norm": 4.34375,
"learning_rate": 8.6810216433692e-07,
"loss": 0.9744,
"step": 4480
},
{
"epoch": 1.511361723615553,
"grad_norm": 3.734375,
"learning_rate": 8.570028366528357e-07,
"loss": 0.9637,
"step": 4490
},
{
"epoch": 1.5147281602423834,
"grad_norm": 3.953125,
"learning_rate": 8.459602261895028e-07,
"loss": 0.9424,
"step": 4500
},
{
"epoch": 1.5180945968692139,
"grad_norm": 4.375,
"learning_rate": 8.349747141442777e-07,
"loss": 0.9672,
"step": 4510
},
{
"epoch": 1.5214610334960446,
"grad_norm": 4.125,
"learning_rate": 8.240466797434465e-07,
"loss": 0.9658,
"step": 4520
},
{
"epoch": 1.524827470122875,
"grad_norm": 3.5625,
"learning_rate": 8.131765002291322e-07,
"loss": 0.9656,
"step": 4530
},
{
"epoch": 1.5281939067497055,
"grad_norm": 3.65625,
"learning_rate": 8.023645508462746e-07,
"loss": 0.9684,
"step": 4540
},
{
"epoch": 1.531560343376536,
"grad_norm": 3.625,
"learning_rate": 7.916112048296754e-07,
"loss": 0.9609,
"step": 4550
},
{
"epoch": 1.5349267800033664,
"grad_norm": 3.765625,
"learning_rate": 7.809168333911141e-07,
"loss": 0.9497,
"step": 4560
},
{
"epoch": 1.5382932166301968,
"grad_norm": 3.84375,
"learning_rate": 7.702818057065303e-07,
"loss": 0.9978,
"step": 4570
},
{
"epoch": 1.5416596532570275,
"grad_norm": 4.09375,
"learning_rate": 7.597064889032888e-07,
"loss": 0.9627,
"step": 4580
},
{
"epoch": 1.545026089883858,
"grad_norm": 4.375,
"learning_rate": 7.491912480474933e-07,
"loss": 0.9539,
"step": 4590
},
{
"epoch": 1.5483925265106886,
"grad_norm": 3.734375,
"learning_rate": 7.387364461313948e-07,
"loss": 0.9645,
"step": 4600
},
{
"epoch": 1.551758963137519,
"grad_norm": 3.6875,
"learning_rate": 7.283424440608558e-07,
"loss": 0.9661,
"step": 4610
},
{
"epoch": 1.5551253997643495,
"grad_norm": 5.03125,
"learning_rate": 7.18009600642893e-07,
"loss": 0.9698,
"step": 4620
},
{
"epoch": 1.55849183639118,
"grad_norm": 4.9375,
"learning_rate": 7.07738272573289e-07,
"loss": 0.9728,
"step": 4630
},
{
"epoch": 1.5618582730180104,
"grad_norm": 3.9375,
"learning_rate": 6.975288144242828e-07,
"loss": 0.9467,
"step": 4640
},
{
"epoch": 1.5652247096448408,
"grad_norm": 4.28125,
"learning_rate": 6.873815786323266e-07,
"loss": 0.9547,
"step": 4650
},
{
"epoch": 1.5685911462716713,
"grad_norm": 3.703125,
"learning_rate": 6.772969154859216e-07,
"loss": 0.9502,
"step": 4660
},
{
"epoch": 1.571957582898502,
"grad_norm": 4.53125,
"learning_rate": 6.672751731135219e-07,
"loss": 0.9623,
"step": 4670
},
{
"epoch": 1.5753240195253324,
"grad_norm": 4.375,
"learning_rate": 6.573166974715242e-07,
"loss": 0.9659,
"step": 4680
},
{
"epoch": 1.578690456152163,
"grad_norm": 4.0625,
"learning_rate": 6.474218323323167e-07,
"loss": 0.9875,
"step": 4690
},
{
"epoch": 1.5820568927789935,
"grad_norm": 3.53125,
"learning_rate": 6.375909192724181e-07,
"loss": 0.9664,
"step": 4700
},
{
"epoch": 1.585423329405824,
"grad_norm": 4.125,
"learning_rate": 6.278242976606833e-07,
"loss": 0.9534,
"step": 4710
},
{
"epoch": 1.5887897660326544,
"grad_norm": 4.03125,
"learning_rate": 6.181223046465901e-07,
"loss": 0.947,
"step": 4720
},
{
"epoch": 1.5921562026594849,
"grad_norm": 3.890625,
"learning_rate": 6.084852751485965e-07,
"loss": 0.9757,
"step": 4730
},
{
"epoch": 1.5955226392863153,
"grad_norm": 4.03125,
"learning_rate": 5.989135418425848e-07,
"loss": 0.9789,
"step": 4740
},
{
"epoch": 1.5988890759131458,
"grad_norm": 4.25,
"learning_rate": 5.89407435150374e-07,
"loss": 0.9728,
"step": 4750
},
{
"epoch": 1.6022555125399764,
"grad_norm": 4.28125,
"learning_rate": 5.799672832283135e-07,
"loss": 0.9658,
"step": 4760
},
{
"epoch": 1.605621949166807,
"grad_norm": 3.859375,
"learning_rate": 5.705934119559559e-07,
"loss": 0.9734,
"step": 4770
},
{
"epoch": 1.6089883857936376,
"grad_norm": 4.15625,
"learning_rate": 5.612861449248073e-07,
"loss": 0.968,
"step": 4780
},
{
"epoch": 1.612354822420468,
"grad_norm": 4.4375,
"learning_rate": 5.52045803427157e-07,
"loss": 0.9465,
"step": 4790
},
{
"epoch": 1.6157212590472985,
"grad_norm": 4.15625,
"learning_rate": 5.428727064449835e-07,
"loss": 1.0007,
"step": 4800
},
{
"epoch": 1.619087695674129,
"grad_norm": 3.609375,
"learning_rate": 5.337671706389478e-07,
"loss": 0.9739,
"step": 4810
},
{
"epoch": 1.6224541323009594,
"grad_norm": 3.640625,
"learning_rate": 5.247295103374589e-07,
"loss": 0.9566,
"step": 4820
},
{
"epoch": 1.6258205689277898,
"grad_norm": 4.34375,
"learning_rate": 5.15760037525824e-07,
"loss": 0.9591,
"step": 4830
},
{
"epoch": 1.6291870055546205,
"grad_norm": 3.984375,
"learning_rate": 5.068590618354779e-07,
"loss": 0.9758,
"step": 4840
},
{
"epoch": 1.632553442181451,
"grad_norm": 3.96875,
"learning_rate": 4.98026890533296e-07,
"loss": 0.9648,
"step": 4850
},
{
"epoch": 1.6359198788082814,
"grad_norm": 4.65625,
"learning_rate": 4.892638285109838e-07,
"loss": 0.9876,
"step": 4860
},
{
"epoch": 1.639286315435112,
"grad_norm": 3.765625,
"learning_rate": 4.805701782745559e-07,
"loss": 0.9649,
"step": 4870
},
{
"epoch": 1.6426527520619425,
"grad_norm": 4.03125,
"learning_rate": 4.7194623993389115e-07,
"loss": 0.9371,
"step": 4880
},
{
"epoch": 1.646019188688773,
"grad_norm": 3.65625,
"learning_rate": 4.633923111923738e-07,
"loss": 0.9475,
"step": 4890
},
{
"epoch": 1.6493856253156034,
"grad_norm": 3.984375,
"learning_rate": 4.549086873366132e-07,
"loss": 0.9662,
"step": 4900
},
{
"epoch": 1.6527520619424338,
"grad_norm": 4.125,
"learning_rate": 4.464956612262575e-07,
"loss": 0.9783,
"step": 4910
},
{
"epoch": 1.6561184985692643,
"grad_norm": 4.90625,
"learning_rate": 4.381535232838749e-07,
"loss": 0.9798,
"step": 4920
},
{
"epoch": 1.659484935196095,
"grad_norm": 4.3125,
"learning_rate": 4.298825614849356e-07,
"loss": 0.968,
"step": 4930
},
{
"epoch": 1.6628513718229254,
"grad_norm": 3.9375,
"learning_rate": 4.2168306134786673e-07,
"loss": 0.9763,
"step": 4940
},
{
"epoch": 1.666217808449756,
"grad_norm": 3.859375,
"learning_rate": 4.135553059241984e-07,
"loss": 0.969,
"step": 4950
},
{
"epoch": 1.6695842450765865,
"grad_norm": 3.96875,
"learning_rate": 4.0549957578878833e-07,
"loss": 0.9601,
"step": 4960
},
{
"epoch": 1.672950681703417,
"grad_norm": 3.78125,
"learning_rate": 3.975161490301424e-07,
"loss": 0.9458,
"step": 4970
},
{
"epoch": 1.6763171183302474,
"grad_norm": 4.46875,
"learning_rate": 3.896053012408102e-07,
"loss": 0.9733,
"step": 4980
},
{
"epoch": 1.6796835549570779,
"grad_norm": 3.875,
"learning_rate": 3.8176730550787437e-07,
"loss": 0.9491,
"step": 4990
},
{
"epoch": 1.6830499915839083,
"grad_norm": 3.75,
"learning_rate": 3.740024324035188e-07,
"loss": 0.9632,
"step": 5000
},
{
"epoch": 1.6864164282107388,
"grad_norm": 4.125,
"learning_rate": 3.6631094997569646e-07,
"loss": 0.9686,
"step": 5010
},
{
"epoch": 1.6897828648375695,
"grad_norm": 3.875,
"learning_rate": 3.586931237388669e-07,
"loss": 0.9435,
"step": 5020
},
{
"epoch": 1.6931493014644,
"grad_norm": 3.984375,
"learning_rate": 3.51149216664837e-07,
"loss": 0.957,
"step": 5030
},
{
"epoch": 1.6965157380912306,
"grad_norm": 4.375,
"learning_rate": 3.4367948917368084e-07,
"loss": 0.9671,
"step": 5040
},
{
"epoch": 1.699882174718061,
"grad_norm": 3.453125,
"learning_rate": 3.362841991247506e-07,
"loss": 0.9502,
"step": 5050
},
{
"epoch": 1.7032486113448915,
"grad_norm": 4.59375,
"learning_rate": 3.289636018077727e-07,
"loss": 0.988,
"step": 5060
},
{
"epoch": 1.706615047971722,
"grad_norm": 4.0625,
"learning_rate": 3.2171794993403866e-07,
"loss": 0.9607,
"step": 5070
},
{
"epoch": 1.7099814845985524,
"grad_norm": 4.125,
"learning_rate": 3.1454749362767876e-07,
"loss": 0.966,
"step": 5080
},
{
"epoch": 1.7133479212253828,
"grad_norm": 4.0,
"learning_rate": 3.074524804170292e-07,
"loss": 0.9621,
"step": 5090
},
{
"epoch": 1.7167143578522133,
"grad_norm": 4.46875,
"learning_rate": 3.004331552260845e-07,
"loss": 0.9409,
"step": 5100
},
{
"epoch": 1.720080794479044,
"grad_norm": 3.546875,
"learning_rate": 2.9348976036604777e-07,
"loss": 0.952,
"step": 5110
},
{
"epoch": 1.7234472311058744,
"grad_norm": 3.984375,
"learning_rate": 2.8662253552695946e-07,
"loss": 0.9521,
"step": 5120
},
{
"epoch": 1.726813667732705,
"grad_norm": 3.515625,
"learning_rate": 2.7983171776942876e-07,
"loss": 0.9861,
"step": 5130
},
{
"epoch": 1.7301801043595355,
"grad_norm": 4.09375,
"learning_rate": 2.731175415164464e-07,
"loss": 0.9539,
"step": 5140
},
{
"epoch": 1.733546540986366,
"grad_norm": 4.25,
"learning_rate": 2.664802385452955e-07,
"loss": 0.9658,
"step": 5150
},
{
"epoch": 1.7369129776131964,
"grad_norm": 3.78125,
"learning_rate": 2.599200379795455e-07,
"loss": 0.9437,
"step": 5160
},
{
"epoch": 1.7402794142400269,
"grad_norm": 3.984375,
"learning_rate": 2.5343716628114994e-07,
"loss": 0.9681,
"step": 5170
},
{
"epoch": 1.7436458508668573,
"grad_norm": 4.1875,
"learning_rate": 2.470318472426209e-07,
"loss": 0.9588,
"step": 5180
},
{
"epoch": 1.747012287493688,
"grad_norm": 3.96875,
"learning_rate": 2.407043019793098e-07,
"loss": 0.9586,
"step": 5190
},
{
"epoch": 1.7503787241205184,
"grad_norm": 3.5,
"learning_rate": 2.3445474892177077e-07,
"loss": 0.9874,
"step": 5200
},
{
"epoch": 1.753745160747349,
"grad_norm": 4.21875,
"learning_rate": 2.2828340380822268e-07,
"loss": 0.9652,
"step": 5210
},
{
"epoch": 1.7571115973741795,
"grad_norm": 3.828125,
"learning_rate": 2.2219047967709879e-07,
"loss": 0.9829,
"step": 5220
},
{
"epoch": 1.76047803400101,
"grad_norm": 4.34375,
"learning_rate": 2.1617618685969538e-07,
"loss": 0.9458,
"step": 5230
},
{
"epoch": 1.7638444706278404,
"grad_norm": 4.125,
"learning_rate": 2.1024073297291032e-07,
"loss": 0.9753,
"step": 5240
},
{
"epoch": 1.767210907254671,
"grad_norm": 4.125,
"learning_rate": 2.0438432291207511e-07,
"loss": 0.9649,
"step": 5250
},
{
"epoch": 1.7705773438815013,
"grad_norm": 4.125,
"learning_rate": 1.986071588438815e-07,
"loss": 0.954,
"step": 5260
},
{
"epoch": 1.7739437805083318,
"grad_norm": 3.765625,
"learning_rate": 1.9290944019940527e-07,
"loss": 0.957,
"step": 5270
},
{
"epoch": 1.7773102171351625,
"grad_norm": 3.984375,
"learning_rate": 1.8729136366721717e-07,
"loss": 0.954,
"step": 5280
},
{
"epoch": 1.780676653761993,
"grad_norm": 4.0,
"learning_rate": 1.8175312318659765e-07,
"loss": 0.9993,
"step": 5290
},
{
"epoch": 1.7840430903888236,
"grad_norm": 4.90625,
"learning_rate": 1.762949099408401e-07,
"loss": 0.9697,
"step": 5300
},
{
"epoch": 1.787409527015654,
"grad_norm": 3.84375,
"learning_rate": 1.7091691235065017e-07,
"loss": 0.9588,
"step": 5310
},
{
"epoch": 1.7907759636424845,
"grad_norm": 3.9375,
"learning_rate": 1.6561931606764308e-07,
"loss": 0.981,
"step": 5320
},
{
"epoch": 1.794142400269315,
"grad_norm": 3.953125,
"learning_rate": 1.6040230396793226e-07,
"loss": 0.9649,
"step": 5330
},
{
"epoch": 1.7975088368961454,
"grad_norm": 3.71875,
"learning_rate": 1.5526605614582162e-07,
"loss": 0.9597,
"step": 5340
},
{
"epoch": 1.8008752735229758,
"grad_norm": 3.765625,
"learning_rate": 1.502107499075814e-07,
"loss": 0.9458,
"step": 5350
},
{
"epoch": 1.8042417101498063,
"grad_norm": 3.890625,
"learning_rate": 1.4523655976533297e-07,
"loss": 0.9793,
"step": 5360
},
{
"epoch": 1.807608146776637,
"grad_norm": 3.84375,
"learning_rate": 1.403436574310227e-07,
"loss": 0.9642,
"step": 5370
},
{
"epoch": 1.8109745834034674,
"grad_norm": 3.75,
"learning_rate": 1.3553221181049403e-07,
"loss": 0.9473,
"step": 5380
},
{
"epoch": 1.814341020030298,
"grad_norm": 4.1875,
"learning_rate": 1.3080238899765702e-07,
"loss": 0.9883,
"step": 5390
},
{
"epoch": 1.8177074566571285,
"grad_norm": 4.1875,
"learning_rate": 1.26154352268755e-07,
"loss": 0.9579,
"step": 5400
},
{
"epoch": 1.821073893283959,
"grad_norm": 3.84375,
"learning_rate": 1.2158826207672807e-07,
"loss": 0.9596,
"step": 5410
},
{
"epoch": 1.8244403299107894,
"grad_norm": 4.15625,
"learning_rate": 1.1710427604567421e-07,
"loss": 0.9611,
"step": 5420
},
{
"epoch": 1.8278067665376199,
"grad_norm": 4.3125,
"learning_rate": 1.1270254896540656e-07,
"loss": 0.9728,
"step": 5430
},
{
"epoch": 1.8311732031644503,
"grad_norm": 4.15625,
"learning_rate": 1.0838323278611368e-07,
"loss": 0.9469,
"step": 5440
},
{
"epoch": 1.8345396397912808,
"grad_norm": 4.875,
"learning_rate": 1.041464766131095e-07,
"loss": 0.9584,
"step": 5450
},
{
"epoch": 1.8379060764181114,
"grad_norm": 4.09375,
"learning_rate": 9.999242670168913e-08,
"loss": 0.9616,
"step": 5460
},
{
"epoch": 1.8412725130449419,
"grad_norm": 3.890625,
"learning_rate": 9.59212264520798e-08,
"loss": 0.9756,
"step": 5470
},
{
"epoch": 1.8446389496717726,
"grad_norm": 4.125,
"learning_rate": 9.19330164044896e-08,
"loss": 0.9389,
"step": 5480
},
{
"epoch": 1.848005386298603,
"grad_norm": 4.21875,
"learning_rate": 8.802793423425604e-08,
"loss": 0.9833,
"step": 5490
},
{
"epoch": 1.8513718229254335,
"grad_norm": 3.765625,
"learning_rate": 8.420611474709461e-08,
"loss": 0.9198,
"step": 5500
},
{
"epoch": 1.854738259552264,
"grad_norm": 4.4375,
"learning_rate": 8.04676898744436e-08,
"loss": 0.9694,
"step": 5510
},
{
"epoch": 1.8581046961790944,
"grad_norm": 3.890625,
"learning_rate": 7.681278866891157e-08,
"loss": 0.9598,
"step": 5520
},
{
"epoch": 1.8614711328059248,
"grad_norm": 3.859375,
"learning_rate": 7.324153729981991e-08,
"loss": 0.9512,
"step": 5530
},
{
"epoch": 1.8648375694327555,
"grad_norm": 4.15625,
"learning_rate": 6.975405904885074e-08,
"loss": 0.9652,
"step": 5540
},
{
"epoch": 1.868204006059586,
"grad_norm": 4.1875,
"learning_rate": 6.635047430578717e-08,
"loss": 0.9519,
"step": 5550
},
{
"epoch": 1.8715704426864166,
"grad_norm": 4.15625,
"learning_rate": 6.30309005643609e-08,
"loss": 0.9573,
"step": 5560
},
{
"epoch": 1.874936879313247,
"grad_norm": 3.984375,
"learning_rate": 5.979545241819456e-08,
"loss": 0.9732,
"step": 5570
},
{
"epoch": 1.8783033159400775,
"grad_norm": 4.375,
"learning_rate": 5.664424155684606e-08,
"loss": 0.9534,
"step": 5580
},
{
"epoch": 1.881669752566908,
"grad_norm": 3.796875,
"learning_rate": 5.35773767619524e-08,
"loss": 0.9714,
"step": 5590
},
{
"epoch": 1.8850361891937384,
"grad_norm": 4.0625,
"learning_rate": 5.059496390347668e-08,
"loss": 0.9429,
"step": 5600
},
{
"epoch": 1.8884026258205688,
"grad_norm": 3.8125,
"learning_rate": 4.769710593605037e-08,
"loss": 0.9424,
"step": 5610
},
{
"epoch": 1.8917690624473993,
"grad_norm": 3.984375,
"learning_rate": 4.488390289542149e-08,
"loss": 0.9557,
"step": 5620
},
{
"epoch": 1.89513549907423,
"grad_norm": 4.21875,
"learning_rate": 4.215545189500042e-08,
"loss": 0.9971,
"step": 5630
},
{
"epoch": 1.8985019357010604,
"grad_norm": 3.90625,
"learning_rate": 3.9511847122508107e-08,
"loss": 0.9691,
"step": 5640
},
{
"epoch": 1.901868372327891,
"grad_norm": 3.875,
"learning_rate": 3.695317983672342e-08,
"loss": 0.9713,
"step": 5650
},
{
"epoch": 1.9052348089547215,
"grad_norm": 3.859375,
"learning_rate": 3.4479538364333984e-08,
"loss": 0.9698,
"step": 5660
},
{
"epoch": 1.908601245581552,
"grad_norm": 3.9375,
"learning_rate": 3.209100809688698e-08,
"loss": 0.9592,
"step": 5670
},
{
"epoch": 1.9119676822083824,
"grad_norm": 4.25,
"learning_rate": 2.9787671487840896e-08,
"loss": 0.9806,
"step": 5680
},
{
"epoch": 1.9153341188352129,
"grad_norm": 3.828125,
"learning_rate": 2.7569608049719533e-08,
"loss": 0.9664,
"step": 5690
},
{
"epoch": 1.9187005554620433,
"grad_norm": 3.46875,
"learning_rate": 2.5436894351366926e-08,
"loss": 0.96,
"step": 5700
},
{
"epoch": 1.9220669920888738,
"grad_norm": 3.71875,
"learning_rate": 2.3389604015304214e-08,
"loss": 0.9766,
"step": 5710
},
{
"epoch": 1.9254334287157044,
"grad_norm": 3.640625,
"learning_rate": 2.142780771518832e-08,
"loss": 0.9732,
"step": 5720
},
{
"epoch": 1.928799865342535,
"grad_norm": 3.703125,
"learning_rate": 1.9551573173372528e-08,
"loss": 0.9726,
"step": 5730
},
{
"epoch": 1.9321663019693656,
"grad_norm": 4.1875,
"learning_rate": 1.776096515856779e-08,
"loss": 0.9566,
"step": 5740
},
{
"epoch": 1.935532738596196,
"grad_norm": 4.40625,
"learning_rate": 1.605604548360784e-08,
"loss": 0.9396,
"step": 5750
},
{
"epoch": 1.9388991752230265,
"grad_norm": 4.03125,
"learning_rate": 1.4436873003314256e-08,
"loss": 0.9493,
"step": 5760
},
{
"epoch": 1.942265611849857,
"grad_norm": 3.828125,
"learning_rate": 1.290350361246584e-08,
"loss": 0.9344,
"step": 5770
},
{
"epoch": 1.9456320484766874,
"grad_norm": 4.03125,
"learning_rate": 1.1455990243869064e-08,
"loss": 0.9567,
"step": 5780
},
{
"epoch": 1.9489984851035178,
"grad_norm": 4.375,
"learning_rate": 1.0094382866529806e-08,
"loss": 0.9512,
"step": 5790
},
{
"epoch": 1.9523649217303485,
"grad_norm": 4.21875,
"learning_rate": 8.818728483929173e-09,
"loss": 0.9678,
"step": 5800
},
{
"epoch": 1.955731358357179,
"grad_norm": 3.890625,
"learning_rate": 7.629071132400634e-09,
"loss": 0.9531,
"step": 5810
},
{
"epoch": 1.9590977949840094,
"grad_norm": 3.828125,
"learning_rate": 6.5254518796104006e-09,
"loss": 0.9878,
"step": 5820
},
{
"epoch": 1.96246423161084,
"grad_norm": 4.46875,
"learning_rate": 5.507908823139119e-09,
"loss": 0.9671,
"step": 5830
},
{
"epoch": 1.9658306682376705,
"grad_norm": 3.875,
"learning_rate": 4.576477089166531e-09,
"loss": 0.9584,
"step": 5840
},
{
"epoch": 1.969197104864501,
"grad_norm": 3.921875,
"learning_rate": 3.731188831259669e-09,
"loss": 0.9612,
"step": 5850
},
{
"epoch": 1.9725635414913314,
"grad_norm": 4.1875,
"learning_rate": 2.972073229262351e-09,
"loss": 0.9478,
"step": 5860
},
{
"epoch": 1.9759299781181618,
"grad_norm": 3.625,
"learning_rate": 2.299156488288212e-09,
"loss": 0.9604,
"step": 5870
},
{
"epoch": 1.9792964147449923,
"grad_norm": 3.984375,
"learning_rate": 1.7124618378155933e-09,
"loss": 0.9548,
"step": 5880
},
{
"epoch": 1.982662851371823,
"grad_norm": 5.375,
"learning_rate": 1.2120095308865153e-09,
"loss": 0.9716,
"step": 5890
},
{
"epoch": 1.9860292879986534,
"grad_norm": 4.46875,
"learning_rate": 7.978168434064071e-10,
"loss": 0.9613,
"step": 5900
},
{
"epoch": 1.989395724625484,
"grad_norm": 4.0,
"learning_rate": 4.698980735484693e-10,
"loss": 0.9701,
"step": 5910
},
{
"epoch": 1.9927621612523145,
"grad_norm": 4.0625,
"learning_rate": 2.2826454125990382e-10,
"loss": 0.9562,
"step": 5920
},
{
"epoch": 1.996128597879145,
"grad_norm": 4.75,
"learning_rate": 7.292458787167e-11,
"loss": 0.9409,
"step": 5930
},
{
"epoch": 1.9994950345059754,
"grad_norm": 4.4375,
"learning_rate": 3.8835758089939e-12,
"loss": 0.9472,
"step": 5940
}
],
"logging_steps": 10,
"max_steps": 5942,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 7.151780936649708e+17,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}