{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 928,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008620689655172414,
      "grad_norm": 0.26599958539009094,
      "learning_rate": 5.319148936170213e-07,
      "loss": 3.7070133686065674,
      "step": 2
    },
    {
      "epoch": 0.017241379310344827,
      "grad_norm": 0.8651937246322632,
      "learning_rate": 1.5957446808510639e-06,
      "loss": 2.565812349319458,
      "step": 4
    },
    {
      "epoch": 0.02586206896551724,
      "grad_norm": 1.8721028566360474,
      "learning_rate": 2.6595744680851065e-06,
      "loss": 2.305459976196289,
      "step": 6
    },
    {
      "epoch": 0.034482758620689655,
      "grad_norm": 0.3591924011707306,
      "learning_rate": 3.723404255319149e-06,
      "loss": 1.9556151628494263,
      "step": 8
    },
    {
      "epoch": 0.04310344827586207,
      "grad_norm": 0.9290058016777039,
      "learning_rate": 4.787234042553191e-06,
      "loss": 1.509597897529602,
      "step": 10
    },
    {
      "epoch": 0.05172413793103448,
      "grad_norm": 0.21377742290496826,
      "learning_rate": 5.851063829787235e-06,
      "loss": 1.7625975608825684,
      "step": 12
    },
    {
      "epoch": 0.0603448275862069,
      "grad_norm": 0.32525238394737244,
      "learning_rate": 6.914893617021277e-06,
      "loss": 1.8272696733474731,
      "step": 14
    },
    {
      "epoch": 0.06896551724137931,
      "grad_norm": 0.1417679488658905,
      "learning_rate": 7.97872340425532e-06,
      "loss": 0.7375114560127258,
      "step": 16
    },
    {
      "epoch": 0.07758620689655173,
      "grad_norm": 0.31613945960998535,
      "learning_rate": 9.042553191489362e-06,
      "loss": 0.8689496517181396,
      "step": 18
    },
    {
      "epoch": 0.08620689655172414,
      "grad_norm": 0.13069696724414825,
      "learning_rate": 1.0106382978723404e-05,
      "loss": 1.5243046283721924,
      "step": 20
    },
    {
      "epoch": 0.09482758620689655,
      "grad_norm": 0.12042803317308426,
      "learning_rate": 1.1170212765957447e-05,
      "loss": 1.672706127166748,
      "step": 22
    },
    {
      "epoch": 0.10344827586206896,
      "grad_norm": 0.24695225059986115,
      "learning_rate": 1.223404255319149e-05,
      "loss": 1.5508843660354614,
      "step": 24
    },
    {
      "epoch": 0.11206896551724138,
      "grad_norm": 0.2681316137313843,
      "learning_rate": 1.3297872340425532e-05,
      "loss": 1.3349249362945557,
      "step": 26
    },
    {
      "epoch": 0.1206896551724138,
      "grad_norm": 2.3104257583618164,
      "learning_rate": 1.4361702127659577e-05,
      "loss": 0.764009952545166,
      "step": 28
    },
    {
      "epoch": 0.12931034482758622,
      "grad_norm": 0.16677935421466827,
      "learning_rate": 1.5425531914893617e-05,
      "loss": 1.4469249248504639,
      "step": 30
    },
    {
      "epoch": 0.13793103448275862,
      "grad_norm": 0.2230503112077713,
      "learning_rate": 1.6489361702127658e-05,
      "loss": 1.205310583114624,
      "step": 32
    },
    {
      "epoch": 0.14655172413793102,
      "grad_norm": 0.1180746778845787,
      "learning_rate": 1.7553191489361703e-05,
      "loss": 1.3685661554336548,
      "step": 34
    },
    {
      "epoch": 0.15517241379310345,
      "grad_norm": 0.12433399260044098,
      "learning_rate": 1.8617021276595745e-05,
      "loss": 1.3990949392318726,
      "step": 36
    },
    {
      "epoch": 0.16379310344827586,
      "grad_norm": 0.20617301762104034,
      "learning_rate": 1.968085106382979e-05,
      "loss": 1.4243614673614502,
      "step": 38
    },
    {
      "epoch": 0.1724137931034483,
      "grad_norm": 0.12327367067337036,
      "learning_rate": 2.074468085106383e-05,
      "loss": 1.4310336112976074,
      "step": 40
    },
    {
      "epoch": 0.1810344827586207,
      "grad_norm": 0.23117953538894653,
      "learning_rate": 2.1808510638297873e-05,
      "loss": 1.6425907611846924,
      "step": 42
    },
    {
      "epoch": 0.1896551724137931,
      "grad_norm": 0.2595376968383789,
      "learning_rate": 2.2872340425531915e-05,
      "loss": 1.0640761852264404,
      "step": 44
    },
    {
      "epoch": 0.19827586206896552,
      "grad_norm": 0.21752837300300598,
      "learning_rate": 2.393617021276596e-05,
      "loss": 1.133437156677246,
      "step": 46
    },
    {
      "epoch": 0.20689655172413793,
      "grad_norm": 0.27208733558654785,
      "learning_rate": 2.5e-05,
      "loss": 1.3954966068267822,
      "step": 48
    },
    {
      "epoch": 0.21551724137931033,
      "grad_norm": 0.1533847600221634,
      "learning_rate": 2.4999713892919425e-05,
      "loss": 1.5042736530303955,
      "step": 50
    },
    {
      "epoch": 0.22413793103448276,
      "grad_norm": 0.8918558955192566,
      "learning_rate": 2.4998855586230096e-05,
      "loss": 0.8469613194465637,
      "step": 52
    },
    {
      "epoch": 0.23275862068965517,
      "grad_norm": 0.2088381052017212,
      "learning_rate": 2.4997425123588484e-05,
      "loss": 1.439496636390686,
      "step": 54
    },
    {
      "epoch": 0.2413793103448276,
      "grad_norm": 0.08935986459255219,
      "learning_rate": 2.4995422577752894e-05,
      "loss": 1.3860807418823242,
      "step": 56
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.27968838810920715,
      "learning_rate": 2.499284805057978e-05,
      "loss": 1.4035521745681763,
      "step": 58
    },
    {
      "epoch": 0.25862068965517243,
      "grad_norm": 0.1754721850156784,
      "learning_rate": 2.4989701673018555e-05,
      "loss": 1.399681568145752,
      "step": 60
    },
    {
      "epoch": 0.2672413793103448,
      "grad_norm": 0.4435209035873413,
      "learning_rate": 2.4985983605104933e-05,
      "loss": 1.099133849143982,
      "step": 62
    },
    {
      "epoch": 0.27586206896551724,
      "grad_norm": 0.09654241800308228,
      "learning_rate": 2.4981694035952797e-05,
      "loss": 1.333014726638794,
      "step": 64
    },
    {
      "epoch": 0.28448275862068967,
      "grad_norm": 0.15928423404693604,
      "learning_rate": 2.497683318374456e-05,
      "loss": 1.316357135772705,
      "step": 66
    },
    {
      "epoch": 0.29310344827586204,
      "grad_norm": 0.09688454866409302,
      "learning_rate": 2.497140129572009e-05,
      "loss": 1.3247264623641968,
      "step": 68
    },
    {
      "epoch": 0.3017241379310345,
      "grad_norm": 0.09453880041837692,
      "learning_rate": 2.4965398648164114e-05,
      "loss": 1.0423731803894043,
      "step": 70
    },
    {
      "epoch": 0.3103448275862069,
      "grad_norm": 0.2843893766403198,
      "learning_rate": 2.495882554639219e-05,
      "loss": 1.1186097860336304,
      "step": 72
    },
    {
      "epoch": 0.31896551724137934,
      "grad_norm": 0.21019835770130157,
      "learning_rate": 2.4951682324735147e-05,
      "loss": 1.0758298635482788,
      "step": 74
    },
    {
      "epoch": 0.3275862068965517,
      "grad_norm": 0.12580540776252747,
      "learning_rate": 2.4943969346522113e-05,
      "loss": 1.177038311958313,
      "step": 76
    },
    {
      "epoch": 0.33620689655172414,
      "grad_norm": 0.06072475388646126,
      "learning_rate": 2.4935687004062e-05,
      "loss": 1.1172969341278076,
      "step": 78
    },
    {
      "epoch": 0.3448275862068966,
      "grad_norm": 0.45244401693344116,
      "learning_rate": 2.492683571862358e-05,
      "loss": 0.925361156463623,
      "step": 80
    },
    {
      "epoch": 0.35344827586206895,
      "grad_norm": 0.12420380115509033,
      "learning_rate": 2.491741594041404e-05,
      "loss": 1.3231241703033447,
      "step": 82
    },
    {
      "epoch": 0.3620689655172414,
      "grad_norm": 0.27926552295684814,
      "learning_rate": 2.4907428148556096e-05,
      "loss": 1.1561291217803955,
      "step": 84
    },
    {
      "epoch": 0.3706896551724138,
      "grad_norm": 0.13415296375751495,
      "learning_rate": 2.48968728510636e-05,
      "loss": 1.3458794355392456,
      "step": 86
    },
    {
      "epoch": 0.3793103448275862,
      "grad_norm": 0.09626008570194244,
      "learning_rate": 2.4885750584815733e-05,
      "loss": 1.2599420547485352,
      "step": 88
    },
    {
      "epoch": 0.3879310344827586,
      "grad_norm": 0.4228016138076782,
      "learning_rate": 2.487406191552967e-05,
      "loss": 1.0761513710021973,
      "step": 90
    },
    {
      "epoch": 0.39655172413793105,
      "grad_norm": 0.09193041175603867,
      "learning_rate": 2.4861807437731816e-05,
      "loss": 0.8632556200027466,
      "step": 92
    },
    {
      "epoch": 0.4051724137931034,
      "grad_norm": 0.4261094629764557,
      "learning_rate": 2.4848987774727572e-05,
      "loss": 1.113997220993042,
      "step": 94
    },
    {
      "epoch": 0.41379310344827586,
      "grad_norm": 0.33825597167015076,
      "learning_rate": 2.4835603578569627e-05,
      "loss": 1.225900650024414,
      "step": 96
    },
    {
      "epoch": 0.4224137931034483,
      "grad_norm": 0.32547953724861145,
      "learning_rate": 2.4821655530024785e-05,
      "loss": 1.2307943105697632,
      "step": 98
    },
    {
      "epoch": 0.43103448275862066,
      "grad_norm": 0.20468811690807343,
      "learning_rate": 2.4807144338539345e-05,
      "loss": 1.150252103805542,
      "step": 100
    },
    {
      "epoch": 0.4396551724137931,
      "grad_norm": 0.23082198202610016,
      "learning_rate": 2.479207074220303e-05,
      "loss": 1.1921312808990479,
      "step": 102
    },
    {
      "epoch": 0.4482758620689655,
      "grad_norm": 0.12201318144798279,
      "learning_rate": 2.4776435507711404e-05,
      "loss": 1.2748147249221802,
      "step": 104
    },
    {
      "epoch": 0.45689655172413796,
      "grad_norm": 0.11089494824409485,
      "learning_rate": 2.476023943032693e-05,
      "loss": 0.961470365524292,
      "step": 106
    },
    {
      "epoch": 0.46551724137931033,
      "grad_norm": 0.21629932522773743,
      "learning_rate": 2.474348333383848e-05,
      "loss": 1.183750867843628,
      "step": 108
    },
    {
      "epoch": 0.47413793103448276,
      "grad_norm": 0.18608631193637848,
      "learning_rate": 2.4726168070519457e-05,
      "loss": 1.1943963766098022,
      "step": 110
    },
    {
      "epoch": 0.4827586206896552,
      "grad_norm": 0.10688410699367523,
      "learning_rate": 2.4708294521084412e-05,
      "loss": 1.4336156845092773,
      "step": 112
    },
    {
      "epoch": 0.49137931034482757,
      "grad_norm": 0.48181021213531494,
      "learning_rate": 2.46898635946443e-05,
      "loss": 0.9574199318885803,
      "step": 114
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.15253476798534393,
      "learning_rate": 2.4670876228660192e-05,
      "loss": 0.8921310901641846,
      "step": 116
    },
    {
      "epoch": 0.5086206896551724,
      "grad_norm": 0.12271574139595032,
      "learning_rate": 2.4651333388895617e-05,
      "loss": 1.2967804670333862,
      "step": 118
    },
    {
      "epoch": 0.5172413793103449,
      "grad_norm": 0.0918755903840065,
      "learning_rate": 2.4631236069367435e-05,
      "loss": 1.2260866165161133,
      "step": 120
    },
    {
      "epoch": 0.5258620689655172,
      "grad_norm": 0.09145987033843994,
      "learning_rate": 2.4610585292295278e-05,
      "loss": 1.3115153312683105,
      "step": 122
    },
    {
      "epoch": 0.5344827586206896,
      "grad_norm": 0.22018814086914062,
      "learning_rate": 2.458938210804955e-05,
      "loss": 1.458106279373169,
      "step": 124
    },
    {
      "epoch": 0.5431034482758621,
      "grad_norm": 0.10475312173366547,
      "learning_rate": 2.456762759509801e-05,
      "loss": 1.589352011680603,
      "step": 126
    },
    {
      "epoch": 0.5517241379310345,
      "grad_norm": 0.074061319231987,
      "learning_rate": 2.454532285995091e-05,
      "loss": 1.3105168342590332,
      "step": 128
    },
    {
      "epoch": 0.5603448275862069,
      "grad_norm": 0.07942219823598862,
      "learning_rate": 2.4522469037104732e-05,
      "loss": 1.2909717559814453,
      "step": 130
    },
    {
      "epoch": 0.5689655172413793,
      "grad_norm": 0.1517823338508606,
      "learning_rate": 2.449906728898445e-05,
      "loss": 1.2390798330307007,
      "step": 132
    },
    {
      "epoch": 0.5775862068965517,
      "grad_norm": 0.3011615574359894,
      "learning_rate": 2.447511880588444e-05,
      "loss": 1.150702714920044,
      "step": 134
    },
    {
      "epoch": 0.5862068965517241,
      "grad_norm": 0.12555329501628876,
      "learning_rate": 2.445062480590792e-05,
      "loss": 0.978553056716919,
      "step": 136
    },
    {
      "epoch": 0.5948275862068966,
      "grad_norm": 0.049786102026700974,
      "learning_rate": 2.4425586534904986e-05,
      "loss": 0.9309189915657043,
      "step": 138
    },
    {
      "epoch": 0.603448275862069,
      "grad_norm": 0.13792221248149872,
      "learning_rate": 2.440000526640926e-05,
      "loss": 0.9206077456474304,
      "step": 140
    },
    {
      "epoch": 0.6120689655172413,
      "grad_norm": 0.08182723075151443,
      "learning_rate": 2.437388230157311e-05,
      "loss": 1.0149747133255005,
      "step": 142
    },
    {
      "epoch": 0.6206896551724138,
      "grad_norm": 0.09423603862524033,
      "learning_rate": 2.4347218969101456e-05,
      "loss": 1.004448413848877,
      "step": 144
    },
    {
      "epoch": 0.6293103448275862,
      "grad_norm": 0.0868963673710823,
      "learning_rate": 2.4320016625184208e-05,
      "loss": 1.251673698425293,
      "step": 146
    },
    {
      "epoch": 0.6379310344827587,
      "grad_norm": 0.24749471247196198,
      "learning_rate": 2.4292276653427266e-05,
      "loss": 1.2662962675094604,
      "step": 148
    },
    {
      "epoch": 0.646551724137931,
      "grad_norm": 0.1099071204662323,
      "learning_rate": 2.4264000464782154e-05,
      "loss": 1.05961275100708,
      "step": 150
    },
    {
      "epoch": 0.6551724137931034,
      "grad_norm": 0.13175362348556519,
      "learning_rate": 2.423518949747426e-05,
      "loss": 0.8592603206634521,
      "step": 152
    },
    {
      "epoch": 0.6637931034482759,
      "grad_norm": 0.25698038935661316,
      "learning_rate": 2.420584521692967e-05,
      "loss": 1.0430630445480347,
      "step": 154
    },
    {
      "epoch": 0.6724137931034483,
      "grad_norm": 1.2381486892700195,
      "learning_rate": 2.4175969115700637e-05,
      "loss": 1.1310160160064697,
      "step": 156
    },
    {
      "epoch": 0.6810344827586207,
      "grad_norm": 0.27287977933883667,
      "learning_rate": 2.4145562713389665e-05,
      "loss": 1.237598180770874,
      "step": 158
    },
    {
      "epoch": 0.6896551724137931,
      "grad_norm": 0.18096014857292175,
      "learning_rate": 2.4114627556572236e-05,
      "loss": 1.2904868125915527,
      "step": 160
    },
    {
      "epoch": 0.6982758620689655,
      "grad_norm": 0.12753747403621674,
      "learning_rate": 2.4083165218718094e-05,
      "loss": 1.5490325689315796,
      "step": 162
    },
    {
      "epoch": 0.7068965517241379,
      "grad_norm": 0.1123153418302536,
      "learning_rate": 2.405117730011127e-05,
      "loss": 0.9844979643821716,
      "step": 164
    },
    {
      "epoch": 0.7155172413793104,
      "grad_norm": 0.14886538684368134,
      "learning_rate": 2.4018665427768652e-05,
      "loss": 1.264622449874878,
      "step": 166
    },
    {
      "epoch": 0.7241379310344828,
      "grad_norm": 0.09712914377450943,
      "learning_rate": 2.3985631255357243e-05,
      "loss": 1.3305325508117676,
      "step": 168
    },
    {
      "epoch": 0.7327586206896551,
      "grad_norm": 0.11504893749952316,
      "learning_rate": 2.3952076463110038e-05,
      "loss": 1.2382071018218994,
      "step": 170
    },
    {
      "epoch": 0.7413793103448276,
      "grad_norm": 0.23046067357063293,
      "learning_rate": 2.3918002757740583e-05,
      "loss": 1.249595046043396,
      "step": 172
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.1944531351327896,
      "learning_rate": 2.3883411872356135e-05,
      "loss": 0.9225230813026428,
      "step": 174
    },
    {
      "epoch": 0.7586206896551724,
      "grad_norm": 0.13637058436870575,
      "learning_rate": 2.3848305566369543e-05,
      "loss": 0.6734743714332581,
      "step": 176
    },
    {
      "epoch": 0.7672413793103449,
      "grad_norm": 0.1864972859621048,
      "learning_rate": 2.3812685625409726e-05,
      "loss": 1.24442458152771,
      "step": 178
    },
    {
      "epoch": 0.7758620689655172,
      "grad_norm": 0.0796503946185112,
      "learning_rate": 2.377655386123088e-05,
      "loss": 1.170575499534607,
      "step": 180
    },
    {
      "epoch": 0.7844827586206896,
      "grad_norm": 0.1396905928850174,
      "learning_rate": 2.3739912111620303e-05,
      "loss": 0.8946050405502319,
      "step": 182
    },
    {
      "epoch": 0.7931034482758621,
      "grad_norm": 0.07272544503211975,
      "learning_rate": 2.3702762240304933e-05,
      "loss": 0.7359169125556946,
      "step": 184
    },
    {
      "epoch": 0.8017241379310345,
      "grad_norm": 0.21276403963565826,
      "learning_rate": 2.366510613685654e-05,
      "loss": 1.377792477607727,
      "step": 186
    },
    {
      "epoch": 0.8103448275862069,
      "grad_norm": 0.6310688853263855,
      "learning_rate": 2.362694571659563e-05,
      "loss": 1.5076112747192383,
      "step": 188
    },
    {
      "epoch": 0.8189655172413793,
      "grad_norm": 0.09929846227169037,
      "learning_rate": 2.3588282920494007e-05,
      "loss": 1.2605912685394287,
      "step": 190
    },
    {
      "epoch": 0.8275862068965517,
      "grad_norm": 0.12678399682044983,
      "learning_rate": 2.354911971507607e-05,
      "loss": 1.2289090156555176,
      "step": 192
    },
    {
      "epoch": 0.8362068965517241,
      "grad_norm": 0.09680190682411194,
      "learning_rate": 2.3509458092318772e-05,
      "loss": 1.2607473134994507,
      "step": 194
    },
    {
      "epoch": 0.8448275862068966,
      "grad_norm": 0.11455738544464111,
      "learning_rate": 2.3469300069550308e-05,
      "loss": 1.2929327487945557,
      "step": 196
    },
    {
      "epoch": 0.853448275862069,
      "grad_norm": 0.22343184053897858,
      "learning_rate": 2.342864768934751e-05,
      "loss": 0.8722022175788879,
      "step": 198
    },
    {
      "epoch": 0.8620689655172413,
      "grad_norm": 0.1278611570596695,
      "learning_rate": 2.338750301943194e-05,
      "loss": 1.184353232383728,
      "step": 200
    },
    {
      "epoch": 0.8706896551724138,
      "grad_norm": 0.1743975579738617,
      "learning_rate": 2.3345868152564734e-05,
      "loss": 1.3033404350280762,
      "step": 202
    },
    {
      "epoch": 0.8793103448275862,
      "grad_norm": 0.09879707545042038,
      "learning_rate": 2.3303745206440156e-05,
      "loss": 0.8301224708557129,
      "step": 204
    },
    {
      "epoch": 0.8879310344827587,
      "grad_norm": 0.07442842423915863,
      "learning_rate": 2.3261136323577873e-05,
      "loss": 1.2247326374053955,
      "step": 206
    },
    {
      "epoch": 0.896551724137931,
      "grad_norm": 0.0767381563782692,
      "learning_rate": 2.3218043671213988e-05,
      "loss": 1.275619626045227,
      "step": 208
    },
    {
      "epoch": 0.9051724137931034,
      "grad_norm": 0.21888962388038635,
      "learning_rate": 2.317446944119081e-05,
      "loss": 0.763530969619751,
      "step": 210
    },
    {
      "epoch": 0.9137931034482759,
      "grad_norm": 0.2872910797595978,
      "learning_rate": 2.313041584984535e-05,
      "loss": 0.8847395181655884,
      "step": 212
    },
    {
      "epoch": 0.9224137931034483,
      "grad_norm": 0.22158221900463104,
      "learning_rate": 2.3085885137896623e-05,
      "loss": 1.139022707939148,
      "step": 214
    },
    {
      "epoch": 0.9310344827586207,
      "grad_norm": 0.11198285222053528,
      "learning_rate": 2.3040879570331648e-05,
      "loss": 1.2234928607940674,
      "step": 216
    },
    {
      "epoch": 0.9396551724137931,
      "grad_norm": 0.09761999547481537,
      "learning_rate": 2.2995401436290252e-05,
      "loss": 0.7140777111053467,
      "step": 218
    },
    {
      "epoch": 0.9482758620689655,
      "grad_norm": 0.22312681376934052,
      "learning_rate": 2.294945304894864e-05,
      "loss": 0.8441833257675171,
      "step": 220
    },
    {
      "epoch": 0.9568965517241379,
      "grad_norm": 0.2743966579437256,
      "learning_rate": 2.2903036745401742e-05,
      "loss": 1.3300198316574097,
      "step": 222
    },
    {
      "epoch": 0.9655172413793104,
      "grad_norm": 0.6734777688980103,
      "learning_rate": 2.2856154886544338e-05,
      "loss": 0.8017496466636658,
      "step": 224
    },
    {
      "epoch": 0.9741379310344828,
      "grad_norm": 0.1572013795375824,
      "learning_rate": 2.280880985695095e-05,
      "loss": 0.8339143991470337,
      "step": 226
    },
    {
      "epoch": 0.9827586206896551,
      "grad_norm": 0.12929460406303406,
      "learning_rate": 2.27610040647546e-05,
      "loss": 1.2216969728469849,
      "step": 228
    },
    {
      "epoch": 0.9913793103448276,
      "grad_norm": 0.24367918074131012,
      "learning_rate": 2.2712739941524287e-05,
      "loss": 0.8863965272903442,
      "step": 230
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.11734236776828766,
      "learning_rate": 2.266401994214133e-05,
      "loss": 0.7733997702598572,
      "step": 232
    },
    {
      "epoch": 1.0086206896551724,
      "grad_norm": 0.3386397063732147,
      "learning_rate": 2.2614846544674486e-05,
      "loss": 0.8990653157234192,
      "step": 234
    },
    {
      "epoch": 1.0172413793103448,
      "grad_norm": 0.20136962831020355,
      "learning_rate": 2.2565222250253925e-05,
      "loss": 0.7269018888473511,
      "step": 236
    },
    {
      "epoch": 1.0258620689655173,
      "grad_norm": 0.27934548258781433,
      "learning_rate": 2.251514958294401e-05,
      "loss": 0.8784797787666321,
      "step": 238
    },
    {
      "epoch": 1.0344827586206897,
      "grad_norm": 0.45081236958503723,
      "learning_rate": 2.246463108961489e-05,
      "loss": 0.5005398988723755,
      "step": 240
    },
    {
      "epoch": 1.043103448275862,
      "grad_norm": 0.10048855096101761,
      "learning_rate": 2.2413669339812997e-05,
      "loss": 1.1350990533828735,
      "step": 242
    },
    {
      "epoch": 1.0517241379310345,
      "grad_norm": 0.21551190316677094,
      "learning_rate": 2.2362266925630318e-05,
      "loss": 0.36204367876052856,
      "step": 244
    },
    {
      "epoch": 1.0603448275862069,
      "grad_norm": 0.19904279708862305,
      "learning_rate": 2.231042646157258e-05,
      "loss": 1.1106600761413574,
      "step": 246
    },
    {
      "epoch": 1.0689655172413792,
      "grad_norm": 0.15923143923282623,
      "learning_rate": 2.2258150584426235e-05,
      "loss": 1.0469825267791748,
      "step": 248
    },
    {
      "epoch": 1.0775862068965518,
      "grad_norm": 0.1504082828760147,
      "learning_rate": 2.220544195312436e-05,
      "loss": 1.0015485286712646,
      "step": 250
    },
    {
      "epoch": 1.0862068965517242,
      "grad_norm": 0.13122083246707916,
      "learning_rate": 2.2152303248611437e-05,
      "loss": 0.36394670605659485,
      "step": 252
    },
    {
      "epoch": 1.0948275862068966,
      "grad_norm": 0.3954187333583832,
      "learning_rate": 2.2098737173706942e-05,
      "loss": 1.0145013332366943,
      "step": 254
    },
    {
      "epoch": 1.103448275862069,
      "grad_norm": 0.24796895682811737,
      "learning_rate": 2.2044746452967906e-05,
      "loss": 0.7123383283615112,
      "step": 256
    },
    {
      "epoch": 1.1120689655172413,
      "grad_norm": 0.653165340423584,
      "learning_rate": 2.1990333832550333e-05,
      "loss": 0.4990229606628418,
      "step": 258
    },
    {
      "epoch": 1.1206896551724137,
      "grad_norm": 0.08383104205131531,
      "learning_rate": 2.1935502080069505e-05,
      "loss": 0.7109271287918091,
      "step": 260
    },
    {
      "epoch": 1.1293103448275863,
      "grad_norm": 0.1300533413887024,
      "learning_rate": 2.1880253984459224e-05,
      "loss": 1.2138782739639282,
      "step": 262
    },
    {
      "epoch": 1.1379310344827587,
      "grad_norm": 0.1580507606267929,
      "learning_rate": 2.1824592355829945e-05,
      "loss": 0.8597405552864075,
      "step": 264
    },
    {
      "epoch": 1.146551724137931,
      "grad_norm": 0.3914850354194641,
      "learning_rate": 2.1768520025325866e-05,
      "loss": 1.1350115537643433,
      "step": 266
    },
    {
      "epoch": 1.1551724137931034,
      "grad_norm": 0.10050661861896515,
      "learning_rate": 2.1712039844980893e-05,
      "loss": 1.1547629833221436,
      "step": 268
    },
    {
      "epoch": 1.1637931034482758,
      "grad_norm": 0.08129733055830002,
      "learning_rate": 2.165515468757362e-05,
      "loss": 1.1632627248764038,
      "step": 270
    },
    {
      "epoch": 1.1724137931034484,
      "grad_norm": 0.10255050659179688,
      "learning_rate": 2.1597867446481155e-05,
      "loss": 1.0813562870025635,
      "step": 272
    },
    {
      "epoch": 1.1810344827586208,
      "grad_norm": 0.30835890769958496,
      "learning_rate": 2.1540181035532004e-05,
      "loss": 0.7284541130065918,
      "step": 274
    },
    {
      "epoch": 1.1896551724137931,
      "grad_norm": 0.08633577823638916,
      "learning_rate": 2.1482098388857834e-05,
      "loss": 0.3468450903892517,
      "step": 276
    },
    {
      "epoch": 1.1982758620689655,
      "grad_norm": 0.09107495844364166,
      "learning_rate": 2.1423622460744242e-05,
      "loss": 1.121955394744873,
      "step": 278
    },
    {
      "epoch": 1.206896551724138,
      "grad_norm": 0.0790572538971901,
      "learning_rate": 2.1364756225480478e-05,
      "loss": 1.0472161769866943,
      "step": 280
    },
    {
      "epoch": 1.2155172413793103,
      "grad_norm": 0.18506139516830444,
      "learning_rate": 2.1305502677208184e-05,
      "loss": 0.5945373773574829,
      "step": 282
    },
    {
      "epoch": 1.2241379310344827,
      "grad_norm": 0.11592728644609451,
      "learning_rate": 2.1245864829769092e-05,
      "loss": 0.7433868646621704,
      "step": 284
    },
    {
      "epoch": 1.2327586206896552,
      "grad_norm": 0.2357645183801651,
      "learning_rate": 2.1185845716551712e-05,
      "loss": 0.8308318853378296,
      "step": 286
    },
    {
      "epoch": 1.2413793103448276,
      "grad_norm": 0.2785407304763794,
      "learning_rate": 2.1125448390337072e-05,
      "loss": 0.3982611894607544,
      "step": 288
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.17765875160694122,
      "learning_rate": 2.106467592314343e-05,
      "loss": 0.9474261403083801,
      "step": 290
    },
    {
      "epoch": 1.2586206896551724,
      "grad_norm": 0.10548825562000275,
      "learning_rate": 2.1003531406070012e-05,
      "loss": 0.927026093006134,
      "step": 292
    },
    {
      "epoch": 1.2672413793103448,
      "grad_norm": 0.32261741161346436,
      "learning_rate": 2.09420179491398e-05,
      "loss": 1.0189390182495117,
      "step": 294
    },
    {
      "epoch": 1.2758620689655173,
      "grad_norm": 0.08806228637695312,
      "learning_rate": 2.0880138681141345e-05,
      "loss": 0.7990611791610718,
      "step": 296
    },
    {
      "epoch": 1.2844827586206897,
      "grad_norm": 0.07184045016765594,
      "learning_rate": 2.081789674946961e-05,
      "loss": 1.1012811660766602,
      "step": 298
    },
    {
      "epoch": 1.293103448275862,
      "grad_norm": 0.09509000182151794,
      "learning_rate": 2.075529531996591e-05,
      "loss": 1.0139261484146118,
      "step": 300
    },
    {
      "epoch": 1.3017241379310345,
      "grad_norm": 0.10540059953927994,
      "learning_rate": 2.0692337576756866e-05,
      "loss": 0.6810011863708496,
      "step": 302
    },
    {
      "epoch": 1.3103448275862069,
      "grad_norm": 0.0546269416809082,
      "learning_rate": 2.0629026722092453e-05,
      "loss": 0.41086244583129883,
      "step": 304
    },
    {
      "epoch": 1.3189655172413794,
      "grad_norm": 0.08956003934144974,
      "learning_rate": 2.056536597618312e-05,
      "loss": 1.3326225280761719,
      "step": 306
    },
    {
      "epoch": 1.3275862068965516,
      "grad_norm": 0.06451157480478287,
      "learning_rate": 2.0501358577036003e-05,
      "loss": 0.6605536341667175,
      "step": 308
    },
    {
      "epoch": 1.3362068965517242,
      "grad_norm": 0.3918032944202423,
      "learning_rate": 2.043700778029024e-05,
      "loss": 1.0086909532546997,
      "step": 310
    },
    {
      "epoch": 1.3448275862068966,
      "grad_norm": 0.09106232970952988,
      "learning_rate": 2.0372316859051357e-05,
      "loss": 0.9816857576370239,
      "step": 312
    },
    {
      "epoch": 1.353448275862069,
      "grad_norm": 0.3334841728210449,
      "learning_rate": 2.0307289103724793e-05,
      "loss": 1.206939458847046,
      "step": 314
    },
    {
      "epoch": 1.3620689655172413,
      "grad_norm": 0.1430651843547821,
      "learning_rate": 2.0241927821848544e-05,
      "loss": 1.3034409284591675,
      "step": 316
    },
    {
      "epoch": 1.3706896551724137,
      "grad_norm": 0.12443620711565018,
      "learning_rate": 2.0176236337924933e-05,
      "loss": 1.0241037607192993,
      "step": 318
    },
    {
      "epoch": 1.3793103448275863,
      "grad_norm": 0.15698978304862976,
      "learning_rate": 2.01102179932515e-05,
      "loss": 0.6907064318656921,
      "step": 320
    },
    {
      "epoch": 1.3879310344827587,
      "grad_norm": 0.10246633738279343,
      "learning_rate": 2.0043876145751067e-05,
      "loss": 1.4752711057662964,
      "step": 322
    },
    {
      "epoch": 1.396551724137931,
      "grad_norm": 0.11384667456150055,
      "learning_rate": 1.997721416980093e-05,
      "loss": 0.4461423456668854,
      "step": 324
    },
    {
      "epoch": 1.4051724137931034,
      "grad_norm": 0.9319630265235901,
      "learning_rate": 1.991023545606124e-05,
      "loss": 0.42419686913490295,
      "step": 326
    },
    {
      "epoch": 1.4137931034482758,
      "grad_norm": 0.1355047971010208,
      "learning_rate": 1.9842943411302527e-05,
      "loss": 0.9030317664146423,
      "step": 328
    },
    {
      "epoch": 1.4224137931034484,
      "grad_norm": 0.12845957279205322,
      "learning_rate": 1.9775341458232436e-05,
      "loss": 0.768023669719696,
      "step": 330
    },
    {
      "epoch": 1.4310344827586206,
      "grad_norm": 0.058132827281951904,
      "learning_rate": 1.9707433035321613e-05,
      "loss": 0.6390933394432068,
      "step": 332
    },
    {
      "epoch": 1.4396551724137931,
      "grad_norm": 0.14224818348884583,
      "learning_rate": 1.9639221596628836e-05,
      "loss": 0.6768423914909363,
      "step": 334
    },
    {
      "epoch": 1.4482758620689655,
      "grad_norm": 0.30597057938575745,
      "learning_rate": 1.957071061162533e-05,
      "loss": 0.710463285446167,
      "step": 336
    },
    {
      "epoch": 1.456896551724138,
      "grad_norm": 0.08679758757352829,
      "learning_rate": 1.9501903565018273e-05,
      "loss": 1.0271871089935303,
      "step": 338
    },
    {
      "epoch": 1.4655172413793103,
      "grad_norm": 0.17182646691799164,
      "learning_rate": 1.943280395657357e-05,
      "loss": 1.2887871265411377,
      "step": 340
    },
    {
      "epoch": 1.4741379310344827,
      "grad_norm": 0.09934200346469879,
      "learning_rate": 1.9363415300937846e-05,
      "loss": 1.1640923023223877,
      "step": 342
    },
    {
      "epoch": 1.4827586206896552,
      "grad_norm": 0.114693783223629,
      "learning_rate": 1.9293741127459665e-05,
      "loss": 0.9469618201255798,
      "step": 344
    },
    {
      "epoch": 1.4913793103448276,
      "grad_norm": 0.1328352689743042,
      "learning_rate": 1.922378498001003e-05,
      "loss": 0.45884159207344055,
      "step": 346
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.3843913972377777,
      "learning_rate": 1.9153550416802107e-05,
      "loss": 0.8928775787353516,
      "step": 348
    },
    {
      "epoch": 1.5086206896551724,
      "grad_norm": 0.2543632686138153,
      "learning_rate": 1.9083041010210273e-05,
      "loss": 1.0797038078308105,
      "step": 350
    },
    {
      "epoch": 1.5172413793103448,
      "grad_norm": 0.1487339287996292,
      "learning_rate": 1.9012260346588394e-05,
      "loss": 1.0333884954452515,
      "step": 352
    },
    {
      "epoch": 1.5258620689655173,
      "grad_norm": 0.14135690033435822,
      "learning_rate": 1.89412120260874e-05,
      "loss": 1.1652461290359497,
      "step": 354
    },
    {
      "epoch": 1.5344827586206895,
      "grad_norm": 0.3962077796459198,
      "learning_rate": 1.8869899662472197e-05,
      "loss": 0.8967676758766174,
      "step": 356
    },
    {
      "epoch": 1.543103448275862,
      "grad_norm": 0.05238835886120796,
      "learning_rate": 1.8798326882937834e-05,
      "loss": 0.7729952335357666,
      "step": 358
    },
    {
      "epoch": 1.5517241379310345,
      "grad_norm": 0.09740617871284485,
      "learning_rate": 1.872649732792502e-05,
      "loss": 1.3566042184829712,
      "step": 360
    },
    {
      "epoch": 1.5603448275862069,
      "grad_norm": 0.06796287000179291,
      "learning_rate": 1.8654414650934968e-05,
      "loss": 1.0945885181427002,
      "step": 362
    },
    {
      "epoch": 1.5689655172413794,
      "grad_norm": 0.08173417299985886,
      "learning_rate": 1.858208251834354e-05,
      "loss": 1.1358948945999146,
      "step": 364
    },
    {
      "epoch": 1.5775862068965516,
      "grad_norm": 0.13112913072109222,
      "learning_rate": 1.8509504609214794e-05,
      "loss": 0.7162285447120667,
      "step": 366
    },
    {
      "epoch": 1.5862068965517242,
      "grad_norm": 0.08164367079734802,
      "learning_rate": 1.8436684615113824e-05,
      "loss": 1.080881953239441,
      "step": 368
    },
    {
      "epoch": 1.5948275862068966,
      "grad_norm": 0.07548147439956665,
      "learning_rate": 1.8363626239919022e-05,
      "loss": 1.054266333580017,
      "step": 370
    },
    {
      "epoch": 1.603448275862069,
      "grad_norm": 0.3800515830516815,
      "learning_rate": 1.8290333199633665e-05,
      "loss": 1.0251119136810303,
      "step": 372
    },
    {
      "epoch": 1.6120689655172413,
      "grad_norm": 0.058936282992362976,
      "learning_rate": 1.821680922219691e-05,
      "loss": 0.7149249315261841,
      "step": 374
    },
    {
      "epoch": 1.6206896551724137,
      "grad_norm": 0.08577758073806763,
      "learning_rate": 1.8143058047294197e-05,
      "loss": 1.0818464756011963,
      "step": 376
    },
    {
      "epoch": 1.6293103448275863,
      "grad_norm": 15.305747985839844,
      "learning_rate": 1.8069083426166995e-05,
      "loss": 1.1972448825836182,
      "step": 378
    },
    {
      "epoch": 1.6379310344827587,
      "grad_norm": 0.08515272289514542,
      "learning_rate": 1.799488912142205e-05,
      "loss": 0.7313613891601562,
      "step": 380
    },
    {
      "epoch": 1.646551724137931,
      "grad_norm": 0.14937828481197357,
      "learning_rate": 1.7920478906839967e-05,
      "loss": 1.3920011520385742,
      "step": 382
    },
    {
      "epoch": 1.6551724137931034,
      "grad_norm": 0.11451909691095352,
      "learning_rate": 1.7845856567183278e-05,
      "loss": 1.0703322887420654,
      "step": 384
    },
    {
      "epoch": 1.6637931034482758,
      "grad_norm": 0.11318498849868774,
      "learning_rate": 1.777102589800394e-05,
      "loss": 0.9152452349662781,
      "step": 386
    },
    {
      "epoch": 1.6724137931034484,
      "grad_norm": 0.11082017421722412,
      "learning_rate": 1.7695990705450276e-05,
      "loss": 1.360831618309021,
      "step": 388
    },
    {
      "epoch": 1.6810344827586206,
      "grad_norm": 0.6852989792823792,
      "learning_rate": 1.7620754806073368e-05,
      "loss": 0.905778706073761,
      "step": 390
    },
    {
      "epoch": 1.6896551724137931,
      "grad_norm": 0.14728988707065582,
      "learning_rate": 1.754532202663297e-05,
      "loss": 0.43694019317626953,
      "step": 392
    },
    {
      "epoch": 1.6982758620689655,
      "grad_norm": 0.48538804054260254,
      "learning_rate": 1.74696962039028e-05,
      "loss": 1.0095391273498535,
      "step": 394
    },
    {
      "epoch": 1.706896551724138,
      "grad_norm": 0.3051482141017914,
      "learning_rate": 1.7393881184475476e-05,
      "loss": 1.0637139081954956,
      "step": 396
    },
    {
      "epoch": 1.7155172413793105,
      "grad_norm": 0.1540980041027069,
      "learning_rate": 1.7317880824566792e-05,
      "loss": 1.042551040649414,
      "step": 398
    },
    {
      "epoch": 1.7241379310344827,
      "grad_norm": 0.16515707969665527,
      "learning_rate": 1.7241698989819604e-05,
      "loss": 1.02712082862854,
      "step": 400
    },
    {
      "epoch": 1.7327586206896552,
      "grad_norm": 0.13348740339279175,
      "learning_rate": 1.7165339555107223e-05,
      "loss": 0.7308106422424316,
      "step": 402
    },
    {
      "epoch": 1.7413793103448276,
      "grad_norm": 0.13249894976615906,
      "learning_rate": 1.7088806404336298e-05,
      "loss": 0.6191306710243225,
      "step": 404
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.26619619131088257,
      "learning_rate": 1.7012103430249292e-05,
      "loss": 0.6533270478248596,
      "step": 406
    },
    {
      "epoch": 1.7586206896551724,
      "grad_norm": 0.07099182903766632,
      "learning_rate": 1.6935234534226475e-05,
      "loss": 0.7554537653923035,
      "step": 408
    },
    {
      "epoch": 1.7672413793103448,
      "grad_norm": 0.12172161787748337,
      "learning_rate": 1.6858203626087468e-05,
      "loss": 0.8455532789230347,
      "step": 410
    },
    {
      "epoch": 1.7758620689655173,
      "grad_norm": 0.11975358426570892,
      "learning_rate": 1.6781014623892406e-05,
      "loss": 0.7798787951469421,
      "step": 412
    },
    {
      "epoch": 1.7844827586206895,
      "grad_norm": 0.11753620207309723,
      "learning_rate": 1.6703671453742642e-05,
      "loss": 0.6053640842437744,
      "step": 414
    },
    {
      "epoch": 1.793103448275862,
      "grad_norm": 0.14249694347381592,
      "learning_rate": 1.6626178049581033e-05,
      "loss": 0.8004411458969116,
      "step": 416
    },
    {
      "epoch": 1.8017241379310345,
      "grad_norm": 0.12212086468935013,
      "learning_rate": 1.654853835299187e-05,
      "loss": 0.9836465716362,
      "step": 418
    },
    {
      "epoch": 1.8103448275862069,
      "grad_norm": 0.4444316625595093,
      "learning_rate": 1.6470756313000386e-05,
      "loss": 1.111708402633667,
      "step": 420
    },
    {
      "epoch": 1.8189655172413794,
      "grad_norm": 0.12780721485614777,
      "learning_rate": 1.6392835885871897e-05,
      "loss": 1.1723716259002686,
      "step": 422
    },
    {
      "epoch": 1.8275862068965516,
      "grad_norm": 0.11766626685857773,
      "learning_rate": 1.631478103491056e-05,
      "loss": 1.006577491760254,
      "step": 424
    },
    {
      "epoch": 1.8362068965517242,
      "grad_norm": 0.10621563345193863,
      "learning_rate": 1.623659573025781e-05,
      "loss": 0.6617158055305481,
      "step": 426
    },
    {
      "epoch": 1.8448275862068966,
      "grad_norm": 0.19948941469192505,
      "learning_rate": 1.61582839486904e-05,
      "loss": 1.0630967617034912,
      "step": 428
    },
    {
      "epoch": 1.853448275862069,
      "grad_norm": 0.08539006859064102,
      "learning_rate": 1.6079849673418144e-05,
      "loss": 1.0498086214065552,
      "step": 430
    },
    {
      "epoch": 1.8620689655172413,
      "grad_norm": 0.15522508323192596,
      "learning_rate": 1.600129689388131e-05,
      "loss": 1.0054386854171753,
      "step": 432
    },
    {
      "epoch": 1.8706896551724137,
      "grad_norm": 0.23854467272758484,
      "learning_rate": 1.5922629605547708e-05,
      "loss": 1.107947826385498,
      "step": 434
    },
    {
      "epoch": 1.8793103448275863,
      "grad_norm": 0.14075548946857452,
      "learning_rate": 1.5843851809709462e-05,
      "loss": 0.6936919093132019,
      "step": 436
    },
    {
      "epoch": 1.8879310344827587,
      "grad_norm": 0.20101788640022278,
      "learning_rate": 1.5764967513279498e-05,
      "loss": 0.9637195467948914,
      "step": 438
    },
    {
      "epoch": 1.896551724137931,
      "grad_norm": 0.09223437309265137,
      "learning_rate": 1.568598072858772e-05,
      "loss": 0.9367073774337769,
      "step": 440
    },
    {
      "epoch": 1.9051724137931034,
      "grad_norm": 0.1555858999490738,
      "learning_rate": 1.5606895473176952e-05,
      "loss": 1.1365821361541748,
      "step": 442
    },
    {
      "epoch": 1.9137931034482758,
      "grad_norm": 0.20968762040138245,
      "learning_rate": 1.5527715769598582e-05,
      "loss": 0.2478710114955902,
      "step": 444
    },
    {
      "epoch": 1.9224137931034484,
      "grad_norm": 0.25190311670303345,
      "learning_rate": 1.544844564520796e-05,
      "loss": 0.636488676071167,
      "step": 446
    },
    {
      "epoch": 1.9310344827586206,
      "grad_norm": 0.09660618007183075,
      "learning_rate": 1.5369089131959546e-05,
      "loss": 1.0994303226470947,
      "step": 448
    },
    {
      "epoch": 1.9396551724137931,
      "grad_norm": 0.09424697607755661,
      "learning_rate": 1.5289650266201842e-05,
      "loss": 1.0804164409637451,
      "step": 450
    },
    {
      "epoch": 1.9482758620689655,
      "grad_norm": 0.12286921590566635,
      "learning_rate": 1.5210133088472096e-05,
      "loss": 0.9485524296760559,
      "step": 452
    },
    {
      "epoch": 1.956896551724138,
      "grad_norm": 0.23227761685848236,
      "learning_rate": 1.5130541643290755e-05,
      "loss": 1.0982067584991455,
      "step": 454
    },
    {
      "epoch": 1.9655172413793105,
      "grad_norm": 0.1315997689962387,
      "learning_rate": 1.5050879978955784e-05,
      "loss": 1.069095492362976,
      "step": 456
    },
    {
      "epoch": 1.9741379310344827,
      "grad_norm": 0.36187416315078735,
      "learning_rate": 1.4971152147336735e-05,
      "loss": 0.29495418071746826,
      "step": 458
    },
    {
      "epoch": 1.9827586206896552,
      "grad_norm": 0.28215736150741577,
      "learning_rate": 1.4891362203668646e-05,
      "loss": 1.0610284805297852,
      "step": 460
    },
    {
      "epoch": 1.9913793103448276,
      "grad_norm": 0.2604519724845886,
      "learning_rate": 1.4811514206345812e-05,
      "loss": 0.6210886240005493,
      "step": 462
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.8820472359657288,
      "learning_rate": 1.4731612216715315e-05,
      "loss": 0.796215295791626,
      "step": 464
    },
    {
      "epoch": 2.0086206896551726,
      "grad_norm": 0.090216726064682,
      "learning_rate": 1.4651660298870496e-05,
      "loss": 0.4970940947532654,
      "step": 466
    },
    {
      "epoch": 2.0172413793103448,
      "grad_norm": 0.17966338992118835,
      "learning_rate": 1.4571662519444208e-05,
      "loss": 0.7648203372955322,
      "step": 468
    },
    {
      "epoch": 2.0258620689655173,
      "grad_norm": 0.8345552682876587,
      "learning_rate": 1.4491622947401976e-05,
      "loss": 0.7012611627578735,
      "step": 470
    },
    {
      "epoch": 2.0344827586206895,
      "grad_norm": 0.14064906537532806,
      "learning_rate": 1.4411545653835056e-05,
      "loss": 0.5726131796836853,
      "step": 472
    },
    {
      "epoch": 2.043103448275862,
      "grad_norm": 0.13400284945964813,
      "learning_rate": 1.433143471175335e-05,
      "loss": 0.5506486296653748,
      "step": 474
    },
    {
      "epoch": 2.0517241379310347,
      "grad_norm": 0.10555726289749146,
      "learning_rate": 1.4251294195878231e-05,
      "loss": 0.5927596688270569,
      "step": 476
    },
    {
      "epoch": 2.060344827586207,
      "grad_norm": 0.10835747420787811,
      "learning_rate": 1.4171128182435312e-05,
      "loss": 0.4717426598072052,
      "step": 478
    },
    {
      "epoch": 2.0689655172413794,
      "grad_norm": 0.14888739585876465,
      "learning_rate": 1.409094074894709e-05,
      "loss": 0.8343894481658936,
      "step": 480
    },
    {
      "epoch": 2.0775862068965516,
      "grad_norm": 0.13630028069019318,
      "learning_rate": 1.4010735974025565e-05,
      "loss": 0.5780265927314758,
      "step": 482
    },
    {
      "epoch": 2.086206896551724,
      "grad_norm": 0.1016228199005127,
      "learning_rate": 1.3930517937164783e-05,
      "loss": 0.628856897354126,
      "step": 484
    },
    {
      "epoch": 2.0948275862068964,
      "grad_norm": 0.09641385078430176,
      "learning_rate": 1.3850290718533334e-05,
      "loss": 0.6124461889266968,
      "step": 486
    },
    {
      "epoch": 2.103448275862069,
      "grad_norm": 0.15994173288345337,
      "learning_rate": 1.3770058398766825e-05,
      "loss": 0.7476667761802673,
      "step": 488
    },
    {
      "epoch": 2.1120689655172415,
      "grad_norm": 0.2043793946504593,
      "learning_rate": 1.3689825058760336e-05,
      "loss": 0.8483129739761353,
      "step": 490
    },
    {
      "epoch": 2.1206896551724137,
      "grad_norm": 0.26761284470558167,
      "learning_rate": 1.3609594779460827e-05,
      "loss": 0.8416441082954407,
      "step": 492
    },
    {
      "epoch": 2.1293103448275863,
      "grad_norm": 0.1123756691813469,
      "learning_rate": 1.3529371641659582e-05,
      "loss": 0.8671878576278687,
      "step": 494
    },
    {
      "epoch": 2.1379310344827585,
      "grad_norm": 0.3551185429096222,
      "learning_rate": 1.344915972578465e-05,
      "loss": 0.919392466545105,
      "step": 496
    },
    {
      "epoch": 2.146551724137931,
      "grad_norm": 0.27395322918891907,
      "learning_rate": 1.3368963111693284e-05,
      "loss": 0.3866020143032074,
      "step": 498
    },
    {
      "epoch": 2.1551724137931036,
      "grad_norm": 0.16738073527812958,
      "learning_rate": 1.3288785878464447e-05,
      "loss": 0.4366337060928345,
      "step": 500
    },
    {
      "epoch": 2.163793103448276,
      "grad_norm": 0.12588895857334137,
      "learning_rate": 1.320863210419131e-05,
      "loss": 0.4676997661590576,
      "step": 502
    },
    {
      "epoch": 2.1724137931034484,
      "grad_norm": 0.08290023356676102,
      "learning_rate": 1.3128505865773855e-05,
      "loss": 0.5133532285690308,
      "step": 504
    },
    {
      "epoch": 2.1810344827586206,
      "grad_norm": 0.30187880992889404,
      "learning_rate": 1.3048411238711478e-05,
      "loss": 0.8324005603790283,
      "step": 506
    },
    {
      "epoch": 2.189655172413793,
      "grad_norm": 0.12202301621437073,
      "learning_rate": 1.296835229689573e-05,
      "loss": 0.49498435854911804,
      "step": 508
    },
    {
      "epoch": 2.1982758620689653,
      "grad_norm": 0.11802731454372406,
      "learning_rate": 1.2888333112403065e-05,
      "loss": 0.863714337348938,
      "step": 510
    },
    {
      "epoch": 2.206896551724138,
      "grad_norm": 0.3642047941684723,
      "learning_rate": 1.2808357755287764e-05,
      "loss": 0.4537322223186493,
      "step": 512
    },
    {
      "epoch": 2.2155172413793105,
      "grad_norm": 0.14233991503715515,
      "learning_rate": 1.2728430293374871e-05,
      "loss": 0.5666446685791016,
      "step": 514
    },
    {
      "epoch": 2.2241379310344827,
      "grad_norm": 0.11399707943201065,
      "learning_rate": 1.2648554792053342e-05,
      "loss": 0.44022682309150696,
      "step": 516
    },
    {
      "epoch": 2.2327586206896552,
      "grad_norm": 0.083638034760952,
      "learning_rate": 1.25687353140692e-05,
      "loss": 0.0981849655508995,
      "step": 518
    },
    {
      "epoch": 2.2413793103448274,
      "grad_norm": 0.7049289345741272,
      "learning_rate": 1.2488975919318963e-05,
      "loss": 0.15166987478733063,
      "step": 520
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.09605469554662704,
      "learning_rate": 1.2409280664643078e-05,
      "loss": 0.8936961889266968,
      "step": 522
    },
    {
      "epoch": 2.2586206896551726,
      "grad_norm": 0.09633728861808777,
      "learning_rate": 1.2329653603619629e-05,
      "loss": 0.6037598848342896,
      "step": 524
    },
    {
      "epoch": 2.2672413793103448,
      "grad_norm": 0.2239246815443039,
      "learning_rate": 1.2250098786358108e-05,
      "loss": 0.5711255669593811,
      "step": 526
    },
    {
      "epoch": 2.2758620689655173,
      "grad_norm": 0.0958482176065445,
      "learning_rate": 1.2170620259293458e-05,
      "loss": 0.5425680875778198,
      "step": 528
    },
    {
      "epoch": 2.2844827586206895,
      "grad_norm": 0.21729297935962677,
      "learning_rate": 1.2091222064980228e-05,
      "loss": 0.5258100628852844,
      "step": 530
    },
    {
      "epoch": 2.293103448275862,
      "grad_norm": 0.10278700292110443,
      "learning_rate": 1.2011908241886966e-05,
      "loss": 0.5525709390640259,
      "step": 532
    },
    {
      "epoch": 2.3017241379310347,
      "grad_norm": 0.13626424968242645,
      "learning_rate": 1.1932682824190799e-05,
      "loss": 0.6124492883682251,
      "step": 534
    },
    {
      "epoch": 2.310344827586207,
      "grad_norm": 0.08790895342826843,
      "learning_rate": 1.1853549841572268e-05,
      "loss": 0.5387366414070129,
      "step": 536
    },
    {
      "epoch": 2.3189655172413794,
      "grad_norm": 0.07563458383083344,
      "learning_rate": 1.177451331901031e-05,
      "loss": 0.32567426562309265,
      "step": 538
    },
    {
      "epoch": 2.3275862068965516,
      "grad_norm": 0.14534932374954224,
      "learning_rate": 1.169557727657761e-05,
      "loss": 0.46518173813819885,
      "step": 540
    },
    {
      "epoch": 2.336206896551724,
      "grad_norm": 0.16097325086593628,
      "learning_rate": 1.1616745729236045e-05,
      "loss": 0.774487316608429,
      "step": 542
    },
    {
      "epoch": 2.344827586206897,
      "grad_norm": 0.22597487270832062,
      "learning_rate": 1.1538022686632536e-05,
      "loss": 0.8165640830993652,
      "step": 544
    },
    {
      "epoch": 2.353448275862069,
      "grad_norm": 0.0724526047706604,
      "learning_rate": 1.145941215289505e-05,
      "loss": 0.19026756286621094,
      "step": 546
    },
    {
      "epoch": 2.3620689655172415,
      "grad_norm": 0.21904027462005615,
      "learning_rate": 1.1380918126428994e-05,
      "loss": 0.8604028820991516,
      "step": 548
    },
    {
      "epoch": 2.3706896551724137,
      "grad_norm": 0.2525452971458435,
      "learning_rate": 1.1302544599713775e-05,
      "loss": 0.4935520589351654,
      "step": 550
    },
    {
      "epoch": 2.3793103448275863,
      "grad_norm": 0.14142468571662903,
      "learning_rate": 1.122429555909979e-05,
      "loss": 0.7264618277549744,
      "step": 552
    },
    {
      "epoch": 2.3879310344827585,
      "grad_norm": 0.09017884731292725,
      "learning_rate": 1.114617498460562e-05,
      "loss": 0.6168904900550842,
      "step": 554
    },
    {
      "epoch": 2.396551724137931,
      "grad_norm": 0.10239855945110321,
      "learning_rate": 1.1068186849715635e-05,
      "loss": 0.6429441571235657,
      "step": 556
    },
    {
      "epoch": 2.405172413793103,
      "grad_norm": 0.14052097499370575,
      "learning_rate": 1.099033512117783e-05,
      "loss": 0.4293447434902191,
      "step": 558
    },
    {
      "epoch": 2.413793103448276,
      "grad_norm": 0.18142321705818176,
      "learning_rate": 1.091262375880214e-05,
      "loss": 0.5241134762763977,
      "step": 560
    },
    {
      "epoch": 2.4224137931034484,
      "grad_norm": 0.24674156308174133,
      "learning_rate": 1.0835056715258956e-05,
      "loss": 0.6372413039207458,
      "step": 562
    },
    {
      "epoch": 2.4310344827586206,
      "grad_norm": 0.16510127484798431,
      "learning_rate": 1.0757637935878126e-05,
      "loss": 0.7959206104278564,
      "step": 564
    },
    {
      "epoch": 2.439655172413793,
      "grad_norm": 0.4282341003417969,
      "learning_rate": 1.0680371358448266e-05,
      "loss": 0.6432863473892212,
      "step": 566
    },
    {
      "epoch": 2.4482758620689653,
      "grad_norm": 0.25148487091064453,
      "learning_rate": 1.0603260913016466e-05,
      "loss": 0.31055551767349243,
      "step": 568
    },
    {
      "epoch": 2.456896551724138,
      "grad_norm": 0.15222759544849396,
      "learning_rate": 1.05263105216884e-05,
      "loss": 0.9433165788650513,
      "step": 570
    },
    {
      "epoch": 2.4655172413793105,
      "grad_norm": 0.46396973729133606,
      "learning_rate": 1.044952409842884e-05,
      "loss": 0.5256112217903137,
      "step": 572
    },
    {
      "epoch": 2.4741379310344827,
      "grad_norm": 0.11589831113815308,
      "learning_rate": 1.0372905548862554e-05,
      "loss": 0.5775177478790283,
      "step": 574
    },
    {
      "epoch": 2.4827586206896552,
      "grad_norm": 0.12475285679101944,
      "learning_rate": 1.0296458770075702e-05,
      "loss": 0.8348941802978516,
      "step": 576
    },
    {
      "epoch": 2.4913793103448274,
      "grad_norm": 0.08721303939819336,
      "learning_rate": 1.0220187650417542e-05,
      "loss": 0.48024982213974,
      "step": 578
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.09861183911561966,
      "learning_rate": 1.0144096069302734e-05,
      "loss": 0.8509948253631592,
      "step": 580
    },
    {
      "epoch": 2.5086206896551726,
      "grad_norm": 0.11032187938690186,
      "learning_rate": 1.0068187897013957e-05,
      "loss": 0.8317974209785461,
      "step": 582
    },
    {
      "epoch": 2.5172413793103448,
      "grad_norm": 0.12203815579414368,
      "learning_rate": 9.992466994505098e-06,
      "loss": 0.6833478808403015,
      "step": 584
    },
    {
      "epoch": 2.5258620689655173,
      "grad_norm": 0.1454346776008606,
      "learning_rate": 9.916937213204835e-06,
      "loss": 0.6194900870323181,
      "step": 586
    },
    {
      "epoch": 2.5344827586206895,
      "grad_norm": 0.10960187017917633,
      "learning_rate": 9.841602394820772e-06,
      "loss": 0.8086000680923462,
      "step": 588
    },
    {
      "epoch": 2.543103448275862,
      "grad_norm": 0.43562886118888855,
      "learning_rate": 9.766466371144e-06,
      "loss": 0.8427749872207642,
      "step": 590
    },
    {
      "epoch": 2.5517241379310347,
      "grad_norm": 0.4390433430671692,
      "learning_rate": 9.691532963854247e-06,
      "loss": 0.7539307475090027,
      "step": 592
    },
    {
      "epoch": 2.560344827586207,
      "grad_norm": 0.08623161911964417,
      "learning_rate": 9.616805984325434e-06,
      "loss": 0.4550768733024597,
      "step": 594
    },
    {
      "epoch": 2.5689655172413794,
      "grad_norm": 0.07664068043231964,
      "learning_rate": 9.542289233431875e-06,
      "loss": 0.4484248459339142,
      "step": 596
    },
    {
      "epoch": 2.5775862068965516,
      "grad_norm": 0.12998774647712708,
      "learning_rate": 9.467986501354905e-06,
      "loss": 0.612811267375946,
      "step": 598
    },
    {
      "epoch": 2.586206896551724,
      "grad_norm": 0.20723579823970795,
      "learning_rate": 9.393901567390133e-06,
      "loss": 0.7580547332763672,
      "step": 600
    },
    {
      "epoch": 2.594827586206897,
      "grad_norm": 0.050058234483003616,
      "learning_rate": 9.320038199755181e-06,
      "loss": 0.5622333288192749,
      "step": 602
    },
    {
      "epoch": 2.603448275862069,
      "grad_norm": 0.2059715837240219,
      "learning_rate": 9.246400155398052e-06,
      "loss": 0.681129515171051,
      "step": 604
    },
    {
      "epoch": 2.612068965517241,
      "grad_norm": 0.1687193214893341,
      "learning_rate": 9.17299117980601e-06,
      "loss": 0.5278960466384888,
      "step": 606
    },
    {
      "epoch": 2.6206896551724137,
      "grad_norm": 0.0910777598619461,
      "learning_rate": 9.09981500681509e-06,
      "loss": 0.8580256104469299,
      "step": 608
    },
    {
      "epoch": 2.6293103448275863,
      "grad_norm": 0.17509189248085022,
      "learning_rate": 9.02687535842018e-06,
      "loss": 0.5199075937271118,
      "step": 610
    },
    {
      "epoch": 2.637931034482759,
      "grad_norm": 0.1662158966064453,
      "learning_rate": 8.9541759445857e-06,
      "loss": 0.8993173837661743,
      "step": 612
    },
    {
      "epoch": 2.646551724137931,
      "grad_norm": 0.12328854948282242,
      "learning_rate": 8.881720463056895e-06,
      "loss": 0.7683141827583313,
      "step": 614
    },
    {
      "epoch": 2.655172413793103,
      "grad_norm": 0.1306183785200119,
      "learning_rate": 8.809512599171781e-06,
      "loss": 0.5064551830291748,
      "step": 616
    },
    {
      "epoch": 2.663793103448276,
      "grad_norm": 0.09901191294193268,
      "learning_rate": 8.737556025673664e-06,
      "loss": 0.8363734483718872,
      "step": 618
    },
    {
      "epoch": 2.6724137931034484,
      "grad_norm": 0.02962668426334858,
      "learning_rate": 8.665854402524356e-06,
      "loss": 0.4550926387310028,
      "step": 620
    },
    {
      "epoch": 2.6810344827586206,
      "grad_norm": 0.14856787025928497,
      "learning_rate": 8.594411376718009e-06,
      "loss": 0.7523635029792786,
      "step": 622
    },
    {
      "epoch": 2.689655172413793,
      "grad_norm": 0.09993023425340652,
      "learning_rate": 8.523230582095595e-06,
      "loss": 0.4919755458831787,
      "step": 624
    },
    {
      "epoch": 2.6982758620689653,
      "grad_norm": 0.09833712875843048,
      "learning_rate": 8.452315639160121e-06,
      "loss": 0.7990663051605225,
      "step": 626
    },
    {
      "epoch": 2.706896551724138,
      "grad_norm": 0.1576949805021286,
      "learning_rate": 8.381670154892429e-06,
      "loss": 0.5231303572654724,
      "step": 628
    },
    {
      "epoch": 2.7155172413793105,
      "grad_norm": 0.07004235684871674,
      "learning_rate": 8.311297722567769e-06,
      "loss": 0.9687949419021606,
      "step": 630
    },
    {
      "epoch": 2.7241379310344827,
      "grad_norm": 0.7200844883918762,
      "learning_rate": 8.241201921573018e-06,
      "loss": 0.42503124475479126,
      "step": 632
    },
    {
      "epoch": 2.7327586206896552,
      "grad_norm": 0.1017894297838211,
      "learning_rate": 8.171386317224612e-06,
      "loss": 0.7973089814186096,
      "step": 634
    },
    {
      "epoch": 2.7413793103448274,
      "grad_norm": 0.11164805293083191,
      "learning_rate": 8.101854460587225e-06,
      "loss": 0.5349655747413635,
      "step": 636
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.15979333221912384,
      "learning_rate": 8.032609888293111e-06,
      "loss": 0.8247781991958618,
      "step": 638
    },
    {
      "epoch": 2.7586206896551726,
      "grad_norm": 0.12597469985485077,
      "learning_rate": 7.963656122362269e-06,
      "loss": 0.4262469708919525,
      "step": 640
    },
    {
      "epoch": 2.7672413793103448,
      "grad_norm": 0.33888915181159973,
      "learning_rate": 7.894996670023252e-06,
      "loss": 0.6027237176895142,
      "step": 642
    },
    {
      "epoch": 2.7758620689655173,
      "grad_norm": 0.10856196284294128,
      "learning_rate": 7.82663502353482e-06,
      "loss": 0.8435148000717163,
      "step": 644
    },
    {
      "epoch": 2.7844827586206895,
      "grad_norm": 0.11784857511520386,
      "learning_rate": 7.758574660008268e-06,
      "loss": 0.7197084426879883,
      "step": 646
    },
    {
      "epoch": 2.793103448275862,
      "grad_norm": 0.09773028641939163,
      "learning_rate": 7.690819041230616e-06,
      "loss": 0.5047876834869385,
      "step": 648
    },
    {
      "epoch": 2.8017241379310347,
      "grad_norm": 0.07827575504779816,
      "learning_rate": 7.623371613488475e-06,
      "loss": 0.6992777585983276,
      "step": 650
    },
    {
      "epoch": 2.810344827586207,
      "grad_norm": 0.28669506311416626,
      "learning_rate": 7.55623580739282e-06,
      "loss": 0.7797609567642212,
      "step": 652
    },
    {
      "epoch": 2.8189655172413794,
      "grad_norm": 0.30535227060317993,
      "learning_rate": 7.489415037704437e-06,
      "loss": 0.7613651156425476,
      "step": 654
    },
    {
      "epoch": 2.8275862068965516,
      "grad_norm": 0.09482012689113617,
      "learning_rate": 7.422912703160281e-06,
      "loss": 0.5962029695510864,
      "step": 656
    },
    {
      "epoch": 2.836206896551724,
      "grad_norm": 0.08960616588592529,
      "learning_rate": 7.3567321863005754e-06,
      "loss": 0.8467994928359985,
      "step": 658
    },
    {
      "epoch": 2.844827586206897,
      "grad_norm": 0.07423507422208786,
      "learning_rate": 7.290876853296791e-06,
      "loss": 0.2079772651195526,
      "step": 660
    },
    {
      "epoch": 2.853448275862069,
      "grad_norm": 0.8393888473510742,
      "learning_rate": 7.225350053780395e-06,
      "loss": 0.3967612087726593,
      "step": 662
    },
    {
      "epoch": 2.862068965517241,
      "grad_norm": 0.17007187008857727,
      "learning_rate": 7.1601551206725165e-06,
      "loss": 0.5651652216911316,
      "step": 664
    },
    {
      "epoch": 2.8706896551724137,
      "grad_norm": 0.09160315245389938,
      "learning_rate": 7.0952953700143915e-06,
      "loss": 0.4929698705673218,
      "step": 666
    },
    {
      "epoch": 2.8793103448275863,
      "grad_norm": 0.5120726823806763,
      "learning_rate": 7.0307741007987234e-06,
      "loss": 0.3763951063156128,
      "step": 668
    },
    {
      "epoch": 2.887931034482759,
      "grad_norm": 0.14259645342826843,
      "learning_rate": 6.9665945948018505e-06,
      "loss": 0.8635995388031006,
      "step": 670
    },
    {
      "epoch": 2.896551724137931,
      "grad_norm": 0.14974215626716614,
      "learning_rate": 6.902760116416865e-06,
      "loss": 0.5167340636253357,
      "step": 672
    },
    {
      "epoch": 2.905172413793103,
      "grad_norm": 0.1967867761850357,
      "learning_rate": 6.83927391248754e-06,
      "loss": 0.7079530358314514,
      "step": 674
    },
    {
      "epoch": 2.913793103448276,
      "grad_norm": 0.7743811011314392,
      "learning_rate": 6.776139212143207e-06,
      "loss": 0.7067210674285889,
      "step": 676
    },
    {
      "epoch": 2.9224137931034484,
      "grad_norm": 0.06761736422777176,
      "learning_rate": 6.713359226634489e-06,
      "loss": 0.8161290287971497,
      "step": 678
    },
    {
      "epoch": 2.9310344827586206,
      "grad_norm": 0.08846219629049301,
      "learning_rate": 6.650937149169996e-06,
      "loss": 0.6068652868270874,
      "step": 680
    },
    {
      "epoch": 2.939655172413793,
      "grad_norm": 0.04591520130634308,
      "learning_rate": 6.588876154753869e-06,
      "loss": 0.2772270143032074,
      "step": 682
    },
    {
      "epoch": 2.9482758620689653,
      "grad_norm": 0.07999559491872787,
      "learning_rate": 6.527179400024325e-06,
      "loss": 0.3592681288719177,
      "step": 684
    },
    {
      "epoch": 2.956896551724138,
      "grad_norm": 0.10797300934791565,
      "learning_rate": 6.46585002309307e-06,
      "loss": 0.7297719120979309,
      "step": 686
    },
    {
      "epoch": 2.9655172413793105,
      "grad_norm": 0.06590516865253448,
      "learning_rate": 6.4048911433857075e-06,
      "loss": 0.7315416932106018,
      "step": 688
    },
    {
      "epoch": 2.9741379310344827,
      "grad_norm": 0.11769416928291321,
      "learning_rate": 6.344305861483049e-06,
      "loss": 0.9271728992462158,
      "step": 690
    },
    {
      "epoch": 2.9827586206896552,
      "grad_norm": 0.08777745068073273,
      "learning_rate": 6.284097258963437e-06,
      "loss": 0.9229543209075928,
      "step": 692
    },
    {
      "epoch": 2.9913793103448274,
      "grad_norm": 0.21901434659957886,
      "learning_rate": 6.224268398245974e-06,
      "loss": 0.503404974937439,
      "step": 694
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.10449380427598953,
      "learning_rate": 6.164822322434786e-06,
      "loss": 0.9257153272628784,
      "step": 696
    },
    {
      "epoch": 3.0086206896551726,
      "grad_norm": 0.09770919382572174,
      "learning_rate": 6.105762055164216e-06,
      "loss": 0.33168870210647583,
      "step": 698
    },
    {
      "epoch": 3.0172413793103448,
      "grad_norm": 0.08789953589439392,
      "learning_rate": 6.047090600445046e-06,
      "loss": 0.5617142915725708,
      "step": 700
    },
    {
      "epoch": 3.0258620689655173,
      "grad_norm": 0.20638461410999298,
      "learning_rate": 5.988810942511703e-06,
      "loss": 0.3617658317089081,
      "step": 702
    },
    {
      "epoch": 3.0344827586206895,
      "grad_norm": 0.1723642200231552,
      "learning_rate": 5.930926045670452e-06,
      "loss": 0.4505208134651184,
      "step": 704
    },
    {
      "epoch": 3.043103448275862,
      "grad_norm": 0.015494012273848057,
      "learning_rate": 5.873438854148648e-06,
      "loss": 0.07196473330259323,
      "step": 706
    },
    {
      "epoch": 3.0517241379310347,
      "grad_norm": 0.3174591660499573,
      "learning_rate": 5.816352291944968e-06,
      "loss": 0.5367481112480164,
      "step": 708
    },
    {
      "epoch": 3.060344827586207,
      "grad_norm": 0.04589797183871269,
      "learning_rate": 5.759669262680674e-06,
      "loss": 0.13180424273014069,
      "step": 710
    },
    {
      "epoch": 3.0689655172413794,
      "grad_norm": 0.1757347136735916,
      "learning_rate": 5.703392649451958e-06,
      "loss": 0.20174063742160797,
      "step": 712
    },
    {
      "epoch": 3.0775862068965516,
      "grad_norm": 0.2314152717590332,
      "learning_rate": 5.647525314683258e-06,
      "loss": 0.7795681357383728,
      "step": 714
    },
    {
      "epoch": 3.086206896551724,
      "grad_norm": 0.12419137358665466,
      "learning_rate": 5.592070099981703e-06,
      "loss": 0.23963159322738647,
      "step": 716
    },
    {
      "epoch": 3.0948275862068964,
      "grad_norm": 0.08404622972011566,
      "learning_rate": 5.53702982599254e-06,
      "loss": 0.5431758761405945,
      "step": 718
    },
    {
      "epoch": 3.103448275862069,
      "grad_norm": 0.055214498192071915,
      "learning_rate": 5.4824072922557125e-06,
      "loss": 0.14490705728530884,
      "step": 720
    },
    {
      "epoch": 3.1120689655172415,
      "grad_norm": 0.09599906206130981,
      "learning_rate": 5.428205277063415e-06,
      "loss": 0.2814076840877533,
      "step": 722
    },
    {
      "epoch": 3.1206896551724137,
      "grad_norm": 0.1080830916762352,
      "learning_rate": 5.374426537318827e-06,
      "loss": 0.4774016737937927,
      "step": 724
    },
    {
      "epoch": 3.1293103448275863,
      "grad_norm": 0.06993446499109268,
      "learning_rate": 5.321073808395846e-06,
      "loss": 0.5318699479103088,
      "step": 726
    },
    {
      "epoch": 3.1379310344827585,
      "grad_norm": 0.2722335755825043,
      "learning_rate": 5.268149803999993e-06,
      "loss": 0.5016626119613647,
      "step": 728
    },
    {
      "epoch": 3.146551724137931,
      "grad_norm": 0.08205164968967438,
      "learning_rate": 5.215657216030353e-06,
      "loss": 0.34347403049468994,
      "step": 730
    },
    {
      "epoch": 3.1551724137931036,
      "grad_norm": 0.08843845874071121,
      "learning_rate": 5.163598714442685e-06,
      "loss": 0.25022614002227783,
      "step": 732
    },
    {
      "epoch": 3.163793103448276,
      "grad_norm": 0.12089242041110992,
      "learning_rate": 5.111976947113585e-06,
      "loss": 0.5590460300445557,
      "step": 734
    },
    {
      "epoch": 3.1724137931034484,
      "grad_norm": 0.02552367001771927,
      "learning_rate": 5.0607945397058415e-06,
      "loss": 0.27467963099479675,
      "step": 736
    },
    {
      "epoch": 3.1810344827586206,
      "grad_norm": 0.6171451807022095,
      "learning_rate": 5.010054095534857e-06,
      "loss": 0.5068318843841553,
      "step": 738
    },
    {
      "epoch": 3.189655172413793,
      "grad_norm": 0.09212607145309448,
      "learning_rate": 4.959758195436251e-06,
      "loss": 0.10915378481149673,
      "step": 740
    },
    {
      "epoch": 3.1982758620689653,
      "grad_norm": 0.03245019167661667,
      "learning_rate": 4.9099093976345804e-06,
      "loss": 0.26452872157096863,
      "step": 742
    },
    {
      "epoch": 3.206896551724138,
      "grad_norm": 0.09578347951173782,
      "learning_rate": 4.860510237613227e-06,
      "loss": 0.41886910796165466,
      "step": 744
    },
    {
      "epoch": 3.2155172413793105,
      "grad_norm": 0.1211688295006752,
      "learning_rate": 4.811563227985424e-06,
      "loss": 0.3879053592681885,
      "step": 746
    },
    {
      "epoch": 3.2241379310344827,
      "grad_norm": 0.09857846051454544,
      "learning_rate": 4.763070858366472e-06,
      "loss": 0.5333488583564758,
      "step": 748
    },
    {
      "epoch": 3.2327586206896552,
      "grad_norm": 0.10673126578330994,
      "learning_rate": 4.7150355952470814e-06,
      "loss": 0.5044701099395752,
      "step": 750
    },
    {
      "epoch": 3.2413793103448274,
      "grad_norm": 0.10853103548288345,
      "learning_rate": 4.667459881867956e-06,
      "loss": 0.29571160674095154,
      "step": 752
    },
    {
      "epoch": 3.25,
      "grad_norm": 0.13593336939811707,
      "learning_rate": 4.6203461380954775e-06,
      "loss": 0.567728579044342,
      "step": 754
    },
    {
      "epoch": 3.2586206896551726,
      "grad_norm": 0.2100203037261963,
      "learning_rate": 4.5736967602986656e-06,
      "loss": 0.3926270306110382,
      "step": 756
    },
    {
      "epoch": 3.2672413793103448,
      "grad_norm": 0.10009382665157318,
      "learning_rate": 4.527514121227256e-06,
      "loss": 0.46212252974510193,
      "step": 758
    },
    {
      "epoch": 3.2758620689655173,
      "grad_norm": 0.03222054988145828,
      "learning_rate": 4.481800569891035e-06,
      "loss": 0.326346218585968,
      "step": 760
    },
    {
      "epoch": 3.2844827586206895,
      "grad_norm": 0.18550357222557068,
      "learning_rate": 4.4365584314403535e-06,
      "loss": 0.42745885252952576,
      "step": 762
    },
    {
      "epoch": 3.293103448275862,
      "grad_norm": 0.44906678795814514,
      "learning_rate": 4.391790007047859e-06,
      "loss": 0.47123435139656067,
      "step": 764
    },
    {
      "epoch": 3.3017241379310347,
      "grad_norm": 0.04999677091836929,
      "learning_rate": 4.347497573791454e-06,
      "loss": 0.28395605087280273,
      "step": 766
    },
    {
      "epoch": 3.310344827586207,
      "grad_norm": 0.15627415478229523,
      "learning_rate": 4.303683384538477e-06,
      "loss": 0.43049338459968567,
      "step": 768
    },
    {
      "epoch": 3.3189655172413794,
      "grad_norm": 0.18941667675971985,
      "learning_rate": 4.260349667831108e-06,
      "loss": 0.14107981324195862,
      "step": 770
    },
    {
      "epoch": 3.3275862068965516,
      "grad_norm": 0.11692267656326294,
      "learning_rate": 4.217498627773029e-06,
      "loss": 0.3092440068721771,
      "step": 772
    },
    {
      "epoch": 3.336206896551724,
      "grad_norm": 0.2257245033979416,
      "learning_rate": 4.175132443917291e-06,
      "loss": 0.5438483357429504,
      "step": 774
    },
    {
      "epoch": 3.344827586206897,
      "grad_norm": 0.10574281215667725,
      "learning_rate": 4.133253271155489e-06,
      "loss": 0.5660426616668701,
      "step": 776
    },
    {
      "epoch": 3.353448275862069,
      "grad_norm": 0.09007226675748825,
      "learning_rate": 4.091863239608131e-06,
      "loss": 0.522563636302948,
      "step": 778
    },
    {
      "epoch": 3.3620689655172415,
      "grad_norm": 0.2889229357242584,
      "learning_rate": 4.050964454516295e-06,
      "loss": 0.3490191698074341,
      "step": 780
    },
    {
      "epoch": 3.3706896551724137,
      "grad_norm": 0.11218395084142685,
      "learning_rate": 4.010558996134562e-06,
      "loss": 0.3295826017856598,
      "step": 782
    },
    {
      "epoch": 3.3793103448275863,
      "grad_norm": 0.08473268151283264,
      "learning_rate": 3.970648919625199e-06,
      "loss": 0.5116720199584961,
      "step": 784
    },
    {
      "epoch": 3.3879310344827585,
      "grad_norm": 0.10021858662366867,
      "learning_rate": 3.9312362549536195e-06,
      "loss": 0.3242761194705963,
      "step": 786
    },
    {
      "epoch": 3.396551724137931,
      "grad_norm": 0.09062958508729935,
      "learning_rate": 3.892323006785148e-06,
      "loss": 0.49483513832092285,
      "step": 788
    },
    {
      "epoch": 3.405172413793103,
      "grad_norm": 0.09515772014856339,
      "learning_rate": 3.853911154383037e-06,
      "loss": 0.4984126091003418,
      "step": 790
    },
    {
      "epoch": 3.413793103448276,
      "grad_norm": 0.0715399831533432,
      "learning_rate": 3.8160026515078165e-06,
      "loss": 0.2216302752494812,
      "step": 792
    },
    {
      "epoch": 3.4224137931034484,
      "grad_norm": 0.125926673412323,
      "learning_rate": 3.7785994263178978e-06,
      "loss": 0.24825817346572876,
      "step": 794
    },
    {
      "epoch": 3.4310344827586206,
      "grad_norm": 0.10221999138593674,
      "learning_rate": 3.741703381271519e-06,
      "loss": 0.6700829863548279,
      "step": 796
    },
    {
      "epoch": 3.439655172413793,
      "grad_norm": 0.18892526626586914,
      "learning_rate": 3.7053163930299625e-06,
      "loss": 0.3420178294181824,
      "step": 798
    },
    {
      "epoch": 3.4482758620689653,
      "grad_norm": 0.13281142711639404,
      "learning_rate": 3.6694403123621143e-06,
      "loss": 0.5975306034088135,
      "step": 800
    },
    {
      "epoch": 3.456896551724138,
      "grad_norm": 0.0972677543759346,
      "learning_rate": 3.634076964050321e-06,
      "loss": 0.4603136479854584,
      "step": 802
    },
    {
      "epoch": 3.4655172413793105,
      "grad_norm": 0.11561470478773117,
      "learning_rate": 3.5992281467975773e-06,
      "loss": 0.4385143518447876,
      "step": 804
    },
    {
      "epoch": 3.4741379310344827,
      "grad_norm": 0.13390004634857178,
      "learning_rate": 3.5648956331360366e-06,
      "loss": 0.4603370428085327,
      "step": 806
    },
    {
      "epoch": 3.4827586206896552,
      "grad_norm": 0.09613792598247528,
      "learning_rate": 3.5310811693368584e-06,
      "loss": 0.5407328605651855,
      "step": 808
    },
    {
      "epoch": 3.4913793103448274,
      "grad_norm": 0.02905660681426525,
      "learning_rate": 3.4977864753213727e-06,
      "loss": 0.14333094656467438,
      "step": 810
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.40168488025665283,
      "learning_rate": 3.4650132445736226e-06,
      "loss": 0.3744501769542694,
      "step": 812
    },
    {
      "epoch": 3.5086206896551726,
      "grad_norm": 0.18611806631088257,
      "learning_rate": 3.4327631440542018e-06,
      "loss": 0.49726125597953796,
      "step": 814
    },
    {
      "epoch": 3.5172413793103448,
      "grad_norm": 0.1080632358789444,
      "learning_rate": 3.401037814115489e-06,
      "loss": 0.441265344619751,
      "step": 816
    },
    {
      "epoch": 3.5258620689655173,
      "grad_norm": 0.17538908123970032,
      "learning_rate": 3.3698388684181987e-06,
      "loss": 0.5781593322753906,
      "step": 818
    },
    {
      "epoch": 3.5344827586206895,
      "grad_norm": 0.08419691771268845,
      "learning_rate": 3.339167893849313e-06,
      "loss": 0.5013993382453918,
      "step": 820
    },
    {
      "epoch": 3.543103448275862,
      "grad_norm": 0.23287096619606018,
      "learning_rate": 3.3090264504413618e-06,
      "loss": 0.2739977240562439,
      "step": 822
    },
    {
      "epoch": 3.5517241379310347,
      "grad_norm": 0.08283835649490356,
      "learning_rate": 3.279416071293083e-06,
      "loss": 0.33216506242752075,
      "step": 824
    },
    {
      "epoch": 3.560344827586207,
      "grad_norm": 0.03011569194495678,
      "learning_rate": 3.2503382624914287e-06,
      "loss": 0.08031827956438065,
      "step": 826
    },
    {
      "epoch": 3.5689655172413794,
      "grad_norm": 0.15765611827373505,
      "learning_rate": 3.2217945030349785e-06,
      "loss": 0.13542333245277405,
      "step": 828
    },
    {
      "epoch": 3.5775862068965516,
      "grad_norm": 0.1197817251086235,
      "learning_rate": 3.1937862447586945e-06,
      "loss": 0.3155873119831085,
      "step": 830
    },
    {
      "epoch": 3.586206896551724,
      "grad_norm": 0.12214954197406769,
      "learning_rate": 3.1663149122600946e-06,
      "loss": 0.32241955399513245,
      "step": 832
    },
    {
      "epoch": 3.594827586206897,
      "grad_norm": 0.11514101922512054,
      "learning_rate": 3.1393819028267712e-06,
      "loss": 0.5564541220664978,
      "step": 834
    },
    {
      "epoch": 3.603448275862069,
      "grad_norm": 0.12636107206344604,
      "learning_rate": 3.112988586365343e-06,
      "loss": 0.3382505476474762,
      "step": 836
    },
    {
      "epoch": 3.612068965517241,
      "grad_norm": 0.13530541956424713,
      "learning_rate": 3.0871363053317565e-06,
      "loss": 0.4953022599220276,
      "step": 838
    },
    {
      "epoch": 3.6206896551724137,
      "grad_norm": 0.21053758263587952,
      "learning_rate": 3.0618263746630195e-06,
      "loss": 0.12510839104652405,
      "step": 840
    },
    {
      "epoch": 3.6293103448275863,
      "grad_norm": 0.20605450868606567,
      "learning_rate": 3.0370600817103054e-06,
      "loss": 0.12819121778011322,
      "step": 842
    },
    {
      "epoch": 3.637931034482759,
      "grad_norm": 0.08630102127790451,
      "learning_rate": 3.0128386861734907e-06,
      "loss": 0.34562334418296814,
      "step": 844
    },
    {
      "epoch": 3.646551724137931,
      "grad_norm": 0.21599891781806946,
      "learning_rate": 2.9891634200370615e-06,
      "loss": 0.2538186311721802,
      "step": 846
    },
    {
      "epoch": 3.655172413793103,
      "grad_norm": 0.08544554561376572,
      "learning_rate": 2.9660354875074694e-06,
      "loss": 0.09849962592124939,
      "step": 848
    },
    {
      "epoch": 3.663793103448276,
      "grad_norm": 0.058089643716812134,
      "learning_rate": 2.9434560649518708e-06,
      "loss": 0.31217214465141296,
      "step": 850
    },
    {
      "epoch": 3.6724137931034484,
      "grad_norm": 0.144460067152977,
      "learning_rate": 2.9214263008382943e-06,
      "loss": 0.5859465003013611,
      "step": 852
    },
    {
      "epoch": 3.6810344827586206,
      "grad_norm": 0.11717334389686584,
      "learning_rate": 2.8999473156772322e-06,
      "loss": 0.32843440771102905,
      "step": 854
    },
    {
      "epoch": 3.689655172413793,
      "grad_norm": 0.42045170068740845,
      "learning_rate": 2.8790202019646327e-06,
      "loss": 0.525081992149353,
      "step": 856
    },
    {
      "epoch": 3.6982758620689653,
      "grad_norm": 0.10836631059646606,
      "learning_rate": 2.858646024126349e-06,
      "loss": 0.5662536025047302,
      "step": 858
    },
    {
      "epoch": 3.706896551724138,
      "grad_norm": 0.044400930404663086,
      "learning_rate": 2.8388258184639895e-06,
      "loss": 0.26983019709587097,
      "step": 860
    },
    {
      "epoch": 3.7155172413793105,
      "grad_norm": 0.20253317058086395,
      "learning_rate": 2.8195605931022044e-06,
      "loss": 0.325879842042923,
      "step": 862
    },
    {
      "epoch": 3.7241379310344827,
      "grad_norm": 0.021246735006570816,
      "learning_rate": 2.8008513279374227e-06,
      "loss": 0.26106154918670654,
      "step": 864
    },
    {
      "epoch": 3.7327586206896552,
      "grad_norm": 0.2902374267578125,
      "learning_rate": 2.7826989745879937e-06,
      "loss": 0.0877346470952034,
      "step": 866
    },
    {
      "epoch": 3.7413793103448274,
      "grad_norm": 0.16706323623657227,
      "learning_rate": 2.7651044563457983e-06,
      "loss": 0.32902559638023376,
      "step": 868
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.06072617322206497,
      "learning_rate": 2.748068668129279e-06,
      "loss": 0.40092846751213074,
      "step": 870
    },
    {
      "epoch": 3.7586206896551726,
      "grad_norm": 0.02342204377055168,
      "learning_rate": 2.7315924764379292e-06,
      "loss": 0.12595315277576447,
      "step": 872
    },
    {
      "epoch": 3.7672413793103448,
      "grad_norm": 0.019321495667099953,
      "learning_rate": 2.715676719308208e-06,
      "loss": 0.30979275703430176,
      "step": 874
    },
    {
      "epoch": 3.7758620689655173,
      "grad_norm": 0.10496604442596436,
      "learning_rate": 2.7003222062709295e-06,
      "loss": 0.5150949358940125,
      "step": 876
    },
    {
      "epoch": 3.7844827586206895,
      "grad_norm": 0.2970009446144104,
      "learning_rate": 2.6855297183100716e-06,
      "loss": 0.41893500089645386,
      "step": 878
    },
    {
      "epoch": 3.793103448275862,
      "grad_norm": 0.1417495757341385,
      "learning_rate": 2.671300007823068e-06,
      "loss": 0.28298234939575195,
      "step": 880
    },
    {
      "epoch": 3.8017241379310347,
      "grad_norm": 0.13228200376033783,
      "learning_rate": 2.657633798582525e-06,
      "loss": 0.29594770073890686,
      "step": 882
    },
    {
      "epoch": 3.810344827586207,
      "grad_norm": 0.23043398559093475,
      "learning_rate": 2.6445317856994187e-06,
      "loss": 0.5765542984008789,
      "step": 884
    },
    {
      "epoch": 3.8189655172413794,
      "grad_norm": 0.19636346399784088,
      "learning_rate": 2.6319946355877313e-06,
      "loss": 0.18173623085021973,
      "step": 886
    },
    {
      "epoch": 3.8275862068965516,
      "grad_norm": 0.09358557313680649,
      "learning_rate": 2.62002298593056e-06,
      "loss": 0.3384171724319458,
      "step": 888
    },
    {
      "epoch": 3.836206896551724,
      "grad_norm": 0.0909893587231636,
      "learning_rate": 2.6086174456476774e-06,
      "loss": 0.10825614631175995,
      "step": 890
    },
    {
      "epoch": 3.844827586206897,
      "grad_norm": 0.15050019323825836,
      "learning_rate": 2.5977785948645674e-06,
      "loss": 0.36479508876800537,
      "step": 892
    },
    {
      "epoch": 3.853448275862069,
      "grad_norm": 0.07940158993005753,
      "learning_rate": 2.58750698488291e-06,
      "loss": 0.4843851923942566,
      "step": 894
    },
    {
      "epoch": 3.862068965517241,
      "grad_norm": 0.08099526911973953,
      "learning_rate": 2.577803138152543e-06,
      "loss": 0.34008628129959106,
      "step": 896
    },
    {
      "epoch": 3.8706896551724137,
      "grad_norm": 0.14477792382240295,
      "learning_rate": 2.56866754824489e-06,
      "loss": 0.2734866142272949,
      "step": 898
    },
    {
      "epoch": 3.8793103448275863,
      "grad_norm": 0.08499010652303696,
      "learning_rate": 2.560100679827858e-06,
      "loss": 0.5301221013069153,
      "step": 900
    },
    {
      "epoch": 3.887931034482759,
      "grad_norm": 0.1628270447254181,
      "learning_rate": 2.5521029686421927e-06,
      "loss": 0.40625,
      "step": 902
    },
    {
      "epoch": 3.896551724137931,
      "grad_norm": 0.1032799556851387,
      "learning_rate": 2.544674821479325e-06,
      "loss": 0.412536084651947,
      "step": 904
    },
    {
      "epoch": 3.905172413793103,
      "grad_norm": 0.19824664294719696,
      "learning_rate": 2.5378166161606777e-06,
      "loss": 0.5061193108558655,
      "step": 906
    },
    {
      "epoch": 3.913793103448276,
      "grad_norm": 0.2031039297580719,
      "learning_rate": 2.5315287015184486e-06,
      "loss": 0.2978528141975403,
      "step": 908
    },
    {
      "epoch": 3.9224137931034484,
      "grad_norm": 0.1463300883769989,
      "learning_rate": 2.525811397377861e-06,
      "loss": 0.21743205189704895,
      "step": 910
    },
    {
      "epoch": 3.9310344827586206,
      "grad_norm": 0.10597418993711472,
      "learning_rate": 2.520664994540907e-06,
      "loss": 0.2807198464870453,
      "step": 912
    },
    {
      "epoch": 3.939655172413793,
      "grad_norm": 0.11543630063533783,
      "learning_rate": 2.5160897547715503e-06,
      "loss": 0.537182629108429,
      "step": 914
    },
    {
      "epoch": 3.9482758620689653,
      "grad_norm": 0.12250090390443802,
      "learning_rate": 2.5120859107824107e-06,
      "loss": 0.4790320098400116,
      "step": 916
    },
    {
      "epoch": 3.956896551724138,
      "grad_norm": 0.1139577180147171,
      "learning_rate": 2.508653666222932e-06,
      "loss": 0.3746393024921417,
      "step": 918
    },
    {
      "epoch": 3.9655172413793105,
      "grad_norm": 0.34079810976982117,
      "learning_rate": 2.5057931956690186e-06,
      "loss": 0.3896881341934204,
      "step": 920
    },
    {
      "epoch": 3.9741379310344827,
      "grad_norm": 0.1102144867181778,
      "learning_rate": 2.5035046446141617e-06,
      "loss": 0.40220990777015686,
      "step": 922
    },
    {
      "epoch": 3.9827586206896552,
      "grad_norm": 0.11032397300004959,
      "learning_rate": 2.501788129462033e-06,
      "loss": 0.5496131777763367,
      "step": 924
    },
    {
      "epoch": 3.9913793103448274,
      "grad_norm": 0.12315075099468231,
      "learning_rate": 2.5006437375205682e-06,
      "loss": 0.5806067585945129,
      "step": 926
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.18088358640670776,
      "learning_rate": 2.5000715269975277e-06,
      "loss": 0.39202290773391724,
      "step": 928
    },
    {
      "epoch": 4.0,
      "step": 928,
      "total_flos": 3.61934117404672e+18,
      "train_loss": 0.7849958032722874,
      "train_runtime": 32566.8601,
      "train_samples_per_second": 1.71,
      "train_steps_per_second": 0.028
    }
  ],
  "logging_steps": 2,
  "max_steps": 928,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 99999,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.61934117404672e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}