tablegpt_judege_model / trainer_state.json
yangsaisai's picture
Upload trainer_state.json with huggingface_hub
0efa105 verified
Raw
History Blame Contribute Delete
56 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.818937939859245,
"eval_steps": 100000,
"global_step": 320,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0025591810620601407,
"grad_norm": 7.682459354400635,
"learning_rate": 0.0,
"loss": 1.2749,
"step": 1
},
{
"epoch": 0.005118362124120281,
"grad_norm": 7.725142478942871,
"learning_rate": 8.333333333333333e-07,
"loss": 1.3597,
"step": 2
},
{
"epoch": 0.007677543186180422,
"grad_norm": 8.06866455078125,
"learning_rate": 1.6666666666666667e-06,
"loss": 1.3199,
"step": 3
},
{
"epoch": 0.010236724248240563,
"grad_norm": 8.006539344787598,
"learning_rate": 2.5e-06,
"loss": 1.2915,
"step": 4
},
{
"epoch": 0.012795905310300703,
"grad_norm": 6.778942584991455,
"learning_rate": 3.3333333333333333e-06,
"loss": 1.2562,
"step": 5
},
{
"epoch": 0.015355086372360844,
"grad_norm": 5.596158504486084,
"learning_rate": 4.166666666666667e-06,
"loss": 1.2589,
"step": 6
},
{
"epoch": 0.017914267434420986,
"grad_norm": 4.587517738342285,
"learning_rate": 5e-06,
"loss": 1.2215,
"step": 7
},
{
"epoch": 0.020473448496481125,
"grad_norm": 4.011584281921387,
"learning_rate": 5.833333333333334e-06,
"loss": 1.1637,
"step": 8
},
{
"epoch": 0.023032629558541268,
"grad_norm": 3.852592945098877,
"learning_rate": 6.666666666666667e-06,
"loss": 1.1721,
"step": 9
},
{
"epoch": 0.025591810620601407,
"grad_norm": 3.982722282409668,
"learning_rate": 7.500000000000001e-06,
"loss": 1.1464,
"step": 10
},
{
"epoch": 0.02815099168266155,
"grad_norm": 7.547529220581055,
"learning_rate": 8.333333333333334e-06,
"loss": 1.099,
"step": 11
},
{
"epoch": 0.030710172744721688,
"grad_norm": 4.825142860412598,
"learning_rate": 9.166666666666666e-06,
"loss": 1.1036,
"step": 12
},
{
"epoch": 0.03326935380678183,
"grad_norm": 4.286171913146973,
"learning_rate": 1e-05,
"loss": 1.1142,
"step": 13
},
{
"epoch": 0.03582853486884197,
"grad_norm": 3.1489102840423584,
"learning_rate": 9.999828225438954e-06,
"loss": 1.0998,
"step": 14
},
{
"epoch": 0.03838771593090211,
"grad_norm": 3.6652419567108154,
"learning_rate": 9.999312913558414e-06,
"loss": 1.0305,
"step": 15
},
{
"epoch": 0.04094689699296225,
"grad_norm": 3.5285732746124268,
"learning_rate": 9.998454099765368e-06,
"loss": 1.1304,
"step": 16
},
{
"epoch": 0.04350607805502239,
"grad_norm": 3.437713146209717,
"learning_rate": 9.997251843068763e-06,
"loss": 1.043,
"step": 17
},
{
"epoch": 0.046065259117082535,
"grad_norm": 3.048311710357666,
"learning_rate": 9.995706226075445e-06,
"loss": 1.0199,
"step": 18
},
{
"epoch": 0.04862444017914268,
"grad_norm": 2.740447521209717,
"learning_rate": 9.993817354984485e-06,
"loss": 0.9405,
"step": 19
},
{
"epoch": 0.05118362124120281,
"grad_norm": 2.4751198291778564,
"learning_rate": 9.991585359579886e-06,
"loss": 0.967,
"step": 20
},
{
"epoch": 0.053742802303262956,
"grad_norm": 2.070446252822876,
"learning_rate": 9.989010393221657e-06,
"loss": 0.9945,
"step": 21
},
{
"epoch": 0.0563019833653231,
"grad_norm": 2.160064458847046,
"learning_rate": 9.986092632835287e-06,
"loss": 1.0185,
"step": 22
},
{
"epoch": 0.05886116442738324,
"grad_norm": 2.309260606765747,
"learning_rate": 9.982832278899582e-06,
"loss": 0.9912,
"step": 23
},
{
"epoch": 0.061420345489443376,
"grad_norm": 2.4544034004211426,
"learning_rate": 9.979229555432884e-06,
"loss": 0.9877,
"step": 24
},
{
"epoch": 0.06397952655150352,
"grad_norm": 2.31300687789917,
"learning_rate": 9.97528470997769e-06,
"loss": 0.9774,
"step": 25
},
{
"epoch": 0.06653870761356366,
"grad_norm": 2.2373909950256348,
"learning_rate": 9.970998013583643e-06,
"loss": 0.9442,
"step": 26
},
{
"epoch": 0.0690978886756238,
"grad_norm": 2.315127372741699,
"learning_rate": 9.966369760788895e-06,
"loss": 0.9787,
"step": 27
},
{
"epoch": 0.07165706973768395,
"grad_norm": 2.379563093185425,
"learning_rate": 9.961400269599885e-06,
"loss": 0.9461,
"step": 28
},
{
"epoch": 0.07421625079974409,
"grad_norm": 2.19171404838562,
"learning_rate": 9.956089881469481e-06,
"loss": 1.0349,
"step": 29
},
{
"epoch": 0.07677543186180422,
"grad_norm": 2.425921678543091,
"learning_rate": 9.950438961273517e-06,
"loss": 1.019,
"step": 30
},
{
"epoch": 0.07933461292386436,
"grad_norm": 2.2654383182525635,
"learning_rate": 9.94444789728573e-06,
"loss": 0.9672,
"step": 31
},
{
"epoch": 0.0818937939859245,
"grad_norm": 2.2154006958007812,
"learning_rate": 9.93811710115107e-06,
"loss": 0.9734,
"step": 32
},
{
"epoch": 0.08445297504798464,
"grad_norm": 2.5069079399108887,
"learning_rate": 9.931447007857433e-06,
"loss": 0.9082,
"step": 33
},
{
"epoch": 0.08701215611004479,
"grad_norm": 2.182760238647461,
"learning_rate": 9.924438075705755e-06,
"loss": 0.9307,
"step": 34
},
{
"epoch": 0.08957133717210493,
"grad_norm": 2.1564674377441406,
"learning_rate": 9.917090786278533e-06,
"loss": 0.9223,
"step": 35
},
{
"epoch": 0.09213051823416507,
"grad_norm": 1.9984478950500488,
"learning_rate": 9.909405644406738e-06,
"loss": 0.9458,
"step": 36
},
{
"epoch": 0.09468969929622521,
"grad_norm": 1.898442268371582,
"learning_rate": 9.901383178135113e-06,
"loss": 0.9263,
"step": 37
},
{
"epoch": 0.09724888035828536,
"grad_norm": 2.001455783843994,
"learning_rate": 9.893023938685911e-06,
"loss": 0.968,
"step": 38
},
{
"epoch": 0.09980806142034548,
"grad_norm": 2.1720826625823975,
"learning_rate": 9.884328500421005e-06,
"loss": 0.9054,
"step": 39
},
{
"epoch": 0.10236724248240563,
"grad_norm": 2.0314197540283203,
"learning_rate": 9.87529746080243e-06,
"loss": 0.9791,
"step": 40
},
{
"epoch": 0.10492642354446577,
"grad_norm": 2.1436691284179688,
"learning_rate": 9.865931440351338e-06,
"loss": 0.8894,
"step": 41
},
{
"epoch": 0.10748560460652591,
"grad_norm": 2.077209711074829,
"learning_rate": 9.856231082605344e-06,
"loss": 0.9522,
"step": 42
},
{
"epoch": 0.11004478566858605,
"grad_norm": 1.8568381071090698,
"learning_rate": 9.846197054074325e-06,
"loss": 0.8813,
"step": 43
},
{
"epoch": 0.1126039667306462,
"grad_norm": 1.9077541828155518,
"learning_rate": 9.835830044194625e-06,
"loss": 0.9361,
"step": 44
},
{
"epoch": 0.11516314779270634,
"grad_norm": 1.7717499732971191,
"learning_rate": 9.825130765281668e-06,
"loss": 0.8744,
"step": 45
},
{
"epoch": 0.11772232885476648,
"grad_norm": 2.042828321456909,
"learning_rate": 9.814099952481031e-06,
"loss": 0.9127,
"step": 46
},
{
"epoch": 0.12028150991682661,
"grad_norm": 1.9925884008407593,
"learning_rate": 9.802738363717927e-06,
"loss": 0.9106,
"step": 47
},
{
"epoch": 0.12284069097888675,
"grad_norm": 2.0410268306732178,
"learning_rate": 9.791046779645121e-06,
"loss": 0.9064,
"step": 48
},
{
"epoch": 0.1253998720409469,
"grad_norm": 1.9194778203964233,
"learning_rate": 9.779026003589304e-06,
"loss": 0.8866,
"step": 49
},
{
"epoch": 0.12795905310300704,
"grad_norm": 2.0479984283447266,
"learning_rate": 9.766676861495888e-06,
"loss": 0.9138,
"step": 50
},
{
"epoch": 0.13051823416506717,
"grad_norm": 2.0392417907714844,
"learning_rate": 9.754000201872258e-06,
"loss": 0.8566,
"step": 51
},
{
"epoch": 0.13307741522712732,
"grad_norm": 2.205040454864502,
"learning_rate": 9.74099689572947e-06,
"loss": 0.8537,
"step": 52
},
{
"epoch": 0.13563659628918745,
"grad_norm": 2.0679237842559814,
"learning_rate": 9.727667836522408e-06,
"loss": 0.8973,
"step": 53
},
{
"epoch": 0.1381957773512476,
"grad_norm": 1.9401817321777344,
"learning_rate": 9.714013940088388e-06,
"loss": 0.8036,
"step": 54
},
{
"epoch": 0.14075495841330773,
"grad_norm": 1.8604705333709717,
"learning_rate": 9.700036144584237e-06,
"loss": 0.8858,
"step": 55
},
{
"epoch": 0.1433141394753679,
"grad_norm": 1.7629581689834595,
"learning_rate": 9.68573541042183e-06,
"loss": 0.852,
"step": 56
},
{
"epoch": 0.14587332053742802,
"grad_norm": 2.052168369293213,
"learning_rate": 9.6711127202021e-06,
"loss": 0.9272,
"step": 57
},
{
"epoch": 0.14843250159948818,
"grad_norm": 1.6959059238433838,
"learning_rate": 9.656169078647525e-06,
"loss": 0.8319,
"step": 58
},
{
"epoch": 0.1509916826615483,
"grad_norm": 1.8046302795410156,
"learning_rate": 9.640905512533092e-06,
"loss": 0.8818,
"step": 59
},
{
"epoch": 0.15355086372360843,
"grad_norm": 1.7874407768249512,
"learning_rate": 9.625323070615751e-06,
"loss": 0.9313,
"step": 60
},
{
"epoch": 0.1561100447856686,
"grad_norm": 1.7794876098632812,
"learning_rate": 9.609422823562345e-06,
"loss": 0.8552,
"step": 61
},
{
"epoch": 0.15866922584772872,
"grad_norm": 3.3993961811065674,
"learning_rate": 9.593205863876062e-06,
"loss": 0.9135,
"step": 62
},
{
"epoch": 0.16122840690978887,
"grad_norm": 2.0278658866882324,
"learning_rate": 9.576673305821353e-06,
"loss": 0.9081,
"step": 63
},
{
"epoch": 0.163787587971849,
"grad_norm": 1.9209970235824585,
"learning_rate": 9.55982628534738e-06,
"loss": 0.841,
"step": 64
},
{
"epoch": 0.16634676903390916,
"grad_norm": 1.702495813369751,
"learning_rate": 9.54266596000996e-06,
"loss": 0.8834,
"step": 65
},
{
"epoch": 0.1689059500959693,
"grad_norm": 1.776999831199646,
"learning_rate": 9.525193508892034e-06,
"loss": 0.9429,
"step": 66
},
{
"epoch": 0.17146513115802944,
"grad_norm": 1.8628787994384766,
"learning_rate": 9.507410132522652e-06,
"loss": 0.7964,
"step": 67
},
{
"epoch": 0.17402431222008957,
"grad_norm": 1.909364104270935,
"learning_rate": 9.489317052794482e-06,
"loss": 0.8362,
"step": 68
},
{
"epoch": 0.1765834932821497,
"grad_norm": 1.8521013259887695,
"learning_rate": 9.470915512879853e-06,
"loss": 0.8874,
"step": 69
},
{
"epoch": 0.17914267434420986,
"grad_norm": 1.777940034866333,
"learning_rate": 9.452206777145343e-06,
"loss": 0.8477,
"step": 70
},
{
"epoch": 0.18170185540626999,
"grad_norm": 1.9356168508529663,
"learning_rate": 9.4331921310649e-06,
"loss": 0.7606,
"step": 71
},
{
"epoch": 0.18426103646833014,
"grad_norm": 1.8606858253479004,
"learning_rate": 9.413872881131519e-06,
"loss": 0.855,
"step": 72
},
{
"epoch": 0.18682021753039027,
"grad_norm": 1.7669318914413452,
"learning_rate": 9.394250354767467e-06,
"loss": 0.8688,
"step": 73
},
{
"epoch": 0.18937939859245043,
"grad_norm": 1.6513736248016357,
"learning_rate": 9.374325900233087e-06,
"loss": 0.8978,
"step": 74
},
{
"epoch": 0.19193857965451055,
"grad_norm": 1.7595405578613281,
"learning_rate": 9.354100886534152e-06,
"loss": 0.8259,
"step": 75
},
{
"epoch": 0.1944977607165707,
"grad_norm": 1.7387309074401855,
"learning_rate": 9.333576703327803e-06,
"loss": 0.8004,
"step": 76
},
{
"epoch": 0.19705694177863084,
"grad_norm": 1.7983427047729492,
"learning_rate": 9.312754760827061e-06,
"loss": 0.8085,
"step": 77
},
{
"epoch": 0.19961612284069097,
"grad_norm": 1.7982909679412842,
"learning_rate": 9.291636489703943e-06,
"loss": 0.8169,
"step": 78
},
{
"epoch": 0.20217530390275112,
"grad_norm": 1.7562614679336548,
"learning_rate": 9.270223340991147e-06,
"loss": 0.8447,
"step": 79
},
{
"epoch": 0.20473448496481125,
"grad_norm": 1.599975347518921,
"learning_rate": 9.248516785982365e-06,
"loss": 0.8606,
"step": 80
},
{
"epoch": 0.2072936660268714,
"grad_norm": 1.976728081703186,
"learning_rate": 9.226518316131176e-06,
"loss": 0.8695,
"step": 81
},
{
"epoch": 0.20985284708893154,
"grad_norm": 2.3505756855010986,
"learning_rate": 9.204229442948585e-06,
"loss": 0.7961,
"step": 82
},
{
"epoch": 0.2124120281509917,
"grad_norm": 2.018191337585449,
"learning_rate": 9.181651697899153e-06,
"loss": 0.8231,
"step": 83
},
{
"epoch": 0.21497120921305182,
"grad_norm": 1.7307785749435425,
"learning_rate": 9.158786632295776e-06,
"loss": 0.7992,
"step": 84
},
{
"epoch": 0.21753039027511195,
"grad_norm": 1.9073212146759033,
"learning_rate": 9.1356358171931e-06,
"loss": 0.7696,
"step": 85
},
{
"epoch": 0.2200895713371721,
"grad_norm": 1.8640539646148682,
"learning_rate": 9.112200843279565e-06,
"loss": 0.8257,
"step": 86
},
{
"epoch": 0.22264875239923224,
"grad_norm": 1.9237738847732544,
"learning_rate": 9.088483320768114e-06,
"loss": 0.7695,
"step": 87
},
{
"epoch": 0.2252079334612924,
"grad_norm": 1.8698316812515259,
"learning_rate": 9.064484879285555e-06,
"loss": 0.8464,
"step": 88
},
{
"epoch": 0.22776711452335252,
"grad_norm": 1.8868035078048706,
"learning_rate": 9.040207167760587e-06,
"loss": 0.8011,
"step": 89
},
{
"epoch": 0.23032629558541268,
"grad_norm": 1.7685335874557495,
"learning_rate": 9.015651854310505e-06,
"loss": 0.7802,
"step": 90
},
{
"epoch": 0.2328854766474728,
"grad_norm": 1.876893162727356,
"learning_rate": 8.99082062612659e-06,
"loss": 0.7589,
"step": 91
},
{
"epoch": 0.23544465770953296,
"grad_norm": 1.7226823568344116,
"learning_rate": 8.965715189358165e-06,
"loss": 0.8311,
"step": 92
},
{
"epoch": 0.2380038387715931,
"grad_norm": 1.6303762197494507,
"learning_rate": 8.940337268995385e-06,
"loss": 0.7922,
"step": 93
},
{
"epoch": 0.24056301983365322,
"grad_norm": 1.7937066555023193,
"learning_rate": 8.9146886087507e-06,
"loss": 0.8152,
"step": 94
},
{
"epoch": 0.24312220089571338,
"grad_norm": 1.8004367351531982,
"learning_rate": 8.888770970939058e-06,
"loss": 0.8162,
"step": 95
},
{
"epoch": 0.2456813819577735,
"grad_norm": 1.6512545347213745,
"learning_rate": 8.862586136356794e-06,
"loss": 0.795,
"step": 96
},
{
"epoch": 0.24824056301983366,
"grad_norm": 1.8923460245132446,
"learning_rate": 8.836135904159301e-06,
"loss": 0.7437,
"step": 97
},
{
"epoch": 0.2507997440818938,
"grad_norm": 1.7516615390777588,
"learning_rate": 8.809422091737388e-06,
"loss": 0.7152,
"step": 98
},
{
"epoch": 0.2533589251439539,
"grad_norm": 1.6280677318572998,
"learning_rate": 8.782446534592412e-06,
"loss": 0.7774,
"step": 99
},
{
"epoch": 0.2559181062060141,
"grad_norm": 1.6521121263504028,
"learning_rate": 8.755211086210172e-06,
"loss": 0.7774,
"step": 100
},
{
"epoch": 0.25847728726807423,
"grad_norm": 1.7474745512008667,
"learning_rate": 8.727717617933545e-06,
"loss": 0.6834,
"step": 101
},
{
"epoch": 0.26103646833013433,
"grad_norm": 1.8764477968215942,
"learning_rate": 8.699968018833903e-06,
"loss": 0.7884,
"step": 102
},
{
"epoch": 0.2635956493921945,
"grad_norm": 1.820784091949463,
"learning_rate": 8.671964195581336e-06,
"loss": 0.8057,
"step": 103
},
{
"epoch": 0.26615483045425464,
"grad_norm": 1.8630762100219727,
"learning_rate": 8.643708072313619e-06,
"loss": 0.7576,
"step": 104
},
{
"epoch": 0.2687140115163148,
"grad_norm": 1.7553012371063232,
"learning_rate": 8.615201590504016e-06,
"loss": 0.807,
"step": 105
},
{
"epoch": 0.2712731925783749,
"grad_norm": 1.6087998151779175,
"learning_rate": 8.586446708827896e-06,
"loss": 0.7319,
"step": 106
},
{
"epoch": 0.27383237364043506,
"grad_norm": 1.8860429525375366,
"learning_rate": 8.557445403028122e-06,
"loss": 0.7882,
"step": 107
},
{
"epoch": 0.2763915547024952,
"grad_norm": 1.8590738773345947,
"learning_rate": 8.52819966577933e-06,
"loss": 0.7881,
"step": 108
},
{
"epoch": 0.27895073576455537,
"grad_norm": 1.6432468891143799,
"learning_rate": 8.498711506550984e-06,
"loss": 0.7186,
"step": 109
},
{
"epoch": 0.28150991682661547,
"grad_norm": 1.6418780088424683,
"learning_rate": 8.468982951469334e-06,
"loss": 0.7721,
"step": 110
},
{
"epoch": 0.2840690978886756,
"grad_norm": 1.6970170736312866,
"learning_rate": 8.439016043178176e-06,
"loss": 0.7377,
"step": 111
},
{
"epoch": 0.2866282789507358,
"grad_norm": 1.8257789611816406,
"learning_rate": 8.408812840698517e-06,
"loss": 0.7748,
"step": 112
},
{
"epoch": 0.2891874600127959,
"grad_norm": 1.8099708557128906,
"learning_rate": 8.378375419287098e-06,
"loss": 0.7326,
"step": 113
},
{
"epoch": 0.29174664107485604,
"grad_norm": 1.7570979595184326,
"learning_rate": 8.3477058702938e-06,
"loss": 0.7606,
"step": 114
},
{
"epoch": 0.2943058221369162,
"grad_norm": 1.608965516090393,
"learning_rate": 8.31680630101795e-06,
"loss": 0.7152,
"step": 115
},
{
"epoch": 0.29686500319897635,
"grad_norm": 1.6170052289962769,
"learning_rate": 8.285678834563524e-06,
"loss": 0.8083,
"step": 116
},
{
"epoch": 0.29942418426103645,
"grad_norm": 1.6714998483657837,
"learning_rate": 8.25432560969328e-06,
"loss": 0.7666,
"step": 117
},
{
"epoch": 0.3019833653230966,
"grad_norm": 1.7456778287887573,
"learning_rate": 8.22274878068179e-06,
"loss": 0.7733,
"step": 118
},
{
"epoch": 0.30454254638515676,
"grad_norm": 1.7775685787200928,
"learning_rate": 8.190950517167438e-06,
"loss": 0.6958,
"step": 119
},
{
"epoch": 0.30710172744721687,
"grad_norm": 1.687903881072998,
"learning_rate": 8.15893300400332e-06,
"loss": 0.7196,
"step": 120
},
{
"epoch": 0.309660908509277,
"grad_norm": 1.6974139213562012,
"learning_rate": 8.126698441107146e-06,
"loss": 0.7,
"step": 121
},
{
"epoch": 0.3122200895713372,
"grad_norm": 1.6907954216003418,
"learning_rate": 8.094249043310074e-06,
"loss": 0.7107,
"step": 122
},
{
"epoch": 0.31477927063339733,
"grad_norm": 1.7220592498779297,
"learning_rate": 8.061587040204528e-06,
"loss": 0.6674,
"step": 123
},
{
"epoch": 0.31733845169545744,
"grad_norm": 1.7368662357330322,
"learning_rate": 8.028714675991008e-06,
"loss": 0.7821,
"step": 124
},
{
"epoch": 0.3198976327575176,
"grad_norm": 1.603564739227295,
"learning_rate": 7.995634209323886e-06,
"loss": 0.7605,
"step": 125
},
{
"epoch": 0.32245681381957775,
"grad_norm": 1.6910033226013184,
"learning_rate": 7.962347913156217e-06,
"loss": 0.7544,
"step": 126
},
{
"epoch": 0.32501599488163785,
"grad_norm": 1.7523442506790161,
"learning_rate": 7.92885807458357e-06,
"loss": 0.7236,
"step": 127
},
{
"epoch": 0.327575175943698,
"grad_norm": 1.5784283876419067,
"learning_rate": 7.895166994686869e-06,
"loss": 0.6806,
"step": 128
},
{
"epoch": 0.33013435700575816,
"grad_norm": 1.863252878189087,
"learning_rate": 7.861276988374303e-06,
"loss": 0.661,
"step": 129
},
{
"epoch": 0.3326935380678183,
"grad_norm": 1.7507966756820679,
"learning_rate": 7.827190384222249e-06,
"loss": 0.7188,
"step": 130
},
{
"epoch": 0.3352527191298784,
"grad_norm": 1.931938648223877,
"learning_rate": 7.7929095243153e-06,
"loss": 0.7304,
"step": 131
},
{
"epoch": 0.3378119001919386,
"grad_norm": 1.8002139329910278,
"learning_rate": 7.758436764085315e-06,
"loss": 0.7354,
"step": 132
},
{
"epoch": 0.34037108125399873,
"grad_norm": 1.644242525100708,
"learning_rate": 7.723774472149601e-06,
"loss": 0.6318,
"step": 133
},
{
"epoch": 0.3429302623160589,
"grad_norm": 1.545650839805603,
"learning_rate": 7.68892503014815e-06,
"loss": 0.6308,
"step": 134
},
{
"epoch": 0.345489443378119,
"grad_norm": 2.2425730228424072,
"learning_rate": 7.653890832580003e-06,
"loss": 0.6681,
"step": 135
},
{
"epoch": 0.34804862444017914,
"grad_norm": 2.2011518478393555,
"learning_rate": 7.61867428663872e-06,
"loss": 0.7387,
"step": 136
},
{
"epoch": 0.3506078055022393,
"grad_norm": 1.6651183366775513,
"learning_rate": 7.583277812046993e-06,
"loss": 0.7516,
"step": 137
},
{
"epoch": 0.3531669865642994,
"grad_norm": 1.749775767326355,
"learning_rate": 7.547703840890377e-06,
"loss": 0.7005,
"step": 138
},
{
"epoch": 0.35572616762635956,
"grad_norm": 1.8172039985656738,
"learning_rate": 7.511954817450181e-06,
"loss": 0.6779,
"step": 139
},
{
"epoch": 0.3582853486884197,
"grad_norm": 1.8840566873550415,
"learning_rate": 7.476033198035532e-06,
"loss": 0.7019,
"step": 140
},
{
"epoch": 0.36084452975047987,
"grad_norm": 1.7813738584518433,
"learning_rate": 7.439941450814591e-06,
"loss": 0.6411,
"step": 141
},
{
"epoch": 0.36340371081253997,
"grad_norm": 2.064899206161499,
"learning_rate": 7.4036820556449704e-06,
"loss": 0.6898,
"step": 142
},
{
"epoch": 0.3659628918746001,
"grad_norm": 2.3556082248687744,
"learning_rate": 7.367257503903349e-06,
"loss": 0.7167,
"step": 143
},
{
"epoch": 0.3685220729366603,
"grad_norm": 1.6402297019958496,
"learning_rate": 7.330670298314282e-06,
"loss": 0.6812,
"step": 144
},
{
"epoch": 0.3710812539987204,
"grad_norm": 1.5648720264434814,
"learning_rate": 7.293922952778239e-06,
"loss": 0.7007,
"step": 145
},
{
"epoch": 0.37364043506078054,
"grad_norm": 1.7159013748168945,
"learning_rate": 7.257017992198879e-06,
"loss": 0.6953,
"step": 146
},
{
"epoch": 0.3761996161228407,
"grad_norm": 1.9053713083267212,
"learning_rate": 7.219957952309568e-06,
"loss": 0.6742,
"step": 147
},
{
"epoch": 0.37875879718490085,
"grad_norm": 1.7517788410186768,
"learning_rate": 7.182745379499134e-06,
"loss": 0.6177,
"step": 148
},
{
"epoch": 0.38131797824696095,
"grad_norm": 1.5937896966934204,
"learning_rate": 7.145382830636925e-06,
"loss": 0.7555,
"step": 149
},
{
"epoch": 0.3838771593090211,
"grad_norm": 1.6669344902038574,
"learning_rate": 7.107872872897112e-06,
"loss": 0.6428,
"step": 150
},
{
"epoch": 0.38643634037108127,
"grad_norm": 1.5842646360397339,
"learning_rate": 7.070218083582306e-06,
"loss": 0.6698,
"step": 151
},
{
"epoch": 0.3889955214331414,
"grad_norm": 1.6338943243026733,
"learning_rate": 7.032421049946467e-06,
"loss": 0.6241,
"step": 152
},
{
"epoch": 0.3915547024952015,
"grad_norm": 1.7281718254089355,
"learning_rate": 6.994484369017142e-06,
"loss": 0.6809,
"step": 153
},
{
"epoch": 0.3941138835572617,
"grad_norm": 1.8216824531555176,
"learning_rate": 6.9564106474170165e-06,
"loss": 0.7112,
"step": 154
},
{
"epoch": 0.39667306461932184,
"grad_norm": 1.8112353086471558,
"learning_rate": 6.9182025011848156e-06,
"loss": 0.673,
"step": 155
},
{
"epoch": 0.39923224568138194,
"grad_norm": 1.7320173978805542,
"learning_rate": 6.879862555595559e-06,
"loss": 0.7106,
"step": 156
},
{
"epoch": 0.4017914267434421,
"grad_norm": 1.792492389678955,
"learning_rate": 6.841393444980177e-06,
"loss": 0.5788,
"step": 157
},
{
"epoch": 0.40435060780550225,
"grad_norm": 1.6218078136444092,
"learning_rate": 6.802797812544503e-06,
"loss": 0.6232,
"step": 158
},
{
"epoch": 0.4069097888675624,
"grad_norm": 1.6211438179016113,
"learning_rate": 6.764078310187667e-06,
"loss": 0.5775,
"step": 159
},
{
"epoch": 0.4094689699296225,
"grad_norm": 2.64115571975708,
"learning_rate": 6.725237598319877e-06,
"loss": 0.6409,
"step": 160
},
{
"epoch": 0.41202815099168266,
"grad_norm": 2.3797988891601562,
"learning_rate": 6.686278345679626e-06,
"loss": 0.6405,
"step": 161
},
{
"epoch": 0.4145873320537428,
"grad_norm": 2.3996291160583496,
"learning_rate": 6.647203229150323e-06,
"loss": 0.6215,
"step": 162
},
{
"epoch": 0.4171465131158029,
"grad_norm": 1.6767221689224243,
"learning_rate": 6.608014933576362e-06,
"loss": 0.6925,
"step": 163
},
{
"epoch": 0.4197056941778631,
"grad_norm": 1.6698813438415527,
"learning_rate": 6.568716151578653e-06,
"loss": 0.6969,
"step": 164
},
{
"epoch": 0.42226487523992323,
"grad_norm": 1.82571280002594,
"learning_rate": 6.529309583369606e-06,
"loss": 0.696,
"step": 165
},
{
"epoch": 0.4248240563019834,
"grad_norm": 1.985872745513916,
"learning_rate": 6.489797936567604e-06,
"loss": 0.6506,
"step": 166
},
{
"epoch": 0.4273832373640435,
"grad_norm": 1.9063838720321655,
"learning_rate": 6.45018392601096e-06,
"loss": 0.6307,
"step": 167
},
{
"epoch": 0.42994241842610365,
"grad_norm": 1.7634496688842773,
"learning_rate": 6.410470273571386e-06,
"loss": 0.5424,
"step": 168
},
{
"epoch": 0.4325015994881638,
"grad_norm": 1.7788817882537842,
"learning_rate": 6.370659707966967e-06,
"loss": 0.6627,
"step": 169
},
{
"epoch": 0.4350607805502239,
"grad_norm": 1.6340923309326172,
"learning_rate": 6.3307549645746756e-06,
"loss": 0.553,
"step": 170
},
{
"epoch": 0.43761996161228406,
"grad_norm": 1.9470851421356201,
"learning_rate": 6.290758785242425e-06,
"loss": 0.6543,
"step": 171
},
{
"epoch": 0.4401791426743442,
"grad_norm": 1.6622216701507568,
"learning_rate": 6.250673918100671e-06,
"loss": 0.6361,
"step": 172
},
{
"epoch": 0.44273832373640437,
"grad_norm": 1.916314721107483,
"learning_rate": 6.210503117373601e-06,
"loss": 0.541,
"step": 173
},
{
"epoch": 0.44529750479846447,
"grad_norm": 1.7608107328414917,
"learning_rate": 6.170249143189878e-06,
"loss": 0.5893,
"step": 174
},
{
"epoch": 0.44785668586052463,
"grad_norm": 2.093899965286255,
"learning_rate": 6.129914761393001e-06,
"loss": 0.5947,
"step": 175
},
{
"epoch": 0.4504158669225848,
"grad_norm": 1.7893019914627075,
"learning_rate": 6.089502743351259e-06,
"loss": 0.6415,
"step": 176
},
{
"epoch": 0.45297504798464494,
"grad_norm": 1.7566556930541992,
"learning_rate": 6.049015865767318e-06,
"loss": 0.7068,
"step": 177
},
{
"epoch": 0.45553422904670504,
"grad_norm": 1.5285935401916504,
"learning_rate": 6.0084569104874276e-06,
"loss": 0.5907,
"step": 178
},
{
"epoch": 0.4580934101087652,
"grad_norm": 1.8482022285461426,
"learning_rate": 5.967828664310283e-06,
"loss": 0.6465,
"step": 179
},
{
"epoch": 0.46065259117082535,
"grad_norm": 1.5722378492355347,
"learning_rate": 5.927133918795548e-06,
"loss": 0.6183,
"step": 180
},
{
"epoch": 0.46321177223288545,
"grad_norm": 1.5821112394332886,
"learning_rate": 5.8863754700720396e-06,
"loss": 0.6004,
"step": 181
},
{
"epoch": 0.4657709532949456,
"grad_norm": 1.552962064743042,
"learning_rate": 5.845556118645613e-06,
"loss": 0.6697,
"step": 182
},
{
"epoch": 0.46833013435700577,
"grad_norm": 1.5706250667572021,
"learning_rate": 5.804678669206738e-06,
"loss": 0.6469,
"step": 183
},
{
"epoch": 0.4708893154190659,
"grad_norm": 1.5389466285705566,
"learning_rate": 5.763745930437787e-06,
"loss": 0.6115,
"step": 184
},
{
"epoch": 0.473448496481126,
"grad_norm": 1.6668463945388794,
"learning_rate": 5.722760714820057e-06,
"loss": 0.54,
"step": 185
},
{
"epoch": 0.4760076775431862,
"grad_norm": 1.6907564401626587,
"learning_rate": 5.681725838440515e-06,
"loss": 0.6831,
"step": 186
},
{
"epoch": 0.47856685860524634,
"grad_norm": 1.6007864475250244,
"learning_rate": 5.640644120798312e-06,
"loss": 0.6649,
"step": 187
},
{
"epoch": 0.48112603966730644,
"grad_norm": 1.5013468265533447,
"learning_rate": 5.599518384611052e-06,
"loss": 0.5745,
"step": 188
},
{
"epoch": 0.4836852207293666,
"grad_norm": 1.569190502166748,
"learning_rate": 5.558351455620852e-06,
"loss": 0.5363,
"step": 189
},
{
"epoch": 0.48624440179142675,
"grad_norm": 1.6902738809585571,
"learning_rate": 5.517146162400171e-06,
"loss": 0.6073,
"step": 190
},
{
"epoch": 0.4888035828534869,
"grad_norm": 1.6453821659088135,
"learning_rate": 5.475905336157473e-06,
"loss": 0.5941,
"step": 191
},
{
"epoch": 0.491362763915547,
"grad_norm": 1.6778477430343628,
"learning_rate": 5.434631810542688e-06,
"loss": 0.5273,
"step": 192
},
{
"epoch": 0.49392194497760716,
"grad_norm": 1.8814613819122314,
"learning_rate": 5.393328421452514e-06,
"loss": 0.5907,
"step": 193
},
{
"epoch": 0.4964811260396673,
"grad_norm": 2.279060125350952,
"learning_rate": 5.351998006835563e-06,
"loss": 0.6219,
"step": 194
},
{
"epoch": 0.4990403071017274,
"grad_norm": 1.5855976343154907,
"learning_rate": 5.3106434064973665e-06,
"loss": 0.6052,
"step": 195
},
{
"epoch": 0.5015994881637876,
"grad_norm": 1.6588892936706543,
"learning_rate": 5.269267461905253e-06,
"loss": 0.601,
"step": 196
},
{
"epoch": 0.5041586692258477,
"grad_norm": 1.602506160736084,
"learning_rate": 5.227873015993108e-06,
"loss": 0.6216,
"step": 197
},
{
"epoch": 0.5067178502879078,
"grad_norm": 1.5565695762634277,
"learning_rate": 5.186462912966047e-06,
"loss": 0.6205,
"step": 198
},
{
"epoch": 0.509277031349968,
"grad_norm": 1.565605640411377,
"learning_rate": 5.145039998104975e-06,
"loss": 0.5435,
"step": 199
},
{
"epoch": 0.5118362124120281,
"grad_norm": 1.9312242269515991,
"learning_rate": 5.1036071175710984e-06,
"loss": 0.6566,
"step": 200
},
{
"epoch": 0.5143953934740882,
"grad_norm": 1.989288568496704,
"learning_rate": 5.062167118210367e-06,
"loss": 0.5612,
"step": 201
},
{
"epoch": 0.5169545745361485,
"grad_norm": 1.5968483686447144,
"learning_rate": 5.020722847357858e-06,
"loss": 0.573,
"step": 202
},
{
"epoch": 0.5195137555982086,
"grad_norm": 1.5364960432052612,
"learning_rate": 4.979277152642145e-06,
"loss": 0.6093,
"step": 203
},
{
"epoch": 0.5220729366602687,
"grad_norm": 1.4956138134002686,
"learning_rate": 4.937832881789635e-06,
"loss": 0.5805,
"step": 204
},
{
"epoch": 0.5246321177223289,
"grad_norm": 1.5566569566726685,
"learning_rate": 4.8963928824289015e-06,
"loss": 0.542,
"step": 205
},
{
"epoch": 0.527191298784389,
"grad_norm": 1.7304389476776123,
"learning_rate": 4.854960001895027e-06,
"loss": 0.6403,
"step": 206
},
{
"epoch": 0.5297504798464492,
"grad_norm": 1.708113431930542,
"learning_rate": 4.813537087033954e-06,
"loss": 0.587,
"step": 207
},
{
"epoch": 0.5323096609085093,
"grad_norm": 1.5466324090957642,
"learning_rate": 4.772126984006892e-06,
"loss": 0.5617,
"step": 208
},
{
"epoch": 0.5348688419705694,
"grad_norm": 1.8750905990600586,
"learning_rate": 4.7307325380947495e-06,
"loss": 0.6119,
"step": 209
},
{
"epoch": 0.5374280230326296,
"grad_norm": 1.6555531024932861,
"learning_rate": 4.689356593502635e-06,
"loss": 0.5196,
"step": 210
},
{
"epoch": 0.5399872040946897,
"grad_norm": 1.6204813718795776,
"learning_rate": 4.648001993164438e-06,
"loss": 0.5631,
"step": 211
},
{
"epoch": 0.5425463851567498,
"grad_norm": 1.6162160634994507,
"learning_rate": 4.606671578547488e-06,
"loss": 0.6036,
"step": 212
},
{
"epoch": 0.54510556621881,
"grad_norm": 1.512351632118225,
"learning_rate": 4.565368189457313e-06,
"loss": 0.5054,
"step": 213
},
{
"epoch": 0.5476647472808701,
"grad_norm": 2.097134828567505,
"learning_rate": 4.524094663842528e-06,
"loss": 0.5553,
"step": 214
},
{
"epoch": 0.5502239283429302,
"grad_norm": 1.6303435564041138,
"learning_rate": 4.482853837599831e-06,
"loss": 0.5383,
"step": 215
},
{
"epoch": 0.5527831094049904,
"grad_norm": 1.5532830953598022,
"learning_rate": 4.441648544379149e-06,
"loss": 0.5538,
"step": 216
},
{
"epoch": 0.5553422904670505,
"grad_norm": 1.4504081010818481,
"learning_rate": 4.400481615388948e-06,
"loss": 0.5123,
"step": 217
},
{
"epoch": 0.5579014715291107,
"grad_norm": 1.6712675094604492,
"learning_rate": 4.359355879201691e-06,
"loss": 0.5813,
"step": 218
},
{
"epoch": 0.5604606525911708,
"grad_norm": 1.5415232181549072,
"learning_rate": 4.318274161559487e-06,
"loss": 0.6498,
"step": 219
},
{
"epoch": 0.5630198336532309,
"grad_norm": 1.6677114963531494,
"learning_rate": 4.277239285179944e-06,
"loss": 0.5192,
"step": 220
},
{
"epoch": 0.5655790147152912,
"grad_norm": 1.721778392791748,
"learning_rate": 4.236254069562214e-06,
"loss": 0.585,
"step": 221
},
{
"epoch": 0.5681381957773513,
"grad_norm": 1.5012080669403076,
"learning_rate": 4.195321330793264e-06,
"loss": 0.5533,
"step": 222
},
{
"epoch": 0.5706973768394114,
"grad_norm": 1.5886061191558838,
"learning_rate": 4.154443881354388e-06,
"loss": 0.5456,
"step": 223
},
{
"epoch": 0.5732565579014716,
"grad_norm": 1.6147302389144897,
"learning_rate": 4.113624529927963e-06,
"loss": 0.5556,
"step": 224
},
{
"epoch": 0.5758157389635317,
"grad_norm": 1.5829546451568604,
"learning_rate": 4.072866081204453e-06,
"loss": 0.5706,
"step": 225
},
{
"epoch": 0.5783749200255918,
"grad_norm": 1.7815481424331665,
"learning_rate": 4.032171335689718e-06,
"loss": 0.5703,
"step": 226
},
{
"epoch": 0.580934101087652,
"grad_norm": 1.7351888418197632,
"learning_rate": 3.991543089512574e-06,
"loss": 0.5189,
"step": 227
},
{
"epoch": 0.5834932821497121,
"grad_norm": 1.560595989227295,
"learning_rate": 3.9509841342326835e-06,
"loss": 0.6063,
"step": 228
},
{
"epoch": 0.5860524632117722,
"grad_norm": 1.6448612213134766,
"learning_rate": 3.910497256648742e-06,
"loss": 0.5401,
"step": 229
},
{
"epoch": 0.5886116442738324,
"grad_norm": 1.5097739696502686,
"learning_rate": 3.870085238607002e-06,
"loss": 0.5877,
"step": 230
},
{
"epoch": 0.5911708253358925,
"grad_norm": 1.6677836179733276,
"learning_rate": 3.829750856810123e-06,
"loss": 0.5796,
"step": 231
},
{
"epoch": 0.5937300063979527,
"grad_norm": 1.6501197814941406,
"learning_rate": 3.7894968826263993e-06,
"loss": 0.532,
"step": 232
},
{
"epoch": 0.5962891874600128,
"grad_norm": 1.4657257795333862,
"learning_rate": 3.7493260818993295e-06,
"loss": 0.5178,
"step": 233
},
{
"epoch": 0.5988483685220729,
"grad_norm": 1.576541781425476,
"learning_rate": 3.7092412147575763e-06,
"loss": 0.5337,
"step": 234
},
{
"epoch": 0.6014075495841331,
"grad_norm": 1.5359219312667847,
"learning_rate": 3.6692450354253244e-06,
"loss": 0.431,
"step": 235
},
{
"epoch": 0.6039667306461932,
"grad_norm": 1.9020817279815674,
"learning_rate": 3.6293402920330345e-06,
"loss": 0.5593,
"step": 236
},
{
"epoch": 0.6065259117082533,
"grad_norm": 1.6816877126693726,
"learning_rate": 3.589529726428615e-06,
"loss": 0.4986,
"step": 237
},
{
"epoch": 0.6090850927703135,
"grad_norm": 1.7360111474990845,
"learning_rate": 3.54981607398904e-06,
"loss": 0.5238,
"step": 238
},
{
"epoch": 0.6116442738323736,
"grad_norm": 1.542158603668213,
"learning_rate": 3.5102020634323974e-06,
"loss": 0.5575,
"step": 239
},
{
"epoch": 0.6142034548944337,
"grad_norm": 1.8407609462738037,
"learning_rate": 3.470690416630395e-06,
"loss": 0.5026,
"step": 240
},
{
"epoch": 0.6167626359564939,
"grad_norm": 1.5505833625793457,
"learning_rate": 3.431283848421347e-06,
"loss": 0.5446,
"step": 241
},
{
"epoch": 0.619321817018554,
"grad_norm": 1.7350224256515503,
"learning_rate": 3.39198506642364e-06,
"loss": 0.5244,
"step": 242
},
{
"epoch": 0.6218809980806143,
"grad_norm": 1.5843605995178223,
"learning_rate": 3.352796770849679e-06,
"loss": 0.4617,
"step": 243
},
{
"epoch": 0.6244401791426744,
"grad_norm": 1.5719588994979858,
"learning_rate": 3.3137216543203747e-06,
"loss": 0.5001,
"step": 244
},
{
"epoch": 0.6269993602047345,
"grad_norm": 1.7105059623718262,
"learning_rate": 3.2747624016801245e-06,
"loss": 0.5415,
"step": 245
},
{
"epoch": 0.6295585412667947,
"grad_norm": 1.7561525106430054,
"learning_rate": 3.2359216898123343e-06,
"loss": 0.5619,
"step": 246
},
{
"epoch": 0.6321177223288548,
"grad_norm": 2.0059330463409424,
"learning_rate": 3.197202187455498e-06,
"loss": 0.5481,
"step": 247
},
{
"epoch": 0.6346769033909149,
"grad_norm": 1.4864341020584106,
"learning_rate": 3.158606555019826e-06,
"loss": 0.5159,
"step": 248
},
{
"epoch": 0.6372360844529751,
"grad_norm": 1.5791542530059814,
"learning_rate": 3.120137444404442e-06,
"loss": 0.5506,
"step": 249
},
{
"epoch": 0.6397952655150352,
"grad_norm": 1.647869348526001,
"learning_rate": 3.081797498815185e-06,
"loss": 0.5004,
"step": 250
},
{
"epoch": 0.6423544465770953,
"grad_norm": 1.6751713752746582,
"learning_rate": 3.0435893525829847e-06,
"loss": 0.5103,
"step": 251
},
{
"epoch": 0.6449136276391555,
"grad_norm": 1.8151626586914062,
"learning_rate": 3.005515630982858e-06,
"loss": 0.4874,
"step": 252
},
{
"epoch": 0.6474728087012156,
"grad_norm": 1.5823078155517578,
"learning_rate": 2.9675789500535328e-06,
"loss": 0.4991,
"step": 253
},
{
"epoch": 0.6500319897632757,
"grad_norm": 1.7204740047454834,
"learning_rate": 2.9297819164176964e-06,
"loss": 0.5046,
"step": 254
},
{
"epoch": 0.6525911708253359,
"grad_norm": 1.5417708158493042,
"learning_rate": 2.8921271271028896e-06,
"loss": 0.4983,
"step": 255
},
{
"epoch": 0.655150351887396,
"grad_norm": 2.2327561378479004,
"learning_rate": 2.854617169363075e-06,
"loss": 0.5921,
"step": 256
},
{
"epoch": 0.6577095329494562,
"grad_norm": 1.7403721809387207,
"learning_rate": 2.817254620500868e-06,
"loss": 0.5774,
"step": 257
},
{
"epoch": 0.6602687140115163,
"grad_norm": 1.547120213508606,
"learning_rate": 2.7800420476904337e-06,
"loss": 0.5383,
"step": 258
},
{
"epoch": 0.6628278950735764,
"grad_norm": 1.4761300086975098,
"learning_rate": 2.7429820078011215e-06,
"loss": 0.5602,
"step": 259
},
{
"epoch": 0.6653870761356366,
"grad_norm": 1.5744880437850952,
"learning_rate": 2.7060770472217635e-06,
"loss": 0.5757,
"step": 260
},
{
"epoch": 0.6679462571976967,
"grad_norm": 2.3688693046569824,
"learning_rate": 2.669329701685719e-06,
"loss": 0.4874,
"step": 261
},
{
"epoch": 0.6705054382597568,
"grad_norm": 1.5973730087280273,
"learning_rate": 2.632742496096651e-06,
"loss": 0.6039,
"step": 262
},
{
"epoch": 0.673064619321817,
"grad_norm": 1.6801444292068481,
"learning_rate": 2.59631794435503e-06,
"loss": 0.4457,
"step": 263
},
{
"epoch": 0.6756238003838771,
"grad_norm": 1.6441094875335693,
"learning_rate": 2.560058549185412e-06,
"loss": 0.507,
"step": 264
},
{
"epoch": 0.6781829814459372,
"grad_norm": 1.648740291595459,
"learning_rate": 2.523966801964468e-06,
"loss": 0.5504,
"step": 265
},
{
"epoch": 0.6807421625079975,
"grad_norm": 1.4132170677185059,
"learning_rate": 2.488045182549819e-06,
"loss": 0.5247,
"step": 266
},
{
"epoch": 0.6833013435700576,
"grad_norm": 1.508416771888733,
"learning_rate": 2.4522961591096246e-06,
"loss": 0.4683,
"step": 267
},
{
"epoch": 0.6858605246321178,
"grad_norm": 1.7160900831222534,
"learning_rate": 2.4167221879530063e-06,
"loss": 0.4942,
"step": 268
},
{
"epoch": 0.6884197056941779,
"grad_norm": 1.9720776081085205,
"learning_rate": 2.381325713361283e-06,
"loss": 0.4986,
"step": 269
},
{
"epoch": 0.690978886756238,
"grad_norm": 1.452165961265564,
"learning_rate": 2.3461091674199998e-06,
"loss": 0.4422,
"step": 270
},
{
"epoch": 0.6935380678182982,
"grad_norm": 1.616641879081726,
"learning_rate": 2.311074969851852e-06,
"loss": 0.5194,
"step": 271
},
{
"epoch": 0.6960972488803583,
"grad_norm": 1.4725862741470337,
"learning_rate": 2.276225527850401e-06,
"loss": 0.5132,
"step": 272
},
{
"epoch": 0.6986564299424184,
"grad_norm": 1.5637017488479614,
"learning_rate": 2.2415632359146855e-06,
"loss": 0.5096,
"step": 273
},
{
"epoch": 0.7012156110044786,
"grad_norm": 2.0925307273864746,
"learning_rate": 2.2070904756847023e-06,
"loss": 0.5367,
"step": 274
},
{
"epoch": 0.7037747920665387,
"grad_norm": 1.4457118511199951,
"learning_rate": 2.1728096157777516e-06,
"loss": 0.4998,
"step": 275
},
{
"epoch": 0.7063339731285988,
"grad_norm": 1.425362229347229,
"learning_rate": 2.1387230116257004e-06,
"loss": 0.47,
"step": 276
},
{
"epoch": 0.708893154190659,
"grad_norm": 1.4573434591293335,
"learning_rate": 2.104833005313131e-06,
"loss": 0.4368,
"step": 277
},
{
"epoch": 0.7114523352527191,
"grad_norm": 1.633025050163269,
"learning_rate": 2.071141925416431e-06,
"loss": 0.4864,
"step": 278
},
{
"epoch": 0.7140115163147792,
"grad_norm": 1.6741235256195068,
"learning_rate": 2.0376520868437838e-06,
"loss": 0.5341,
"step": 279
},
{
"epoch": 0.7165706973768394,
"grad_norm": 1.5268086194992065,
"learning_rate": 2.0043657906761144e-06,
"loss": 0.4713,
"step": 280
},
{
"epoch": 0.7191298784388995,
"grad_norm": 1.6337690353393555,
"learning_rate": 1.971285324008994e-06,
"loss": 0.502,
"step": 281
},
{
"epoch": 0.7216890595009597,
"grad_norm": 1.707281470298767,
"learning_rate": 1.938412959795472e-06,
"loss": 0.4848,
"step": 282
},
{
"epoch": 0.7242482405630198,
"grad_norm": 1.4948759078979492,
"learning_rate": 1.9057509566899268e-06,
"loss": 0.4637,
"step": 283
},
{
"epoch": 0.7268074216250799,
"grad_norm": 1.5909830331802368,
"learning_rate": 1.873301558892855e-06,
"loss": 0.4133,
"step": 284
},
{
"epoch": 0.7293666026871402,
"grad_norm": 1.5825873613357544,
"learning_rate": 1.8410669959966804e-06,
"loss": 0.4676,
"step": 285
},
{
"epoch": 0.7319257837492003,
"grad_norm": 1.4792473316192627,
"learning_rate": 1.8090494828325634e-06,
"loss": 0.4517,
"step": 286
},
{
"epoch": 0.7344849648112604,
"grad_norm": 1.6525112390518188,
"learning_rate": 1.7772512193182096e-06,
"loss": 0.4793,
"step": 287
},
{
"epoch": 0.7370441458733206,
"grad_norm": 3.011749267578125,
"learning_rate": 1.745674390306722e-06,
"loss": 0.5269,
"step": 288
},
{
"epoch": 0.7396033269353807,
"grad_norm": 1.5214685201644897,
"learning_rate": 1.7143211654364761e-06,
"loss": 0.5309,
"step": 289
},
{
"epoch": 0.7421625079974408,
"grad_norm": 1.545071005821228,
"learning_rate": 1.6831936989820507e-06,
"loss": 0.4268,
"step": 290
},
{
"epoch": 0.744721689059501,
"grad_norm": 1.5162274837493896,
"learning_rate": 1.6522941297061996e-06,
"loss": 0.4596,
"step": 291
},
{
"epoch": 0.7472808701215611,
"grad_norm": 1.4541124105453491,
"learning_rate": 1.6216245807129005e-06,
"loss": 0.4311,
"step": 292
},
{
"epoch": 0.7498400511836213,
"grad_norm": 2.05718731880188,
"learning_rate": 1.5911871593014838e-06,
"loss": 0.4577,
"step": 293
},
{
"epoch": 0.7523992322456814,
"grad_norm": 2.3280277252197266,
"learning_rate": 1.5609839568218248e-06,
"loss": 0.4765,
"step": 294
},
{
"epoch": 0.7549584133077415,
"grad_norm": 1.4779192209243774,
"learning_rate": 1.5310170485306675e-06,
"loss": 0.4647,
"step": 295
},
{
"epoch": 0.7575175943698017,
"grad_norm": 1.3130896091461182,
"learning_rate": 1.5012884934490168e-06,
"loss": 0.4476,
"step": 296
},
{
"epoch": 0.7600767754318618,
"grad_norm": 1.6979542970657349,
"learning_rate": 1.4718003342206722e-06,
"loss": 0.4898,
"step": 297
},
{
"epoch": 0.7626359564939219,
"grad_norm": 1.580965518951416,
"learning_rate": 1.44255459697188e-06,
"loss": 0.4565,
"step": 298
},
{
"epoch": 0.7651951375559821,
"grad_norm": 1.6245533227920532,
"learning_rate": 1.413553291172106e-06,
"loss": 0.5063,
"step": 299
},
{
"epoch": 0.7677543186180422,
"grad_norm": 1.4497474431991577,
"learning_rate": 1.3847984094959843e-06,
"loss": 0.4968,
"step": 300
},
{
"epoch": 0.7703134996801023,
"grad_norm": 1.5835232734680176,
"learning_rate": 1.3562919276863846e-06,
"loss": 0.4664,
"step": 301
},
{
"epoch": 0.7728726807421625,
"grad_norm": 1.6141891479492188,
"learning_rate": 1.3280358044186648e-06,
"loss": 0.4485,
"step": 302
},
{
"epoch": 0.7754318618042226,
"grad_norm": 1.5265198945999146,
"learning_rate": 1.3000319811660967e-06,
"loss": 0.4546,
"step": 303
},
{
"epoch": 0.7779910428662828,
"grad_norm": 1.6432017087936401,
"learning_rate": 1.2722823820664577e-06,
"loss": 0.4247,
"step": 304
},
{
"epoch": 0.780550223928343,
"grad_norm": 1.6264252662658691,
"learning_rate": 1.2447889137898295e-06,
"loss": 0.5032,
"step": 305
},
{
"epoch": 0.783109404990403,
"grad_norm": 1.7414301633834839,
"learning_rate": 1.2175534654075887e-06,
"loss": 0.4517,
"step": 306
},
{
"epoch": 0.7856685860524633,
"grad_norm": 1.4454889297485352,
"learning_rate": 1.190577908262614e-06,
"loss": 0.4992,
"step": 307
},
{
"epoch": 0.7882277671145234,
"grad_norm": 1.5262882709503174,
"learning_rate": 1.1638640958407e-06,
"loss": 0.501,
"step": 308
},
{
"epoch": 0.7907869481765835,
"grad_norm": 1.644844651222229,
"learning_rate": 1.1374138636432054e-06,
"loss": 0.4567,
"step": 309
},
{
"epoch": 0.7933461292386437,
"grad_norm": 1.6340616941452026,
"learning_rate": 1.1112290290609444e-06,
"loss": 0.4374,
"step": 310
},
{
"epoch": 0.7959053103007038,
"grad_norm": 1.4388090372085571,
"learning_rate": 1.085311391249299e-06,
"loss": 0.431,
"step": 311
},
{
"epoch": 0.7984644913627639,
"grad_norm": 1.9001237154006958,
"learning_rate": 1.0596627310046165e-06,
"loss": 0.4767,
"step": 312
},
{
"epoch": 0.8010236724248241,
"grad_norm": 1.5323541164398193,
"learning_rate": 1.0342848106418368e-06,
"loss": 0.4444,
"step": 313
},
{
"epoch": 0.8035828534868842,
"grad_norm": 1.5047707557678223,
"learning_rate": 1.0091793738734113e-06,
"loss": 0.4322,
"step": 314
},
{
"epoch": 0.8061420345489443,
"grad_norm": 1.3882431983947754,
"learning_rate": 9.843481456894949e-07,
"loss": 0.3974,
"step": 315
},
{
"epoch": 0.8087012156110045,
"grad_norm": 1.5407518148422241,
"learning_rate": 9.59792832239415e-07,
"loss": 0.4345,
"step": 316
},
{
"epoch": 0.8112603966730646,
"grad_norm": 1.5471398830413818,
"learning_rate": 9.35515120714447e-07,
"loss": 0.4478,
"step": 317
},
{
"epoch": 0.8138195777351248,
"grad_norm": 1.770851731300354,
"learning_rate": 9.115166792318858e-07,
"loss": 0.3907,
"step": 318
},
{
"epoch": 0.8163787587971849,
"grad_norm": 1.5951015949249268,
"learning_rate": 8.877991567204353e-07,
"loss": 0.4732,
"step": 319
},
{
"epoch": 0.818937939859245,
"grad_norm": 1.4890408515930176,
"learning_rate": 8.643641828069004e-07,
"loss": 0.4236,
"step": 320
}
],
"logging_steps": 1.0,
"max_steps": 391,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 80,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 6.693790773915156e+17,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}