diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,31943 +1,3992 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 1.0,
+  "epoch": 0.9999069680900549,
   "eval_steps": 100,
-  "global_step": 21570,
+  "global_step": 2687,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.0002318034306907742,
-      "grad_norm": 6.929919242858887,
-      "learning_rate": 4.636068613815485e-08,
-      "loss": 1.3729,
+      "epoch": 0.0018606381989022234,
+      "grad_norm": 0.025198739022016525,
+      "learning_rate": 3.7174721189591085e-07,
+      "loss": 0.7637,
       "step": 5
     },
     {
-      "epoch": 0.0004636068613815484,
-      "grad_norm": 7.42155647277832,
-      "learning_rate": 9.27213722763097e-08,
-      "loss": 1.33,
+      "epoch": 0.003721276397804447,
+      "grad_norm": 0.017377199605107307,
+      "learning_rate": 7.434944237918217e-07,
+      "loss": 0.7786,
       "step": 10
     },
     {
-      "epoch": 0.0006954102920723226,
-      "grad_norm": 7.66618537902832,
-      "learning_rate": 1.3908205841446455e-07,
-      "loss": 1.1782,
+      "epoch": 0.0055819145967066705,
+      "grad_norm": 0.019743537530303,
+      "learning_rate": 1.1152416356877324e-06,
+      "loss": 0.7174,
       "step": 15
     },
     {
-      "epoch": 0.0009272137227630969,
-      "grad_norm": 7.257485389709473,
-      "learning_rate": 1.854427445526194e-07,
-      "loss": 1.2896,
+      "epoch": 0.007442552795608894,
+      "grad_norm": 0.017805561423301697,
+      "learning_rate": 1.4869888475836434e-06,
+      "loss": 0.724,
       "step": 20
     },
     {
-      "epoch": 0.0011590171534538712,
-      "grad_norm": 9.955284118652344,
-      "learning_rate": 2.3180343069077424e-07,
-      "loss": 1.3364,
+      "epoch": 0.009303190994511117,
+      "grad_norm": 0.017246991395950317,
+      "learning_rate": 1.858736059479554e-06,
+      "loss": 0.7668,
       "step": 25
     },
     {
-      "epoch": 0.0013908205841446453,
-      "grad_norm": 10.206951141357422,
-      "learning_rate": 2.781641168289291e-07,
-      "loss": 1.4175,
+      "epoch": 0.011163829193413341,
+      "grad_norm": 0.023152414709329605,
+      "learning_rate": 2.2304832713754648e-06,
+      "loss": 0.781,
       "step": 30
     },
     {
-      "epoch": 0.0016226240148354196,
-      "grad_norm": 7.223724365234375,
-      "learning_rate": 3.2452480296708393e-07,
-      "loss": 1.2606,
+      "epoch": 0.013024467392315565,
+      "grad_norm": 0.019485710188746452,
+      "learning_rate": 2.6022304832713758e-06,
+      "loss": 0.7593,
       "step": 35
     },
     {
-      "epoch": 0.0018544274455261937,
-      "grad_norm": 7.419017791748047,
-      "learning_rate": 3.708854891052388e-07,
-      "loss": 1.2551,
+      "epoch": 0.014885105591217787,
+      "grad_norm": 0.01784881390631199,
+      "learning_rate": 2.973977695167287e-06,
+      "loss": 0.762,
       "step": 40
     },
     {
-      "epoch": 0.002086230876216968,
-      "grad_norm": 7.923377990722656,
-      "learning_rate": 4.1724617524339365e-07,
-      "loss": 1.3281,
+      "epoch": 0.01674574379012001,
+      "grad_norm": 0.01972021535038948,
+      "learning_rate": 3.3457249070631974e-06,
+      "loss": 0.7489,
       "step": 45
     },
     {
-      "epoch": 0.0023180343069077423,
-      "grad_norm": 8.805521011352539,
-      "learning_rate": 4.636068613815485e-07,
-      "loss": 1.2294,
+      "epoch": 0.018606381989022234,
+      "grad_norm": 0.019333072006702423,
+      "learning_rate": 3.717472118959108e-06,
+      "loss": 0.7461,
       "step": 50
     },
     {
-      "epoch": 0.0025498377375985167,
-      "grad_norm": 7.392923355102539,
-      "learning_rate": 5.099675475197033e-07,
-      "loss": 1.1609,
+      "epoch": 0.02046702018792446,
+      "grad_norm": 0.015675414353609085,
+      "learning_rate": 4.089219330855019e-06,
+      "loss": 0.7621,
       "step": 55
     },
     {
-      "epoch": 0.0027816411682892906,
-      "grad_norm": 7.491943836212158,
-      "learning_rate": 5.563282336578582e-07,
-      "loss": 1.2333,
+      "epoch": 0.022327658386826682,
+      "grad_norm": 0.021361010149121284,
+      "learning_rate": 4.4609665427509296e-06,
+      "loss": 0.7677,
       "step": 60
     },
     {
-      "epoch": 0.003013444598980065,
-      "grad_norm": 7.105735778808594,
-      "learning_rate": 6.026889197960131e-07,
-      "loss": 1.4028,
+      "epoch": 0.024188296585728904,
+      "grad_norm": 0.021466689184308052,
+      "learning_rate": 4.83271375464684e-06,
+      "loss": 0.7857,
       "step": 65
     },
     {
-      "epoch": 0.003245248029670839,
-      "grad_norm": 5.720520973205566,
-      "learning_rate": 6.490496059341679e-07,
-      "loss": 1.2407,
+      "epoch": 0.02604893478463113,
+      "grad_norm": 0.019599348306655884,
+      "learning_rate": 5.2044609665427516e-06,
+      "loss": 0.7308,
       "step": 70
     },
     {
-      "epoch": 0.0034770514603616135,
-      "grad_norm": 6.424951076507568,
-      "learning_rate": 6.954102920723226e-07,
-      "loss": 1.1028,
+      "epoch": 0.027909572983533353,
+      "grad_norm": 0.015855278819799423,
+      "learning_rate": 5.576208178438662e-06,
+      "loss": 0.7423,
       "step": 75
     },
     {
-      "epoch": 0.0037088548910523874,
-      "grad_norm": 7.229697227478027,
-      "learning_rate": 7.417709782104776e-07,
-      "loss": 1.2738,
+      "epoch": 0.029770211182435575,
+      "grad_norm": 0.019443219527602196,
+      "learning_rate": 5.947955390334574e-06,
+      "loss": 0.7782,
       "step": 80
     },
     {
-      "epoch": 0.003940658321743162,
-      "grad_norm": 6.052042484283447,
-      "learning_rate": 7.881316643486324e-07,
-      "loss": 1.2327,
+      "epoch": 0.0316308493813378,
+      "grad_norm": 0.01575319655239582,
+      "learning_rate": 6.319702602230484e-06,
+      "loss": 0.7494,
       "step": 85
     },
     {
-      "epoch": 0.004172461752433936,
-      "grad_norm": 6.341604709625244,
-      "learning_rate": 8.344923504867873e-07,
-      "loss": 1.3533,
+      "epoch": 0.03349148758024002,
+      "grad_norm": 0.0159482192248106,
+      "learning_rate": 6.691449814126395e-06,
+      "loss": 0.6999,
       "step": 90
     },
     {
-      "epoch": 0.00440426518312471,
-      "grad_norm": 6.311711311340332,
-      "learning_rate": 8.808530366249421e-07,
-      "loss": 1.2228,
+      "epoch": 0.03535212577914225,
+      "grad_norm": 0.016392122954130173,
+      "learning_rate": 7.063197026022306e-06,
+      "loss": 0.6966,
       "step": 95
     },
     {
-      "epoch": 0.004636068613815485,
-      "grad_norm": 6.693397045135498,
-      "learning_rate": 9.27213722763097e-07,
-      "loss": 1.2429,
+      "epoch": 0.03721276397804447,
+      "grad_norm": 0.017619600519537926,
+      "learning_rate": 7.434944237918216e-06,
+      "loss": 0.7659,
       "step": 100
     },
     {
-      "epoch": 0.004636068613815485,
-      "eval_loss": 1.2549220323562622,
-      "eval_runtime": 11.2838,
-      "eval_samples_per_second": 11.255,
-      "eval_steps_per_second": 11.255,
+      "epoch": 0.03721276397804447,
+      "eval_loss": 0.760661780834198,
+      "eval_runtime": 26.8914,
+      "eval_samples_per_second": 4.723,
+      "eval_steps_per_second": 4.723,
       "step": 100
     },
     {
-      "epoch": 0.004867872044506259,
-      "grad_norm": 7.291402339935303,
-      "learning_rate": 9.735744089012517e-07,
-      "loss": 1.3052,
+      "epoch": 0.039073402176946694,
+      "grad_norm": 0.02020377479493618,
+      "learning_rate": 7.806691449814127e-06,
+      "loss": 0.7475,
       "step": 105
     },
     {
-      "epoch": 0.005099675475197033,
-      "grad_norm": 6.058638095855713,
-      "learning_rate": 1.0199350950394066e-06,
-      "loss": 1.1783,
+      "epoch": 0.04093404037584892,
+      "grad_norm": 0.018723690882325172,
+      "learning_rate": 8.178438661710038e-06,
+      "loss": 0.7233,
       "step": 110
     },
     {
-      "epoch": 0.005331478905887807,
-      "grad_norm": 6.480159282684326,
-      "learning_rate": 1.0662957811775615e-06,
-      "loss": 1.2306,
+      "epoch": 0.04279467857475114,
+      "grad_norm": 0.017194446176290512,
+      "learning_rate": 8.550185873605949e-06,
+      "loss": 0.7655,
       "step": 115
     },
     {
-      "epoch": 0.005563282336578581,
-      "grad_norm": 6.4310221672058105,
-      "learning_rate": 1.1126564673157164e-06,
-      "loss": 1.0381,
+      "epoch": 0.044655316773653364,
+      "grad_norm": 0.013910962268710136,
+      "learning_rate": 8.921933085501859e-06,
+      "loss": 0.7493,
       "step": 120
     },
     {
-      "epoch": 0.005795085767269356,
-      "grad_norm": 5.661340236663818,
-      "learning_rate": 1.1590171534538713e-06,
-      "loss": 1.1524,
+      "epoch": 0.04651595497255559,
+      "grad_norm": 0.01600288413465023,
+      "learning_rate": 9.29368029739777e-06,
+      "loss": 0.7528,
       "step": 125
     },
     {
-      "epoch": 0.00602688919796013,
-      "grad_norm": 6.941490173339844,
-      "learning_rate": 1.2053778395920262e-06,
-      "loss": 1.2525,
+      "epoch": 0.04837659317145781,
+      "grad_norm": 0.01554879080504179,
+      "learning_rate": 9.66542750929368e-06,
+      "loss": 0.6947,
       "step": 130
     },
     {
-      "epoch": 0.006258692628650904,
-      "grad_norm": 8.481853485107422,
-      "learning_rate": 1.2517385257301808e-06,
-      "loss": 1.2434,
+      "epoch": 0.050237231370360035,
+      "grad_norm": 0.015109645202755928,
+      "learning_rate": 1.0037174721189591e-05,
+      "loss": 0.6647,
       "step": 135
     },
     {
-      "epoch": 0.006490496059341678,
-      "grad_norm": 6.217038154602051,
-      "learning_rate": 1.2980992118683357e-06,
-      "loss": 1.3233,
+      "epoch": 0.05209786956926226,
+      "grad_norm": 0.015620779246091843,
+      "learning_rate": 1.0408921933085503e-05,
+      "loss": 0.7298,
       "step": 140
     },
     {
-      "epoch": 0.006722299490032452,
-      "grad_norm": 5.551589012145996,
-      "learning_rate": 1.3444598980064908e-06,
-      "loss": 1.1833,
+      "epoch": 0.05395850776816448,
+      "grad_norm": 0.013970567844808102,
+      "learning_rate": 1.0780669144981412e-05,
+      "loss": 0.7485,
       "step": 145
     },
     {
-      "epoch": 0.006954102920723227,
-      "grad_norm": 5.535802364349365,
-      "learning_rate": 1.3908205841446453e-06,
-      "loss": 1.2354,
+      "epoch": 0.055819145967066705,
+      "grad_norm": 0.014776401221752167,
+      "learning_rate": 1.1152416356877324e-05,
+      "loss": 0.6767,
       "step": 150
     },
     {
-      "epoch": 0.007185906351414001,
-      "grad_norm": 6.194558143615723,
-      "learning_rate": 1.4371812702828004e-06,
-      "loss": 1.2508,
+      "epoch": 0.057679784165968924,
+      "grad_norm": 0.014368101954460144,
+      "learning_rate": 1.1524163568773235e-05,
+      "loss": 0.7037,
       "step": 155
     },
     {
-      "epoch": 0.007417709782104775,
-      "grad_norm": 6.735708236694336,
-      "learning_rate": 1.4835419564209553e-06,
-      "loss": 1.2612,
+      "epoch": 0.05954042236487115,
+      "grad_norm": 0.01449244562536478,
+      "learning_rate": 1.1895910780669147e-05,
+      "loss": 0.7305,
       "step": 160
     },
     {
-      "epoch": 0.0076495132127955496,
-      "grad_norm": 6.2094855308532715,
-      "learning_rate": 1.52990264255911e-06,
-      "loss": 1.2216,
+      "epoch": 0.061401060563773376,
+      "grad_norm": 0.015350298024713993,
+      "learning_rate": 1.2267657992565056e-05,
+      "loss": 0.6804,
       "step": 165
     },
     {
-      "epoch": 0.007881316643486324,
-      "grad_norm": 5.640394687652588,
-      "learning_rate": 1.5762633286972648e-06,
-      "loss": 0.9614,
+      "epoch": 0.0632616987626756,
+      "grad_norm": 0.014395203441381454,
+      "learning_rate": 1.2639405204460968e-05,
+      "loss": 0.6832,
       "step": 170
     },
     {
-      "epoch": 0.008113120074177098,
-      "grad_norm": 6.062097549438477,
-      "learning_rate": 1.6226240148354197e-06,
-      "loss": 1.0738,
+      "epoch": 0.06512233696157782,
+      "grad_norm": 0.015541068278253078,
+      "learning_rate": 1.3011152416356879e-05,
+      "loss": 0.7168,
       "step": 175
     },
     {
-      "epoch": 0.008344923504867872,
-      "grad_norm": 7.769379615783691,
-      "learning_rate": 1.6689847009735746e-06,
-      "loss": 1.1195,
+      "epoch": 0.06698297516048005,
+      "grad_norm": 0.012579885311424732,
+      "learning_rate": 1.338289962825279e-05,
+      "loss": 0.693,
       "step": 180
     },
     {
-      "epoch": 0.008576726935558646,
-      "grad_norm": 6.438906192779541,
-      "learning_rate": 1.7153453871117293e-06,
-      "loss": 1.3437,
+      "epoch": 0.06884361335938227,
+      "grad_norm": 0.014168789610266685,
+      "learning_rate": 1.37546468401487e-05,
+      "loss": 0.7056,
       "step": 185
     },
     {
-      "epoch": 0.00880853036624942,
-      "grad_norm": 5.559536933898926,
-      "learning_rate": 1.7617060732498842e-06,
-      "loss": 1.2253,
+      "epoch": 0.0707042515582845,
+      "grad_norm": 0.012087655253708363,
+      "learning_rate": 1.4126394052044612e-05,
+      "loss": 0.7006,
       "step": 190
     },
     {
-      "epoch": 0.009040333796940195,
-      "grad_norm": 6.249836444854736,
-      "learning_rate": 1.808066759388039e-06,
-      "loss": 1.1692,
+      "epoch": 0.07256488975718671,
+      "grad_norm": 0.014433121308684349,
+      "learning_rate": 1.4498141263940521e-05,
+      "loss": 0.7016,
       "step": 195
     },
     {
-      "epoch": 0.00927213722763097,
-      "grad_norm": 5.227753162384033,
-      "learning_rate": 1.854427445526194e-06,
-      "loss": 1.0196,
+      "epoch": 0.07442552795608894,
+      "grad_norm": 0.015074139460921288,
+      "learning_rate": 1.4869888475836432e-05,
+      "loss": 0.6912,
       "step": 200
     },
     {
-      "epoch": 0.00927213722763097,
-      "eval_loss": 1.1532349586486816,
-      "eval_runtime": 11.1843,
-      "eval_samples_per_second": 11.355,
-      "eval_steps_per_second": 11.355,
+      "epoch": 0.07442552795608894,
+      "eval_loss": 0.7027233242988586,
+      "eval_runtime": 26.5611,
+      "eval_samples_per_second": 4.781,
+      "eval_steps_per_second": 4.781,
       "step": 200
     },
     {
-      "epoch": 0.009503940658321743,
-      "grad_norm": 4.921353340148926,
-      "learning_rate": 1.9007881316643486e-06,
-      "loss": 0.9627,
+      "epoch": 0.07628616615499116,
+      "grad_norm": 0.013617471791803837,
+      "learning_rate": 1.5241635687732344e-05,
+      "loss": 0.7139,
       "step": 205
     },
     {
-      "epoch": 0.009735744089012517,
-      "grad_norm": 8.843475341796875,
-      "learning_rate": 1.9471488178025035e-06,
-      "loss": 1.1093,
+      "epoch": 0.07814680435389339,
+      "grad_norm": 0.013274065218865871,
+      "learning_rate": 1.5613382899628255e-05,
+      "loss": 0.6955,
       "step": 210
     },
     {
-      "epoch": 0.009967547519703291,
-      "grad_norm": 5.638999938964844,
-      "learning_rate": 1.9935095039406584e-06,
-      "loss": 1.1378,
+      "epoch": 0.08000744255279561,
+      "grad_norm": 0.013781987130641937,
+      "learning_rate": 1.5985130111524165e-05,
+      "loss": 0.65,
       "step": 215
     },
     {
-      "epoch": 0.010199350950394067,
-      "grad_norm": 6.697763919830322,
-      "learning_rate": 2.0398701900788133e-06,
-      "loss": 1.1164,
+      "epoch": 0.08186808075169784,
+      "grad_norm": 0.01373015996068716,
+      "learning_rate": 1.6356877323420076e-05,
+      "loss": 0.6681,
       "step": 220
     },
     {
-      "epoch": 0.01043115438108484,
-      "grad_norm": 5.590152263641357,
-      "learning_rate": 2.086230876216968e-06,
-      "loss": 1.1577,
+      "epoch": 0.08372871895060005,
+      "grad_norm": 0.01403126772493124,
+      "learning_rate": 1.6728624535315986e-05,
+      "loss": 0.6981,
       "step": 225
     },
     {
-      "epoch": 0.010662957811775614,
-      "grad_norm": 8.595314979553223,
-      "learning_rate": 2.132591562355123e-06,
-      "loss": 0.9921,
+      "epoch": 0.08558935714950228,
+      "grad_norm": 0.01332685723900795,
+      "learning_rate": 1.7100371747211897e-05,
+      "loss": 0.6517,
       "step": 230
     },
     {
-      "epoch": 0.010894761242466388,
-      "grad_norm": 6.406239986419678,
-      "learning_rate": 2.178952248493278e-06,
-      "loss": 1.0457,
+      "epoch": 0.0874499953484045,
+      "grad_norm": 0.013317782431840897,
+      "learning_rate": 1.7472118959107808e-05,
+      "loss": 0.6835,
       "step": 235
     },
     {
-      "epoch": 0.011126564673157162,
-      "grad_norm": 6.516043663024902,
-      "learning_rate": 2.225312934631433e-06,
-      "loss": 1.3394,
+      "epoch": 0.08931063354730673,
+      "grad_norm": 0.015096917748451233,
+      "learning_rate": 1.7843866171003718e-05,
+      "loss": 0.7138,
       "step": 240
     },
     {
-      "epoch": 0.011358368103847936,
-      "grad_norm": 8.184934616088867,
-      "learning_rate": 2.2716736207695877e-06,
-      "loss": 1.1182,
+      "epoch": 0.09117127174620895,
+      "grad_norm": 0.013502717949450016,
+      "learning_rate": 1.8215613382899632e-05,
+      "loss": 0.6722,
       "step": 245
     },
     {
-      "epoch": 0.011590171534538712,
-      "grad_norm": 6.349449157714844,
-      "learning_rate": 2.3180343069077426e-06,
-      "loss": 1.0817,
+      "epoch": 0.09303190994511118,
+      "grad_norm": 0.01279054582118988,
+      "learning_rate": 1.858736059479554e-05,
+      "loss": 0.6323,
       "step": 250
     },
     {
-      "epoch": 0.011821974965229486,
-      "grad_norm": 8.29964828491211,
-      "learning_rate": 2.3643949930458975e-06,
-      "loss": 0.9925,
+      "epoch": 0.09489254814401339,
+      "grad_norm": 0.011777005158364773,
+      "learning_rate": 1.8959107806691453e-05,
+      "loss": 0.6332,
       "step": 255
     },
     {
-      "epoch": 0.01205377839592026,
-      "grad_norm": 4.632545471191406,
-      "learning_rate": 2.4107556791840523e-06,
-      "loss": 0.9848,
+      "epoch": 0.09675318634291562,
+      "grad_norm": 0.01320689544081688,
+      "learning_rate": 1.933085501858736e-05,
+      "loss": 0.7034,
       "step": 260
     },
     {
-      "epoch": 0.012285581826611033,
-      "grad_norm": 9.396188735961914,
-      "learning_rate": 2.457116365322207e-06,
-      "loss": 1.1873,
+      "epoch": 0.09861382454181784,
+      "grad_norm": 0.012140162289142609,
+      "learning_rate": 1.970260223048327e-05,
+      "loss": 0.6154,
       "step": 265
     },
     {
-      "epoch": 0.012517385257301807,
-      "grad_norm": 9.827582359313965,
-      "learning_rate": 2.5034770514603617e-06,
-      "loss": 1.219,
+      "epoch": 0.10047446274072007,
+      "grad_norm": 0.01772845722734928,
+      "learning_rate": 1.9999991559715313e-05,
+      "loss": 0.7119,
       "step": 270
     },
     {
-      "epoch": 0.012749188687992583,
-      "grad_norm": 6.420851230621338,
-      "learning_rate": 2.5498377375985166e-06,
-      "loss": 1.1635,
+      "epoch": 0.1023351009396223,
+      "grad_norm": 0.012948377057909966,
+      "learning_rate": 1.999969615124717e-05,
+      "loss": 0.6554,
       "step": 275
     },
     {
-      "epoch": 0.012980992118683357,
-      "grad_norm": 9.211454391479492,
-      "learning_rate": 2.5961984237366715e-06,
-      "loss": 1.1646,
+      "epoch": 0.10419573913852452,
+      "grad_norm": 0.013341420330107212,
+      "learning_rate": 1.9998978742792098e-05,
+      "loss": 0.636,
       "step": 280
     },
     {
-      "epoch": 0.01321279554937413,
-      "grad_norm": 5.8738322257995605,
-      "learning_rate": 2.6425591098748268e-06,
-      "loss": 1.0619,
+      "epoch": 0.10605637733742673,
+      "grad_norm": 0.012272336520254612,
+      "learning_rate": 1.999783936462566e-05,
+      "loss": 0.6182,
       "step": 285
     },
     {
-      "epoch": 0.013444598980064905,
-      "grad_norm": 5.047887325286865,
-      "learning_rate": 2.6889197960129816e-06,
-      "loss": 0.8132,
+      "epoch": 0.10791701553632896,
+      "grad_norm": 0.012329615652561188,
+      "learning_rate": 1.999627806483107e-05,
+      "loss": 0.601,
       "step": 290
     },
     {
-      "epoch": 0.013676402410755678,
-      "grad_norm": 6.454254150390625,
-      "learning_rate": 2.7352804821511357e-06,
-      "loss": 1.0644,
+      "epoch": 0.10977765373523118,
+      "grad_norm": 0.01313008088618517,
+      "learning_rate": 1.999429490929718e-05,
+      "loss": 0.7002,
       "step": 295
     },
     {
-      "epoch": 0.013908205841446454,
-      "grad_norm": 12.117624282836914,
-      "learning_rate": 2.7816411682892906e-06,
-      "loss": 1.1196,
+      "epoch": 0.11163829193413341,
+      "grad_norm": 0.01170238945633173,
+      "learning_rate": 1.9991889981715696e-05,
+      "loss": 0.6784,
       "step": 300
     },
     {
-      "epoch": 0.013908205841446454,
-      "eval_loss": 1.1029967069625854,
-      "eval_runtime": 11.2775,
-      "eval_samples_per_second": 11.261,
-      "eval_steps_per_second": 11.261,
+      "epoch": 0.11163829193413341,
+      "eval_loss": 0.676296055316925,
+      "eval_runtime": 26.7211,
+      "eval_samples_per_second": 4.753,
+      "eval_steps_per_second": 4.753,
       "step": 300
     },
     {
-      "epoch": 0.014140009272137228,
-      "grad_norm": 4.559245586395264,
-      "learning_rate": 2.828001854427446e-06,
-      "loss": 1.0169,
+      "epoch": 0.11349893013303564,
+      "grad_norm": 0.012631393037736416,
+      "learning_rate": 1.9989063383577644e-05,
+      "loss": 0.6332,
       "step": 305
     },
     {
-      "epoch": 0.014371812702828002,
-      "grad_norm": 6.905066013336182,
-      "learning_rate": 2.8743625405656008e-06,
-      "loss": 1.1846,
+      "epoch": 0.11535956833193785,
+      "grad_norm": 0.012392008677124977,
+      "learning_rate": 1.998581523416908e-05,
+      "loss": 0.713,
       "step": 310
     },
     {
-      "epoch": 0.014603616133518776,
-      "grad_norm": 5.575958251953125,
-      "learning_rate": 2.9207232267037557e-06,
-      "loss": 1.0439,
+      "epoch": 0.11722020653084007,
+      "grad_norm": 0.012342042289674282,
+      "learning_rate": 1.998214567056607e-05,
+      "loss": 0.6072,
       "step": 315
     },
     {
-      "epoch": 0.01483541956420955,
-      "grad_norm": 7.027660369873047,
-      "learning_rate": 2.9670839128419105e-06,
-      "loss": 1.2515,
+      "epoch": 0.1190808447297423,
+      "grad_norm": 0.011801215820014477,
+      "learning_rate": 1.9978054847628908e-05,
+      "loss": 0.6553,
       "step": 320
     },
     {
-      "epoch": 0.015067222994900325,
-      "grad_norm": 5.862566947937012,
-      "learning_rate": 3.0134445989800654e-06,
-      "loss": 1.067,
+      "epoch": 0.12094148292864453,
+      "grad_norm": 0.013543626293540001,
+      "learning_rate": 1.997354293799555e-05,
+      "loss": 0.638,
       "step": 325
     },
     {
-      "epoch": 0.015299026425591099,
-      "grad_norm": 5.865589618682861,
-      "learning_rate": 3.05980528511822e-06,
-      "loss": 1.1834,
+      "epoch": 0.12280212112754675,
+      "grad_norm": 0.012367943301796913,
+      "learning_rate": 1.9968610132074372e-05,
+      "loss": 0.6867,
       "step": 330
     },
     {
-      "epoch": 0.015530829856281873,
-      "grad_norm": 6.330904960632324,
-      "learning_rate": 3.1061659712563748e-06,
-      "loss": 1.2675,
+      "epoch": 0.12466275932644898,
+      "grad_norm": 0.01382706593722105,
+      "learning_rate": 1.99632566380361e-05,
+      "loss": 0.67,
       "step": 335
     },
     {
-      "epoch": 0.01576263328697265,
-      "grad_norm": 6.37352180480957,
-      "learning_rate": 3.1525266573945297e-06,
-      "loss": 1.0757,
+      "epoch": 0.1265233975253512,
+      "grad_norm": 0.01336714904755354,
+      "learning_rate": 1.9957482681805036e-05,
+      "loss": 0.6615,
       "step": 340
     },
     {
-      "epoch": 0.015994436717663423,
-      "grad_norm": 6.554245948791504,
-      "learning_rate": 3.1988873435326845e-06,
-      "loss": 1.0759,
+      "epoch": 0.12838403572425341,
+      "grad_norm": 0.012786686420440674,
+      "learning_rate": 1.9951288507049532e-05,
+      "loss": 0.6343,
       "step": 345
     },
     {
-      "epoch": 0.016226240148354196,
-      "grad_norm": 5.791826248168945,
-      "learning_rate": 3.2452480296708394e-06,
-      "loss": 1.0212,
+      "epoch": 0.13024467392315564,
+      "grad_norm": 0.014750728383660316,
+      "learning_rate": 1.9944674375171697e-05,
+      "loss": 0.6478,
       "step": 350
     },
     {
-      "epoch": 0.01645804357904497,
-      "grad_norm": 5.290635108947754,
-      "learning_rate": 3.2916087158089943e-06,
-      "loss": 0.9677,
+      "epoch": 0.13210531212205787,
+      "grad_norm": 0.01330367662012577,
+      "learning_rate": 1.9937640565296372e-05,
+      "loss": 0.6844,
       "step": 355
     },
     {
-      "epoch": 0.016689847009735744,
-      "grad_norm": 6.6352739334106445,
-      "learning_rate": 3.337969401947149e-06,
-      "loss": 1.1298,
+      "epoch": 0.1339659503209601,
+      "grad_norm": 0.011332959868013859,
+      "learning_rate": 1.9930187374259338e-05,
+      "loss": 0.6188,
       "step": 360
     },
     {
-      "epoch": 0.016921650440426518,
-      "grad_norm": 5.054739952087402,
-      "learning_rate": 3.384330088085304e-06,
-      "loss": 1.1997,
+      "epoch": 0.13582658851986232,
+      "grad_norm": 0.013507647439837456,
+      "learning_rate": 1.992231511659481e-05,
+      "loss": 0.6844,
       "step": 365
     },
     {
-      "epoch": 0.017153453871117292,
-      "grad_norm": 6.105016231536865,
-      "learning_rate": 3.4306907742234585e-06,
-      "loss": 1.0421,
+      "epoch": 0.13768722671876454,
+      "grad_norm": 0.01494019664824009,
+      "learning_rate": 1.991402412452214e-05,
+      "loss": 0.6616,
       "step": 370
     },
     {
-      "epoch": 0.017385257301808066,
-      "grad_norm": 7.15664529800415,
-      "learning_rate": 3.4770514603616134e-06,
-      "loss": 1.0932,
+      "epoch": 0.13954786491766677,
+      "grad_norm": 0.012762832455337048,
+      "learning_rate": 1.9905314747931816e-05,
+      "loss": 0.6797,
       "step": 375
     },
     {
-      "epoch": 0.01761706073249884,
-      "grad_norm": 5.2151103019714355,
-      "learning_rate": 3.5234121464997683e-06,
-      "loss": 1.0914,
+      "epoch": 0.141408503116569,
+      "grad_norm": 0.014913683757185936,
+      "learning_rate": 1.989618735437069e-05,
+      "loss": 0.6268,
       "step": 380
     },
     {
-      "epoch": 0.017848864163189614,
-      "grad_norm": 4.892900466918945,
-      "learning_rate": 3.569772832637923e-06,
-      "loss": 0.8586,
+      "epoch": 0.1432691413154712,
+      "grad_norm": 0.012221734039485455,
+      "learning_rate": 1.9886642329026457e-05,
+      "loss": 0.6587,
       "step": 385
     },
     {
-      "epoch": 0.01808066759388039,
-      "grad_norm": 5.810879230499268,
-      "learning_rate": 3.616133518776078e-06,
-      "loss": 1.009,
+      "epoch": 0.14512977951437342,
+      "grad_norm": 0.01553855836391449,
+      "learning_rate": 1.9876680074711417e-05,
+      "loss": 0.6403,
       "step": 390
     },
     {
-      "epoch": 0.018312471024571165,
-      "grad_norm": 5.979573726654053,
-      "learning_rate": 3.662494204914233e-06,
-      "loss": 1.0977,
+      "epoch": 0.14699041771327565,
+      "grad_norm": 0.013682518154382706,
+      "learning_rate": 1.986630101184546e-05,
+      "loss": 0.6287,
       "step": 395
     },
     {
-      "epoch": 0.01854427445526194,
-      "grad_norm": 6.757267951965332,
-      "learning_rate": 3.708854891052388e-06,
-      "loss": 0.9805,
+      "epoch": 0.14885105591217787,
+      "grad_norm": 0.013481502421200275,
+      "learning_rate": 1.9855505578438343e-05,
+      "loss": 0.6757,
       "step": 400
     },
     {
-      "epoch": 0.01854427445526194,
-      "eval_loss": 1.0770819187164307,
-      "eval_runtime": 11.2853,
-      "eval_samples_per_second": 11.254,
-      "eval_steps_per_second": 11.254,
+      "epoch": 0.14885105591217787,
+      "eval_loss": 0.6647208333015442,
+      "eval_runtime": 27.129,
+      "eval_samples_per_second": 4.681,
+      "eval_steps_per_second": 4.681,
       "step": 400
     },
     {
-      "epoch": 0.018776077885952713,
-      "grad_norm": 7.137505531311035,
-      "learning_rate": 3.755215577190543e-06,
-      "loss": 1.0624,
+      "epoch": 0.1507116941110801,
+      "grad_norm": 0.01228385604918003,
+      "learning_rate": 1.984429423007117e-05,
+      "loss": 0.6277,
       "step": 405
     },
     {
-      "epoch": 0.019007881316643487,
-      "grad_norm": 6.788882255554199,
-      "learning_rate": 3.801576263328697e-06,
-      "loss": 1.1705,
+      "epoch": 0.15257233230998232,
+      "grad_norm": 0.014119746163487434,
+      "learning_rate": 1.9832667439877217e-05,
+      "loss": 0.615,
       "step": 410
     },
     {
-      "epoch": 0.01923968474733426,
-      "grad_norm": 5.606616020202637,
-      "learning_rate": 3.8479369494668525e-06,
-      "loss": 1.1273,
+      "epoch": 0.15443297050888455,
+      "grad_norm": 0.014395875856280327,
+      "learning_rate": 1.9820625698521918e-05,
+      "loss": 0.6417,
       "step": 415
     },
     {
-      "epoch": 0.019471488178025034,
-      "grad_norm": 6.507837772369385,
-      "learning_rate": 3.894297635605007e-06,
-      "loss": 1.1154,
+      "epoch": 0.15629360870778677,
+      "grad_norm": 0.013175971806049347,
+      "learning_rate": 1.9808169514182182e-05,
+      "loss": 0.6509,
       "step": 420
     },
     {
-      "epoch": 0.019703291608715808,
-      "grad_norm": 6.080861568450928,
-      "learning_rate": 3.940658321743162e-06,
-      "loss": 1.1084,
+      "epoch": 0.158154246906689,
+      "grad_norm": 0.015295376069843769,
+      "learning_rate": 1.9795299412524948e-05,
+      "loss": 0.6275,
       "step": 425
     },
     {
-      "epoch": 0.019935095039406582,
-      "grad_norm": 5.980709075927734,
-      "learning_rate": 3.987019007881317e-06,
-      "loss": 1.1777,
+      "epoch": 0.16001488510559123,
+      "grad_norm": 0.014611025340855122,
+      "learning_rate": 1.9782015936684987e-05,
+      "loss": 0.6627,
       "step": 430
     },
     {
-      "epoch": 0.020166898470097356,
-      "grad_norm": 5.779782772064209,
-      "learning_rate": 4.033379694019472e-06,
-      "loss": 1.0581,
+      "epoch": 0.16187552330449345,
+      "grad_norm": 0.01412207167595625,
+      "learning_rate": 1.9768319647242e-05,
+      "loss": 0.6362,
       "step": 435
     },
     {
-      "epoch": 0.020398701900788133,
-      "grad_norm": 6.758626461029053,
-      "learning_rate": 4.0797403801576265e-06,
-      "loss": 1.1615,
+      "epoch": 0.16373616150339568,
+      "grad_norm": 0.012070410884916782,
+      "learning_rate": 1.9754211122196945e-05,
+      "loss": 0.6429,
       "step": 440
     },
     {
-      "epoch": 0.020630505331478907,
-      "grad_norm": 5.383087635040283,
-      "learning_rate": 4.126101066295782e-06,
-      "loss": 0.9877,
+      "epoch": 0.16559679970229788,
+      "grad_norm": 0.013232079334557056,
+      "learning_rate": 1.9739690956947652e-05,
+      "loss": 0.6941,
       "step": 445
     },
     {
-      "epoch": 0.02086230876216968,
-      "grad_norm": 5.77476692199707,
-      "learning_rate": 4.172461752433936e-06,
-      "loss": 1.0031,
+      "epoch": 0.1674574379012001,
+      "grad_norm": 0.012606708332896233,
+      "learning_rate": 1.972475976426369e-05,
+      "loss": 0.6554,
       "step": 450
     },
     {
-      "epoch": 0.021094112192860455,
-      "grad_norm": 5.7638726234436035,
-      "learning_rate": 4.218822438572091e-06,
-      "loss": 0.9153,
+      "epoch": 0.16931807610010233,
+      "grad_norm": 0.012638423591852188,
+      "learning_rate": 1.9709418174260523e-05,
+      "loss": 0.645,
       "step": 455
     },
     {
-      "epoch": 0.02132591562355123,
-      "grad_norm": 6.627535820007324,
-      "learning_rate": 4.265183124710246e-06,
-      "loss": 1.1059,
+      "epoch": 0.17117871429900455,
+      "grad_norm": 0.013119902461767197,
+      "learning_rate": 1.9693666834372896e-05,
+      "loss": 0.6128,
       "step": 460
     },
     {
-      "epoch": 0.021557719054242003,
-      "grad_norm": 6.557712554931641,
-      "learning_rate": 4.3115438108484005e-06,
-      "loss": 1.2056,
+      "epoch": 0.17303935249790678,
+      "grad_norm": 0.011363113299012184,
+      "learning_rate": 1.9677506409327532e-05,
+      "loss": 0.6294,
       "step": 465
     },
     {
-      "epoch": 0.021789522484932777,
-      "grad_norm": 5.442972660064697,
-      "learning_rate": 4.357904496986556e-06,
-      "loss": 0.9902,
+      "epoch": 0.174899990696809,
+      "grad_norm": 0.014238959178328514,
+      "learning_rate": 1.9660937581115073e-05,
+      "loss": 0.6647,
       "step": 470
     },
     {
-      "epoch": 0.02202132591562355,
-      "grad_norm": 6.064192771911621,
-      "learning_rate": 4.404265183124711e-06,
-      "loss": 1.3211,
+      "epoch": 0.17676062889571123,
+      "grad_norm": 0.013214629143476486,
+      "learning_rate": 1.9643961048961283e-05,
+      "loss": 0.6037,
       "step": 475
     },
     {
-      "epoch": 0.022253129346314324,
-      "grad_norm": 11.156740188598633,
-      "learning_rate": 4.450625869262866e-06,
-      "loss": 1.091,
+      "epoch": 0.17862126709461346,
+      "grad_norm": 0.012312485836446285,
+      "learning_rate": 1.9626577529297573e-05,
+      "loss": 0.6703,
       "step": 480
     },
     {
-      "epoch": 0.0224849327770051,
-      "grad_norm": 7.064011096954346,
-      "learning_rate": 4.49698655540102e-06,
-      "loss": 1.0799,
+      "epoch": 0.18048190529351568,
+      "grad_norm": 0.012318914756178856,
+      "learning_rate": 1.9608787755730746e-05,
+      "loss": 0.6141,
       "step": 485
     },
     {
-      "epoch": 0.022716736207695872,
-      "grad_norm": 6.225513458251953,
-      "learning_rate": 4.543347241539175e-06,
-      "loss": 1.1523,
+      "epoch": 0.1823425434924179,
+      "grad_norm": 0.01374764647334814,
+      "learning_rate": 1.9590592479012022e-05,
+      "loss": 0.673,
       "step": 490
     },
     {
-      "epoch": 0.02294853963838665,
-      "grad_norm": 9.263200759887695,
-      "learning_rate": 4.58970792767733e-06,
-      "loss": 1.1255,
+      "epoch": 0.18420318169132013,
+      "grad_norm": 0.012634415179491043,
+      "learning_rate": 1.9571992467005395e-05,
+      "loss": 0.6135,
       "step": 495
     },
     {
-      "epoch": 0.023180343069077423,
-      "grad_norm": 5.4971699714660645,
-      "learning_rate": 4.636068613815485e-06,
-      "loss": 1.0001,
+      "epoch": 0.18606381989022236,
+      "grad_norm": 0.012813772074878216,
+      "learning_rate": 1.9552988504655194e-05,
+      "loss": 0.6648,
       "step": 500
     },
     {
-      "epoch": 0.023180343069077423,
-      "eval_loss": 1.0599620342254639,
-      "eval_runtime": 11.2721,
-      "eval_samples_per_second": 11.267,
-      "eval_steps_per_second": 11.267,
+      "epoch": 0.18606381989022236,
+      "eval_loss": 0.6582558751106262,
+      "eval_runtime": 26.9621,
+      "eval_samples_per_second": 4.71,
+      "eval_steps_per_second": 4.71,
       "step": 500
     },
     {
-      "epoch": 0.023412146499768197,
-      "grad_norm": 23.34360694885254,
-      "learning_rate": 4.68242929995364e-06,
-      "loss": 0.8703,
+      "epoch": 0.18792445808912456,
+      "grad_norm": 0.012235240079462528,
+      "learning_rate": 1.9533581393952978e-05,
+      "loss": 0.6108,
       "step": 505
     },
     {
-      "epoch": 0.02364394993045897,
-      "grad_norm": 6.617534637451172,
-      "learning_rate": 4.728789986091795e-06,
-      "loss": 1.0301,
+      "epoch": 0.18978509628802678,
+      "grad_norm": 0.012589952908456326,
+      "learning_rate": 1.951377195390367e-05,
+      "loss": 0.6218,
       "step": 510
     },
     {
-      "epoch": 0.023875753361149745,
-      "grad_norm": 6.7085771560668945,
-      "learning_rate": 4.775150672229949e-06,
-      "loss": 0.9515,
+      "epoch": 0.191645734486929,
+      "grad_norm": 0.012783786281943321,
+      "learning_rate": 1.9493561020491024e-05,
+      "loss": 0.6668,
       "step": 515
     },
     {
-      "epoch": 0.02410755679184052,
-      "grad_norm": 7.118703365325928,
-      "learning_rate": 4.821511358368105e-06,
-      "loss": 1.1126,
+      "epoch": 0.19350637268583123,
+      "grad_norm": 0.013828632421791553,
+      "learning_rate": 1.9472949446642318e-05,
+      "loss": 0.6081,
       "step": 520
     },
     {
-      "epoch": 0.024339360222531293,
-      "grad_norm": 5.29957914352417,
-      "learning_rate": 4.867872044506259e-06,
-      "loss": 0.9504,
+      "epoch": 0.19536701088473346,
+      "grad_norm": 0.011929171159863472,
+      "learning_rate": 1.945193810219237e-05,
+      "loss": 0.6329,
       "step": 525
     },
     {
-      "epoch": 0.024571163653222067,
-      "grad_norm": 5.3429436683654785,
-      "learning_rate": 4.914232730644414e-06,
-      "loss": 0.9477,
+      "epoch": 0.1972276490836357,
+      "grad_norm": 0.014584473334252834,
+      "learning_rate": 1.9430527873846826e-05,
+      "loss": 0.7017,
       "step": 530
     },
     {
-      "epoch": 0.02480296708391284,
-      "grad_norm": 5.861160755157471,
-      "learning_rate": 4.960593416782569e-06,
-      "loss": 0.9679,
+      "epoch": 0.1990882872825379,
+      "grad_norm": 0.01474926806986332,
+      "learning_rate": 1.9408719665144756e-05,
+      "loss": 0.632,
       "step": 535
     },
     {
-      "epoch": 0.025034770514603615,
-      "grad_norm": 5.540627956390381,
-      "learning_rate": 5.006954102920723e-06,
-      "loss": 1.049,
+      "epoch": 0.20094892548144014,
+      "grad_norm": 0.015552829019725323,
+      "learning_rate": 1.9386514396420503e-05,
+      "loss": 0.6757,
       "step": 540
     },
     {
-      "epoch": 0.025266573945294392,
-      "grad_norm": 5.6925950050354,
-      "learning_rate": 5.053314789058879e-06,
-      "loss": 1.08,
+      "epoch": 0.20280956368034236,
+      "grad_norm": 0.013232480734586716,
+      "learning_rate": 1.9363913004764847e-05,
+      "loss": 0.6722,
       "step": 545
     },
     {
-      "epoch": 0.025498377375985166,
-      "grad_norm": 5.670673847198486,
-      "learning_rate": 5.099675475197033e-06,
-      "loss": 0.993,
+      "epoch": 0.2046702018792446,
+      "grad_norm": 0.012858827598392963,
+      "learning_rate": 1.9340916443985465e-05,
+      "loss": 0.6231,
       "step": 550
     },
     {
-      "epoch": 0.02573018080667594,
-      "grad_norm": 6.216073513031006,
-      "learning_rate": 5.1460361613351884e-06,
-      "loss": 1.18,
+      "epoch": 0.20653084007814682,
+      "grad_norm": 0.012365566566586494,
+      "learning_rate": 1.9317525684566686e-05,
+      "loss": 0.5986,
       "step": 555
     },
     {
-      "epoch": 0.025961984237366714,
-      "grad_norm": 6.3086113929748535,
-      "learning_rate": 5.192396847473343e-06,
-      "loss": 0.9918,
+      "epoch": 0.20839147827704904,
+      "grad_norm": 0.01528852991759777,
+      "learning_rate": 1.9293741713628518e-05,
+      "loss": 0.6537,
       "step": 560
     },
     {
-      "epoch": 0.026193787668057487,
-      "grad_norm": 7.024346828460693,
-      "learning_rate": 5.238757533611498e-06,
-      "loss": 1.0553,
+      "epoch": 0.21025211647595124,
+      "grad_norm": 0.014512522146105766,
+      "learning_rate": 1.9269565534885003e-05,
+      "loss": 0.6527,
       "step": 565
     },
     {
-      "epoch": 0.02642559109874826,
-      "grad_norm": 4.924084186553955,
-      "learning_rate": 5.2851182197496535e-06,
-      "loss": 0.9521,
+      "epoch": 0.21211275467485347,
+      "grad_norm": 0.013798325322568417,
+      "learning_rate": 1.9244998168601848e-05,
+      "loss": 0.6148,
       "step": 570
     },
     {
-      "epoch": 0.026657394529439035,
-      "grad_norm": 5.455867290496826,
-      "learning_rate": 5.331478905887808e-06,
-      "loss": 0.875,
+      "epoch": 0.2139733928737557,
+      "grad_norm": 0.013186248019337654,
+      "learning_rate": 1.9220040651553388e-05,
+      "loss": 0.6106,
       "step": 575
     },
     {
-      "epoch": 0.02688919796012981,
-      "grad_norm": 4.741636276245117,
-      "learning_rate": 5.377839592025963e-06,
-      "loss": 0.9555,
+      "epoch": 0.21583403107265792,
+      "grad_norm": 0.013986771926283836,
+      "learning_rate": 1.9194694036978807e-05,
+      "loss": 0.6654,
       "step": 580
     },
     {
-      "epoch": 0.027121001390820583,
-      "grad_norm": 6.165698528289795,
-      "learning_rate": 5.424200278164117e-06,
-      "loss": 1.0817,
+      "epoch": 0.21769466927156014,
+      "grad_norm": 0.016201818361878395,
+      "learning_rate": 1.9168959394537708e-05,
+      "loss": 0.6306,
       "step": 585
     },
     {
-      "epoch": 0.027352804821511357,
-      "grad_norm": 5.713921070098877,
-      "learning_rate": 5.470560964302271e-06,
-      "loss": 0.9599,
+      "epoch": 0.21955530747046237,
+      "grad_norm": 0.013889294117689133,
+      "learning_rate": 1.9142837810264972e-05,
+      "loss": 0.6749,
       "step": 590
     },
     {
-      "epoch": 0.027584608252202134,
-      "grad_norm": 4.89811372756958,
-      "learning_rate": 5.516921650440427e-06,
-      "loss": 1.0632,
+      "epoch": 0.2214159456693646,
+      "grad_norm": 0.013025142252445221,
+      "learning_rate": 1.911633038652491e-05,
+      "loss": 0.6075,
       "step": 595
     },
     {
-      "epoch": 0.027816411682892908,
-      "grad_norm": 6.775177478790283,
-      "learning_rate": 5.563282336578581e-06,
-      "loss": 0.9467,
+      "epoch": 0.22327658386826682,
+      "grad_norm": 0.013716059736907482,
+      "learning_rate": 1.9089438241964764e-05,
+      "loss": 0.6516,
       "step": 600
     },
     {
-      "epoch": 0.027816411682892908,
-      "eval_loss": 1.0482524633407593,
-      "eval_runtime": 11.1015,
-      "eval_samples_per_second": 11.44,
-      "eval_steps_per_second": 11.44,
+      "epoch": 0.22327658386826682,
+      "eval_loss": 0.6547934412956238,
+      "eval_runtime": 26.7928,
+      "eval_samples_per_second": 4.74,
+      "eval_steps_per_second": 4.74,
       "step": 600
     },
     {
-      "epoch": 0.028048215113583682,
-      "grad_norm": 5.133644104003906,
-      "learning_rate": 5.6096430227167365e-06,
-      "loss": 1.0033,
+      "epoch": 0.22513722206716905,
+      "grad_norm": 0.016251811757683754,
+      "learning_rate": 1.906216251146748e-05,
+      "loss": 0.6265,
       "step": 605
     },
     {
-      "epoch": 0.028280018544274456,
-      "grad_norm": 6.307579040527344,
-      "learning_rate": 5.656003708854892e-06,
-      "loss": 1.0006,
+      "epoch": 0.22699786026607127,
+      "grad_norm": 0.013359563425183296,
+      "learning_rate": 1.9034504346103825e-05,
+      "loss": 0.6052,
       "step": 610
     },
     {
-      "epoch": 0.02851182197496523,
-      "grad_norm": 6.110098838806152,
-      "learning_rate": 5.702364394993046e-06,
-      "loss": 1.1813,
+      "epoch": 0.2288584984649735,
+      "grad_norm": 0.012794552370905876,
+      "learning_rate": 1.9006464913083807e-05,
+      "loss": 0.613,
       "step": 615
     },
     {
-      "epoch": 0.028743625405656004,
-      "grad_norm": 7.244579315185547,
-      "learning_rate": 5.7487250811312015e-06,
-      "loss": 1.1448,
+      "epoch": 0.2307191366638757,
+      "grad_norm": 0.012533072382211685,
+      "learning_rate": 1.897804539570742e-05,
+      "loss": 0.6735,
       "step": 620
     },
     {
-      "epoch": 0.028975428836346778,
-      "grad_norm": 5.6296305656433105,
-      "learning_rate": 5.795085767269356e-06,
-      "loss": 1.0722,
+      "epoch": 0.23257977486277792,
+      "grad_norm": 0.013286220841109753,
+      "learning_rate": 1.8949246993314694e-05,
+      "loss": 0.6692,
       "step": 625
     },
     {
-      "epoch": 0.02920723226703755,
-      "grad_norm": 5.689414978027344,
-      "learning_rate": 5.841446453407511e-06,
-      "loss": 1.0578,
+      "epoch": 0.23444041306168015,
+      "grad_norm": 0.013466808013617992,
+      "learning_rate": 1.892007092123511e-05,
+      "loss": 0.6513,
       "step": 630
     },
     {
-      "epoch": 0.029439035697728325,
-      "grad_norm": 7.251938343048096,
-      "learning_rate": 5.887807139545666e-06,
-      "loss": 0.9588,
+      "epoch": 0.23630105126058237,
+      "grad_norm": 0.012991335242986679,
+      "learning_rate": 1.8890518410736275e-05,
+      "loss": 0.6405,
       "step": 635
     },
     {
-      "epoch": 0.0296708391284191,
-      "grad_norm": 6.593122482299805,
-      "learning_rate": 5.934167825683821e-06,
-      "loss": 1.1287,
+      "epoch": 0.2381616894594846,
+      "grad_norm": 0.013223089277744293,
+      "learning_rate": 1.8860590708971997e-05,
+      "loss": 0.6488,
       "step": 640
     },
     {
-      "epoch": 0.029902642559109873,
-      "grad_norm": 5.741578102111816,
-      "learning_rate": 5.9805285118219755e-06,
-      "loss": 0.9135,
+      "epoch": 0.24002232765838682,
+      "grad_norm": 0.012394067831337452,
+      "learning_rate": 1.8830289078929618e-05,
+      "loss": 0.6131,
       "step": 645
     },
     {
-      "epoch": 0.03013444598980065,
-      "grad_norm": 7.281170845031738,
-      "learning_rate": 6.026889197960131e-06,
-      "loss": 0.9585,
+      "epoch": 0.24188296585728905,
+      "grad_norm": 0.013721502386033535,
+      "learning_rate": 1.8799614799376743e-05,
+      "loss": 0.681,
       "step": 650
     },
     {
-      "epoch": 0.030366249420491424,
-      "grad_norm": 5.263773441314697,
-      "learning_rate": 6.073249884098285e-06,
-      "loss": 1.0363,
+      "epoch": 0.24374360405619128,
+      "grad_norm": 0.012726777233183384,
+      "learning_rate": 1.8768569164807272e-05,
+      "loss": 0.6837,
       "step": 655
     },
     {
-      "epoch": 0.030598052851182198,
-      "grad_norm": 6.059519290924072,
-      "learning_rate": 6.11961057023644e-06,
-      "loss": 0.9297,
+      "epoch": 0.2456042422550935,
+      "grad_norm": 0.013720668852329254,
+      "learning_rate": 1.8737153485386737e-05,
+      "loss": 0.6007,
       "step": 660
     },
     {
-      "epoch": 0.030829856281872972,
-      "grad_norm": 5.903897762298584,
-      "learning_rate": 6.165971256374594e-06,
-      "loss": 0.9503,
+      "epoch": 0.24746488045399573,
+      "grad_norm": 0.012646087445318699,
+      "learning_rate": 1.8705369086897063e-05,
+      "loss": 0.6545,
       "step": 665
     },
     {
-      "epoch": 0.031061659712563746,
-      "grad_norm": 5.2086944580078125,
-      "learning_rate": 6.2123319425127495e-06,
-      "loss": 0.9917,
+      "epoch": 0.24932551865289795,
+      "grad_norm": 0.013658811338245869,
+      "learning_rate": 1.8673217310680578e-05,
+      "loss": 0.6379,
       "step": 670
     },
     {
-      "epoch": 0.03129346314325452,
-      "grad_norm": 4.655348777770996,
-      "learning_rate": 6.258692628650904e-06,
-      "loss": 0.9246,
+      "epoch": 0.2511861568518002,
+      "grad_norm": 0.012248256243765354,
+      "learning_rate": 1.864069951358342e-05,
+      "loss": 0.6356,
       "step": 675
     },
     {
-      "epoch": 0.0315252665739453,
-      "grad_norm": 6.055633544921875,
-      "learning_rate": 6.305053314789059e-06,
-      "loss": 0.9003,
+      "epoch": 0.2530467950507024,
+      "grad_norm": 0.0133894681930542,
+      "learning_rate": 1.860781706789829e-05,
+      "loss": 0.6737,
       "step": 680
     },
     {
-      "epoch": 0.03175707000463607,
-      "grad_norm": 5.528895378112793,
-      "learning_rate": 6.351414000927214e-06,
-      "loss": 1.1059,
+      "epoch": 0.25490743324960463,
+      "grad_norm": 0.01430124044418335,
+      "learning_rate": 1.857457136130651e-05,
+      "loss": 0.6169,
       "step": 685
     },
     {
-      "epoch": 0.031988873435326845,
-      "grad_norm": 5.239342212677002,
-      "learning_rate": 6.397774687065369e-06,
-      "loss": 1.0502,
+      "epoch": 0.25676807144850683,
+      "grad_norm": 0.013437042012810707,
+      "learning_rate": 1.854096379681949e-05,
+      "loss": 0.6021,
       "step": 690
     },
     {
-      "epoch": 0.032220676866017615,
-      "grad_norm": 6.7139763832092285,
-      "learning_rate": 6.4441353732035235e-06,
-      "loss": 1.0828,
+      "epoch": 0.2586287096474091,
+      "grad_norm": 0.011497768573462963,
+      "learning_rate": 1.8506995792719498e-05,
+      "loss": 0.6119,
       "step": 695
     },
     {
-      "epoch": 0.03245248029670839,
-      "grad_norm": 5.635594844818115,
-      "learning_rate": 6.490496059341679e-06,
-      "loss": 1.0016,
+      "epoch": 0.2604893478463113,
+      "grad_norm": 0.01319235097616911,
+      "learning_rate": 1.8472668782499817e-05,
+      "loss": 0.627,
       "step": 700
     },
     {
-      "epoch": 0.03245248029670839,
-      "eval_loss": 1.0429729223251343,
-      "eval_runtime": 11.2781,
-      "eval_samples_per_second": 11.261,
-      "eval_steps_per_second": 11.261,
+      "epoch": 0.2604893478463113,
+      "eval_loss": 0.6522720456123352,
+      "eval_runtime": 26.7201,
+      "eval_samples_per_second": 4.753,
+      "eval_steps_per_second": 4.753,
       "step": 700
     },
     {
-      "epoch": 0.03268428372739916,
-      "grad_norm": 5.803088665008545,
-      "learning_rate": 6.536856745479834e-06,
-      "loss": 1.1353,
+      "epoch": 0.26234998604521353,
+      "grad_norm": 0.014044429175555706,
+      "learning_rate": 1.843798421480426e-05,
+      "loss": 0.6244,
       "step": 705
     },
     {
-      "epoch": 0.03291608715808994,
-      "grad_norm": 7.0808305740356445,
-      "learning_rate": 6.583217431617989e-06,
-      "loss": 0.814,
+      "epoch": 0.26421062424411573,
+      "grad_norm": 0.011311609297990799,
+      "learning_rate": 1.8402943553365998e-05,
+      "loss": 0.5975,
       "step": 710
     },
     {
-      "epoch": 0.03314789058878071,
-      "grad_norm": 5.607944011688232,
-      "learning_rate": 6.629578117756144e-06,
-      "loss": 1.0398,
+      "epoch": 0.26607126244301793,
+      "grad_norm": 0.012355692684650421,
+      "learning_rate": 1.8367548276945846e-05,
+      "loss": 0.6009,
       "step": 715
     },
     {
-      "epoch": 0.03337969401947149,
-      "grad_norm": 4.771095275878906,
-      "learning_rate": 6.675938803894298e-06,
-      "loss": 0.975,
+      "epoch": 0.2679319006419202,
+      "grad_norm": 0.012264437042176723,
+      "learning_rate": 1.83317998792698e-05,
+      "loss": 0.6103,
       "step": 720
     },
     {
-      "epoch": 0.033611497450162266,
-      "grad_norm": 6.333477020263672,
-      "learning_rate": 6.722299490032454e-06,
-      "loss": 0.8296,
+      "epoch": 0.2697925388408224,
+      "grad_norm": 0.012037084437906742,
+      "learning_rate": 1.8295699868966038e-05,
+      "loss": 0.5602,
       "step": 725
     },
     {
-      "epoch": 0.033843300880853036,
-      "grad_norm": 5.489449501037598,
-      "learning_rate": 6.768660176170608e-06,
-      "loss": 1.0049,
+      "epoch": 0.27165317703972464,
+      "grad_norm": 0.012773050926625729,
+      "learning_rate": 1.8259249769501237e-05,
+      "loss": 0.6215,
       "step": 730
     },
     {
-      "epoch": 0.034075104311543813,
-      "grad_norm": 6.153010845184326,
-      "learning_rate": 6.8150208623087635e-06,
-      "loss": 0.921,
+      "epoch": 0.27351381523862683,
+      "grad_norm": 0.012415550649166107,
+      "learning_rate": 1.8222451119116288e-05,
+      "loss": 0.6364,
       "step": 735
     },
     {
-      "epoch": 0.034306907742234584,
-      "grad_norm": 5.430294036865234,
-      "learning_rate": 6.861381548446917e-06,
-      "loss": 0.9989,
+      "epoch": 0.2753744534375291,
+      "grad_norm": 0.01302468404173851,
+      "learning_rate": 1.8185305470761366e-05,
+      "loss": 0.5994,
       "step": 740
     },
     {
-      "epoch": 0.03453871117292536,
-      "grad_norm": 5.40978479385376,
-      "learning_rate": 6.907742234585072e-06,
-      "loss": 0.9716,
+      "epoch": 0.2772350916364313,
+      "grad_norm": 0.013447316363453865,
+      "learning_rate": 1.814781439203043e-05,
+      "loss": 0.6458,
       "step": 745
     },
     {
-      "epoch": 0.03477051460361613,
-      "grad_norm": 7.3993754386901855,
-      "learning_rate": 6.954102920723227e-06,
-      "loss": 0.9539,
+      "epoch": 0.27909572983533354,
+      "grad_norm": 0.012098093517124653,
+      "learning_rate": 1.8109979465095014e-05,
+      "loss": 0.6357,
       "step": 750
     },
     {
-      "epoch": 0.03500231803430691,
-      "grad_norm": 6.530559539794922,
-      "learning_rate": 7.000463606861382e-06,
-      "loss": 0.8866,
+      "epoch": 0.28095636803423574,
+      "grad_norm": 0.012987499125301838,
+      "learning_rate": 1.8071802286637505e-05,
+      "loss": 0.6248,
       "step": 755
     },
     {
-      "epoch": 0.03523412146499768,
-      "grad_norm": 5.703915596008301,
-      "learning_rate": 7.046824292999537e-06,
-      "loss": 0.9307,
+      "epoch": 0.282817006233138,
+      "grad_norm": 0.011747024022042751,
+      "learning_rate": 1.8033284467783742e-05,
+      "loss": 0.6202,
       "step": 760
     },
     {
-      "epoch": 0.03546592489568846,
-      "grad_norm": 5.743514060974121,
-      "learning_rate": 7.093184979137692e-06,
-      "loss": 0.9864,
+      "epoch": 0.2846776444320402,
+      "grad_norm": 0.01332057174295187,
+      "learning_rate": 1.7994427634035016e-05,
+      "loss": 0.6347,
       "step": 765
     },
     {
-      "epoch": 0.03569772832637923,
-      "grad_norm": 5.572900295257568,
-      "learning_rate": 7.139545665275846e-06,
-      "loss": 0.9106,
+      "epoch": 0.2865382826309424,
+      "grad_norm": 0.013383504003286362,
+      "learning_rate": 1.795523342519948e-05,
+      "loss": 0.6001,
       "step": 770
     },
     {
-      "epoch": 0.035929531757070005,
-      "grad_norm": 6.122991561889648,
-      "learning_rate": 7.185906351414002e-06,
-      "loss": 1.051,
+      "epoch": 0.28839892082984464,
+      "grad_norm": 0.013648821040987968,
+      "learning_rate": 1.7915703495322967e-05,
+      "loss": 0.6399,
       "step": 775
     },
     {
-      "epoch": 0.03616133518776078,
-      "grad_norm": 6.111961364746094,
-      "learning_rate": 7.232267037552156e-06,
-      "loss": 1.1246,
+      "epoch": 0.29025955902874684,
+      "grad_norm": 0.012947522103786469,
+      "learning_rate": 1.7875839512619148e-05,
+      "loss": 0.6298,
       "step": 780
     },
     {
-      "epoch": 0.03639313861845155,
-      "grad_norm": 6.494778633117676,
-      "learning_rate": 7.2786277236903115e-06,
-      "loss": 1.057,
+      "epoch": 0.2921201972276491,
+      "grad_norm": 0.01334394421428442,
+      "learning_rate": 1.7835643159399156e-05,
+      "loss": 0.6418,
       "step": 785
     },
     {
-      "epoch": 0.03662494204914233,
-      "grad_norm": 5.405693054199219,
-      "learning_rate": 7.324988409828466e-06,
-      "loss": 1.0847,
+      "epoch": 0.2939808354265513,
+      "grad_norm": 0.014045110903680325,
+      "learning_rate": 1.7795116132000587e-05,
+      "loss": 0.6403,
       "step": 790
     },
     {
-      "epoch": 0.0368567454798331,
-      "grad_norm": 5.207508087158203,
-      "learning_rate": 7.371349095966621e-06,
-      "loss": 1.0974,
+      "epoch": 0.29584147362545354,
+      "grad_norm": 0.015219368040561676,
+      "learning_rate": 1.7754260140715918e-05,
+      "loss": 0.6277,
       "step": 795
     },
     {
-      "epoch": 0.03708854891052388,
-      "grad_norm": 6.345035076141357,
-      "learning_rate": 7.417709782104776e-06,
-      "loss": 1.0643,
+      "epoch": 0.29770211182435574,
+      "grad_norm": 0.01307649165391922,
+      "learning_rate": 1.771307690972031e-05,
+      "loss": 0.6271,
       "step": 800
     },
     {
-      "epoch": 0.03708854891052388,
-      "eval_loss": 1.040871024131775,
-      "eval_runtime": 11.2912,
-      "eval_samples_per_second": 11.248,
-      "eval_steps_per_second": 11.248,
+      "epoch": 0.29770211182435574,
+      "eval_loss": 0.6502260565757751,
+      "eval_runtime": 26.7042,
+      "eval_samples_per_second": 4.756,
+      "eval_steps_per_second": 4.756,
       "step": 800
     },
     {
-      "epoch": 0.03732035234121465,
-      "grad_norm": 5.8186492919921875,
-      "learning_rate": 7.464070468242931e-06,
-      "loss": 1.0505,
+      "epoch": 0.299562750023258,
+      "grad_norm": 0.013835963793098927,
+      "learning_rate": 1.7671568176998865e-05,
+      "loss": 0.6286,
       "step": 805
     },
     {
-      "epoch": 0.037552155771905425,
-      "grad_norm": 6.559715747833252,
-      "learning_rate": 7.510431154381086e-06,
-      "loss": 0.9892,
+      "epoch": 0.3014233882221602,
+      "grad_norm": 0.013574733398854733,
+      "learning_rate": 1.762973569427328e-05,
+      "loss": 0.6462,
       "step": 810
     },
     {
-      "epoch": 0.037783959202596196,
-      "grad_norm": 5.379362106323242,
-      "learning_rate": 7.55679184051924e-06,
-      "loss": 1.1113,
+      "epoch": 0.30328402642106245,
+      "grad_norm": 0.01131366565823555,
+      "learning_rate": 1.758758122692791e-05,
+      "loss": 0.6167,
       "step": 815
     },
     {
-      "epoch": 0.03801576263328697,
-      "grad_norm": 5.991211414337158,
-      "learning_rate": 7.603152526657394e-06,
-      "loss": 1.0382,
+      "epoch": 0.30514466461996465,
+      "grad_norm": 0.013482702895998955,
+      "learning_rate": 1.7545106553935277e-05,
+      "loss": 0.6413,
       "step": 820
     },
     {
-      "epoch": 0.03824756606397774,
-      "grad_norm": 10.996150016784668,
-      "learning_rate": 7.64951321279555e-06,
-      "loss": 0.963,
+      "epoch": 0.3070053028188669,
+      "grad_norm": 0.012861824594438076,
+      "learning_rate": 1.7502313467780988e-05,
+      "loss": 0.6027,
       "step": 825
     },
     {
-      "epoch": 0.03847936949466852,
-      "grad_norm": 6.428567409515381,
-      "learning_rate": 7.695873898933705e-06,
-      "loss": 1.0694,
+      "epoch": 0.3088659410177691,
+      "grad_norm": 0.012556380592286587,
+      "learning_rate": 1.7459203774388097e-05,
+      "loss": 0.6603,
       "step": 830
     },
     {
-      "epoch": 0.0387111729253593,
-      "grad_norm": 6.335524082183838,
-      "learning_rate": 7.742234585071859e-06,
-      "loss": 1.15,
+      "epoch": 0.3107265792166713,
+      "grad_norm": 0.012278062291443348,
+      "learning_rate": 1.7415779293040887e-05,
+      "loss": 0.5803,
       "step": 835
     },
     {
-      "epoch": 0.03894297635605007,
-      "grad_norm": 6.979407787322998,
-      "learning_rate": 7.788595271210014e-06,
-      "loss": 1.0347,
+      "epoch": 0.31258721741557355,
+      "grad_norm": 0.012213567271828651,
+      "learning_rate": 1.7372041856308098e-05,
+      "loss": 0.6624,
       "step": 840
     },
     {
-      "epoch": 0.039174779786740846,
-      "grad_norm": 6.700367450714111,
-      "learning_rate": 7.83495595734817e-06,
-      "loss": 0.9949,
+      "epoch": 0.31444785561447575,
+      "grad_norm": 0.0131307952105999,
+      "learning_rate": 1.7327993309965583e-05,
+      "loss": 0.6447,
       "step": 845
     },
     {
-      "epoch": 0.039406583217431616,
-      "grad_norm": 5.5337605476379395,
-      "learning_rate": 7.881316643486325e-06,
-      "loss": 0.9346,
+      "epoch": 0.316308493813378,
+      "grad_norm": 0.0121999466791749,
+      "learning_rate": 1.7283635512918423e-05,
+      "loss": 0.6451,
       "step": 850
     },
     {
-      "epoch": 0.039638386648122394,
-      "grad_norm": 6.536638259887695,
-      "learning_rate": 7.92767732962448e-06,
-      "loss": 0.8486,
+      "epoch": 0.3181691320122802,
+      "grad_norm": 0.012872702442109585,
+      "learning_rate": 1.7238970337122484e-05,
+      "loss": 0.5724,
       "step": 855
     },
     {
-      "epoch": 0.039870190078813164,
-      "grad_norm": 5.657130241394043,
-      "learning_rate": 7.974038015762633e-06,
-      "loss": 0.9584,
+      "epoch": 0.32002977021118245,
+      "grad_norm": 0.013137550093233585,
+      "learning_rate": 1.7193999667505387e-05,
+      "loss": 0.6459,
       "step": 860
     },
     {
-      "epoch": 0.04010199350950394,
-      "grad_norm": 5.246485233306885,
-      "learning_rate": 8.020398701900789e-06,
-      "loss": 0.9707,
+      "epoch": 0.32189040841008465,
+      "grad_norm": 0.013948196545243263,
+      "learning_rate": 1.7148725401887002e-05,
+      "loss": 0.651,
       "step": 865
     },
     {
-      "epoch": 0.04033379694019471,
-      "grad_norm": 8.451269149780273,
-      "learning_rate": 8.066759388038944e-06,
-      "loss": 0.9778,
+      "epoch": 0.3237510466089869,
+      "grad_norm": 0.012517811730504036,
+      "learning_rate": 1.710314945089933e-05,
+      "loss": 0.6114,
       "step": 870
     },
     {
-      "epoch": 0.04056560037088549,
-      "grad_norm": 5.41778564453125,
-      "learning_rate": 8.1131200741771e-06,
-      "loss": 1.0868,
+      "epoch": 0.3256116848078891,
+      "grad_norm": 0.014199101366102695,
+      "learning_rate": 1.7057273737905887e-05,
+      "loss": 0.6405,
       "step": 875
     },
     {
-      "epoch": 0.04079740380157627,
-      "grad_norm": 5.614444255828857,
-      "learning_rate": 8.159480760315253e-06,
-      "loss": 1.0862,
+      "epoch": 0.32747232300679135,
+      "grad_norm": 0.012591714970767498,
+      "learning_rate": 1.7011100198920528e-05,
+      "loss": 0.6767,
       "step": 880
     },
     {
-      "epoch": 0.04102920723226704,
-      "grad_norm": 5.727239608764648,
-      "learning_rate": 8.205841446453408e-06,
-      "loss": 1.028,
+      "epoch": 0.32933296120569355,
+      "grad_norm": 0.012114683166146278,
+      "learning_rate": 1.6964630782525743e-05,
+      "loss": 0.6037,
       "step": 885
     },
     {
-      "epoch": 0.041261010662957814,
-      "grad_norm": 6.3700056076049805,
-      "learning_rate": 8.252202132591564e-06,
-      "loss": 1.1344,
+      "epoch": 0.33119359940459575,
+      "grad_norm": 0.013139299117028713,
+      "learning_rate": 1.6917867449790432e-05,
+      "loss": 0.643,
       "step": 890
     },
     {
-      "epoch": 0.041492814093648585,
-      "grad_norm": 5.90902042388916,
-      "learning_rate": 8.298562818729717e-06,
-      "loss": 1.0694,
+      "epoch": 0.333054237603498,
+      "grad_norm": 0.012391779571771622,
+      "learning_rate": 1.6870812174187136e-05,
+      "loss": 0.647,
       "step": 895
     },
     {
-      "epoch": 0.04172461752433936,
-      "grad_norm": 7.204585075378418,
-      "learning_rate": 8.344923504867873e-06,
-      "loss": 1.1645,
+      "epoch": 0.3349148758024002,
+      "grad_norm": 0.013360656797885895,
+      "learning_rate": 1.6823466941508762e-05,
+      "loss": 0.725,
       "step": 900
     },
     {
-      "epoch": 0.04172461752433936,
-      "eval_loss": 1.035333275794983,
-      "eval_runtime": 11.2783,
-      "eval_samples_per_second": 11.261,
-      "eval_steps_per_second": 11.261,
+      "epoch": 0.3349148758024002,
+      "eval_loss": 0.6488396525382996,
+      "eval_runtime": 27.3774,
+      "eval_samples_per_second": 4.639,
+      "eval_steps_per_second": 4.639,
       "step": 900
     },
     {
-      "epoch": 0.04195642095503013,
-      "grad_norm": 7.7777018547058105,
-      "learning_rate": 8.391284191006028e-06,
-      "loss": 1.1797,
+      "epoch": 0.33677551400130246,
+      "grad_norm": 0.013424807228147984,
+      "learning_rate": 1.677583374978478e-05,
+      "loss": 0.6342,
       "step": 905
     },
     {
-      "epoch": 0.04218822438572091,
-      "grad_norm": 6.48301362991333,
-      "learning_rate": 8.437644877144181e-06,
-      "loss": 0.9009,
+      "epoch": 0.33863615220020465,
+      "grad_norm": 0.01573541946709156,
+      "learning_rate": 1.6727914609196895e-05,
+      "loss": 0.6562,
       "step": 910
     },
     {
-      "epoch": 0.04242002781641168,
-      "grad_norm": 5.617315769195557,
-      "learning_rate": 8.484005563282337e-06,
-      "loss": 0.9737,
+      "epoch": 0.3404967903991069,
+      "grad_norm": 0.011363287456333637,
+      "learning_rate": 1.6679711541994227e-05,
+      "loss": 0.6492,
       "step": 915
     },
     {
-      "epoch": 0.04265183124710246,
-      "grad_norm": 5.954869747161865,
-      "learning_rate": 8.530366249420492e-06,
-      "loss": 0.9795,
+      "epoch": 0.3423574285980091,
+      "grad_norm": 0.012154373340308666,
+      "learning_rate": 1.6631226582407954e-05,
+      "loss": 0.602,
       "step": 920
     },
     {
-      "epoch": 0.04288363467779323,
-      "grad_norm": 5.577536582946777,
-      "learning_rate": 8.576726935558647e-06,
-      "loss": 1.1313,
+      "epoch": 0.34421806679691136,
+      "grad_norm": 0.01744014024734497,
+      "learning_rate": 1.658246177656548e-05,
+      "loss": 0.6318,
       "step": 925
     },
     {
-      "epoch": 0.043115438108484005,
-      "grad_norm": 5.356098175048828,
-      "learning_rate": 8.623087621696801e-06,
-      "loss": 0.9889,
+      "epoch": 0.34607870499581356,
+      "grad_norm": 0.012465902604162693,
+      "learning_rate": 1.6533419182404078e-05,
+      "loss": 0.6522,
       "step": 930
     },
     {
-      "epoch": 0.04334724153917478,
-      "grad_norm": 6.9520087242126465,
-      "learning_rate": 8.669448307834956e-06,
-      "loss": 1.1183,
+      "epoch": 0.3479393431947158,
+      "grad_norm": 0.015055039897561073,
+      "learning_rate": 1.6484100869584044e-05,
+      "loss": 0.6376,
       "step": 935
     },
     {
-      "epoch": 0.04357904496986555,
-      "grad_norm": 5.181753635406494,
-      "learning_rate": 8.715808993973112e-06,
-      "loss": 0.9116,
+      "epoch": 0.349799981393618,
+      "grad_norm": 0.012351201847195625,
+      "learning_rate": 1.6434508919401357e-05,
+      "loss": 0.6206,
       "step": 940
     },
     {
-      "epoch": 0.04381084840055633,
-      "grad_norm": 6.152126312255859,
-      "learning_rate": 8.762169680111267e-06,
-      "loss": 1.0747,
+      "epoch": 0.3516606195925202,
+      "grad_norm": 0.012793191708624363,
+      "learning_rate": 1.6384645424699835e-05,
+      "loss": 0.6182,
       "step": 945
     },
     {
-      "epoch": 0.0440426518312471,
-      "grad_norm": 6.403101921081543,
-      "learning_rate": 8.808530366249422e-06,
-      "loss": 0.9084,
+      "epoch": 0.35352125779142246,
+      "grad_norm": 0.012946651317179203,
+      "learning_rate": 1.6334512489782833e-05,
+      "loss": 0.5839,
       "step": 950
     },
     {
-      "epoch": 0.04427445526193788,
-      "grad_norm": 5.676062107086182,
-      "learning_rate": 8.854891052387576e-06,
-      "loss": 0.8152,
+      "epoch": 0.35538189599032466,
+      "grad_norm": 0.012998082675039768,
+      "learning_rate": 1.628411223032442e-05,
+      "loss": 0.6517,
       "step": 955
     },
     {
-      "epoch": 0.04450625869262865,
-      "grad_norm": 5.6627936363220215,
-      "learning_rate": 8.901251738525731e-06,
-      "loss": 1.0508,
+      "epoch": 0.3572425341892269,
+      "grad_norm": 0.012614963576197624,
+      "learning_rate": 1.6233446773280113e-05,
+      "loss": 0.6235,
       "step": 960
     },
     {
-      "epoch": 0.044738062123319426,
-      "grad_norm": 5.71261739730835,
-      "learning_rate": 8.947612424663886e-06,
-      "loss": 1.0253,
+      "epoch": 0.3591031723881291,
+      "grad_norm": 0.012318151071667671,
+      "learning_rate": 1.6182518256797095e-05,
+      "loss": 0.664,
       "step": 965
     },
     {
-      "epoch": 0.0449698655540102,
-      "grad_norm": 7.736510753631592,
-      "learning_rate": 8.99397311080204e-06,
-      "loss": 1.0148,
+      "epoch": 0.36096381058703136,
+      "grad_norm": 0.012551162391901016,
+      "learning_rate": 1.6131328830123997e-05,
+      "loss": 0.6317,
       "step": 970
     },
     {
-      "epoch": 0.045201668984700974,
-      "grad_norm": 6.378748416900635,
-      "learning_rate": 9.040333796940195e-06,
-      "loss": 1.0871,
+      "epoch": 0.36282444878593356,
+      "grad_norm": 0.013372802175581455,
+      "learning_rate": 1.60798806535202e-05,
+      "loss": 0.6418,
       "step": 975
     },
     {
-      "epoch": 0.045433472415391744,
-      "grad_norm": 6.469996452331543,
-      "learning_rate": 9.08669448307835e-06,
-      "loss": 0.9835,
+      "epoch": 0.3646850869848358,
+      "grad_norm": 0.011675640940666199,
+      "learning_rate": 1.6028175898164665e-05,
+      "loss": 0.6118,
       "step": 980
     },
     {
-      "epoch": 0.04566527584608252,
-      "grad_norm": 5.87582540512085,
-      "learning_rate": 9.133055169216504e-06,
-      "loss": 0.8729,
+      "epoch": 0.366545725183738,
+      "grad_norm": 0.013295911252498627,
+      "learning_rate": 1.5976216746064294e-05,
+      "loss": 0.6217,
       "step": 985
     },
     {
-      "epoch": 0.0458970792767733,
-      "grad_norm": 6.7339606285095215,
-      "learning_rate": 9.17941585535466e-06,
-      "loss": 1.0141,
+      "epoch": 0.36840636338264027,
+      "grad_norm": 0.012895721010863781,
+      "learning_rate": 1.5924005389961866e-05,
+      "loss": 0.6436,
       "step": 990
     },
     {
-      "epoch": 0.04612888270746407,
-      "grad_norm": 6.447714328765869,
-      "learning_rate": 9.225776541492815e-06,
-      "loss": 0.8944,
+      "epoch": 0.37026700158154247,
+      "grad_norm": 0.014572090469300747,
+      "learning_rate": 1.5871544033243488e-05,
+      "loss": 0.6342,
       "step": 995
     },
     {
-      "epoch": 0.04636068613815485,
-      "grad_norm": 5.313723564147949,
-      "learning_rate": 9.27213722763097e-06,
-      "loss": 0.926,
+      "epoch": 0.3721276397804447,
+      "grad_norm": 0.012393898330628872,
+      "learning_rate": 1.581883488984562e-05,
+      "loss": 0.6218,
       "step": 1000
     },
     {
-      "epoch": 0.04636068613815485,
-      "eval_loss": 1.0359811782836914,
-      "eval_runtime": 11.2916,
-      "eval_samples_per_second": 11.247,
-      "eval_steps_per_second": 11.247,
+      "epoch": 0.3721276397804447,
+      "eval_loss": 0.6476932168006897,
+      "eval_runtime": 27.1488,
+      "eval_samples_per_second": 4.678,
+      "eval_steps_per_second": 4.678,
       "step": 1000
     },
     {
-      "epoch": 0.04659248956884562,
-      "grad_norm": 5.054633140563965,
-      "learning_rate": 9.318497913769124e-06,
-      "loss": 0.9767,
+      "epoch": 0.3739882779793469,
+      "grad_norm": 0.014692210592329502,
+      "learning_rate": 1.5765880184161625e-05,
+      "loss": 0.6216,
       "step": 1005
     },
     {
-      "epoch": 0.046824292999536395,
-      "grad_norm": 5.397629737854004,
-      "learning_rate": 9.36485859990728e-06,
-      "loss": 1.0489,
+      "epoch": 0.3758489161782491,
+      "grad_norm": 0.012152746319770813,
+      "learning_rate": 1.5712682150947926e-05,
+      "loss": 0.6243,
       "step": 1010
     },
     {
-      "epoch": 0.047056096430227165,
-      "grad_norm": 4.538703441619873,
-      "learning_rate": 9.411219286045434e-06,
-      "loss": 0.854,
+      "epoch": 0.37770955437715137,
+      "grad_norm": 0.012929155491292477,
+      "learning_rate": 1.5659243035229657e-05,
+      "loss": 0.6493,
       "step": 1015
     },
     {
-      "epoch": 0.04728789986091794,
-      "grad_norm": 5.5513529777526855,
-      "learning_rate": 9.45757997218359e-06,
-      "loss": 0.9744,
+      "epoch": 0.37957019257605357,
+      "grad_norm": 0.0136475944891572,
+      "learning_rate": 1.5605565092205973e-05,
+      "loss": 0.6506,
       "step": 1020
     },
     {
-      "epoch": 0.04751970329160871,
-      "grad_norm": 5.287125110626221,
-      "learning_rate": 9.503940658321743e-06,
-      "loss": 0.993,
+      "epoch": 0.3814308307749558,
+      "grad_norm": 0.014008302241563797,
+      "learning_rate": 1.5551650587154815e-05,
+      "loss": 0.6429,
       "step": 1025
     },
     {
-      "epoch": 0.04775150672229949,
-      "grad_norm": 5.260387897491455,
-      "learning_rate": 9.550301344459899e-06,
-      "loss": 0.9027,
+      "epoch": 0.383291468973858,
+      "grad_norm": 0.014000017195940018,
+      "learning_rate": 1.5497501795337366e-05,
+      "loss": 0.6277,
       "step": 1030
     },
     {
-      "epoch": 0.04798331015299027,
-      "grad_norm": 7.19069242477417,
-      "learning_rate": 9.596662030598054e-06,
-      "loss": 0.783,
+      "epoch": 0.38515210717276027,
+      "grad_norm": 0.012146887369453907,
+      "learning_rate": 1.5443121001901994e-05,
+      "loss": 0.635,
       "step": 1035
     },
     {
-      "epoch": 0.04821511358368104,
-      "grad_norm": 5.297065734863281,
-      "learning_rate": 9.64302271673621e-06,
-      "loss": 0.9357,
+      "epoch": 0.38701274537166247,
+      "grad_norm": 0.013878699392080307,
+      "learning_rate": 1.5388510501787855e-05,
+      "loss": 0.6416,
       "step": 1040
     },
     {
-      "epoch": 0.048446917014371815,
-      "grad_norm": 5.547972202301025,
-      "learning_rate": 9.689383402874365e-06,
-      "loss": 1.1059,
+      "epoch": 0.3888733835705647,
+      "grad_norm": 0.011823480948805809,
+      "learning_rate": 1.5333672599628005e-05,
+      "loss": 0.637,
       "step": 1045
     },
     {
-      "epoch": 0.048678720445062586,
-      "grad_norm": 5.573083400726318,
-      "learning_rate": 9.735744089012518e-06,
-      "loss": 1.0051,
+      "epoch": 0.3907340217694669,
+      "grad_norm": 0.012524113990366459,
+      "learning_rate": 1.527860960965216e-05,
+      "loss": 0.6763,
       "step": 1050
     },
     {
-      "epoch": 0.04891052387575336,
-      "grad_norm": 6.3478546142578125,
-      "learning_rate": 9.782104775150672e-06,
-      "loss": 1.2446,
+      "epoch": 0.3925946599683692,
+      "grad_norm": 0.013192784041166306,
+      "learning_rate": 1.5223323855589027e-05,
+      "loss": 0.6501,
       "step": 1055
     },
     {
-      "epoch": 0.049142327306444133,
-      "grad_norm": 5.411722183227539,
-      "learning_rate": 9.828465461288827e-06,
-      "loss": 0.9051,
+      "epoch": 0.3944552981672714,
+      "grad_norm": 0.012439992278814316,
+      "learning_rate": 1.5167817670568253e-05,
+      "loss": 0.5886,
       "step": 1060
     },
     {
-      "epoch": 0.04937413073713491,
-      "grad_norm": 5.237289905548096,
-      "learning_rate": 9.874826147426983e-06,
-      "loss": 1.1089,
+      "epoch": 0.39631593636617357,
+      "grad_norm": 0.013605669140815735,
+      "learning_rate": 1.5112093397021945e-05,
+      "loss": 0.5925,
       "step": 1065
     },
     {
-      "epoch": 0.04960593416782568,
-      "grad_norm": 6.6244354248046875,
-      "learning_rate": 9.921186833565138e-06,
-      "loss": 1.0403,
+      "epoch": 0.3981765745650758,
+      "grad_norm": 0.01384530495852232,
+      "learning_rate": 1.5056153386585828e-05,
+      "loss": 0.6607,
       "step": 1070
     },
     {
-      "epoch": 0.04983773759851646,
-      "grad_norm": 5.800417423248291,
-      "learning_rate": 9.967547519703291e-06,
-      "loss": 1.2312,
+      "epoch": 0.400037212763978,
+      "grad_norm": 0.014060786925256252,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.6458,
       "step": 1075
     },
     {
-      "epoch": 0.05006954102920723,
-      "grad_norm": 5.320492744445801,
-      "learning_rate": 1.0013908205841447e-05,
-      "loss": 0.8337,
+      "epoch": 0.4018978509628803,
+      "grad_norm": 0.012853951193392277,
+      "learning_rate": 1.494363560700931e-05,
+      "loss": 0.6028,
       "step": 1080
     },
     {
-      "epoch": 0.050301344459898006,
-      "grad_norm": 5.400928974151611,
-      "learning_rate": 1.0060268891979604e-05,
-      "loss": 0.884,
+      "epoch": 0.4037584891617825,
+      "grad_norm": 0.012846381403505802,
+      "learning_rate": 1.4887062586263334e-05,
+      "loss": 0.6543,
       "step": 1085
     },
     {
-      "epoch": 0.050533147890588784,
-      "grad_norm": 5.387668132781982,
-      "learning_rate": 1.0106629578117757e-05,
-      "loss": 0.9599,
+      "epoch": 0.40561912736068473,
+      "grad_norm": 0.013001542538404465,
+      "learning_rate": 1.4830283325216026e-05,
+      "loss": 0.5654,
       "step": 1090
     },
     {
-      "epoch": 0.050764951321279554,
-      "grad_norm": 5.861815929412842,
-      "learning_rate": 1.0152990264255911e-05,
-      "loss": 0.9738,
+      "epoch": 0.4074797655595869,
+      "grad_norm": 0.012333175167441368,
+      "learning_rate": 1.477330022002493e-05,
+      "loss": 0.6465,
       "step": 1095
     },
     {
-      "epoch": 0.05099675475197033,
-      "grad_norm": 6.088605880737305,
-      "learning_rate": 1.0199350950394066e-05,
-      "loss": 1.1038,
+      "epoch": 0.4093404037584892,
+      "grad_norm": 0.01342159602791071,
+      "learning_rate": 1.4716115675450078e-05,
+      "loss": 0.6168,
       "step": 1100
     },
     {
-      "epoch": 0.05099675475197033,
-      "eval_loss": 1.0365773439407349,
-      "eval_runtime": 11.2883,
-      "eval_samples_per_second": 11.251,
-      "eval_steps_per_second": 11.251,
+      "epoch": 0.4093404037584892,
+      "eval_loss": 0.6468775272369385,
+      "eval_runtime": 26.8263,
+      "eval_samples_per_second": 4.734,
+      "eval_steps_per_second": 4.734,
       "step": 1100
     },
     {
-      "epoch": 0.0512285581826611,
-      "grad_norm": 5.065865516662598,
-      "learning_rate": 1.024571163653222e-05,
-      "loss": 0.8697,
+      "epoch": 0.4112010419573914,
+      "grad_norm": 0.0139292748644948,
+      "learning_rate": 1.4658732104752507e-05,
+      "loss": 0.634,
       "step": 1105
     },
     {
-      "epoch": 0.05146036161335188,
-      "grad_norm": 6.684823513031006,
-      "learning_rate": 1.0292072322670377e-05,
-      "loss": 1.1078,
+      "epoch": 0.41306168015629363,
+      "grad_norm": 0.013169731944799423,
+      "learning_rate": 1.4601151929592403e-05,
+      "loss": 0.6227,
       "step": 1110
     },
     {
-      "epoch": 0.05169216504404265,
-      "grad_norm": 5.413392066955566,
-      "learning_rate": 1.033843300880853e-05,
-      "loss": 0.9671,
+      "epoch": 0.41492231835519583,
+      "grad_norm": 0.013705245219171047,
+      "learning_rate": 1.4543377579926915e-05,
+      "loss": 0.6441,
       "step": 1115
     },
     {
-      "epoch": 0.05192396847473343,
-      "grad_norm": 5.247165679931641,
-      "learning_rate": 1.0384793694946686e-05,
-      "loss": 0.8268,
+      "epoch": 0.4167829565540981,
+      "grad_norm": 0.013035726733505726,
+      "learning_rate": 1.4485411493907617e-05,
+      "loss": 0.6498,
       "step": 1120
     },
     {
-      "epoch": 0.0521557719054242,
-      "grad_norm": 5.909581661224365,
-      "learning_rate": 1.0431154381084841e-05,
-      "loss": 1.0279,
+      "epoch": 0.4186435947530003,
+      "grad_norm": 0.01190096139907837,
+      "learning_rate": 1.442725611777758e-05,
+      "loss": 0.6285,
       "step": 1125
     },
     {
-      "epoch": 0.052387575336114975,
-      "grad_norm": 5.5670366287231445,
-      "learning_rate": 1.0477515067222996e-05,
-      "loss": 1.0836,
+      "epoch": 0.4205042329519025,
+      "grad_norm": 0.013753347098827362,
+      "learning_rate": 1.4368913905768178e-05,
+      "loss": 0.6541,
       "step": 1130
     },
     {
-      "epoch": 0.052619378766805745,
-      "grad_norm": 6.892451763153076,
-      "learning_rate": 1.052387575336115e-05,
-      "loss": 0.8755,
+      "epoch": 0.42236487115080473,
+      "grad_norm": 0.012330746278166771,
+      "learning_rate": 1.4310387319995492e-05,
+      "loss": 0.6721,
       "step": 1135
     },
     {
-      "epoch": 0.05285118219749652,
-      "grad_norm": 7.612436294555664,
-      "learning_rate": 1.0570236439499307e-05,
-      "loss": 1.0173,
+      "epoch": 0.42422550934970693,
+      "grad_norm": 0.01278294064104557,
+      "learning_rate": 1.4251678830356408e-05,
+      "loss": 0.6589,
       "step": 1140
     },
     {
-      "epoch": 0.0530829856281873,
-      "grad_norm": 5.493641376495361,
-      "learning_rate": 1.061659712563746e-05,
-      "loss": 1.0388,
+      "epoch": 0.4260861475486092,
+      "grad_norm": 0.012772184796631336,
+      "learning_rate": 1.41927909144244e-05,
+      "loss": 0.6411,
       "step": 1145
     },
     {
-      "epoch": 0.05331478905887807,
-      "grad_norm": 5.2113471031188965,
-      "learning_rate": 1.0662957811775616e-05,
-      "loss": 0.9213,
+      "epoch": 0.4279467857475114,
+      "grad_norm": 0.012047790922224522,
+      "learning_rate": 1.413372605734495e-05,
+      "loss": 0.5759,
       "step": 1150
     },
     {
-      "epoch": 0.05354659248956885,
-      "grad_norm": 6.035247802734375,
-      "learning_rate": 1.070931849791377e-05,
-      "loss": 0.9501,
+      "epoch": 0.42980742394641364,
+      "grad_norm": 0.014543715864419937,
+      "learning_rate": 1.4074486751730687e-05,
+      "loss": 0.6578,
       "step": 1155
     },
     {
-      "epoch": 0.05377839592025962,
-      "grad_norm": 6.905342102050781,
-      "learning_rate": 1.0755679184051927e-05,
-      "loss": 1.0902,
+      "epoch": 0.43166806214531583,
+      "grad_norm": 0.013436605222523212,
+      "learning_rate": 1.4015075497556193e-05,
+      "loss": 0.5876,
       "step": 1160
     },
     {
-      "epoch": 0.054010199350950396,
-      "grad_norm": 4.778216361999512,
-      "learning_rate": 1.080203987019008e-05,
-      "loss": 1.0163,
+      "epoch": 0.4335287003442181,
+      "grad_norm": 0.011584432795643806,
+      "learning_rate": 1.3955494802052498e-05,
+      "loss": 0.656,
       "step": 1165
     },
     {
-      "epoch": 0.054242002781641166,
-      "grad_norm": 5.389718055725098,
-      "learning_rate": 1.0848400556328234e-05,
-      "loss": 0.9602,
+      "epoch": 0.4353893385431203,
+      "grad_norm": 0.012196795083582401,
+      "learning_rate": 1.3895747179601275e-05,
+      "loss": 0.6562,
       "step": 1170
     },
     {
-      "epoch": 0.05447380621233194,
-      "grad_norm": 5.472878932952881,
-      "learning_rate": 1.0894761242466389e-05,
-      "loss": 1.1119,
+      "epoch": 0.43724997674202254,
+      "grad_norm": 0.011762428097426891,
+      "learning_rate": 1.3835835151628728e-05,
+      "loss": 0.5918,
       "step": 1175
     },
     {
-      "epoch": 0.054705609643022714,
-      "grad_norm": 6.158339500427246,
-      "learning_rate": 1.0941121928604543e-05,
-      "loss": 1.152,
+      "epoch": 0.43911061494092474,
+      "grad_norm": 0.014100808650255203,
+      "learning_rate": 1.3775761246499177e-05,
+      "loss": 0.6216,
       "step": 1180
     },
     {
-      "epoch": 0.05493741307371349,
-      "grad_norm": 6.443728923797607,
-      "learning_rate": 1.09874826147427e-05,
-      "loss": 1.0942,
+      "epoch": 0.44097125313982694,
+      "grad_norm": 0.011724433861672878,
+      "learning_rate": 1.3715527999408376e-05,
+      "loss": 0.6434,
       "step": 1185
     },
     {
-      "epoch": 0.05516921650440427,
-      "grad_norm": 6.741264343261719,
-      "learning_rate": 1.1033843300880853e-05,
-      "loss": 1.0458,
+      "epoch": 0.4428318913387292,
+      "grad_norm": 0.013045977801084518,
+      "learning_rate": 1.365513795227651e-05,
+      "loss": 0.5915,
       "step": 1190
     },
     {
-      "epoch": 0.05540101993509504,
-      "grad_norm": 5.246371269226074,
-      "learning_rate": 1.1080203987019009e-05,
-      "loss": 0.849,
+      "epoch": 0.4446925295376314,
+      "grad_norm": 0.013339078053832054,
+      "learning_rate": 1.359459365364092e-05,
+      "loss": 0.6148,
       "step": 1195
     },
     {
-      "epoch": 0.055632823365785816,
-      "grad_norm": 7.0242509841918945,
-      "learning_rate": 1.1126564673157162e-05,
-      "loss": 1.0244,
+      "epoch": 0.44655316773653364,
+      "grad_norm": 0.011851554736495018,
+      "learning_rate": 1.3533897658548571e-05,
+      "loss": 0.6294,
       "step": 1200
     },
     {
-      "epoch": 0.055632823365785816,
-      "eval_loss": 1.035953164100647,
-      "eval_runtime": 11.2796,
-      "eval_samples_per_second": 11.259,
-      "eval_steps_per_second": 11.259,
+      "epoch": 0.44655316773653364,
+      "eval_loss": 0.6459712982177734,
+      "eval_runtime": 27.1858,
+      "eval_samples_per_second": 4.672,
+      "eval_steps_per_second": 4.672,
       "step": 1200
     },
     {
-      "epoch": 0.05586462679647659,
-      "grad_norm": 5.66998815536499,
-      "learning_rate": 1.117292535929532e-05,
-      "loss": 0.9911,
+      "epoch": 0.44841380593543584,
+      "grad_norm": 0.014516811817884445,
+      "learning_rate": 1.3473052528448203e-05,
+      "loss": 0.6052,
       "step": 1205
     },
     {
-      "epoch": 0.056096430227167364,
-      "grad_norm": 6.682615756988525,
-      "learning_rate": 1.1219286045433473e-05,
-      "loss": 0.9718,
+      "epoch": 0.4502744441343381,
+      "grad_norm": 0.013740907423198223,
+      "learning_rate": 1.341206083108225e-05,
+      "loss": 0.6035,
       "step": 1210
     },
     {
-      "epoch": 0.056328233657858134,
-      "grad_norm": 6.9051513671875,
-      "learning_rate": 1.1265646731571628e-05,
-      "loss": 1.1582,
+      "epoch": 0.4521350823332403,
+      "grad_norm": 0.013699905015528202,
+      "learning_rate": 1.3350925140378465e-05,
+      "loss": 0.64,
       "step": 1215
     },
     {
-      "epoch": 0.05656003708854891,
-      "grad_norm": 5.607714653015137,
-      "learning_rate": 1.1312007417709784e-05,
-      "loss": 0.9506,
+      "epoch": 0.45399572053214254,
+      "grad_norm": 0.013219136744737625,
+      "learning_rate": 1.328964803634131e-05,
+      "loss": 0.6102,
       "step": 1220
     },
     {
-      "epoch": 0.05679184051923968,
-      "grad_norm": 6.658040523529053,
-      "learning_rate": 1.1358368103847939e-05,
-      "loss": 1.1766,
+      "epoch": 0.45585635873104474,
+      "grad_norm": 0.012603058479726315,
+      "learning_rate": 1.3228232104943073e-05,
+      "loss": 0.5452,
       "step": 1225
     },
     {
-      "epoch": 0.05702364394993046,
-      "grad_norm": 7.252627849578857,
-      "learning_rate": 1.1404728789986092e-05,
-      "loss": 0.9402,
+      "epoch": 0.457716996929947,
+      "grad_norm": 0.012072132900357246,
+      "learning_rate": 1.3166679938014728e-05,
+      "loss": 0.5864,
       "step": 1230
     },
     {
-      "epoch": 0.05725544738062123,
-      "grad_norm": 6.398477554321289,
-      "learning_rate": 1.1451089476124248e-05,
-      "loss": 0.8712,
+      "epoch": 0.4595776351288492,
+      "grad_norm": 0.01093310210853815,
+      "learning_rate": 1.3104994133136563e-05,
+      "loss": 0.6122,
       "step": 1235
     },
     {
-      "epoch": 0.05748725081131201,
-      "grad_norm": 6.501800060272217,
-      "learning_rate": 1.1497450162262403e-05,
-      "loss": 1.164,
+      "epoch": 0.4614382733277514,
+      "grad_norm": 0.01245942059904337,
+      "learning_rate": 1.3043177293528571e-05,
+      "loss": 0.5889,
       "step": 1240
     },
     {
-      "epoch": 0.057719054242002785,
-      "grad_norm": 5.963634967803955,
-      "learning_rate": 1.1543810848400557e-05,
-      "loss": 1.0071,
+      "epoch": 0.46329891152665365,
+      "grad_norm": 0.013127041980624199,
+      "learning_rate": 1.2981232027940562e-05,
+      "loss": 0.6225,
       "step": 1245
     },
     {
-      "epoch": 0.057950857672693555,
-      "grad_norm": 6.0244317054748535,
-      "learning_rate": 1.1590171534538712e-05,
-      "loss": 1.1575,
+      "epoch": 0.46515954972555584,
+      "grad_norm": 0.018618909642100334,
+      "learning_rate": 1.2919160950542095e-05,
+      "loss": 0.6189,
       "step": 1250
     },
     {
-      "epoch": 0.05818266110338433,
-      "grad_norm": 7.449973106384277,
-      "learning_rate": 1.1636532220676866e-05,
-      "loss": 0.8669,
+      "epoch": 0.4670201879244581,
+      "grad_norm": 0.013042682781815529,
+      "learning_rate": 1.2856966680812148e-05,
+      "loss": 0.674,
       "step": 1255
     },
     {
-      "epoch": 0.0584144645340751,
-      "grad_norm": 6.4726080894470215,
-      "learning_rate": 1.1682892906815023e-05,
-      "loss": 0.9558,
+      "epoch": 0.4688808261233603,
+      "grad_norm": 0.013208975084125996,
+      "learning_rate": 1.2794651843428575e-05,
+      "loss": 0.6084,
       "step": 1260
     },
     {
-      "epoch": 0.05864626796476588,
-      "grad_norm": 4.924577713012695,
-      "learning_rate": 1.1729253592953176e-05,
-      "loss": 0.9858,
+      "epoch": 0.47074146432226255,
+      "grad_norm": 0.012473770417273045,
+      "learning_rate": 1.2732219068157335e-05,
+      "loss": 0.5748,
       "step": 1265
     },
     {
-      "epoch": 0.05887807139545665,
-      "grad_norm": 7.104029178619385,
-      "learning_rate": 1.1775614279091332e-05,
-      "loss": 1.0259,
+      "epoch": 0.47260210252116475,
+      "grad_norm": 0.013629582710564137,
+      "learning_rate": 1.2669670989741519e-05,
+      "loss": 0.6358,
       "step": 1270
     },
     {
-      "epoch": 0.05910987482614743,
-      "grad_norm": 5.916851997375488,
-      "learning_rate": 1.1821974965229485e-05,
-      "loss": 1.0703,
+      "epoch": 0.474462740720067,
+      "grad_norm": 0.014257396571338177,
+      "learning_rate": 1.2607010247790158e-05,
+      "loss": 0.6794,
       "step": 1275
     },
     {
-      "epoch": 0.0593416782568382,
-      "grad_norm": 5.41799783706665,
-      "learning_rate": 1.1868335651367642e-05,
-      "loss": 0.9557,
+      "epoch": 0.4763233789189692,
+      "grad_norm": 0.0164741612970829,
+      "learning_rate": 1.2544239486666831e-05,
+      "loss": 0.6647,
       "step": 1280
     },
     {
-      "epoch": 0.059573481687528976,
-      "grad_norm": 5.087590217590332,
-      "learning_rate": 1.1914696337505796e-05,
-      "loss": 0.8486,
+      "epoch": 0.47818401711787145,
+      "grad_norm": 0.012896180152893066,
+      "learning_rate": 1.2481361355378066e-05,
+      "loss": 0.6413,
       "step": 1285
     },
     {
-      "epoch": 0.059805285118219746,
-      "grad_norm": 4.912267208099365,
-      "learning_rate": 1.1961057023643951e-05,
-      "loss": 1.0576,
+      "epoch": 0.48004465531677365,
+      "grad_norm": 0.012296337634325027,
+      "learning_rate": 1.2418378507461544e-05,
+      "loss": 0.62,
       "step": 1290
     },
     {
-      "epoch": 0.060037088548910524,
-      "grad_norm": 5.457602024078369,
-      "learning_rate": 1.2007417709782105e-05,
-      "loss": 1.0238,
+      "epoch": 0.4819052935156759,
+      "grad_norm": 0.01213445421308279,
+      "learning_rate": 1.2355293600874132e-05,
+      "loss": 0.6611,
       "step": 1295
     },
     {
-      "epoch": 0.0602688919796013,
-      "grad_norm": 5.606520652770996,
-      "learning_rate": 1.2053778395920262e-05,
-      "loss": 0.9371,
+      "epoch": 0.4837659317145781,
+      "grad_norm": 0.01198558695614338,
+      "learning_rate": 1.229210929787969e-05,
+      "loss": 0.6438,
       "step": 1300
     },
     {
-      "epoch": 0.0602688919796013,
-      "eval_loss": 1.0377291440963745,
-      "eval_runtime": 11.2632,
-      "eval_samples_per_second": 11.276,
-      "eval_steps_per_second": 11.276,
+      "epoch": 0.4837659317145781,
+      "eval_loss": 0.6453238725662231,
+      "eval_runtime": 26.6037,
+      "eval_samples_per_second": 4.774,
+      "eval_steps_per_second": 4.774,
       "step": 1300
     },
     {
-      "epoch": 0.06050069541029207,
-      "grad_norm": 5.769026279449463,
-      "learning_rate": 1.2100139082058415e-05,
-      "loss": 0.9785,
+      "epoch": 0.4856265699134803,
+      "grad_norm": 0.012105841189622879,
+      "learning_rate": 1.2228828264936755e-05,
+      "loss": 0.675,
       "step": 1305
     },
     {
-      "epoch": 0.06073249884098285,
-      "grad_norm": 4.725026607513428,
-      "learning_rate": 1.214649976819657e-05,
-      "loss": 0.7743,
+      "epoch": 0.48748720811238255,
+      "grad_norm": 0.012657279148697853,
+      "learning_rate": 1.2165453172585964e-05,
+      "loss": 0.6066,
       "step": 1310
     },
     {
-      "epoch": 0.06096430227167362,
-      "grad_norm": 5.6914567947387695,
-      "learning_rate": 1.2192860454334726e-05,
-      "loss": 0.926,
+      "epoch": 0.48934784631128475,
+      "grad_norm": 0.01320530753582716,
+      "learning_rate": 1.2101986695337407e-05,
+      "loss": 0.6578,
       "step": 1315
     },
     {
-      "epoch": 0.061196105702364396,
-      "grad_norm": 5.947300434112549,
-      "learning_rate": 1.223922114047288e-05,
-      "loss": 0.8786,
+      "epoch": 0.491208484510187,
+      "grad_norm": 0.012736879289150238,
+      "learning_rate": 1.2038431511557715e-05,
+      "loss": 0.6596,
       "step": 1320
     },
     {
-      "epoch": 0.06142790913305517,
-      "grad_norm": 4.879319190979004,
-      "learning_rate": 1.2285581826611035e-05,
-      "loss": 0.9271,
+      "epoch": 0.4930691227090892,
+      "grad_norm": 0.013182558119297028,
+      "learning_rate": 1.197479030335706e-05,
+      "loss": 0.595,
       "step": 1325
     },
     {
-      "epoch": 0.061659712563745944,
-      "grad_norm": 5.405600070953369,
-      "learning_rate": 1.2331942512749188e-05,
-      "loss": 1.0455,
+      "epoch": 0.49492976090799146,
+      "grad_norm": 0.013970241881906986,
+      "learning_rate": 1.1911065756475953e-05,
+      "loss": 0.6525,
       "step": 1330
     },
     {
-      "epoch": 0.061891515994436715,
-      "grad_norm": 6.316072463989258,
-      "learning_rate": 1.2378303198887345e-05,
-      "loss": 0.9825,
+      "epoch": 0.49679039910689365,
+      "grad_norm": 0.012158108875155449,
+      "learning_rate": 1.1847260560171895e-05,
+      "loss": 0.576,
       "step": 1335
     },
     {
-      "epoch": 0.06212331942512749,
-      "grad_norm": 5.218293190002441,
-      "learning_rate": 1.2424663885025499e-05,
-      "loss": 0.9407,
+      "epoch": 0.4986510373057959,
+      "grad_norm": 0.012398924678564072,
+      "learning_rate": 1.1783377407105907e-05,
+      "loss": 0.6039,
       "step": 1340
     },
     {
-      "epoch": 0.06235512285581827,
-      "grad_norm": 5.403647422790527,
-      "learning_rate": 1.2471024571163654e-05,
-      "loss": 0.9509,
+      "epoch": 0.5005116755046981,
+      "grad_norm": 0.013791786506772041,
+      "learning_rate": 1.1719418993228883e-05,
+      "loss": 0.6585,
       "step": 1345
     },
     {
-      "epoch": 0.06258692628650904,
-      "grad_norm": 5.6688666343688965,
-      "learning_rate": 1.2517385257301808e-05,
-      "loss": 0.9351,
+      "epoch": 0.5023723137036004,
+      "grad_norm": 0.011280239559710026,
+      "learning_rate": 1.1655388017667812e-05,
+      "loss": 0.5919,
       "step": 1350
     },
     {
-      "epoch": 0.06281872971719982,
-      "grad_norm": 4.872259616851807,
-      "learning_rate": 1.2563745943439965e-05,
-      "loss": 0.9504,
+      "epoch": 0.5042329519025026,
+      "grad_norm": 0.015436794608831406,
+      "learning_rate": 1.159128718261189e-05,
+      "loss": 0.6632,
       "step": 1355
     },
     {
-      "epoch": 0.0630505331478906,
-      "grad_norm": 5.114622592926025,
-      "learning_rate": 1.2610106629578119e-05,
-      "loss": 0.9407,
+      "epoch": 0.5060935901014048,
+      "grad_norm": 0.015739573165774345,
+      "learning_rate": 1.1527119193198466e-05,
+      "loss": 0.6384,
       "step": 1360
     },
     {
-      "epoch": 0.06328233657858136,
-      "grad_norm": 6.169310092926025,
-      "learning_rate": 1.2656467315716274e-05,
-      "loss": 1.2146,
+      "epoch": 0.507954228300307,
+      "grad_norm": 0.01318281702697277,
+      "learning_rate": 1.146288675739889e-05,
+      "loss": 0.6312,
       "step": 1365
     },
     {
-      "epoch": 0.06351414000927214,
-      "grad_norm": 6.065832138061523,
-      "learning_rate": 1.2702828001854428e-05,
-      "loss": 1.0371,
+      "epoch": 0.5098148664992093,
+      "grad_norm": 0.01480270829051733,
+      "learning_rate": 1.1398592585904234e-05,
+      "loss": 0.6453,
       "step": 1370
     },
     {
-      "epoch": 0.06374594343996291,
-      "grad_norm": 7.362551689147949,
-      "learning_rate": 1.2749188687992585e-05,
-      "loss": 0.9644,
+      "epoch": 0.5116755046981114,
+      "grad_norm": 0.011967113241553307,
+      "learning_rate": 1.133423939201089e-05,
+      "loss": 0.6335,
       "step": 1375
     },
     {
-      "epoch": 0.06397774687065369,
-      "grad_norm": 6.0073065757751465,
-      "learning_rate": 1.2795549374130738e-05,
-      "loss": 1.1034,
+      "epoch": 0.5135361428970137,
+      "grad_norm": 0.013129732571542263,
+      "learning_rate": 1.1269829891506081e-05,
+      "loss": 0.5852,
       "step": 1380
     },
     {
-      "epoch": 0.06420955030134445,
-      "grad_norm": 6.123709201812744,
-      "learning_rate": 1.2841910060268893e-05,
-      "loss": 1.0149,
+      "epoch": 0.5153967810959159,
+      "grad_norm": 0.0132956113666296,
+      "learning_rate": 1.1205366802553231e-05,
+      "loss": 0.6477,
       "step": 1385
     },
     {
-      "epoch": 0.06444135373203523,
-      "grad_norm": 5.83848237991333,
-      "learning_rate": 1.2888270746407047e-05,
-      "loss": 1.1497,
+      "epoch": 0.5172574192948182,
+      "grad_norm": 0.01314165536314249,
+      "learning_rate": 1.1140852845577273e-05,
+      "loss": 0.6441,
       "step": 1390
     },
     {
-      "epoch": 0.06467315716272601,
-      "grad_norm": 5.201710224151611,
-      "learning_rate": 1.2934631432545204e-05,
-      "loss": 0.8791,
+      "epoch": 0.5191180574937203,
+      "grad_norm": 0.01178740430623293,
+      "learning_rate": 1.1076290743149827e-05,
+      "loss": 0.6035,
       "step": 1395
     },
     {
-      "epoch": 0.06490496059341679,
-      "grad_norm": 5.282418727874756,
-      "learning_rate": 1.2980992118683358e-05,
-      "loss": 0.9191,
+      "epoch": 0.5209786956926226,
+      "grad_norm": 0.013940893113613129,
+      "learning_rate": 1.1011683219874324e-05,
+      "loss": 0.6492,
       "step": 1400
     },
     {
-      "epoch": 0.06490496059341679,
-      "eval_loss": 1.0392991304397583,
-      "eval_runtime": 11.2723,
-      "eval_samples_per_second": 11.267,
-      "eval_steps_per_second": 11.267,
+      "epoch": 0.5209786956926226,
+      "eval_loss": 0.644782543182373,
+      "eval_runtime": 26.5697,
+      "eval_samples_per_second": 4.78,
+      "eval_steps_per_second": 4.78,
       "step": 1400
     },
     {
-      "epoch": 0.06513676402410756,
-      "grad_norm": 6.327620506286621,
-      "learning_rate": 1.3027352804821511e-05,
-      "loss": 0.9029,
+      "epoch": 0.5228393338915248,
+      "grad_norm": 0.013468287885189056,
+      "learning_rate": 1.0947033002271001e-05,
+      "loss": 0.6135,
       "step": 1405
     },
     {
-      "epoch": 0.06536856745479833,
-      "grad_norm": 5.448641777038574,
-      "learning_rate": 1.3073713490959668e-05,
-      "loss": 1.0854,
+      "epoch": 0.5246999720904271,
+      "grad_norm": 0.015101495198905468,
+      "learning_rate": 1.0882342818661859e-05,
+      "loss": 0.6449,
       "step": 1410
     },
     {
-      "epoch": 0.0656003708854891,
-      "grad_norm": 6.067873001098633,
-      "learning_rate": 1.3120074177097822e-05,
-      "loss": 1.0033,
+      "epoch": 0.5265606102893292,
+      "grad_norm": 0.013072527013719082,
+      "learning_rate": 1.0817615399055513e-05,
+      "loss": 0.6252,
       "step": 1415
     },
     {
-      "epoch": 0.06583217431617988,
-      "grad_norm": 4.918057918548584,
-      "learning_rate": 1.3166434863235977e-05,
-      "loss": 1.1301,
+      "epoch": 0.5284212484882315,
+      "grad_norm": 0.01259040180593729,
+      "learning_rate": 1.075285347503198e-05,
+      "loss": 0.6274,
       "step": 1420
     },
     {
-      "epoch": 0.06606397774687066,
-      "grad_norm": 5.890987873077393,
-      "learning_rate": 1.3212795549374131e-05,
-      "loss": 0.9813,
+      "epoch": 0.5302818866871337,
+      "grad_norm": 0.014084907248616219,
+      "learning_rate": 1.0688059779627417e-05,
+      "loss": 0.6298,
       "step": 1425
     },
     {
-      "epoch": 0.06629578117756142,
-      "grad_norm": 5.978640556335449,
-      "learning_rate": 1.3259156235512288e-05,
-      "loss": 1.0217,
+      "epoch": 0.5321425248860359,
+      "grad_norm": 0.015118095092475414,
+      "learning_rate": 1.0623237047218771e-05,
+      "loss": 0.6638,
       "step": 1430
     },
     {
-      "epoch": 0.0665275846082522,
-      "grad_norm": 7.252511024475098,
-      "learning_rate": 1.3305516921650441e-05,
-      "loss": 0.9992,
+      "epoch": 0.5340031630849381,
+      "grad_norm": 0.013168774545192719,
+      "learning_rate": 1.0558388013408378e-05,
+      "loss": 0.6134,
       "step": 1435
     },
     {
-      "epoch": 0.06675938803894298,
-      "grad_norm": 5.379759788513184,
-      "learning_rate": 1.3351877607788597e-05,
-      "loss": 0.9253,
+      "epoch": 0.5358638012838404,
+      "grad_norm": 0.011875185184180737,
+      "learning_rate": 1.0493515414908542e-05,
+      "loss": 0.6396,
       "step": 1440
     },
     {
-      "epoch": 0.06699119146963375,
-      "grad_norm": 6.269161701202393,
-      "learning_rate": 1.339823829392675e-05,
-      "loss": 1.0769,
+      "epoch": 0.5377244394827426,
+      "grad_norm": 0.013676362112164497,
+      "learning_rate": 1.0428621989426016e-05,
+      "loss": 0.6286,
       "step": 1445
     },
     {
-      "epoch": 0.06722299490032453,
-      "grad_norm": 6.233831882476807,
-      "learning_rate": 1.3444598980064907e-05,
-      "loss": 1.0805,
+      "epoch": 0.5395850776816448,
+      "grad_norm": 0.012775209732353687,
+      "learning_rate": 1.0363710475546483e-05,
+      "loss": 0.6156,
       "step": 1450
     },
     {
-      "epoch": 0.0674547983310153,
-      "grad_norm": 6.35414981842041,
-      "learning_rate": 1.3490959666203061e-05,
-      "loss": 0.9839,
+      "epoch": 0.541445715880547,
+      "grad_norm": 0.014763396233320236,
+      "learning_rate": 1.0298783612618977e-05,
+      "loss": 0.6713,
       "step": 1455
     },
     {
-      "epoch": 0.06768660176170607,
-      "grad_norm": 5.051734924316406,
-      "learning_rate": 1.3537320352341216e-05,
-      "loss": 1.1025,
+      "epoch": 0.5433063540794493,
+      "grad_norm": 0.014012233354151249,
+      "learning_rate": 1.0233844140640287e-05,
+      "loss": 0.5887,
       "step": 1460
     },
     {
-      "epoch": 0.06791840519239685,
-      "grad_norm": 6.6913886070251465,
-      "learning_rate": 1.358368103847937e-05,
-      "loss": 1.2106,
+      "epoch": 0.5451669922783515,
+      "grad_norm": 0.013994456268846989,
+      "learning_rate": 1.0168894800139311e-05,
+      "loss": 0.6509,
       "step": 1465
     },
     {
-      "epoch": 0.06815020862308763,
-      "grad_norm": 5.920834064483643,
-      "learning_rate": 1.3630041724617527e-05,
-      "loss": 0.9518,
+      "epoch": 0.5470276304772537,
+      "grad_norm": 0.014792009256780148,
+      "learning_rate": 1.0103938332061422e-05,
+      "loss": 0.6434,
       "step": 1470
     },
     {
-      "epoch": 0.06838201205377839,
-      "grad_norm": 4.299671173095703,
-      "learning_rate": 1.367640241075568e-05,
-      "loss": 0.7529,
+      "epoch": 0.5488882686761559,
+      "grad_norm": 0.01255644578486681,
+      "learning_rate": 1.0038977477652779e-05,
+      "loss": 0.6407,
       "step": 1475
     },
     {
-      "epoch": 0.06861381548446917,
-      "grad_norm": 5.374516010284424,
-      "learning_rate": 1.3722763096893834e-05,
-      "loss": 1.1624,
+      "epoch": 0.5507489068750582,
+      "grad_norm": 0.013431582599878311,
+      "learning_rate": 9.974014978344646e-06,
+      "loss": 0.6528,
       "step": 1480
     },
     {
-      "epoch": 0.06884561891515995,
-      "grad_norm": 6.2488932609558105,
-      "learning_rate": 1.376912378303199e-05,
-      "loss": 1.0323,
+      "epoch": 0.5526095450739603,
+      "grad_norm": 0.0123568931594491,
+      "learning_rate": 9.909053575637717e-06,
+      "loss": 0.602,
       "step": 1485
     },
     {
-      "epoch": 0.06907742234585072,
-      "grad_norm": 6.374792575836182,
-      "learning_rate": 1.3815484469170145e-05,
-      "loss": 1.1237,
+      "epoch": 0.5544701832728626,
+      "grad_norm": 0.014569776132702827,
+      "learning_rate": 9.844096010986392e-06,
+      "loss": 0.6268,
       "step": 1490
     },
     {
-      "epoch": 0.06930922577654149,
-      "grad_norm": 5.536932468414307,
-      "learning_rate": 1.38618451553083e-05,
-      "loss": 1.1524,
+      "epoch": 0.5563308214717648,
+      "grad_norm": 0.012518184259533882,
+      "learning_rate": 9.779145025683114e-06,
+      "loss": 0.5936,
       "step": 1495
     },
     {
-      "epoch": 0.06954102920723226,
-      "grad_norm": 5.977290153503418,
-      "learning_rate": 1.3908205841446454e-05,
-      "loss": 0.9357,
+      "epoch": 0.5581914596706671,
+      "grad_norm": 0.011615007184445858,
+      "learning_rate": 9.714203360742666e-06,
+      "loss": 0.6275,
       "step": 1500
     },
     {
-      "epoch": 0.06954102920723226,
-      "eval_loss": 1.0378950834274292,
-      "eval_runtime": 11.2663,
-      "eval_samples_per_second": 11.273,
-      "eval_steps_per_second": 11.273,
+      "epoch": 0.5581914596706671,
+      "eval_loss": 0.6442868709564209,
+      "eval_runtime": 26.5597,
+      "eval_samples_per_second": 4.782,
+      "eval_steps_per_second": 4.782,
       "step": 1500
     },
     {
-      "epoch": 0.06977283263792304,
-      "grad_norm": 5.865034580230713,
-      "learning_rate": 1.3954566527584609e-05,
-      "loss": 1.0989,
+      "epoch": 0.5600520978695692,
+      "grad_norm": 0.013426556251943111,
+      "learning_rate": 9.649273756786486e-06,
+      "loss": 0.6291,
       "step": 1505
     },
     {
-      "epoch": 0.07000463606861382,
-      "grad_norm": 5.796088695526123,
-      "learning_rate": 1.4000927213722764e-05,
-      "loss": 0.9742,
+      "epoch": 0.5619127360684715,
+      "grad_norm": 0.01306986529380083,
+      "learning_rate": 9.584358953927043e-06,
+      "loss": 0.6211,
       "step": 1510
     },
     {
-      "epoch": 0.0702364394993046,
-      "grad_norm": 0.5797013640403748,
-      "learning_rate": 1.404728789986092e-05,
-      "loss": 0.9583,
+      "epoch": 0.5637733742673737,
+      "grad_norm": 0.012594708241522312,
+      "learning_rate": 9.519461691652169e-06,
+      "loss": 0.5803,
       "step": 1515
     },
     {
-      "epoch": 0.07046824292999536,
-      "grad_norm": 6.2530341148376465,
-      "learning_rate": 1.4093648585999073e-05,
-      "loss": 1.0374,
+      "epoch": 0.565634012466276,
+      "grad_norm": 0.014830299653112888,
+      "learning_rate": 9.454584708709462e-06,
+      "loss": 0.5976,
       "step": 1520
     },
     {
-      "epoch": 0.07070004636068614,
-      "grad_norm": 5.679171562194824,
-      "learning_rate": 1.414000927213723e-05,
-      "loss": 1.1167,
+      "epoch": 0.5674946506651781,
+      "grad_norm": 0.013333864510059357,
+      "learning_rate": 9.389730742990714e-06,
+      "loss": 0.6154,
       "step": 1525
     },
     {
-      "epoch": 0.07093184979137691,
-      "grad_norm": 4.820891380310059,
-      "learning_rate": 1.4186369958275384e-05,
-      "loss": 0.9248,
+      "epoch": 0.5693552888640804,
+      "grad_norm": 0.013150627724826336,
+      "learning_rate": 9.324902531416348e-06,
+      "loss": 0.581,
       "step": 1530
     },
     {
-      "epoch": 0.07116365322206769,
-      "grad_norm": 5.5291948318481445,
-      "learning_rate": 1.423273064441354e-05,
-      "loss": 1.014,
+      "epoch": 0.5712159270629826,
+      "grad_norm": 0.012874056585133076,
+      "learning_rate": 9.260102809819939e-06,
+      "loss": 0.6224,
       "step": 1535
     },
     {
-      "epoch": 0.07139545665275845,
-      "grad_norm": 5.817046165466309,
-      "learning_rate": 1.4279091330551693e-05,
-      "loss": 1.2119,
+      "epoch": 0.5730765652618848,
+      "grad_norm": 0.012678616680204868,
+      "learning_rate": 9.195334312832742e-06,
+      "loss": 0.6705,
       "step": 1540
     },
     {
-      "epoch": 0.07162726008344923,
-      "grad_norm": 5.100951194763184,
-      "learning_rate": 1.432545201668985e-05,
-      "loss": 0.8246,
+      "epoch": 0.574937203460787,
+      "grad_norm": 0.011814710684120655,
+      "learning_rate": 9.1305997737683e-06,
+      "loss": 0.6103,
       "step": 1545
     },
     {
-      "epoch": 0.07185906351414001,
-      "grad_norm": 8.208178520202637,
-      "learning_rate": 1.4371812702828003e-05,
-      "loss": 1.1136,
+      "epoch": 0.5767978416596893,
+      "grad_norm": 0.013010908849537373,
+      "learning_rate": 9.065901924507085e-06,
+      "loss": 0.655,
       "step": 1550
     },
     {
-      "epoch": 0.07209086694483079,
-      "grad_norm": 5.660605430603027,
-      "learning_rate": 1.4418173388966157e-05,
-      "loss": 0.8556,
+      "epoch": 0.5786584798585915,
+      "grad_norm": 0.013622297905385494,
+      "learning_rate": 9.001243495381207e-06,
+      "loss": 0.5961,
       "step": 1555
     },
     {
-      "epoch": 0.07232267037552156,
-      "grad_norm": 7.057506084442139,
-      "learning_rate": 1.4464534075104312e-05,
-      "loss": 1.0404,
+      "epoch": 0.5805191180574937,
+      "grad_norm": 0.013876644894480705,
+      "learning_rate": 8.936627215059206e-06,
+      "loss": 0.6789,
       "step": 1560
     },
     {
-      "epoch": 0.07255447380621233,
-      "grad_norm": 5.514895915985107,
-      "learning_rate": 1.4510894761242466e-05,
-      "loss": 0.9745,
+      "epoch": 0.5823797562563959,
+      "grad_norm": 0.01281541958451271,
+      "learning_rate": 8.872055810430881e-06,
+      "loss": 0.6567,
       "step": 1565
     },
     {
-      "epoch": 0.0727862772369031,
-      "grad_norm": 5.650478363037109,
-      "learning_rate": 1.4557255447380623e-05,
-      "loss": 0.9403,
+      "epoch": 0.5842403944552982,
+      "grad_norm": 0.012182512320578098,
+      "learning_rate": 8.80753200649222e-06,
+      "loss": 0.6338,
       "step": 1570
     },
     {
-      "epoch": 0.07301808066759388,
-      "grad_norm": 5.063937187194824,
-      "learning_rate": 1.4603616133518777e-05,
-      "loss": 1.0381,
+      "epoch": 0.5861010326542004,
+      "grad_norm": 0.012718496844172478,
+      "learning_rate": 8.743058526230409e-06,
+      "loss": 0.6151,
       "step": 1575
     },
     {
-      "epoch": 0.07324988409828466,
-      "grad_norm": 4.689387798309326,
-      "learning_rate": 1.4649976819656932e-05,
-      "loss": 0.9246,
+      "epoch": 0.5879616708531026,
+      "grad_norm": 0.01301120687276125,
+      "learning_rate": 8.678638090508897e-06,
+      "loss": 0.6147,
       "step": 1580
     },
     {
-      "epoch": 0.07348168752897542,
-      "grad_norm": 6.5369720458984375,
-      "learning_rate": 1.4696337505795087e-05,
-      "loss": 1.1306,
+      "epoch": 0.5898223090520048,
+      "grad_norm": 0.011715607717633247,
+      "learning_rate": 8.614273417952593e-06,
+      "loss": 0.5776,
       "step": 1585
     },
     {
-      "epoch": 0.0737134909596662,
-      "grad_norm": 6.205171585083008,
-      "learning_rate": 1.4742698191933242e-05,
-      "loss": 1.0382,
+      "epoch": 0.5916829472509071,
+      "grad_norm": 0.013846187852323055,
+      "learning_rate": 8.549967224833131e-06,
+      "loss": 0.6604,
       "step": 1590
     },
     {
-      "epoch": 0.07394529439035698,
-      "grad_norm": 4.940855026245117,
-      "learning_rate": 1.4789058878071396e-05,
-      "loss": 0.9282,
+      "epoch": 0.5935435854498092,
+      "grad_norm": 0.013312350027263165,
+      "learning_rate": 8.485722224954237e-06,
+      "loss": 0.6395,
       "step": 1595
     },
     {
-      "epoch": 0.07417709782104775,
-      "grad_norm": 5.930394649505615,
-      "learning_rate": 1.4835419564209551e-05,
-      "loss": 0.9031,
+      "epoch": 0.5954042236487115,
+      "grad_norm": 0.014313463121652603,
+      "learning_rate": 8.421541129537194e-06,
+      "loss": 0.6848,
       "step": 1600
     },
     {
-      "epoch": 0.07417709782104775,
-      "eval_loss": 1.041509985923767,
-      "eval_runtime": 11.2652,
-      "eval_samples_per_second": 11.274,
-      "eval_steps_per_second": 11.274,
+      "epoch": 0.5954042236487115,
+      "eval_loss": 0.6439012885093689,
+      "eval_runtime": 26.5935,
+      "eval_samples_per_second": 4.776,
+      "eval_steps_per_second": 4.776,
       "step": 1600
     },
     {
-      "epoch": 0.07440890125173853,
-      "grad_norm": 5.346956729888916,
-      "learning_rate": 1.4881780250347707e-05,
-      "loss": 0.9515,
+      "epoch": 0.5972648618476137,
+      "grad_norm": 0.012196023017168045,
+      "learning_rate": 8.357426647106451e-06,
+      "loss": 0.6079,
       "step": 1605
     },
     {
-      "epoch": 0.0746407046824293,
-      "grad_norm": 7.964709758758545,
-      "learning_rate": 1.4928140936485862e-05,
-      "loss": 1.0475,
+      "epoch": 0.599125500046516,
+      "grad_norm": 0.0135956397280097,
+      "learning_rate": 8.293381483375293e-06,
+      "loss": 0.6463,
       "step": 1610
     },
     {
-      "epoch": 0.07487250811312007,
-      "grad_norm": 5.0009331703186035,
-      "learning_rate": 1.4974501622624016e-05,
-      "loss": 1.1008,
+      "epoch": 0.6009861382454181,
+      "grad_norm": 0.011982797645032406,
+      "learning_rate": 8.229408341131665e-06,
+      "loss": 0.6113,
       "step": 1615
     },
     {
-      "epoch": 0.07510431154381085,
-      "grad_norm": 5.313481330871582,
-      "learning_rate": 1.5020862308762173e-05,
-      "loss": 0.9793,
+      "epoch": 0.6028467764443204,
+      "grad_norm": 0.014127896167337894,
+      "learning_rate": 8.165509920124125e-06,
+      "loss": 0.6602,
       "step": 1620
     },
     {
-      "epoch": 0.07533611497450163,
-      "grad_norm": 4.500079154968262,
-      "learning_rate": 1.5067222994900326e-05,
-      "loss": 1.067,
+      "epoch": 0.6047074146432226,
+      "grad_norm": 0.012588880024850368,
+      "learning_rate": 8.10168891694789e-06,
+      "loss": 0.6376,
       "step": 1625
     },
     {
-      "epoch": 0.07556791840519239,
-      "grad_norm": 5.636900901794434,
-      "learning_rate": 1.511358368103848e-05,
-      "loss": 1.081,
+      "epoch": 0.6065680528421249,
+      "grad_norm": 0.012559432536363602,
+      "learning_rate": 8.037948024931039e-06,
+      "loss": 0.6336,
       "step": 1630
     },
     {
-      "epoch": 0.07579972183588317,
-      "grad_norm": 5.819832801818848,
-      "learning_rate": 1.5159944367176635e-05,
-      "loss": 1.0421,
+      "epoch": 0.608428691041027,
+      "grad_norm": 0.012455436401069164,
+      "learning_rate": 7.974289934020879e-06,
+      "loss": 0.6403,
       "step": 1635
     },
     {
-      "epoch": 0.07603152526657395,
-      "grad_norm": 4.920659065246582,
-      "learning_rate": 1.5206305053314789e-05,
-      "loss": 0.8251,
+      "epoch": 0.6102893292399293,
+      "grad_norm": 0.01241991762071848,
+      "learning_rate": 7.91071733067038e-06,
+      "loss": 0.6518,
       "step": 1640
     },
     {
-      "epoch": 0.07626332869726472,
-      "grad_norm": 7.103638172149658,
-      "learning_rate": 1.5252665739452946e-05,
-      "loss": 1.1794,
+      "epoch": 0.6121499674388315,
+      "grad_norm": 0.012328005395829678,
+      "learning_rate": 7.84723289772484e-06,
+      "loss": 0.6162,
       "step": 1645
     },
     {
-      "epoch": 0.07649513212795549,
-      "grad_norm": 6.366265296936035,
-      "learning_rate": 1.52990264255911e-05,
-      "loss": 0.9501,
+      "epoch": 0.6140106056377338,
+      "grad_norm": 0.012782512232661247,
+      "learning_rate": 7.783839314308656e-06,
+      "loss": 0.6624,
       "step": 1650
     },
     {
-      "epoch": 0.07672693555864626,
-      "grad_norm": 7.341405868530273,
-      "learning_rate": 1.5345387111729255e-05,
-      "loss": 0.8921,
+      "epoch": 0.6158712438366359,
+      "grad_norm": 0.012050081044435501,
+      "learning_rate": 7.720539255712252e-06,
+      "loss": 0.6565,
       "step": 1655
     },
     {
-      "epoch": 0.07695873898933704,
-      "grad_norm": 8.745707511901855,
-      "learning_rate": 1.539174779786741e-05,
-      "loss": 0.9825,
+      "epoch": 0.6177318820355382,
+      "grad_norm": 0.013794245198369026,
+      "learning_rate": 7.657335393279179e-06,
+      "loss": 0.6475,
       "step": 1660
     },
     {
-      "epoch": 0.07719054242002782,
-      "grad_norm": 6.279496192932129,
-      "learning_rate": 1.5438108484005565e-05,
-      "loss": 1.1137,
+      "epoch": 0.6195925202344404,
+      "grad_norm": 0.012430761009454727,
+      "learning_rate": 7.594230394293404e-06,
+      "loss": 0.5821,
       "step": 1665
     },
     {
-      "epoch": 0.0774223458507186,
-      "grad_norm": 4.467720031738281,
-      "learning_rate": 1.5484469170143717e-05,
-      "loss": 0.8885,
+      "epoch": 0.6214531584333426,
+      "grad_norm": 0.013150406070053577,
+      "learning_rate": 7.531226921866715e-06,
+      "loss": 0.6023,
       "step": 1670
     },
     {
-      "epoch": 0.07765414928140936,
-      "grad_norm": 4.987480163574219,
-      "learning_rate": 1.5530829856281876e-05,
-      "loss": 1.022,
+      "epoch": 0.6233137966322448,
+      "grad_norm": 0.013466687873005867,
+      "learning_rate": 7.468327634826354e-06,
+      "loss": 0.637,
       "step": 1675
     },
     {
-      "epoch": 0.07788595271210014,
-      "grad_norm": 5.3170061111450195,
-      "learning_rate": 1.5577190542420028e-05,
-      "loss": 1.0107,
+      "epoch": 0.6251744348311471,
+      "grad_norm": 0.014000273309648037,
+      "learning_rate": 7.405535187602809e-06,
+      "loss": 0.6113,
       "step": 1680
     },
     {
-      "epoch": 0.07811775614279091,
-      "grad_norm": 5.079998970031738,
-      "learning_rate": 1.5623551228558183e-05,
-      "loss": 1.0645,
+      "epoch": 0.6270350730300494,
+      "grad_norm": 0.013736380264163017,
+      "learning_rate": 7.3428522301177894e-06,
+      "loss": 0.5914,
       "step": 1685
     },
     {
-      "epoch": 0.07834955957348169,
-      "grad_norm": 6.98558235168457,
-      "learning_rate": 1.566991191469634e-05,
-      "loss": 1.0992,
+      "epoch": 0.6288957112289515,
+      "grad_norm": 0.013854081742465496,
+      "learning_rate": 7.2802814076723896e-06,
+      "loss": 0.6744,
       "step": 1690
     },
     {
-      "epoch": 0.07858136300417246,
-      "grad_norm": 5.634239673614502,
-      "learning_rate": 1.5716272600834494e-05,
-      "loss": 0.8674,
+      "epoch": 0.6307563494278537,
+      "grad_norm": 0.012866591103374958,
+      "learning_rate": 7.217825360835475e-06,
+      "loss": 0.6209,
       "step": 1695
     },
     {
-      "epoch": 0.07881316643486323,
-      "grad_norm": 6.679068088531494,
-      "learning_rate": 1.576263328697265e-05,
-      "loss": 0.9038,
+      "epoch": 0.632616987626756,
+      "grad_norm": 0.012788123451173306,
+      "learning_rate": 7.155486725332224e-06,
+      "loss": 0.5764,
       "step": 1700
     },
     {
-      "epoch": 0.07881316643486323,
-      "eval_loss": 1.0476100444793701,
-      "eval_runtime": 11.2706,
-      "eval_samples_per_second": 11.268,
-      "eval_steps_per_second": 11.268,
+      "epoch": 0.632616987626756,
+      "eval_loss": 0.6434745192527771,
+      "eval_runtime": 26.5737,
+      "eval_samples_per_second": 4.779,
+      "eval_steps_per_second": 4.779,
       "step": 1700
     },
     {
-      "epoch": 0.07904496986555401,
-      "grad_norm": 4.895448207855225,
-      "learning_rate": 1.5808993973110804e-05,
-      "loss": 0.9676,
+      "epoch": 0.6344776258256583,
+      "grad_norm": 0.013442011550068855,
+      "learning_rate": 7.093268131932905e-06,
+      "loss": 0.6522,
       "step": 1705
     },
     {
-      "epoch": 0.07927677329624479,
-      "grad_norm": 5.62677001953125,
-      "learning_rate": 1.585535465924896e-05,
-      "loss": 1.1671,
+      "epoch": 0.6363382640245604,
+      "grad_norm": 0.013659958727657795,
+      "learning_rate": 7.03117220634187e-06,
+      "loss": 0.5949,
       "step": 1710
     },
     {
-      "epoch": 0.07950857672693556,
-      "grad_norm": 4.569333076477051,
-      "learning_rate": 1.590171534538711e-05,
-      "loss": 0.8581,
+      "epoch": 0.6381989022234626,
+      "grad_norm": 0.013367819599807262,
+      "learning_rate": 6.9692015690867135e-06,
+      "loss": 0.5959,
       "step": 1715
     },
     {
-      "epoch": 0.07974038015762633,
-      "grad_norm": 6.019289970397949,
-      "learning_rate": 1.5948076031525267e-05,
-      "loss": 0.9214,
+      "epoch": 0.6400595404223649,
+      "grad_norm": 0.013543561100959778,
+      "learning_rate": 6.9073588354077125e-06,
+      "loss": 0.6539,
       "step": 1720
     },
     {
-      "epoch": 0.0799721835883171,
-      "grad_norm": 5.4483537673950195,
-      "learning_rate": 1.5994436717663422e-05,
-      "loss": 1.16,
+      "epoch": 0.641920178621267,
+      "grad_norm": 0.014200146310031414,
+      "learning_rate": 6.845646615147445e-06,
+      "loss": 0.6438,
       "step": 1725
     },
     {
-      "epoch": 0.08020398701900788,
-      "grad_norm": 5.327470302581787,
-      "learning_rate": 1.6040797403801578e-05,
-      "loss": 0.8904,
+      "epoch": 0.6437808168201693,
+      "grad_norm": 0.012410931289196014,
+      "learning_rate": 6.784067512640666e-06,
+      "loss": 0.6035,
       "step": 1730
     },
     {
-      "epoch": 0.08043579044969866,
-      "grad_norm": 4.499557971954346,
-      "learning_rate": 1.6087158089939733e-05,
-      "loss": 1.0687,
+      "epoch": 0.6456414550190716,
+      "grad_norm": 0.01299245934933424,
+      "learning_rate": 6.7226241266043735e-06,
+      "loss": 0.6507,
       "step": 1735
     },
     {
-      "epoch": 0.08066759388038942,
-      "grad_norm": 4.95378303527832,
-      "learning_rate": 1.6133518776077888e-05,
-      "loss": 0.9192,
+      "epoch": 0.6475020932179738,
+      "grad_norm": 0.013046164996922016,
+      "learning_rate": 6.661319050028167e-06,
+      "loss": 0.6277,
       "step": 1740
     },
     {
-      "epoch": 0.0808993973110802,
-      "grad_norm": 5.148226737976074,
-      "learning_rate": 1.617987946221604e-05,
-      "loss": 0.9066,
+      "epoch": 0.649362731416876,
+      "grad_norm": 0.013011117465794086,
+      "learning_rate": 6.600154870064812e-06,
+      "loss": 0.6415,
       "step": 1745
     },
     {
-      "epoch": 0.08113120074177098,
-      "grad_norm": 5.059067726135254,
-      "learning_rate": 1.62262401483542e-05,
-      "loss": 1.0501,
+      "epoch": 0.6512233696157782,
+      "grad_norm": 0.012195507064461708,
+      "learning_rate": 6.53913416792105e-06,
+      "loss": 0.5718,
       "step": 1750
     },
     {
-      "epoch": 0.08136300417246176,
-      "grad_norm": 4.736515998840332,
-      "learning_rate": 1.627260083449235e-05,
-      "loss": 1.0785,
+      "epoch": 0.6530840078146805,
+      "grad_norm": 0.012789854779839516,
+      "learning_rate": 6.478259518748675e-06,
+      "loss": 0.5963,
       "step": 1755
     },
     {
-      "epoch": 0.08159480760315253,
-      "grad_norm": 5.576144695281982,
-      "learning_rate": 1.6318961520630506e-05,
-      "loss": 1.0075,
+      "epoch": 0.6549446460135827,
+      "grad_norm": 0.01217294204980135,
+      "learning_rate": 6.41753349153587e-06,
+      "loss": 0.6361,
       "step": 1760
     },
     {
-      "epoch": 0.0818266110338433,
-      "grad_norm": 4.682522773742676,
-      "learning_rate": 1.636532220676866e-05,
-      "loss": 1.0356,
+      "epoch": 0.6568052842124849,
+      "grad_norm": 0.014139696955680847,
+      "learning_rate": 6.356958648998762e-06,
+      "loss": 0.6321,
       "step": 1765
     },
     {
-      "epoch": 0.08205841446453407,
-      "grad_norm": 5.505351543426514,
-      "learning_rate": 1.6411682892906817e-05,
-      "loss": 0.8688,
+      "epoch": 0.6586659224113871,
+      "grad_norm": 0.014064906165003777,
+      "learning_rate": 6.296537547473302e-06,
+      "loss": 0.6519,
       "step": 1770
     },
     {
-      "epoch": 0.08229021789522485,
-      "grad_norm": 5.349116325378418,
-      "learning_rate": 1.6458043579044972e-05,
-      "loss": 1.0601,
+      "epoch": 0.6605265606102894,
+      "grad_norm": 0.012787656858563423,
+      "learning_rate": 6.236272736807378e-06,
+      "loss": 0.6033,
       "step": 1775
     },
     {
-      "epoch": 0.08252202132591563,
-      "grad_norm": 4.7216572761535645,
-      "learning_rate": 1.6504404265183127e-05,
-      "loss": 0.8931,
+      "epoch": 0.6623871988091915,
+      "grad_norm": 0.012475831434130669,
+      "learning_rate": 6.176166760253196e-06,
+      "loss": 0.5947,
       "step": 1780
     },
     {
-      "epoch": 0.08275382475660639,
-      "grad_norm": 6.225747585296631,
-      "learning_rate": 1.6550764951321283e-05,
-      "loss": 1.1836,
+      "epoch": 0.6642478370080938,
+      "grad_norm": 0.013439202681183815,
+      "learning_rate": 6.116222154359952e-06,
+      "loss": 0.636,
       "step": 1785
     },
     {
-      "epoch": 0.08298562818729717,
-      "grad_norm": 4.889942646026611,
-      "learning_rate": 1.6597125637459435e-05,
-      "loss": 0.9129,
+      "epoch": 0.666108475206996,
+      "grad_norm": 0.01295219361782074,
+      "learning_rate": 6.056441448866817e-06,
+      "loss": 0.6203,
       "step": 1790
     },
     {
-      "epoch": 0.08321743161798795,
-      "grad_norm": 4.839940071105957,
-      "learning_rate": 1.664348632359759e-05,
-      "loss": 0.9188,
+      "epoch": 0.6679691134058983,
+      "grad_norm": 0.012275603599846363,
+      "learning_rate": 5.996827166596129e-06,
+      "loss": 0.6743,
       "step": 1795
     },
     {
-      "epoch": 0.08344923504867872,
-      "grad_norm": 4.898179531097412,
-      "learning_rate": 1.6689847009735745e-05,
-      "loss": 1.1024,
+      "epoch": 0.6698297516048004,
+      "grad_norm": 0.01329441275447607,
+      "learning_rate": 5.937381823346964e-06,
+      "loss": 0.5975,
       "step": 1800
     },
     {
-      "epoch": 0.08344923504867872,
-      "eval_loss": 1.0464279651641846,
-      "eval_runtime": 11.2783,
-      "eval_samples_per_second": 11.261,
-      "eval_steps_per_second": 11.261,
+      "epoch": 0.6698297516048004,
+      "eval_loss": 0.6431623101234436,
+      "eval_runtime": 26.5644,
+      "eval_samples_per_second": 4.781,
+      "eval_steps_per_second": 4.781,
       "step": 1800
     },
     {
-      "epoch": 0.08368103847936949,
-      "grad_norm": 6.300966739654541,
-      "learning_rate": 1.67362076958739e-05,
-      "loss": 1.006,
+      "epoch": 0.6716903898037027,
+      "grad_norm": 0.013007073663175106,
+      "learning_rate": 5.878107927788962e-06,
+      "loss": 0.6165,
       "step": 1805
     },
     {
-      "epoch": 0.08391284191006027,
-      "grad_norm": 5.7721476554870605,
-      "learning_rate": 1.6782568382012056e-05,
-      "loss": 0.9676,
+      "epoch": 0.6735510280026049,
+      "grad_norm": 0.013040522113442421,
+      "learning_rate": 5.819007981356441e-06,
+      "loss": 0.6107,
       "step": 1810
     },
     {
-      "epoch": 0.08414464534075104,
-      "grad_norm": 5.889309406280518,
-      "learning_rate": 1.682892906815021e-05,
-      "loss": 1.0873,
+      "epoch": 0.6754116662015072,
+      "grad_norm": 0.014542641118168831,
+      "learning_rate": 5.760084478142842e-06,
+      "loss": 0.6284,
       "step": 1815
     },
     {
-      "epoch": 0.08437644877144182,
-      "grad_norm": 6.087542533874512,
-      "learning_rate": 1.6875289754288363e-05,
-      "loss": 1.129,
+      "epoch": 0.6772723044004093,
+      "grad_norm": 0.01339266262948513,
+      "learning_rate": 5.701339904795486e-06,
+      "loss": 0.6228,
       "step": 1820
     },
     {
-      "epoch": 0.0846082522021326,
-      "grad_norm": 5.6989545822143555,
-      "learning_rate": 1.692165044042652e-05,
-      "loss": 1.0761,
+      "epoch": 0.6791329425993116,
+      "grad_norm": 0.013091943226754665,
+      "learning_rate": 5.642776740410618e-06,
+      "loss": 0.5995,
       "step": 1825
     },
     {
-      "epoch": 0.08484005563282336,
-      "grad_norm": 6.22220516204834,
-      "learning_rate": 1.6968011126564674e-05,
-      "loss": 1.1475,
+      "epoch": 0.6809935807982138,
+      "grad_norm": 0.013321259059011936,
+      "learning_rate": 5.584397456428785e-06,
+      "loss": 0.627,
       "step": 1830
     },
     {
-      "epoch": 0.08507185906351414,
-      "grad_norm": 5.317427635192871,
-      "learning_rate": 1.701437181270283e-05,
-      "loss": 0.9838,
+      "epoch": 0.682854218997116,
+      "grad_norm": 0.012838364578783512,
+      "learning_rate": 5.5262045165305615e-06,
+      "loss": 0.658,
       "step": 1835
     },
     {
-      "epoch": 0.08530366249420492,
-      "grad_norm": 4.751984119415283,
-      "learning_rate": 1.7060732498840984e-05,
-      "loss": 1.0279,
+      "epoch": 0.6847148571960182,
+      "grad_norm": 0.01306592021137476,
+      "learning_rate": 5.468200376532552e-06,
+      "loss": 0.6756,
       "step": 1840
     },
     {
-      "epoch": 0.08553546592489569,
-      "grad_norm": 6.037285804748535,
-      "learning_rate": 1.710709318497914e-05,
-      "loss": 1.0937,
+      "epoch": 0.6865754953949205,
+      "grad_norm": 0.014314945787191391,
+      "learning_rate": 5.410387484283767e-06,
+      "loss": 0.6598,
       "step": 1845
     },
     {
-      "epoch": 0.08576726935558646,
-      "grad_norm": 4.684612274169922,
-      "learning_rate": 1.7153453871117295e-05,
-      "loss": 0.9422,
+      "epoch": 0.6884361335938227,
+      "grad_norm": 0.013973386958241463,
+      "learning_rate": 5.352768279562315e-06,
+      "loss": 0.6535,
       "step": 1850
     },
     {
-      "epoch": 0.08599907278627723,
-      "grad_norm": 4.810619831085205,
-      "learning_rate": 1.719981455725545e-05,
-      "loss": 0.9663,
+      "epoch": 0.6902967717927249,
+      "grad_norm": 0.012864621356129646,
+      "learning_rate": 5.295345193972445e-06,
+      "loss": 0.6422,
       "step": 1855
     },
     {
-      "epoch": 0.08623087621696801,
-      "grad_norm": 5.514533042907715,
-      "learning_rate": 1.7246175243393602e-05,
-      "loss": 1.0154,
+      "epoch": 0.6921574099916271,
+      "grad_norm": 0.014392906799912453,
+      "learning_rate": 5.238120650841925e-06,
+      "loss": 0.664,
       "step": 1860
     },
     {
-      "epoch": 0.08646267964765879,
-      "grad_norm": 6.909518718719482,
-      "learning_rate": 1.7292535929531757e-05,
-      "loss": 1.0768,
+      "epoch": 0.6940180481905294,
+      "grad_norm": 0.01327193807810545,
+      "learning_rate": 5.18109706511978e-06,
+      "loss": 0.6179,
       "step": 1865
     },
     {
-      "epoch": 0.08669448307834957,
-      "grad_norm": 4.488744258880615,
-      "learning_rate": 1.7338896615669913e-05,
-      "loss": 1.0446,
+      "epoch": 0.6958786863894316,
+      "grad_norm": 0.01355487760156393,
+      "learning_rate": 5.124276843274372e-06,
+      "loss": 0.605,
       "step": 1870
     },
     {
-      "epoch": 0.08692628650904033,
-      "grad_norm": 4.692446708679199,
-      "learning_rate": 1.7385257301808068e-05,
-      "loss": 1.0563,
+      "epoch": 0.6977393245883338,
+      "grad_norm": 0.012646518647670746,
+      "learning_rate": 5.067662383191845e-06,
+      "loss": 0.6608,
       "step": 1875
     },
     {
-      "epoch": 0.0871580899397311,
-      "grad_norm": 5.291960716247559,
-      "learning_rate": 1.7431617987946223e-05,
-      "loss": 1.0662,
+      "epoch": 0.699599962787236,
+      "grad_norm": 0.013363865204155445,
+      "learning_rate": 5.011256074074945e-06,
+      "loss": 0.6975,
       "step": 1880
     },
     {
-      "epoch": 0.08738989337042188,
-      "grad_norm": 5.532649517059326,
-      "learning_rate": 1.747797867408438e-05,
-      "loss": 1.1875,
+      "epoch": 0.7014606009861383,
+      "grad_norm": 0.013032359071075916,
+      "learning_rate": 4.955060296342163e-06,
+      "loss": 0.6041,
       "step": 1885
     },
     {
-      "epoch": 0.08762169680111266,
-      "grad_norm": 5.574441909790039,
-      "learning_rate": 1.7524339360222534e-05,
-      "loss": 1.1473,
+      "epoch": 0.7033212391850404,
+      "grad_norm": 0.012049228884279728,
+      "learning_rate": 4.899077421527304e-06,
+      "loss": 0.6078,
       "step": 1890
     },
     {
-      "epoch": 0.08785350023180342,
-      "grad_norm": 6.044393539428711,
-      "learning_rate": 1.7570700046360686e-05,
-      "loss": 1.1049,
+      "epoch": 0.7051818773839427,
+      "grad_norm": 0.013416007161140442,
+      "learning_rate": 4.843309812179405e-06,
+      "loss": 0.6514,
       "step": 1895
     },
     {
-      "epoch": 0.0880853036624942,
-      "grad_norm": 4.733377456665039,
-      "learning_rate": 1.7617060732498845e-05,
-      "loss": 0.9169,
+      "epoch": 0.7070425155828449,
+      "grad_norm": 0.013701760210096836,
+      "learning_rate": 4.787759821763017e-06,
+      "loss": 0.6606,
       "step": 1900
     },
     {
-      "epoch": 0.0880853036624942,
-      "eval_loss": 1.0482640266418457,
-      "eval_runtime": 11.2598,
-      "eval_samples_per_second": 11.279,
-      "eval_steps_per_second": 11.279,
+      "epoch": 0.7070425155828449,
+      "eval_loss": 0.6429811716079712,
+      "eval_runtime": 26.563,
+      "eval_samples_per_second": 4.781,
+      "eval_steps_per_second": 4.781,
       "step": 1900
     },
     {
-      "epoch": 0.08831710709318498,
-      "grad_norm": 6.185121536254883,
-      "learning_rate": 1.7663421418636996e-05,
-      "loss": 0.8974,
+      "epoch": 0.7089031537817472,
+      "grad_norm": 0.013226517476141453,
+      "learning_rate": 4.732429794558887e-06,
+      "loss": 0.6391,
       "step": 1905
     },
     {
-      "epoch": 0.08854891052387576,
-      "grad_norm": 4.563868999481201,
-      "learning_rate": 1.7709782104775152e-05,
-      "loss": 0.9664,
+      "epoch": 0.7107637919806493,
+      "grad_norm": 0.012964668683707714,
+      "learning_rate": 4.677322065565039e-06,
+      "loss": 0.6768,
       "step": 1910
     },
     {
-      "epoch": 0.08878071395456653,
-      "grad_norm": 4.728658676147461,
-      "learning_rate": 1.7756142790913307e-05,
-      "loss": 0.9351,
+      "epoch": 0.7126244301795516,
+      "grad_norm": 0.01224041823297739,
+      "learning_rate": 4.622438960398234e-06,
+      "loss": 0.645,
       "step": 1915
     },
     {
-      "epoch": 0.0890125173852573,
-      "grad_norm": 5.058661460876465,
-      "learning_rate": 1.7802503477051462e-05,
-      "loss": 1.0738,
+      "epoch": 0.7144850683784538,
+      "grad_norm": 0.013205330818891525,
+      "learning_rate": 4.567782795195816e-06,
+      "loss": 0.6051,
       "step": 1920
     },
     {
-      "epoch": 0.08924432081594807,
-      "grad_norm": 5.769558906555176,
-      "learning_rate": 1.7848864163189618e-05,
-      "loss": 0.8531,
+      "epoch": 0.7163457065773561,
+      "grad_norm": 0.01323428563773632,
+      "learning_rate": 4.5133558765179576e-06,
+      "loss": 0.6113,
       "step": 1925
     },
     {
-      "epoch": 0.08947612424663885,
-      "grad_norm": 5.377395153045654,
-      "learning_rate": 1.7895224849327773e-05,
-      "loss": 0.958,
+      "epoch": 0.7182063447762582,
+      "grad_norm": 0.012786868028342724,
+      "learning_rate": 4.459160501250358e-06,
+      "loss": 0.6677,
       "step": 1930
     },
     {
-      "epoch": 0.08970792767732963,
-      "grad_norm": 4.951549530029297,
-      "learning_rate": 1.7941585535465925e-05,
-      "loss": 0.9338,
+      "epoch": 0.7200669829751605,
+      "grad_norm": 0.010466697625815868,
+      "learning_rate": 4.405198956507272e-06,
+      "loss": 0.586,
       "step": 1935
     },
     {
-      "epoch": 0.0899397311080204,
-      "grad_norm": 4.409064769744873,
-      "learning_rate": 1.798794622160408e-05,
-      "loss": 0.9537,
+      "epoch": 0.7219276211740627,
+      "grad_norm": 0.012410039082169533,
+      "learning_rate": 4.35147351953501e-06,
+      "loss": 0.6111,
       "step": 1940
     },
     {
-      "epoch": 0.09017153453871117,
-      "grad_norm": 6.281864166259766,
-      "learning_rate": 1.8034306907742236e-05,
-      "loss": 1.0452,
+      "epoch": 0.723788259372965,
+      "grad_norm": 0.013155271299183369,
+      "learning_rate": 4.297986457615836e-06,
+      "loss": 0.6404,
       "step": 1945
     },
     {
-      "epoch": 0.09040333796940195,
-      "grad_norm": 5.833890438079834,
-      "learning_rate": 1.808066759388039e-05,
-      "loss": 1.3076,
+      "epoch": 0.7256488975718671,
+      "grad_norm": 0.01472384575754404,
+      "learning_rate": 4.244740027972275e-06,
+      "loss": 0.6553,
       "step": 1950
     },
     {
-      "epoch": 0.09063514140009273,
-      "grad_norm": 4.426032066345215,
-      "learning_rate": 1.8127028280018546e-05,
-      "loss": 1.0576,
+      "epoch": 0.7275095357707694,
+      "grad_norm": 0.012972739525139332,
+      "learning_rate": 4.191736477671864e-06,
+      "loss": 0.6613,
       "step": 1955
     },
     {
-      "epoch": 0.09086694483078349,
-      "grad_norm": 4.635809898376465,
-      "learning_rate": 1.81733889661567e-05,
-      "loss": 0.8691,
+      "epoch": 0.7293701739696716,
+      "grad_norm": 0.013699792325496674,
+      "learning_rate": 4.138978043532332e-06,
+      "loss": 0.6178,
       "step": 1960
     },
     {
-      "epoch": 0.09109874826147427,
-      "grad_norm": 5.994819641113281,
-      "learning_rate": 1.8219749652294857e-05,
-      "loss": 1.1195,
+      "epoch": 0.7312308121685738,
+      "grad_norm": 0.012072841636836529,
+      "learning_rate": 4.086466952027171e-06,
+      "loss": 0.5865,
       "step": 1965
     },
     {
-      "epoch": 0.09133055169216504,
-      "grad_norm": 5.217397689819336,
-      "learning_rate": 1.826611033843301e-05,
-      "loss": 0.9707,
+      "epoch": 0.733091450367476,
+      "grad_norm": 0.013510013930499554,
+      "learning_rate": 4.034205419191709e-06,
+      "loss": 0.6387,
       "step": 1970
     },
     {
-      "epoch": 0.09156235512285582,
-      "grad_norm": 5.001087665557861,
-      "learning_rate": 1.8312471024571164e-05,
-      "loss": 0.9911,
+      "epoch": 0.7349520885663783,
+      "grad_norm": 0.01405192632228136,
+      "learning_rate": 3.982195650529583e-06,
+      "loss": 0.5677,
       "step": 1975
     },
     {
-      "epoch": 0.0917941585535466,
-      "grad_norm": 4.645782470703125,
-      "learning_rate": 1.835883171070932e-05,
-      "loss": 0.8967,
+      "epoch": 0.7368127267652805,
+      "grad_norm": 0.012013067491352558,
+      "learning_rate": 3.930439840919652e-06,
+      "loss": 0.5868,
       "step": 1980
     },
     {
-      "epoch": 0.09202596198423736,
-      "grad_norm": 5.851307392120361,
-      "learning_rate": 1.8405192396847475e-05,
-      "loss": 1.0551,
+      "epoch": 0.7386733649641827,
+      "grad_norm": 0.01283415500074625,
+      "learning_rate": 3.878940174523371e-06,
+      "loss": 0.597,
       "step": 1985
     },
     {
-      "epoch": 0.09225776541492814,
-      "grad_norm": 5.592276096343994,
-      "learning_rate": 1.845155308298563e-05,
-      "loss": 1.1209,
+      "epoch": 0.7405340031630849,
+      "grad_norm": 0.013413486070930958,
+      "learning_rate": 3.827698824692643e-06,
+      "loss": 0.6074,
       "step": 1990
     },
     {
-      "epoch": 0.09248956884561892,
-      "grad_norm": 5.643247604370117,
-      "learning_rate": 1.8497913769123785e-05,
-      "loss": 1.1347,
+      "epoch": 0.7423946413619872,
+      "grad_norm": 0.013137887232005596,
+      "learning_rate": 3.776717953878064e-06,
+      "loss": 0.6599,
       "step": 1995
     },
     {
-      "epoch": 0.0927213722763097,
-      "grad_norm": 6.246877670288086,
-      "learning_rate": 1.854427445526194e-05,
-      "loss": 0.9848,
+      "epoch": 0.7442552795608894,
+      "grad_norm": 0.013435564935207367,
+      "learning_rate": 3.725999713537689e-06,
+      "loss": 0.6191,
       "step": 2000
     },
     {
-      "epoch": 0.0927213722763097,
-      "eval_loss": 1.0543617010116577,
-      "eval_runtime": 11.2709,
-      "eval_samples_per_second": 11.268,
-      "eval_steps_per_second": 11.268,
+      "epoch": 0.7442552795608894,
+      "eval_loss": 0.6428595185279846,
+      "eval_runtime": 26.6061,
+      "eval_samples_per_second": 4.773,
+      "eval_steps_per_second": 4.773,
       "step": 2000
     },
     {
-      "epoch": 0.09295317570700046,
-      "grad_norm": 6.111345291137695,
-      "learning_rate": 1.8590635141400096e-05,
-      "loss": 1.1706,
+      "epoch": 0.7461159177597916,
+      "grad_norm": 0.013417122885584831,
+      "learning_rate": 3.6755462440462288e-06,
+      "loss": 0.6012,
       "step": 2005
     },
     {
-      "epoch": 0.09318497913769123,
-      "grad_norm": 4.53964900970459,
-      "learning_rate": 1.8636995827538248e-05,
-      "loss": 1.0575,
+      "epoch": 0.7479765559586938,
+      "grad_norm": 0.01378430612385273,
+      "learning_rate": 3.625359674604725e-06,
+      "loss": 0.607,
       "step": 2010
     },
     {
-      "epoch": 0.09341678256838201,
-      "grad_norm": 5.4626922607421875,
-      "learning_rate": 1.8683356513676406e-05,
-      "loss": 1.0724,
+      "epoch": 0.7498371941575961,
+      "grad_norm": 0.012614194303750992,
+      "learning_rate": 3.5754421231506953e-06,
+      "loss": 0.6364,
       "step": 2015
     },
     {
-      "epoch": 0.09364858599907279,
-      "grad_norm": 5.114150047302246,
-      "learning_rate": 1.872971719981456e-05,
-      "loss": 0.9194,
+      "epoch": 0.7516978323564982,
+      "grad_norm": 0.013113941997289658,
+      "learning_rate": 3.5257956962687545e-06,
+      "loss": 0.6148,
       "step": 2020
     },
     {
-      "epoch": 0.09388038942976357,
-      "grad_norm": 5.33641242980957,
-      "learning_rate": 1.8776077885952714e-05,
-      "loss": 1.2183,
+      "epoch": 0.7535584705554005,
+      "grad_norm": 0.012877865694463253,
+      "learning_rate": 3.476422489101713e-06,
+      "loss": 0.6369,
       "step": 2025
     },
     {
-      "epoch": 0.09411219286045433,
-      "grad_norm": 5.04241418838501,
-      "learning_rate": 1.882243857209087e-05,
-      "loss": 1.046,
+      "epoch": 0.7554191087543027,
+      "grad_norm": 0.013343026861548424,
+      "learning_rate": 3.427324585262156e-06,
+      "loss": 0.6257,
       "step": 2030
     },
     {
-      "epoch": 0.09434399629114511,
-      "grad_norm": 5.5946431159973145,
-      "learning_rate": 1.886879925822902e-05,
-      "loss": 1.12,
+      "epoch": 0.757279746953205,
+      "grad_norm": 0.01313395518809557,
+      "learning_rate": 3.3785040567445282e-06,
+      "loss": 0.6301,
       "step": 2035
     },
     {
-      "epoch": 0.09457579972183588,
-      "grad_norm": 4.880713939666748,
-      "learning_rate": 1.891515994436718e-05,
-      "loss": 0.9931,
+      "epoch": 0.7591403851521071,
+      "grad_norm": 0.013584131374955177,
+      "learning_rate": 3.329962963837661e-06,
+      "loss": 0.6244,
       "step": 2040
     },
     {
-      "epoch": 0.09480760315252666,
-      "grad_norm": 5.912775993347168,
-      "learning_rate": 1.896152063050533e-05,
-      "loss": 0.9105,
+      "epoch": 0.7610010233510094,
+      "grad_norm": 0.011685586534440517,
+      "learning_rate": 3.281703355037854e-06,
+      "loss": 0.5818,
       "step": 2045
     },
     {
-      "epoch": 0.09503940658321743,
-      "grad_norm": 5.8085198402404785,
-      "learning_rate": 1.9007881316643487e-05,
-      "loss": 1.1281,
+      "epoch": 0.7628616615499116,
+      "grad_norm": 0.013454140163958073,
+      "learning_rate": 3.233727266962425e-06,
+      "loss": 0.6151,
       "step": 2050
     },
     {
-      "epoch": 0.0952712100139082,
-      "grad_norm": 5.165389060974121,
-      "learning_rate": 1.9054242002781642e-05,
-      "loss": 0.9402,
+      "epoch": 0.7647222997488139,
+      "grad_norm": 0.012276149354875088,
+      "learning_rate": 3.186036724263748e-06,
+      "loss": 0.621,
       "step": 2055
     },
     {
-      "epoch": 0.09550301344459898,
-      "grad_norm": 4.676323890686035,
-      "learning_rate": 1.9100602688919797e-05,
-      "loss": 1.0616,
+      "epoch": 0.766582937947716,
+      "grad_norm": 0.013796020299196243,
+      "learning_rate": 3.138633739543805e-06,
+      "loss": 0.6466,
       "step": 2060
     },
     {
-      "epoch": 0.09573481687528976,
-      "grad_norm": 4.0118088722229,
-      "learning_rate": 1.9146963375057953e-05,
-      "loss": 0.8473,
+      "epoch": 0.7684435761466183,
+      "grad_norm": 0.013075289316475391,
+      "learning_rate": 3.0915203132692805e-06,
+      "loss": 0.6116,
       "step": 2065
     },
     {
-      "epoch": 0.09596662030598054,
-      "grad_norm": 5.034766674041748,
-      "learning_rate": 1.9193324061196108e-05,
-      "loss": 1.167,
+      "epoch": 0.7703042143455205,
+      "grad_norm": 0.014096383936703205,
+      "learning_rate": 3.0446984336871144e-06,
+      "loss": 0.5877,
       "step": 2070
     },
     {
-      "epoch": 0.0961984237366713,
-      "grad_norm": 5.5457763671875,
-      "learning_rate": 1.9239684747334263e-05,
-      "loss": 1.3238,
+      "epoch": 0.7721648525444227,
+      "grad_norm": 0.013074836693704128,
+      "learning_rate": 2.998170076740601e-06,
+      "loss": 0.5829,
       "step": 2075
     },
     {
-      "epoch": 0.09643022716736208,
-      "grad_norm": 5.167206764221191,
-      "learning_rate": 1.928604543347242e-05,
-      "loss": 1.2318,
+      "epoch": 0.7740254907433249,
+      "grad_norm": 0.01671191304922104,
+      "learning_rate": 2.951937205986004e-06,
+      "loss": 0.6439,
       "step": 2080
     },
     {
-      "epoch": 0.09666203059805285,
-      "grad_norm": 5.160457134246826,
-      "learning_rate": 1.933240611961057e-05,
-      "loss": 1.0846,
+      "epoch": 0.7758861289422272,
+      "grad_norm": 0.013207321055233479,
+      "learning_rate": 2.9060017725096943e-06,
+      "loss": 0.621,
       "step": 2085
     },
     {
-      "epoch": 0.09689383402874363,
-      "grad_norm": 6.138757705688477,
-      "learning_rate": 1.937876680574873e-05,
-      "loss": 0.84,
+      "epoch": 0.7777467671411294,
+      "grad_norm": 0.014768741093575954,
+      "learning_rate": 2.8603657148458053e-06,
+      "loss": 0.6272,
       "step": 2090
     },
     {
-      "epoch": 0.0971256374594344,
-      "grad_norm": 5.376162052154541,
-      "learning_rate": 1.942512749188688e-05,
-      "loss": 0.8957,
+      "epoch": 0.7796074053400316,
+      "grad_norm": 0.015293040312826633,
+      "learning_rate": 2.8150309588944304e-06,
+      "loss": 0.6388,
       "step": 2095
     },
     {
-      "epoch": 0.09735744089012517,
-      "grad_norm": 6.612864971160889,
-      "learning_rate": 1.9471488178025037e-05,
-      "loss": 0.983,
+      "epoch": 0.7814680435389338,
+      "grad_norm": 0.01271377969533205,
+      "learning_rate": 2.769999417840341e-06,
+      "loss": 0.6249,
       "step": 2100
     },
     {
-      "epoch": 0.09735744089012517,
-      "eval_loss": 1.0578044652938843,
-      "eval_runtime": 11.2841,
-      "eval_samples_per_second": 11.255,
-      "eval_steps_per_second": 11.255,
+      "epoch": 0.7814680435389338,
+      "eval_loss": 0.6426796913146973,
+      "eval_runtime": 26.6067,
+      "eval_samples_per_second": 4.773,
+      "eval_steps_per_second": 4.773,
       "step": 2100
     },
     {
-      "epoch": 0.09758924432081595,
-      "grad_norm": 4.397307395935059,
-      "learning_rate": 1.9517848864163192e-05,
-      "loss": 0.9275,
+      "epoch": 0.7833286817378361,
+      "grad_norm": 0.013834511861205101,
+      "learning_rate": 2.7252729920722564e-06,
+      "loss": 0.6254,
       "step": 2105
     },
     {
-      "epoch": 0.09782104775150673,
-      "grad_norm": 5.631065368652344,
-      "learning_rate": 1.9564209550301344e-05,
-      "loss": 1.1583,
+      "epoch": 0.7851893199367384,
+      "grad_norm": 0.020366957411170006,
+      "learning_rate": 2.680853569102633e-06,
+      "loss": 0.632,
       "step": 2110
     },
     {
-      "epoch": 0.09805285118219749,
-      "grad_norm": 4.9445600509643555,
-      "learning_rate": 1.9610570236439502e-05,
-      "loss": 1.1222,
+      "epoch": 0.7870499581356405,
+      "grad_norm": 0.012656195089221,
+      "learning_rate": 2.6367430234880286e-06,
+      "loss": 0.6274,
       "step": 2115
     },
     {
-      "epoch": 0.09828465461288827,
-      "grad_norm": 5.1628265380859375,
-      "learning_rate": 1.9656930922577654e-05,
-      "loss": 1.2511,
+      "epoch": 0.7889105963345427,
+      "grad_norm": 0.01308165118098259,
+      "learning_rate": 2.5929432167499658e-06,
+      "loss": 0.6457,
       "step": 2120
     },
     {
-      "epoch": 0.09851645804357904,
-      "grad_norm": 4.909499168395996,
-      "learning_rate": 1.970329160871581e-05,
-      "loss": 1.0582,
+      "epoch": 0.790771234533445,
+      "grad_norm": 0.012838170863687992,
+      "learning_rate": 2.5494559972963928e-06,
+      "loss": 0.6436,
       "step": 2125
     },
     {
-      "epoch": 0.09874826147426982,
-      "grad_norm": 6.675722599029541,
-      "learning_rate": 1.9749652294853965e-05,
-      "loss": 1.1201,
+      "epoch": 0.7926318727323471,
+      "grad_norm": 0.012966095469892025,
+      "learning_rate": 2.5062832003436833e-06,
+      "loss": 0.6449,
       "step": 2130
     },
     {
-      "epoch": 0.0989800649049606,
-      "grad_norm": 4.577010154724121,
-      "learning_rate": 1.979601298099212e-05,
-      "loss": 0.8593,
+      "epoch": 0.7944925109312494,
+      "grad_norm": 0.013781096786260605,
+      "learning_rate": 2.463426647839173e-06,
+      "loss": 0.584,
       "step": 2135
     },
     {
-      "epoch": 0.09921186833565136,
-      "grad_norm": 5.183391571044922,
-      "learning_rate": 1.9842373667130276e-05,
-      "loss": 0.8958,
+      "epoch": 0.7963531491301516,
+      "grad_norm": 0.012445746921002865,
+      "learning_rate": 2.420888148384265e-06,
+      "loss": 0.6397,
       "step": 2140
     },
     {
-      "epoch": 0.09944367176634214,
-      "grad_norm": 4.433844089508057,
-      "learning_rate": 1.988873435326843e-05,
-      "loss": 1.0494,
+      "epoch": 0.7982137873290539,
+      "grad_norm": 0.012599524110555649,
+      "learning_rate": 2.378669497158138e-06,
+      "loss": 0.5974,
       "step": 2145
     },
     {
-      "epoch": 0.09967547519703292,
-      "grad_norm": 4.70166015625,
-      "learning_rate": 1.9935095039406583e-05,
-      "loss": 1.0033,
+      "epoch": 0.800074425527956,
+      "grad_norm": 0.011499716900289059,
+      "learning_rate": 2.3367724758419495e-06,
+      "loss": 0.5552,
       "step": 2150
     },
     {
-      "epoch": 0.0999072786277237,
-      "grad_norm": 5.131468772888184,
-      "learning_rate": 1.998145572554474e-05,
-      "loss": 1.0701,
+      "epoch": 0.8019350637268583,
+      "grad_norm": 0.012771397829055786,
+      "learning_rate": 2.2951988525436695e-06,
+      "loss": 0.666,
       "step": 2155
     },
     {
-      "epoch": 0.10013908205841446,
-      "grad_norm": 5.585656642913818,
-      "learning_rate": 1.999999882150718e-05,
-      "loss": 1.1144,
+      "epoch": 0.8037957019257606,
+      "grad_norm": 0.013215843588113785,
+      "learning_rate": 2.2539503817234553e-06,
+      "loss": 0.5925,
       "step": 2160
     },
     {
-      "epoch": 0.10037088548910524,
-      "grad_norm": 4.321651935577393,
-      "learning_rate": 1.999999161960761e-05,
-      "loss": 1.0022,
+      "epoch": 0.8056563401246628,
+      "grad_norm": 0.014011417515575886,
+      "learning_rate": 2.2130288041196135e-06,
+      "loss": 0.6216,
       "step": 2165
     },
     {
-      "epoch": 0.10060268891979601,
-      "grad_norm": 4.363278388977051,
-      "learning_rate": 1.9999977870531412e-05,
-      "loss": 0.8248,
+      "epoch": 0.807516978323565,
+      "grad_norm": 0.014169846661388874,
+      "learning_rate": 2.1724358466751394e-06,
+      "loss": 0.625,
       "step": 2170
     },
     {
-      "epoch": 0.10083449235048679,
-      "grad_norm": 4.530508041381836,
-      "learning_rate": 1.9999957574287587e-05,
-      "loss": 0.8732,
+      "epoch": 0.8093776165224672,
+      "grad_norm": 0.013908619992434978,
+      "learning_rate": 2.132173222464834e-06,
+      "loss": 0.6641,
       "step": 2175
     },
     {
-      "epoch": 0.10106629578117757,
-      "grad_norm": 5.886618137359619,
-      "learning_rate": 1.9999930730889426e-05,
-      "loss": 1.0238,
+      "epoch": 0.8112382547213695,
+      "grad_norm": 0.01255475077778101,
+      "learning_rate": 2.092242630623016e-06,
+      "loss": 0.6135,
       "step": 2180
     },
     {
-      "epoch": 0.10129809921186833,
-      "grad_norm": 5.83167028427124,
-      "learning_rate": 1.99998973403545e-05,
-      "loss": 1.1206,
+      "epoch": 0.8130988929202716,
+      "grad_norm": 0.01320129819214344,
+      "learning_rate": 2.0526457562718074e-06,
+      "loss": 0.5893,
       "step": 2185
     },
     {
-      "epoch": 0.10152990264255911,
-      "grad_norm": 4.87026834487915,
-      "learning_rate": 1.9999857402704675e-05,
-      "loss": 1.1022,
+      "epoch": 0.8149595311191739,
+      "grad_norm": 0.012163982726633549,
+      "learning_rate": 2.013384270450036e-06,
+      "loss": 0.6552,
       "step": 2190
     },
     {
-      "epoch": 0.10176170607324989,
-      "grad_norm": 5.260519504547119,
-      "learning_rate": 1.9999810917966097e-05,
-      "loss": 0.8485,
+      "epoch": 0.8168201693180761,
+      "grad_norm": 0.012990654446184635,
+      "learning_rate": 1.974459830042691e-06,
+      "loss": 0.6309,
       "step": 2195
     },
     {
-      "epoch": 0.10199350950394066,
-      "grad_norm": 5.333953857421875,
-      "learning_rate": 1.99997578861692e-05,
-      "loss": 1.1512,
+      "epoch": 0.8186808075169784,
+      "grad_norm": 0.015373739413917065,
+      "learning_rate": 1.9358740777110154e-06,
+      "loss": 0.6761,
       "step": 2200
     },
     {
-      "epoch": 0.10199350950394066,
-      "eval_loss": 1.0549308061599731,
-      "eval_runtime": 11.2621,
-      "eval_samples_per_second": 11.277,
-      "eval_steps_per_second": 11.277,
+      "epoch": 0.8186808075169784,
+      "eval_loss": 0.6425639390945435,
+      "eval_runtime": 26.5817,
+      "eval_samples_per_second": 4.778,
+      "eval_steps_per_second": 4.778,
       "step": 2200
     },
     {
-      "epoch": 0.10222531293463143,
-      "grad_norm": 4.999898910522461,
-      "learning_rate": 1.9999698307348702e-05,
-      "loss": 1.0005,
+      "epoch": 0.8205414457158805,
+      "grad_norm": 0.012748058885335922,
+      "learning_rate": 1.8976286418231916e-06,
+      "loss": 0.6313,
       "step": 2205
     },
     {
-      "epoch": 0.1024571163653222,
-      "grad_norm": 5.3397932052612305,
-      "learning_rate": 1.9999632181543614e-05,
-      "loss": 1.048,
+      "epoch": 0.8224020839147828,
+      "grad_norm": 0.014108446426689625,
+      "learning_rate": 1.8597251363856061e-06,
+      "loss": 0.5997,
       "step": 2210
     },
     {
-      "epoch": 0.10268891979601298,
-      "grad_norm": 4.57288122177124,
-      "learning_rate": 1.9999559508797233e-05,
-      "loss": 1.0049,
+      "epoch": 0.824262722113685,
+      "grad_norm": 0.0133186811581254,
+      "learning_rate": 1.8221651609747337e-06,
+      "loss": 0.6732,
       "step": 2215
     },
     {
-      "epoch": 0.10292072322670376,
-      "grad_norm": 4.630931854248047,
-      "learning_rate": 1.9999480289157135e-05,
-      "loss": 0.9457,
+      "epoch": 0.8261233603125873,
+      "grad_norm": 0.012968887574970722,
+      "learning_rate": 1.7849503006696566e-06,
+      "loss": 0.6129,
       "step": 2220
     },
     {
-      "epoch": 0.10315252665739454,
-      "grad_norm": 5.614922046661377,
-      "learning_rate": 1.9999394522675186e-05,
-      "loss": 1.082,
+      "epoch": 0.8279839985114894,
+      "grad_norm": 0.012473355047404766,
+      "learning_rate": 1.7480821259851488e-06,
+      "loss": 0.6131,
       "step": 2225
     },
     {
-      "epoch": 0.1033843300880853,
-      "grad_norm": 5.246216297149658,
-      "learning_rate": 1.999930220940754e-05,
-      "loss": 0.9773,
+      "epoch": 0.8298446367103917,
+      "grad_norm": 0.012497167102992535,
+      "learning_rate": 1.7115621928054105e-06,
+      "loss": 0.6199,
       "step": 2230
     },
     {
-      "epoch": 0.10361613351877608,
-      "grad_norm": 4.738265514373779,
-      "learning_rate": 1.9999203349414635e-05,
-      "loss": 0.9073,
+      "epoch": 0.8317052749092939,
+      "grad_norm": 0.014741111546754837,
+      "learning_rate": 1.6753920423184022e-06,
+      "loss": 0.6159,
       "step": 2235
     },
     {
-      "epoch": 0.10384793694946685,
-      "grad_norm": 5.955806255340576,
-      "learning_rate": 1.9999097942761197e-05,
-      "loss": 0.9572,
+      "epoch": 0.8335659131081962,
+      "grad_norm": 0.013057067058980465,
+      "learning_rate": 1.6395732009508058e-06,
+      "loss": 0.5898,
       "step": 2240
     },
     {
-      "epoch": 0.10407974038015763,
-      "grad_norm": 5.307494163513184,
-      "learning_rate": 1.999898598951624e-05,
-      "loss": 1.1539,
+      "epoch": 0.8354265513070983,
+      "grad_norm": 0.014299440197646618,
+      "learning_rate": 1.60410718030361e-06,
+      "loss": 0.6486,
       "step": 2245
     },
     {
-      "epoch": 0.1043115438108484,
-      "grad_norm": 5.4604172706604,
-      "learning_rate": 1.9998867489753057e-05,
-      "loss": 1.1064,
+      "epoch": 0.8372871895060006,
+      "grad_norm": 0.013178296387195587,
+      "learning_rate": 1.568995477088323e-06,
+      "loss": 0.6254,
       "step": 2250
     },
     {
-      "epoch": 0.10454334724153917,
-      "grad_norm": 5.4470319747924805,
-      "learning_rate": 1.9998742443549237e-05,
-      "loss": 1.1014,
+      "epoch": 0.8391478277049028,
+      "grad_norm": 0.013072814792394638,
+      "learning_rate": 1.5342395730637904e-06,
+      "loss": 0.6385,
       "step": 2255
     },
     {
-      "epoch": 0.10477515067222995,
-      "grad_norm": 4.63338565826416,
-      "learning_rate": 1.999861085098665e-05,
-      "loss": 1.0084,
+      "epoch": 0.841008465903805,
+      "grad_norm": 0.014187455177307129,
+      "learning_rate": 1.4998409349736841e-06,
+      "loss": 0.6458,
       "step": 2260
     },
     {
-      "epoch": 0.10500695410292073,
-      "grad_norm": 4.390408039093018,
-      "learning_rate": 1.9998472712151446e-05,
-      "loss": 1.0617,
+      "epoch": 0.8428691041027072,
+      "grad_norm": 0.011844666674733162,
+      "learning_rate": 1.4658010144846001e-06,
+      "loss": 0.6556,
       "step": 2265
     },
     {
-      "epoch": 0.10523875753361149,
-      "grad_norm": 5.576091766357422,
-      "learning_rate": 1.999832802713407e-05,
-      "loss": 1.1452,
+      "epoch": 0.8447297423016095,
+      "grad_norm": 0.013700997456908226,
+      "learning_rate": 1.432121248124786e-06,
+      "loss": 0.6381,
       "step": 2270
     },
     {
-      "epoch": 0.10547056096430227,
-      "grad_norm": 5.20993185043335,
-      "learning_rate": 1.9998176796029253e-05,
-      "loss": 1.0616,
+      "epoch": 0.8465903805005117,
+      "grad_norm": 0.014204096049070358,
+      "learning_rate": 1.3988030572235212e-06,
+      "loss": 0.624,
       "step": 2275
     },
     {
-      "epoch": 0.10570236439499305,
-      "grad_norm": 4.806192874908447,
-      "learning_rate": 1.9998019018936007e-05,
-      "loss": 1.0339,
+      "epoch": 0.8484510186994139,
+      "grad_norm": 0.013039126060903072,
+      "learning_rate": 1.3658478478511416e-06,
+      "loss": 0.593,
       "step": 2280
     },
     {
-      "epoch": 0.10593416782568382,
-      "grad_norm": 4.287393569946289,
-      "learning_rate": 1.999785469595763e-05,
-      "loss": 0.9953,
+      "epoch": 0.8503116568983161,
+      "grad_norm": 0.014153816737234592,
+      "learning_rate": 1.333257010759702e-06,
+      "loss": 0.5909,
       "step": 2285
     },
     {
-      "epoch": 0.1061659712563746,
-      "grad_norm": 5.184103488922119,
-      "learning_rate": 1.999768382720171e-05,
-      "loss": 0.9038,
+      "epoch": 0.8521722950972184,
+      "grad_norm": 0.012623702175915241,
+      "learning_rate": 1.3010319213242762e-06,
+      "loss": 0.6006,
       "step": 2290
     },
     {
-      "epoch": 0.10639777468706536,
-      "grad_norm": 6.341686725616455,
-      "learning_rate": 1.9997506412780117e-05,
-      "loss": 1.1514,
+      "epoch": 0.8540329332961206,
+      "grad_norm": 0.013684497214853764,
+      "learning_rate": 1.2691739394849089e-06,
+      "loss": 0.6112,
       "step": 2295
     },
     {
-      "epoch": 0.10662957811775614,
-      "grad_norm": 5.4436750411987305,
-      "learning_rate": 1.9997322452809003e-05,
-      "loss": 1.1623,
+      "epoch": 0.8558935714950228,
+      "grad_norm": 0.012923210859298706,
+      "learning_rate": 1.2376844096892526e-06,
+      "loss": 0.6239,
       "step": 2300
     },
     {
-      "epoch": 0.10662957811775614,
-      "eval_loss": 1.0587830543518066,
-      "eval_runtime": 11.261,
-      "eval_samples_per_second": 11.278,
-      "eval_steps_per_second": 11.278,
+      "epoch": 0.8558935714950228,
+      "eval_loss": 0.6425209641456604,
+      "eval_runtime": 26.5582,
+      "eval_samples_per_second": 4.782,
+      "eval_steps_per_second": 4.782,
       "step": 2300
     },
     {
-      "epoch": 0.10686138154844692,
-      "grad_norm": 5.4230427742004395,
-      "learning_rate": 1.9997131947408818e-05,
-      "loss": 0.967,
+      "epoch": 0.857754209693925,
+      "grad_norm": 0.01325127761811018,
+      "learning_rate": 1.2065646608357972e-06,
+      "loss": 0.6537,
       "step": 2305
     },
     {
-      "epoch": 0.1070931849791377,
-      "grad_norm": 4.601907253265381,
-      "learning_rate": 1.9996934896704282e-05,
-      "loss": 1.1419,
+      "epoch": 0.8596148478928273,
+      "grad_norm": 0.01215402316302061,
+      "learning_rate": 1.1758160062178093e-06,
+      "loss": 0.634,
       "step": 2310
     },
     {
-      "epoch": 0.10732498840982846,
-      "grad_norm": 4.844226837158203,
-      "learning_rate": 1.999673130082441e-05,
-      "loss": 0.9171,
+      "epoch": 0.8614754860917294,
+      "grad_norm": 0.015918321907520294,
+      "learning_rate": 1.1454397434679022e-06,
+      "loss": 0.6296,
       "step": 2315
     },
     {
-      "epoch": 0.10755679184051924,
-      "grad_norm": 5.482775688171387,
-      "learning_rate": 1.9996521159902502e-05,
-      "loss": 1.1142,
+      "epoch": 0.8633361242906317,
+      "grad_norm": 0.011971482075750828,
+      "learning_rate": 1.1154371545032738e-06,
+      "loss": 0.5983,
       "step": 2320
     },
     {
-      "epoch": 0.10778859527121001,
-      "grad_norm": 5.575770854949951,
-      "learning_rate": 1.999630447407614e-05,
-      "loss": 1.0038,
+      "epoch": 0.8651967624895339,
+      "grad_norm": 0.01542737614363432,
+      "learning_rate": 1.0858095054716111e-06,
+      "loss": 0.6468,
       "step": 2325
     },
     {
-      "epoch": 0.10802039870190079,
-      "grad_norm": 4.347164630889893,
-      "learning_rate": 1.9996081243487194e-05,
-      "loss": 0.9837,
+      "epoch": 0.8670574006884362,
+      "grad_norm": 0.012823596596717834,
+      "learning_rate": 1.0565580466976566e-06,
+      "loss": 0.6222,
       "step": 2330
     },
     {
-      "epoch": 0.10825220213259157,
-      "grad_norm": 5.454910755157471,
-      "learning_rate": 1.999585146828181e-05,
-      "loss": 1.11,
+      "epoch": 0.8689180388873383,
+      "grad_norm": 0.012822597287595272,
+      "learning_rate": 1.027684012630441e-06,
+      "loss": 0.6385,
       "step": 2335
     },
     {
-      "epoch": 0.10848400556328233,
-      "grad_norm": 4.132719039916992,
-      "learning_rate": 1.9995615148610436e-05,
-      "loss": 0.8614,
+      "epoch": 0.8707786770862406,
+      "grad_norm": 0.014365943148732185,
+      "learning_rate": 9.991886217911851e-07,
+      "loss": 0.6541,
       "step": 2340
     },
     {
-      "epoch": 0.10871580899397311,
-      "grad_norm": 4.780175685882568,
-      "learning_rate": 1.999537228462779e-05,
-      "loss": 0.9514,
+      "epoch": 0.8726393152851428,
+      "grad_norm": 0.013635417446494102,
+      "learning_rate": 9.710730767218913e-07,
+      "loss": 0.6323,
       "step": 2345
     },
     {
-      "epoch": 0.10894761242466389,
-      "grad_norm": 5.469396114349365,
-      "learning_rate": 1.9995122876492873e-05,
-      "loss": 0.8916,
+      "epoch": 0.8744999534840451,
+      "grad_norm": 0.012839280068874359,
+      "learning_rate": 9.433385639345705e-07,
+      "loss": 0.5944,
       "step": 2350
     },
     {
-      "epoch": 0.10917941585535466,
-      "grad_norm": 5.702508449554443,
-      "learning_rate": 1.9994866924368988e-05,
-      "loss": 1.1915,
+      "epoch": 0.8763605916829472,
+      "grad_norm": 0.0141257019713521,
+      "learning_rate": 9.159862538611908e-07,
+      "loss": 0.6632,
       "step": 2355
     },
     {
-      "epoch": 0.10941121928604543,
-      "grad_norm": 4.359443187713623,
-      "learning_rate": 1.9994604428423706e-05,
-      "loss": 0.9538,
+      "epoch": 0.8782212298818495,
+      "grad_norm": 0.013571621850132942,
+      "learning_rate": 8.890173008042768e-07,
+      "loss": 0.6086,
       "step": 2360
     },
     {
-      "epoch": 0.1096430227167362,
-      "grad_norm": 5.570619106292725,
-      "learning_rate": 1.999433538882889e-05,
-      "loss": 0.9006,
+      "epoch": 0.8800818680807517,
+      "grad_norm": 0.01433682069182396,
+      "learning_rate": 8.624328428881945e-07,
+      "loss": 0.6482,
       "step": 2365
     },
     {
-      "epoch": 0.10987482614742698,
-      "grad_norm": 4.556485176086426,
-      "learning_rate": 1.999405980576068e-05,
-      "loss": 1.0132,
+      "epoch": 0.8819425062796539,
+      "grad_norm": 0.014090972021222115,
+      "learning_rate": 8.36234002011117e-07,
+      "loss": 0.6335,
       "step": 2370
     },
     {
-      "epoch": 0.11010662957811776,
-      "grad_norm": 4.931366443634033,
-      "learning_rate": 1.999377767939951e-05,
-      "loss": 0.9837,
+      "epoch": 0.8838031444785561,
+      "grad_norm": 0.013004067353904247,
+      "learning_rate": 8.10421883797694e-07,
+      "loss": 0.6145,
       "step": 2375
     },
     {
-      "epoch": 0.11033843300880854,
-      "grad_norm": 5.512187480926514,
-      "learning_rate": 1.9993489009930097e-05,
-      "loss": 0.9279,
+      "epoch": 0.8856637826774584,
+      "grad_norm": 0.01391025260090828,
+      "learning_rate": 7.849975775523777e-07,
+      "loss": 0.6415,
       "step": 2380
     },
     {
-      "epoch": 0.1105702364394993,
-      "grad_norm": 5.395881652832031,
-      "learning_rate": 1.999319379754143e-05,
-      "loss": 1.0187,
+      "epoch": 0.8875244208763606,
+      "grad_norm": 0.012305272743105888,
+      "learning_rate": 7.599621562134596e-07,
+      "loss": 0.6462,
       "step": 2385
     },
     {
-      "epoch": 0.11080203987019008,
-      "grad_norm": 5.412979602813721,
-      "learning_rate": 1.9992892042426794e-05,
-      "loss": 1.1533,
+      "epoch": 0.8893850590752628,
+      "grad_norm": 0.013947544619441032,
+      "learning_rate": 7.35316676307789e-07,
+      "loss": 0.6266,
       "step": 2390
     },
     {
-      "epoch": 0.11103384330088086,
-      "grad_norm": 4.9955735206604,
-      "learning_rate": 1.9992583744783754e-05,
-      "loss": 1.0025,
+      "epoch": 0.891245697274165,
+      "grad_norm": 0.016055511310696602,
+      "learning_rate": 7.110621779061889e-07,
+      "loss": 0.6501,
       "step": 2395
     },
     {
-      "epoch": 0.11126564673157163,
-      "grad_norm": 5.231233596801758,
-      "learning_rate": 1.9992268904814154e-05,
-      "loss": 1.0544,
+      "epoch": 0.8931063354730673,
+      "grad_norm": 0.01238927897065878,
+      "learning_rate": 6.871996845795581e-07,
+      "loss": 0.5895,
       "step": 2400
     },
     {
-      "epoch": 0.11126564673157163,
-      "eval_loss": 1.052207589149475,
-      "eval_runtime": 11.2784,
-      "eval_samples_per_second": 11.26,
-      "eval_steps_per_second": 11.26,
+      "epoch": 0.8931063354730673,
+      "eval_loss": 0.6424703598022461,
+      "eval_runtime": 26.6049,
+      "eval_samples_per_second": 4.774,
+      "eval_steps_per_second": 4.774,
       "step": 2400
     },
     {
-      "epoch": 0.1114974501622624,
-      "grad_norm": 4.9348015785217285,
-      "learning_rate": 1.999194752272413e-05,
-      "loss": 1.1822,
+      "epoch": 0.8949669736719695,
+      "grad_norm": 0.012343904934823513,
+      "learning_rate": 6.637302033556891e-07,
+      "loss": 0.6105,
       "step": 2405
     },
     {
-      "epoch": 0.11172925359295317,
-      "grad_norm": 4.779728412628174,
-      "learning_rate": 1.9991619598724096e-05,
-      "loss": 1.08,
+      "epoch": 0.8968276118708717,
+      "grad_norm": 0.01289752684533596,
+      "learning_rate": 6.40654724676748e-07,
+      "loss": 0.6265,
       "step": 2410
     },
     {
-      "epoch": 0.11196105702364395,
-      "grad_norm": 4.476454734802246,
-      "learning_rate": 1.999128513302875e-05,
-      "loss": 1.0141,
+      "epoch": 0.8986882500697739,
+      "grad_norm": 0.013334677554666996,
+      "learning_rate": 6.179742223574936e-07,
+      "loss": 0.6261,
       "step": 2415
     },
     {
-      "epoch": 0.11219286045433473,
-      "grad_norm": 4.818485736846924,
-      "learning_rate": 1.999094412585707e-05,
-      "loss": 0.9972,
+      "epoch": 0.9005488882686762,
+      "grad_norm": 0.013512643985450268,
+      "learning_rate": 5.956896535441803e-07,
+      "loss": 0.5797,
       "step": 2420
     },
     {
-      "epoch": 0.11242466388502549,
-      "grad_norm": 4.220043182373047,
-      "learning_rate": 1.9990596577432325e-05,
-      "loss": 0.9463,
+      "epoch": 0.9024095264675783,
+      "grad_norm": 0.01403987966477871,
+      "learning_rate": 5.738019586741573e-07,
+      "loss": 0.616,
       "step": 2425
     },
     {
-      "epoch": 0.11265646731571627,
-      "grad_norm": 4.722184181213379,
-      "learning_rate": 1.9990242487982056e-05,
-      "loss": 1.0425,
+      "epoch": 0.9042701646664806,
+      "grad_norm": 0.013224196620285511,
+      "learning_rate": 5.523120614361821e-07,
+      "loss": 0.608,
       "step": 2430
     },
     {
-      "epoch": 0.11288827074640705,
-      "grad_norm": 4.847781658172607,
-      "learning_rate": 1.9989881857738093e-05,
-      "loss": 1.0659,
+      "epoch": 0.9061308028653828,
+      "grad_norm": 0.01317799836397171,
+      "learning_rate": 5.312208687314502e-07,
+      "loss": 0.6206,
       "step": 2435
     },
     {
-      "epoch": 0.11312007417709782,
-      "grad_norm": 4.4079813957214355,
-      "learning_rate": 1.9989514686936545e-05,
-      "loss": 1.0153,
+      "epoch": 0.9079914410642851,
+      "grad_norm": 0.01357815321534872,
+      "learning_rate": 5.105292706353093e-07,
+      "loss": 0.6759,
       "step": 2440
     },
     {
-      "epoch": 0.1133518776077886,
-      "grad_norm": 3.875089168548584,
-      "learning_rate": 1.9989140975817813e-05,
-      "loss": 0.9605,
+      "epoch": 0.9098520792631872,
+      "grad_norm": 0.012360199354588985,
+      "learning_rate": 4.902381403597046e-07,
+      "loss": 0.6182,
       "step": 2445
     },
     {
-      "epoch": 0.11358368103847936,
-      "grad_norm": 4.662908554077148,
-      "learning_rate": 1.9988760724626564e-05,
-      "loss": 1.0402,
+      "epoch": 0.9117127174620895,
+      "grad_norm": 0.012978832237422466,
+      "learning_rate": 4.703483342163262e-07,
+      "loss": 0.5892,
       "step": 2450
     },
     {
-      "epoch": 0.11381548446917014,
-      "grad_norm": 5.138006687164307,
-      "learning_rate": 1.998837393361176e-05,
-      "loss": 1.1952,
+      "epoch": 0.9135733556609917,
+      "grad_norm": 0.013323403894901276,
+      "learning_rate": 4.5086069158047143e-07,
+      "loss": 0.636,
       "step": 2455
     },
     {
-      "epoch": 0.11404728789986092,
-      "grad_norm": 5.774982452392578,
-      "learning_rate": 1.998798060302664e-05,
-      "loss": 0.9564,
+      "epoch": 0.915433993859894,
+      "grad_norm": 0.013127562589943409,
+      "learning_rate": 4.3177603485562327e-07,
+      "loss": 0.5847,
       "step": 2460
     },
     {
-      "epoch": 0.1142790913305517,
-      "grad_norm": 4.6387553215026855,
-      "learning_rate": 1.998758073312872e-05,
-      "loss": 0.9307,
+      "epoch": 0.9172946320587961,
+      "grad_norm": 0.012560434639453888,
+      "learning_rate": 4.1309516943874196e-07,
+      "loss": 0.6073,
       "step": 2465
     },
     {
-      "epoch": 0.11451089476124246,
-      "grad_norm": 4.885598659515381,
-      "learning_rate": 1.9987174324179814e-05,
-      "loss": 0.8062,
+      "epoch": 0.9191552702576984,
+      "grad_norm": 0.01229447964578867,
+      "learning_rate": 3.9481888368627764e-07,
+      "loss": 0.5846,
       "step": 2470
     },
     {
-      "epoch": 0.11474269819193324,
-      "grad_norm": 5.628473281860352,
-      "learning_rate": 1.9986761376445992e-05,
-      "loss": 1.1763,
+      "epoch": 0.9210159084566006,
+      "grad_norm": 0.012554515153169632,
+      "learning_rate": 3.7694794888090025e-07,
+      "loss": 0.6186,
       "step": 2475
     },
     {
-      "epoch": 0.11497450162262401,
-      "grad_norm": 5.386045932769775,
-      "learning_rate": 1.9986341890197622e-05,
-      "loss": 0.9168,
+      "epoch": 0.9228765466555028,
+      "grad_norm": 0.011958773247897625,
+      "learning_rate": 3.594831191989523e-07,
+      "loss": 0.6217,
       "step": 2480
     },
     {
-      "epoch": 0.11520630505331479,
-      "grad_norm": 4.536678791046143,
-      "learning_rate": 1.9985915865709352e-05,
-      "loss": 1.0792,
+      "epoch": 0.924737184854405,
+      "grad_norm": 0.01367896143347025,
+      "learning_rate": 3.424251316786165e-07,
+      "loss": 0.6572,
       "step": 2485
     },
     {
-      "epoch": 0.11543810848400557,
-      "grad_norm": 5.205998420715332,
-      "learning_rate": 1.998548330326011e-05,
-      "loss": 1.1665,
+      "epoch": 0.9265978230533073,
+      "grad_norm": 0.013063084334135056,
+      "learning_rate": 3.2577470618881726e-07,
+      "loss": 0.5973,
       "step": 2490
     },
     {
-      "epoch": 0.11566991191469633,
-      "grad_norm": 3.94810152053833,
-      "learning_rate": 1.9985044203133096e-05,
-      "loss": 1.0869,
+      "epoch": 0.9284584612522095,
+      "grad_norm": 0.012596881948411465,
+      "learning_rate": 3.095325453988385e-07,
+      "loss": 0.6153,
       "step": 2495
     },
     {
-      "epoch": 0.11590171534538711,
-      "grad_norm": 4.870349407196045,
-      "learning_rate": 1.99845985656158e-05,
-      "loss": 1.0749,
+      "epoch": 0.9303190994511117,
+      "grad_norm": 0.013333328068256378,
+      "learning_rate": 2.9369933474867496e-07,
+      "loss": 0.6337,
       "step": 2500
     },
     {
-      "epoch": 0.11590171534538711,
-      "eval_loss": 1.0539507865905762,
-      "eval_runtime": 11.2662,
-      "eval_samples_per_second": 11.273,
-      "eval_steps_per_second": 11.273,
+      "epoch": 0.9303190994511117,
+      "eval_loss": 0.6424322128295898,
+      "eval_runtime": 26.5954,
+      "eval_samples_per_second": 4.775,
+      "eval_steps_per_second": 4.775,
       "step": 2500
     },
     {
-      "epoch": 0.11613351877607789,
-      "grad_norm": 4.539355278015137,
-      "learning_rate": 1.9984146390999993e-05,
-      "loss": 0.9024,
+      "epoch": 0.9321797376500139,
+      "grad_norm": 0.012263627722859383,
+      "learning_rate": 2.7827574242009434e-07,
+      "loss": 0.6258,
       "step": 2505
     },
     {
-      "epoch": 0.11636532220676866,
-      "grad_norm": 5.9270219802856445,
-      "learning_rate": 1.9983687679581717e-05,
-      "loss": 1.0351,
+      "epoch": 0.9340403758489162,
+      "grad_norm": 0.01295017171651125,
+      "learning_rate": 2.632624193084499e-07,
+      "loss": 0.6071,
       "step": 2510
     },
     {
-      "epoch": 0.11659712563745943,
-      "grad_norm": 5.80509614944458,
-      "learning_rate": 1.99832224316613e-05,
-      "loss": 0.9956,
+      "epoch": 0.9359010140478184,
+      "grad_norm": 0.013952870853245258,
+      "learning_rate": 2.48659998995211e-07,
+      "loss": 0.6594,
       "step": 2515
     },
     {
-      "epoch": 0.1168289290681502,
-      "grad_norm": 4.957856178283691,
-      "learning_rate": 1.9982750647543347e-05,
-      "loss": 1.0678,
+      "epoch": 0.9377616522467206,
+      "grad_norm": 0.01351676881313324,
+      "learning_rate": 2.344690977212205e-07,
+      "loss": 0.6072,
       "step": 2520
     },
     {
-      "epoch": 0.11706073249884098,
-      "grad_norm": 5.07832670211792,
-      "learning_rate": 1.9982272327536748e-05,
-      "loss": 1.209,
+      "epoch": 0.9396222904456228,
+      "grad_norm": 0.012661050073802471,
+      "learning_rate": 2.2069031436068643e-07,
+      "loss": 0.6152,
       "step": 2525
     },
     {
-      "epoch": 0.11729253592953176,
-      "grad_norm": 5.460022926330566,
-      "learning_rate": 1.9981787471954663e-05,
-      "loss": 0.8889,
+      "epoch": 0.9414829286445251,
+      "grad_norm": 0.013936568051576614,
+      "learning_rate": 2.0732423039591998e-07,
+      "loss": 0.6149,
       "step": 2530
     },
     {
-      "epoch": 0.11752433936022254,
-      "grad_norm": 5.039451599121094,
-      "learning_rate": 1.9981296081114535e-05,
-      "loss": 1.1051,
+      "epoch": 0.9433435668434274,
+      "grad_norm": 0.012601389549672604,
+      "learning_rate": 1.9437140989278624e-07,
+      "loss": 0.6205,
       "step": 2535
     },
     {
-      "epoch": 0.1177561427909133,
-      "grad_norm": 5.151773452758789,
-      "learning_rate": 1.9980798155338092e-05,
-      "loss": 1.0199,
+      "epoch": 0.9452042050423295,
+      "grad_norm": 0.011978083290159702,
+      "learning_rate": 1.8183239947690112e-07,
+      "loss": 0.6252,
       "step": 2540
     },
     {
-      "epoch": 0.11798794622160408,
-      "grad_norm": 5.761076927185059,
-      "learning_rate": 1.9980293694951333e-05,
-      "loss": 1.0285,
+      "epoch": 0.9470648432412317,
+      "grad_norm": 0.012075323611497879,
+      "learning_rate": 1.6970772831056637e-07,
+      "loss": 0.6438,
       "step": 2545
     },
     {
-      "epoch": 0.11821974965229486,
-      "grad_norm": 5.472416400909424,
-      "learning_rate": 1.9979782700284532e-05,
-      "loss": 0.9648,
+      "epoch": 0.948925481440134,
+      "grad_norm": 0.012837120331823826,
+      "learning_rate": 1.5799790807043857e-07,
+      "loss": 0.6304,
       "step": 2550
     },
     {
-      "epoch": 0.11845155308298563,
-      "grad_norm": 3.8476479053497314,
-      "learning_rate": 1.997926517167225e-05,
-      "loss": 1.0118,
+      "epoch": 0.9507861196390361,
+      "grad_norm": 0.012269029393792152,
+      "learning_rate": 1.467034329259287e-07,
+      "loss": 0.5938,
       "step": 2555
     },
     {
-      "epoch": 0.1186833565136764,
-      "grad_norm": 4.49735164642334,
-      "learning_rate": 1.997874110945333e-05,
-      "loss": 1.0989,
+      "epoch": 0.9526467578379384,
+      "grad_norm": 0.01306453812867403,
+      "learning_rate": 1.358247795183587e-07,
+      "loss": 0.5981,
       "step": 2560
     },
     {
-      "epoch": 0.11891515994436717,
-      "grad_norm": 3.996696710586548,
-      "learning_rate": 1.9978210513970877e-05,
-      "loss": 0.8726,
+      "epoch": 0.9545073960368406,
+      "grad_norm": 0.012171389535069466,
+      "learning_rate": 1.2536240694083658e-07,
+      "loss": 0.5599,
       "step": 2565
     },
     {
-      "epoch": 0.11914696337505795,
-      "grad_norm": 4.320407390594482,
-      "learning_rate": 1.9977673385572284e-05,
-      "loss": 0.9728,
+      "epoch": 0.9563680342357429,
+      "grad_norm": 0.013338044285774231,
+      "learning_rate": 1.1531675671888621e-07,
+      "loss": 0.6366,
       "step": 2570
     },
     {
-      "epoch": 0.11937876680574873,
-      "grad_norm": 7.207583427429199,
-      "learning_rate": 1.9977129724609214e-05,
-      "loss": 1.1305,
+      "epoch": 0.958228672434645,
+      "grad_norm": 0.012387475930154324,
+      "learning_rate": 1.0568825279181572e-07,
+      "loss": 0.6376,
       "step": 2575
     },
     {
-      "epoch": 0.11961057023643949,
-      "grad_norm": 5.2364821434021,
-      "learning_rate": 1.9976579531437616e-05,
-      "loss": 1.0501,
+      "epoch": 0.9600893106335473,
+      "grad_norm": 0.013054094277322292,
+      "learning_rate": 9.647730149482614e-08,
+      "loss": 0.6386,
       "step": 2580
     },
     {
-      "epoch": 0.11984237366713027,
-      "grad_norm": 4.274256706237793,
-      "learning_rate": 1.9976022806417714e-05,
-      "loss": 1.0153,
+      "epoch": 0.9619499488324496,
+      "grad_norm": 0.01319506112486124,
+      "learning_rate": 8.768429154185853e-08,
+      "loss": 0.6067,
       "step": 2585
     },
     {
-      "epoch": 0.12007417709782105,
-      "grad_norm": 5.584084510803223,
-      "learning_rate": 1.9975459549914e-05,
-      "loss": 0.974,
+      "epoch": 0.9638105870313518,
+      "grad_norm": 0.013576803728938103,
+      "learning_rate": 7.930959400919924e-08,
+      "loss": 0.6258,
       "step": 2590
     },
     {
-      "epoch": 0.12030598052851182,
-      "grad_norm": 5.529228687286377,
-      "learning_rate": 1.9974889762295254e-05,
-      "loss": 1.0517,
+      "epoch": 0.965671225230254,
+      "grad_norm": 0.013341645710170269,
+      "learning_rate": 7.135356231981028e-08,
+      "loss": 0.6344,
       "step": 2595
     },
     {
-      "epoch": 0.1205377839592026,
-      "grad_norm": 4.252645015716553,
-      "learning_rate": 1.997431344393452e-05,
-      "loss": 1.0758,
+      "epoch": 0.9675318634291562,
+      "grad_norm": 0.013685373589396477,
+      "learning_rate": 6.381653222842011e-08,
+      "loss": 0.6144,
       "step": 2600
     },
     {
-      "epoch": 0.1205377839592026,
-      "eval_loss": 1.0526373386383057,
-      "eval_runtime": 11.2657,
-      "eval_samples_per_second": 11.273,
-      "eval_steps_per_second": 11.273,
+      "epoch": 0.9675318634291562,
+      "eval_loss": 0.642424464225769,
+      "eval_runtime": 26.5716,
+      "eval_samples_per_second": 4.78,
+      "eval_steps_per_second": 4.78,
       "step": 2600
     },
     {
-      "epoch": 0.12076958738989337,
-      "grad_norm": 4.629552364349365,
-      "learning_rate": 1.9973730595209134e-05,
-      "loss": 1.0771,
+      "epoch": 0.9693925016280585,
+      "grad_norm": 0.018206071108579636,
+      "learning_rate": 5.6698821807354975e-08,
+      "loss": 0.5716,
       "step": 2605
     },
     {
-      "epoch": 0.12100139082058414,
-      "grad_norm": 4.153298377990723,
-      "learning_rate": 1.9973141216500685e-05,
-      "loss": 1.1439,
+      "epoch": 0.9712531398269606,
+      "grad_norm": 0.014082850888371468,
+      "learning_rate": 5.000073143310969e-08,
+      "loss": 0.641,
       "step": 2610
     },
     {
-      "epoch": 0.12123319425127492,
-      "grad_norm": 4.761916637420654,
-      "learning_rate": 1.997254530819506e-05,
-      "loss": 1.0997,
+      "epoch": 0.9731137780258629,
+      "grad_norm": 0.01310745719820261,
+      "learning_rate": 4.3722543773681016e-08,
+      "loss": 0.6206,
       "step": 2615
     },
     {
-      "epoch": 0.1214649976819657,
-      "grad_norm": 6.3712544441223145,
-      "learning_rate": 1.9971942870682404e-05,
-      "loss": 1.0405,
+      "epoch": 0.9749744162247651,
+      "grad_norm": 0.013652559369802475,
+      "learning_rate": 3.7864523776628414e-08,
+      "loss": 0.6405,
       "step": 2620
     },
     {
-      "epoch": 0.12169680111265646,
-      "grad_norm": 5.3124518394470215,
-      "learning_rate": 1.9971333904357152e-05,
-      "loss": 1.1737,
+      "epoch": 0.9768350544236674,
+      "grad_norm": 0.01430275198072195,
+      "learning_rate": 3.242691865790071e-08,
+      "loss": 0.6191,
       "step": 2625
     },
     {
-      "epoch": 0.12192860454334724,
-      "grad_norm": 4.761479377746582,
-      "learning_rate": 1.9970718409618e-05,
-      "loss": 1.0439,
+      "epoch": 0.9786956926225695,
+      "grad_norm": 0.013149100355803967,
+      "learning_rate": 2.7409957891397775e-08,
+      "loss": 0.6676,
       "step": 2630
     },
     {
-      "epoch": 0.12216040797403802,
-      "grad_norm": 5.314613342285156,
-      "learning_rate": 1.9970096386867922e-05,
-      "loss": 1.0017,
+      "epoch": 0.9805563308214718,
+      "grad_norm": 0.01429106667637825,
+      "learning_rate": 2.2813853199292745e-08,
+      "loss": 0.656,
       "step": 2635
     },
     {
-      "epoch": 0.12239221140472879,
-      "grad_norm": 5.509566307067871,
-      "learning_rate": 1.996946783651417e-05,
-      "loss": 1.0059,
+      "epoch": 0.982416969020374,
+      "grad_norm": 0.012532561086118221,
+      "learning_rate": 1.8638798543090253e-08,
+      "loss": 0.6065,
       "step": 2640
     },
     {
-      "epoch": 0.12262401483541957,
-      "grad_norm": 5.131368637084961,
-      "learning_rate": 1.996883275896827e-05,
-      "loss": 0.8546,
+      "epoch": 0.9842776072192763,
+      "grad_norm": 0.01296111661940813,
+      "learning_rate": 1.4884970115444097e-08,
+      "loss": 0.604,
       "step": 2645
     },
     {
-      "epoch": 0.12285581826611033,
-      "grad_norm": 4.782067775726318,
-      "learning_rate": 1.9968191154646012e-05,
-      "loss": 0.9647,
+      "epoch": 0.9861382454181784,
+      "grad_norm": 0.012804310768842697,
+      "learning_rate": 1.1552526332723191e-08,
+      "loss": 0.6236,
       "step": 2650
     },
     {
-      "epoch": 0.12308762169680111,
-      "grad_norm": 5.031082630157471,
-      "learning_rate": 1.996754302396747e-05,
-      "loss": 1.0906,
+      "epoch": 0.9879988836170807,
+      "grad_norm": 0.012863220646977425,
+      "learning_rate": 8.641607828324682e-09,
+      "loss": 0.6245,
       "step": 2655
     },
     {
-      "epoch": 0.12331942512749189,
-      "grad_norm": 4.008341312408447,
-      "learning_rate": 1.996688836735699e-05,
-      "loss": 1.0645,
+      "epoch": 0.9898595218159829,
+      "grad_norm": 0.013312343508005142,
+      "learning_rate": 6.152337446736489e-09,
+      "loss": 0.6561,
       "step": 2660
     },
     {
-      "epoch": 0.12355122855818267,
-      "grad_norm": 4.894102096557617,
-      "learning_rate": 1.9966227185243184e-05,
-      "loss": 0.9666,
+      "epoch": 0.991720160014885,
+      "grad_norm": 0.011470803059637547,
+      "learning_rate": 4.0848202383581e-09,
+      "loss": 0.5847,
       "step": 2665
     },
     {
-      "epoch": 0.12378303198887343,
-      "grad_norm": 4.749420166015625,
-      "learning_rate": 1.996555947805894e-05,
-      "loss": 1.2034,
+      "epoch": 0.9935807982137873,
+      "grad_norm": 0.013984150253236294,
+      "learning_rate": 2.4391434550652403e-09,
+      "loss": 0.5881,
       "step": 2670
     },
     {
-      "epoch": 0.1240148354195642,
-      "grad_norm": 5.051076889038086,
-      "learning_rate": 1.9964885246241417e-05,
-      "loss": 0.9919,
+      "epoch": 0.9954414364126896,
+      "grad_norm": 0.012489933520555496,
+      "learning_rate": 1.2153765465250378e-09,
+      "loss": 0.6653,
       "step": 2675
     },
     {
-      "epoch": 0.12424663885025498,
-      "grad_norm": 5.225306987762451,
-      "learning_rate": 1.996420449023205e-05,
-      "loss": 0.9872,
+      "epoch": 0.9973020746115918,
+      "grad_norm": 0.012180610559880733,
+      "learning_rate": 4.1357115726947674e-10,
+      "loss": 0.629,
       "step": 2680
     },
     {
-      "epoch": 0.12447844228094576,
-      "grad_norm": 4.617201328277588,
-      "learning_rate": 1.9963517210476543e-05,
-      "loss": 1.0147,
+      "epoch": 0.999162712810494,
+      "grad_norm": 0.014413848519325256,
+      "learning_rate": 3.376112451158875e-11,
+      "loss": 0.6525,
       "step": 2685
     },
     {
-      "epoch": 0.12471024571163654,
-      "grad_norm": 4.746542453765869,
-      "learning_rate": 1.9962823407424868e-05,
-      "loss": 0.8903,
-      "step": 2690
-    },
-    {
-      "epoch": 0.1249420491423273,
-      "grad_norm": 4.461585521697998,
-      "learning_rate": 1.9962123081531268e-05,
-      "loss": 1.0642,
-      "step": 2695
-    },
-    {
-      "epoch": 0.12517385257301808,
-      "grad_norm": 4.527534008026123,
-      "learning_rate": 1.996141623325426e-05,
-      "loss": 1.001,
-      "step": 2700
-    },
-    {
-      "epoch": 0.12517385257301808,
-      "eval_loss": 1.0466318130493164,
-      "eval_runtime": 11.2744,
-      "eval_samples_per_second": 11.264,
-      "eval_steps_per_second": 11.264,
-      "step": 2700
-    },
-    {
-      "epoch": 0.12540565600370884,
-      "grad_norm": 5.0327982902526855,
-      "learning_rate": 1.996070286305664e-05,
-      "loss": 0.9662,
-      "step": 2705
-    },
-    {
-      "epoch": 0.12563745943439963,
-      "grad_norm": 4.870184421539307,
-      "learning_rate": 1.9959982971405454e-05,
-      "loss": 0.9642,
-      "step": 2710
-    },
-    {
-      "epoch": 0.1258692628650904,
-      "grad_norm": 4.229808807373047,
-      "learning_rate": 1.995925655877203e-05,
-      "loss": 1.1574,
-      "step": 2715
-    },
-    {
-      "epoch": 0.1261010662957812,
-      "grad_norm": 4.645317077636719,
-      "learning_rate": 1.9958523625631964e-05,
-      "loss": 0.8886,
-      "step": 2720
-    },
-    {
-      "epoch": 0.12633286972647195,
-      "grad_norm": 4.935416221618652,
-      "learning_rate": 1.9957784172465124e-05,
-      "loss": 0.9621,
-      "step": 2725
-    },
-    {
-      "epoch": 0.12656467315716272,
-      "grad_norm": 4.857961654663086,
-      "learning_rate": 1.9957038199755637e-05,
-      "loss": 1.0029,
-      "step": 2730
-    },
-    {
-      "epoch": 0.1267964765878535,
-      "grad_norm": 5.052342891693115,
-      "learning_rate": 1.9956285707991913e-05,
-      "loss": 1.0644,
-      "step": 2735
-    },
-    {
-      "epoch": 0.12702828001854427,
-      "grad_norm": 4.7845354080200195,
-      "learning_rate": 1.9955526697666615e-05,
-      "loss": 1.0014,
-      "step": 2740
-    },
-    {
-      "epoch": 0.12726008344923506,
-      "grad_norm": 5.164909362792969,
-      "learning_rate": 1.9954761169276686e-05,
-      "loss": 1.0553,
-      "step": 2745
-    },
-    {
-      "epoch": 0.12749188687992583,
-      "grad_norm": 4.757340431213379,
-      "learning_rate": 1.9953989123323327e-05,
-      "loss": 1.1321,
-      "step": 2750
-    },
-    {
-      "epoch": 0.1277236903106166,
-      "grad_norm": 5.6578688621521,
-      "learning_rate": 1.995321056031201e-05,
-      "loss": 0.9902,
-      "step": 2755
-    },
-    {
-      "epoch": 0.12795549374130738,
-      "grad_norm": 4.892259120941162,
-      "learning_rate": 1.9952425480752483e-05,
-      "loss": 1.061,
-      "step": 2760
-    },
-    {
-      "epoch": 0.12818729717199814,
-      "grad_norm": 5.429172992706299,
-      "learning_rate": 1.9951633885158745e-05,
-      "loss": 1.0976,
-      "step": 2765
-    },
-    {
-      "epoch": 0.1284191006026889,
-      "grad_norm": 3.9769465923309326,
-      "learning_rate": 1.9950835774049066e-05,
-      "loss": 0.9902,
-      "step": 2770
-    },
-    {
-      "epoch": 0.1286509040333797,
-      "grad_norm": 5.092761993408203,
-      "learning_rate": 1.995003114794599e-05,
-      "loss": 1.114,
-      "step": 2775
-    },
-    {
-      "epoch": 0.12888270746407046,
-      "grad_norm": 5.163008689880371,
-      "learning_rate": 1.9949220007376312e-05,
-      "loss": 1.2063,
-      "step": 2780
-    },
-    {
-      "epoch": 0.12911451089476125,
-      "grad_norm": 4.68748140335083,
-      "learning_rate": 1.994840235287111e-05,
-      "loss": 1.1674,
-      "step": 2785
-    },
-    {
-      "epoch": 0.12934631432545202,
-      "grad_norm": 5.405124664306641,
-      "learning_rate": 1.9947578184965712e-05,
-      "loss": 1.0991,
-      "step": 2790
-    },
-    {
-      "epoch": 0.12957811775614278,
-      "grad_norm": 5.273200511932373,
-      "learning_rate": 1.9946747504199718e-05,
-      "loss": 0.8909,
-      "step": 2795
-    },
-    {
-      "epoch": 0.12980992118683357,
-      "grad_norm": 4.340039253234863,
-      "learning_rate": 1.9945910311116986e-05,
-      "loss": 0.9397,
-      "step": 2800
-    },
-    {
-      "epoch": 0.12980992118683357,
-      "eval_loss": 1.0486783981323242,
-      "eval_runtime": 11.2737,
-      "eval_samples_per_second": 11.265,
-      "eval_steps_per_second": 11.265,
-      "step": 2800
-    },
-    {
-      "epoch": 0.13004172461752433,
-      "grad_norm": 4.297397136688232,
-      "learning_rate": 1.9945066606265645e-05,
-      "loss": 0.9343,
-      "step": 2805
-    },
-    {
-      "epoch": 0.13027352804821513,
-      "grad_norm": 4.575728416442871,
-      "learning_rate": 1.9944216390198082e-05,
-      "loss": 0.9489,
-      "step": 2810
-    },
-    {
-      "epoch": 0.1305053314789059,
-      "grad_norm": 5.547889232635498,
-      "learning_rate": 1.9943359663470954e-05,
-      "loss": 1.0477,
-      "step": 2815
-    },
-    {
-      "epoch": 0.13073713490959665,
-      "grad_norm": 5.181078910827637,
-      "learning_rate": 1.994249642664517e-05,
-      "loss": 1.0363,
-      "step": 2820
-    },
-    {
-      "epoch": 0.13096893834028744,
-      "grad_norm": 4.966522693634033,
-      "learning_rate": 1.9941626680285906e-05,
-      "loss": 1.0665,
-      "step": 2825
-    },
-    {
-      "epoch": 0.1312007417709782,
-      "grad_norm": 6.055009841918945,
-      "learning_rate": 1.9940750424962603e-05,
-      "loss": 0.9614,
-      "step": 2830
-    },
-    {
-      "epoch": 0.13143254520166897,
-      "grad_norm": 4.3711628913879395,
-      "learning_rate": 1.9939867661248964e-05,
-      "loss": 1.0237,
-      "step": 2835
-    },
-    {
-      "epoch": 0.13166434863235976,
-      "grad_norm": 4.811582565307617,
-      "learning_rate": 1.993897838972295e-05,
-      "loss": 0.9741,
-      "step": 2840
-    },
-    {
-      "epoch": 0.13189615206305053,
-      "grad_norm": 4.890336036682129,
-      "learning_rate": 1.9938082610966775e-05,
-      "loss": 0.9146,
-      "step": 2845
-    },
-    {
-      "epoch": 0.13212795549374132,
-      "grad_norm": 4.723270416259766,
-      "learning_rate": 1.9937180325566934e-05,
-      "loss": 1.1937,
-      "step": 2850
-    },
-    {
-      "epoch": 0.13235975892443208,
-      "grad_norm": 3.966965675354004,
-      "learning_rate": 1.9936271534114157e-05,
-      "loss": 0.9725,
-      "step": 2855
-    },
-    {
-      "epoch": 0.13259156235512284,
-      "grad_norm": 5.717922687530518,
-      "learning_rate": 1.9935356237203457e-05,
-      "loss": 1.0297,
-      "step": 2860
-    },
-    {
-      "epoch": 0.13282336578581364,
-      "grad_norm": 5.139970302581787,
-      "learning_rate": 1.9934434435434095e-05,
-      "loss": 1.0494,
-      "step": 2865
-    },
-    {
-      "epoch": 0.1330551692165044,
-      "grad_norm": 4.475926399230957,
-      "learning_rate": 1.9933506129409583e-05,
-      "loss": 0.928,
-      "step": 2870
-    },
-    {
-      "epoch": 0.1332869726471952,
-      "grad_norm": 4.684488296508789,
-      "learning_rate": 1.9932571319737706e-05,
-      "loss": 1.1672,
-      "step": 2875
-    },
-    {
-      "epoch": 0.13351877607788595,
-      "grad_norm": 4.571727275848389,
-      "learning_rate": 1.99316300070305e-05,
-      "loss": 0.9693,
-      "step": 2880
-    },
-    {
-      "epoch": 0.13375057950857672,
-      "grad_norm": 3.749621629714966,
-      "learning_rate": 1.9930682191904262e-05,
-      "loss": 0.8496,
-      "step": 2885
-    },
-    {
-      "epoch": 0.1339823829392675,
-      "grad_norm": 4.513935565948486,
-      "learning_rate": 1.9929727874979536e-05,
-      "loss": 1.1165,
-      "step": 2890
-    },
-    {
-      "epoch": 0.13421418636995827,
-      "grad_norm": 4.019639492034912,
-      "learning_rate": 1.9928767056881137e-05,
-      "loss": 1.0072,
-      "step": 2895
-    },
-    {
-      "epoch": 0.13444598980064906,
-      "grad_norm": 5.7689948081970215,
-      "learning_rate": 1.9927799738238128e-05,
-      "loss": 1.0774,
-      "step": 2900
-    },
-    {
-      "epoch": 0.13444598980064906,
-      "eval_loss": 1.0431135892868042,
-      "eval_runtime": 11.2686,
-      "eval_samples_per_second": 11.27,
-      "eval_steps_per_second": 11.27,
-      "step": 2900
-    },
-    {
-      "epoch": 0.13467779323133983,
-      "grad_norm": 4.070760726928711,
-      "learning_rate": 1.9926825919683836e-05,
-      "loss": 0.9854,
-      "step": 2905
-    },
-    {
-      "epoch": 0.1349095966620306,
-      "grad_norm": 4.487087249755859,
-      "learning_rate": 1.992584560185583e-05,
-      "loss": 0.9863,
-      "step": 2910
-    },
-    {
-      "epoch": 0.13514140009272138,
-      "grad_norm": 4.818826675415039,
-      "learning_rate": 1.9924858785395943e-05,
-      "loss": 0.9824,
-      "step": 2915
-    },
-    {
-      "epoch": 0.13537320352341214,
-      "grad_norm": 4.67568826675415,
-      "learning_rate": 1.9923865470950266e-05,
-      "loss": 1.0129,
-      "step": 2920
-    },
-    {
-      "epoch": 0.1356050069541029,
-      "grad_norm": 4.589109897613525,
-      "learning_rate": 1.9922865659169133e-05,
-      "loss": 1.073,
-      "step": 2925
-    },
-    {
-      "epoch": 0.1358368103847937,
-      "grad_norm": 4.670289039611816,
-      "learning_rate": 1.992185935070715e-05,
-      "loss": 1.058,
-      "step": 2930
-    },
-    {
-      "epoch": 0.13606861381548446,
-      "grad_norm": 4.4648542404174805,
-      "learning_rate": 1.9920846546223155e-05,
-      "loss": 0.8634,
-      "step": 2935
-    },
-    {
-      "epoch": 0.13630041724617525,
-      "grad_norm": 5.334418296813965,
-      "learning_rate": 1.9919827246380254e-05,
-      "loss": 0.9094,
-      "step": 2940
-    },
-    {
-      "epoch": 0.13653222067686602,
-      "grad_norm": 4.683016777038574,
-      "learning_rate": 1.9918801451845798e-05,
-      "loss": 0.9466,
-      "step": 2945
-    },
-    {
-      "epoch": 0.13676402410755678,
-      "grad_norm": 5.748208045959473,
-      "learning_rate": 1.99177691632914e-05,
-      "loss": 1.061,
-      "step": 2950
-    },
-    {
-      "epoch": 0.13699582753824757,
-      "grad_norm": 4.584070205688477,
-      "learning_rate": 1.9916730381392913e-05,
-      "loss": 1.0056,
-      "step": 2955
-    },
-    {
-      "epoch": 0.13722763096893834,
-      "grad_norm": 5.5280561447143555,
-      "learning_rate": 1.9915685106830444e-05,
-      "loss": 1.0739,
-      "step": 2960
-    },
-    {
-      "epoch": 0.13745943439962913,
-      "grad_norm": 4.314013481140137,
-      "learning_rate": 1.9914633340288362e-05,
-      "loss": 1.0037,
-      "step": 2965
-    },
-    {
-      "epoch": 0.1376912378303199,
-      "grad_norm": 5.107608795166016,
-      "learning_rate": 1.991357508245527e-05,
-      "loss": 1.0779,
-      "step": 2970
-    },
-    {
-      "epoch": 0.13792304126101065,
-      "grad_norm": 5.307722091674805,
-      "learning_rate": 1.9912510334024034e-05,
-      "loss": 1.1289,
-      "step": 2975
-    },
-    {
-      "epoch": 0.13815484469170144,
-      "grad_norm": 6.47243595123291,
-      "learning_rate": 1.991143909569176e-05,
-      "loss": 1.1177,
-      "step": 2980
-    },
-    {
-      "epoch": 0.1383866481223922,
-      "grad_norm": 5.38716459274292,
-      "learning_rate": 1.9910361368159803e-05,
-      "loss": 1.0031,
-      "step": 2985
-    },
-    {
-      "epoch": 0.13861845155308297,
-      "grad_norm": 4.5754194259643555,
-      "learning_rate": 1.990927715213378e-05,
-      "loss": 1.0479,
-      "step": 2990
-    },
-    {
-      "epoch": 0.13885025498377376,
-      "grad_norm": 5.582353591918945,
-      "learning_rate": 1.9908186448323546e-05,
-      "loss": 0.9586,
-      "step": 2995
-    },
-    {
-      "epoch": 0.13908205841446453,
-      "grad_norm": 7.415589809417725,
-      "learning_rate": 1.99070892574432e-05,
-      "loss": 0.9956,
-      "step": 3000
-    },
-    {
-      "epoch": 0.13908205841446453,
-      "eval_loss": 1.0443520545959473,
-      "eval_runtime": 11.2822,
-      "eval_samples_per_second": 11.257,
-      "eval_steps_per_second": 11.257,
-      "step": 3000
-    },
-    {
-      "epoch": 0.13931386184515532,
-      "grad_norm": 4.259057521820068,
-      "learning_rate": 1.99059855802111e-05,
-      "loss": 0.9444,
-      "step": 3005
-    },
-    {
-      "epoch": 0.13954566527584608,
-      "grad_norm": 4.621634006500244,
-      "learning_rate": 1.990487541734983e-05,
-      "loss": 0.9122,
-      "step": 3010
-    },
-    {
-      "epoch": 0.13977746870653684,
-      "grad_norm": 4.235317707061768,
-      "learning_rate": 1.9903758769586247e-05,
-      "loss": 1.0184,
-      "step": 3015
-    },
-    {
-      "epoch": 0.14000927213722764,
-      "grad_norm": 4.515482425689697,
-      "learning_rate": 1.9902635637651435e-05,
-      "loss": 1.0877,
-      "step": 3020
-    },
-    {
-      "epoch": 0.1402410755679184,
-      "grad_norm": 4.261685371398926,
-      "learning_rate": 1.990150602228073e-05,
-      "loss": 1.034,
-      "step": 3025
-    },
-    {
-      "epoch": 0.1404728789986092,
-      "grad_norm": 5.3254289627075195,
-      "learning_rate": 1.990036992421371e-05,
-      "loss": 1.1411,
-      "step": 3030
-    },
-    {
-      "epoch": 0.14070468242929995,
-      "grad_norm": 5.06146764755249,
-      "learning_rate": 1.9899227344194204e-05,
-      "loss": 0.9846,
-      "step": 3035
-    },
-    {
-      "epoch": 0.14093648585999072,
-      "grad_norm": 4.180016994476318,
-      "learning_rate": 1.9898078282970274e-05,
-      "loss": 1.0153,
-      "step": 3040
-    },
-    {
-      "epoch": 0.1411682892906815,
-      "grad_norm": 4.471304893493652,
-      "learning_rate": 1.989692274129424e-05,
-      "loss": 1.0025,
-      "step": 3045
-    },
-    {
-      "epoch": 0.14140009272137227,
-      "grad_norm": 4.3698811531066895,
-      "learning_rate": 1.9895760719922644e-05,
-      "loss": 0.9501,
-      "step": 3050
-    },
-    {
-      "epoch": 0.14163189615206306,
-      "grad_norm": 4.26015567779541,
-      "learning_rate": 1.989459221961629e-05,
-      "loss": 1.0604,
-      "step": 3055
-    },
-    {
-      "epoch": 0.14186369958275383,
-      "grad_norm": 6.512139320373535,
-      "learning_rate": 1.9893417241140215e-05,
-      "loss": 1.0173,
-      "step": 3060
-    },
-    {
-      "epoch": 0.1420955030134446,
-      "grad_norm": 3.8038687705993652,
-      "learning_rate": 1.98922357852637e-05,
-      "loss": 1.0969,
-      "step": 3065
-    },
-    {
-      "epoch": 0.14232730644413538,
-      "grad_norm": 4.5116143226623535,
-      "learning_rate": 1.9891047852760264e-05,
-      "loss": 1.0758,
-      "step": 3070
-    },
-    {
-      "epoch": 0.14255910987482615,
-      "grad_norm": 5.065197944641113,
-      "learning_rate": 1.9889853444407666e-05,
-      "loss": 1.2611,
-      "step": 3075
-    },
-    {
-      "epoch": 0.1427909133055169,
-      "grad_norm": 5.6167778968811035,
-      "learning_rate": 1.9888652560987915e-05,
-      "loss": 0.9903,
-      "step": 3080
-    },
-    {
-      "epoch": 0.1430227167362077,
-      "grad_norm": 4.782162666320801,
-      "learning_rate": 1.9887445203287242e-05,
-      "loss": 1.0061,
-      "step": 3085
-    },
-    {
-      "epoch": 0.14325452016689846,
-      "grad_norm": 4.842726230621338,
-      "learning_rate": 1.9886231372096133e-05,
-      "loss": 1.2085,
-      "step": 3090
-    },
-    {
-      "epoch": 0.14348632359758925,
-      "grad_norm": 4.182819843292236,
-      "learning_rate": 1.98850110682093e-05,
-      "loss": 0.9167,
-      "step": 3095
-    },
-    {
-      "epoch": 0.14371812702828002,
-      "grad_norm": 4.709084510803223,
-      "learning_rate": 1.98837842924257e-05,
-      "loss": 1.0333,
-      "step": 3100
-    },
-    {
-      "epoch": 0.14371812702828002,
-      "eval_loss": 1.0421561002731323,
-      "eval_runtime": 11.2658,
-      "eval_samples_per_second": 11.273,
-      "eval_steps_per_second": 11.273,
-      "step": 3100
-    },
-    {
-      "epoch": 0.14394993045897078,
-      "grad_norm": 4.043697834014893,
-      "learning_rate": 1.988255104554853e-05,
-      "loss": 1.1004,
-      "step": 3105
-    },
-    {
-      "epoch": 0.14418173388966157,
-      "grad_norm": 5.348196029663086,
-      "learning_rate": 1.9881311328385213e-05,
-      "loss": 1.152,
-      "step": 3110
-    },
-    {
-      "epoch": 0.14441353732035234,
-      "grad_norm": 4.3684821128845215,
-      "learning_rate": 1.9880065141747417e-05,
-      "loss": 1.0153,
-      "step": 3115
-    },
-    {
-      "epoch": 0.14464534075104313,
-      "grad_norm": 4.7167253494262695,
-      "learning_rate": 1.987881248645104e-05,
-      "loss": 0.9573,
-      "step": 3120
-    },
-    {
-      "epoch": 0.1448771441817339,
-      "grad_norm": 5.030665874481201,
-      "learning_rate": 1.987755336331622e-05,
-      "loss": 1.0171,
-      "step": 3125
-    },
-    {
-      "epoch": 0.14510894761242465,
-      "grad_norm": 5.634954929351807,
-      "learning_rate": 1.9876287773167333e-05,
-      "loss": 0.9415,
-      "step": 3130
-    },
-    {
-      "epoch": 0.14534075104311545,
-      "grad_norm": 4.053722858428955,
-      "learning_rate": 1.9875015716832978e-05,
-      "loss": 1.039,
-      "step": 3135
-    },
-    {
-      "epoch": 0.1455725544738062,
-      "grad_norm": 6.290874481201172,
-      "learning_rate": 1.9873737195145993e-05,
-      "loss": 1.0583,
-      "step": 3140
-    },
-    {
-      "epoch": 0.14580435790449697,
-      "grad_norm": 4.683097839355469,
-      "learning_rate": 1.9872452208943453e-05,
-      "loss": 1.0219,
-      "step": 3145
-    },
-    {
-      "epoch": 0.14603616133518776,
-      "grad_norm": 5.337057590484619,
-      "learning_rate": 1.9871160759066662e-05,
-      "loss": 0.9462,
-      "step": 3150
-    },
-    {
-      "epoch": 0.14626796476587853,
-      "grad_norm": 5.12256383895874,
-      "learning_rate": 1.9869862846361152e-05,
-      "loss": 0.9631,
-      "step": 3155
-    },
-    {
-      "epoch": 0.14649976819656932,
-      "grad_norm": 4.650876522064209,
-      "learning_rate": 1.9868558471676693e-05,
-      "loss": 1.1056,
-      "step": 3160
-    },
-    {
-      "epoch": 0.14673157162726008,
-      "grad_norm": 5.170388698577881,
-      "learning_rate": 1.9867247635867285e-05,
-      "loss": 1.2239,
-      "step": 3165
-    },
-    {
-      "epoch": 0.14696337505795085,
-      "grad_norm": 3.694185495376587,
-      "learning_rate": 1.9865930339791147e-05,
-      "loss": 0.924,
-      "step": 3170
-    },
-    {
-      "epoch": 0.14719517848864164,
-      "grad_norm": 7.866819858551025,
-      "learning_rate": 1.986460658431075e-05,
-      "loss": 1.1734,
-      "step": 3175
-    },
-    {
-      "epoch": 0.1474269819193324,
-      "grad_norm": 4.429941654205322,
-      "learning_rate": 1.9863276370292766e-05,
-      "loss": 1.0446,
-      "step": 3180
-    },
-    {
-      "epoch": 0.1476587853500232,
-      "grad_norm": 5.235711097717285,
-      "learning_rate": 1.9861939698608123e-05,
-      "loss": 1.053,
-      "step": 3185
-    },
-    {
-      "epoch": 0.14789058878071396,
-      "grad_norm": 4.496561527252197,
-      "learning_rate": 1.986059657013196e-05,
-      "loss": 1.0764,
-      "step": 3190
-    },
-    {
-      "epoch": 0.14812239221140472,
-      "grad_norm": 4.243244647979736,
-      "learning_rate": 1.9859246985743646e-05,
-      "loss": 0.9841,
-      "step": 3195
-    },
-    {
-      "epoch": 0.1483541956420955,
-      "grad_norm": 4.924827575683594,
-      "learning_rate": 1.985789094632678e-05,
-      "loss": 0.9833,
-      "step": 3200
-    },
-    {
-      "epoch": 0.1483541956420955,
-      "eval_loss": 1.0363918542861938,
-      "eval_runtime": 11.2718,
-      "eval_samples_per_second": 11.267,
-      "eval_steps_per_second": 11.267,
-      "step": 3200
-    },
-    {
-      "epoch": 0.14858599907278627,
-      "grad_norm": 4.187185287475586,
-      "learning_rate": 1.9856528452769182e-05,
-      "loss": 0.941,
-      "step": 3205
-    },
-    {
-      "epoch": 0.14881780250347706,
-      "grad_norm": 4.020816802978516,
-      "learning_rate": 1.9855159505962907e-05,
-      "loss": 0.95,
-      "step": 3210
-    },
-    {
-      "epoch": 0.14904960593416783,
-      "grad_norm": 4.336887359619141,
-      "learning_rate": 1.9853784106804227e-05,
-      "loss": 0.7706,
-      "step": 3215
-    },
-    {
-      "epoch": 0.1492814093648586,
-      "grad_norm": 4.014132976531982,
-      "learning_rate": 1.9852402256193638e-05,
-      "loss": 0.917,
-      "step": 3220
-    },
-    {
-      "epoch": 0.14951321279554938,
-      "grad_norm": 3.605818510055542,
-      "learning_rate": 1.9851013955035868e-05,
-      "loss": 1.0043,
-      "step": 3225
-    },
-    {
-      "epoch": 0.14974501622624015,
-      "grad_norm": 4.072787284851074,
-      "learning_rate": 1.9849619204239863e-05,
-      "loss": 0.9526,
-      "step": 3230
-    },
-    {
-      "epoch": 0.1499768196569309,
-      "grad_norm": 4.913269519805908,
-      "learning_rate": 1.9848218004718783e-05,
-      "loss": 1.0963,
-      "step": 3235
-    },
-    {
-      "epoch": 0.1502086230876217,
-      "grad_norm": 4.653753280639648,
-      "learning_rate": 1.9846810357390033e-05,
-      "loss": 1.1541,
-      "step": 3240
-    },
-    {
-      "epoch": 0.15044042651831246,
-      "grad_norm": 4.507899761199951,
-      "learning_rate": 1.9845396263175214e-05,
-      "loss": 1.108,
-      "step": 3245
-    },
-    {
-      "epoch": 0.15067222994900326,
-      "grad_norm": 6.352303504943848,
-      "learning_rate": 1.984397572300016e-05,
-      "loss": 0.9772,
-      "step": 3250
-    },
-    {
-      "epoch": 0.15090403337969402,
-      "grad_norm": 4.838869094848633,
-      "learning_rate": 1.984254873779493e-05,
-      "loss": 1.0617,
-      "step": 3255
-    },
-    {
-      "epoch": 0.15113583681038478,
-      "grad_norm": 4.51560115814209,
-      "learning_rate": 1.9841115308493797e-05,
-      "loss": 0.7006,
-      "step": 3260
-    },
-    {
-      "epoch": 0.15136764024107557,
-      "grad_norm": 5.014158725738525,
-      "learning_rate": 1.9839675436035245e-05,
-      "loss": 0.9927,
-      "step": 3265
-    },
-    {
-      "epoch": 0.15159944367176634,
-      "grad_norm": 4.209564685821533,
-      "learning_rate": 1.9838229121361995e-05,
-      "loss": 1.0364,
-      "step": 3270
-    },
-    {
-      "epoch": 0.15183124710245713,
-      "grad_norm": 4.790988445281982,
-      "learning_rate": 1.983677636542097e-05,
-      "loss": 0.9416,
-      "step": 3275
-    },
-    {
-      "epoch": 0.1520630505331479,
-      "grad_norm": 3.9592926502227783,
-      "learning_rate": 1.9835317169163314e-05,
-      "loss": 1.053,
-      "step": 3280
-    },
-    {
-      "epoch": 0.15229485396383866,
-      "grad_norm": 3.99648118019104,
-      "learning_rate": 1.9833851533544396e-05,
-      "loss": 0.9706,
-      "step": 3285
-    },
-    {
-      "epoch": 0.15252665739452945,
-      "grad_norm": 4.372700214385986,
-      "learning_rate": 1.9832379459523793e-05,
-      "loss": 0.9676,
-      "step": 3290
-    },
-    {
-      "epoch": 0.1527584608252202,
-      "grad_norm": 5.660383224487305,
-      "learning_rate": 1.983090094806529e-05,
-      "loss": 0.9255,
-      "step": 3295
-    },
-    {
-      "epoch": 0.15299026425591097,
-      "grad_norm": 4.162616729736328,
-      "learning_rate": 1.9829416000136905e-05,
-      "loss": 1.0264,
-      "step": 3300
-    },
-    {
-      "epoch": 0.15299026425591097,
-      "eval_loss": 1.038940668106079,
-      "eval_runtime": 11.2668,
-      "eval_samples_per_second": 11.272,
-      "eval_steps_per_second": 11.272,
-      "step": 3300
-    },
-    {
-      "epoch": 0.15322206768660177,
-      "grad_norm": 4.065362453460693,
-      "learning_rate": 1.9827924616710858e-05,
-      "loss": 0.909,
-      "step": 3305
-    },
-    {
-      "epoch": 0.15345387111729253,
-      "grad_norm": 4.5612592697143555,
-      "learning_rate": 1.9826426798763577e-05,
-      "loss": 1.2016,
-      "step": 3310
-    },
-    {
-      "epoch": 0.15368567454798332,
-      "grad_norm": 4.262414932250977,
-      "learning_rate": 1.9824922547275726e-05,
-      "loss": 0.9091,
-      "step": 3315
-    },
-    {
-      "epoch": 0.15391747797867408,
-      "grad_norm": 5.065141201019287,
-      "learning_rate": 1.9823411863232155e-05,
-      "loss": 1.0608,
-      "step": 3320
-    },
-    {
-      "epoch": 0.15414928140936485,
-      "grad_norm": 4.390530109405518,
-      "learning_rate": 1.9821894747621936e-05,
-      "loss": 1.1014,
-      "step": 3325
-    },
-    {
-      "epoch": 0.15438108484005564,
-      "grad_norm": 5.454830169677734,
-      "learning_rate": 1.9820371201438358e-05,
-      "loss": 1.0328,
-      "step": 3330
-    },
-    {
-      "epoch": 0.1546128882707464,
-      "grad_norm": 3.939898729324341,
-      "learning_rate": 1.981884122567891e-05,
-      "loss": 0.9734,
-      "step": 3335
-    },
-    {
-      "epoch": 0.1548446917014372,
-      "grad_norm": 4.209561824798584,
-      "learning_rate": 1.98173048213453e-05,
-      "loss": 0.7803,
-      "step": 3340
-    },
-    {
-      "epoch": 0.15507649513212796,
-      "grad_norm": 4.415044784545898,
-      "learning_rate": 1.981576198944343e-05,
-      "loss": 0.9674,
-      "step": 3345
-    },
-    {
-      "epoch": 0.15530829856281872,
-      "grad_norm": 4.023775100708008,
-      "learning_rate": 1.981421273098343e-05,
-      "loss": 0.9258,
-      "step": 3350
-    },
-    {
-      "epoch": 0.1555401019935095,
-      "grad_norm": 4.346290588378906,
-      "learning_rate": 1.9812657046979625e-05,
-      "loss": 1.1134,
-      "step": 3355
-    },
-    {
-      "epoch": 0.15577190542420027,
-      "grad_norm": 3.649505853652954,
-      "learning_rate": 1.9811094938450548e-05,
-      "loss": 0.9605,
-      "step": 3360
-    },
-    {
-      "epoch": 0.15600370885489107,
-      "grad_norm": 5.957110404968262,
-      "learning_rate": 1.9809526406418943e-05,
-      "loss": 1.1297,
-      "step": 3365
-    },
-    {
-      "epoch": 0.15623551228558183,
-      "grad_norm": 4.172776699066162,
-      "learning_rate": 1.980795145191175e-05,
-      "loss": 1.0034,
-      "step": 3370
-    },
-    {
-      "epoch": 0.1564673157162726,
-      "grad_norm": 4.410923004150391,
-      "learning_rate": 1.9806370075960128e-05,
-      "loss": 1.0254,
-      "step": 3375
-    },
-    {
-      "epoch": 0.15669911914696338,
-      "grad_norm": 4.170181751251221,
-      "learning_rate": 1.9804782279599424e-05,
-      "loss": 1.1244,
-      "step": 3380
-    },
-    {
-      "epoch": 0.15693092257765415,
-      "grad_norm": 4.1127119064331055,
-      "learning_rate": 1.980318806386921e-05,
-      "loss": 0.9039,
-      "step": 3385
-    },
-    {
-      "epoch": 0.1571627260083449,
-      "grad_norm": 4.409833908081055,
-      "learning_rate": 1.9801587429813237e-05,
-      "loss": 1.0573,
-      "step": 3390
-    },
-    {
-      "epoch": 0.1573945294390357,
-      "grad_norm": 4.1619086265563965,
-      "learning_rate": 1.979998037847947e-05,
-      "loss": 0.9595,
-      "step": 3395
-    },
-    {
-      "epoch": 0.15762633286972647,
-      "grad_norm": 4.631487846374512,
-      "learning_rate": 1.9798366910920082e-05,
-      "loss": 1.0714,
-      "step": 3400
-    },
-    {
-      "epoch": 0.15762633286972647,
-      "eval_loss": 1.0348224639892578,
-      "eval_runtime": 11.2693,
-      "eval_samples_per_second": 11.27,
-      "eval_steps_per_second": 11.27,
-      "step": 3400
-    },
-    {
-      "epoch": 0.15785813630041726,
-      "grad_norm": 3.5335793495178223,
-      "learning_rate": 1.9796747028191433e-05,
-      "loss": 0.6737,
-      "step": 3405
-    },
-    {
-      "epoch": 0.15808993973110802,
-      "grad_norm": 4.976848602294922,
-      "learning_rate": 1.9795120731354092e-05,
-      "loss": 1.1097,
-      "step": 3410
-    },
-    {
-      "epoch": 0.15832174316179878,
-      "grad_norm": 4.47586727142334,
-      "learning_rate": 1.9793488021472827e-05,
-      "loss": 0.9923,
-      "step": 3415
-    },
-    {
-      "epoch": 0.15855354659248957,
-      "grad_norm": 4.697174072265625,
-      "learning_rate": 1.97918488996166e-05,
-      "loss": 1.0532,
-      "step": 3420
-    },
-    {
-      "epoch": 0.15878535002318034,
-      "grad_norm": 4.547231674194336,
-      "learning_rate": 1.9790203366858572e-05,
-      "loss": 0.9237,
-      "step": 3425
-    },
-    {
-      "epoch": 0.15901715345387113,
-      "grad_norm": 4.481298923492432,
-      "learning_rate": 1.978855142427611e-05,
-      "loss": 1.1366,
-      "step": 3430
-    },
-    {
-      "epoch": 0.1592489568845619,
-      "grad_norm": 4.639220714569092,
-      "learning_rate": 1.9786893072950766e-05,
-      "loss": 1.0086,
-      "step": 3435
-    },
-    {
-      "epoch": 0.15948076031525266,
-      "grad_norm": 4.7241339683532715,
-      "learning_rate": 1.978522831396829e-05,
-      "loss": 1.0335,
-      "step": 3440
-    },
-    {
-      "epoch": 0.15971256374594345,
-      "grad_norm": 6.761361598968506,
-      "learning_rate": 1.978355714841864e-05,
-      "loss": 1.0096,
-      "step": 3445
-    },
-    {
-      "epoch": 0.1599443671766342,
-      "grad_norm": 11.095553398132324,
-      "learning_rate": 1.978187957739595e-05,
-      "loss": 1.0299,
-      "step": 3450
-    },
-    {
-      "epoch": 0.16017617060732497,
-      "grad_norm": 4.084702491760254,
-      "learning_rate": 1.978019560199856e-05,
-      "loss": 1.0295,
-      "step": 3455
-    },
-    {
-      "epoch": 0.16040797403801577,
-      "grad_norm": 4.840928554534912,
-      "learning_rate": 1.9778505223328996e-05,
-      "loss": 0.9896,
-      "step": 3460
-    },
-    {
-      "epoch": 0.16063977746870653,
-      "grad_norm": 4.553187370300293,
-      "learning_rate": 1.977680844249398e-05,
-      "loss": 0.9175,
-      "step": 3465
-    },
-    {
-      "epoch": 0.16087158089939732,
-      "grad_norm": 3.998581647872925,
-      "learning_rate": 1.977510526060443e-05,
-      "loss": 1.0071,
-      "step": 3470
-    },
-    {
-      "epoch": 0.16110338433008808,
-      "grad_norm": 3.791153907775879,
-      "learning_rate": 1.977339567877545e-05,
-      "loss": 1.0841,
-      "step": 3475
-    },
-    {
-      "epoch": 0.16133518776077885,
-      "grad_norm": 5.333462715148926,
-      "learning_rate": 1.9771679698126325e-05,
-      "loss": 1.0645,
-      "step": 3480
-    },
-    {
-      "epoch": 0.16156699119146964,
-      "grad_norm": 4.272615909576416,
-      "learning_rate": 1.976995731978055e-05,
-      "loss": 0.8177,
-      "step": 3485
-    },
-    {
-      "epoch": 0.1617987946221604,
-      "grad_norm": 4.884770393371582,
-      "learning_rate": 1.9768228544865793e-05,
-      "loss": 0.9687,
-      "step": 3490
-    },
-    {
-      "epoch": 0.1620305980528512,
-      "grad_norm": 4.274031162261963,
-      "learning_rate": 1.9766493374513912e-05,
-      "loss": 1.0413,
-      "step": 3495
-    },
-    {
-      "epoch": 0.16226240148354196,
-      "grad_norm": 3.577754497528076,
-      "learning_rate": 1.9764751809860957e-05,
-      "loss": 1.0992,
-      "step": 3500
-    },
-    {
-      "epoch": 0.16226240148354196,
-      "eval_loss": 1.0341907739639282,
-      "eval_runtime": 11.2731,
-      "eval_samples_per_second": 11.266,
-      "eval_steps_per_second": 11.266,
-      "step": 3500
-    },
-    {
-      "epoch": 0.16249420491423272,
-      "grad_norm": 3.879619598388672,
-      "learning_rate": 1.9763003852047164e-05,
-      "loss": 0.9959,
-      "step": 3505
-    },
-    {
-      "epoch": 0.1627260083449235,
-      "grad_norm": 4.179292678833008,
-      "learning_rate": 1.976124950221695e-05,
-      "loss": 1.1132,
-      "step": 3510
-    },
-    {
-      "epoch": 0.16295781177561428,
-      "grad_norm": 4.479405879974365,
-      "learning_rate": 1.975948876151892e-05,
-      "loss": 1.1255,
-      "step": 3515
-    },
-    {
-      "epoch": 0.16318961520630507,
-      "grad_norm": 4.490816116333008,
-      "learning_rate": 1.9757721631105865e-05,
-      "loss": 0.8673,
-      "step": 3520
-    },
-    {
-      "epoch": 0.16342141863699583,
-      "grad_norm": 6.789510726928711,
-      "learning_rate": 1.9755948112134758e-05,
-      "loss": 0.9336,
-      "step": 3525
-    },
-    {
-      "epoch": 0.1636532220676866,
-      "grad_norm": 3.9455666542053223,
-      "learning_rate": 1.975416820576675e-05,
-      "loss": 0.8529,
-      "step": 3530
-    },
-    {
-      "epoch": 0.16388502549837738,
-      "grad_norm": 4.85823917388916,
-      "learning_rate": 1.9752381913167178e-05,
-      "loss": 1.1189,
-      "step": 3535
-    },
-    {
-      "epoch": 0.16411682892906815,
-      "grad_norm": 5.231497764587402,
-      "learning_rate": 1.9750589235505565e-05,
-      "loss": 0.9141,
-      "step": 3540
-    },
-    {
-      "epoch": 0.1643486323597589,
-      "grad_norm": 5.203330993652344,
-      "learning_rate": 1.9748790173955606e-05,
-      "loss": 0.8673,
-      "step": 3545
-    },
-    {
-      "epoch": 0.1645804357904497,
-      "grad_norm": 3.671243190765381,
-      "learning_rate": 1.974698472969518e-05,
-      "loss": 0.9806,
-      "step": 3550
-    },
-    {
-      "epoch": 0.16481223922114047,
-      "grad_norm": 4.87835693359375,
-      "learning_rate": 1.974517290390635e-05,
-      "loss": 1.1254,
-      "step": 3555
-    },
-    {
-      "epoch": 0.16504404265183126,
-      "grad_norm": 4.269375801086426,
-      "learning_rate": 1.9743354697775343e-05,
-      "loss": 0.875,
-      "step": 3560
-    },
-    {
-      "epoch": 0.16527584608252202,
-      "grad_norm": 4.071610927581787,
-      "learning_rate": 1.974153011249257e-05,
-      "loss": 1.0405,
-      "step": 3565
-    },
-    {
-      "epoch": 0.16550764951321278,
-      "grad_norm": 4.637951374053955,
-      "learning_rate": 1.9739699149252634e-05,
-      "loss": 1.088,
-      "step": 3570
-    },
-    {
-      "epoch": 0.16573945294390358,
-      "grad_norm": 3.6990163326263428,
-      "learning_rate": 1.9737861809254285e-05,
-      "loss": 0.9877,
-      "step": 3575
-    },
-    {
-      "epoch": 0.16597125637459434,
-      "grad_norm": 3.8444910049438477,
-      "learning_rate": 1.9736018093700477e-05,
-      "loss": 0.882,
-      "step": 3580
-    },
-    {
-      "epoch": 0.16620305980528513,
-      "grad_norm": 4.383355617523193,
-      "learning_rate": 1.973416800379831e-05,
-      "loss": 1.1287,
-      "step": 3585
-    },
-    {
-      "epoch": 0.1664348632359759,
-      "grad_norm": 4.884894371032715,
-      "learning_rate": 1.9732311540759075e-05,
-      "loss": 0.9952,
-      "step": 3590
-    },
-    {
-      "epoch": 0.16666666666666666,
-      "grad_norm": 5.813953399658203,
-      "learning_rate": 1.973044870579824e-05,
-      "loss": 0.9972,
-      "step": 3595
-    },
-    {
-      "epoch": 0.16689847009735745,
-      "grad_norm": 6.32355260848999,
-      "learning_rate": 1.972857950013543e-05,
-      "loss": 1.1215,
-      "step": 3600
-    },
-    {
-      "epoch": 0.16689847009735745,
-      "eval_loss": 1.0363401174545288,
-      "eval_runtime": 11.2664,
-      "eval_samples_per_second": 11.273,
-      "eval_steps_per_second": 11.273,
-      "step": 3600
-    },
-    {
-      "epoch": 0.1671302735280482,
-      "grad_norm": 4.845508575439453,
-      "learning_rate": 1.972670392499445e-05,
-      "loss": 0.9931,
-      "step": 3605
-    },
-    {
-      "epoch": 0.16736207695873898,
-      "grad_norm": 4.5534515380859375,
-      "learning_rate": 1.9724821981603276e-05,
-      "loss": 1.0293,
-      "step": 3610
-    },
-    {
-      "epoch": 0.16759388038942977,
-      "grad_norm": 4.300433158874512,
-      "learning_rate": 1.9722933671194045e-05,
-      "loss": 0.933,
-      "step": 3615
-    },
-    {
-      "epoch": 0.16782568382012053,
-      "grad_norm": 4.074310302734375,
-      "learning_rate": 1.972103899500307e-05,
-      "loss": 0.9394,
-      "step": 3620
-    },
-    {
-      "epoch": 0.16805748725081132,
-      "grad_norm": 4.055570602416992,
-      "learning_rate": 1.9719137954270832e-05,
-      "loss": 1.0515,
-      "step": 3625
-    },
-    {
-      "epoch": 0.16828929068150209,
-      "grad_norm": 3.8710527420043945,
-      "learning_rate": 1.9717230550241978e-05,
-      "loss": 1.118,
-      "step": 3630
-    },
-    {
-      "epoch": 0.16852109411219285,
-      "grad_norm": 5.327306747436523,
-      "learning_rate": 1.9715316784165316e-05,
-      "loss": 1.1147,
-      "step": 3635
-    },
-    {
-      "epoch": 0.16875289754288364,
-      "grad_norm": 3.6472907066345215,
-      "learning_rate": 1.971339665729383e-05,
-      "loss": 0.9233,
-      "step": 3640
-    },
-    {
-      "epoch": 0.1689847009735744,
-      "grad_norm": 3.9819483757019043,
-      "learning_rate": 1.9711470170884652e-05,
-      "loss": 1.0589,
-      "step": 3645
-    },
-    {
-      "epoch": 0.1692165044042652,
-      "grad_norm": 4.43086051940918,
-      "learning_rate": 1.9709537326199096e-05,
-      "loss": 0.9735,
-      "step": 3650
-    },
-    {
-      "epoch": 0.16944830783495596,
-      "grad_norm": 5.653796195983887,
-      "learning_rate": 1.9707598124502627e-05,
-      "loss": 1.087,
-      "step": 3655
-    },
-    {
-      "epoch": 0.16968011126564672,
-      "grad_norm": 4.229031562805176,
-      "learning_rate": 1.9705652567064877e-05,
-      "loss": 1.0886,
-      "step": 3660
-    },
-    {
-      "epoch": 0.1699119146963375,
-      "grad_norm": 4.481167316436768,
-      "learning_rate": 1.9703700655159638e-05,
-      "loss": 0.9242,
-      "step": 3665
-    },
-    {
-      "epoch": 0.17014371812702828,
-      "grad_norm": 5.274721622467041,
-      "learning_rate": 1.970174239006486e-05,
-      "loss": 0.9642,
-      "step": 3670
-    },
-    {
-      "epoch": 0.17037552155771907,
-      "grad_norm": 5.67561674118042,
-      "learning_rate": 1.969977777306266e-05,
-      "loss": 1.1584,
-      "step": 3675
-    },
-    {
-      "epoch": 0.17060732498840983,
-      "grad_norm": 4.051698207855225,
-      "learning_rate": 1.9697806805439302e-05,
-      "loss": 0.9111,
-      "step": 3680
-    },
-    {
-      "epoch": 0.1708391284191006,
-      "grad_norm": 4.867814064025879,
-      "learning_rate": 1.969582948848522e-05,
-      "loss": 0.9635,
-      "step": 3685
-    },
-    {
-      "epoch": 0.17107093184979139,
-      "grad_norm": 4.764296531677246,
-      "learning_rate": 1.9693845823494997e-05,
-      "loss": 0.8434,
-      "step": 3690
-    },
-    {
-      "epoch": 0.17130273528048215,
-      "grad_norm": 3.813321352005005,
-      "learning_rate": 1.969185581176737e-05,
-      "loss": 0.9985,
-      "step": 3695
-    },
-    {
-      "epoch": 0.1715345387111729,
-      "grad_norm": 4.2770280838012695,
-      "learning_rate": 1.9689859454605246e-05,
-      "loss": 0.9844,
-      "step": 3700
-    },
-    {
-      "epoch": 0.1715345387111729,
-      "eval_loss": 1.0277271270751953,
-      "eval_runtime": 11.275,
-      "eval_samples_per_second": 11.264,
-      "eval_steps_per_second": 11.264,
-      "step": 3700
-    },
-    {
-      "epoch": 0.1717663421418637,
-      "grad_norm": 4.8459391593933105,
-      "learning_rate": 1.968785675331567e-05,
-      "loss": 1.0516,
-      "step": 3705
-    },
-    {
-      "epoch": 0.17199814557255447,
-      "grad_norm": 4.204828262329102,
-      "learning_rate": 1.9685847709209847e-05,
-      "loss": 1.08,
-      "step": 3710
-    },
-    {
-      "epoch": 0.17222994900324526,
-      "grad_norm": 4.045187950134277,
-      "learning_rate": 1.9683832323603137e-05,
-      "loss": 0.9182,
-      "step": 3715
-    },
-    {
-      "epoch": 0.17246175243393602,
-      "grad_norm": 4.368062973022461,
-      "learning_rate": 1.9681810597815047e-05,
-      "loss": 0.9901,
-      "step": 3720
-    },
-    {
-      "epoch": 0.17269355586462679,
-      "grad_norm": 4.184208393096924,
-      "learning_rate": 1.9679782533169237e-05,
-      "loss": 0.9044,
-      "step": 3725
-    },
-    {
-      "epoch": 0.17292535929531758,
-      "grad_norm": 4.531180381774902,
-      "learning_rate": 1.9677748130993524e-05,
-      "loss": 0.9238,
-      "step": 3730
-    },
-    {
-      "epoch": 0.17315716272600834,
-      "grad_norm": 3.6048779487609863,
-      "learning_rate": 1.9675707392619864e-05,
-      "loss": 1.026,
-      "step": 3735
-    },
-    {
-      "epoch": 0.17338896615669913,
-      "grad_norm": 4.324881553649902,
-      "learning_rate": 1.9673660319384362e-05,
-      "loss": 0.8431,
-      "step": 3740
-    },
-    {
-      "epoch": 0.1736207695873899,
-      "grad_norm": 5.622114181518555,
-      "learning_rate": 1.967160691262728e-05,
-      "loss": 0.9655,
-      "step": 3745
-    },
-    {
-      "epoch": 0.17385257301808066,
-      "grad_norm": 4.113274097442627,
-      "learning_rate": 1.9669547173693017e-05,
-      "loss": 0.7703,
-      "step": 3750
-    },
-    {
-      "epoch": 0.17408437644877145,
-      "grad_norm": 4.276374340057373,
-      "learning_rate": 1.9667481103930125e-05,
-      "loss": 1.1605,
-      "step": 3755
-    },
-    {
-      "epoch": 0.1743161798794622,
-      "grad_norm": 3.935689687728882,
-      "learning_rate": 1.9665408704691295e-05,
-      "loss": 1.0532,
-      "step": 3760
-    },
-    {
-      "epoch": 0.17454798331015298,
-      "grad_norm": 3.852186679840088,
-      "learning_rate": 1.9663329977333365e-05,
-      "loss": 0.8158,
-      "step": 3765
-    },
-    {
-      "epoch": 0.17477978674084377,
-      "grad_norm": 5.546504497528076,
-      "learning_rate": 1.9661244923217314e-05,
-      "loss": 1.0553,
-      "step": 3770
-    },
-    {
-      "epoch": 0.17501159017153453,
-      "grad_norm": 4.326759338378906,
-      "learning_rate": 1.9659153543708268e-05,
-      "loss": 1.0558,
-      "step": 3775
-    },
-    {
-      "epoch": 0.17524339360222532,
-      "grad_norm": 5.423635959625244,
-      "learning_rate": 1.9657055840175493e-05,
-      "loss": 0.9978,
-      "step": 3780
-    },
-    {
-      "epoch": 0.17547519703291609,
-      "grad_norm": 4.628658771514893,
-      "learning_rate": 1.9654951813992387e-05,
-      "loss": 1.0361,
-      "step": 3785
-    },
-    {
-      "epoch": 0.17570700046360685,
-      "grad_norm": 5.572815418243408,
-      "learning_rate": 1.96528414665365e-05,
-      "loss": 0.8554,
-      "step": 3790
-    },
-    {
-      "epoch": 0.17593880389429764,
-      "grad_norm": 3.796617269515991,
-      "learning_rate": 1.965072479918951e-05,
-      "loss": 1.012,
-      "step": 3795
-    },
-    {
-      "epoch": 0.1761706073249884,
-      "grad_norm": 4.33388614654541,
-      "learning_rate": 1.9648601813337244e-05,
-      "loss": 0.9588,
-      "step": 3800
-    },
-    {
-      "epoch": 0.1761706073249884,
-      "eval_loss": 1.0285202264785767,
-      "eval_runtime": 11.2827,
-      "eval_samples_per_second": 11.256,
-      "eval_steps_per_second": 11.256,
-      "step": 3800
-    },
-    {
-      "epoch": 0.1764024107556792,
-      "grad_norm": 8.51945972442627,
-      "learning_rate": 1.9646472510369656e-05,
-      "loss": 1.0992,
-      "step": 3805
-    },
-    {
-      "epoch": 0.17663421418636996,
-      "grad_norm": 4.519212245941162,
-      "learning_rate": 1.9644336891680836e-05,
-      "loss": 1.0471,
-      "step": 3810
-    },
-    {
-      "epoch": 0.17686601761706072,
-      "grad_norm": 3.9827346801757812,
-      "learning_rate": 1.964219495866902e-05,
-      "loss": 0.8874,
-      "step": 3815
-    },
-    {
-      "epoch": 0.1770978210477515,
-      "grad_norm": 4.4609694480896,
-      "learning_rate": 1.9640046712736565e-05,
-      "loss": 1.0301,
-      "step": 3820
-    },
-    {
-      "epoch": 0.17732962447844228,
-      "grad_norm": 3.9472250938415527,
-      "learning_rate": 1.9637892155289967e-05,
-      "loss": 0.8735,
-      "step": 3825
-    },
-    {
-      "epoch": 0.17756142790913307,
-      "grad_norm": 3.6204352378845215,
-      "learning_rate": 1.9635731287739857e-05,
-      "loss": 0.9376,
-      "step": 3830
-    },
-    {
-      "epoch": 0.17779323133982383,
-      "grad_norm": 3.9136033058166504,
-      "learning_rate": 1.963356411150099e-05,
-      "loss": 0.9396,
-      "step": 3835
-    },
-    {
-      "epoch": 0.1780250347705146,
-      "grad_norm": 3.8638482093811035,
-      "learning_rate": 1.963139062799226e-05,
-      "loss": 0.9546,
-      "step": 3840
-    },
-    {
-      "epoch": 0.1782568382012054,
-      "grad_norm": 3.9721662998199463,
-      "learning_rate": 1.9629210838636685e-05,
-      "loss": 1.0491,
-      "step": 3845
-    },
-    {
-      "epoch": 0.17848864163189615,
-      "grad_norm": 5.257063865661621,
-      "learning_rate": 1.962702474486141e-05,
-      "loss": 1.1241,
-      "step": 3850
-    },
-    {
-      "epoch": 0.1787204450625869,
-      "grad_norm": 4.413024425506592,
-      "learning_rate": 1.9624832348097717e-05,
-      "loss": 1.1912,
-      "step": 3855
-    },
-    {
-      "epoch": 0.1789522484932777,
-      "grad_norm": 5.4773759841918945,
-      "learning_rate": 1.9622633649780998e-05,
-      "loss": 0.9792,
-      "step": 3860
-    },
-    {
-      "epoch": 0.17918405192396847,
-      "grad_norm": 4.147509574890137,
-      "learning_rate": 1.962042865135079e-05,
-      "loss": 0.923,
-      "step": 3865
-    },
-    {
-      "epoch": 0.17941585535465926,
-      "grad_norm": 3.813749313354492,
-      "learning_rate": 1.9618217354250742e-05,
-      "loss": 0.9268,
-      "step": 3870
-    },
-    {
-      "epoch": 0.17964765878535002,
-      "grad_norm": 5.082292079925537,
-      "learning_rate": 1.9615999759928626e-05,
-      "loss": 1.0358,
-      "step": 3875
-    },
-    {
-      "epoch": 0.1798794622160408,
-      "grad_norm": 4.669640064239502,
-      "learning_rate": 1.9613775869836355e-05,
-      "loss": 1.0233,
-      "step": 3880
-    },
-    {
-      "epoch": 0.18011126564673158,
-      "grad_norm": 4.9730424880981445,
-      "learning_rate": 1.9611545685429937e-05,
-      "loss": 1.0794,
-      "step": 3885
-    },
-    {
-      "epoch": 0.18034306907742234,
-      "grad_norm": 4.652693271636963,
-      "learning_rate": 1.9609309208169518e-05,
-      "loss": 0.9949,
-      "step": 3890
-    },
-    {
-      "epoch": 0.18057487250811313,
-      "grad_norm": 3.6250641345977783,
-      "learning_rate": 1.9607066439519363e-05,
-      "loss": 0.8387,
-      "step": 3895
-    },
-    {
-      "epoch": 0.1808066759388039,
-      "grad_norm": 4.949042320251465,
-      "learning_rate": 1.960481738094785e-05,
-      "loss": 0.9968,
-      "step": 3900
-    },
-    {
-      "epoch": 0.1808066759388039,
-      "eval_loss": 1.0276681184768677,
-      "eval_runtime": 11.2655,
-      "eval_samples_per_second": 11.273,
-      "eval_steps_per_second": 11.273,
-      "step": 3900
-    },
-    {
-      "epoch": 0.18103847936949466,
-      "grad_norm": 3.579059362411499,
-      "learning_rate": 1.9602562033927478e-05,
-      "loss": 0.9648,
-      "step": 3905
-    },
-    {
-      "epoch": 0.18127028280018545,
-      "grad_norm": 4.2248358726501465,
-      "learning_rate": 1.960030039993487e-05,
-      "loss": 0.9858,
-      "step": 3910
-    },
-    {
-      "epoch": 0.18150208623087621,
-      "grad_norm": 4.166600227355957,
-      "learning_rate": 1.9598032480450757e-05,
-      "loss": 1.079,
-      "step": 3915
-    },
-    {
-      "epoch": 0.18173388966156698,
-      "grad_norm": 4.941333293914795,
-      "learning_rate": 1.9595758276959982e-05,
-      "loss": 0.9412,
-      "step": 3920
-    },
-    {
-      "epoch": 0.18196569309225777,
-      "grad_norm": 5.1348981857299805,
-      "learning_rate": 1.959347779095151e-05,
-      "loss": 1.0761,
-      "step": 3925
-    },
-    {
-      "epoch": 0.18219749652294853,
-      "grad_norm": 4.545997142791748,
-      "learning_rate": 1.9591191023918417e-05,
-      "loss": 1.055,
-      "step": 3930
-    },
-    {
-      "epoch": 0.18242929995363932,
-      "grad_norm": 4.699060916900635,
-      "learning_rate": 1.9588897977357894e-05,
-      "loss": 1.1684,
-      "step": 3935
-    },
-    {
-      "epoch": 0.1826611033843301,
-      "grad_norm": 5.560849666595459,
-      "learning_rate": 1.9586598652771237e-05,
-      "loss": 0.9208,
-      "step": 3940
-    },
-    {
-      "epoch": 0.18289290681502085,
-      "grad_norm": 4.416356086730957,
-      "learning_rate": 1.9584293051663855e-05,
-      "loss": 0.9471,
-      "step": 3945
-    },
-    {
-      "epoch": 0.18312471024571164,
-      "grad_norm": 3.3139140605926514,
-      "learning_rate": 1.958198117554527e-05,
-      "loss": 1.0265,
-      "step": 3950
-    },
-    {
-      "epoch": 0.1833565136764024,
-      "grad_norm": 3.708087205886841,
-      "learning_rate": 1.957966302592911e-05,
-      "loss": 1.0605,
-      "step": 3955
-    },
-    {
-      "epoch": 0.1835883171070932,
-      "grad_norm": 4.243253707885742,
-      "learning_rate": 1.9577338604333102e-05,
-      "loss": 1.0028,
-      "step": 3960
-    },
-    {
-      "epoch": 0.18382012053778396,
-      "grad_norm": 4.022925853729248,
-      "learning_rate": 1.9575007912279096e-05,
-      "loss": 0.8775,
-      "step": 3965
-    },
-    {
-      "epoch": 0.18405192396847472,
-      "grad_norm": 4.368210315704346,
-      "learning_rate": 1.9572670951293034e-05,
-      "loss": 1.1574,
-      "step": 3970
-    },
-    {
-      "epoch": 0.18428372739916551,
-      "grad_norm": 3.763110399246216,
-      "learning_rate": 1.957032772290497e-05,
-      "loss": 0.9621,
-      "step": 3975
-    },
-    {
-      "epoch": 0.18451553082985628,
-      "grad_norm": 4.240961074829102,
-      "learning_rate": 1.9567978228649052e-05,
-      "loss": 1.1428,
-      "step": 3980
-    },
-    {
-      "epoch": 0.18474733426054707,
-      "grad_norm": 4.0695037841796875,
-      "learning_rate": 1.9565622470063544e-05,
-      "loss": 0.8761,
-      "step": 3985
-    },
-    {
-      "epoch": 0.18497913769123783,
-      "grad_norm": 4.663343906402588,
-      "learning_rate": 1.9563260448690805e-05,
-      "loss": 1.1419,
-      "step": 3990
-    },
-    {
-      "epoch": 0.1852109411219286,
-      "grad_norm": 3.7018463611602783,
-      "learning_rate": 1.9560892166077285e-05,
-      "loss": 0.9401,
-      "step": 3995
-    },
-    {
-      "epoch": 0.1854427445526194,
-      "grad_norm": 4.6532511711120605,
-      "learning_rate": 1.9558517623773547e-05,
-      "loss": 0.8992,
-      "step": 4000
-    },
-    {
-      "epoch": 0.1854427445526194,
-      "eval_loss": 1.0249594449996948,
-      "eval_runtime": 11.2634,
-      "eval_samples_per_second": 11.275,
-      "eval_steps_per_second": 11.275,
-      "step": 4000
-    },
-    {
-      "epoch": 0.18567454798331015,
-      "grad_norm": 4.233008861541748,
-      "learning_rate": 1.955613682333425e-05,
-      "loss": 1.0099,
-      "step": 4005
-    },
-    {
-      "epoch": 0.18590635141400091,
-      "grad_norm": 4.297733306884766,
-      "learning_rate": 1.955374976631814e-05,
-      "loss": 0.8721,
-      "step": 4010
-    },
-    {
-      "epoch": 0.1861381548446917,
-      "grad_norm": 5.088667869567871,
-      "learning_rate": 1.9551356454288073e-05,
-      "loss": 0.9625,
-      "step": 4015
-    },
-    {
-      "epoch": 0.18636995827538247,
-      "grad_norm": 4.837838172912598,
-      "learning_rate": 1.954895688881099e-05,
-      "loss": 1.07,
-      "step": 4020
-    },
-    {
-      "epoch": 0.18660176170607326,
-      "grad_norm": 3.7905993461608887,
-      "learning_rate": 1.954655107145793e-05,
-      "loss": 1.0712,
-      "step": 4025
-    },
-    {
-      "epoch": 0.18683356513676402,
-      "grad_norm": 3.836812734603882,
-      "learning_rate": 1.954413900380403e-05,
-      "loss": 1.1433,
-      "step": 4030
-    },
-    {
-      "epoch": 0.1870653685674548,
-      "grad_norm": 3.796400547027588,
-      "learning_rate": 1.9541720687428512e-05,
-      "loss": 0.7793,
-      "step": 4035
-    },
-    {
-      "epoch": 0.18729717199814558,
-      "grad_norm": 4.4649553298950195,
-      "learning_rate": 1.9539296123914693e-05,
-      "loss": 1.0256,
-      "step": 4040
-    },
-    {
-      "epoch": 0.18752897542883634,
-      "grad_norm": 3.9846534729003906,
-      "learning_rate": 1.9536865314849973e-05,
-      "loss": 0.9158,
-      "step": 4045
-    },
-    {
-      "epoch": 0.18776077885952713,
-      "grad_norm": 5.033636093139648,
-      "learning_rate": 1.9534428261825857e-05,
-      "loss": 1.1844,
-      "step": 4050
-    },
-    {
-      "epoch": 0.1879925822902179,
-      "grad_norm": 4.519948959350586,
-      "learning_rate": 1.9531984966437916e-05,
-      "loss": 1.1649,
-      "step": 4055
-    },
-    {
-      "epoch": 0.18822438572090866,
-      "grad_norm": 5.361178874969482,
-      "learning_rate": 1.952953543028583e-05,
-      "loss": 0.9993,
-      "step": 4060
-    },
-    {
-      "epoch": 0.18845618915159945,
-      "grad_norm": 3.7113776206970215,
-      "learning_rate": 1.9527079654973353e-05,
-      "loss": 0.9374,
-      "step": 4065
-    },
-    {
-      "epoch": 0.18868799258229021,
-      "grad_norm": 4.310559272766113,
-      "learning_rate": 1.952461764210832e-05,
-      "loss": 0.8647,
-      "step": 4070
-    },
-    {
-      "epoch": 0.18891979601298098,
-      "grad_norm": 3.929408073425293,
-      "learning_rate": 1.9522149393302664e-05,
-      "loss": 0.9928,
-      "step": 4075
-    },
-    {
-      "epoch": 0.18915159944367177,
-      "grad_norm": 4.187215805053711,
-      "learning_rate": 1.9519674910172387e-05,
-      "loss": 1.2086,
-      "step": 4080
-    },
-    {
-      "epoch": 0.18938340287436253,
-      "grad_norm": 5.6810302734375,
-      "learning_rate": 1.9517194194337578e-05,
-      "loss": 1.0479,
-      "step": 4085
-    },
-    {
-      "epoch": 0.18961520630505332,
-      "grad_norm": 4.767360687255859,
-      "learning_rate": 1.951470724742241e-05,
-      "loss": 1.1135,
-      "step": 4090
-    },
-    {
-      "epoch": 0.1898470097357441,
-      "grad_norm": 5.183935642242432,
-      "learning_rate": 1.951221407105513e-05,
-      "loss": 1.1425,
-      "step": 4095
-    },
-    {
-      "epoch": 0.19007881316643485,
-      "grad_norm": 4.624146938323975,
-      "learning_rate": 1.9509714666868066e-05,
-      "loss": 1.1632,
-      "step": 4100
-    },
-    {
-      "epoch": 0.19007881316643485,
-      "eval_loss": 1.0201714038848877,
-      "eval_runtime": 11.2685,
-      "eval_samples_per_second": 11.27,
-      "eval_steps_per_second": 11.27,
-      "step": 4100
-    },
-    {
-      "epoch": 0.19031061659712564,
-      "grad_norm": 6.600108623504639,
-      "learning_rate": 1.9507209036497626e-05,
-      "loss": 1.063,
-      "step": 4105
-    },
-    {
-      "epoch": 0.1905424200278164,
-      "grad_norm": 7.14796257019043,
-      "learning_rate": 1.950469718158429e-05,
-      "loss": 0.9141,
-      "step": 4110
-    },
-    {
-      "epoch": 0.1907742234585072,
-      "grad_norm": 5.579740524291992,
-      "learning_rate": 1.9502179103772612e-05,
-      "loss": 0.9236,
-      "step": 4115
-    },
-    {
-      "epoch": 0.19100602688919796,
-      "grad_norm": 3.9887657165527344,
-      "learning_rate": 1.9499654804711227e-05,
-      "loss": 0.972,
-      "step": 4120
-    },
-    {
-      "epoch": 0.19123783031988872,
-      "grad_norm": 5.2029595375061035,
-      "learning_rate": 1.949712428605284e-05,
-      "loss": 1.1145,
-      "step": 4125
-    },
-    {
-      "epoch": 0.19146963375057952,
-      "grad_norm": 3.8869788646698,
-      "learning_rate": 1.949458754945423e-05,
-      "loss": 0.8691,
-      "step": 4130
-    },
-    {
-      "epoch": 0.19170143718127028,
-      "grad_norm": 4.547691822052002,
-      "learning_rate": 1.9492044596576238e-05,
-      "loss": 0.8942,
-      "step": 4135
-    },
-    {
-      "epoch": 0.19193324061196107,
-      "grad_norm": 4.375545978546143,
-      "learning_rate": 1.9489495429083787e-05,
-      "loss": 1.0702,
-      "step": 4140
-    },
-    {
-      "epoch": 0.19216504404265183,
-      "grad_norm": 4.513903617858887,
-      "learning_rate": 1.948694004864586e-05,
-      "loss": 1.0058,
-      "step": 4145
-    },
-    {
-      "epoch": 0.1923968474733426,
-      "grad_norm": 3.5358970165252686,
-      "learning_rate": 1.948437845693551e-05,
-      "loss": 0.9213,
-      "step": 4150
-    },
-    {
-      "epoch": 0.1926286509040334,
-      "grad_norm": 4.308315277099609,
-      "learning_rate": 1.9481810655629862e-05,
-      "loss": 0.9508,
-      "step": 4155
-    },
-    {
-      "epoch": 0.19286045433472415,
-      "grad_norm": 4.383856773376465,
-      "learning_rate": 1.9479236646410104e-05,
-      "loss": 1.0954,
-      "step": 4160
-    },
-    {
-      "epoch": 0.19309225776541492,
-      "grad_norm": 4.728183746337891,
-      "learning_rate": 1.947665643096148e-05,
-      "loss": 1.0281,
-      "step": 4165
-    },
-    {
-      "epoch": 0.1933240611961057,
-      "grad_norm": 4.869402885437012,
-      "learning_rate": 1.9474070010973307e-05,
-      "loss": 0.8735,
-      "step": 4170
-    },
-    {
-      "epoch": 0.19355586462679647,
-      "grad_norm": 3.7020456790924072,
-      "learning_rate": 1.9471477388138962e-05,
-      "loss": 0.8899,
-      "step": 4175
-    },
-    {
-      "epoch": 0.19378766805748726,
-      "grad_norm": 4.01472806930542,
-      "learning_rate": 1.9468878564155882e-05,
-      "loss": 0.973,
-      "step": 4180
-    },
-    {
-      "epoch": 0.19401947148817802,
-      "grad_norm": 4.646851062774658,
-      "learning_rate": 1.9466273540725565e-05,
-      "loss": 1.1544,
-      "step": 4185
-    },
-    {
-      "epoch": 0.1942512749188688,
-      "grad_norm": 4.423065662384033,
-      "learning_rate": 1.9463662319553568e-05,
-      "loss": 0.9554,
-      "step": 4190
-    },
-    {
-      "epoch": 0.19448307834955958,
-      "grad_norm": 4.54693078994751,
-      "learning_rate": 1.9461044902349497e-05,
-      "loss": 0.923,
-      "step": 4195
-    },
-    {
-      "epoch": 0.19471488178025034,
-      "grad_norm": 4.460963726043701,
-      "learning_rate": 1.9458421290827035e-05,
-      "loss": 1.0284,
-      "step": 4200
-    },
-    {
-      "epoch": 0.19471488178025034,
-      "eval_loss": 1.022770881652832,
-      "eval_runtime": 11.2826,
-      "eval_samples_per_second": 11.256,
-      "eval_steps_per_second": 11.256,
-      "step": 4200
-    },
-    {
-      "epoch": 0.19494668521094113,
-      "grad_norm": 3.523135185241699,
-      "learning_rate": 1.9455791486703905e-05,
-      "loss": 0.9056,
-      "step": 4205
-    },
-    {
-      "epoch": 0.1951784886416319,
-      "grad_norm": 3.8507754802703857,
-      "learning_rate": 1.945315549170188e-05,
-      "loss": 0.7916,
-      "step": 4210
-    },
-    {
-      "epoch": 0.19541029207232266,
-      "grad_norm": 14.194623947143555,
-      "learning_rate": 1.94505133075468e-05,
-      "loss": 1.075,
-      "step": 4215
-    },
-    {
-      "epoch": 0.19564209550301345,
-      "grad_norm": 3.8128092288970947,
-      "learning_rate": 1.944786493596855e-05,
-      "loss": 1.0419,
-      "step": 4220
-    },
-    {
-      "epoch": 0.19587389893370422,
-      "grad_norm": 4.251338481903076,
-      "learning_rate": 1.9445210378701067e-05,
-      "loss": 1.1187,
-      "step": 4225
-    },
-    {
-      "epoch": 0.19610570236439498,
-      "grad_norm": 4.296945571899414,
-      "learning_rate": 1.9442549637482342e-05,
-      "loss": 0.9089,
-      "step": 4230
-    },
-    {
-      "epoch": 0.19633750579508577,
-      "grad_norm": 4.671785831451416,
-      "learning_rate": 1.9439882714054403e-05,
-      "loss": 0.9709,
-      "step": 4235
-    },
-    {
-      "epoch": 0.19656930922577653,
-      "grad_norm": 4.192035675048828,
-      "learning_rate": 1.943720961016334e-05,
-      "loss": 0.9624,
-      "step": 4240
-    },
-    {
-      "epoch": 0.19680111265646733,
-      "grad_norm": 5.137444019317627,
-      "learning_rate": 1.943453032755928e-05,
-      "loss": 1.0318,
-      "step": 4245
-    },
-    {
-      "epoch": 0.1970329160871581,
-      "grad_norm": 4.235700607299805,
-      "learning_rate": 1.9431844867996396e-05,
-      "loss": 0.897,
-      "step": 4250
-    },
-    {
-      "epoch": 0.19726471951784885,
-      "grad_norm": 4.324668884277344,
-      "learning_rate": 1.9429153233232912e-05,
-      "loss": 0.9075,
-      "step": 4255
-    },
-    {
-      "epoch": 0.19749652294853964,
-      "grad_norm": 5.024466514587402,
-      "learning_rate": 1.9426455425031085e-05,
-      "loss": 1.1617,
-      "step": 4260
-    },
-    {
-      "epoch": 0.1977283263792304,
-      "grad_norm": 3.738459348678589,
-      "learning_rate": 1.9423751445157225e-05,
-      "loss": 0.8962,
-      "step": 4265
-    },
-    {
-      "epoch": 0.1979601298099212,
-      "grad_norm": 4.7824625968933105,
-      "learning_rate": 1.9421041295381674e-05,
-      "loss": 1.0028,
-      "step": 4270
-    },
-    {
-      "epoch": 0.19819193324061196,
-      "grad_norm": 5.4419264793396,
-      "learning_rate": 1.9418324977478815e-05,
-      "loss": 1.1338,
-      "step": 4275
-    },
-    {
-      "epoch": 0.19842373667130273,
-      "grad_norm": 3.802666425704956,
-      "learning_rate": 1.941560249322707e-05,
-      "loss": 1.0708,
-      "step": 4280
-    },
-    {
-      "epoch": 0.19865554010199352,
-      "grad_norm": 4.5152788162231445,
-      "learning_rate": 1.9412873844408905e-05,
-      "loss": 1.0177,
-      "step": 4285
-    },
-    {
-      "epoch": 0.19888734353268428,
-      "grad_norm": 4.837648391723633,
-      "learning_rate": 1.941013903281081e-05,
-      "loss": 1.099,
-      "step": 4290
-    },
-    {
-      "epoch": 0.19911914696337507,
-      "grad_norm": 6.359385013580322,
-      "learning_rate": 1.940739806022332e-05,
-      "loss": 1.1031,
-      "step": 4295
-    },
-    {
-      "epoch": 0.19935095039406583,
-      "grad_norm": 3.1812539100646973,
-      "learning_rate": 1.9404650928440996e-05,
-      "loss": 0.9961,
-      "step": 4300
-    },
-    {
-      "epoch": 0.19935095039406583,
-      "eval_loss": 1.0196547508239746,
-      "eval_runtime": 11.2671,
-      "eval_samples_per_second": 11.272,
-      "eval_steps_per_second": 11.272,
-      "step": 4300
-    },
-    {
-      "epoch": 0.1995827538247566,
-      "grad_norm": 3.9984710216522217,
-      "learning_rate": 1.940189763926244e-05,
-      "loss": 1.1403,
-      "step": 4305
-    },
-    {
-      "epoch": 0.1998145572554474,
-      "grad_norm": 4.785546779632568,
-      "learning_rate": 1.9399138194490273e-05,
-      "loss": 1.0475,
-      "step": 4310
-    },
-    {
-      "epoch": 0.20004636068613815,
-      "grad_norm": 4.264800548553467,
-      "learning_rate": 1.9396372595931165e-05,
-      "loss": 1.01,
-      "step": 4315
-    },
-    {
-      "epoch": 0.20027816411682892,
-      "grad_norm": 4.709249496459961,
-      "learning_rate": 1.9393600845395792e-05,
-      "loss": 0.8559,
-      "step": 4320
-    },
-    {
-      "epoch": 0.2005099675475197,
-      "grad_norm": 4.430532455444336,
-      "learning_rate": 1.9390822944698873e-05,
-      "loss": 0.8757,
-      "step": 4325
-    },
-    {
-      "epoch": 0.20074177097821047,
-      "grad_norm": 3.601813554763794,
-      "learning_rate": 1.9388038895659156e-05,
-      "loss": 1.0746,
-      "step": 4330
-    },
-    {
-      "epoch": 0.20097357440890126,
-      "grad_norm": 5.1407880783081055,
-      "learning_rate": 1.9385248700099402e-05,
-      "loss": 0.9955,
-      "step": 4335
-    },
-    {
-      "epoch": 0.20120537783959203,
-      "grad_norm": 4.2756829261779785,
-      "learning_rate": 1.9382452359846408e-05,
-      "loss": 0.9426,
-      "step": 4340
-    },
-    {
-      "epoch": 0.2014371812702828,
-      "grad_norm": 4.615556716918945,
-      "learning_rate": 1.937964987673098e-05,
-      "loss": 1.0964,
-      "step": 4345
-    },
-    {
-      "epoch": 0.20166898470097358,
-      "grad_norm": 4.2694244384765625,
-      "learning_rate": 1.937684125258797e-05,
-      "loss": 0.9078,
-      "step": 4350
-    },
-    {
-      "epoch": 0.20190078813166434,
-      "grad_norm": 3.894665241241455,
-      "learning_rate": 1.9374026489256217e-05,
-      "loss": 0.7885,
-      "step": 4355
-    },
-    {
-      "epoch": 0.20213259156235514,
-      "grad_norm": 3.197622537612915,
-      "learning_rate": 1.9371205588578613e-05,
-      "loss": 0.9278,
-      "step": 4360
-    },
-    {
-      "epoch": 0.2023643949930459,
-      "grad_norm": 4.220120906829834,
-      "learning_rate": 1.9368378552402046e-05,
-      "loss": 0.8669,
-      "step": 4365
-    },
-    {
-      "epoch": 0.20259619842373666,
-      "grad_norm": 4.15202522277832,
-      "learning_rate": 1.9365545382577433e-05,
-      "loss": 0.9815,
-      "step": 4370
-    },
-    {
-      "epoch": 0.20282800185442745,
-      "grad_norm": 4.307468414306641,
-      "learning_rate": 1.9362706080959694e-05,
-      "loss": 1.1502,
-      "step": 4375
-    },
-    {
-      "epoch": 0.20305980528511822,
-      "grad_norm": 4.232968807220459,
-      "learning_rate": 1.9359860649407775e-05,
-      "loss": 1.1307,
-      "step": 4380
-    },
-    {
-      "epoch": 0.20329160871580898,
-      "grad_norm": 4.220963954925537,
-      "learning_rate": 1.935700908978463e-05,
-      "loss": 1.1395,
-      "step": 4385
-    },
-    {
-      "epoch": 0.20352341214649977,
-      "grad_norm": 3.8981432914733887,
-      "learning_rate": 1.9354151403957234e-05,
-      "loss": 1.035,
-      "step": 4390
-    },
-    {
-      "epoch": 0.20375521557719053,
-      "grad_norm": 5.082888603210449,
-      "learning_rate": 1.935128759379656e-05,
-      "loss": 0.9727,
-      "step": 4395
-    },
-    {
-      "epoch": 0.20398701900788133,
-      "grad_norm": 4.414380073547363,
-      "learning_rate": 1.9348417661177597e-05,
-      "loss": 1.0242,
-      "step": 4400
-    },
-    {
-      "epoch": 0.20398701900788133,
-      "eval_loss": 1.0201265811920166,
-      "eval_runtime": 11.2759,
-      "eval_samples_per_second": 11.263,
-      "eval_steps_per_second": 11.263,
-      "step": 4400
-    },
-    {
-      "epoch": 0.2042188224385721,
-      "grad_norm": 3.861715793609619,
-      "learning_rate": 1.9345541607979347e-05,
-      "loss": 0.9632,
-      "step": 4405
-    },
-    {
-      "epoch": 0.20445062586926285,
-      "grad_norm": 3.9358503818511963,
-      "learning_rate": 1.9342659436084808e-05,
-      "loss": 0.8809,
-      "step": 4410
-    },
-    {
-      "epoch": 0.20468242929995364,
-      "grad_norm": 4.636179447174072,
-      "learning_rate": 1.9339771147380993e-05,
-      "loss": 1.0324,
-      "step": 4415
-    },
-    {
-      "epoch": 0.2049142327306444,
-      "grad_norm": 3.6778595447540283,
-      "learning_rate": 1.9336876743758918e-05,
-      "loss": 1.1124,
-      "step": 4420
-    },
-    {
-      "epoch": 0.2051460361613352,
-      "grad_norm": 3.6706385612487793,
-      "learning_rate": 1.9333976227113598e-05,
-      "loss": 0.9638,
-      "step": 4425
-    },
-    {
-      "epoch": 0.20537783959202596,
-      "grad_norm": 4.206617832183838,
-      "learning_rate": 1.9331069599344058e-05,
-      "loss": 1.0878,
-      "step": 4430
-    },
-    {
-      "epoch": 0.20560964302271673,
-      "grad_norm": 4.7787017822265625,
-      "learning_rate": 1.9328156862353322e-05,
-      "loss": 0.9389,
-      "step": 4435
-    },
-    {
-      "epoch": 0.20584144645340752,
-      "grad_norm": 4.6298933029174805,
-      "learning_rate": 1.9325238018048405e-05,
-      "loss": 1.1185,
-      "step": 4440
-    },
-    {
-      "epoch": 0.20607324988409828,
-      "grad_norm": 4.0797858238220215,
-      "learning_rate": 1.9322313068340338e-05,
-      "loss": 0.9152,
-      "step": 4445
-    },
-    {
-      "epoch": 0.20630505331478907,
-      "grad_norm": 3.8898723125457764,
-      "learning_rate": 1.9319382015144127e-05,
-      "loss": 1.027,
-      "step": 4450
-    },
-    {
-      "epoch": 0.20653685674547984,
-      "grad_norm": 3.4612269401550293,
-      "learning_rate": 1.9316444860378792e-05,
-      "loss": 0.9548,
-      "step": 4455
-    },
-    {
-      "epoch": 0.2067686601761706,
-      "grad_norm": 3.6337077617645264,
-      "learning_rate": 1.9313501605967343e-05,
-      "loss": 0.8258,
-      "step": 4460
-    },
-    {
-      "epoch": 0.2070004636068614,
-      "grad_norm": 4.511351585388184,
-      "learning_rate": 1.9310552253836784e-05,
-      "loss": 1.0118,
-      "step": 4465
-    },
-    {
-      "epoch": 0.20723226703755215,
-      "grad_norm": 4.938009738922119,
-      "learning_rate": 1.93075968059181e-05,
-      "loss": 0.9311,
-      "step": 4470
-    },
-    {
-      "epoch": 0.20746407046824292,
-      "grad_norm": 4.323369026184082,
-      "learning_rate": 1.9304635264146287e-05,
-      "loss": 1.0904,
-      "step": 4475
-    },
-    {
-      "epoch": 0.2076958738989337,
-      "grad_norm": 4.58251428604126,
-      "learning_rate": 1.9301667630460312e-05,
-      "loss": 1.0012,
-      "step": 4480
-    },
-    {
-      "epoch": 0.20792767732962447,
-      "grad_norm": 3.8780105113983154,
-      "learning_rate": 1.9298693906803146e-05,
-      "loss": 0.978,
-      "step": 4485
-    },
-    {
-      "epoch": 0.20815948076031526,
-      "grad_norm": 4.129722595214844,
-      "learning_rate": 1.9295714095121733e-05,
-      "loss": 1.1244,
-      "step": 4490
-    },
-    {
-      "epoch": 0.20839128419100603,
-      "grad_norm": 4.224957466125488,
-      "learning_rate": 1.9292728197367017e-05,
-      "loss": 0.9721,
-      "step": 4495
-    },
-    {
-      "epoch": 0.2086230876216968,
-      "grad_norm": 4.470384120941162,
-      "learning_rate": 1.9289736215493914e-05,
-      "loss": 1.0162,
-      "step": 4500
-    },
-    {
-      "epoch": 0.2086230876216968,
-      "eval_loss": 1.0174404382705688,
-      "eval_runtime": 11.2755,
-      "eval_samples_per_second": 11.263,
-      "eval_steps_per_second": 11.263,
-      "step": 4500
-    },
-    {
-      "epoch": 0.20885489105238758,
-      "grad_norm": 3.653148651123047,
-      "learning_rate": 1.9286738151461333e-05,
-      "loss": 0.8662,
-      "step": 4505
-    },
-    {
-      "epoch": 0.20908669448307834,
-      "grad_norm": 4.83407735824585,
-      "learning_rate": 1.9283734007232156e-05,
-      "loss": 1.0202,
-      "step": 4510
-    },
-    {
-      "epoch": 0.20931849791376914,
-      "grad_norm": 5.023587226867676,
-      "learning_rate": 1.9280723784773257e-05,
-      "loss": 1.0915,
-      "step": 4515
-    },
-    {
-      "epoch": 0.2095503013444599,
-      "grad_norm": 5.459712982177734,
-      "learning_rate": 1.9277707486055487e-05,
-      "loss": 1.1079,
-      "step": 4520
-    },
-    {
-      "epoch": 0.20978210477515066,
-      "grad_norm": 4.731924057006836,
-      "learning_rate": 1.9274685113053656e-05,
-      "loss": 1.1386,
-      "step": 4525
-    },
-    {
-      "epoch": 0.21001390820584145,
-      "grad_norm": 3.6431119441986084,
-      "learning_rate": 1.9271656667746583e-05,
-      "loss": 0.9088,
-      "step": 4530
-    },
-    {
-      "epoch": 0.21024571163653222,
-      "grad_norm": 3.588373899459839,
-      "learning_rate": 1.9268622152117035e-05,
-      "loss": 0.8291,
-      "step": 4535
-    },
-    {
-      "epoch": 0.21047751506722298,
-      "grad_norm": 5.668548107147217,
-      "learning_rate": 1.926558156815177e-05,
-      "loss": 1.1071,
-      "step": 4540
-    },
-    {
-      "epoch": 0.21070931849791377,
-      "grad_norm": 5.053746223449707,
-      "learning_rate": 1.9262534917841514e-05,
-      "loss": 0.8661,
-      "step": 4545
-    },
-    {
-      "epoch": 0.21094112192860454,
-      "grad_norm": 3.778334140777588,
-      "learning_rate": 1.9259482203180963e-05,
-      "loss": 0.9204,
-      "step": 4550
-    },
-    {
-      "epoch": 0.21117292535929533,
-      "grad_norm": 4.436891555786133,
-      "learning_rate": 1.9256423426168785e-05,
-      "loss": 0.8769,
-      "step": 4555
-    },
-    {
-      "epoch": 0.2114047287899861,
-      "grad_norm": 4.076458930969238,
-      "learning_rate": 1.9253358588807617e-05,
-      "loss": 0.9674,
-      "step": 4560
-    },
-    {
-      "epoch": 0.21163653222067685,
-      "grad_norm": 4.523109436035156,
-      "learning_rate": 1.9250287693104068e-05,
-      "loss": 0.904,
-      "step": 4565
-    },
-    {
-      "epoch": 0.21186833565136765,
-      "grad_norm": 4.488249778747559,
-      "learning_rate": 1.92472107410687e-05,
-      "loss": 0.9186,
-      "step": 4570
-    },
-    {
-      "epoch": 0.2121001390820584,
-      "grad_norm": 4.525340557098389,
-      "learning_rate": 1.9244127734716058e-05,
-      "loss": 0.8192,
-      "step": 4575
-    },
-    {
-      "epoch": 0.2123319425127492,
-      "grad_norm": 3.548717498779297,
-      "learning_rate": 1.9241038676064637e-05,
-      "loss": 0.9903,
-      "step": 4580
-    },
-    {
-      "epoch": 0.21256374594343996,
-      "grad_norm": 4.380881309509277,
-      "learning_rate": 1.9237943567136905e-05,
-      "loss": 1.0492,
-      "step": 4585
-    },
-    {
-      "epoch": 0.21279554937413073,
-      "grad_norm": 4.449493408203125,
-      "learning_rate": 1.9234842409959283e-05,
-      "loss": 1.1464,
-      "step": 4590
-    },
-    {
-      "epoch": 0.21302735280482152,
-      "grad_norm": 4.247632026672363,
-      "learning_rate": 1.9231735206562154e-05,
-      "loss": 0.8259,
-      "step": 4595
-    },
-    {
-      "epoch": 0.21325915623551228,
-      "grad_norm": 3.8416435718536377,
-      "learning_rate": 1.9228621958979862e-05,
-      "loss": 1.0981,
-      "step": 4600
-    },
-    {
-      "epoch": 0.21325915623551228,
-      "eval_loss": 1.0158852338790894,
-      "eval_runtime": 11.2758,
-      "eval_samples_per_second": 11.263,
-      "eval_steps_per_second": 11.263,
-      "step": 4600
-    },
-    {
-      "epoch": 0.21349095966620307,
-      "grad_norm": 4.114621162414551,
-      "learning_rate": 1.922550266925071e-05,
-      "loss": 0.9712,
-      "step": 4605
-    },
-    {
-      "epoch": 0.21372276309689384,
-      "grad_norm": 4.465586185455322,
-      "learning_rate": 1.9222377339416945e-05,
-      "loss": 0.9613,
-      "step": 4610
-    },
-    {
-      "epoch": 0.2139545665275846,
-      "grad_norm": 5.107663154602051,
-      "learning_rate": 1.921924597152479e-05,
-      "loss": 1.0306,
-      "step": 4615
-    },
-    {
-      "epoch": 0.2141863699582754,
-      "grad_norm": 3.389920949935913,
-      "learning_rate": 1.9216108567624397e-05,
-      "loss": 0.7988,
-      "step": 4620
-    },
-    {
-      "epoch": 0.21441817338896615,
-      "grad_norm": 4.314112186431885,
-      "learning_rate": 1.921296512976989e-05,
-      "loss": 0.9554,
-      "step": 4625
-    },
-    {
-      "epoch": 0.21464997681965692,
-      "grad_norm": 4.409841060638428,
-      "learning_rate": 1.920981566001933e-05,
-      "loss": 0.9666,
-      "step": 4630
-    },
-    {
-      "epoch": 0.2148817802503477,
-      "grad_norm": 3.8194494247436523,
-      "learning_rate": 1.9206660160434733e-05,
-      "loss": 1.0888,
-      "step": 4635
-    },
-    {
-      "epoch": 0.21511358368103847,
-      "grad_norm": 4.668557167053223,
-      "learning_rate": 1.9203498633082062e-05,
-      "loss": 0.9607,
-      "step": 4640
-    },
-    {
-      "epoch": 0.21534538711172926,
-      "grad_norm": 4.180658340454102,
-      "learning_rate": 1.9200331080031228e-05,
-      "loss": 1.1876,
-      "step": 4645
-    },
-    {
-      "epoch": 0.21557719054242003,
-      "grad_norm": 3.8917407989501953,
-      "learning_rate": 1.9197157503356084e-05,
-      "loss": 0.8905,
-      "step": 4650
-    },
-    {
-      "epoch": 0.2158089939731108,
-      "grad_norm": 3.4583301544189453,
-      "learning_rate": 1.919397790513443e-05,
-      "loss": 1.0806,
-      "step": 4655
-    },
-    {
-      "epoch": 0.21604079740380158,
-      "grad_norm": 4.301894664764404,
-      "learning_rate": 1.9190792287448007e-05,
-      "loss": 1.0075,
-      "step": 4660
-    },
-    {
-      "epoch": 0.21627260083449235,
-      "grad_norm": 3.784499168395996,
-      "learning_rate": 1.9187600652382495e-05,
-      "loss": 0.8791,
-      "step": 4665
-    },
-    {
-      "epoch": 0.21650440426518314,
-      "grad_norm": 4.241847515106201,
-      "learning_rate": 1.9184403002027516e-05,
-      "loss": 0.8495,
-      "step": 4670
-    },
-    {
-      "epoch": 0.2167362076958739,
-      "grad_norm": 5.300185203552246,
-      "learning_rate": 1.918119933847663e-05,
-      "loss": 0.9325,
-      "step": 4675
-    },
-    {
-      "epoch": 0.21696801112656466,
-      "grad_norm": 4.148271560668945,
-      "learning_rate": 1.9177989663827336e-05,
-      "loss": 1.131,
-      "step": 4680
-    },
-    {
-      "epoch": 0.21719981455725546,
-      "grad_norm": 4.2981696128845215,
-      "learning_rate": 1.9174773980181066e-05,
-      "loss": 1.1398,
-      "step": 4685
-    },
-    {
-      "epoch": 0.21743161798794622,
-      "grad_norm": 4.249345302581787,
-      "learning_rate": 1.917155228964318e-05,
-      "loss": 0.9213,
-      "step": 4690
-    },
-    {
-      "epoch": 0.21766342141863698,
-      "grad_norm": 3.92807674407959,
-      "learning_rate": 1.9168324594322988e-05,
-      "loss": 0.9589,
-      "step": 4695
-    },
-    {
-      "epoch": 0.21789522484932777,
-      "grad_norm": 4.179681777954102,
-      "learning_rate": 1.9165090896333717e-05,
-      "loss": 1.0402,
-      "step": 4700
-    },
-    {
-      "epoch": 0.21789522484932777,
-      "eval_loss": 1.0108238458633423,
-      "eval_runtime": 11.2618,
-      "eval_samples_per_second": 11.277,
-      "eval_steps_per_second": 11.277,
-      "step": 4700
-    },
-    {
-      "epoch": 0.21812702828001854,
-      "grad_norm": 4.779577732086182,
-      "learning_rate": 1.9161851197792523e-05,
-      "loss": 0.9088,
-      "step": 4705
-    },
-    {
-      "epoch": 0.21835883171070933,
-      "grad_norm": 4.182905673980713,
-      "learning_rate": 1.91586055008205e-05,
-      "loss": 1.0959,
-      "step": 4710
-    },
-    {
-      "epoch": 0.2185906351414001,
-      "grad_norm": 5.201434135437012,
-      "learning_rate": 1.9155353807542666e-05,
-      "loss": 1.1631,
-      "step": 4715
-    },
-    {
-      "epoch": 0.21882243857209085,
-      "grad_norm": 4.404604911804199,
-      "learning_rate": 1.915209612008796e-05,
-      "loss": 1.1433,
-      "step": 4720
-    },
-    {
-      "epoch": 0.21905424200278165,
-      "grad_norm": 3.8333659172058105,
-      "learning_rate": 1.9148832440589252e-05,
-      "loss": 0.9584,
-      "step": 4725
-    },
-    {
-      "epoch": 0.2192860454334724,
-      "grad_norm": 4.002883434295654,
-      "learning_rate": 1.9145562771183333e-05,
-      "loss": 1.0783,
-      "step": 4730
-    },
-    {
-      "epoch": 0.2195178488641632,
-      "grad_norm": 4.002244472503662,
-      "learning_rate": 1.9142287114010915e-05,
-      "loss": 1.1606,
-      "step": 4735
-    },
-    {
-      "epoch": 0.21974965229485396,
-      "grad_norm": 4.178402900695801,
-      "learning_rate": 1.9139005471216627e-05,
-      "loss": 1.0444,
-      "step": 4740
-    },
-    {
-      "epoch": 0.21998145572554473,
-      "grad_norm": 4.380372047424316,
-      "learning_rate": 1.9135717844949023e-05,
-      "loss": 1.001,
-      "step": 4745
-    },
-    {
-      "epoch": 0.22021325915623552,
-      "grad_norm": 5.0303168296813965,
-      "learning_rate": 1.9132424237360572e-05,
-      "loss": 0.9649,
-      "step": 4750
-    },
-    {
-      "epoch": 0.22044506258692628,
-      "grad_norm": 4.722301483154297,
-      "learning_rate": 1.912912465060766e-05,
-      "loss": 1.0605,
-      "step": 4755
-    },
-    {
-      "epoch": 0.22067686601761707,
-      "grad_norm": 4.6185102462768555,
-      "learning_rate": 1.9125819086850582e-05,
-      "loss": 0.9996,
-      "step": 4760
-    },
-    {
-      "epoch": 0.22090866944830784,
-      "grad_norm": 4.793702602386475,
-      "learning_rate": 1.9122507548253555e-05,
-      "loss": 1.0098,
-      "step": 4765
-    },
-    {
-      "epoch": 0.2211404728789986,
-      "grad_norm": 4.518040180206299,
-      "learning_rate": 1.9119190036984702e-05,
-      "loss": 0.8762,
-      "step": 4770
-    },
-    {
-      "epoch": 0.2213722763096894,
-      "grad_norm": 3.88997745513916,
-      "learning_rate": 1.9115866555216058e-05,
-      "loss": 1.1243,
-      "step": 4775
-    },
-    {
-      "epoch": 0.22160407974038016,
-      "grad_norm": 0.39716479182243347,
-      "learning_rate": 1.9112537105123567e-05,
-      "loss": 0.9426,
-      "step": 4780
-    },
-    {
-      "epoch": 0.22183588317107092,
-      "grad_norm": 3.8836007118225098,
-      "learning_rate": 1.910920168888708e-05,
-      "loss": 0.9133,
-      "step": 4785
-    },
-    {
-      "epoch": 0.2220676866017617,
-      "grad_norm": 4.020860195159912,
-      "learning_rate": 1.910586030869036e-05,
-      "loss": 1.0884,
-      "step": 4790
-    },
-    {
-      "epoch": 0.22229949003245247,
-      "grad_norm": 3.6284425258636475,
-      "learning_rate": 1.9102512966721063e-05,
-      "loss": 1.0384,
-      "step": 4795
-    },
-    {
-      "epoch": 0.22253129346314326,
-      "grad_norm": 4.15256929397583,
-      "learning_rate": 1.9099159665170758e-05,
-      "loss": 0.9094,
-      "step": 4800
-    },
-    {
-      "epoch": 0.22253129346314326,
-      "eval_loss": 1.0125683546066284,
-      "eval_runtime": 11.2696,
-      "eval_samples_per_second": 11.269,
-      "eval_steps_per_second": 11.269,
-      "step": 4800
-    },
-    {
-      "epoch": 0.22276309689383403,
-      "grad_norm": 5.169234275817871,
-      "learning_rate": 1.909580040623491e-05,
-      "loss": 1.1009,
-      "step": 4805
-    },
-    {
-      "epoch": 0.2229949003245248,
-      "grad_norm": 3.830695867538452,
-      "learning_rate": 1.909243519211289e-05,
-      "loss": 0.8057,
-      "step": 4810
-    },
-    {
-      "epoch": 0.22322670375521558,
-      "grad_norm": 3.351536989212036,
-      "learning_rate": 1.9089064025007958e-05,
-      "loss": 0.8803,
-      "step": 4815
-    },
-    {
-      "epoch": 0.22345850718590635,
-      "grad_norm": 3.44328236579895,
-      "learning_rate": 1.908568690712729e-05,
-      "loss": 0.8897,
-      "step": 4820
-    },
-    {
-      "epoch": 0.22369031061659714,
-      "grad_norm": 3.884352922439575,
-      "learning_rate": 1.908230384068194e-05,
-      "loss": 1.0575,
-      "step": 4825
-    },
-    {
-      "epoch": 0.2239221140472879,
-      "grad_norm": 3.6071789264678955,
-      "learning_rate": 1.907891482788686e-05,
-      "loss": 1.0257,
-      "step": 4830
-    },
-    {
-      "epoch": 0.22415391747797866,
-      "grad_norm": 4.853216648101807,
-      "learning_rate": 1.9075519870960906e-05,
-      "loss": 0.9984,
-      "step": 4835
-    },
-    {
-      "epoch": 0.22438572090866946,
-      "grad_norm": 4.0743632316589355,
-      "learning_rate": 1.9072118972126814e-05,
-      "loss": 1.0508,
-      "step": 4840
-    },
-    {
-      "epoch": 0.22461752433936022,
-      "grad_norm": 3.989567995071411,
-      "learning_rate": 1.9068712133611216e-05,
-      "loss": 0.9234,
-      "step": 4845
-    },
-    {
-      "epoch": 0.22484932777005098,
-      "grad_norm": 4.845396518707275,
-      "learning_rate": 1.9065299357644628e-05,
-      "loss": 0.8949,
-      "step": 4850
-    },
-    {
-      "epoch": 0.22508113120074177,
-      "grad_norm": 3.8650968074798584,
-      "learning_rate": 1.9061880646461457e-05,
-      "loss": 0.9957,
-      "step": 4855
-    },
-    {
-      "epoch": 0.22531293463143254,
-      "grad_norm": 4.8118181228637695,
-      "learning_rate": 1.90584560023e-05,
-      "loss": 0.8788,
-      "step": 4860
-    },
-    {
-      "epoch": 0.22554473806212333,
-      "grad_norm": 3.8270719051361084,
-      "learning_rate": 1.905502542740243e-05,
-      "loss": 1.0374,
-      "step": 4865
-    },
-    {
-      "epoch": 0.2257765414928141,
-      "grad_norm": 4.53631067276001,
-      "learning_rate": 1.9051588924014804e-05,
-      "loss": 0.9582,
-      "step": 4870
-    },
-    {
-      "epoch": 0.22600834492350486,
-      "grad_norm": 3.948395013809204,
-      "learning_rate": 1.904814649438707e-05,
-      "loss": 0.9951,
-      "step": 4875
-    },
-    {
-      "epoch": 0.22624014835419565,
-      "grad_norm": 7.741011142730713,
-      "learning_rate": 1.904469814077305e-05,
-      "loss": 0.9671,
-      "step": 4880
-    },
-    {
-      "epoch": 0.2264719517848864,
-      "grad_norm": 4.1785712242126465,
-      "learning_rate": 1.9041243865430436e-05,
-      "loss": 1.0577,
-      "step": 4885
-    },
-    {
-      "epoch": 0.2267037552155772,
-      "grad_norm": 4.016406059265137,
-      "learning_rate": 1.903778367062081e-05,
-      "loss": 0.8957,
-      "step": 4890
-    },
-    {
-      "epoch": 0.22693555864626797,
-      "grad_norm": 3.7880759239196777,
-      "learning_rate": 1.9034317558609624e-05,
-      "loss": 1.0002,
-      "step": 4895
-    },
-    {
-      "epoch": 0.22716736207695873,
-      "grad_norm": 3.6266255378723145,
-      "learning_rate": 1.9030845531666203e-05,
-      "loss": 0.7714,
-      "step": 4900
-    },
-    {
-      "epoch": 0.22716736207695873,
-      "eval_loss": 1.0120409727096558,
-      "eval_runtime": 11.2606,
-      "eval_samples_per_second": 11.278,
-      "eval_steps_per_second": 11.278,
-      "step": 4900
-    },
-    {
-      "epoch": 0.22739916550764952,
-      "grad_norm": 4.234470367431641,
-      "learning_rate": 1.902736759206375e-05,
-      "loss": 1.0156,
-      "step": 4905
-    },
-    {
-      "epoch": 0.22763096893834028,
-      "grad_norm": 3.4304628372192383,
-      "learning_rate": 1.9023883742079337e-05,
-      "loss": 1.0283,
-      "step": 4910
-    },
-    {
-      "epoch": 0.22786277236903107,
-      "grad_norm": 4.131404399871826,
-      "learning_rate": 1.9020393983993895e-05,
-      "loss": 0.9577,
-      "step": 4915
-    },
-    {
-      "epoch": 0.22809457579972184,
-      "grad_norm": 3.733011484146118,
-      "learning_rate": 1.9016898320092242e-05,
-      "loss": 0.9742,
-      "step": 4920
-    },
-    {
-      "epoch": 0.2283263792304126,
-      "grad_norm": 4.668416976928711,
-      "learning_rate": 1.9013396752663044e-05,
-      "loss": 1.0222,
-      "step": 4925
-    },
-    {
-      "epoch": 0.2285581826611034,
-      "grad_norm": 4.214738845825195,
-      "learning_rate": 1.9009889283998847e-05,
-      "loss": 0.9623,
-      "step": 4930
-    },
-    {
-      "epoch": 0.22878998609179416,
-      "grad_norm": 3.838366985321045,
-      "learning_rate": 1.900637591639605e-05,
-      "loss": 0.9191,
-      "step": 4935
-    },
-    {
-      "epoch": 0.22902178952248492,
-      "grad_norm": 3.8003344535827637,
-      "learning_rate": 1.9002856652154925e-05,
-      "loss": 0.8639,
-      "step": 4940
-    },
-    {
-      "epoch": 0.2292535929531757,
-      "grad_norm": 3.651313543319702,
-      "learning_rate": 1.8999331493579594e-05,
-      "loss": 1.012,
-      "step": 4945
-    },
-    {
-      "epoch": 0.22948539638386647,
-      "grad_norm": 4.221668720245361,
-      "learning_rate": 1.8995800442978044e-05,
-      "loss": 0.9683,
-      "step": 4950
-    },
-    {
-      "epoch": 0.22971719981455727,
-      "grad_norm": 4.224164009094238,
-      "learning_rate": 1.8992263502662114e-05,
-      "loss": 1.0671,
-      "step": 4955
-    },
-    {
-      "epoch": 0.22994900324524803,
-      "grad_norm": 4.996068000793457,
-      "learning_rate": 1.8988720674947512e-05,
-      "loss": 1.1553,
-      "step": 4960
-    },
-    {
-      "epoch": 0.2301808066759388,
-      "grad_norm": 4.583253860473633,
-      "learning_rate": 1.8985171962153784e-05,
-      "loss": 0.9906,
-      "step": 4965
-    },
-    {
-      "epoch": 0.23041261010662958,
-      "grad_norm": 4.157360553741455,
-      "learning_rate": 1.898161736660434e-05,
-      "loss": 1.017,
-      "step": 4970
-    },
-    {
-      "epoch": 0.23064441353732035,
-      "grad_norm": 3.966968297958374,
-      "learning_rate": 1.897805689062644e-05,
-      "loss": 1.0456,
-      "step": 4975
-    },
-    {
-      "epoch": 0.23087621696801114,
-      "grad_norm": 5.055805683135986,
-      "learning_rate": 1.8974490536551186e-05,
-      "loss": 0.9645,
-      "step": 4980
-    },
-    {
-      "epoch": 0.2311080203987019,
-      "grad_norm": 5.330160617828369,
-      "learning_rate": 1.897091830671354e-05,
-      "loss": 0.9688,
-      "step": 4985
-    },
-    {
-      "epoch": 0.23133982382939267,
-      "grad_norm": 4.764847755432129,
-      "learning_rate": 1.896734020345231e-05,
-      "loss": 1.0223,
-      "step": 4990
-    },
-    {
-      "epoch": 0.23157162726008346,
-      "grad_norm": 3.9281156063079834,
-      "learning_rate": 1.896375622911014e-05,
-      "loss": 0.9495,
-      "step": 4995
-    },
-    {
-      "epoch": 0.23180343069077422,
-      "grad_norm": 4.123696327209473,
-      "learning_rate": 1.896016638603352e-05,
-      "loss": 0.8876,
-      "step": 5000
-    },
-    {
-      "epoch": 0.23180343069077422,
-      "eval_loss": 1.0095674991607666,
-      "eval_runtime": 11.2932,
-      "eval_samples_per_second": 11.246,
-      "eval_steps_per_second": 11.246,
-      "step": 5000
-    },
-    {
-      "epoch": 0.23203523412146498,
-      "grad_norm": 4.128748416900635,
-      "learning_rate": 1.8956570676572793e-05,
-      "loss": 0.9017,
-      "step": 5005
-    },
-    {
-      "epoch": 0.23226703755215578,
-      "grad_norm": 4.058050155639648,
-      "learning_rate": 1.895296910308213e-05,
-      "loss": 0.9553,
-      "step": 5010
-    },
-    {
-      "epoch": 0.23249884098284654,
-      "grad_norm": 4.250258445739746,
-      "learning_rate": 1.8949361667919552e-05,
-      "loss": 0.9679,
-      "step": 5015
-    },
-    {
-      "epoch": 0.23273064441353733,
-      "grad_norm": 5.300534248352051,
-      "learning_rate": 1.8945748373446904e-05,
-      "loss": 0.9784,
-      "step": 5020
-    },
-    {
-      "epoch": 0.2329624478442281,
-      "grad_norm": 4.7261881828308105,
-      "learning_rate": 1.894212922202989e-05,
-      "loss": 1.0984,
-      "step": 5025
-    },
-    {
-      "epoch": 0.23319425127491886,
-      "grad_norm": 4.31865930557251,
-      "learning_rate": 1.893850421603802e-05,
-      "loss": 1.1129,
-      "step": 5030
-    },
-    {
-      "epoch": 0.23342605470560965,
-      "grad_norm": 3.7286322116851807,
-      "learning_rate": 1.8934873357844654e-05,
-      "loss": 0.9771,
-      "step": 5035
-    },
-    {
-      "epoch": 0.2336578581363004,
-      "grad_norm": 3.7164711952209473,
-      "learning_rate": 1.8931236649826988e-05,
-      "loss": 0.8744,
-      "step": 5040
-    },
-    {
-      "epoch": 0.2338896615669912,
-      "grad_norm": 4.368716716766357,
-      "learning_rate": 1.892759409436603e-05,
-      "loss": 0.9397,
-      "step": 5045
-    },
-    {
-      "epoch": 0.23412146499768197,
-      "grad_norm": 3.451289415359497,
-      "learning_rate": 1.8923945693846642e-05,
-      "loss": 0.9761,
-      "step": 5050
-    },
-    {
-      "epoch": 0.23435326842837273,
-      "grad_norm": 4.144486904144287,
-      "learning_rate": 1.8920291450657487e-05,
-      "loss": 0.959,
-      "step": 5055
-    },
-    {
-      "epoch": 0.23458507185906352,
-      "grad_norm": 4.257894515991211,
-      "learning_rate": 1.8916631367191072e-05,
-      "loss": 0.8989,
-      "step": 5060
-    },
-    {
-      "epoch": 0.23481687528975428,
-      "grad_norm": 4.326287269592285,
-      "learning_rate": 1.891296544584371e-05,
-      "loss": 0.9968,
-      "step": 5065
-    },
-    {
-      "epoch": 0.23504867872044508,
-      "grad_norm": 5.085713863372803,
-      "learning_rate": 1.890929368901556e-05,
-      "loss": 0.9061,
-      "step": 5070
-    },
-    {
-      "epoch": 0.23528048215113584,
-      "grad_norm": 3.4437763690948486,
-      "learning_rate": 1.8905616099110578e-05,
-      "loss": 0.9122,
-      "step": 5075
-    },
-    {
-      "epoch": 0.2355122855818266,
-      "grad_norm": 3.675459384918213,
-      "learning_rate": 1.890193267853655e-05,
-      "loss": 0.9841,
-      "step": 5080
-    },
-    {
-      "epoch": 0.2357440890125174,
-      "grad_norm": 3.7504916191101074,
-      "learning_rate": 1.8898243429705082e-05,
-      "loss": 0.9003,
-      "step": 5085
-    },
-    {
-      "epoch": 0.23597589244320816,
-      "grad_norm": 5.602362155914307,
-      "learning_rate": 1.889454835503159e-05,
-      "loss": 1.0204,
-      "step": 5090
-    },
-    {
-      "epoch": 0.23620769587389892,
-      "grad_norm": 4.032191753387451,
-      "learning_rate": 1.8890847456935305e-05,
-      "loss": 0.8905,
-      "step": 5095
-    },
-    {
-      "epoch": 0.2364394993045897,
-      "grad_norm": 3.700024127960205,
-      "learning_rate": 1.888714073783928e-05,
-      "loss": 0.9313,
-      "step": 5100
-    },
-    {
-      "epoch": 0.2364394993045897,
-      "eval_loss": 1.0069371461868286,
-      "eval_runtime": 11.262,
-      "eval_samples_per_second": 11.277,
-      "eval_steps_per_second": 11.277,
-      "step": 5100
-    },
-    {
-      "epoch": 0.23667130273528048,
-      "grad_norm": 3.9474759101867676,
-      "learning_rate": 1.8883428200170365e-05,
-      "loss": 0.9212,
-      "step": 5105
-    },
-    {
-      "epoch": 0.23690310616597127,
-      "grad_norm": 4.089510440826416,
-      "learning_rate": 1.8879709846359227e-05,
-      "loss": 1.0069,
-      "step": 5110
-    },
-    {
-      "epoch": 0.23713490959666203,
-      "grad_norm": 3.7710094451904297,
-      "learning_rate": 1.887598567884034e-05,
-      "loss": 0.9985,
-      "step": 5115
-    },
-    {
-      "epoch": 0.2373667130273528,
-      "grad_norm": 4.190286636352539,
-      "learning_rate": 1.8872255700051988e-05,
-      "loss": 0.8677,
-      "step": 5120
-    },
-    {
-      "epoch": 0.23759851645804358,
-      "grad_norm": 3.9296810626983643,
-      "learning_rate": 1.8868519912436247e-05,
-      "loss": 0.9758,
-      "step": 5125
-    },
-    {
-      "epoch": 0.23783031988873435,
-      "grad_norm": 3.989487409591675,
-      "learning_rate": 1.8864778318439013e-05,
-      "loss": 0.9516,
-      "step": 5130
-    },
-    {
-      "epoch": 0.23806212331942514,
-      "grad_norm": 4.058485984802246,
-      "learning_rate": 1.8861030920509976e-05,
-      "loss": 0.9176,
-      "step": 5135
-    },
-    {
-      "epoch": 0.2382939267501159,
-      "grad_norm": 4.617326736450195,
-      "learning_rate": 1.885727772110262e-05,
-      "loss": 0.9109,
-      "step": 5140
-    },
-    {
-      "epoch": 0.23852573018080667,
-      "grad_norm": 4.54904317855835,
-      "learning_rate": 1.8853518722674237e-05,
-      "loss": 1.1414,
-      "step": 5145
-    },
-    {
-      "epoch": 0.23875753361149746,
-      "grad_norm": 3.805340528488159,
-      "learning_rate": 1.8849753927685915e-05,
-      "loss": 1.0677,
-      "step": 5150
-    },
-    {
-      "epoch": 0.23898933704218822,
-      "grad_norm": 4.674151420593262,
-      "learning_rate": 1.8845983338602524e-05,
-      "loss": 0.9734,
-      "step": 5155
-    },
-    {
-      "epoch": 0.23922114047287898,
-      "grad_norm": 4.656925201416016,
-      "learning_rate": 1.884220695789275e-05,
-      "loss": 1.083,
-      "step": 5160
-    },
-    {
-      "epoch": 0.23945294390356978,
-      "grad_norm": 4.5245256423950195,
-      "learning_rate": 1.8838424788029047e-05,
-      "loss": 1.0778,
-      "step": 5165
-    },
-    {
-      "epoch": 0.23968474733426054,
-      "grad_norm": 6.588644981384277,
-      "learning_rate": 1.8834636831487673e-05,
-      "loss": 1.0181,
-      "step": 5170
-    },
-    {
-      "epoch": 0.23991655076495133,
-      "grad_norm": 3.7076587677001953,
-      "learning_rate": 1.8830843090748678e-05,
-      "loss": 0.9473,
-      "step": 5175
-    },
-    {
-      "epoch": 0.2401483541956421,
-      "grad_norm": 4.623571395874023,
-      "learning_rate": 1.8827043568295887e-05,
-      "loss": 1.0638,
-      "step": 5180
-    },
-    {
-      "epoch": 0.24038015762633286,
-      "grad_norm": 4.6283135414123535,
-      "learning_rate": 1.8823238266616922e-05,
-      "loss": 1.0713,
-      "step": 5185
-    },
-    {
-      "epoch": 0.24061196105702365,
-      "grad_norm": 3.885566473007202,
-      "learning_rate": 1.8819427188203177e-05,
-      "loss": 0.9161,
-      "step": 5190
-    },
-    {
-      "epoch": 0.2408437644877144,
-      "grad_norm": 4.2143402099609375,
-      "learning_rate": 1.881561033554984e-05,
-      "loss": 1.1575,
-      "step": 5195
-    },
-    {
-      "epoch": 0.2410755679184052,
-      "grad_norm": 4.28688907623291,
-      "learning_rate": 1.881178771115587e-05,
-      "loss": 1.0844,
-      "step": 5200
-    },
-    {
-      "epoch": 0.2410755679184052,
-      "eval_loss": 1.0090265274047852,
-      "eval_runtime": 11.2654,
-      "eval_samples_per_second": 11.273,
-      "eval_steps_per_second": 11.273,
-      "step": 5200
-    },
-    {
-      "epoch": 0.24130737134909597,
-      "grad_norm": 6.4583306312561035,
-      "learning_rate": 1.880795931752401e-05,
-      "loss": 1.1072,
-      "step": 5205
-    },
-    {
-      "epoch": 0.24153917477978673,
-      "grad_norm": 3.783980369567871,
-      "learning_rate": 1.8804125157160777e-05,
-      "loss": 1.0074,
-      "step": 5210
-    },
-    {
-      "epoch": 0.24177097821047752,
-      "grad_norm": 3.996311902999878,
-      "learning_rate": 1.8800285232576473e-05,
-      "loss": 1.0447,
-      "step": 5215
-    },
-    {
-      "epoch": 0.24200278164116829,
-      "grad_norm": 4.2546467781066895,
-      "learning_rate": 1.879643954628516e-05,
-      "loss": 0.9928,
-      "step": 5220
-    },
-    {
-      "epoch": 0.24223458507185908,
-      "grad_norm": 3.8486084938049316,
-      "learning_rate": 1.8792588100804677e-05,
-      "loss": 1.0198,
-      "step": 5225
-    },
-    {
-      "epoch": 0.24246638850254984,
-      "grad_norm": 3.533219575881958,
-      "learning_rate": 1.8788730898656645e-05,
-      "loss": 0.9151,
-      "step": 5230
-    },
-    {
-      "epoch": 0.2426981919332406,
-      "grad_norm": 4.661604404449463,
-      "learning_rate": 1.878486794236643e-05,
-      "loss": 1.0086,
-      "step": 5235
-    },
-    {
-      "epoch": 0.2429299953639314,
-      "grad_norm": 3.634969472885132,
-      "learning_rate": 1.8780999234463195e-05,
-      "loss": 1.144,
-      "step": 5240
-    },
-    {
-      "epoch": 0.24316179879462216,
-      "grad_norm": 4.38676643371582,
-      "learning_rate": 1.8777124777479846e-05,
-      "loss": 1.0801,
-      "step": 5245
-    },
-    {
-      "epoch": 0.24339360222531292,
-      "grad_norm": 3.5356669425964355,
-      "learning_rate": 1.8773244573953058e-05,
-      "loss": 0.6998,
-      "step": 5250
-    },
-    {
-      "epoch": 0.2436254056560037,
-      "grad_norm": 3.8418920040130615,
-      "learning_rate": 1.8769358626423276e-05,
-      "loss": 0.818,
-      "step": 5255
-    },
-    {
-      "epoch": 0.24385720908669448,
-      "grad_norm": 4.476934432983398,
-      "learning_rate": 1.87654669374347e-05,
-      "loss": 1.0859,
-      "step": 5260
-    },
-    {
-      "epoch": 0.24408901251738527,
-      "grad_norm": 4.2773566246032715,
-      "learning_rate": 1.8761569509535284e-05,
-      "loss": 1.1232,
-      "step": 5265
-    },
-    {
-      "epoch": 0.24432081594807603,
-      "grad_norm": 3.773895740509033,
-      "learning_rate": 1.8757666345276754e-05,
-      "loss": 0.8429,
-      "step": 5270
-    },
-    {
-      "epoch": 0.2445526193787668,
-      "grad_norm": 3.9892804622650146,
-      "learning_rate": 1.8753757447214574e-05,
-      "loss": 1.0794,
-      "step": 5275
-    },
-    {
-      "epoch": 0.24478442280945759,
-      "grad_norm": 4.0765767097473145,
-      "learning_rate": 1.874984281790798e-05,
-      "loss": 0.9931,
-      "step": 5280
-    },
-    {
-      "epoch": 0.24501622624014835,
-      "grad_norm": 3.730342388153076,
-      "learning_rate": 1.8745922459919944e-05,
-      "loss": 0.8929,
-      "step": 5285
-    },
-    {
-      "epoch": 0.24524802967083914,
-      "grad_norm": 4.401835918426514,
-      "learning_rate": 1.8741996375817196e-05,
-      "loss": 1.0135,
-      "step": 5290
-    },
-    {
-      "epoch": 0.2454798331015299,
-      "grad_norm": 3.8997461795806885,
-      "learning_rate": 1.8738064568170217e-05,
-      "loss": 0.9449,
-      "step": 5295
-    },
-    {
-      "epoch": 0.24571163653222067,
-      "grad_norm": 4.102304935455322,
-      "learning_rate": 1.8734127039553235e-05,
-      "loss": 0.8591,
-      "step": 5300
-    },
-    {
-      "epoch": 0.24571163653222067,
-      "eval_loss": 1.0059479475021362,
-      "eval_runtime": 11.265,
-      "eval_samples_per_second": 11.274,
-      "eval_steps_per_second": 11.274,
-      "step": 5300
-    },
-    {
-      "epoch": 0.24594343996291146,
-      "grad_norm": 4.207913875579834,
-      "learning_rate": 1.8730183792544218e-05,
-      "loss": 1.1101,
-      "step": 5305
-    },
-    {
-      "epoch": 0.24617524339360222,
-      "grad_norm": 3.844038963317871,
-      "learning_rate": 1.8726234829724883e-05,
-      "loss": 0.9573,
-      "step": 5310
-    },
-    {
-      "epoch": 0.24640704682429299,
-      "grad_norm": 4.369620323181152,
-      "learning_rate": 1.8722280153680688e-05,
-      "loss": 0.9645,
-      "step": 5315
-    },
-    {
-      "epoch": 0.24663885025498378,
-      "grad_norm": 5.992620944976807,
-      "learning_rate": 1.871831976700083e-05,
-      "loss": 1.0597,
-      "step": 5320
-    },
-    {
-      "epoch": 0.24687065368567454,
-      "grad_norm": 5.34641695022583,
-      "learning_rate": 1.8714353672278246e-05,
-      "loss": 1.0403,
-      "step": 5325
-    },
-    {
-      "epoch": 0.24710245711636533,
-      "grad_norm": 3.386537551879883,
-      "learning_rate": 1.8710381872109616e-05,
-      "loss": 0.8237,
-      "step": 5330
-    },
-    {
-      "epoch": 0.2473342605470561,
-      "grad_norm": 3.749030828475952,
-      "learning_rate": 1.8706404369095344e-05,
-      "loss": 0.9373,
-      "step": 5335
-    },
-    {
-      "epoch": 0.24756606397774686,
-      "grad_norm": 3.7837107181549072,
-      "learning_rate": 1.8702421165839578e-05,
-      "loss": 1.0784,
-      "step": 5340
-    },
-    {
-      "epoch": 0.24779786740843765,
-      "grad_norm": 4.194408416748047,
-      "learning_rate": 1.869843226495019e-05,
-      "loss": 1.0148,
-      "step": 5345
-    },
-    {
-      "epoch": 0.2480296708391284,
-      "grad_norm": 4.585747241973877,
-      "learning_rate": 1.869443766903878e-05,
-      "loss": 0.8059,
-      "step": 5350
-    },
-    {
-      "epoch": 0.2482614742698192,
-      "grad_norm": 3.960101842880249,
-      "learning_rate": 1.8690437380720697e-05,
-      "loss": 0.9932,
-      "step": 5355
-    },
-    {
-      "epoch": 0.24849327770050997,
-      "grad_norm": 3.7180113792419434,
-      "learning_rate": 1.8686431402614996e-05,
-      "loss": 1.0071,
-      "step": 5360
-    },
-    {
-      "epoch": 0.24872508113120073,
-      "grad_norm": 3.8417141437530518,
-      "learning_rate": 1.8682419737344465e-05,
-      "loss": 0.9517,
-      "step": 5365
-    },
-    {
-      "epoch": 0.24895688456189152,
-      "grad_norm": 4.174343109130859,
-      "learning_rate": 1.8678402387535608e-05,
-      "loss": 0.9137,
-      "step": 5370
-    },
-    {
-      "epoch": 0.2491886879925823,
-      "grad_norm": 3.7449638843536377,
-      "learning_rate": 1.8674379355818664e-05,
-      "loss": 0.8484,
-      "step": 5375
-    },
-    {
-      "epoch": 0.24942049142327308,
-      "grad_norm": 3.8778748512268066,
-      "learning_rate": 1.8670350644827585e-05,
-      "loss": 0.9259,
-      "step": 5380
-    },
-    {
-      "epoch": 0.24965229485396384,
-      "grad_norm": 4.755029678344727,
-      "learning_rate": 1.8666316257200035e-05,
-      "loss": 1.1386,
-      "step": 5385
-    },
-    {
-      "epoch": 0.2498840982846546,
-      "grad_norm": 4.226649761199951,
-      "learning_rate": 1.8662276195577407e-05,
-      "loss": 0.9777,
-      "step": 5390
-    },
-    {
-      "epoch": 0.25011590171534537,
-      "grad_norm": 4.788735866546631,
-      "learning_rate": 1.8658230462604804e-05,
-      "loss": 0.9213,
-      "step": 5395
-    },
-    {
-      "epoch": 0.25034770514603616,
-      "grad_norm": 4.801352500915527,
-      "learning_rate": 1.8654179060931035e-05,
-      "loss": 1.0886,
-      "step": 5400
-    },
-    {
-      "epoch": 0.25034770514603616,
-      "eval_loss": 1.0038655996322632,
-      "eval_runtime": 11.266,
-      "eval_samples_per_second": 11.273,
-      "eval_steps_per_second": 11.273,
-      "step": 5400
-    },
-    {
-      "epoch": 0.25057950857672695,
-      "grad_norm": 4.064159870147705,
-      "learning_rate": 1.8650121993208632e-05,
-      "loss": 0.8917,
-      "step": 5405
-    },
-    {
-      "epoch": 0.2508113120074177,
-      "grad_norm": 3.9093940258026123,
-      "learning_rate": 1.864605926209383e-05,
-      "loss": 1.015,
-      "step": 5410
-    },
-    {
-      "epoch": 0.2510431154381085,
-      "grad_norm": 3.832493305206299,
-      "learning_rate": 1.8641990870246568e-05,
-      "loss": 0.893,
-      "step": 5415
-    },
-    {
-      "epoch": 0.25127491886879927,
-      "grad_norm": 3.6580631732940674,
-      "learning_rate": 1.86379168203305e-05,
-      "loss": 0.9733,
-      "step": 5420
-    },
-    {
-      "epoch": 0.25150672229949006,
-      "grad_norm": 4.236906051635742,
-      "learning_rate": 1.8633837115012985e-05,
-      "loss": 0.8722,
-      "step": 5425
-    },
-    {
-      "epoch": 0.2517385257301808,
-      "grad_norm": 4.54565954208374,
-      "learning_rate": 1.8629751756965073e-05,
-      "loss": 1.0042,
-      "step": 5430
-    },
-    {
-      "epoch": 0.2519703291608716,
-      "grad_norm": 4.301955699920654,
-      "learning_rate": 1.8625660748861528e-05,
-      "loss": 0.8718,
-      "step": 5435
-    },
-    {
-      "epoch": 0.2522021325915624,
-      "grad_norm": 3.71239972114563,
-      "learning_rate": 1.8621564093380805e-05,
-      "loss": 1.0586,
-      "step": 5440
-    },
-    {
-      "epoch": 0.2524339360222531,
-      "grad_norm": 4.631463050842285,
-      "learning_rate": 1.861746179320506e-05,
-      "loss": 0.8433,
-      "step": 5445
-    },
-    {
-      "epoch": 0.2526657394529439,
-      "grad_norm": 3.705151319503784,
-      "learning_rate": 1.861335385102014e-05,
-      "loss": 0.9236,
-      "step": 5450
-    },
-    {
-      "epoch": 0.2528975428836347,
-      "grad_norm": 4.138430595397949,
-      "learning_rate": 1.8609240269515595e-05,
-      "loss": 0.8659,
-      "step": 5455
-    },
-    {
-      "epoch": 0.25312934631432543,
-      "grad_norm": 3.9012696743011475,
-      "learning_rate": 1.8605121051384657e-05,
-      "loss": 0.8742,
-      "step": 5460
-    },
-    {
-      "epoch": 0.2533611497450162,
-      "grad_norm": 5.521048069000244,
-      "learning_rate": 1.8600996199324253e-05,
-      "loss": 1.1732,
-      "step": 5465
-    },
-    {
-      "epoch": 0.253592953175707,
-      "grad_norm": 3.5360827445983887,
-      "learning_rate": 1.8596865716035005e-05,
-      "loss": 0.9477,
-      "step": 5470
-    },
-    {
-      "epoch": 0.25382475660639775,
-      "grad_norm": 3.5689191818237305,
-      "learning_rate": 1.859272960422121e-05,
-      "loss": 0.9334,
-      "step": 5475
-    },
-    {
-      "epoch": 0.25405656003708854,
-      "grad_norm": 4.756504058837891,
-      "learning_rate": 1.8588587866590858e-05,
-      "loss": 1.0424,
-      "step": 5480
-    },
-    {
-      "epoch": 0.25428836346777933,
-      "grad_norm": 3.3674094676971436,
-      "learning_rate": 1.8584440505855622e-05,
-      "loss": 0.7284,
-      "step": 5485
-    },
-    {
-      "epoch": 0.2545201668984701,
-      "grad_norm": 4.216386795043945,
-      "learning_rate": 1.8580287524730846e-05,
-      "loss": 0.9278,
-      "step": 5490
-    },
-    {
-      "epoch": 0.25475197032916086,
-      "grad_norm": 6.581062316894531,
-      "learning_rate": 1.857612892593557e-05,
-      "loss": 1.0525,
-      "step": 5495
-    },
-    {
-      "epoch": 0.25498377375985165,
-      "grad_norm": 3.360948085784912,
-      "learning_rate": 1.8571964712192504e-05,
-      "loss": 0.8063,
-      "step": 5500
-    },
-    {
-      "epoch": 0.25498377375985165,
-      "eval_loss": 1.003278136253357,
-      "eval_runtime": 11.2614,
-      "eval_samples_per_second": 11.277,
-      "eval_steps_per_second": 11.277,
-      "step": 5500
-    },
-    {
-      "epoch": 0.25521557719054244,
-      "grad_norm": 3.8659751415252686,
-      "learning_rate": 1.8567794886228034e-05,
-      "loss": 0.8668,
-      "step": 5505
-    },
-    {
-      "epoch": 0.2554473806212332,
-      "grad_norm": 4.503422737121582,
-      "learning_rate": 1.856361945077222e-05,
-      "loss": 1.0136,
-      "step": 5510
-    },
-    {
-      "epoch": 0.25567918405192397,
-      "grad_norm": 3.9764323234558105,
-      "learning_rate": 1.8559438408558795e-05,
-      "loss": 1.0744,
-      "step": 5515
-    },
-    {
-      "epoch": 0.25591098748261476,
-      "grad_norm": 3.8709752559661865,
-      "learning_rate": 1.8555251762325164e-05,
-      "loss": 0.8849,
-      "step": 5520
-    },
-    {
-      "epoch": 0.2561427909133055,
-      "grad_norm": 4.205710411071777,
-      "learning_rate": 1.8551059514812402e-05,
-      "loss": 0.841,
-      "step": 5525
-    },
-    {
-      "epoch": 0.2563745943439963,
-      "grad_norm": 3.547140121459961,
-      "learning_rate": 1.8546861668765245e-05,
-      "loss": 1.0945,
-      "step": 5530
-    },
-    {
-      "epoch": 0.2566063977746871,
-      "grad_norm": 3.594226360321045,
-      "learning_rate": 1.8542658226932107e-05,
-      "loss": 1.0526,
-      "step": 5535
-    },
-    {
-      "epoch": 0.2568382012053778,
-      "grad_norm": 3.7059407234191895,
-      "learning_rate": 1.8538449192065052e-05,
-      "loss": 1.078,
-      "step": 5540
-    },
-    {
-      "epoch": 0.2570700046360686,
-      "grad_norm": 3.7842459678649902,
-      "learning_rate": 1.8534234566919815e-05,
-      "loss": 0.8654,
-      "step": 5545
-    },
-    {
-      "epoch": 0.2573018080667594,
-      "grad_norm": 3.989053964614868,
-      "learning_rate": 1.8530014354255788e-05,
-      "loss": 1.0133,
-      "step": 5550
-    },
-    {
-      "epoch": 0.2575336114974502,
-      "grad_norm": 4.875503063201904,
-      "learning_rate": 1.8525788556836016e-05,
-      "loss": 0.9915,
-      "step": 5555
-    },
-    {
-      "epoch": 0.2577654149281409,
-      "grad_norm": 4.440192699432373,
-      "learning_rate": 1.852155717742721e-05,
-      "loss": 1.0343,
-      "step": 5560
-    },
-    {
-      "epoch": 0.2579972183588317,
-      "grad_norm": 4.262231349945068,
-      "learning_rate": 1.851732021879973e-05,
-      "loss": 1.0658,
-      "step": 5565
-    },
-    {
-      "epoch": 0.2582290217895225,
-      "grad_norm": 3.8692004680633545,
-      "learning_rate": 1.851307768372759e-05,
-      "loss": 0.9289,
-      "step": 5570
-    },
-    {
-      "epoch": 0.25846082522021324,
-      "grad_norm": 3.9327476024627686,
-      "learning_rate": 1.850882957498846e-05,
-      "loss": 1.0081,
-      "step": 5575
-    },
-    {
-      "epoch": 0.25869262865090403,
-      "grad_norm": 4.598284721374512,
-      "learning_rate": 1.8504575895363645e-05,
-      "loss": 1.076,
-      "step": 5580
-    },
-    {
-      "epoch": 0.2589244320815948,
-      "grad_norm": 4.525602340698242,
-      "learning_rate": 1.850031664763811e-05,
-      "loss": 0.9856,
-      "step": 5585
-    },
-    {
-      "epoch": 0.25915623551228556,
-      "grad_norm": 4.555290222167969,
-      "learning_rate": 1.8496051834600465e-05,
-      "loss": 1.101,
-      "step": 5590
-    },
-    {
-      "epoch": 0.25938803894297635,
-      "grad_norm": 4.705940246582031,
-      "learning_rate": 1.8491781459042956e-05,
-      "loss": 1.0118,
-      "step": 5595
-    },
-    {
-      "epoch": 0.25961984237366714,
-      "grad_norm": 4.4292521476745605,
-      "learning_rate": 1.848750552376148e-05,
-      "loss": 1.0232,
-      "step": 5600
-    },
-    {
-      "epoch": 0.25961984237366714,
-      "eval_loss": 1.0037119388580322,
-      "eval_runtime": 11.2651,
-      "eval_samples_per_second": 11.274,
-      "eval_steps_per_second": 11.274,
-      "step": 5600
-    },
-    {
-      "epoch": 0.2598516458043579,
-      "grad_norm": 3.5477733612060547,
-      "learning_rate": 1.848322403155557e-05,
-      "loss": 1.073,
-      "step": 5605
-    },
-    {
-      "epoch": 0.26008344923504867,
-      "grad_norm": 4.175229549407959,
-      "learning_rate": 1.8478936985228394e-05,
-      "loss": 1.0311,
-      "step": 5610
-    },
-    {
-      "epoch": 0.26031525266573946,
-      "grad_norm": 4.131260395050049,
-      "learning_rate": 1.8474644387586757e-05,
-      "loss": 0.9146,
-      "step": 5615
-    },
-    {
-      "epoch": 0.26054705609643025,
-      "grad_norm": 3.8669774532318115,
-      "learning_rate": 1.8470346241441105e-05,
-      "loss": 0.9011,
-      "step": 5620
-    },
-    {
-      "epoch": 0.260778859527121,
-      "grad_norm": 4.5490546226501465,
-      "learning_rate": 1.8466042549605518e-05,
-      "loss": 0.9329,
-      "step": 5625
-    },
-    {
-      "epoch": 0.2610106629578118,
-      "grad_norm": 4.513428688049316,
-      "learning_rate": 1.8461733314897688e-05,
-      "loss": 1.1291,
-      "step": 5630
-    },
-    {
-      "epoch": 0.26124246638850257,
-      "grad_norm": 4.3655500411987305,
-      "learning_rate": 1.8457418540138958e-05,
-      "loss": 1.094,
-      "step": 5635
-    },
-    {
-      "epoch": 0.2614742698191933,
-      "grad_norm": 4.182011127471924,
-      "learning_rate": 1.8453098228154287e-05,
-      "loss": 1.0215,
-      "step": 5640
-    },
-    {
-      "epoch": 0.2617060732498841,
-      "grad_norm": 4.210841655731201,
-      "learning_rate": 1.8448772381772266e-05,
-      "loss": 1.0594,
-      "step": 5645
-    },
-    {
-      "epoch": 0.2619378766805749,
-      "grad_norm": 3.889146327972412,
-      "learning_rate": 1.84444410038251e-05,
-      "loss": 1.0411,
-      "step": 5650
-    },
-    {
-      "epoch": 0.2621696801112656,
-      "grad_norm": 4.405303955078125,
-      "learning_rate": 1.8440104097148625e-05,
-      "loss": 1.0183,
-      "step": 5655
-    },
-    {
-      "epoch": 0.2624014835419564,
-      "grad_norm": 3.846829891204834,
-      "learning_rate": 1.843576166458229e-05,
-      "loss": 0.8977,
-      "step": 5660
-    },
-    {
-      "epoch": 0.2626332869726472,
-      "grad_norm": 3.1990158557891846,
-      "learning_rate": 1.843141370896917e-05,
-      "loss": 0.9122,
-      "step": 5665
-    },
-    {
-      "epoch": 0.26286509040333794,
-      "grad_norm": 4.093268394470215,
-      "learning_rate": 1.842706023315594e-05,
-      "loss": 1.1009,
-      "step": 5670
-    },
-    {
-      "epoch": 0.26309689383402873,
-      "grad_norm": 4.303419589996338,
-      "learning_rate": 1.8422701239992916e-05,
-      "loss": 0.8344,
-      "step": 5675
-    },
-    {
-      "epoch": 0.2633286972647195,
-      "grad_norm": 3.7140135765075684,
-      "learning_rate": 1.841833673233399e-05,
-      "loss": 1.0971,
-      "step": 5680
-    },
-    {
-      "epoch": 0.2635605006954103,
-      "grad_norm": 4.93548059463501,
-      "learning_rate": 1.8413966713036703e-05,
-      "loss": 0.9318,
-      "step": 5685
-    },
-    {
-      "epoch": 0.26379230412610105,
-      "grad_norm": 3.852540969848633,
-      "learning_rate": 1.8409591184962176e-05,
-      "loss": 0.8139,
-      "step": 5690
-    },
-    {
-      "epoch": 0.26402410755679184,
-      "grad_norm": 4.363709449768066,
-      "learning_rate": 1.840521015097515e-05,
-      "loss": 0.9556,
-      "step": 5695
-    },
-    {
-      "epoch": 0.26425591098748263,
-      "grad_norm": 4.695869445800781,
-      "learning_rate": 1.840082361394397e-05,
-      "loss": 1.0201,
-      "step": 5700
-    },
-    {
-      "epoch": 0.26425591098748263,
-      "eval_loss": 1.0051652193069458,
-      "eval_runtime": 11.2696,
-      "eval_samples_per_second": 11.269,
-      "eval_steps_per_second": 11.269,
-      "step": 5700
-    },
-    {
-      "epoch": 0.26448771441817337,
-      "grad_norm": 3.8231186866760254,
-      "learning_rate": 1.8396431576740574e-05,
-      "loss": 1.0544,
-      "step": 5705
-    },
-    {
-      "epoch": 0.26471951784886416,
-      "grad_norm": 3.3638997077941895,
-      "learning_rate": 1.8392034042240516e-05,
-      "loss": 0.9932,
-      "step": 5710
-    },
-    {
-      "epoch": 0.26495132127955495,
-      "grad_norm": 3.88324236869812,
-      "learning_rate": 1.838763101332294e-05,
-      "loss": 0.8402,
-      "step": 5715
-    },
-    {
-      "epoch": 0.2651831247102457,
-      "grad_norm": 4.111605644226074,
-      "learning_rate": 1.8383222492870587e-05,
-      "loss": 0.9163,
-      "step": 5720
-    },
-    {
-      "epoch": 0.2654149281409365,
-      "grad_norm": 3.5883824825286865,
-      "learning_rate": 1.83788084837698e-05,
-      "loss": 1.0904,
-      "step": 5725
-    },
-    {
-      "epoch": 0.26564673157162727,
-      "grad_norm": 3.9498956203460693,
-      "learning_rate": 1.8374388988910507e-05,
-      "loss": 1.1557,
-      "step": 5730
-    },
-    {
-      "epoch": 0.26587853500231806,
-      "grad_norm": 3.6644833087921143,
-      "learning_rate": 1.8369964011186236e-05,
-      "loss": 0.9459,
-      "step": 5735
-    },
-    {
-      "epoch": 0.2661103384330088,
-      "grad_norm": 4.248614311218262,
-      "learning_rate": 1.8365533553494092e-05,
-      "loss": 1.0257,
-      "step": 5740
-    },
-    {
-      "epoch": 0.2663421418636996,
-      "grad_norm": 3.679971694946289,
-      "learning_rate": 1.8361097618734787e-05,
-      "loss": 1.1086,
-      "step": 5745
-    },
-    {
-      "epoch": 0.2665739452943904,
-      "grad_norm": 4.014171600341797,
-      "learning_rate": 1.8356656209812602e-05,
-      "loss": 0.8715,
-      "step": 5750
-    },
-    {
-      "epoch": 0.2668057487250811,
-      "grad_norm": 4.839527130126953,
-      "learning_rate": 1.8352209329635412e-05,
-      "loss": 1.1297,
-      "step": 5755
-    },
-    {
-      "epoch": 0.2670375521557719,
-      "grad_norm": 3.6321723461151123,
-      "learning_rate": 1.8347756981114665e-05,
-      "loss": 0.9076,
-      "step": 5760
-    },
-    {
-      "epoch": 0.2672693555864627,
-      "grad_norm": 3.7051403522491455,
-      "learning_rate": 1.83432991671654e-05,
-      "loss": 0.9269,
-      "step": 5765
-    },
-    {
-      "epoch": 0.26750115901715343,
-      "grad_norm": 4.2225341796875,
-      "learning_rate": 1.8338835890706227e-05,
-      "loss": 1.0015,
-      "step": 5770
-    },
-    {
-      "epoch": 0.2677329624478442,
-      "grad_norm": 4.238531112670898,
-      "learning_rate": 1.8334367154659334e-05,
-      "loss": 0.997,
-      "step": 5775
-    },
-    {
-      "epoch": 0.267964765878535,
-      "grad_norm": 4.749455451965332,
-      "learning_rate": 1.832989296195048e-05,
-      "loss": 0.9658,
-      "step": 5780
-    },
-    {
-      "epoch": 0.26819656930922575,
-      "grad_norm": 4.021186351776123,
-      "learning_rate": 1.832541331550901e-05,
-      "loss": 1.0967,
-      "step": 5785
-    },
-    {
-      "epoch": 0.26842837273991654,
-      "grad_norm": 4.332131385803223,
-      "learning_rate": 1.8320928218267827e-05,
-      "loss": 0.9285,
-      "step": 5790
-    },
-    {
-      "epoch": 0.26866017617060733,
-      "grad_norm": 3.860887050628662,
-      "learning_rate": 1.8316437673163396e-05,
-      "loss": 1.0157,
-      "step": 5795
-    },
-    {
-      "epoch": 0.2688919796012981,
-      "grad_norm": 3.988381862640381,
-      "learning_rate": 1.831194168313577e-05,
-      "loss": 0.9248,
-      "step": 5800
-    },
-    {
-      "epoch": 0.2688919796012981,
-      "eval_loss": 1.0011706352233887,
-      "eval_runtime": 11.2794,
-      "eval_samples_per_second": 11.259,
-      "eval_steps_per_second": 11.259,
-      "step": 5800
-    },
-    {
-      "epoch": 0.26912378303198886,
-      "grad_norm": 4.354172706604004,
-      "learning_rate": 1.830744025112855e-05,
-      "loss": 0.9361,
-      "step": 5805
-    },
-    {
-      "epoch": 0.26935558646267965,
-      "grad_norm": 3.436056137084961,
-      "learning_rate": 1.8302933380088912e-05,
-      "loss": 1.0889,
-      "step": 5810
-    },
-    {
-      "epoch": 0.26958738989337044,
-      "grad_norm": 4.711020469665527,
-      "learning_rate": 1.8298421072967577e-05,
-      "loss": 0.9565,
-      "step": 5815
-    },
-    {
-      "epoch": 0.2698191933240612,
-      "grad_norm": 4.042435169219971,
-      "learning_rate": 1.829390333271884e-05,
-      "loss": 1.0529,
-      "step": 5820
-    },
-    {
-      "epoch": 0.27005099675475197,
-      "grad_norm": 4.613515853881836,
-      "learning_rate": 1.8289380162300552e-05,
-      "loss": 0.9848,
-      "step": 5825
-    },
-    {
-      "epoch": 0.27028280018544276,
-      "grad_norm": 4.459861755371094,
-      "learning_rate": 1.8284851564674105e-05,
-      "loss": 0.859,
-      "step": 5830
-    },
-    {
-      "epoch": 0.2705146036161335,
-      "grad_norm": 4.675287246704102,
-      "learning_rate": 1.828031754280446e-05,
-      "loss": 1.1043,
-      "step": 5835
-    },
-    {
-      "epoch": 0.2707464070468243,
-      "grad_norm": 4.442194938659668,
-      "learning_rate": 1.8275778099660127e-05,
-      "loss": 0.9316,
-      "step": 5840
-    },
-    {
-      "epoch": 0.2709782104775151,
-      "grad_norm": 3.762488842010498,
-      "learning_rate": 1.8271233238213155e-05,
-      "loss": 1.006,
-      "step": 5845
-    },
-    {
-      "epoch": 0.2712100139082058,
-      "grad_norm": 4.232035160064697,
-      "learning_rate": 1.826668296143915e-05,
-      "loss": 1.0029,
-      "step": 5850
-    },
-    {
-      "epoch": 0.2714418173388966,
-      "grad_norm": 4.225643634796143,
-      "learning_rate": 1.8262127272317264e-05,
-      "loss": 1.1048,
-      "step": 5855
-    },
-    {
-      "epoch": 0.2716736207695874,
-      "grad_norm": 4.082033157348633,
-      "learning_rate": 1.8257566173830192e-05,
-      "loss": 0.8627,
-      "step": 5860
-    },
-    {
-      "epoch": 0.2719054242002782,
-      "grad_norm": 4.129724025726318,
-      "learning_rate": 1.825299966896416e-05,
-      "loss": 1.0083,
-      "step": 5865
-    },
-    {
-      "epoch": 0.2721372276309689,
-      "grad_norm": 3.868680953979492,
-      "learning_rate": 1.8248427760708943e-05,
-      "loss": 0.9134,
-      "step": 5870
-    },
-    {
-      "epoch": 0.2723690310616597,
-      "grad_norm": 4.29145622253418,
-      "learning_rate": 1.824385045205786e-05,
-      "loss": 1.0508,
-      "step": 5875
-    },
-    {
-      "epoch": 0.2726008344923505,
-      "grad_norm": 3.5361058712005615,
-      "learning_rate": 1.823926774600775e-05,
-      "loss": 0.9143,
-      "step": 5880
-    },
-    {
-      "epoch": 0.27283263792304124,
-      "grad_norm": 3.730649948120117,
-      "learning_rate": 1.8234679645559e-05,
-      "loss": 0.9608,
-      "step": 5885
-    },
-    {
-      "epoch": 0.27306444135373203,
-      "grad_norm": 4.100008487701416,
-      "learning_rate": 1.8230086153715523e-05,
-      "loss": 0.9398,
-      "step": 5890
-    },
-    {
-      "epoch": 0.2732962447844228,
-      "grad_norm": 4.048903465270996,
-      "learning_rate": 1.8225487273484758e-05,
-      "loss": 0.9586,
-      "step": 5895
-    },
-    {
-      "epoch": 0.27352804821511356,
-      "grad_norm": 3.6291110515594482,
-      "learning_rate": 1.822088300787767e-05,
-      "loss": 0.9457,
-      "step": 5900
-    },
-    {
-      "epoch": 0.27352804821511356,
-      "eval_loss": 1.0006332397460938,
-      "eval_runtime": 11.2615,
-      "eval_samples_per_second": 11.277,
-      "eval_steps_per_second": 11.277,
-      "step": 5900
-    },
-    {
-      "epoch": 0.27375985164580435,
-      "grad_norm": 4.81033182144165,
-      "learning_rate": 1.8216273359908775e-05,
-      "loss": 0.9129,
-      "step": 5905
-    },
-    {
-      "epoch": 0.27399165507649514,
-      "grad_norm": 3.627592086791992,
-      "learning_rate": 1.8211658332596075e-05,
-      "loss": 0.8996,
-      "step": 5910
-    },
-    {
-      "epoch": 0.2742234585071859,
-      "grad_norm": 4.121501445770264,
-      "learning_rate": 1.8207037928961116e-05,
-      "loss": 0.9794,
-      "step": 5915
-    },
-    {
-      "epoch": 0.27445526193787667,
-      "grad_norm": 4.973855018615723,
-      "learning_rate": 1.8202412152028966e-05,
-      "loss": 0.8467,
-      "step": 5920
-    },
-    {
-      "epoch": 0.27468706536856746,
-      "grad_norm": 4.146650314331055,
-      "learning_rate": 1.8197781004828205e-05,
-      "loss": 0.9895,
-      "step": 5925
-    },
-    {
-      "epoch": 0.27491886879925825,
-      "grad_norm": 3.674971580505371,
-      "learning_rate": 1.819314449039092e-05,
-      "loss": 0.9533,
-      "step": 5930
-    },
-    {
-      "epoch": 0.275150672229949,
-      "grad_norm": 4.400362014770508,
-      "learning_rate": 1.8188502611752733e-05,
-      "loss": 0.9267,
-      "step": 5935
-    },
-    {
-      "epoch": 0.2753824756606398,
-      "grad_norm": 4.569714069366455,
-      "learning_rate": 1.8183855371952764e-05,
-      "loss": 1.0855,
-      "step": 5940
-    },
-    {
-      "epoch": 0.27561427909133057,
-      "grad_norm": 4.336934566497803,
-      "learning_rate": 1.817920277403364e-05,
-      "loss": 1.104,
-      "step": 5945
-    },
-    {
-      "epoch": 0.2758460825220213,
-      "grad_norm": 4.304383277893066,
-      "learning_rate": 1.8174544821041508e-05,
-      "loss": 0.9737,
-      "step": 5950
-    },
-    {
-      "epoch": 0.2760778859527121,
-      "grad_norm": 4.5948991775512695,
-      "learning_rate": 1.8169881516026005e-05,
-      "loss": 1.0474,
-      "step": 5955
-    },
-    {
-      "epoch": 0.2763096893834029,
-      "grad_norm": 3.7741198539733887,
-      "learning_rate": 1.8165212862040294e-05,
-      "loss": 1.0487,
-      "step": 5960
-    },
-    {
-      "epoch": 0.2765414928140936,
-      "grad_norm": 5.039603233337402,
-      "learning_rate": 1.816053886214102e-05,
-      "loss": 1.0309,
-      "step": 5965
-    },
-    {
-      "epoch": 0.2767732962447844,
-      "grad_norm": 4.001100540161133,
-      "learning_rate": 1.8155859519388343e-05,
-      "loss": 0.9451,
-      "step": 5970
-    },
-    {
-      "epoch": 0.2770050996754752,
-      "grad_norm": 3.5335066318511963,
-      "learning_rate": 1.81511748368459e-05,
-      "loss": 0.9249,
-      "step": 5975
-    },
-    {
-      "epoch": 0.27723690310616594,
-      "grad_norm": 3.4955151081085205,
-      "learning_rate": 1.8146484817580852e-05,
-      "loss": 0.9852,
-      "step": 5980
-    },
-    {
-      "epoch": 0.27746870653685674,
-      "grad_norm": 6.319854736328125,
-      "learning_rate": 1.8141789464663833e-05,
-      "loss": 1.0451,
-      "step": 5985
-    },
-    {
-      "epoch": 0.2777005099675475,
-      "grad_norm": 4.266619682312012,
-      "learning_rate": 1.813708878116898e-05,
-      "loss": 0.9854,
-      "step": 5990
-    },
-    {
-      "epoch": 0.2779323133982383,
-      "grad_norm": 3.541940927505493,
-      "learning_rate": 1.8132382770173913e-05,
-      "loss": 1.021,
-      "step": 5995
-    },
-    {
-      "epoch": 0.27816411682892905,
-      "grad_norm": 3.38006591796875,
-      "learning_rate": 1.8127671434759745e-05,
-      "loss": 1.0177,
-      "step": 6000
-    },
-    {
-      "epoch": 0.27816411682892905,
-      "eval_loss": 1.0024383068084717,
-      "eval_runtime": 11.261,
-      "eval_samples_per_second": 11.278,
-      "eval_steps_per_second": 11.278,
-      "step": 6000
-    },
-    {
-      "epoch": 0.27839592025961984,
-      "grad_norm": 4.122401714324951,
-      "learning_rate": 1.812295477801107e-05,
-      "loss": 0.8534,
-      "step": 6005
-    },
-    {
-      "epoch": 0.27862772369031064,
-      "grad_norm": 5.3609137535095215,
-      "learning_rate": 1.8118232803015972e-05,
-      "loss": 1.0622,
-      "step": 6010
-    },
-    {
-      "epoch": 0.27885952712100137,
-      "grad_norm": 4.3532185554504395,
-      "learning_rate": 1.8113505512866014e-05,
-      "loss": 1.0705,
-      "step": 6015
-    },
-    {
-      "epoch": 0.27909133055169216,
-      "grad_norm": 4.290230751037598,
-      "learning_rate": 1.8108772910656237e-05,
-      "loss": 0.8422,
-      "step": 6020
-    },
-    {
-      "epoch": 0.27932313398238295,
-      "grad_norm": 4.357093334197998,
-      "learning_rate": 1.8104034999485164e-05,
-      "loss": 0.9788,
-      "step": 6025
-    },
-    {
-      "epoch": 0.2795549374130737,
-      "grad_norm": 4.413539409637451,
-      "learning_rate": 1.809929178245479e-05,
-      "loss": 1.0253,
-      "step": 6030
-    },
-    {
-      "epoch": 0.2797867408437645,
-      "grad_norm": 3.7482516765594482,
-      "learning_rate": 1.809454326267059e-05,
-      "loss": 0.9227,
-      "step": 6035
-    },
-    {
-      "epoch": 0.28001854427445527,
-      "grad_norm": 3.96517014503479,
-      "learning_rate": 1.80897894432415e-05,
-      "loss": 1.0275,
-      "step": 6040
-    },
-    {
-      "epoch": 0.28025034770514606,
-      "grad_norm": 3.922147512435913,
-      "learning_rate": 1.8085030327279933e-05,
-      "loss": 0.945,
-      "step": 6045
-    },
-    {
-      "epoch": 0.2804821511358368,
-      "grad_norm": 4.065891742706299,
-      "learning_rate": 1.8080265917901773e-05,
-      "loss": 0.8729,
-      "step": 6050
-    },
-    {
-      "epoch": 0.2807139545665276,
-      "grad_norm": 5.4653191566467285,
-      "learning_rate": 1.8075496218226362e-05,
-      "loss": 1.1503,
-      "step": 6055
-    },
-    {
-      "epoch": 0.2809457579972184,
-      "grad_norm": 4.193495750427246,
-      "learning_rate": 1.807072123137651e-05,
-      "loss": 0.9621,
-      "step": 6060
-    },
-    {
-      "epoch": 0.2811775614279091,
-      "grad_norm": 7.294194221496582,
-      "learning_rate": 1.806594096047849e-05,
-      "loss": 0.9805,
-      "step": 6065
-    },
-    {
-      "epoch": 0.2814093648585999,
-      "grad_norm": 3.062164783477783,
-      "learning_rate": 1.8061155408662032e-05,
-      "loss": 0.9644,
-      "step": 6070
-    },
-    {
-      "epoch": 0.2816411682892907,
-      "grad_norm": 3.978872776031494,
-      "learning_rate": 1.8056364579060322e-05,
-      "loss": 0.9448,
-      "step": 6075
-    },
-    {
-      "epoch": 0.28187297171998144,
-      "grad_norm": 3.817619800567627,
-      "learning_rate": 1.8051568474810005e-05,
-      "loss": 1.0054,
-      "step": 6080
-    },
-    {
-      "epoch": 0.2821047751506722,
-      "grad_norm": 3.854076385498047,
-      "learning_rate": 1.8046767099051177e-05,
-      "loss": 0.982,
-      "step": 6085
-    },
-    {
-      "epoch": 0.282336578581363,
-      "grad_norm": 4.013674259185791,
-      "learning_rate": 1.8041960454927384e-05,
-      "loss": 0.7673,
-      "step": 6090
-    },
-    {
-      "epoch": 0.28256838201205375,
-      "grad_norm": 4.976582050323486,
-      "learning_rate": 1.803714854558563e-05,
-      "loss": 1.0388,
-      "step": 6095
-    },
-    {
-      "epoch": 0.28280018544274454,
-      "grad_norm": 4.081470489501953,
-      "learning_rate": 1.803233137417635e-05,
-      "loss": 0.9529,
-      "step": 6100
-    },
-    {
-      "epoch": 0.28280018544274454,
-      "eval_loss": 0.9999383687973022,
-      "eval_runtime": 11.265,
-      "eval_samples_per_second": 11.274,
-      "eval_steps_per_second": 11.274,
-      "step": 6100
-    },
-    {
-      "epoch": 0.28303198887343534,
-      "grad_norm": 10.968564987182617,
-      "learning_rate": 1.8027508943853443e-05,
-      "loss": 0.9481,
-      "step": 6105
-    },
-    {
-      "epoch": 0.2832637923041261,
-      "grad_norm": 3.7944722175598145,
-      "learning_rate": 1.8022681257774234e-05,
-      "loss": 1.2888,
-      "step": 6110
-    },
-    {
-      "epoch": 0.28349559573481686,
-      "grad_norm": 4.5122551918029785,
-      "learning_rate": 1.8017848319099504e-05,
-      "loss": 0.9062,
-      "step": 6115
-    },
-    {
-      "epoch": 0.28372739916550765,
-      "grad_norm": 3.6060822010040283,
-      "learning_rate": 1.801301013099346e-05,
-      "loss": 1.1642,
-      "step": 6120
-    },
-    {
-      "epoch": 0.28395920259619845,
-      "grad_norm": 4.519002914428711,
-      "learning_rate": 1.800816669662376e-05,
-      "loss": 1.1781,
-      "step": 6125
-    },
-    {
-      "epoch": 0.2841910060268892,
-      "grad_norm": 3.604435443878174,
-      "learning_rate": 1.800331801916148e-05,
-      "loss": 0.8987,
-      "step": 6130
-    },
-    {
-      "epoch": 0.28442280945758,
-      "grad_norm": 4.327789783477783,
-      "learning_rate": 1.7998464101781145e-05,
-      "loss": 0.9225,
-      "step": 6135
-    },
-    {
-      "epoch": 0.28465461288827076,
-      "grad_norm": 4.079405784606934,
-      "learning_rate": 1.7993604947660693e-05,
-      "loss": 0.9651,
-      "step": 6140
-    },
-    {
-      "epoch": 0.2848864163189615,
-      "grad_norm": 4.200125217437744,
-      "learning_rate": 1.798874055998151e-05,
-      "loss": 0.9448,
-      "step": 6145
-    },
-    {
-      "epoch": 0.2851182197496523,
-      "grad_norm": 4.063922882080078,
-      "learning_rate": 1.7983870941928397e-05,
-      "loss": 0.8086,
-      "step": 6150
-    },
-    {
-      "epoch": 0.2853500231803431,
-      "grad_norm": 4.742971897125244,
-      "learning_rate": 1.7978996096689578e-05,
-      "loss": 0.9615,
-      "step": 6155
-    },
-    {
-      "epoch": 0.2855818266110338,
-      "grad_norm": 3.8731369972229004,
-      "learning_rate": 1.7974116027456707e-05,
-      "loss": 1.1075,
-      "step": 6160
-    },
-    {
-      "epoch": 0.2858136300417246,
-      "grad_norm": 4.0479044914245605,
-      "learning_rate": 1.796923073742485e-05,
-      "loss": 1.0039,
-      "step": 6165
-    },
-    {
-      "epoch": 0.2860454334724154,
-      "grad_norm": 4.324695587158203,
-      "learning_rate": 1.7964340229792506e-05,
-      "loss": 1.0035,
-      "step": 6170
-    },
-    {
-      "epoch": 0.2862772369031062,
-      "grad_norm": 4.0023627281188965,
-      "learning_rate": 1.7959444507761564e-05,
-      "loss": 1.0187,
-      "step": 6175
-    },
-    {
-      "epoch": 0.2865090403337969,
-      "grad_norm": 3.887049913406372,
-      "learning_rate": 1.7954543574537356e-05,
-      "loss": 0.9609,
-      "step": 6180
-    },
-    {
-      "epoch": 0.2867408437644877,
-      "grad_norm": 4.166080474853516,
-      "learning_rate": 1.79496374333286e-05,
-      "loss": 0.9904,
-      "step": 6185
-    },
-    {
-      "epoch": 0.2869726471951785,
-      "grad_norm": 4.174224853515625,
-      "learning_rate": 1.7944726087347446e-05,
-      "loss": 0.966,
-      "step": 6190
-    },
-    {
-      "epoch": 0.28720445062586925,
-      "grad_norm": 5.26371955871582,
-      "learning_rate": 1.7939809539809437e-05,
-      "loss": 0.9457,
-      "step": 6195
-    },
-    {
-      "epoch": 0.28743625405656004,
-      "grad_norm": 3.897738218307495,
-      "learning_rate": 1.7934887793933527e-05,
-      "loss": 0.8693,
-      "step": 6200
-    },
-    {
-      "epoch": 0.28743625405656004,
-      "eval_loss": 0.9928064942359924,
-      "eval_runtime": 11.2728,
-      "eval_samples_per_second": 11.266,
-      "eval_steps_per_second": 11.266,
-      "step": 6200
-    },
-    {
-      "epoch": 0.28766805748725083,
-      "grad_norm": 4.036500453948975,
-      "learning_rate": 1.7929960852942073e-05,
-      "loss": 0.8726,
-      "step": 6205
-    },
-    {
-      "epoch": 0.28789986091794156,
-      "grad_norm": 3.5420334339141846,
-      "learning_rate": 1.7925028720060833e-05,
-      "loss": 0.9567,
-      "step": 6210
-    },
-    {
-      "epoch": 0.28813166434863235,
-      "grad_norm": 3.9448046684265137,
-      "learning_rate": 1.7920091398518963e-05,
-      "loss": 0.8335,
-      "step": 6215
-    },
-    {
-      "epoch": 0.28836346777932315,
-      "grad_norm": 3.8117644786834717,
-      "learning_rate": 1.7915148891549016e-05,
-      "loss": 0.8017,
-      "step": 6220
-    },
-    {
-      "epoch": 0.2885952712100139,
-      "grad_norm": 3.843318462371826,
-      "learning_rate": 1.791020120238695e-05,
-      "loss": 0.931,
-      "step": 6225
-    },
-    {
-      "epoch": 0.2888270746407047,
-      "grad_norm": 4.147006511688232,
-      "learning_rate": 1.7905248334272094e-05,
-      "loss": 1.0015,
-      "step": 6230
-    },
-    {
-      "epoch": 0.28905887807139546,
-      "grad_norm": 3.9577856063842773,
-      "learning_rate": 1.790029029044719e-05,
-      "loss": 0.9748,
-      "step": 6235
-    },
-    {
-      "epoch": 0.28929068150208626,
-      "grad_norm": 4.157485008239746,
-      "learning_rate": 1.7895327074158356e-05,
-      "loss": 0.989,
-      "step": 6240
-    },
-    {
-      "epoch": 0.289522484932777,
-      "grad_norm": 4.6592020988464355,
-      "learning_rate": 1.7890358688655105e-05,
-      "loss": 0.8493,
-      "step": 6245
-    },
-    {
-      "epoch": 0.2897542883634678,
-      "grad_norm": 3.6401262283325195,
-      "learning_rate": 1.7885385137190324e-05,
-      "loss": 0.9636,
-      "step": 6250
-    },
-    {
-      "epoch": 0.2899860917941586,
-      "grad_norm": 3.0835177898406982,
-      "learning_rate": 1.788040642302029e-05,
-      "loss": 0.9383,
-      "step": 6255
-    },
-    {
-      "epoch": 0.2902178952248493,
-      "grad_norm": 3.7407875061035156,
-      "learning_rate": 1.7875422549404657e-05,
-      "loss": 0.9174,
-      "step": 6260
-    },
-    {
-      "epoch": 0.2904496986555401,
-      "grad_norm": 3.3547024726867676,
-      "learning_rate": 1.7870433519606457e-05,
-      "loss": 0.8579,
-      "step": 6265
-    },
-    {
-      "epoch": 0.2906815020862309,
-      "grad_norm": 5.221685886383057,
-      "learning_rate": 1.78654393368921e-05,
-      "loss": 1.237,
-      "step": 6270
-    },
-    {
-      "epoch": 0.2909133055169216,
-      "grad_norm": 3.961782693862915,
-      "learning_rate": 1.786044000453137e-05,
-      "loss": 0.7314,
-      "step": 6275
-    },
-    {
-      "epoch": 0.2911451089476124,
-      "grad_norm": 3.9514427185058594,
-      "learning_rate": 1.785543552579742e-05,
-      "loss": 0.9881,
-      "step": 6280
-    },
-    {
-      "epoch": 0.2913769123783032,
-      "grad_norm": 3.950453281402588,
-      "learning_rate": 1.785042590396677e-05,
-      "loss": 0.9165,
-      "step": 6285
-    },
-    {
-      "epoch": 0.29160871580899395,
-      "grad_norm": 4.681061267852783,
-      "learning_rate": 1.7845411142319316e-05,
-      "loss": 0.9462,
-      "step": 6290
-    },
-    {
-      "epoch": 0.29184051923968474,
-      "grad_norm": 4.256533145904541,
-      "learning_rate": 1.784039124413831e-05,
-      "loss": 1.0572,
-      "step": 6295
-    },
-    {
-      "epoch": 0.29207232267037553,
-      "grad_norm": 4.432834148406982,
-      "learning_rate": 1.7835366212710372e-05,
-      "loss": 0.9016,
-      "step": 6300
-    },
-    {
-      "epoch": 0.29207232267037553,
-      "eval_loss": 0.9896853566169739,
-      "eval_runtime": 11.2569,
-      "eval_samples_per_second": 11.282,
-      "eval_steps_per_second": 11.282,
-      "step": 6300
-    },
-    {
-      "epoch": 0.2923041261010663,
-      "grad_norm": 3.48752760887146,
-      "learning_rate": 1.783033605132548e-05,
-      "loss": 0.7874,
-      "step": 6305
-    },
-    {
-      "epoch": 0.29253592953175706,
-      "grad_norm": 3.671842336654663,
-      "learning_rate": 1.7825300763276974e-05,
-      "loss": 0.7982,
-      "step": 6310
-    },
-    {
-      "epoch": 0.29276773296244785,
-      "grad_norm": 4.298096656799316,
-      "learning_rate": 1.7820260351861545e-05,
-      "loss": 1.0207,
-      "step": 6315
-    },
-    {
-      "epoch": 0.29299953639313864,
-      "grad_norm": 4.248196125030518,
-      "learning_rate": 1.7815214820379253e-05,
-      "loss": 0.9549,
-      "step": 6320
-    },
-    {
-      "epoch": 0.2932313398238294,
-      "grad_norm": 3.7347469329833984,
-      "learning_rate": 1.781016417213349e-05,
-      "loss": 0.9005,
-      "step": 6325
-    },
-    {
-      "epoch": 0.29346314325452016,
-      "grad_norm": 3.76611328125,
-      "learning_rate": 1.7805108410431006e-05,
-      "loss": 0.9894,
-      "step": 6330
-    },
-    {
-      "epoch": 0.29369494668521096,
-      "grad_norm": 3.8931164741516113,
-      "learning_rate": 1.7800047538581902e-05,
-      "loss": 0.9377,
-      "step": 6335
-    },
-    {
-      "epoch": 0.2939267501159017,
-      "grad_norm": 3.719175100326538,
-      "learning_rate": 1.779498155989963e-05,
-      "loss": 0.856,
-      "step": 6340
-    },
-    {
-      "epoch": 0.2941585535465925,
-      "grad_norm": 4.287708282470703,
-      "learning_rate": 1.778991047770097e-05,
-      "loss": 1.0496,
-      "step": 6345
-    },
-    {
-      "epoch": 0.2943903569772833,
-      "grad_norm": 3.886796474456787,
-      "learning_rate": 1.7784834295306056e-05,
-      "loss": 0.956,
-      "step": 6350
-    },
-    {
-      "epoch": 0.29462216040797407,
-      "grad_norm": 3.6003739833831787,
-      "learning_rate": 1.7779753016038356e-05,
-      "loss": 0.9431,
-      "step": 6355
-    },
-    {
-      "epoch": 0.2948539638386648,
-      "grad_norm": 4.9724297523498535,
-      "learning_rate": 1.7774666643224675e-05,
-      "loss": 1.1131,
-      "step": 6360
-    },
-    {
-      "epoch": 0.2950857672693556,
-      "grad_norm": 4.584454536437988,
-      "learning_rate": 1.776957518019515e-05,
-      "loss": 0.9687,
-      "step": 6365
-    },
-    {
-      "epoch": 0.2953175707000464,
-      "grad_norm": 4.142701625823975,
-      "learning_rate": 1.776447863028327e-05,
-      "loss": 1.0858,
-      "step": 6370
-    },
-    {
-      "epoch": 0.2955493741307371,
-      "grad_norm": 5.801587104797363,
-      "learning_rate": 1.775937699682582e-05,
-      "loss": 1.0839,
-      "step": 6375
-    },
-    {
-      "epoch": 0.2957811775614279,
-      "grad_norm": 3.5847275257110596,
-      "learning_rate": 1.7754270283162943e-05,
-      "loss": 0.9185,
-      "step": 6380
-    },
-    {
-      "epoch": 0.2960129809921187,
-      "grad_norm": 3.590195417404175,
-      "learning_rate": 1.77491584926381e-05,
-      "loss": 0.9499,
-      "step": 6385
-    },
-    {
-      "epoch": 0.29624478442280944,
-      "grad_norm": 3.7471694946289062,
-      "learning_rate": 1.774404162859806e-05,
-      "loss": 1.0203,
-      "step": 6390
-    },
-    {
-      "epoch": 0.29647658785350023,
-      "grad_norm": 7.472535610198975,
-      "learning_rate": 1.7738919694392944e-05,
-      "loss": 1.1356,
-      "step": 6395
-    },
-    {
-      "epoch": 0.296708391284191,
-      "grad_norm": 3.6303653717041016,
-      "learning_rate": 1.7733792693376167e-05,
-      "loss": 0.8752,
-      "step": 6400
-    },
-    {
-      "epoch": 0.296708391284191,
-      "eval_loss": 0.989250898361206,
-      "eval_runtime": 11.2677,
-      "eval_samples_per_second": 11.271,
-      "eval_steps_per_second": 11.271,
-      "step": 6400
-    },
-    {
-      "epoch": 0.29694019471488176,
-      "grad_norm": 4.146617889404297,
-      "learning_rate": 1.7728660628904463e-05,
-      "loss": 0.9494,
-      "step": 6405
-    },
-    {
-      "epoch": 0.29717199814557255,
-      "grad_norm": 4.808352947235107,
-      "learning_rate": 1.77235235043379e-05,
-      "loss": 0.969,
-      "step": 6410
-    },
-    {
-      "epoch": 0.29740380157626334,
-      "grad_norm": 4.137024402618408,
-      "learning_rate": 1.7718381323039844e-05,
-      "loss": 1.027,
-      "step": 6415
-    },
-    {
-      "epoch": 0.29763560500695413,
-      "grad_norm": 4.234049320220947,
-      "learning_rate": 1.7713234088376973e-05,
-      "loss": 1.0522,
-      "step": 6420
-    },
-    {
-      "epoch": 0.29786740843764486,
-      "grad_norm": 3.660369873046875,
-      "learning_rate": 1.770808180371927e-05,
-      "loss": 0.848,
-      "step": 6425
-    },
-    {
-      "epoch": 0.29809921186833566,
-      "grad_norm": 3.89237904548645,
-      "learning_rate": 1.7702924472440038e-05,
-      "loss": 1.0321,
-      "step": 6430
-    },
-    {
-      "epoch": 0.29833101529902645,
-      "grad_norm": 4.689309597015381,
-      "learning_rate": 1.7697762097915875e-05,
-      "loss": 0.9966,
-      "step": 6435
-    },
-    {
-      "epoch": 0.2985628187297172,
-      "grad_norm": 5.034631729125977,
-      "learning_rate": 1.769259468352668e-05,
-      "loss": 1.1245,
-      "step": 6440
-    },
-    {
-      "epoch": 0.298794622160408,
-      "grad_norm": 3.535649299621582,
-      "learning_rate": 1.768742223265565e-05,
-      "loss": 0.8608,
-      "step": 6445
-    },
-    {
-      "epoch": 0.29902642559109877,
-      "grad_norm": 3.9463391304016113,
-      "learning_rate": 1.7682244748689283e-05,
-      "loss": 0.8687,
-      "step": 6450
-    },
-    {
-      "epoch": 0.2992582290217895,
-      "grad_norm": 3.97773814201355,
-      "learning_rate": 1.767706223501738e-05,
-      "loss": 0.9634,
-      "step": 6455
-    },
-    {
-      "epoch": 0.2994900324524803,
-      "grad_norm": 3.951277017593384,
-      "learning_rate": 1.7671874695033015e-05,
-      "loss": 0.9828,
-      "step": 6460
-    },
-    {
-      "epoch": 0.2997218358831711,
-      "grad_norm": 3.4690914154052734,
-      "learning_rate": 1.7666682132132575e-05,
-      "loss": 0.9453,
-      "step": 6465
-    },
-    {
-      "epoch": 0.2999536393138618,
-      "grad_norm": 4.310832977294922,
-      "learning_rate": 1.766148454971572e-05,
-      "loss": 1.0056,
-      "step": 6470
-    },
-    {
-      "epoch": 0.3001854427445526,
-      "grad_norm": 5.2137861251831055,
-      "learning_rate": 1.7656281951185406e-05,
-      "loss": 1.1368,
-      "step": 6475
-    },
-    {
-      "epoch": 0.3004172461752434,
-      "grad_norm": 3.1809844970703125,
-      "learning_rate": 1.7651074339947864e-05,
-      "loss": 0.7993,
-      "step": 6480
-    },
-    {
-      "epoch": 0.3006490496059342,
-      "grad_norm": 4.750578880310059,
-      "learning_rate": 1.7645861719412617e-05,
-      "loss": 0.9543,
-      "step": 6485
-    },
-    {
-      "epoch": 0.30088085303662493,
-      "grad_norm": 3.670433282852173,
-      "learning_rate": 1.764064409299246e-05,
-      "loss": 0.9679,
-      "step": 6490
-    },
-    {
-      "epoch": 0.3011126564673157,
-      "grad_norm": 4.144705772399902,
-      "learning_rate": 1.7635421464103468e-05,
-      "loss": 1.0732,
-      "step": 6495
-    },
-    {
-      "epoch": 0.3013444598980065,
-      "grad_norm": 3.8694875240325928,
-      "learning_rate": 1.7630193836164996e-05,
-      "loss": 1.1033,
-      "step": 6500
-    },
-    {
-      "epoch": 0.3013444598980065,
-      "eval_loss": 0.9887648820877075,
-      "eval_runtime": 11.2829,
-      "eval_samples_per_second": 11.256,
-      "eval_steps_per_second": 11.256,
-      "step": 6500
-    },
-    {
-      "epoch": 0.30157626332869725,
-      "grad_norm": 3.5212910175323486,
-      "learning_rate": 1.7624961212599654e-05,
-      "loss": 0.871,
-      "step": 6505
-    },
-    {
-      "epoch": 0.30180806675938804,
-      "grad_norm": 3.8312153816223145,
-      "learning_rate": 1.761972359683335e-05,
-      "loss": 1.0764,
-      "step": 6510
-    },
-    {
-      "epoch": 0.30203987019007883,
-      "grad_norm": 4.861583709716797,
-      "learning_rate": 1.761448099229524e-05,
-      "loss": 0.9775,
-      "step": 6515
-    },
-    {
-      "epoch": 0.30227167362076957,
-      "grad_norm": 4.358705043792725,
-      "learning_rate": 1.7609233402417755e-05,
-      "loss": 0.9921,
-      "step": 6520
-    },
-    {
-      "epoch": 0.30250347705146036,
-      "grad_norm": 4.940891742706299,
-      "learning_rate": 1.7603980830636586e-05,
-      "loss": 0.9865,
-      "step": 6525
-    },
-    {
-      "epoch": 0.30273528048215115,
-      "grad_norm": 4.035562992095947,
-      "learning_rate": 1.759872328039069e-05,
-      "loss": 1.0127,
-      "step": 6530
-    },
-    {
-      "epoch": 0.3029670839128419,
-      "grad_norm": 3.8091330528259277,
-      "learning_rate": 1.7593460755122274e-05,
-      "loss": 1.0808,
-      "step": 6535
-    },
-    {
-      "epoch": 0.3031988873435327,
-      "grad_norm": 4.551588535308838,
-      "learning_rate": 1.7588193258276815e-05,
-      "loss": 0.9015,
-      "step": 6540
-    },
-    {
-      "epoch": 0.30343069077422347,
-      "grad_norm": 6.152681350708008,
-      "learning_rate": 1.758292079330304e-05,
-      "loss": 1.126,
-      "step": 6545
-    },
-    {
-      "epoch": 0.30366249420491426,
-      "grad_norm": 4.223917007446289,
-      "learning_rate": 1.7577643363652925e-05,
-      "loss": 0.9735,
-      "step": 6550
-    },
-    {
-      "epoch": 0.303894297635605,
-      "grad_norm": 4.873682975769043,
-      "learning_rate": 1.75723609727817e-05,
-      "loss": 0.9722,
-      "step": 6555
-    },
-    {
-      "epoch": 0.3041261010662958,
-      "grad_norm": 4.19580602645874,
-      "learning_rate": 1.756707362414784e-05,
-      "loss": 1.0676,
-      "step": 6560
-    },
-    {
-      "epoch": 0.3043579044969866,
-      "grad_norm": 4.814520835876465,
-      "learning_rate": 1.7561781321213078e-05,
-      "loss": 1.0338,
-      "step": 6565
-    },
-    {
-      "epoch": 0.3045897079276773,
-      "grad_norm": 3.395010471343994,
-      "learning_rate": 1.755648406744237e-05,
-      "loss": 0.973,
-      "step": 6570
-    },
-    {
-      "epoch": 0.3048215113583681,
-      "grad_norm": 3.860635757446289,
-      "learning_rate": 1.7551181866303926e-05,
-      "loss": 0.8789,
-      "step": 6575
-    },
-    {
-      "epoch": 0.3050533147890589,
-      "grad_norm": 4.463065147399902,
-      "learning_rate": 1.7545874721269198e-05,
-      "loss": 1.0307,
-      "step": 6580
-    },
-    {
-      "epoch": 0.30528511821974963,
-      "grad_norm": 5.383005619049072,
-      "learning_rate": 1.754056263581287e-05,
-      "loss": 0.9404,
-      "step": 6585
-    },
-    {
-      "epoch": 0.3055169216504404,
-      "grad_norm": 4.1449360847473145,
-      "learning_rate": 1.7535245613412862e-05,
-      "loss": 0.9562,
-      "step": 6590
-    },
-    {
-      "epoch": 0.3057487250811312,
-      "grad_norm": 5.013772487640381,
-      "learning_rate": 1.752992365755032e-05,
-      "loss": 1.0953,
-      "step": 6595
-    },
-    {
-      "epoch": 0.30598052851182195,
-      "grad_norm": 3.9730842113494873,
-      "learning_rate": 1.7524596771709628e-05,
-      "loss": 1.0541,
-      "step": 6600
-    },
-    {
-      "epoch": 0.30598052851182195,
-      "eval_loss": 0.9859225749969482,
-      "eval_runtime": 11.2765,
-      "eval_samples_per_second": 11.262,
-      "eval_steps_per_second": 11.262,
-      "step": 6600
-    },
-    {
-      "epoch": 0.30621233194251274,
-      "grad_norm": 3.646376132965088,
-      "learning_rate": 1.7519264959378398e-05,
-      "loss": 0.8244,
-      "step": 6605
-    },
-    {
-      "epoch": 0.30644413537320353,
-      "grad_norm": 3.8406198024749756,
-      "learning_rate": 1.7513928224047464e-05,
-      "loss": 0.9946,
-      "step": 6610
-    },
-    {
-      "epoch": 0.3066759388038943,
-      "grad_norm": 3.9793317317962646,
-      "learning_rate": 1.7508586569210883e-05,
-      "loss": 0.8445,
-      "step": 6615
-    },
-    {
-      "epoch": 0.30690774223458506,
-      "grad_norm": 4.22384786605835,
-      "learning_rate": 1.750323999836593e-05,
-      "loss": 0.9408,
-      "step": 6620
-    },
-    {
-      "epoch": 0.30713954566527585,
-      "grad_norm": 3.759800910949707,
-      "learning_rate": 1.7497888515013106e-05,
-      "loss": 0.8005,
-      "step": 6625
-    },
-    {
-      "epoch": 0.30737134909596664,
-      "grad_norm": 3.72318959236145,
-      "learning_rate": 1.7492532122656124e-05,
-      "loss": 0.9359,
-      "step": 6630
-    },
-    {
-      "epoch": 0.3076031525266574,
-      "grad_norm": 3.6016530990600586,
-      "learning_rate": 1.7487170824801912e-05,
-      "loss": 1.0068,
-      "step": 6635
-    },
-    {
-      "epoch": 0.30783495595734817,
-      "grad_norm": 4.295738220214844,
-      "learning_rate": 1.7481804624960607e-05,
-      "loss": 1.0878,
-      "step": 6640
-    },
-    {
-      "epoch": 0.30806675938803896,
-      "grad_norm": 3.8684656620025635,
-      "learning_rate": 1.7476433526645562e-05,
-      "loss": 1.0067,
-      "step": 6645
-    },
-    {
-      "epoch": 0.3082985628187297,
-      "grad_norm": 3.390023708343506,
-      "learning_rate": 1.7471057533373332e-05,
-      "loss": 0.903,
-      "step": 6650
-    },
-    {
-      "epoch": 0.3085303662494205,
-      "grad_norm": 4.495607376098633,
-      "learning_rate": 1.746567664866367e-05,
-      "loss": 0.7868,
-      "step": 6655
-    },
-    {
-      "epoch": 0.3087621696801113,
-      "grad_norm": 4.227192401885986,
-      "learning_rate": 1.7460290876039552e-05,
-      "loss": 1.1199,
-      "step": 6660
-    },
-    {
-      "epoch": 0.30899397311080207,
-      "grad_norm": 3.549609422683716,
-      "learning_rate": 1.7454900219027133e-05,
-      "loss": 0.9409,
-      "step": 6665
-    },
-    {
-      "epoch": 0.3092257765414928,
-      "grad_norm": 5.658492088317871,
-      "learning_rate": 1.744950468115578e-05,
-      "loss": 1.086,
-      "step": 6670
-    },
-    {
-      "epoch": 0.3094575799721836,
-      "grad_norm": 3.934952974319458,
-      "learning_rate": 1.744410426595804e-05,
-      "loss": 0.998,
-      "step": 6675
-    },
-    {
-      "epoch": 0.3096893834028744,
-      "grad_norm": 3.8127615451812744,
-      "learning_rate": 1.743869897696967e-05,
-      "loss": 0.9748,
-      "step": 6680
-    },
-    {
-      "epoch": 0.3099211868335651,
-      "grad_norm": 3.870283603668213,
-      "learning_rate": 1.7433288817729614e-05,
-      "loss": 1.1462,
-      "step": 6685
-    },
-    {
-      "epoch": 0.3101529902642559,
-      "grad_norm": 4.023786544799805,
-      "learning_rate": 1.7427873791779994e-05,
-      "loss": 0.9699,
-      "step": 6690
-    },
-    {
-      "epoch": 0.3103847936949467,
-      "grad_norm": 3.8653597831726074,
-      "learning_rate": 1.7422453902666135e-05,
-      "loss": 0.8618,
-      "step": 6695
-    },
-    {
-      "epoch": 0.31061659712563744,
-      "grad_norm": 4.728411674499512,
-      "learning_rate": 1.7417029153936528e-05,
-      "loss": 1.0816,
-      "step": 6700
-    },
-    {
-      "epoch": 0.31061659712563744,
-      "eval_loss": 0.9867631196975708,
-      "eval_runtime": 11.2651,
-      "eval_samples_per_second": 11.274,
-      "eval_steps_per_second": 11.274,
-      "step": 6700
-    },
-    {
-      "epoch": 0.31084840055632823,
-      "grad_norm": 4.564522743225098,
-      "learning_rate": 1.7411599549142863e-05,
-      "loss": 1.1131,
-      "step": 6705
-    },
-    {
-      "epoch": 0.311080203987019,
-      "grad_norm": 4.069855213165283,
-      "learning_rate": 1.740616509183999e-05,
-      "loss": 0.9792,
-      "step": 6710
-    },
-    {
-      "epoch": 0.31131200741770976,
-      "grad_norm": 3.651477575302124,
-      "learning_rate": 1.7400725785585965e-05,
-      "loss": 0.9195,
-      "step": 6715
-    },
-    {
-      "epoch": 0.31154381084840055,
-      "grad_norm": 4.551967144012451,
-      "learning_rate": 1.7395281633941986e-05,
-      "loss": 0.9555,
-      "step": 6720
-    },
-    {
-      "epoch": 0.31177561427909134,
-      "grad_norm": 3.895969867706299,
-      "learning_rate": 1.738983264047244e-05,
-      "loss": 1.0506,
-      "step": 6725
-    },
-    {
-      "epoch": 0.31200741770978213,
-      "grad_norm": 3.545581817626953,
-      "learning_rate": 1.738437880874489e-05,
-      "loss": 0.8803,
-      "step": 6730
-    },
-    {
-      "epoch": 0.31223922114047287,
-      "grad_norm": 3.2424161434173584,
-      "learning_rate": 1.737892014233005e-05,
-      "loss": 1.0651,
-      "step": 6735
-    },
-    {
-      "epoch": 0.31247102457116366,
-      "grad_norm": 4.444201469421387,
-      "learning_rate": 1.737345664480181e-05,
-      "loss": 0.9039,
-      "step": 6740
-    },
-    {
-      "epoch": 0.31270282800185445,
-      "grad_norm": 4.671426773071289,
-      "learning_rate": 1.736798831973723e-05,
-      "loss": 1.0772,
-      "step": 6745
-    },
-    {
-      "epoch": 0.3129346314325452,
-      "grad_norm": 3.8473031520843506,
-      "learning_rate": 1.7362515170716512e-05,
-      "loss": 1.0927,
-      "step": 6750
-    },
-    {
-      "epoch": 0.313166434863236,
-      "grad_norm": 4.373749732971191,
-      "learning_rate": 1.735703720132303e-05,
-      "loss": 0.992,
-      "step": 6755
-    },
-    {
-      "epoch": 0.31339823829392677,
-      "grad_norm": 4.11835241317749,
-      "learning_rate": 1.735155441514331e-05,
-      "loss": 0.8926,
-      "step": 6760
-    },
-    {
-      "epoch": 0.3136300417246175,
-      "grad_norm": 3.815171957015991,
-      "learning_rate": 1.7346066815767033e-05,
-      "loss": 0.9104,
-      "step": 6765
-    },
-    {
-      "epoch": 0.3138618451553083,
-      "grad_norm": 3.835723638534546,
-      "learning_rate": 1.734057440678703e-05,
-      "loss": 0.9115,
-      "step": 6770
-    },
-    {
-      "epoch": 0.3140936485859991,
-      "grad_norm": 3.634661912918091,
-      "learning_rate": 1.7335077191799277e-05,
-      "loss": 0.7149,
-      "step": 6775
-    },
-    {
-      "epoch": 0.3143254520166898,
-      "grad_norm": 4.546589374542236,
-      "learning_rate": 1.7329575174402907e-05,
-      "loss": 1.0683,
-      "step": 6780
-    },
-    {
-      "epoch": 0.3145572554473806,
-      "grad_norm": 3.3564836978912354,
-      "learning_rate": 1.732406835820019e-05,
-      "loss": 0.8881,
-      "step": 6785
-    },
-    {
-      "epoch": 0.3147890588780714,
-      "grad_norm": 3.6788740158081055,
-      "learning_rate": 1.7318556746796532e-05,
-      "loss": 1.0333,
-      "step": 6790
-    },
-    {
-      "epoch": 0.3150208623087622,
-      "grad_norm": 3.0859086513519287,
-      "learning_rate": 1.7313040343800493e-05,
-      "loss": 0.8799,
-      "step": 6795
-    },
-    {
-      "epoch": 0.31525266573945293,
-      "grad_norm": 4.490921497344971,
-      "learning_rate": 1.7307519152823763e-05,
-      "loss": 0.8285,
-      "step": 6800
-    },
-    {
-      "epoch": 0.31525266573945293,
-      "eval_loss": 0.9875900745391846,
-      "eval_runtime": 11.2755,
-      "eval_samples_per_second": 11.263,
-      "eval_steps_per_second": 11.263,
-      "step": 6800
-    },
-    {
-      "epoch": 0.3154844691701437,
-      "grad_norm": 4.825845241546631,
-      "learning_rate": 1.730199317748116e-05,
-      "loss": 1.0106,
-      "step": 6805
-    },
-    {
-      "epoch": 0.3157162726008345,
-      "grad_norm": 4.728207588195801,
-      "learning_rate": 1.7296462421390647e-05,
-      "loss": 0.9964,
-      "step": 6810
-    },
-    {
-      "epoch": 0.31594807603152525,
-      "grad_norm": 4.3029279708862305,
-      "learning_rate": 1.7290926888173306e-05,
-      "loss": 0.9904,
-      "step": 6815
-    },
-    {
-      "epoch": 0.31617987946221604,
-      "grad_norm": 3.814422130584717,
-      "learning_rate": 1.7285386581453353e-05,
-      "loss": 0.9931,
-      "step": 6820
-    },
-    {
-      "epoch": 0.31641168289290683,
-      "grad_norm": 4.1385393142700195,
-      "learning_rate": 1.7279841504858127e-05,
-      "loss": 1.0007,
-      "step": 6825
-    },
-    {
-      "epoch": 0.31664348632359757,
-      "grad_norm": 3.7476789951324463,
-      "learning_rate": 1.727429166201809e-05,
-      "loss": 0.9299,
-      "step": 6830
-    },
-    {
-      "epoch": 0.31687528975428836,
-      "grad_norm": 3.564824104309082,
-      "learning_rate": 1.726873705656683e-05,
-      "loss": 0.9108,
-      "step": 6835
-    },
-    {
-      "epoch": 0.31710709318497915,
-      "grad_norm": 4.3677215576171875,
-      "learning_rate": 1.7263177692141045e-05,
-      "loss": 0.9922,
-      "step": 6840
-    },
-    {
-      "epoch": 0.3173388966156699,
-      "grad_norm": 3.237433433532715,
-      "learning_rate": 1.725761357238055e-05,
-      "loss": 0.7999,
-      "step": 6845
-    },
-    {
-      "epoch": 0.3175707000463607,
-      "grad_norm": 5.139328479766846,
-      "learning_rate": 1.7252044700928275e-05,
-      "loss": 0.9433,
-      "step": 6850
-    },
-    {
-      "epoch": 0.31780250347705147,
-      "grad_norm": 3.4519400596618652,
-      "learning_rate": 1.724647108143027e-05,
-      "loss": 0.8964,
-      "step": 6855
-    },
-    {
-      "epoch": 0.31803430690774226,
-      "grad_norm": 3.80947208404541,
-      "learning_rate": 1.724089271753568e-05,
-      "loss": 1.0493,
-      "step": 6860
-    },
-    {
-      "epoch": 0.318266110338433,
-      "grad_norm": 4.166898250579834,
-      "learning_rate": 1.7235309612896756e-05,
-      "loss": 1.0334,
-      "step": 6865
-    },
-    {
-      "epoch": 0.3184979137691238,
-      "grad_norm": 3.7955734729766846,
-      "learning_rate": 1.7229721771168864e-05,
-      "loss": 1.0212,
-      "step": 6870
-    },
-    {
-      "epoch": 0.3187297171998146,
-      "grad_norm": 3.6304619312286377,
-      "learning_rate": 1.7224129196010467e-05,
-      "loss": 0.948,
-      "step": 6875
-    },
-    {
-      "epoch": 0.3189615206305053,
-      "grad_norm": 3.9213900566101074,
-      "learning_rate": 1.7218531891083122e-05,
-      "loss": 0.97,
-      "step": 6880
-    },
-    {
-      "epoch": 0.3191933240611961,
-      "grad_norm": 5.918650150299072,
-      "learning_rate": 1.721292986005149e-05,
-      "loss": 1.0927,
-      "step": 6885
-    },
-    {
-      "epoch": 0.3194251274918869,
-      "grad_norm": 3.4678218364715576,
-      "learning_rate": 1.720732310658332e-05,
-      "loss": 0.8423,
-      "step": 6890
-    },
-    {
-      "epoch": 0.31965693092257763,
-      "grad_norm": 4.282207489013672,
-      "learning_rate": 1.7201711634349457e-05,
-      "loss": 1.029,
-      "step": 6895
-    },
-    {
-      "epoch": 0.3198887343532684,
-      "grad_norm": 4.10146951675415,
-      "learning_rate": 1.7196095447023833e-05,
-      "loss": 0.8673,
-      "step": 6900
-    },
-    {
-      "epoch": 0.3198887343532684,
-      "eval_loss": 0.9845543503761292,
-      "eval_runtime": 11.2805,
-      "eval_samples_per_second": 11.258,
-      "eval_steps_per_second": 11.258,
-      "step": 6900
-    },
-    {
-      "epoch": 0.3201205377839592,
-      "grad_norm": 4.028499603271484,
-      "learning_rate": 1.7190474548283468e-05,
-      "loss": 0.8934,
-      "step": 6905
-    },
-    {
-      "epoch": 0.32035234121464995,
-      "grad_norm": 3.83030104637146,
-      "learning_rate": 1.718484894180847e-05,
-      "loss": 1.0745,
-      "step": 6910
-    },
-    {
-      "epoch": 0.32058414464534074,
-      "grad_norm": 3.5020344257354736,
-      "learning_rate": 1.717921863128202e-05,
-      "loss": 0.8391,
-      "step": 6915
-    },
-    {
-      "epoch": 0.32081594807603153,
-      "grad_norm": 3.863657236099243,
-      "learning_rate": 1.717358362039039e-05,
-      "loss": 0.9929,
-      "step": 6920
-    },
-    {
-      "epoch": 0.3210477515067223,
-      "grad_norm": 3.93176007270813,
-      "learning_rate": 1.716794391282292e-05,
-      "loss": 1.0097,
-      "step": 6925
-    },
-    {
-      "epoch": 0.32127955493741306,
-      "grad_norm": 4.715558052062988,
-      "learning_rate": 1.7162299512272036e-05,
-      "loss": 0.9322,
-      "step": 6930
-    },
-    {
-      "epoch": 0.32151135836810385,
-      "grad_norm": 3.9522275924682617,
-      "learning_rate": 1.7156650422433225e-05,
-      "loss": 0.8993,
-      "step": 6935
-    },
-    {
-      "epoch": 0.32174316179879464,
-      "grad_norm": 5.151059627532959,
-      "learning_rate": 1.715099664700505e-05,
-      "loss": 0.8975,
-      "step": 6940
-    },
-    {
-      "epoch": 0.3219749652294854,
-      "grad_norm": 3.8080761432647705,
-      "learning_rate": 1.7145338189689136e-05,
-      "loss": 0.8875,
-      "step": 6945
-    },
-    {
-      "epoch": 0.32220676866017617,
-      "grad_norm": 3.906095504760742,
-      "learning_rate": 1.7139675054190186e-05,
-      "loss": 0.944,
-      "step": 6950
-    },
-    {
-      "epoch": 0.32243857209086696,
-      "grad_norm": 4.30042028427124,
-      "learning_rate": 1.7134007244215955e-05,
-      "loss": 0.9607,
-      "step": 6955
-    },
-    {
-      "epoch": 0.3226703755215577,
-      "grad_norm": 3.7272627353668213,
-      "learning_rate": 1.7128334763477256e-05,
-      "loss": 0.9211,
-      "step": 6960
-    },
-    {
-      "epoch": 0.3229021789522485,
-      "grad_norm": 4.208070755004883,
-      "learning_rate": 1.7122657615687975e-05,
-      "loss": 0.9681,
-      "step": 6965
-    },
-    {
-      "epoch": 0.3231339823829393,
-      "grad_norm": 3.3243024349212646,
-      "learning_rate": 1.711697580456504e-05,
-      "loss": 0.9908,
-      "step": 6970
-    },
-    {
-      "epoch": 0.32336578581363007,
-      "grad_norm": 3.6350746154785156,
-      "learning_rate": 1.711128933382843e-05,
-      "loss": 1.0272,
-      "step": 6975
-    },
-    {
-      "epoch": 0.3235975892443208,
-      "grad_norm": 5.2530198097229,
-      "learning_rate": 1.710559820720119e-05,
-      "loss": 1.0747,
-      "step": 6980
-    },
-    {
-      "epoch": 0.3238293926750116,
-      "grad_norm": 3.2672336101531982,
-      "learning_rate": 1.7099902428409404e-05,
-      "loss": 0.8968,
-      "step": 6985
-    },
-    {
-      "epoch": 0.3240611961057024,
-      "grad_norm": 3.2137134075164795,
-      "learning_rate": 1.7094202001182192e-05,
-      "loss": 1.0142,
-      "step": 6990
-    },
-    {
-      "epoch": 0.3242929995363931,
-      "grad_norm": 4.092966079711914,
-      "learning_rate": 1.708849692925174e-05,
-      "loss": 0.9918,
-      "step": 6995
-    },
-    {
-      "epoch": 0.3245248029670839,
-      "grad_norm": 5.9110331535339355,
-      "learning_rate": 1.7082787216353255e-05,
-      "loss": 1.0513,
-      "step": 7000
-    },
-    {
-      "epoch": 0.3245248029670839,
-      "eval_loss": 0.9831182956695557,
-      "eval_runtime": 11.2664,
-      "eval_samples_per_second": 11.272,
-      "eval_steps_per_second": 11.272,
-      "step": 7000
-    },
-    {
-      "epoch": 0.3247566063977747,
-      "grad_norm": 4.733374118804932,
-      "learning_rate": 1.707707286622499e-05,
-      "loss": 1.0291,
-      "step": 7005
-    },
-    {
-      "epoch": 0.32498840982846544,
-      "grad_norm": 3.4680347442626953,
-      "learning_rate": 1.7071353882608238e-05,
-      "loss": 1.0134,
-      "step": 7010
-    },
-    {
-      "epoch": 0.32522021325915623,
-      "grad_norm": 3.5740184783935547,
-      "learning_rate": 1.706563026924732e-05,
-      "loss": 0.9281,
-      "step": 7015
-    },
-    {
-      "epoch": 0.325452016689847,
-      "grad_norm": 4.178410530090332,
-      "learning_rate": 1.7059902029889585e-05,
-      "loss": 0.9851,
-      "step": 7020
-    },
-    {
-      "epoch": 0.32568382012053776,
-      "grad_norm": 3.8983752727508545,
-      "learning_rate": 1.7054169168285425e-05,
-      "loss": 0.9198,
-      "step": 7025
-    },
-    {
-      "epoch": 0.32591562355122855,
-      "grad_norm": 3.6942498683929443,
-      "learning_rate": 1.7048431688188244e-05,
-      "loss": 0.9314,
-      "step": 7030
-    },
-    {
-      "epoch": 0.32614742698191934,
-      "grad_norm": 3.4959239959716797,
-      "learning_rate": 1.7042689593354472e-05,
-      "loss": 0.8442,
-      "step": 7035
-    },
-    {
-      "epoch": 0.32637923041261013,
-      "grad_norm": 3.6562867164611816,
-      "learning_rate": 1.7036942887543566e-05,
-      "loss": 1.0384,
-      "step": 7040
-    },
-    {
-      "epoch": 0.32661103384330087,
-      "grad_norm": 3.508683204650879,
-      "learning_rate": 1.7031191574518e-05,
-      "loss": 1.0261,
-      "step": 7045
-    },
-    {
-      "epoch": 0.32684283727399166,
-      "grad_norm": 4.6104278564453125,
-      "learning_rate": 1.702543565804326e-05,
-      "loss": 0.9351,
-      "step": 7050
-    },
-    {
-      "epoch": 0.32707464070468245,
-      "grad_norm": 3.4227206707000732,
-      "learning_rate": 1.7019675141887855e-05,
-      "loss": 0.8968,
-      "step": 7055
-    },
-    {
-      "epoch": 0.3273064441353732,
-      "grad_norm": 3.3952324390411377,
-      "learning_rate": 1.7013910029823292e-05,
-      "loss": 0.8983,
-      "step": 7060
-    },
-    {
-      "epoch": 0.327538247566064,
-      "grad_norm": 4.351296424865723,
-      "learning_rate": 1.7008140325624104e-05,
-      "loss": 0.9035,
-      "step": 7065
-    },
-    {
-      "epoch": 0.32777005099675477,
-      "grad_norm": 3.7643845081329346,
-      "learning_rate": 1.7002366033067813e-05,
-      "loss": 0.8031,
-      "step": 7070
-    },
-    {
-      "epoch": 0.3280018544274455,
-      "grad_norm": 4.087846279144287,
-      "learning_rate": 1.699658715593496e-05,
-      "loss": 1.1498,
-      "step": 7075
-    },
-    {
-      "epoch": 0.3282336578581363,
-      "grad_norm": 4.362499237060547,
-      "learning_rate": 1.699080369800908e-05,
-      "loss": 0.9128,
-      "step": 7080
-    },
-    {
-      "epoch": 0.3284654612888271,
-      "grad_norm": 3.442183494567871,
-      "learning_rate": 1.6985015663076702e-05,
-      "loss": 0.8206,
-      "step": 7085
-    },
-    {
-      "epoch": 0.3286972647195178,
-      "grad_norm": 3.769672155380249,
-      "learning_rate": 1.697922305492736e-05,
-      "loss": 1.019,
-      "step": 7090
-    },
-    {
-      "epoch": 0.3289290681502086,
-      "grad_norm": 4.2746381759643555,
-      "learning_rate": 1.6973425877353587e-05,
-      "loss": 0.8872,
-      "step": 7095
-    },
-    {
-      "epoch": 0.3291608715808994,
-      "grad_norm": 4.174293041229248,
-      "learning_rate": 1.6967624134150895e-05,
-      "loss": 0.9736,
-      "step": 7100
-    },
-    {
-      "epoch": 0.3291608715808994,
-      "eval_loss": 0.9836241602897644,
-      "eval_runtime": 11.2694,
-      "eval_samples_per_second": 11.269,
-      "eval_steps_per_second": 11.269,
-      "step": 7100
-    },
-    {
-      "epoch": 0.3293926750115902,
-      "grad_norm": 2.938143491744995,
-      "learning_rate": 1.696181782911779e-05,
-      "loss": 0.8206,
-      "step": 7105
-    },
-    {
-      "epoch": 0.32962447844228093,
-      "grad_norm": 3.7101032733917236,
-      "learning_rate": 1.6956006966055772e-05,
-      "loss": 0.9207,
-      "step": 7110
-    },
-    {
-      "epoch": 0.3298562818729717,
-      "grad_norm": 3.974308490753174,
-      "learning_rate": 1.6950191548769314e-05,
-      "loss": 0.9115,
-      "step": 7115
-    },
-    {
-      "epoch": 0.3300880853036625,
-      "grad_norm": 5.249458312988281,
-      "learning_rate": 1.6944371581065873e-05,
-      "loss": 1.0737,
-      "step": 7120
-    },
-    {
-      "epoch": 0.33031988873435325,
-      "grad_norm": 4.166738033294678,
-      "learning_rate": 1.693854706675589e-05,
-      "loss": 0.9348,
-      "step": 7125
-    },
-    {
-      "epoch": 0.33055169216504404,
-      "grad_norm": 4.688779830932617,
-      "learning_rate": 1.6932718009652785e-05,
-      "loss": 0.8953,
-      "step": 7130
-    },
-    {
-      "epoch": 0.33078349559573483,
-      "grad_norm": 3.903956890106201,
-      "learning_rate": 1.6926884413572942e-05,
-      "loss": 0.9307,
-      "step": 7135
-    },
-    {
-      "epoch": 0.33101529902642557,
-      "grad_norm": 4.132503509521484,
-      "learning_rate": 1.6921046282335724e-05,
-      "loss": 0.969,
-      "step": 7140
-    },
-    {
-      "epoch": 0.33124710245711636,
-      "grad_norm": 3.627603054046631,
-      "learning_rate": 1.691520361976347e-05,
-      "loss": 0.8196,
-      "step": 7145
-    },
-    {
-      "epoch": 0.33147890588780715,
-      "grad_norm": 3.7588562965393066,
-      "learning_rate": 1.690935642968146e-05,
-      "loss": 0.8648,
-      "step": 7150
-    },
-    {
-      "epoch": 0.3317107093184979,
-      "grad_norm": 4.168130874633789,
-      "learning_rate": 1.6903504715917972e-05,
-      "loss": 1.1597,
-      "step": 7155
-    },
-    {
-      "epoch": 0.3319425127491887,
-      "grad_norm": 3.6999406814575195,
-      "learning_rate": 1.6897648482304226e-05,
-      "loss": 0.7546,
-      "step": 7160
-    },
-    {
-      "epoch": 0.33217431617987947,
-      "grad_norm": 3.8351409435272217,
-      "learning_rate": 1.68917877326744e-05,
-      "loss": 0.9968,
-      "step": 7165
-    },
-    {
-      "epoch": 0.33240611961057026,
-      "grad_norm": 3.701868772506714,
-      "learning_rate": 1.6885922470865636e-05,
-      "loss": 0.9861,
-      "step": 7170
-    },
-    {
-      "epoch": 0.332637923041261,
-      "grad_norm": 3.982545852661133,
-      "learning_rate": 1.6880052700718028e-05,
-      "loss": 1.0263,
-      "step": 7175
-    },
-    {
-      "epoch": 0.3328697264719518,
-      "grad_norm": 4.025231838226318,
-      "learning_rate": 1.687417842607462e-05,
-      "loss": 0.8908,
-      "step": 7180
-    },
-    {
-      "epoch": 0.3331015299026426,
-      "grad_norm": 3.8443965911865234,
-      "learning_rate": 1.686829965078141e-05,
-      "loss": 0.8625,
-      "step": 7185
-    },
-    {
-      "epoch": 0.3333333333333333,
-      "grad_norm": 4.3348517417907715,
-      "learning_rate": 1.686241637868734e-05,
-      "loss": 1.0526,
-      "step": 7190
-    },
-    {
-      "epoch": 0.3335651367640241,
-      "grad_norm": 4.766390800476074,
-      "learning_rate": 1.6856528613644284e-05,
-      "loss": 0.9305,
-      "step": 7195
-    },
-    {
-      "epoch": 0.3337969401947149,
-      "grad_norm": 4.653823375701904,
-      "learning_rate": 1.6850636359507084e-05,
-      "loss": 0.985,
-      "step": 7200
-    },
-    {
-      "epoch": 0.3337969401947149,
-      "eval_loss": 0.9839121103286743,
-      "eval_runtime": 11.2603,
-      "eval_samples_per_second": 11.279,
-      "eval_steps_per_second": 11.279,
-      "step": 7200
-    },
-    {
-      "epoch": 0.33402874362540563,
-      "grad_norm": 4.698121070861816,
-      "learning_rate": 1.68447396201335e-05,
-      "loss": 0.8599,
-      "step": 7205
-    },
-    {
-      "epoch": 0.3342605470560964,
-      "grad_norm": 3.780442953109741,
-      "learning_rate": 1.6838838399384233e-05,
-      "loss": 0.8712,
-      "step": 7210
-    },
-    {
-      "epoch": 0.3344923504867872,
-      "grad_norm": 3.3134045600891113,
-      "learning_rate": 1.6832932701122917e-05,
-      "loss": 0.9061,
-      "step": 7215
-    },
-    {
-      "epoch": 0.33472415391747795,
-      "grad_norm": 4.624858379364014,
-      "learning_rate": 1.682702252921613e-05,
-      "loss": 0.8832,
-      "step": 7220
-    },
-    {
-      "epoch": 0.33495595734816874,
-      "grad_norm": 3.465853214263916,
-      "learning_rate": 1.682110788753336e-05,
-      "loss": 0.7583,
-      "step": 7225
-    },
-    {
-      "epoch": 0.33518776077885953,
-      "grad_norm": 4.856706142425537,
-      "learning_rate": 1.6815188779947035e-05,
-      "loss": 1.1742,
-      "step": 7230
-    },
-    {
-      "epoch": 0.3354195642095503,
-      "grad_norm": 3.952514171600342,
-      "learning_rate": 1.68092652103325e-05,
-      "loss": 1.0064,
-      "step": 7235
-    },
-    {
-      "epoch": 0.33565136764024106,
-      "grad_norm": 3.868814706802368,
-      "learning_rate": 1.6803337182568025e-05,
-      "loss": 0.9385,
-      "step": 7240
-    },
-    {
-      "epoch": 0.33588317107093185,
-      "grad_norm": 3.863393783569336,
-      "learning_rate": 1.67974047005348e-05,
-      "loss": 1.0178,
-      "step": 7245
-    },
-    {
-      "epoch": 0.33611497450162264,
-      "grad_norm": 4.632596492767334,
-      "learning_rate": 1.6791467768116927e-05,
-      "loss": 1.0524,
-      "step": 7250
-    },
-    {
-      "epoch": 0.3363467779323134,
-      "grad_norm": 3.71608829498291,
-      "learning_rate": 1.6785526389201422e-05,
-      "loss": 1.0079,
-      "step": 7255
-    },
-    {
-      "epoch": 0.33657858136300417,
-      "grad_norm": 4.460151672363281,
-      "learning_rate": 1.6779580567678218e-05,
-      "loss": 0.8563,
-      "step": 7260
-    },
-    {
-      "epoch": 0.33681038479369496,
-      "grad_norm": 3.5773184299468994,
-      "learning_rate": 1.677363030744015e-05,
-      "loss": 0.9136,
-      "step": 7265
-    },
-    {
-      "epoch": 0.3370421882243857,
-      "grad_norm": 4.082240104675293,
-      "learning_rate": 1.6767675612382964e-05,
-      "loss": 1.0243,
-      "step": 7270
-    },
-    {
-      "epoch": 0.3372739916550765,
-      "grad_norm": 4.144321918487549,
-      "learning_rate": 1.6761716486405305e-05,
-      "loss": 0.9109,
-      "step": 7275
-    },
-    {
-      "epoch": 0.3375057950857673,
-      "grad_norm": 3.912290334701538,
-      "learning_rate": 1.6755752933408717e-05,
-      "loss": 0.8532,
-      "step": 7280
-    },
-    {
-      "epoch": 0.33773759851645807,
-      "grad_norm": 4.518337249755859,
-      "learning_rate": 1.6749784957297656e-05,
-      "loss": 0.8114,
-      "step": 7285
-    },
-    {
-      "epoch": 0.3379694019471488,
-      "grad_norm": 4.082379341125488,
-      "learning_rate": 1.674381256197946e-05,
-      "loss": 0.9259,
-      "step": 7290
-    },
-    {
-      "epoch": 0.3382012053778396,
-      "grad_norm": 3.4231996536254883,
-      "learning_rate": 1.6737835751364364e-05,
-      "loss": 0.9061,
-      "step": 7295
-    },
-    {
-      "epoch": 0.3384330088085304,
-      "grad_norm": 4.965606212615967,
-      "learning_rate": 1.6731854529365494e-05,
-      "loss": 1.0355,
-      "step": 7300
-    },
-    {
-      "epoch": 0.3384330088085304,
-      "eval_loss": 0.9803537130355835,
-      "eval_runtime": 11.265,
-      "eval_samples_per_second": 11.274,
-      "eval_steps_per_second": 11.274,
-      "step": 7300
-    },
-    {
-      "epoch": 0.3386648122392211,
-      "grad_norm": 4.2138237953186035,
-      "learning_rate": 1.6725868899898873e-05,
-      "loss": 0.9846,
-      "step": 7305
-    },
-    {
-      "epoch": 0.3388966156699119,
-      "grad_norm": 4.691622734069824,
-      "learning_rate": 1.671987886688339e-05,
-      "loss": 1.0037,
-      "step": 7310
-    },
-    {
-      "epoch": 0.3391284191006027,
-      "grad_norm": 3.772075653076172,
-      "learning_rate": 1.6713884434240838e-05,
-      "loss": 0.9368,
-      "step": 7315
-    },
-    {
-      "epoch": 0.33936022253129344,
-      "grad_norm": 3.8454747200012207,
-      "learning_rate": 1.6707885605895876e-05,
-      "loss": 0.9976,
-      "step": 7320
-    },
-    {
-      "epoch": 0.33959202596198423,
-      "grad_norm": 3.014082193374634,
-      "learning_rate": 1.6701882385776047e-05,
-      "loss": 0.9104,
-      "step": 7325
-    },
-    {
-      "epoch": 0.339823829392675,
-      "grad_norm": 4.067466735839844,
-      "learning_rate": 1.6695874777811768e-05,
-      "loss": 1.014,
-      "step": 7330
-    },
-    {
-      "epoch": 0.34005563282336576,
-      "grad_norm": 4.160797595977783,
-      "learning_rate": 1.6689862785936336e-05,
-      "loss": 0.9069,
-      "step": 7335
-    },
-    {
-      "epoch": 0.34028743625405655,
-      "grad_norm": 3.3378775119781494,
-      "learning_rate": 1.66838464140859e-05,
-      "loss": 0.7564,
-      "step": 7340
-    },
-    {
-      "epoch": 0.34051923968474734,
-      "grad_norm": 3.7165188789367676,
-      "learning_rate": 1.6677825666199497e-05,
-      "loss": 0.8907,
-      "step": 7345
-    },
-    {
-      "epoch": 0.34075104311543813,
-      "grad_norm": 4.068546772003174,
-      "learning_rate": 1.667180054621902e-05,
-      "loss": 0.8771,
-      "step": 7350
-    },
-    {
-      "epoch": 0.34098284654612887,
-      "grad_norm": 4.538022994995117,
-      "learning_rate": 1.6665771058089218e-05,
-      "loss": 1.0441,
-      "step": 7355
-    },
-    {
-      "epoch": 0.34121464997681966,
-      "grad_norm": 3.9702048301696777,
-      "learning_rate": 1.6659737205757713e-05,
-      "loss": 0.989,
-      "step": 7360
-    },
-    {
-      "epoch": 0.34144645340751045,
-      "grad_norm": 3.565910816192627,
-      "learning_rate": 1.6653698993174975e-05,
-      "loss": 0.939,
-      "step": 7365
-    },
-    {
-      "epoch": 0.3416782568382012,
-      "grad_norm": 3.3991103172302246,
-      "learning_rate": 1.6647656424294334e-05,
-      "loss": 0.9982,
-      "step": 7370
-    },
-    {
-      "epoch": 0.341910060268892,
-      "grad_norm": 3.2477567195892334,
-      "learning_rate": 1.6641609503071968e-05,
-      "loss": 0.9894,
-      "step": 7375
-    },
-    {
-      "epoch": 0.34214186369958277,
-      "grad_norm": 3.702672004699707,
-      "learning_rate": 1.663555823346691e-05,
-      "loss": 0.9833,
-      "step": 7380
-    },
-    {
-      "epoch": 0.3423736671302735,
-      "grad_norm": 3.8544015884399414,
-      "learning_rate": 1.6629502619441028e-05,
-      "loss": 1.0198,
-      "step": 7385
-    },
-    {
-      "epoch": 0.3426054705609643,
-      "grad_norm": 3.860546827316284,
-      "learning_rate": 1.6623442664959054e-05,
-      "loss": 0.849,
-      "step": 7390
-    },
-    {
-      "epoch": 0.3428372739916551,
-      "grad_norm": 4.790958881378174,
-      "learning_rate": 1.6617378373988544e-05,
-      "loss": 0.8904,
-      "step": 7395
-    },
-    {
-      "epoch": 0.3430690774223458,
-      "grad_norm": 4.495256423950195,
-      "learning_rate": 1.6611309750499898e-05,
-      "loss": 1.0719,
-      "step": 7400
-    },
-    {
-      "epoch": 0.3430690774223458,
-      "eval_loss": 0.9776688814163208,
-      "eval_runtime": 11.2736,
-      "eval_samples_per_second": 11.265,
-      "eval_steps_per_second": 11.265,
-      "step": 7400
-    },
-    {
-      "epoch": 0.3433008808530366,
-      "grad_norm": 3.7262792587280273,
-      "learning_rate": 1.6605236798466358e-05,
-      "loss": 0.8616,
-      "step": 7405
-    },
-    {
-      "epoch": 0.3435326842837274,
-      "grad_norm": 3.547213554382324,
-      "learning_rate": 1.6599159521863995e-05,
-      "loss": 0.9388,
-      "step": 7410
-    },
-    {
-      "epoch": 0.3437644877144182,
-      "grad_norm": 3.846677541732788,
-      "learning_rate": 1.6593077924671712e-05,
-      "loss": 1.0564,
-      "step": 7415
-    },
-    {
-      "epoch": 0.34399629114510893,
-      "grad_norm": 3.696018695831299,
-      "learning_rate": 1.6586992010871245e-05,
-      "loss": 1.0205,
-      "step": 7420
-    },
-    {
-      "epoch": 0.3442280945757997,
-      "grad_norm": 4.1325249671936035,
-      "learning_rate": 1.658090178444715e-05,
-      "loss": 0.8804,
-      "step": 7425
-    },
-    {
-      "epoch": 0.3444598980064905,
-      "grad_norm": 4.22160005569458,
-      "learning_rate": 1.657480724938681e-05,
-      "loss": 0.9673,
-      "step": 7430
-    },
-    {
-      "epoch": 0.34469170143718125,
-      "grad_norm": 3.7792885303497314,
-      "learning_rate": 1.6568708409680422e-05,
-      "loss": 0.9713,
-      "step": 7435
-    },
-    {
-      "epoch": 0.34492350486787204,
-      "grad_norm": 3.9127721786499023,
-      "learning_rate": 1.6562605269321017e-05,
-      "loss": 0.8601,
-      "step": 7440
-    },
-    {
-      "epoch": 0.34515530829856284,
-      "grad_norm": 3.9979143142700195,
-      "learning_rate": 1.6556497832304425e-05,
-      "loss": 1.0057,
-      "step": 7445
-    },
-    {
-      "epoch": 0.34538711172925357,
-      "grad_norm": 4.151058673858643,
-      "learning_rate": 1.6550386102629302e-05,
-      "loss": 1.0317,
-      "step": 7450
-    },
-    {
-      "epoch": 0.34561891515994436,
-      "grad_norm": 4.163232803344727,
-      "learning_rate": 1.6544270084297107e-05,
-      "loss": 1.0535,
-      "step": 7455
-    },
-    {
-      "epoch": 0.34585071859063515,
-      "grad_norm": 3.3977575302124023,
-      "learning_rate": 1.6538149781312102e-05,
-      "loss": 0.9496,
-      "step": 7460
-    },
-    {
-      "epoch": 0.3460825220213259,
-      "grad_norm": 4.182466983795166,
-      "learning_rate": 1.6532025197681368e-05,
-      "loss": 0.9746,
-      "step": 7465
-    },
-    {
-      "epoch": 0.3463143254520167,
-      "grad_norm": 4.1667256355285645,
-      "learning_rate": 1.652589633741478e-05,
-      "loss": 0.9128,
-      "step": 7470
-    },
-    {
-      "epoch": 0.34654612888270747,
-      "grad_norm": 3.694902181625366,
-      "learning_rate": 1.6519763204525015e-05,
-      "loss": 0.8044,
-      "step": 7475
-    },
-    {
-      "epoch": 0.34677793231339826,
-      "grad_norm": 4.3444085121154785,
-      "learning_rate": 1.6513625803027545e-05,
-      "loss": 0.9057,
-      "step": 7480
-    },
-    {
-      "epoch": 0.347009735744089,
-      "grad_norm": 4.2170090675354,
-      "learning_rate": 1.650748413694064e-05,
-      "loss": 1.0288,
-      "step": 7485
-    },
-    {
-      "epoch": 0.3472415391747798,
-      "grad_norm": 3.5477216243743896,
-      "learning_rate": 1.650133821028536e-05,
-      "loss": 0.9128,
-      "step": 7490
-    },
-    {
-      "epoch": 0.3474733426054706,
-      "grad_norm": 4.178347587585449,
-      "learning_rate": 1.6495188027085554e-05,
-      "loss": 1.012,
-      "step": 7495
-    },
-    {
-      "epoch": 0.3477051460361613,
-      "grad_norm": 4.44584321975708,
-      "learning_rate": 1.648903359136786e-05,
-      "loss": 1.1048,
-      "step": 7500
-    },
-    {
-      "epoch": 0.3477051460361613,
-      "eval_loss": 0.9784466624259949,
-      "eval_runtime": 11.2673,
-      "eval_samples_per_second": 11.272,
-      "eval_steps_per_second": 11.272,
-      "step": 7500
-    },
-    {
-      "epoch": 0.3479369494668521,
-      "grad_norm": 3.767441511154175,
-      "learning_rate": 1.64828749071617e-05,
-      "loss": 1.0245,
-      "step": 7505
-    },
-    {
-      "epoch": 0.3481687528975429,
-      "grad_norm": 3.552576780319214,
-      "learning_rate": 1.6476711978499275e-05,
-      "loss": 0.8817,
-      "step": 7510
-    },
-    {
-      "epoch": 0.34840055632823363,
-      "grad_norm": 4.1061906814575195,
-      "learning_rate": 1.6470544809415568e-05,
-      "loss": 1.0079,
-      "step": 7515
-    },
-    {
-      "epoch": 0.3486323597589244,
-      "grad_norm": 4.06695032119751,
-      "learning_rate": 1.6464373403948336e-05,
-      "loss": 0.957,
-      "step": 7520
-    },
-    {
-      "epoch": 0.3488641631896152,
-      "grad_norm": 3.781661033630371,
-      "learning_rate": 1.6458197766138115e-05,
-      "loss": 0.9984,
-      "step": 7525
-    },
-    {
-      "epoch": 0.34909596662030595,
-      "grad_norm": 3.849099636077881,
-      "learning_rate": 1.6452017900028198e-05,
-      "loss": 0.9009,
-      "step": 7530
-    },
-    {
-      "epoch": 0.34932777005099674,
-      "grad_norm": 4.14891242980957,
-      "learning_rate": 1.6445833809664664e-05,
-      "loss": 1.1266,
-      "step": 7535
-    },
-    {
-      "epoch": 0.34955957348168754,
-      "grad_norm": 5.304938316345215,
-      "learning_rate": 1.6439645499096347e-05,
-      "loss": 0.9744,
-      "step": 7540
-    },
-    {
-      "epoch": 0.3497913769123783,
-      "grad_norm": 3.5674333572387695,
-      "learning_rate": 1.6433452972374846e-05,
-      "loss": 0.9666,
-      "step": 7545
-    },
-    {
-      "epoch": 0.35002318034306906,
-      "grad_norm": 3.3488993644714355,
-      "learning_rate": 1.6427256233554523e-05,
-      "loss": 0.8965,
-      "step": 7550
-    },
-    {
-      "epoch": 0.35025498377375985,
-      "grad_norm": 3.5717201232910156,
-      "learning_rate": 1.6421055286692492e-05,
-      "loss": 0.8758,
-      "step": 7555
-    },
-    {
-      "epoch": 0.35048678720445064,
-      "grad_norm": 4.32782506942749,
-      "learning_rate": 1.641485013584863e-05,
-      "loss": 1.021,
-      "step": 7560
-    },
-    {
-      "epoch": 0.3507185906351414,
-      "grad_norm": 4.122309684753418,
-      "learning_rate": 1.640864078508556e-05,
-      "loss": 0.93,
-      "step": 7565
-    },
-    {
-      "epoch": 0.35095039406583217,
-      "grad_norm": 3.43186616897583,
-      "learning_rate": 1.640242723846866e-05,
-      "loss": 0.9018,
-      "step": 7570
-    },
-    {
-      "epoch": 0.35118219749652296,
-      "grad_norm": 4.272997856140137,
-      "learning_rate": 1.6396209500066045e-05,
-      "loss": 0.9095,
-      "step": 7575
-    },
-    {
-      "epoch": 0.3514140009272137,
-      "grad_norm": 3.9995079040527344,
-      "learning_rate": 1.6389987573948588e-05,
-      "loss": 0.9397,
-      "step": 7580
-    },
-    {
-      "epoch": 0.3516458043579045,
-      "grad_norm": 3.7181596755981445,
-      "learning_rate": 1.6383761464189895e-05,
-      "loss": 0.7581,
-      "step": 7585
-    },
-    {
-      "epoch": 0.3518776077885953,
-      "grad_norm": 3.570861577987671,
-      "learning_rate": 1.6377531174866315e-05,
-      "loss": 0.973,
-      "step": 7590
-    },
-    {
-      "epoch": 0.3521094112192861,
-      "grad_norm": 3.842984676361084,
-      "learning_rate": 1.637129671005693e-05,
-      "loss": 0.939,
-      "step": 7595
-    },
-    {
-      "epoch": 0.3523412146499768,
-      "grad_norm": 3.6865546703338623,
-      "learning_rate": 1.6365058073843562e-05,
-      "loss": 0.958,
-      "step": 7600
-    },
-    {
-      "epoch": 0.3523412146499768,
-      "eval_loss": 0.9766842722892761,
-      "eval_runtime": 11.2741,
-      "eval_samples_per_second": 11.265,
-      "eval_steps_per_second": 11.265,
-      "step": 7600
-    },
-    {
-      "epoch": 0.3525730180806676,
-      "grad_norm": 3.9954993724823,
-      "learning_rate": 1.6358815270310752e-05,
-      "loss": 1.032,
-      "step": 7605
-    },
-    {
-      "epoch": 0.3528048215113584,
-      "grad_norm": 3.9146227836608887,
-      "learning_rate": 1.6352568303545782e-05,
-      "loss": 0.9987,
-      "step": 7610
-    },
-    {
-      "epoch": 0.3530366249420491,
-      "grad_norm": 3.8388328552246094,
-      "learning_rate": 1.6346317177638658e-05,
-      "loss": 0.8106,
-      "step": 7615
-    },
-    {
-      "epoch": 0.3532684283727399,
-      "grad_norm": 3.9061410427093506,
-      "learning_rate": 1.6340061896682098e-05,
-      "loss": 1.0071,
-      "step": 7620
-    },
-    {
-      "epoch": 0.3535002318034307,
-      "grad_norm": 5.165677547454834,
-      "learning_rate": 1.6333802464771557e-05,
-      "loss": 1.1079,
-      "step": 7625
-    },
-    {
-      "epoch": 0.35373203523412144,
-      "grad_norm": 4.240540027618408,
-      "learning_rate": 1.632753888600519e-05,
-      "loss": 0.9281,
-      "step": 7630
-    },
-    {
-      "epoch": 0.35396383866481224,
-      "grad_norm": 3.3363611698150635,
-      "learning_rate": 1.6321271164483884e-05,
-      "loss": 0.9122,
-      "step": 7635
-    },
-    {
-      "epoch": 0.354195642095503,
-      "grad_norm": 3.6181395053863525,
-      "learning_rate": 1.6314999304311222e-05,
-      "loss": 0.7473,
-      "step": 7640
-    },
-    {
-      "epoch": 0.35442744552619376,
-      "grad_norm": 3.771057605743408,
-      "learning_rate": 1.630872330959352e-05,
-      "loss": 1.0243,
-      "step": 7645
-    },
-    {
-      "epoch": 0.35465924895688455,
-      "grad_norm": 4.966143608093262,
-      "learning_rate": 1.630244318443977e-05,
-      "loss": 0.9069,
-      "step": 7650
-    },
-    {
-      "epoch": 0.35489105238757535,
-      "grad_norm": 4.111125946044922,
-      "learning_rate": 1.629615893296169e-05,
-      "loss": 0.9339,
-      "step": 7655
-    },
-    {
-      "epoch": 0.35512285581826614,
-      "grad_norm": 3.705050230026245,
-      "learning_rate": 1.6289870559273696e-05,
-      "loss": 0.8188,
-      "step": 7660
-    },
-    {
-      "epoch": 0.3553546592489569,
-      "grad_norm": 3.890373945236206,
-      "learning_rate": 1.62835780674929e-05,
-      "loss": 0.7916,
-      "step": 7665
-    },
-    {
-      "epoch": 0.35558646267964766,
-      "grad_norm": 3.7870006561279297,
-      "learning_rate": 1.627728146173911e-05,
-      "loss": 0.8877,
-      "step": 7670
-    },
-    {
-      "epoch": 0.35581826611033845,
-      "grad_norm": 4.223922252655029,
-      "learning_rate": 1.627098074613483e-05,
-      "loss": 0.9675,
-      "step": 7675
-    },
-    {
-      "epoch": 0.3560500695410292,
-      "grad_norm": 3.460359573364258,
-      "learning_rate": 1.626467592480525e-05,
-      "loss": 0.956,
-      "step": 7680
-    },
-    {
-      "epoch": 0.35628187297172,
-      "grad_norm": 4.657145977020264,
-      "learning_rate": 1.625836700187825e-05,
-      "loss": 0.8836,
-      "step": 7685
-    },
-    {
-      "epoch": 0.3565136764024108,
-      "grad_norm": 3.332714319229126,
-      "learning_rate": 1.625205398148441e-05,
-      "loss": 0.9026,
-      "step": 7690
-    },
-    {
-      "epoch": 0.3567454798331015,
-      "grad_norm": 3.8605220317840576,
-      "learning_rate": 1.624573686775696e-05,
-      "loss": 1.0878,
-      "step": 7695
-    },
-    {
-      "epoch": 0.3569772832637923,
-      "grad_norm": 3.559641122817993,
-      "learning_rate": 1.6239415664831844e-05,
-      "loss": 0.9073,
-      "step": 7700
-    },
-    {
-      "epoch": 0.3569772832637923,
-      "eval_loss": 0.9736750721931458,
-      "eval_runtime": 11.2674,
-      "eval_samples_per_second": 11.271,
-      "eval_steps_per_second": 11.271,
-      "step": 7700
-    },
-    {
-      "epoch": 0.3572090866944831,
-      "grad_norm": 3.865588426589966,
-      "learning_rate": 1.623309037684766e-05,
-      "loss": 1.0368,
-      "step": 7705
-    },
-    {
-      "epoch": 0.3574408901251738,
-      "grad_norm": 6.948596000671387,
-      "learning_rate": 1.6226761007945696e-05,
-      "loss": 0.8942,
-      "step": 7710
-    },
-    {
-      "epoch": 0.3576726935558646,
-      "grad_norm": 4.2951812744140625,
-      "learning_rate": 1.62204275622699e-05,
-      "loss": 0.7803,
-      "step": 7715
-    },
-    {
-      "epoch": 0.3579044969865554,
-      "grad_norm": 3.882315158843994,
-      "learning_rate": 1.62140900439669e-05,
-      "loss": 0.8272,
-      "step": 7720
-    },
-    {
-      "epoch": 0.3581363004172462,
-      "grad_norm": 4.093032360076904,
-      "learning_rate": 1.6207748457185974e-05,
-      "loss": 0.8589,
-      "step": 7725
-    },
-    {
-      "epoch": 0.35836810384793694,
-      "grad_norm": 3.319401264190674,
-      "learning_rate": 1.6201402806079086e-05,
-      "loss": 0.7305,
-      "step": 7730
-    },
-    {
-      "epoch": 0.3585999072786277,
-      "grad_norm": 3.839540958404541,
-      "learning_rate": 1.6195053094800844e-05,
-      "loss": 1.0484,
-      "step": 7735
-    },
-    {
-      "epoch": 0.3588317107093185,
-      "grad_norm": 3.2904021739959717,
-      "learning_rate": 1.6188699327508518e-05,
-      "loss": 0.9034,
-      "step": 7740
-    },
-    {
-      "epoch": 0.35906351414000925,
-      "grad_norm": 3.4221081733703613,
-      "learning_rate": 1.6182341508362044e-05,
-      "loss": 0.9592,
-      "step": 7745
-    },
-    {
-      "epoch": 0.35929531757070005,
-      "grad_norm": 4.381012916564941,
-      "learning_rate": 1.6175979641523986e-05,
-      "loss": 1.008,
-      "step": 7750
-    },
-    {
-      "epoch": 0.35952712100139084,
-      "grad_norm": 4.044552326202393,
-      "learning_rate": 1.6169613731159586e-05,
-      "loss": 0.8802,
-      "step": 7755
-    },
-    {
-      "epoch": 0.3597589244320816,
-      "grad_norm": 3.523761510848999,
-      "learning_rate": 1.6163243781436723e-05,
-      "loss": 0.9593,
-      "step": 7760
-    },
-    {
-      "epoch": 0.35999072786277236,
-      "grad_norm": 3.548126697540283,
-      "learning_rate": 1.615686979652591e-05,
-      "loss": 0.8895,
-      "step": 7765
-    },
-    {
-      "epoch": 0.36022253129346316,
-      "grad_norm": 4.415576457977295,
-      "learning_rate": 1.615049178060032e-05,
-      "loss": 0.9599,
-      "step": 7770
-    },
-    {
-      "epoch": 0.3604543347241539,
-      "grad_norm": 3.6763060092926025,
-      "learning_rate": 1.6144109737835748e-05,
-      "loss": 1.0321,
-      "step": 7775
-    },
-    {
-      "epoch": 0.3606861381548447,
-      "grad_norm": 3.7065987586975098,
-      "learning_rate": 1.613772367241064e-05,
-      "loss": 0.9211,
-      "step": 7780
-    },
-    {
-      "epoch": 0.3609179415855355,
-      "grad_norm": 3.909865379333496,
-      "learning_rate": 1.6131333588506064e-05,
-      "loss": 0.9941,
-      "step": 7785
-    },
-    {
-      "epoch": 0.36114974501622626,
-      "grad_norm": 3.2999017238616943,
-      "learning_rate": 1.6124939490305732e-05,
-      "loss": 0.8126,
-      "step": 7790
-    },
-    {
-      "epoch": 0.361381548446917,
-      "grad_norm": 4.182840347290039,
-      "learning_rate": 1.6118541381995966e-05,
-      "loss": 0.9713,
-      "step": 7795
-    },
-    {
-      "epoch": 0.3616133518776078,
-      "grad_norm": 4.793496131896973,
-      "learning_rate": 1.6112139267765737e-05,
-      "loss": 0.9375,
-      "step": 7800
-    },
-    {
-      "epoch": 0.3616133518776078,
-      "eval_loss": 0.9727134108543396,
-      "eval_runtime": 11.2846,
-      "eval_samples_per_second": 11.254,
-      "eval_steps_per_second": 11.254,
-      "step": 7800
-    },
-    {
-      "epoch": 0.3618451553082986,
-      "grad_norm": 3.9969582557678223,
-      "learning_rate": 1.6105733151806615e-05,
-      "loss": 1.021,
-      "step": 7805
-    },
-    {
-      "epoch": 0.3620769587389893,
-      "grad_norm": 3.9675846099853516,
-      "learning_rate": 1.60993230383128e-05,
-      "loss": 0.9999,
-      "step": 7810
-    },
-    {
-      "epoch": 0.3623087621696801,
-      "grad_norm": 3.7798831462860107,
-      "learning_rate": 1.609290893148112e-05,
-      "loss": 1.0193,
-      "step": 7815
-    },
-    {
-      "epoch": 0.3625405656003709,
-      "grad_norm": 4.016881942749023,
-      "learning_rate": 1.6086490835511e-05,
-      "loss": 1.0227,
-      "step": 7820
-    },
-    {
-      "epoch": 0.36277236903106164,
-      "grad_norm": 3.7668068408966064,
-      "learning_rate": 1.6080068754604486e-05,
-      "loss": 1.0612,
-      "step": 7825
-    },
-    {
-      "epoch": 0.36300417246175243,
-      "grad_norm": 3.260951280593872,
-      "learning_rate": 1.6073642692966233e-05,
-      "loss": 0.955,
-      "step": 7830
-    },
-    {
-      "epoch": 0.3632359758924432,
-      "grad_norm": 3.218118190765381,
-      "learning_rate": 1.6067212654803503e-05,
-      "loss": 0.7994,
-      "step": 7835
-    },
-    {
-      "epoch": 0.36346777932313395,
-      "grad_norm": 4.213292121887207,
-      "learning_rate": 1.606077864432615e-05,
-      "loss": 0.909,
-      "step": 7840
-    },
-    {
-      "epoch": 0.36369958275382475,
-      "grad_norm": 4.328354835510254,
-      "learning_rate": 1.6054340665746645e-05,
-      "loss": 0.9793,
-      "step": 7845
-    },
-    {
-      "epoch": 0.36393138618451554,
-      "grad_norm": 3.9034736156463623,
-      "learning_rate": 1.6047898723280045e-05,
-      "loss": 0.9643,
-      "step": 7850
-    },
-    {
-      "epoch": 0.36416318961520633,
-      "grad_norm": 3.579296350479126,
-      "learning_rate": 1.6041452821144013e-05,
-      "loss": 0.8492,
-      "step": 7855
-    },
-    {
-      "epoch": 0.36439499304589706,
-      "grad_norm": 6.530494689941406,
-      "learning_rate": 1.6035002963558796e-05,
-      "loss": 1.1487,
-      "step": 7860
-    },
-    {
-      "epoch": 0.36462679647658786,
-      "grad_norm": 4.108696937561035,
-      "learning_rate": 1.602854915474723e-05,
-      "loss": 0.911,
-      "step": 7865
-    },
-    {
-      "epoch": 0.36485859990727865,
-      "grad_norm": 4.311334133148193,
-      "learning_rate": 1.6022091398934746e-05,
-      "loss": 0.9403,
-      "step": 7870
-    },
-    {
-      "epoch": 0.3650904033379694,
-      "grad_norm": 3.4167885780334473,
-      "learning_rate": 1.601562970034935e-05,
-      "loss": 0.9623,
-      "step": 7875
-    },
-    {
-      "epoch": 0.3653222067686602,
-      "grad_norm": 3.9069101810455322,
-      "learning_rate": 1.600916406322164e-05,
-      "loss": 0.8251,
-      "step": 7880
-    },
-    {
-      "epoch": 0.36555401019935096,
-      "grad_norm": 3.9749691486358643,
-      "learning_rate": 1.600269449178478e-05,
-      "loss": 0.8434,
-      "step": 7885
-    },
-    {
-      "epoch": 0.3657858136300417,
-      "grad_norm": 3.524258852005005,
-      "learning_rate": 1.599622099027452e-05,
-      "loss": 0.8075,
-      "step": 7890
-    },
-    {
-      "epoch": 0.3660176170607325,
-      "grad_norm": 3.6314921379089355,
-      "learning_rate": 1.5989743562929177e-05,
-      "loss": 0.9493,
-      "step": 7895
-    },
-    {
-      "epoch": 0.3662494204914233,
-      "grad_norm": 3.23356294631958,
-      "learning_rate": 1.598326221398964e-05,
-      "loss": 0.9669,
-      "step": 7900
-    },
-    {
-      "epoch": 0.3662494204914233,
-      "eval_loss": 0.9721471071243286,
-      "eval_runtime": 11.2707,
-      "eval_samples_per_second": 11.268,
-      "eval_steps_per_second": 11.268,
-      "step": 7900
-    },
-    {
-      "epoch": 0.3664812239221141,
-      "grad_norm": 4.47834587097168,
-      "learning_rate": 1.5976776947699368e-05,
-      "loss": 1.0061,
-      "step": 7905
-    },
-    {
-      "epoch": 0.3667130273528048,
-      "grad_norm": 4.536191463470459,
-      "learning_rate": 1.5970287768304385e-05,
-      "loss": 1.0296,
-      "step": 7910
-    },
-    {
-      "epoch": 0.3669448307834956,
-      "grad_norm": 3.6170647144317627,
-      "learning_rate": 1.596379468005327e-05,
-      "loss": 0.8788,
-      "step": 7915
-    },
-    {
-      "epoch": 0.3671766342141864,
-      "grad_norm": 3.2994272708892822,
-      "learning_rate": 1.5957297687197172e-05,
-      "loss": 0.8066,
-      "step": 7920
-    },
-    {
-      "epoch": 0.36740843764487713,
-      "grad_norm": 3.139523983001709,
-      "learning_rate": 1.5950796793989785e-05,
-      "loss": 0.882,
-      "step": 7925
-    },
-    {
-      "epoch": 0.3676402410755679,
-      "grad_norm": 3.833423376083374,
-      "learning_rate": 1.5944292004687362e-05,
-      "loss": 0.9673,
-      "step": 7930
-    },
-    {
-      "epoch": 0.3678720445062587,
-      "grad_norm": 4.070364952087402,
-      "learning_rate": 1.5937783323548715e-05,
-      "loss": 0.9485,
-      "step": 7935
-    },
-    {
-      "epoch": 0.36810384793694945,
-      "grad_norm": 4.206231117248535,
-      "learning_rate": 1.593127075483519e-05,
-      "loss": 1.0304,
-      "step": 7940
-    },
-    {
-      "epoch": 0.36833565136764024,
-      "grad_norm": 4.32648229598999,
-      "learning_rate": 1.5924754302810682e-05,
-      "loss": 1.0254,
-      "step": 7945
-    },
-    {
-      "epoch": 0.36856745479833103,
-      "grad_norm": 3.5725724697113037,
-      "learning_rate": 1.591823397174163e-05,
-      "loss": 0.9672,
-      "step": 7950
-    },
-    {
-      "epoch": 0.36879925822902176,
-      "grad_norm": 4.0060954093933105,
-      "learning_rate": 1.5911709765897027e-05,
-      "loss": 0.9852,
-      "step": 7955
-    },
-    {
-      "epoch": 0.36903106165971256,
-      "grad_norm": 4.210195541381836,
-      "learning_rate": 1.5905181689548373e-05,
-      "loss": 0.9808,
-      "step": 7960
-    },
-    {
-      "epoch": 0.36926286509040335,
-      "grad_norm": 4.058366775512695,
-      "learning_rate": 1.5898649746969728e-05,
-      "loss": 0.8752,
-      "step": 7965
-    },
-    {
-      "epoch": 0.36949466852109414,
-      "grad_norm": 3.9418766498565674,
-      "learning_rate": 1.589211394243767e-05,
-      "loss": 1.0223,
-      "step": 7970
-    },
-    {
-      "epoch": 0.3697264719517849,
-      "grad_norm": 3.892645835876465,
-      "learning_rate": 1.5885574280231308e-05,
-      "loss": 0.8914,
-      "step": 7975
-    },
-    {
-      "epoch": 0.36995827538247567,
-      "grad_norm": 3.822187900543213,
-      "learning_rate": 1.587903076463228e-05,
-      "loss": 1.0347,
-      "step": 7980
-    },
-    {
-      "epoch": 0.37019007881316646,
-      "grad_norm": 4.411693096160889,
-      "learning_rate": 1.587248339992475e-05,
-      "loss": 1.0486,
-      "step": 7985
-    },
-    {
-      "epoch": 0.3704218822438572,
-      "grad_norm": 3.2753002643585205,
-      "learning_rate": 1.5865932190395386e-05,
-      "loss": 0.801,
-      "step": 7990
-    },
-    {
-      "epoch": 0.370653685674548,
-      "grad_norm": 3.6129369735717773,
-      "learning_rate": 1.5859377140333393e-05,
-      "loss": 0.8998,
-      "step": 7995
-    },
-    {
-      "epoch": 0.3708854891052388,
-      "grad_norm": 3.9310081005096436,
-      "learning_rate": 1.5852818254030474e-05,
-      "loss": 0.922,
-      "step": 8000
-    },
-    {
-      "epoch": 0.3708854891052388,
-      "eval_loss": 0.9707987904548645,
-      "eval_runtime": 11.2704,
-      "eval_samples_per_second": 11.268,
-      "eval_steps_per_second": 11.268,
-      "step": 8000
-    },
-    {
-      "epoch": 0.3711172925359295,
-      "grad_norm": 4.156801700592041,
-      "learning_rate": 1.5846255535780863e-05,
-      "loss": 1.0529,
-      "step": 8005
-    },
-    {
-      "epoch": 0.3713490959666203,
-      "grad_norm": 5.164330005645752,
-      "learning_rate": 1.583968898988128e-05,
-      "loss": 1.0731,
-      "step": 8010
-    },
-    {
-      "epoch": 0.3715808993973111,
-      "grad_norm": 3.661679983139038,
-      "learning_rate": 1.5833118620630966e-05,
-      "loss": 0.9838,
-      "step": 8015
-    },
-    {
-      "epoch": 0.37181270282800183,
-      "grad_norm": 4.266704082489014,
-      "learning_rate": 1.5826544432331666e-05,
-      "loss": 0.8833,
-      "step": 8020
-    },
-    {
-      "epoch": 0.3720445062586926,
-      "grad_norm": 5.135181903839111,
-      "learning_rate": 1.5819966429287612e-05,
-      "loss": 1.0761,
-      "step": 8025
-    },
-    {
-      "epoch": 0.3722763096893834,
-      "grad_norm": 3.6553714275360107,
-      "learning_rate": 1.581338461580555e-05,
-      "loss": 0.7655,
-      "step": 8030
-    },
-    {
-      "epoch": 0.3725081131200742,
-      "grad_norm": 4.098663806915283,
-      "learning_rate": 1.5806798996194704e-05,
-      "loss": 1.0577,
-      "step": 8035
-    },
-    {
-      "epoch": 0.37273991655076494,
-      "grad_norm": 3.716402769088745,
-      "learning_rate": 1.5800209574766813e-05,
-      "loss": 0.9381,
-      "step": 8040
-    },
-    {
-      "epoch": 0.37297171998145573,
-      "grad_norm": 4.095868110656738,
-      "learning_rate": 1.5793616355836076e-05,
-      "loss": 1.0957,
-      "step": 8045
-    },
-    {
-      "epoch": 0.3732035234121465,
-      "grad_norm": 2.871941328048706,
-      "learning_rate": 1.5787019343719206e-05,
-      "loss": 0.6745,
-      "step": 8050
-    },
-    {
-      "epoch": 0.37343532684283726,
-      "grad_norm": 4.028416633605957,
-      "learning_rate": 1.578041854273538e-05,
-      "loss": 1.1041,
-      "step": 8055
-    },
-    {
-      "epoch": 0.37366713027352805,
-      "grad_norm": 3.7373595237731934,
-      "learning_rate": 1.5773813957206265e-05,
-      "loss": 1.1335,
-      "step": 8060
-    },
-    {
-      "epoch": 0.37389893370421884,
-      "grad_norm": 3.2081689834594727,
-      "learning_rate": 1.5767205591456e-05,
-      "loss": 0.8999,
-      "step": 8065
-    },
-    {
-      "epoch": 0.3741307371349096,
-      "grad_norm": 3.564337730407715,
-      "learning_rate": 1.576059344981121e-05,
-      "loss": 0.816,
-      "step": 8070
-    },
-    {
-      "epoch": 0.37436254056560037,
-      "grad_norm": 4.3315653800964355,
-      "learning_rate": 1.575397753660098e-05,
-      "loss": 0.9207,
-      "step": 8075
-    },
-    {
-      "epoch": 0.37459434399629116,
-      "grad_norm": 4.162485599517822,
-      "learning_rate": 1.5747357856156864e-05,
-      "loss": 0.9209,
-      "step": 8080
-    },
-    {
-      "epoch": 0.3748261474269819,
-      "grad_norm": 3.711111545562744,
-      "learning_rate": 1.5740734412812894e-05,
-      "loss": 0.7482,
-      "step": 8085
-    },
-    {
-      "epoch": 0.3750579508576727,
-      "grad_norm": 3.6291377544403076,
-      "learning_rate": 1.5734107210905557e-05,
-      "loss": 0.8207,
-      "step": 8090
-    },
-    {
-      "epoch": 0.3752897542883635,
-      "grad_norm": 3.9952619075775146,
-      "learning_rate": 1.5727476254773805e-05,
-      "loss": 1.0326,
-      "step": 8095
-    },
-    {
-      "epoch": 0.37552155771905427,
-      "grad_norm": 3.7901878356933594,
-      "learning_rate": 1.572084154875904e-05,
-      "loss": 0.8866,
-      "step": 8100
-    },
-    {
-      "epoch": 0.37552155771905427,
-      "eval_loss": 0.9721558094024658,
-      "eval_runtime": 11.2673,
-      "eval_samples_per_second": 11.272,
-      "eval_steps_per_second": 11.272,
-      "step": 8100
-    },
-    {
-      "epoch": 0.375753361149745,
-      "grad_norm": 3.700965642929077,
-      "learning_rate": 1.5714203097205126e-05,
-      "loss": 0.9772,
-      "step": 8105
-    },
-    {
-      "epoch": 0.3759851645804358,
-      "grad_norm": 3.4462482929229736,
-      "learning_rate": 1.5707560904458385e-05,
-      "loss": 0.8462,
-      "step": 8110
-    },
-    {
-      "epoch": 0.3762169680111266,
-      "grad_norm": 4.038230895996094,
-      "learning_rate": 1.570091497486758e-05,
-      "loss": 0.9706,
-      "step": 8115
-    },
-    {
-      "epoch": 0.3764487714418173,
-      "grad_norm": 3.681793689727783,
-      "learning_rate": 1.5694265312783915e-05,
-      "loss": 0.9543,
-      "step": 8120
-    },
-    {
-      "epoch": 0.3766805748725081,
-      "grad_norm": 3.430973529815674,
-      "learning_rate": 1.5687611922561048e-05,
-      "loss": 0.9888,
-      "step": 8125
-    },
-    {
-      "epoch": 0.3769123783031989,
-      "grad_norm": 4.354652404785156,
-      "learning_rate": 1.568095480855508e-05,
-      "loss": 0.9147,
-      "step": 8130
-    },
-    {
-      "epoch": 0.37714418173388964,
-      "grad_norm": 4.103789806365967,
-      "learning_rate": 1.5674293975124537e-05,
-      "loss": 1.02,
-      "step": 8135
-    },
-    {
-      "epoch": 0.37737598516458043,
-      "grad_norm": 3.2359514236450195,
-      "learning_rate": 1.5667629426630392e-05,
-      "loss": 0.8205,
-      "step": 8140
-    },
-    {
-      "epoch": 0.3776077885952712,
-      "grad_norm": 3.833833932876587,
-      "learning_rate": 1.5660961167436043e-05,
-      "loss": 0.8056,
-      "step": 8145
-    },
-    {
-      "epoch": 0.37783959202596196,
-      "grad_norm": 3.0617690086364746,
-      "learning_rate": 1.565428920190733e-05,
-      "loss": 0.9386,
-      "step": 8150
-    },
-    {
-      "epoch": 0.37807139545665275,
-      "grad_norm": 3.7850069999694824,
-      "learning_rate": 1.5647613534412495e-05,
-      "loss": 0.9045,
-      "step": 8155
-    },
-    {
-      "epoch": 0.37830319888734354,
-      "grad_norm": 4.245758056640625,
-      "learning_rate": 1.5640934169322236e-05,
-      "loss": 0.922,
-      "step": 8160
-    },
-    {
-      "epoch": 0.37853500231803433,
-      "grad_norm": 3.542757749557495,
-      "learning_rate": 1.5634251111009644e-05,
-      "loss": 1.0869,
-      "step": 8165
-    },
-    {
-      "epoch": 0.37876680574872507,
-      "grad_norm": 3.3576743602752686,
-      "learning_rate": 1.5627564363850244e-05,
-      "loss": 0.9502,
-      "step": 8170
-    },
-    {
-      "epoch": 0.37899860917941586,
-      "grad_norm": 3.7136809825897217,
-      "learning_rate": 1.5620873932221966e-05,
-      "loss": 0.8285,
-      "step": 8175
-    },
-    {
-      "epoch": 0.37923041261010665,
-      "grad_norm": 4.323848724365234,
-      "learning_rate": 1.5614179820505162e-05,
-      "loss": 1.0334,
-      "step": 8180
-    },
-    {
-      "epoch": 0.3794622160407974,
-      "grad_norm": 3.9784998893737793,
-      "learning_rate": 1.5607482033082586e-05,
-      "loss": 0.9102,
-      "step": 8185
-    },
-    {
-      "epoch": 0.3796940194714882,
-      "grad_norm": 3.1557223796844482,
-      "learning_rate": 1.56007805743394e-05,
-      "loss": 0.8081,
-      "step": 8190
-    },
-    {
-      "epoch": 0.37992582290217897,
-      "grad_norm": 4.566058158874512,
-      "learning_rate": 1.5594075448663183e-05,
-      "loss": 0.9558,
-      "step": 8195
-    },
-    {
-      "epoch": 0.3801576263328697,
-      "grad_norm": 3.514949321746826,
-      "learning_rate": 1.5587366660443887e-05,
-      "loss": 1.0877,
-      "step": 8200
-    },
-    {
-      "epoch": 0.3801576263328697,
-      "eval_loss": 0.9680636525154114,
-      "eval_runtime": 11.2709,
-      "eval_samples_per_second": 11.268,
-      "eval_steps_per_second": 11.268,
-      "step": 8200
-    },
-    {
-      "epoch": 0.3803894297635605,
-      "grad_norm": 3.738237142562866,
-      "learning_rate": 1.5580654214073885e-05,
-      "loss": 0.857,
-      "step": 8205
-    },
-    {
-      "epoch": 0.3806212331942513,
-      "grad_norm": 4.238804817199707,
-      "learning_rate": 1.5573938113947938e-05,
-      "loss": 0.8178,
-      "step": 8210
-    },
-    {
-      "epoch": 0.3808530366249421,
-      "grad_norm": 4.3957438468933105,
-      "learning_rate": 1.5567218364463198e-05,
-      "loss": 0.9658,
-      "step": 8215
-    },
-    {
-      "epoch": 0.3810848400556328,
-      "grad_norm": 3.63651967048645,
-      "learning_rate": 1.556049497001921e-05,
-      "loss": 0.9853,
-      "step": 8220
-    },
-    {
-      "epoch": 0.3813166434863236,
-      "grad_norm": 3.9356980323791504,
-      "learning_rate": 1.5553767935017895e-05,
-      "loss": 0.7605,
-      "step": 8225
-    },
-    {
-      "epoch": 0.3815484469170144,
-      "grad_norm": 4.169957637786865,
-      "learning_rate": 1.5547037263863577e-05,
-      "loss": 1.0644,
-      "step": 8230
-    },
-    {
-      "epoch": 0.38178025034770513,
-      "grad_norm": 3.485525131225586,
-      "learning_rate": 1.554030296096294e-05,
-      "loss": 0.9829,
-      "step": 8235
-    },
-    {
-      "epoch": 0.3820120537783959,
-      "grad_norm": 4.001947402954102,
-      "learning_rate": 1.553356503072506e-05,
-      "loss": 0.8294,
-      "step": 8240
-    },
-    {
-      "epoch": 0.3822438572090867,
-      "grad_norm": 3.6511752605438232,
-      "learning_rate": 1.552682347756138e-05,
-      "loss": 0.8348,
-      "step": 8245
-    },
-    {
-      "epoch": 0.38247566063977745,
-      "grad_norm": 3.49690580368042,
-      "learning_rate": 1.5520078305885712e-05,
-      "loss": 0.929,
-      "step": 8250
-    },
-    {
-      "epoch": 0.38270746407046824,
-      "grad_norm": 4.088393211364746,
-      "learning_rate": 1.5513329520114254e-05,
-      "loss": 0.9723,
-      "step": 8255
-    },
-    {
-      "epoch": 0.38293926750115903,
-      "grad_norm": 3.4010307788848877,
-      "learning_rate": 1.5506577124665552e-05,
-      "loss": 0.8488,
-      "step": 8260
-    },
-    {
-      "epoch": 0.38317107093184977,
-      "grad_norm": 4.285269260406494,
-      "learning_rate": 1.5499821123960525e-05,
-      "loss": 0.9262,
-      "step": 8265
-    },
-    {
-      "epoch": 0.38340287436254056,
-      "grad_norm": 4.1062517166137695,
-      "learning_rate": 1.5493061522422447e-05,
-      "loss": 0.8384,
-      "step": 8270
-    },
-    {
-      "epoch": 0.38363467779323135,
-      "grad_norm": 3.7376549243927,
-      "learning_rate": 1.548629832447696e-05,
-      "loss": 1.0614,
-      "step": 8275
-    },
-    {
-      "epoch": 0.38386648122392214,
-      "grad_norm": 4.19257926940918,
-      "learning_rate": 1.547953153455204e-05,
-      "loss": 1.16,
-      "step": 8280
-    },
-    {
-      "epoch": 0.3840982846546129,
-      "grad_norm": 4.338336944580078,
-      "learning_rate": 1.547276115707803e-05,
-      "loss": 1.0549,
-      "step": 8285
-    },
-    {
-      "epoch": 0.38433008808530367,
-      "grad_norm": 3.7592556476593018,
-      "learning_rate": 1.5465987196487634e-05,
-      "loss": 0.7895,
-      "step": 8290
-    },
-    {
-      "epoch": 0.38456189151599446,
-      "grad_norm": 3.735196352005005,
-      "learning_rate": 1.545920965721587e-05,
-      "loss": 0.9698,
-      "step": 8295
-    },
-    {
-      "epoch": 0.3847936949466852,
-      "grad_norm": 3.527911424636841,
-      "learning_rate": 1.5452428543700126e-05,
-      "loss": 0.9027,
-      "step": 8300
-    },
-    {
-      "epoch": 0.3847936949466852,
-      "eval_loss": 0.9673840403556824,
-      "eval_runtime": 11.2732,
-      "eval_samples_per_second": 11.266,
-      "eval_steps_per_second": 11.266,
-      "step": 8300
-    },
-    {
-      "epoch": 0.385025498377376,
-      "grad_norm": 4.203237533569336,
-      "learning_rate": 1.544564386038012e-05,
-      "loss": 1.004,
-      "step": 8305
-    },
-    {
-      "epoch": 0.3852573018080668,
-      "grad_norm": 4.254677772521973,
-      "learning_rate": 1.5438855611697903e-05,
-      "loss": 0.8579,
-      "step": 8310
-    },
-    {
-      "epoch": 0.3854891052387575,
-      "grad_norm": 3.9172377586364746,
-      "learning_rate": 1.5432063802097866e-05,
-      "loss": 0.8802,
-      "step": 8315
-    },
-    {
-      "epoch": 0.3857209086694483,
-      "grad_norm": 4.3555498123168945,
-      "learning_rate": 1.5425268436026736e-05,
-      "loss": 0.8698,
-      "step": 8320
-    },
-    {
-      "epoch": 0.3859527121001391,
-      "grad_norm": 4.431722640991211,
-      "learning_rate": 1.541846951793356e-05,
-      "loss": 1.0022,
-      "step": 8325
-    },
-    {
-      "epoch": 0.38618451553082983,
-      "grad_norm": 4.228847026824951,
-      "learning_rate": 1.541166705226971e-05,
-      "loss": 1.0159,
-      "step": 8330
-    },
-    {
-      "epoch": 0.3864163189615206,
-      "grad_norm": 3.549771547317505,
-      "learning_rate": 1.5404861043488883e-05,
-      "loss": 0.8531,
-      "step": 8335
-    },
-    {
-      "epoch": 0.3866481223922114,
-      "grad_norm": 4.803309917449951,
-      "learning_rate": 1.5398051496047105e-05,
-      "loss": 1.0774,
-      "step": 8340
-    },
-    {
-      "epoch": 0.3868799258229022,
-      "grad_norm": 3.5839805603027344,
-      "learning_rate": 1.5391238414402704e-05,
-      "loss": 1.0094,
-      "step": 8345
-    },
-    {
-      "epoch": 0.38711172925359294,
-      "grad_norm": 3.267014980316162,
-      "learning_rate": 1.5384421803016337e-05,
-      "loss": 0.9279,
-      "step": 8350
-    },
-    {
-      "epoch": 0.38734353268428373,
-      "grad_norm": 4.285528659820557,
-      "learning_rate": 1.5377601666350953e-05,
-      "loss": 0.9292,
-      "step": 8355
-    },
-    {
-      "epoch": 0.3875753361149745,
-      "grad_norm": 4.261288166046143,
-      "learning_rate": 1.537077800887182e-05,
-      "loss": 0.9634,
-      "step": 8360
-    },
-    {
-      "epoch": 0.38780713954566526,
-      "grad_norm": 4.2175517082214355,
-      "learning_rate": 1.5363950835046524e-05,
-      "loss": 0.94,
-      "step": 8365
-    },
-    {
-      "epoch": 0.38803894297635605,
-      "grad_norm": 4.262879371643066,
-      "learning_rate": 1.5357120149344923e-05,
-      "loss": 1.0298,
-      "step": 8370
-    },
-    {
-      "epoch": 0.38827074640704684,
-      "grad_norm": 3.9581894874572754,
-      "learning_rate": 1.5350285956239204e-05,
-      "loss": 0.8537,
-      "step": 8375
-    },
-    {
-      "epoch": 0.3885025498377376,
-      "grad_norm": 3.6455395221710205,
-      "learning_rate": 1.5343448260203833e-05,
-      "loss": 0.8997,
-      "step": 8380
-    },
-    {
-      "epoch": 0.38873435326842837,
-      "grad_norm": 3.51712703704834,
-      "learning_rate": 1.533660706571557e-05,
-      "loss": 0.9503,
-      "step": 8385
-    },
-    {
-      "epoch": 0.38896615669911916,
-      "grad_norm": 3.7901172637939453,
-      "learning_rate": 1.5329762377253482e-05,
-      "loss": 0.887,
-      "step": 8390
-    },
-    {
-      "epoch": 0.3891979601298099,
-      "grad_norm": 3.1369612216949463,
-      "learning_rate": 1.5322914199298897e-05,
-      "loss": 0.8935,
-      "step": 8395
-    },
-    {
-      "epoch": 0.3894297635605007,
-      "grad_norm": 3.3814847469329834,
-      "learning_rate": 1.5316062536335448e-05,
-      "loss": 0.8281,
-      "step": 8400
-    },
-    {
-      "epoch": 0.3894297635605007,
-      "eval_loss": 0.9663224816322327,
-      "eval_runtime": 11.2673,
-      "eval_samples_per_second": 11.272,
-      "eval_steps_per_second": 11.272,
-      "step": 8400
-    },
-    {
-      "epoch": 0.3896615669911915,
-      "grad_norm": 3.347869396209717,
-      "learning_rate": 1.5309207392849044e-05,
-      "loss": 1.0086,
-      "step": 8405
-    },
-    {
-      "epoch": 0.38989337042188227,
-      "grad_norm": 4.276888847351074,
-      "learning_rate": 1.5302348773327876e-05,
-      "loss": 1.1375,
-      "step": 8410
-    },
-    {
-      "epoch": 0.390125173852573,
-      "grad_norm": 4.132003307342529,
-      "learning_rate": 1.5295486682262394e-05,
-      "loss": 1.0415,
-      "step": 8415
-    },
-    {
-      "epoch": 0.3903569772832638,
-      "grad_norm": 3.471198320388794,
-      "learning_rate": 1.5288621124145352e-05,
-      "loss": 0.8592,
-      "step": 8420
-    },
-    {
-      "epoch": 0.3905887807139546,
-      "grad_norm": 4.1366448402404785,
-      "learning_rate": 1.528175210347174e-05,
-      "loss": 1.0408,
-      "step": 8425
-    },
-    {
-      "epoch": 0.3908205841446453,
-      "grad_norm": 3.8777146339416504,
-      "learning_rate": 1.527487962473884e-05,
-      "loss": 0.9605,
-      "step": 8430
-    },
-    {
-      "epoch": 0.3910523875753361,
-      "grad_norm": 4.086522579193115,
-      "learning_rate": 1.526800369244619e-05,
-      "loss": 1.0024,
-      "step": 8435
-    },
-    {
-      "epoch": 0.3912841910060269,
-      "grad_norm": 3.3036468029022217,
-      "learning_rate": 1.526112431109558e-05,
-      "loss": 0.9449,
-      "step": 8440
-    },
-    {
-      "epoch": 0.39151599443671764,
-      "grad_norm": 3.620948314666748,
-      "learning_rate": 1.525424148519107e-05,
-      "loss": 0.8556,
-      "step": 8445
-    },
-    {
-      "epoch": 0.39174779786740843,
-      "grad_norm": 4.023210048675537,
-      "learning_rate": 1.5247355219238977e-05,
-      "loss": 1.0359,
-      "step": 8450
-    },
-    {
-      "epoch": 0.3919796012980992,
-      "grad_norm": 3.1166470050811768,
-      "learning_rate": 1.5240465517747857e-05,
-      "loss": 0.8803,
-      "step": 8455
-    },
-    {
-      "epoch": 0.39221140472878996,
-      "grad_norm": 3.456226110458374,
-      "learning_rate": 1.5233572385228527e-05,
-      "loss": 0.9047,
-      "step": 8460
-    },
-    {
-      "epoch": 0.39244320815948075,
-      "grad_norm": 6.233022689819336,
-      "learning_rate": 1.5226675826194046e-05,
-      "loss": 0.9472,
-      "step": 8465
-    },
-    {
-      "epoch": 0.39267501159017154,
-      "grad_norm": 4.138795375823975,
-      "learning_rate": 1.5219775845159716e-05,
-      "loss": 1.0286,
-      "step": 8470
-    },
-    {
-      "epoch": 0.39290681502086233,
-      "grad_norm": 3.463423728942871,
-      "learning_rate": 1.5212872446643082e-05,
-      "loss": 1.0471,
-      "step": 8475
-    },
-    {
-      "epoch": 0.39313861845155307,
-      "grad_norm": 3.417653799057007,
-      "learning_rate": 1.5205965635163921e-05,
-      "loss": 0.7668,
-      "step": 8480
-    },
-    {
-      "epoch": 0.39337042188224386,
-      "grad_norm": 3.8116729259490967,
-      "learning_rate": 1.5199055415244253e-05,
-      "loss": 0.9211,
-      "step": 8485
-    },
-    {
-      "epoch": 0.39360222531293465,
-      "grad_norm": 3.489842176437378,
-      "learning_rate": 1.519214179140832e-05,
-      "loss": 0.8637,
-      "step": 8490
-    },
-    {
-      "epoch": 0.3938340287436254,
-      "grad_norm": 4.423672676086426,
-      "learning_rate": 1.51852247681826e-05,
-      "loss": 1.0781,
-      "step": 8495
-    },
-    {
-      "epoch": 0.3940658321743162,
-      "grad_norm": 3.542827606201172,
-      "learning_rate": 1.5178304350095796e-05,
-      "loss": 0.9339,
-      "step": 8500
-    },
-    {
-      "epoch": 0.3940658321743162,
-      "eval_loss": 0.9644456505775452,
-      "eval_runtime": 11.2654,
-      "eval_samples_per_second": 11.274,
-      "eval_steps_per_second": 11.274,
-      "step": 8500
-    },
-    {
-      "epoch": 0.39429763560500697,
-      "grad_norm": 3.8117423057556152,
-      "learning_rate": 1.5171380541678829e-05,
-      "loss": 0.8443,
-      "step": 8505
-    },
-    {
-      "epoch": 0.3945294390356977,
-      "grad_norm": 3.482234001159668,
-      "learning_rate": 1.516445334746484e-05,
-      "loss": 0.8924,
-      "step": 8510
-    },
-    {
-      "epoch": 0.3947612424663885,
-      "grad_norm": 3.5414836406707764,
-      "learning_rate": 1.5157522771989195e-05,
-      "loss": 1.0507,
-      "step": 8515
-    },
-    {
-      "epoch": 0.3949930458970793,
-      "grad_norm": 2.974254846572876,
-      "learning_rate": 1.5150588819789467e-05,
-      "loss": 0.8076,
-      "step": 8520
-    },
-    {
-      "epoch": 0.3952248493277701,
-      "grad_norm": 3.456716775894165,
-      "learning_rate": 1.5143651495405439e-05,
-      "loss": 0.9779,
-      "step": 8525
-    },
-    {
-      "epoch": 0.3954566527584608,
-      "grad_norm": 3.6141698360443115,
-      "learning_rate": 1.5136710803379103e-05,
-      "loss": 0.9303,
-      "step": 8530
-    },
-    {
-      "epoch": 0.3956884561891516,
-      "grad_norm": 3.409841537475586,
-      "learning_rate": 1.5129766748254656e-05,
-      "loss": 0.8588,
-      "step": 8535
-    },
-    {
-      "epoch": 0.3959202596198424,
-      "grad_norm": 4.912245750427246,
-      "learning_rate": 1.5122819334578498e-05,
-      "loss": 0.8972,
-      "step": 8540
-    },
-    {
-      "epoch": 0.39615206305053313,
-      "grad_norm": 3.854921340942383,
-      "learning_rate": 1.511586856689923e-05,
-      "loss": 1.1203,
-      "step": 8545
-    },
-    {
-      "epoch": 0.3963838664812239,
-      "grad_norm": 4.43947696685791,
-      "learning_rate": 1.510891444976764e-05,
-      "loss": 0.9963,
-      "step": 8550
-    },
-    {
-      "epoch": 0.3966156699119147,
-      "grad_norm": 3.782301902770996,
-      "learning_rate": 1.5101956987736724e-05,
-      "loss": 1.0837,
-      "step": 8555
-    },
-    {
-      "epoch": 0.39684747334260545,
-      "grad_norm": 3.896822214126587,
-      "learning_rate": 1.509499618536165e-05,
-      "loss": 0.8543,
-      "step": 8560
-    },
-    {
-      "epoch": 0.39707927677329624,
-      "grad_norm": 3.10201096534729,
-      "learning_rate": 1.5088032047199787e-05,
-      "loss": 0.8756,
-      "step": 8565
-    },
-    {
-      "epoch": 0.39731108020398703,
-      "grad_norm": 4.653040885925293,
-      "learning_rate": 1.508106457781069e-05,
-      "loss": 0.9539,
-      "step": 8570
-    },
-    {
-      "epoch": 0.39754288363467777,
-      "grad_norm": 3.215571880340576,
-      "learning_rate": 1.507409378175607e-05,
-      "loss": 0.9654,
-      "step": 8575
-    },
-    {
-      "epoch": 0.39777468706536856,
-      "grad_norm": 3.752126932144165,
-      "learning_rate": 1.506711966359985e-05,
-      "loss": 0.9448,
-      "step": 8580
-    },
-    {
-      "epoch": 0.39800649049605935,
-      "grad_norm": 3.5548033714294434,
-      "learning_rate": 1.5060142227908106e-05,
-      "loss": 0.7547,
-      "step": 8585
-    },
-    {
-      "epoch": 0.39823829392675014,
-      "grad_norm": 2.9906864166259766,
-      "learning_rate": 1.5053161479249092e-05,
-      "loss": 0.8918,
-      "step": 8590
-    },
-    {
-      "epoch": 0.3984700973574409,
-      "grad_norm": 3.0783543586730957,
-      "learning_rate": 1.5046177422193234e-05,
-      "loss": 0.8722,
-      "step": 8595
-    },
-    {
-      "epoch": 0.39870190078813167,
-      "grad_norm": 3.2667911052703857,
-      "learning_rate": 1.5039190061313118e-05,
-      "loss": 0.9629,
-      "step": 8600
-    },
-    {
-      "epoch": 0.39870190078813167,
-      "eval_loss": 0.9621018767356873,
-      "eval_runtime": 11.278,
-      "eval_samples_per_second": 11.261,
-      "eval_steps_per_second": 11.261,
-      "step": 8600
-    },
-    {
-      "epoch": 0.39893370421882246,
-      "grad_norm": 3.2560768127441406,
-      "learning_rate": 1.50321994011835e-05,
-      "loss": 0.8151,
-      "step": 8605
-    },
-    {
-      "epoch": 0.3991655076495132,
-      "grad_norm": 3.382565498352051,
-      "learning_rate": 1.5025205446381286e-05,
-      "loss": 1.0409,
-      "step": 8610
-    },
-    {
-      "epoch": 0.399397311080204,
-      "grad_norm": 3.8526535034179688,
-      "learning_rate": 1.5018208201485555e-05,
-      "loss": 0.8708,
-      "step": 8615
-    },
-    {
-      "epoch": 0.3996291145108948,
-      "grad_norm": 3.978084087371826,
-      "learning_rate": 1.5011207671077523e-05,
-      "loss": 0.9014,
-      "step": 8620
-    },
-    {
-      "epoch": 0.3998609179415855,
-      "grad_norm": 3.4769928455352783,
-      "learning_rate": 1.5004203859740569e-05,
-      "loss": 0.9161,
-      "step": 8625
-    },
-    {
-      "epoch": 0.4000927213722763,
-      "grad_norm": 3.813260078430176,
-      "learning_rate": 1.4997196772060215e-05,
-      "loss": 0.9638,
-      "step": 8630
-    },
-    {
-      "epoch": 0.4003245248029671,
-      "grad_norm": 4.105741500854492,
-      "learning_rate": 1.4990186412624126e-05,
-      "loss": 1.0067,
-      "step": 8635
-    },
-    {
-      "epoch": 0.40055632823365783,
-      "grad_norm": 3.5164148807525635,
-      "learning_rate": 1.4983172786022115e-05,
-      "loss": 0.9244,
-      "step": 8640
-    },
-    {
-      "epoch": 0.4007881316643486,
-      "grad_norm": 3.283759593963623,
-      "learning_rate": 1.4976155896846129e-05,
-      "loss": 0.8354,
-      "step": 8645
-    },
-    {
-      "epoch": 0.4010199350950394,
-      "grad_norm": 4.097290992736816,
-      "learning_rate": 1.4969135749690259e-05,
-      "loss": 0.8365,
-      "step": 8650
-    },
-    {
-      "epoch": 0.4012517385257302,
-      "grad_norm": 4.539755344390869,
-      "learning_rate": 1.4962112349150714e-05,
-      "loss": 1.0438,
-      "step": 8655
-    },
-    {
-      "epoch": 0.40148354195642094,
-      "grad_norm": 3.291273355484009,
-      "learning_rate": 1.4955085699825849e-05,
-      "loss": 0.8505,
-      "step": 8660
-    },
-    {
-      "epoch": 0.40171534538711173,
-      "grad_norm": 3.5600454807281494,
-      "learning_rate": 1.4948055806316138e-05,
-      "loss": 0.9365,
-      "step": 8665
-    },
-    {
-      "epoch": 0.4019471488178025,
-      "grad_norm": 3.5624420642852783,
-      "learning_rate": 1.4941022673224178e-05,
-      "loss": 0.9191,
-      "step": 8670
-    },
-    {
-      "epoch": 0.40217895224849326,
-      "grad_norm": 3.605802297592163,
-      "learning_rate": 1.4933986305154691e-05,
-      "loss": 1.0921,
-      "step": 8675
-    },
-    {
-      "epoch": 0.40241075567918405,
-      "grad_norm": 3.824154853820801,
-      "learning_rate": 1.4926946706714516e-05,
-      "loss": 0.9617,
-      "step": 8680
-    },
-    {
-      "epoch": 0.40264255910987484,
-      "grad_norm": 3.4327552318573,
-      "learning_rate": 1.4919903882512605e-05,
-      "loss": 0.8381,
-      "step": 8685
-    },
-    {
-      "epoch": 0.4028743625405656,
-      "grad_norm": 3.498324394226074,
-      "learning_rate": 1.4912857837160023e-05,
-      "loss": 0.8875,
-      "step": 8690
-    },
-    {
-      "epoch": 0.40310616597125637,
-      "grad_norm": 3.320936918258667,
-      "learning_rate": 1.4905808575269942e-05,
-      "loss": 0.7844,
-      "step": 8695
-    },
-    {
-      "epoch": 0.40333796940194716,
-      "grad_norm": 4.480831146240234,
-      "learning_rate": 1.4898756101457652e-05,
-      "loss": 0.902,
-      "step": 8700
-    },
-    {
-      "epoch": 0.40333796940194716,
-      "eval_loss": 0.9607927799224854,
-      "eval_runtime": 11.2715,
-      "eval_samples_per_second": 11.267,
-      "eval_steps_per_second": 11.267,
-      "step": 8700
-    },
-    {
-      "epoch": 0.4035697728326379,
-      "grad_norm": 3.961334466934204,
-      "learning_rate": 1.4891700420340522e-05,
-      "loss": 1.0543,
-      "step": 8705
-    },
-    {
-      "epoch": 0.4038015762633287,
-      "grad_norm": 9.081817626953125,
-      "learning_rate": 1.4884641536538045e-05,
-      "loss": 1.1713,
-      "step": 8710
-    },
-    {
-      "epoch": 0.4040333796940195,
-      "grad_norm": 4.37900447845459,
-      "learning_rate": 1.48775794546718e-05,
-      "loss": 0.9349,
-      "step": 8715
-    },
-    {
-      "epoch": 0.40426518312471027,
-      "grad_norm": 4.0761399269104,
-      "learning_rate": 1.4870514179365453e-05,
-      "loss": 0.9451,
-      "step": 8720
-    },
-    {
-      "epoch": 0.404496986555401,
-      "grad_norm": 3.5387260913848877,
-      "learning_rate": 1.486344571524478e-05,
-      "loss": 0.8872,
-      "step": 8725
-    },
-    {
-      "epoch": 0.4047287899860918,
-      "grad_norm": 3.9236161708831787,
-      "learning_rate": 1.4856374066937622e-05,
-      "loss": 0.9091,
-      "step": 8730
-    },
-    {
-      "epoch": 0.4049605934167826,
-      "grad_norm": 3.637903928756714,
-      "learning_rate": 1.4849299239073924e-05,
-      "loss": 0.932,
-      "step": 8735
-    },
-    {
-      "epoch": 0.4051923968474733,
-      "grad_norm": 3.5043904781341553,
-      "learning_rate": 1.4842221236285701e-05,
-      "loss": 0.9343,
-      "step": 8740
-    },
-    {
-      "epoch": 0.4054242002781641,
-      "grad_norm": 4.224621295928955,
-      "learning_rate": 1.483514006320705e-05,
-      "loss": 0.8623,
-      "step": 8745
-    },
-    {
-      "epoch": 0.4056560037088549,
-      "grad_norm": 3.410630226135254,
-      "learning_rate": 1.4828055724474144e-05,
-      "loss": 0.8286,
-      "step": 8750
-    },
-    {
-      "epoch": 0.40588780713954564,
-      "grad_norm": 4.416448593139648,
-      "learning_rate": 1.4820968224725229e-05,
-      "loss": 0.93,
-      "step": 8755
-    },
-    {
-      "epoch": 0.40611961057023643,
-      "grad_norm": 3.3819997310638428,
-      "learning_rate": 1.4813877568600625e-05,
-      "loss": 0.9855,
-      "step": 8760
-    },
-    {
-      "epoch": 0.4063514140009272,
-      "grad_norm": 3.6573407649993896,
-      "learning_rate": 1.4806783760742703e-05,
-      "loss": 0.9538,
-      "step": 8765
-    },
-    {
-      "epoch": 0.40658321743161796,
-      "grad_norm": 3.8681700229644775,
-      "learning_rate": 1.4799686805795919e-05,
-      "loss": 1.0216,
-      "step": 8770
-    },
-    {
-      "epoch": 0.40681502086230875,
-      "grad_norm": 3.9974756240844727,
-      "learning_rate": 1.4792586708406776e-05,
-      "loss": 0.9717,
-      "step": 8775
-    },
-    {
-      "epoch": 0.40704682429299954,
-      "grad_norm": 4.307816028594971,
-      "learning_rate": 1.4785483473223829e-05,
-      "loss": 0.9331,
-      "step": 8780
-    },
-    {
-      "epoch": 0.40727862772369033,
-      "grad_norm": 4.006975173950195,
-      "learning_rate": 1.4778377104897703e-05,
-      "loss": 1.0737,
-      "step": 8785
-    },
-    {
-      "epoch": 0.40751043115438107,
-      "grad_norm": 3.7386765480041504,
-      "learning_rate": 1.4771267608081068e-05,
-      "loss": 0.98,
-      "step": 8790
-    },
-    {
-      "epoch": 0.40774223458507186,
-      "grad_norm": 3.2965736389160156,
-      "learning_rate": 1.4764154987428639e-05,
-      "loss": 0.9451,
-      "step": 8795
-    },
-    {
-      "epoch": 0.40797403801576265,
-      "grad_norm": 3.6365609169006348,
-      "learning_rate": 1.4757039247597176e-05,
-      "loss": 0.8254,
-      "step": 8800
-    },
-    {
-      "epoch": 0.40797403801576265,
-      "eval_loss": 0.9614850878715515,
-      "eval_runtime": 11.2707,
-      "eval_samples_per_second": 11.268,
-      "eval_steps_per_second": 11.268,
-      "step": 8800
-    },
-    {
-      "epoch": 0.4082058414464534,
-      "grad_norm": 3.285236358642578,
-      "learning_rate": 1.4749920393245485e-05,
-      "loss": 1.0057,
-      "step": 8805
-    },
-    {
-      "epoch": 0.4084376448771442,
-      "grad_norm": 3.88112473487854,
-      "learning_rate": 1.4742798429034409e-05,
-      "loss": 0.8818,
-      "step": 8810
-    },
-    {
-      "epoch": 0.40866944830783497,
-      "grad_norm": 3.2408924102783203,
-      "learning_rate": 1.4735673359626828e-05,
-      "loss": 0.8564,
-      "step": 8815
-    },
-    {
-      "epoch": 0.4089012517385257,
-      "grad_norm": 5.373213291168213,
-      "learning_rate": 1.4728545189687658e-05,
-      "loss": 1.0474,
-      "step": 8820
-    },
-    {
-      "epoch": 0.4091330551692165,
-      "grad_norm": 3.371821880340576,
-      "learning_rate": 1.4721413923883838e-05,
-      "loss": 0.8351,
-      "step": 8825
-    },
-    {
-      "epoch": 0.4093648585999073,
-      "grad_norm": 3.3529322147369385,
-      "learning_rate": 1.4714279566884335e-05,
-      "loss": 0.869,
-      "step": 8830
-    },
-    {
-      "epoch": 0.4095966620305981,
-      "grad_norm": 4.656132698059082,
-      "learning_rate": 1.4707142123360149e-05,
-      "loss": 1.1344,
-      "step": 8835
-    },
-    {
-      "epoch": 0.4098284654612888,
-      "grad_norm": 4.094824314117432,
-      "learning_rate": 1.4700001597984291e-05,
-      "loss": 0.9209,
-      "step": 8840
-    },
-    {
-      "epoch": 0.4100602688919796,
-      "grad_norm": 3.9256341457366943,
-      "learning_rate": 1.469285799543179e-05,
-      "loss": 0.8983,
-      "step": 8845
-    },
-    {
-      "epoch": 0.4102920723226704,
-      "grad_norm": 3.937293529510498,
-      "learning_rate": 1.4685711320379698e-05,
-      "loss": 0.9011,
-      "step": 8850
-    },
-    {
-      "epoch": 0.41052387575336113,
-      "grad_norm": 3.690925359725952,
-      "learning_rate": 1.4678561577507074e-05,
-      "loss": 0.9901,
-      "step": 8855
-    },
-    {
-      "epoch": 0.4107556791840519,
-      "grad_norm": 4.106996536254883,
-      "learning_rate": 1.467140877149498e-05,
-      "loss": 0.9249,
-      "step": 8860
-    },
-    {
-      "epoch": 0.4109874826147427,
-      "grad_norm": 3.8499739170074463,
-      "learning_rate": 1.4664252907026488e-05,
-      "loss": 1.0555,
-      "step": 8865
-    },
-    {
-      "epoch": 0.41121928604543345,
-      "grad_norm": 4.457788944244385,
-      "learning_rate": 1.4657093988786677e-05,
-      "loss": 0.9393,
-      "step": 8870
-    },
-    {
-      "epoch": 0.41145108947612424,
-      "grad_norm": 3.177001476287842,
-      "learning_rate": 1.4649932021462616e-05,
-      "loss": 0.8521,
-      "step": 8875
-    },
-    {
-      "epoch": 0.41168289290681503,
-      "grad_norm": 4.380415916442871,
-      "learning_rate": 1.4642767009743384e-05,
-      "loss": 0.9663,
-      "step": 8880
-    },
-    {
-      "epoch": 0.41191469633750577,
-      "grad_norm": 4.102474212646484,
-      "learning_rate": 1.463559895832004e-05,
-      "loss": 0.9828,
-      "step": 8885
-    },
-    {
-      "epoch": 0.41214649976819656,
-      "grad_norm": 7.0587663650512695,
-      "learning_rate": 1.4628427871885632e-05,
-      "loss": 0.9628,
-      "step": 8890
-    },
-    {
-      "epoch": 0.41237830319888735,
-      "grad_norm": 3.6423287391662598,
-      "learning_rate": 1.4621253755135207e-05,
-      "loss": 0.7434,
-      "step": 8895
-    },
-    {
-      "epoch": 0.41261010662957814,
-      "grad_norm": 4.276022434234619,
-      "learning_rate": 1.4614076612765788e-05,
-      "loss": 0.8113,
-      "step": 8900
-    },
-    {
-      "epoch": 0.41261010662957814,
-      "eval_loss": 0.9591353535652161,
-      "eval_runtime": 11.2682,
-      "eval_samples_per_second": 11.271,
-      "eval_steps_per_second": 11.271,
-      "step": 8900
-    },
-    {
-      "epoch": 0.4128419100602689,
-      "grad_norm": 3.499518871307373,
-      "learning_rate": 1.4606896449476387e-05,
-      "loss": 0.8564,
-      "step": 8905
-    },
-    {
-      "epoch": 0.41307371349095967,
-      "grad_norm": 3.5929625034332275,
-      "learning_rate": 1.4599713269967976e-05,
-      "loss": 1.0022,
-      "step": 8910
-    },
-    {
-      "epoch": 0.41330551692165046,
-      "grad_norm": 4.23135232925415,
-      "learning_rate": 1.4592527078943523e-05,
-      "loss": 1.1004,
-      "step": 8915
-    },
-    {
-      "epoch": 0.4135373203523412,
-      "grad_norm": 3.6771299839019775,
-      "learning_rate": 1.4585337881107956e-05,
-      "loss": 1.0565,
-      "step": 8920
-    },
-    {
-      "epoch": 0.413769123783032,
-      "grad_norm": 4.86899471282959,
-      "learning_rate": 1.457814568116817e-05,
-      "loss": 0.9799,
-      "step": 8925
-    },
-    {
-      "epoch": 0.4140009272137228,
-      "grad_norm": 3.4616496562957764,
-      "learning_rate": 1.4570950483833036e-05,
-      "loss": 0.9112,
-      "step": 8930
-    },
-    {
-      "epoch": 0.4142327306444135,
-      "grad_norm": 3.190037250518799,
-      "learning_rate": 1.4563752293813371e-05,
-      "loss": 0.9892,
-      "step": 8935
-    },
-    {
-      "epoch": 0.4144645340751043,
-      "grad_norm": 4.035621166229248,
-      "learning_rate": 1.455655111582197e-05,
-      "loss": 0.8924,
-      "step": 8940
-    },
-    {
-      "epoch": 0.4146963375057951,
-      "grad_norm": 3.7359254360198975,
-      "learning_rate": 1.4549346954573574e-05,
-      "loss": 0.9515,
-      "step": 8945
-    },
-    {
-      "epoch": 0.41492814093648583,
-      "grad_norm": 4.1890997886657715,
-      "learning_rate": 1.4542139814784874e-05,
-      "loss": 0.913,
-      "step": 8950
-    },
-    {
-      "epoch": 0.4151599443671766,
-      "grad_norm": 4.180643558502197,
-      "learning_rate": 1.453492970117452e-05,
-      "loss": 0.8282,
-      "step": 8955
-    },
-    {
-      "epoch": 0.4153917477978674,
-      "grad_norm": 3.5779945850372314,
-      "learning_rate": 1.45277166184631e-05,
-      "loss": 0.9543,
-      "step": 8960
-    },
-    {
-      "epoch": 0.4156235512285582,
-      "grad_norm": 3.7106773853302,
-      "learning_rate": 1.4520500571373157e-05,
-      "loss": 1.0066,
-      "step": 8965
-    },
-    {
-      "epoch": 0.41585535465924894,
-      "grad_norm": 3.7494590282440186,
-      "learning_rate": 1.4513281564629163e-05,
-      "loss": 0.9777,
-      "step": 8970
-    },
-    {
-      "epoch": 0.41608715808993973,
-      "grad_norm": 3.517127513885498,
-      "learning_rate": 1.4506059602957537e-05,
-      "loss": 0.9611,
-      "step": 8975
-    },
-    {
-      "epoch": 0.4163189615206305,
-      "grad_norm": 4.559821605682373,
-      "learning_rate": 1.449883469108663e-05,
-      "loss": 0.9203,
-      "step": 8980
-    },
-    {
-      "epoch": 0.41655076495132126,
-      "grad_norm": 3.6281211376190186,
-      "learning_rate": 1.4491606833746714e-05,
-      "loss": 0.9143,
-      "step": 8985
-    },
-    {
-      "epoch": 0.41678256838201205,
-      "grad_norm": 4.122176170349121,
-      "learning_rate": 1.4484376035670008e-05,
-      "loss": 1.0841,
-      "step": 8990
-    },
-    {
-      "epoch": 0.41701437181270284,
-      "grad_norm": 4.669403553009033,
-      "learning_rate": 1.4477142301590647e-05,
-      "loss": 0.7889,
-      "step": 8995
-    },
-    {
-      "epoch": 0.4172461752433936,
-      "grad_norm": 3.9678564071655273,
-      "learning_rate": 1.4469905636244686e-05,
-      "loss": 0.8821,
-      "step": 9000
-    },
-    {
-      "epoch": 0.4172461752433936,
-      "eval_loss": 0.9594786167144775,
-      "eval_runtime": 11.271,
-      "eval_samples_per_second": 11.268,
-      "eval_steps_per_second": 11.268,
-      "step": 9000
-    },
-    {
-      "epoch": 0.41747797867408437,
-      "grad_norm": 3.803729295730591,
-      "learning_rate": 1.44626660443701e-05,
-      "loss": 0.9246,
-      "step": 9005
-    },
-    {
-      "epoch": 0.41770978210477516,
-      "grad_norm": 4.379030227661133,
-      "learning_rate": 1.4455423530706782e-05,
-      "loss": 1.1268,
-      "step": 9010
-    },
-    {
-      "epoch": 0.4179415855354659,
-      "grad_norm": 3.691082715988159,
-      "learning_rate": 1.4448178099996539e-05,
-      "loss": 0.9361,
-      "step": 9015
-    },
-    {
-      "epoch": 0.4181733889661567,
-      "grad_norm": 3.965627908706665,
-      "learning_rate": 1.4440929756983086e-05,
-      "loss": 0.997,
-      "step": 9020
-    },
-    {
-      "epoch": 0.4184051923968475,
-      "grad_norm": 4.264866352081299,
-      "learning_rate": 1.443367850641205e-05,
-      "loss": 0.8876,
-      "step": 9025
-    },
-    {
-      "epoch": 0.41863699582753827,
-      "grad_norm": 4.311303615570068,
-      "learning_rate": 1.4426424353030947e-05,
-      "loss": 0.8455,
-      "step": 9030
-    },
-    {
-      "epoch": 0.418868799258229,
-      "grad_norm": 3.237496852874756,
-      "learning_rate": 1.441916730158921e-05,
-      "loss": 0.8728,
-      "step": 9035
-    },
-    {
-      "epoch": 0.4191006026889198,
-      "grad_norm": 2.8361425399780273,
-      "learning_rate": 1.4411907356838165e-05,
-      "loss": 0.7104,
-      "step": 9040
-    },
-    {
-      "epoch": 0.4193324061196106,
-      "grad_norm": 4.360476493835449,
-      "learning_rate": 1.440464452353102e-05,
-      "loss": 0.8732,
-      "step": 9045
-    },
-    {
-      "epoch": 0.4195642095503013,
-      "grad_norm": 3.61860728263855,
-      "learning_rate": 1.4397378806422895e-05,
-      "loss": 0.8852,
-      "step": 9050
-    },
-    {
-      "epoch": 0.4197960129809921,
-      "grad_norm": 4.051429271697998,
-      "learning_rate": 1.439011021027078e-05,
-      "loss": 1.1358,
-      "step": 9055
-    },
-    {
-      "epoch": 0.4200278164116829,
-      "grad_norm": 3.6325385570526123,
-      "learning_rate": 1.4382838739833562e-05,
-      "loss": 0.8702,
-      "step": 9060
-    },
-    {
-      "epoch": 0.42025961984237364,
-      "grad_norm": 3.3029043674468994,
-      "learning_rate": 1.4375564399872005e-05,
-      "loss": 0.8015,
-      "step": 9065
-    },
-    {
-      "epoch": 0.42049142327306444,
-      "grad_norm": 4.105788707733154,
-      "learning_rate": 1.4368287195148746e-05,
-      "loss": 0.9538,
-      "step": 9070
-    },
-    {
-      "epoch": 0.4207232267037552,
-      "grad_norm": 3.321793556213379,
-      "learning_rate": 1.4361007130428311e-05,
-      "loss": 0.9167,
-      "step": 9075
-    },
-    {
-      "epoch": 0.42095503013444596,
-      "grad_norm": 4.301546096801758,
-      "learning_rate": 1.4353724210477084e-05,
-      "loss": 0.9535,
-      "step": 9080
-    },
-    {
-      "epoch": 0.42118683356513675,
-      "grad_norm": 3.8021814823150635,
-      "learning_rate": 1.4346438440063333e-05,
-      "loss": 0.876,
-      "step": 9085
-    },
-    {
-      "epoch": 0.42141863699582754,
-      "grad_norm": 3.5097808837890625,
-      "learning_rate": 1.433914982395718e-05,
-      "loss": 0.9506,
-      "step": 9090
-    },
-    {
-      "epoch": 0.42165044042651834,
-      "grad_norm": 3.8643198013305664,
-      "learning_rate": 1.4331858366930611e-05,
-      "loss": 0.9854,
-      "step": 9095
-    },
-    {
-      "epoch": 0.42188224385720907,
-      "grad_norm": 3.721323251724243,
-      "learning_rate": 1.4324564073757486e-05,
-      "loss": 0.8642,
-      "step": 9100
-    },
-    {
-      "epoch": 0.42188224385720907,
-      "eval_loss": 0.9581558108329773,
-      "eval_runtime": 11.2675,
-      "eval_samples_per_second": 11.271,
-      "eval_steps_per_second": 11.271,
-      "step": 9100
-    },
-    {
-      "epoch": 0.42211404728789986,
-      "grad_norm": 3.54001522064209,
-      "learning_rate": 1.43172669492135e-05,
-      "loss": 0.8244,
-      "step": 9105
-    },
-    {
-      "epoch": 0.42234585071859065,
-      "grad_norm": 3.663177490234375,
-      "learning_rate": 1.4309966998076222e-05,
-      "loss": 0.9079,
-      "step": 9110
-    },
-    {
-      "epoch": 0.4225776541492814,
-      "grad_norm": 3.5418756008148193,
-      "learning_rate": 1.4302664225125058e-05,
-      "loss": 0.8807,
-      "step": 9115
-    },
-    {
-      "epoch": 0.4228094575799722,
-      "grad_norm": 3.5246732234954834,
-      "learning_rate": 1.429535863514127e-05,
-      "loss": 0.7437,
-      "step": 9120
-    },
-    {
-      "epoch": 0.423041261010663,
-      "grad_norm": 3.5707225799560547,
-      "learning_rate": 1.428805023290796e-05,
-      "loss": 1.0353,
-      "step": 9125
-    },
-    {
-      "epoch": 0.4232730644413537,
-      "grad_norm": 3.847414255142212,
-      "learning_rate": 1.4280739023210067e-05,
-      "loss": 1.1087,
-      "step": 9130
-    },
-    {
-      "epoch": 0.4235048678720445,
-      "grad_norm": 4.29426383972168,
-      "learning_rate": 1.427342501083438e-05,
-      "loss": 0.7605,
-      "step": 9135
-    },
-    {
-      "epoch": 0.4237366713027353,
-      "grad_norm": 3.425589084625244,
-      "learning_rate": 1.426610820056951e-05,
-      "loss": 0.9359,
-      "step": 9140
-    },
-    {
-      "epoch": 0.4239684747334261,
-      "grad_norm": 3.226741075515747,
-      "learning_rate": 1.4258788597205913e-05,
-      "loss": 0.8194,
-      "step": 9145
-    },
-    {
-      "epoch": 0.4242002781641168,
-      "grad_norm": 3.313603401184082,
-      "learning_rate": 1.425146620553586e-05,
-      "loss": 0.8958,
-      "step": 9150
-    },
-    {
-      "epoch": 0.4244320815948076,
-      "grad_norm": 4.064041614532471,
-      "learning_rate": 1.4244141030353458e-05,
-      "loss": 1.008,
-      "step": 9155
-    },
-    {
-      "epoch": 0.4246638850254984,
-      "grad_norm": 4.129899024963379,
-      "learning_rate": 1.4236813076454632e-05,
-      "loss": 0.8858,
-      "step": 9160
-    },
-    {
-      "epoch": 0.42489568845618914,
-      "grad_norm": 3.3487892150878906,
-      "learning_rate": 1.4229482348637125e-05,
-      "loss": 0.8874,
-      "step": 9165
-    },
-    {
-      "epoch": 0.4251274918868799,
-      "grad_norm": 3.243873357772827,
-      "learning_rate": 1.4222148851700504e-05,
-      "loss": 0.8868,
-      "step": 9170
-    },
-    {
-      "epoch": 0.4253592953175707,
-      "grad_norm": 4.211264133453369,
-      "learning_rate": 1.4214812590446135e-05,
-      "loss": 0.8756,
-      "step": 9175
-    },
-    {
-      "epoch": 0.42559109874826145,
-      "grad_norm": 3.718165159225464,
-      "learning_rate": 1.4207473569677203e-05,
-      "loss": 0.9177,
-      "step": 9180
-    },
-    {
-      "epoch": 0.42582290217895225,
-      "grad_norm": 3.3769257068634033,
-      "learning_rate": 1.4200131794198706e-05,
-      "loss": 0.8602,
-      "step": 9185
-    },
-    {
-      "epoch": 0.42605470560964304,
-      "grad_norm": 3.4048376083374023,
-      "learning_rate": 1.4192787268817427e-05,
-      "loss": 0.9522,
-      "step": 9190
-    },
-    {
-      "epoch": 0.42628650904033377,
-      "grad_norm": 4.399397850036621,
-      "learning_rate": 1.418543999834197e-05,
-      "loss": 0.8503,
-      "step": 9195
-    },
-    {
-      "epoch": 0.42651831247102456,
-      "grad_norm": 3.5596532821655273,
-      "learning_rate": 1.4178089987582722e-05,
-      "loss": 0.7514,
-      "step": 9200
-    },
-    {
-      "epoch": 0.42651831247102456,
-      "eval_loss": 0.957184374332428,
-      "eval_runtime": 11.2793,
-      "eval_samples_per_second": 11.26,
-      "eval_steps_per_second": 11.26,
-      "step": 9200
-    },
-    {
-      "epoch": 0.42675011590171535,
-      "grad_norm": 3.185898780822754,
-      "learning_rate": 1.4170737241351873e-05,
-      "loss": 0.9724,
-      "step": 9205
-    },
-    {
-      "epoch": 0.42698191933240615,
-      "grad_norm": 3.8426990509033203,
-      "learning_rate": 1.4163381764463395e-05,
-      "loss": 0.841,
-      "step": 9210
-    },
-    {
-      "epoch": 0.4272137227630969,
-      "grad_norm": 3.9944119453430176,
-      "learning_rate": 1.4156023561733054e-05,
-      "loss": 0.9992,
-      "step": 9215
-    },
-    {
-      "epoch": 0.4274455261937877,
-      "grad_norm": 3.3030200004577637,
-      "learning_rate": 1.41486626379784e-05,
-      "loss": 1.078,
-      "step": 9220
-    },
-    {
-      "epoch": 0.42767732962447846,
-      "grad_norm": 4.052797317504883,
-      "learning_rate": 1.4141298998018765e-05,
-      "loss": 0.7569,
-      "step": 9225
-    },
-    {
-      "epoch": 0.4279091330551692,
-      "grad_norm": 5.390350341796875,
-      "learning_rate": 1.4133932646675258e-05,
-      "loss": 0.9488,
-      "step": 9230
-    },
-    {
-      "epoch": 0.42814093648586,
-      "grad_norm": 3.9626481533050537,
-      "learning_rate": 1.412656358877076e-05,
-      "loss": 1.0712,
-      "step": 9235
-    },
-    {
-      "epoch": 0.4283727399165508,
-      "grad_norm": 3.662187337875366,
-      "learning_rate": 1.411919182912993e-05,
-      "loss": 0.8196,
-      "step": 9240
-    },
-    {
-      "epoch": 0.4286045433472415,
-      "grad_norm": 3.5590271949768066,
-      "learning_rate": 1.4111817372579198e-05,
-      "loss": 1.1157,
-      "step": 9245
-    },
-    {
-      "epoch": 0.4288363467779323,
-      "grad_norm": 3.3557329177856445,
-      "learning_rate": 1.4104440223946742e-05,
-      "loss": 0.8617,
-      "step": 9250
-    },
-    {
-      "epoch": 0.4290681502086231,
-      "grad_norm": 4.487248420715332,
-      "learning_rate": 1.4097060388062529e-05,
-      "loss": 0.9544,
-      "step": 9255
-    },
-    {
-      "epoch": 0.42929995363931384,
-      "grad_norm": 5.409276962280273,
-      "learning_rate": 1.4089677869758266e-05,
-      "loss": 1.0187,
-      "step": 9260
-    },
-    {
-      "epoch": 0.4295317570700046,
-      "grad_norm": 3.230283737182617,
-      "learning_rate": 1.4082292673867421e-05,
-      "loss": 0.971,
-      "step": 9265
-    },
-    {
-      "epoch": 0.4297635605006954,
-      "grad_norm": 4.221251010894775,
-      "learning_rate": 1.407490480522522e-05,
-      "loss": 0.928,
-      "step": 9270
-    },
-    {
-      "epoch": 0.4299953639313862,
-      "grad_norm": 4.350439071655273,
-      "learning_rate": 1.4067514268668631e-05,
-      "loss": 1.002,
-      "step": 9275
-    },
-    {
-      "epoch": 0.43022716736207695,
-      "grad_norm": 3.369349479675293,
-      "learning_rate": 1.4060121069036379e-05,
-      "loss": 0.8569,
-      "step": 9280
-    },
-    {
-      "epoch": 0.43045897079276774,
-      "grad_norm": 3.8549163341522217,
-      "learning_rate": 1.405272521116892e-05,
-      "loss": 0.9794,
-      "step": 9285
-    },
-    {
-      "epoch": 0.43069077422345853,
-      "grad_norm": 4.107554912567139,
-      "learning_rate": 1.404532669990846e-05,
-      "loss": 0.9398,
-      "step": 9290
-    },
-    {
-      "epoch": 0.43092257765414926,
-      "grad_norm": 3.9353766441345215,
-      "learning_rate": 1.4037925540098937e-05,
-      "loss": 1.0514,
-      "step": 9295
-    },
-    {
-      "epoch": 0.43115438108484005,
-      "grad_norm": 3.31166934967041,
-      "learning_rate": 1.4030521736586025e-05,
-      "loss": 0.6717,
-      "step": 9300
-    },
-    {
-      "epoch": 0.43115438108484005,
-      "eval_loss": 0.954317033290863,
-      "eval_runtime": 11.2662,
-      "eval_samples_per_second": 11.273,
-      "eval_steps_per_second": 11.273,
-      "step": 9300
-    },
-    {
-      "epoch": 0.43138618451553085,
-      "grad_norm": 3.7155702114105225,
-      "learning_rate": 1.4023115294217135e-05,
-      "loss": 0.9786,
-      "step": 9305
-    },
-    {
-      "epoch": 0.4316179879462216,
-      "grad_norm": 3.2847890853881836,
-      "learning_rate": 1.4015706217841392e-05,
-      "loss": 0.8072,
-      "step": 9310
-    },
-    {
-      "epoch": 0.4318497913769124,
-      "grad_norm": 4.270799160003662,
-      "learning_rate": 1.400829451230966e-05,
-      "loss": 0.9943,
-      "step": 9315
-    },
-    {
-      "epoch": 0.43208159480760316,
-      "grad_norm": 3.409757614135742,
-      "learning_rate": 1.4000880182474512e-05,
-      "loss": 0.8742,
-      "step": 9320
-    },
-    {
-      "epoch": 0.4323133982382939,
-      "grad_norm": 3.892242908477783,
-      "learning_rate": 1.3993463233190245e-05,
-      "loss": 1.1398,
-      "step": 9325
-    },
-    {
-      "epoch": 0.4325452016689847,
-      "grad_norm": 4.129483222961426,
-      "learning_rate": 1.3986043669312874e-05,
-      "loss": 0.945,
-      "step": 9330
-    },
-    {
-      "epoch": 0.4327770050996755,
-      "grad_norm": 3.7721927165985107,
-      "learning_rate": 1.397862149570012e-05,
-      "loss": 1.0127,
-      "step": 9335
-    },
-    {
-      "epoch": 0.4330088085303663,
-      "grad_norm": 4.086781978607178,
-      "learning_rate": 1.397119671721142e-05,
-      "loss": 0.9437,
-      "step": 9340
-    },
-    {
-      "epoch": 0.433240611961057,
-      "grad_norm": 3.799339532852173,
-      "learning_rate": 1.3963769338707905e-05,
-      "loss": 0.8334,
-      "step": 9345
-    },
-    {
-      "epoch": 0.4334724153917478,
-      "grad_norm": 3.543330669403076,
-      "learning_rate": 1.3956339365052419e-05,
-      "loss": 0.9997,
-      "step": 9350
-    },
-    {
-      "epoch": 0.4337042188224386,
-      "grad_norm": 5.324821949005127,
-      "learning_rate": 1.3948906801109498e-05,
-      "loss": 0.8411,
-      "step": 9355
-    },
-    {
-      "epoch": 0.4339360222531293,
-      "grad_norm": 3.7983787059783936,
-      "learning_rate": 1.3941471651745382e-05,
-      "loss": 0.9135,
-      "step": 9360
-    },
-    {
-      "epoch": 0.4341678256838201,
-      "grad_norm": 3.554536819458008,
-      "learning_rate": 1.3934033921827994e-05,
-      "loss": 0.9295,
-      "step": 9365
-    },
-    {
-      "epoch": 0.4343996291145109,
-      "grad_norm": 4.4289140701293945,
-      "learning_rate": 1.3926593616226951e-05,
-      "loss": 1.0013,
-      "step": 9370
-    },
-    {
-      "epoch": 0.43463143254520165,
-      "grad_norm": 4.294192314147949,
-      "learning_rate": 1.3919150739813566e-05,
-      "loss": 1.0827,
-      "step": 9375
-    },
-    {
-      "epoch": 0.43486323597589244,
-      "grad_norm": 3.2097225189208984,
-      "learning_rate": 1.3911705297460812e-05,
-      "loss": 0.9832,
-      "step": 9380
-    },
-    {
-      "epoch": 0.43509503940658323,
-      "grad_norm": 3.6051383018493652,
-      "learning_rate": 1.3904257294043362e-05,
-      "loss": 0.9388,
-      "step": 9385
-    },
-    {
-      "epoch": 0.43532684283727396,
-      "grad_norm": 4.199221134185791,
-      "learning_rate": 1.3896806734437563e-05,
-      "loss": 0.8745,
-      "step": 9390
-    },
-    {
-      "epoch": 0.43555864626796476,
-      "grad_norm": 3.3234703540802,
-      "learning_rate": 1.3889353623521426e-05,
-      "loss": 0.9371,
-      "step": 9395
-    },
-    {
-      "epoch": 0.43579044969865555,
-      "grad_norm": 3.5717806816101074,
-      "learning_rate": 1.3881897966174642e-05,
-      "loss": 0.8621,
-      "step": 9400
-    },
-    {
-      "epoch": 0.43579044969865555,
-      "eval_loss": 0.9543979167938232,
-      "eval_runtime": 11.2671,
-      "eval_samples_per_second": 11.272,
-      "eval_steps_per_second": 11.272,
-      "step": 9400
-    },
-    {
-      "epoch": 0.43602225312934634,
-      "grad_norm": 3.5517239570617676,
-      "learning_rate": 1.3874439767278562e-05,
-      "loss": 0.9488,
-      "step": 9405
-    },
-    {
-      "epoch": 0.4362540565600371,
-      "grad_norm": 3.392153739929199,
-      "learning_rate": 1.3866979031716212e-05,
-      "loss": 0.8369,
-      "step": 9410
-    },
-    {
-      "epoch": 0.43648585999072786,
-      "grad_norm": 4.374027729034424,
-      "learning_rate": 1.3859515764372265e-05,
-      "loss": 1.0121,
-      "step": 9415
-    },
-    {
-      "epoch": 0.43671766342141866,
-      "grad_norm": 3.541436195373535,
-      "learning_rate": 1.385204997013306e-05,
-      "loss": 0.9504,
-      "step": 9420
-    },
-    {
-      "epoch": 0.4369494668521094,
-      "grad_norm": 3.709672451019287,
-      "learning_rate": 1.3844581653886587e-05,
-      "loss": 1.021,
-      "step": 9425
-    },
-    {
-      "epoch": 0.4371812702828002,
-      "grad_norm": 3.851954460144043,
-      "learning_rate": 1.383711082052249e-05,
-      "loss": 1.0206,
-      "step": 9430
-    },
-    {
-      "epoch": 0.437413073713491,
-      "grad_norm": 4.461718559265137,
-      "learning_rate": 1.3829637474932058e-05,
-      "loss": 0.9523,
-      "step": 9435
-    },
-    {
-      "epoch": 0.4376448771441817,
-      "grad_norm": 3.9250519275665283,
-      "learning_rate": 1.3822161622008233e-05,
-      "loss": 0.9426,
-      "step": 9440
-    },
-    {
-      "epoch": 0.4378766805748725,
-      "grad_norm": 3.6661217212677,
-      "learning_rate": 1.381468326664558e-05,
-      "loss": 0.9269,
-      "step": 9445
-    },
-    {
-      "epoch": 0.4381084840055633,
-      "grad_norm": 4.122355937957764,
-      "learning_rate": 1.3807202413740329e-05,
-      "loss": 0.9779,
-      "step": 9450
-    },
-    {
-      "epoch": 0.4383402874362541,
-      "grad_norm": 3.240402936935425,
-      "learning_rate": 1.3799719068190318e-05,
-      "loss": 0.8976,
-      "step": 9455
-    },
-    {
-      "epoch": 0.4385720908669448,
-      "grad_norm": 4.14400577545166,
-      "learning_rate": 1.3792233234895037e-05,
-      "loss": 0.9346,
-      "step": 9460
-    },
-    {
-      "epoch": 0.4388038942976356,
-      "grad_norm": 3.896681308746338,
-      "learning_rate": 1.3784744918755593e-05,
-      "loss": 1.093,
-      "step": 9465
-    },
-    {
-      "epoch": 0.4390356977283264,
-      "grad_norm": 3.402827262878418,
-      "learning_rate": 1.3777254124674726e-05,
-      "loss": 1.0033,
-      "step": 9470
-    },
-    {
-      "epoch": 0.43926750115901714,
-      "grad_norm": 3.659036874771118,
-      "learning_rate": 1.3769760857556795e-05,
-      "loss": 0.9373,
-      "step": 9475
-    },
-    {
-      "epoch": 0.43949930458970793,
-      "grad_norm": 4.025357723236084,
-      "learning_rate": 1.3762265122307778e-05,
-      "loss": 1.0778,
-      "step": 9480
-    },
-    {
-      "epoch": 0.4397311080203987,
-      "grad_norm": 3.3651130199432373,
-      "learning_rate": 1.375476692383527e-05,
-      "loss": 0.8191,
-      "step": 9485
-    },
-    {
-      "epoch": 0.43996291145108946,
-      "grad_norm": 3.4289755821228027,
-      "learning_rate": 1.3747266267048475e-05,
-      "loss": 0.7334,
-      "step": 9490
-    },
-    {
-      "epoch": 0.44019471488178025,
-      "grad_norm": 3.8313558101654053,
-      "learning_rate": 1.3739763156858215e-05,
-      "loss": 0.8722,
-      "step": 9495
-    },
-    {
-      "epoch": 0.44042651831247104,
-      "grad_norm": 3.7963833808898926,
-      "learning_rate": 1.3732257598176908e-05,
-      "loss": 0.832,
-      "step": 9500
-    },
-    {
-      "epoch": 0.44042651831247104,
-      "eval_loss": 0.9525100588798523,
-      "eval_runtime": 11.2607,
-      "eval_samples_per_second": 11.278,
-      "eval_steps_per_second": 11.278,
-      "step": 9500
-    },
-    {
-      "epoch": 0.4406583217431618,
-      "grad_norm": 4.198210716247559,
-      "learning_rate": 1.372474959591858e-05,
-      "loss": 1.0779,
-      "step": 9505
-    },
-    {
-      "epoch": 0.44089012517385257,
-      "grad_norm": 3.5892603397369385,
-      "learning_rate": 1.3717239154998862e-05,
-      "loss": 0.8437,
-      "step": 9510
-    },
-    {
-      "epoch": 0.44112192860454336,
-      "grad_norm": 4.606617450714111,
-      "learning_rate": 1.370972628033497e-05,
-      "loss": 1.0619,
-      "step": 9515
-    },
-    {
-      "epoch": 0.44135373203523415,
-      "grad_norm": 2.997912883758545,
-      "learning_rate": 1.3702210976845725e-05,
-      "loss": 0.8647,
-      "step": 9520
-    },
-    {
-      "epoch": 0.4415855354659249,
-      "grad_norm": 3.43404483795166,
-      "learning_rate": 1.3694693249451532e-05,
-      "loss": 0.766,
-      "step": 9525
-    },
-    {
-      "epoch": 0.4418173388966157,
-      "grad_norm": 3.4902102947235107,
-      "learning_rate": 1.3687173103074378e-05,
-      "loss": 0.8688,
-      "step": 9530
-    },
-    {
-      "epoch": 0.44204914232730647,
-      "grad_norm": 3.934156656265259,
-      "learning_rate": 1.3679650542637848e-05,
-      "loss": 0.9175,
-      "step": 9535
-    },
-    {
-      "epoch": 0.4422809457579972,
-      "grad_norm": 4.285318374633789,
-      "learning_rate": 1.3672125573067097e-05,
-      "loss": 0.918,
-      "step": 9540
-    },
-    {
-      "epoch": 0.442512749188688,
-      "grad_norm": 3.6897430419921875,
-      "learning_rate": 1.3664598199288857e-05,
-      "loss": 0.829,
-      "step": 9545
-    },
-    {
-      "epoch": 0.4427445526193788,
-      "grad_norm": 4.011903285980225,
-      "learning_rate": 1.3657068426231439e-05,
-      "loss": 1.1165,
-      "step": 9550
-    },
-    {
-      "epoch": 0.4429763560500695,
-      "grad_norm": 4.198677062988281,
-      "learning_rate": 1.3649536258824722e-05,
-      "loss": 0.8271,
-      "step": 9555
-    },
-    {
-      "epoch": 0.4432081594807603,
-      "grad_norm": 4.13018274307251,
-      "learning_rate": 1.3642001702000155e-05,
-      "loss": 0.8224,
-      "step": 9560
-    },
-    {
-      "epoch": 0.4434399629114511,
-      "grad_norm": 3.2623791694641113,
-      "learning_rate": 1.3634464760690742e-05,
-      "loss": 0.8145,
-      "step": 9565
-    },
-    {
-      "epoch": 0.44367176634214184,
-      "grad_norm": 4.026551246643066,
-      "learning_rate": 1.3626925439831067e-05,
-      "loss": 0.9373,
-      "step": 9570
-    },
-    {
-      "epoch": 0.44390356977283263,
-      "grad_norm": 4.671341896057129,
-      "learning_rate": 1.361938374435725e-05,
-      "loss": 0.9916,
-      "step": 9575
-    },
-    {
-      "epoch": 0.4441353732035234,
-      "grad_norm": 3.9089975357055664,
-      "learning_rate": 1.3611839679206986e-05,
-      "loss": 0.9412,
-      "step": 9580
-    },
-    {
-      "epoch": 0.4443671766342142,
-      "grad_norm": 4.026656150817871,
-      "learning_rate": 1.3604293249319507e-05,
-      "loss": 0.7684,
-      "step": 9585
-    },
-    {
-      "epoch": 0.44459898006490495,
-      "grad_norm": 3.824301242828369,
-      "learning_rate": 1.3596744459635595e-05,
-      "loss": 1.0024,
-      "step": 9590
-    },
-    {
-      "epoch": 0.44483078349559574,
-      "grad_norm": 3.724039316177368,
-      "learning_rate": 1.3589193315097589e-05,
-      "loss": 0.8391,
-      "step": 9595
-    },
-    {
-      "epoch": 0.44506258692628653,
-      "grad_norm": 3.2751495838165283,
-      "learning_rate": 1.358163982064935e-05,
-      "loss": 0.8337,
-      "step": 9600
-    },
-    {
-      "epoch": 0.44506258692628653,
-      "eval_loss": 0.9512506127357483,
-      "eval_runtime": 11.2694,
-      "eval_samples_per_second": 11.269,
-      "eval_steps_per_second": 11.269,
-      "step": 9600
-    },
-    {
-      "epoch": 0.44529439035697727,
-      "grad_norm": 3.5212228298187256,
-      "learning_rate": 1.3574083981236295e-05,
-      "loss": 1.0335,
-      "step": 9605
-    },
-    {
-      "epoch": 0.44552619378766806,
-      "grad_norm": 4.416726112365723,
-      "learning_rate": 1.3566525801805369e-05,
-      "loss": 0.9456,
-      "step": 9610
-    },
-    {
-      "epoch": 0.44575799721835885,
-      "grad_norm": 3.1660563945770264,
-      "learning_rate": 1.3558965287305049e-05,
-      "loss": 0.7768,
-      "step": 9615
-    },
-    {
-      "epoch": 0.4459898006490496,
-      "grad_norm": 3.384840488433838,
-      "learning_rate": 1.3551402442685342e-05,
-      "loss": 0.9298,
-      "step": 9620
-    },
-    {
-      "epoch": 0.4462216040797404,
-      "grad_norm": 3.863919496536255,
-      "learning_rate": 1.3543837272897778e-05,
-      "loss": 0.7726,
-      "step": 9625
-    },
-    {
-      "epoch": 0.44645340751043117,
-      "grad_norm": 3.7253286838531494,
-      "learning_rate": 1.3536269782895416e-05,
-      "loss": 0.8242,
-      "step": 9630
-    },
-    {
-      "epoch": 0.4466852109411219,
-      "grad_norm": 3.352973461151123,
-      "learning_rate": 1.3528699977632824e-05,
-      "loss": 0.8054,
-      "step": 9635
-    },
-    {
-      "epoch": 0.4469170143718127,
-      "grad_norm": 3.536757707595825,
-      "learning_rate": 1.3521127862066095e-05,
-      "loss": 0.7579,
-      "step": 9640
-    },
-    {
-      "epoch": 0.4471488178025035,
-      "grad_norm": 3.5800118446350098,
-      "learning_rate": 1.351355344115283e-05,
-      "loss": 0.8839,
-      "step": 9645
-    },
-    {
-      "epoch": 0.4473806212331943,
-      "grad_norm": 3.774768829345703,
-      "learning_rate": 1.3505976719852138e-05,
-      "loss": 0.8917,
-      "step": 9650
-    },
-    {
-      "epoch": 0.447612424663885,
-      "grad_norm": 3.8715434074401855,
-      "learning_rate": 1.3498397703124644e-05,
-      "loss": 0.903,
-      "step": 9655
-    },
-    {
-      "epoch": 0.4478442280945758,
-      "grad_norm": 4.232203483581543,
-      "learning_rate": 1.3490816395932456e-05,
-      "loss": 0.9655,
-      "step": 9660
-    },
-    {
-      "epoch": 0.4480760315252666,
-      "grad_norm": 3.0055248737335205,
-      "learning_rate": 1.3483232803239203e-05,
-      "loss": 0.8359,
-      "step": 9665
-    },
-    {
-      "epoch": 0.44830783495595733,
-      "grad_norm": 3.7039334774017334,
-      "learning_rate": 1.3475646930010005e-05,
-      "loss": 0.8554,
-      "step": 9670
-    },
-    {
-      "epoch": 0.4485396383866481,
-      "grad_norm": 3.22282338142395,
-      "learning_rate": 1.346805878121146e-05,
-      "loss": 0.9855,
-      "step": 9675
-    },
-    {
-      "epoch": 0.4487714418173389,
-      "grad_norm": 4.586467266082764,
-      "learning_rate": 1.3460468361811674e-05,
-      "loss": 0.8068,
-      "step": 9680
-    },
-    {
-      "epoch": 0.44900324524802965,
-      "grad_norm": 3.3284201622009277,
-      "learning_rate": 1.345287567678023e-05,
-      "loss": 0.892,
-      "step": 9685
-    },
-    {
-      "epoch": 0.44923504867872044,
-      "grad_norm": 3.7659194469451904,
-      "learning_rate": 1.34452807310882e-05,
-      "loss": 0.9078,
-      "step": 9690
-    },
-    {
-      "epoch": 0.44946685210941123,
-      "grad_norm": 4.3242363929748535,
-      "learning_rate": 1.3437683529708132e-05,
-      "loss": 0.904,
-      "step": 9695
-    },
-    {
-      "epoch": 0.44969865554010197,
-      "grad_norm": 4.269758701324463,
-      "learning_rate": 1.3430084077614052e-05,
-      "loss": 0.9889,
-      "step": 9700
-    },
-    {
-      "epoch": 0.44969865554010197,
-      "eval_loss": 0.9509567618370056,
-      "eval_runtime": 11.2534,
-      "eval_samples_per_second": 11.285,
-      "eval_steps_per_second": 11.285,
-      "step": 9700
-    },
-    {
-      "epoch": 0.44993045897079276,
-      "grad_norm": 3.9716100692749023,
-      "learning_rate": 1.3422482379781459e-05,
-      "loss": 0.8453,
-      "step": 9705
-    },
-    {
-      "epoch": 0.45016226240148355,
-      "grad_norm": 4.124273300170898,
-      "learning_rate": 1.3414878441187321e-05,
-      "loss": 0.885,
-      "step": 9710
-    },
-    {
-      "epoch": 0.45039406583217434,
-      "grad_norm": 3.850370168685913,
-      "learning_rate": 1.3407272266810082e-05,
-      "loss": 1.081,
-      "step": 9715
-    },
-    {
-      "epoch": 0.4506258692628651,
-      "grad_norm": 3.7931642532348633,
-      "learning_rate": 1.3399663861629638e-05,
-      "loss": 0.9646,
-      "step": 9720
-    },
-    {
-      "epoch": 0.45085767269355587,
-      "grad_norm": 3.9382190704345703,
-      "learning_rate": 1.3392053230627349e-05,
-      "loss": 0.7776,
-      "step": 9725
-    },
-    {
-      "epoch": 0.45108947612424666,
-      "grad_norm": 4.051333427429199,
-      "learning_rate": 1.3384440378786039e-05,
-      "loss": 0.9445,
-      "step": 9730
-    },
-    {
-      "epoch": 0.4513212795549374,
-      "grad_norm": 4.055787563323975,
-      "learning_rate": 1.3376825311089974e-05,
-      "loss": 0.9952,
-      "step": 9735
-    },
-    {
-      "epoch": 0.4515530829856282,
-      "grad_norm": 3.4830150604248047,
-      "learning_rate": 1.3369208032524884e-05,
-      "loss": 0.9384,
-      "step": 9740
-    },
-    {
-      "epoch": 0.451784886416319,
-      "grad_norm": 3.9164631366729736,
-      "learning_rate": 1.3361588548077935e-05,
-      "loss": 0.8746,
-      "step": 9745
-    },
-    {
-      "epoch": 0.4520166898470097,
-      "grad_norm": 3.741633415222168,
-      "learning_rate": 1.3353966862737746e-05,
-      "loss": 0.9874,
-      "step": 9750
-    },
-    {
-      "epoch": 0.4522484932777005,
-      "grad_norm": 3.608835220336914,
-      "learning_rate": 1.3346342981494372e-05,
-      "loss": 0.9664,
-      "step": 9755
-    },
-    {
-      "epoch": 0.4524802967083913,
-      "grad_norm": 3.453850746154785,
-      "learning_rate": 1.3338716909339304e-05,
-      "loss": 0.8703,
-      "step": 9760
-    },
-    {
-      "epoch": 0.4527121001390821,
-      "grad_norm": 3.682974100112915,
-      "learning_rate": 1.3331088651265473e-05,
-      "loss": 1.0097,
-      "step": 9765
-    },
-    {
-      "epoch": 0.4529439035697728,
-      "grad_norm": 3.790799140930176,
-      "learning_rate": 1.3323458212267238e-05,
-      "loss": 0.8946,
-      "step": 9770
-    },
-    {
-      "epoch": 0.4531757070004636,
-      "grad_norm": 3.7424209117889404,
-      "learning_rate": 1.3315825597340388e-05,
-      "loss": 0.872,
-      "step": 9775
-    },
-    {
-      "epoch": 0.4534075104311544,
-      "grad_norm": 4.28916597366333,
-      "learning_rate": 1.3308190811482134e-05,
-      "loss": 0.9537,
-      "step": 9780
-    },
-    {
-      "epoch": 0.45363931386184514,
-      "grad_norm": 4.0296220779418945,
-      "learning_rate": 1.3300553859691103e-05,
-      "loss": 1.019,
-      "step": 9785
-    },
-    {
-      "epoch": 0.45387111729253593,
-      "grad_norm": 3.4381442070007324,
-      "learning_rate": 1.3292914746967361e-05,
-      "loss": 0.8505,
-      "step": 9790
-    },
-    {
-      "epoch": 0.4541029207232267,
-      "grad_norm": 3.793199300765991,
-      "learning_rate": 1.3285273478312359e-05,
-      "loss": 0.8457,
-      "step": 9795
-    },
-    {
-      "epoch": 0.45433472415391746,
-      "grad_norm": 3.8284950256347656,
-      "learning_rate": 1.3277630058728984e-05,
-      "loss": 1.0618,
-      "step": 9800
-    },
-    {
-      "epoch": 0.45433472415391746,
-      "eval_loss": 0.9499906301498413,
-      "eval_runtime": 11.2829,
-      "eval_samples_per_second": 11.256,
-      "eval_steps_per_second": 11.256,
-      "step": 9800
-    },
-    {
-      "epoch": 0.45456652758460825,
-      "grad_norm": 3.412783622741699,
-      "learning_rate": 1.3269984493221516e-05,
-      "loss": 0.8846,
-      "step": 9805
-    },
-    {
-      "epoch": 0.45479833101529904,
-      "grad_norm": 3.8301942348480225,
-      "learning_rate": 1.3262336786795654e-05,
-      "loss": 0.982,
-      "step": 9810
-    },
-    {
-      "epoch": 0.4550301344459898,
-      "grad_norm": 3.887139081954956,
-      "learning_rate": 1.3254686944458484e-05,
-      "loss": 0.8593,
-      "step": 9815
-    },
-    {
-      "epoch": 0.45526193787668057,
-      "grad_norm": 3.891221761703491,
-      "learning_rate": 1.32470349712185e-05,
-      "loss": 0.9703,
-      "step": 9820
-    },
-    {
-      "epoch": 0.45549374130737136,
-      "grad_norm": 3.008439540863037,
-      "learning_rate": 1.3239380872085586e-05,
-      "loss": 0.7827,
-      "step": 9825
-    },
-    {
-      "epoch": 0.45572554473806215,
-      "grad_norm": 3.8640429973602295,
-      "learning_rate": 1.3231724652071021e-05,
-      "loss": 0.9659,
-      "step": 9830
-    },
-    {
-      "epoch": 0.4559573481687529,
-      "grad_norm": 3.8008694648742676,
-      "learning_rate": 1.3224066316187472e-05,
-      "loss": 0.8439,
-      "step": 9835
-    },
-    {
-      "epoch": 0.4561891515994437,
-      "grad_norm": 3.950207471847534,
-      "learning_rate": 1.3216405869448991e-05,
-      "loss": 1.0957,
-      "step": 9840
-    },
-    {
-      "epoch": 0.45642095503013447,
-      "grad_norm": 4.314381122589111,
-      "learning_rate": 1.3208743316871013e-05,
-      "loss": 0.9362,
-      "step": 9845
-    },
-    {
-      "epoch": 0.4566527584608252,
-      "grad_norm": 3.4835643768310547,
-      "learning_rate": 1.3201078663470348e-05,
-      "loss": 1.0829,
-      "step": 9850
-    },
-    {
-      "epoch": 0.456884561891516,
-      "grad_norm": 3.5081305503845215,
-      "learning_rate": 1.3193411914265186e-05,
-      "loss": 0.8306,
-      "step": 9855
-    },
-    {
-      "epoch": 0.4571163653222068,
-      "grad_norm": 3.41742205619812,
-      "learning_rate": 1.318574307427509e-05,
-      "loss": 0.7945,
-      "step": 9860
-    },
-    {
-      "epoch": 0.4573481687528975,
-      "grad_norm": 4.183412075042725,
-      "learning_rate": 1.3178072148520981e-05,
-      "loss": 0.9223,
-      "step": 9865
-    },
-    {
-      "epoch": 0.4575799721835883,
-      "grad_norm": 3.7634990215301514,
-      "learning_rate": 1.3170399142025163e-05,
-      "loss": 0.7524,
-      "step": 9870
-    },
-    {
-      "epoch": 0.4578117756142791,
-      "grad_norm": 3.36533784866333,
-      "learning_rate": 1.3162724059811291e-05,
-      "loss": 0.8724,
-      "step": 9875
-    },
-    {
-      "epoch": 0.45804357904496984,
-      "grad_norm": 3.756016254425049,
-      "learning_rate": 1.3155046906904376e-05,
-      "loss": 0.9082,
-      "step": 9880
-    },
-    {
-      "epoch": 0.45827538247566063,
-      "grad_norm": 3.022764205932617,
-      "learning_rate": 1.3147367688330795e-05,
-      "loss": 0.8565,
-      "step": 9885
-    },
-    {
-      "epoch": 0.4585071859063514,
-      "grad_norm": 3.910804271697998,
-      "learning_rate": 1.3139686409118268e-05,
-      "loss": 0.994,
-      "step": 9890
-    },
-    {
-      "epoch": 0.4587389893370422,
-      "grad_norm": 4.257674694061279,
-      "learning_rate": 1.3132003074295874e-05,
-      "loss": 0.8303,
-      "step": 9895
-    },
-    {
-      "epoch": 0.45897079276773295,
-      "grad_norm": 4.139712810516357,
-      "learning_rate": 1.3124317688894026e-05,
-      "loss": 0.9909,
-      "step": 9900
-    },
-    {
-      "epoch": 0.45897079276773295,
-      "eval_loss": 0.948937714099884,
-      "eval_runtime": 11.2699,
-      "eval_samples_per_second": 11.269,
-      "eval_steps_per_second": 11.269,
-      "step": 9900
-    },
-    {
-      "epoch": 0.45920259619842374,
-      "grad_norm": 3.3769872188568115,
-      "learning_rate": 1.3116630257944492e-05,
-      "loss": 0.9206,
-      "step": 9905
-    },
-    {
-      "epoch": 0.45943439962911453,
-      "grad_norm": 3.6999621391296387,
-      "learning_rate": 1.3108940786480366e-05,
-      "loss": 0.9053,
-      "step": 9910
-    },
-    {
-      "epoch": 0.45966620305980527,
-      "grad_norm": 3.6005308628082275,
-      "learning_rate": 1.3101249279536089e-05,
-      "loss": 0.8128,
-      "step": 9915
-    },
-    {
-      "epoch": 0.45989800649049606,
-      "grad_norm": 3.3132901191711426,
-      "learning_rate": 1.3093555742147435e-05,
-      "loss": 0.8088,
-      "step": 9920
-    },
-    {
-      "epoch": 0.46012980992118685,
-      "grad_norm": 3.400157928466797,
-      "learning_rate": 1.3085860179351493e-05,
-      "loss": 0.9798,
-      "step": 9925
-    },
-    {
-      "epoch": 0.4603616133518776,
-      "grad_norm": 3.298585891723633,
-      "learning_rate": 1.3078162596186697e-05,
-      "loss": 1.0104,
-      "step": 9930
-    },
-    {
-      "epoch": 0.4605934167825684,
-      "grad_norm": 4.665667533874512,
-      "learning_rate": 1.3070462997692794e-05,
-      "loss": 0.9341,
-      "step": 9935
-    },
-    {
-      "epoch": 0.46082522021325917,
-      "grad_norm": 3.343885660171509,
-      "learning_rate": 1.3062761388910842e-05,
-      "loss": 1.0162,
-      "step": 9940
-    },
-    {
-      "epoch": 0.4610570236439499,
-      "grad_norm": 3.290959596633911,
-      "learning_rate": 1.3055057774883237e-05,
-      "loss": 0.9058,
-      "step": 9945
-    },
-    {
-      "epoch": 0.4612888270746407,
-      "grad_norm": 3.9871811866760254,
-      "learning_rate": 1.3047352160653668e-05,
-      "loss": 1.0865,
-      "step": 9950
-    },
-    {
-      "epoch": 0.4615206305053315,
-      "grad_norm": 4.483913421630859,
-      "learning_rate": 1.3039644551267143e-05,
-      "loss": 0.9857,
-      "step": 9955
-    },
-    {
-      "epoch": 0.4617524339360223,
-      "grad_norm": 3.8120338916778564,
-      "learning_rate": 1.3031934951769975e-05,
-      "loss": 0.9333,
-      "step": 9960
-    },
-    {
-      "epoch": 0.461984237366713,
-      "grad_norm": 3.768869638442993,
-      "learning_rate": 1.3024223367209777e-05,
-      "loss": 0.8497,
-      "step": 9965
-    },
-    {
-      "epoch": 0.4622160407974038,
-      "grad_norm": 3.1632208824157715,
-      "learning_rate": 1.3016509802635464e-05,
-      "loss": 0.8643,
-      "step": 9970
-    },
-    {
-      "epoch": 0.4624478442280946,
-      "grad_norm": 3.730023145675659,
-      "learning_rate": 1.3008794263097248e-05,
-      "loss": 0.7617,
-      "step": 9975
-    },
-    {
-      "epoch": 0.46267964765878533,
-      "grad_norm": 3.803635597229004,
-      "learning_rate": 1.3001076753646637e-05,
-      "loss": 0.8348,
-      "step": 9980
-    },
-    {
-      "epoch": 0.4629114510894761,
-      "grad_norm": 3.9034042358398438,
-      "learning_rate": 1.299335727933642e-05,
-      "loss": 0.9941,
-      "step": 9985
-    },
-    {
-      "epoch": 0.4631432545201669,
-      "grad_norm": 4.450758934020996,
-      "learning_rate": 1.2985635845220678e-05,
-      "loss": 1.0976,
-      "step": 9990
-    },
-    {
-      "epoch": 0.46337505795085765,
-      "grad_norm": 3.4911108016967773,
-      "learning_rate": 1.2977912456354778e-05,
-      "loss": 0.872,
-      "step": 9995
-    },
-    {
-      "epoch": 0.46360686138154844,
-      "grad_norm": 3.8963191509246826,
-      "learning_rate": 1.2970187117795357e-05,
-      "loss": 1.0407,
-      "step": 10000
-    },
-    {
-      "epoch": 0.46360686138154844,
-      "eval_loss": 0.9476349949836731,
-      "eval_runtime": 11.2604,
-      "eval_samples_per_second": 11.278,
-      "eval_steps_per_second": 11.278,
-      "step": 10000
-    },
-    {
-      "epoch": 0.46383866481223923,
-      "grad_norm": 4.193517208099365,
-      "learning_rate": 1.2962459834600341e-05,
-      "loss": 0.9288,
-      "step": 10005
-    },
-    {
-      "epoch": 0.46407046824292997,
-      "grad_norm": 3.453949213027954,
-      "learning_rate": 1.295473061182892e-05,
-      "loss": 0.9205,
-      "step": 10010
-    },
-    {
-      "epoch": 0.46430227167362076,
-      "grad_norm": 3.316931962966919,
-      "learning_rate": 1.2946999454541558e-05,
-      "loss": 0.8677,
-      "step": 10015
-    },
-    {
-      "epoch": 0.46453407510431155,
-      "grad_norm": 3.576442241668701,
-      "learning_rate": 1.2939266367799983e-05,
-      "loss": 0.953,
-      "step": 10020
-    },
-    {
-      "epoch": 0.46476587853500234,
-      "grad_norm": 3.715101957321167,
-      "learning_rate": 1.2931531356667191e-05,
-      "loss": 0.8921,
-      "step": 10025
-    },
-    {
-      "epoch": 0.4649976819656931,
-      "grad_norm": 4.151893138885498,
-      "learning_rate": 1.292379442620743e-05,
-      "loss": 0.8068,
-      "step": 10030
-    },
-    {
-      "epoch": 0.46522948539638387,
-      "grad_norm": 3.853817939758301,
-      "learning_rate": 1.2916055581486211e-05,
-      "loss": 0.9439,
-      "step": 10035
-    },
-    {
-      "epoch": 0.46546128882707466,
-      "grad_norm": 3.5883684158325195,
-      "learning_rate": 1.29083148275703e-05,
-      "loss": 0.8318,
-      "step": 10040
-    },
-    {
-      "epoch": 0.4656930922577654,
-      "grad_norm": 4.212533950805664,
-      "learning_rate": 1.2900572169527705e-05,
-      "loss": 1.0225,
-      "step": 10045
-    },
-    {
-      "epoch": 0.4659248956884562,
-      "grad_norm": 2.86691951751709,
-      "learning_rate": 1.289282761242769e-05,
-      "loss": 0.8968,
-      "step": 10050
-    },
-    {
-      "epoch": 0.466156699119147,
-      "grad_norm": 2.563044548034668,
-      "learning_rate": 1.2885081161340751e-05,
-      "loss": 0.8134,
-      "step": 10055
-    },
-    {
-      "epoch": 0.4663885025498377,
-      "grad_norm": 3.6845993995666504,
-      "learning_rate": 1.2877332821338635e-05,
-      "loss": 0.9026,
-      "step": 10060
-    },
-    {
-      "epoch": 0.4666203059805285,
-      "grad_norm": 4.9104413986206055,
-      "learning_rate": 1.2869582597494324e-05,
-      "loss": 1.0064,
-      "step": 10065
-    },
-    {
-      "epoch": 0.4668521094112193,
-      "grad_norm": 3.9013454914093018,
-      "learning_rate": 1.2861830494882021e-05,
-      "loss": 0.9372,
-      "step": 10070
-    },
-    {
-      "epoch": 0.4670839128419101,
-      "grad_norm": 3.2289061546325684,
-      "learning_rate": 1.285407651857718e-05,
-      "loss": 0.8886,
-      "step": 10075
-    },
-    {
-      "epoch": 0.4673157162726008,
-      "grad_norm": 3.8512096405029297,
-      "learning_rate": 1.2846320673656464e-05,
-      "loss": 0.9159,
-      "step": 10080
-    },
-    {
-      "epoch": 0.4675475197032916,
-      "grad_norm": 4.084563732147217,
-      "learning_rate": 1.2838562965197766e-05,
-      "loss": 0.8895,
-      "step": 10085
-    },
-    {
-      "epoch": 0.4677793231339824,
-      "grad_norm": 4.185845375061035,
-      "learning_rate": 1.2830803398280203e-05,
-      "loss": 0.9423,
-      "step": 10090
-    },
-    {
-      "epoch": 0.46801112656467314,
-      "grad_norm": 3.8135392665863037,
-      "learning_rate": 1.2823041977984102e-05,
-      "loss": 0.8725,
-      "step": 10095
-    },
-    {
-      "epoch": 0.46824292999536393,
-      "grad_norm": 4.087751865386963,
-      "learning_rate": 1.2815278709391004e-05,
-      "loss": 0.9895,
-      "step": 10100
-    },
-    {
-      "epoch": 0.46824292999536393,
-      "eval_loss": 0.9459433555603027,
-      "eval_runtime": 11.2767,
-      "eval_samples_per_second": 11.262,
-      "eval_steps_per_second": 11.262,
-      "step": 10100
-    },
-    {
-      "epoch": 0.4684747334260547,
-      "grad_norm": 3.648947238922119,
-      "learning_rate": 1.2807513597583665e-05,
-      "loss": 1.0151,
-      "step": 10105
-    },
-    {
-      "epoch": 0.46870653685674546,
-      "grad_norm": 3.339587926864624,
-      "learning_rate": 1.2799746647646045e-05,
-      "loss": 0.9785,
-      "step": 10110
-    },
-    {
-      "epoch": 0.46893834028743625,
-      "grad_norm": 3.7756528854370117,
-      "learning_rate": 1.279197786466331e-05,
-      "loss": 0.7827,
-      "step": 10115
-    },
-    {
-      "epoch": 0.46917014371812704,
-      "grad_norm": 4.143910884857178,
-      "learning_rate": 1.2784207253721817e-05,
-      "loss": 0.954,
-      "step": 10120
-    },
-    {
-      "epoch": 0.4694019471488178,
-      "grad_norm": 2.996795892715454,
-      "learning_rate": 1.2776434819909136e-05,
-      "loss": 0.6456,
-      "step": 10125
-    },
-    {
-      "epoch": 0.46963375057950857,
-      "grad_norm": 4.004779815673828,
-      "learning_rate": 1.2768660568314012e-05,
-      "loss": 1.0712,
-      "step": 10130
-    },
-    {
-      "epoch": 0.46986555401019936,
-      "grad_norm": 3.2952988147735596,
-      "learning_rate": 1.2760884504026392e-05,
-      "loss": 0.7785,
-      "step": 10135
-    },
-    {
-      "epoch": 0.47009735744089015,
-      "grad_norm": 5.342103958129883,
-      "learning_rate": 1.275310663213741e-05,
-      "loss": 1.0066,
-      "step": 10140
-    },
-    {
-      "epoch": 0.4703291608715809,
-      "grad_norm": 4.140252113342285,
-      "learning_rate": 1.2745326957739375e-05,
-      "loss": 0.959,
-      "step": 10145
-    },
-    {
-      "epoch": 0.4705609643022717,
-      "grad_norm": 3.928713321685791,
-      "learning_rate": 1.2737545485925786e-05,
-      "loss": 0.9485,
-      "step": 10150
-    },
-    {
-      "epoch": 0.47079276773296247,
-      "grad_norm": 4.4672136306762695,
-      "learning_rate": 1.2729762221791312e-05,
-      "loss": 0.9565,
-      "step": 10155
-    },
-    {
-      "epoch": 0.4710245711636532,
-      "grad_norm": 3.2409780025482178,
-      "learning_rate": 1.27219771704318e-05,
-      "loss": 0.8986,
-      "step": 10160
-    },
-    {
-      "epoch": 0.471256374594344,
-      "grad_norm": 4.780323505401611,
-      "learning_rate": 1.2714190336944264e-05,
-      "loss": 0.9055,
-      "step": 10165
-    },
-    {
-      "epoch": 0.4714881780250348,
-      "grad_norm": 3.7296030521392822,
-      "learning_rate": 1.2706401726426884e-05,
-      "loss": 0.8048,
-      "step": 10170
-    },
-    {
-      "epoch": 0.4717199814557255,
-      "grad_norm": 3.707808017730713,
-      "learning_rate": 1.2698611343979007e-05,
-      "loss": 0.9201,
-      "step": 10175
-    },
-    {
-      "epoch": 0.4719517848864163,
-      "grad_norm": 3.560077667236328,
-      "learning_rate": 1.2690819194701134e-05,
-      "loss": 0.9069,
-      "step": 10180
-    },
-    {
-      "epoch": 0.4721835883171071,
-      "grad_norm": 4.014420032501221,
-      "learning_rate": 1.2683025283694936e-05,
-      "loss": 0.9935,
-      "step": 10185
-    },
-    {
-      "epoch": 0.47241539174779784,
-      "grad_norm": 3.263787031173706,
-      "learning_rate": 1.267522961606322e-05,
-      "loss": 0.8071,
-      "step": 10190
-    },
-    {
-      "epoch": 0.47264719517848863,
-      "grad_norm": 4.139646053314209,
-      "learning_rate": 1.2667432196909955e-05,
-      "loss": 0.9272,
-      "step": 10195
-    },
-    {
-      "epoch": 0.4728789986091794,
-      "grad_norm": 4.009764671325684,
-      "learning_rate": 1.2659633031340255e-05,
-      "loss": 0.9658,
-      "step": 10200
-    },
-    {
-      "epoch": 0.4728789986091794,
-      "eval_loss": 0.9445049166679382,
-      "eval_runtime": 11.2699,
-      "eval_samples_per_second": 11.269,
-      "eval_steps_per_second": 11.269,
-      "step": 10200
-    },
-    {
-      "epoch": 0.4731108020398702,
-      "grad_norm": 3.5922906398773193,
-      "learning_rate": 1.2651832124460369e-05,
-      "loss": 0.9533,
-      "step": 10205
-    },
-    {
-      "epoch": 0.47334260547056095,
-      "grad_norm": 3.384347677230835,
-      "learning_rate": 1.26440294813777e-05,
-      "loss": 0.7947,
-      "step": 10210
-    },
-    {
-      "epoch": 0.47357440890125174,
-      "grad_norm": 3.7490670680999756,
-      "learning_rate": 1.2636225107200779e-05,
-      "loss": 0.7939,
-      "step": 10215
-    },
-    {
-      "epoch": 0.47380621233194253,
-      "grad_norm": 4.333261966705322,
-      "learning_rate": 1.262841900703927e-05,
-      "loss": 0.7959,
-      "step": 10220
-    },
-    {
-      "epoch": 0.47403801576263327,
-      "grad_norm": 4.149163722991943,
-      "learning_rate": 1.262061118600397e-05,
-      "loss": 0.9461,
-      "step": 10225
-    },
-    {
-      "epoch": 0.47426981919332406,
-      "grad_norm": 3.4442434310913086,
-      "learning_rate": 1.26128016492068e-05,
-      "loss": 0.874,
-      "step": 10230
-    },
-    {
-      "epoch": 0.47450162262401485,
-      "grad_norm": 3.540757894515991,
-      "learning_rate": 1.260499040176081e-05,
-      "loss": 0.9761,
-      "step": 10235
-    },
-    {
-      "epoch": 0.4747334260547056,
-      "grad_norm": 4.07343864440918,
-      "learning_rate": 1.2597177448780162e-05,
-      "loss": 1.0432,
-      "step": 10240
-    },
-    {
-      "epoch": 0.4749652294853964,
-      "grad_norm": 3.954550266265869,
-      "learning_rate": 1.258936279538014e-05,
-      "loss": 0.9347,
-      "step": 10245
-    },
-    {
-      "epoch": 0.47519703291608717,
-      "grad_norm": 4.400870323181152,
-      "learning_rate": 1.2581546446677142e-05,
-      "loss": 1.2181,
-      "step": 10250
-    },
-    {
-      "epoch": 0.4754288363467779,
-      "grad_norm": 3.343297004699707,
-      "learning_rate": 1.2573728407788667e-05,
-      "loss": 0.7962,
-      "step": 10255
-    },
-    {
-      "epoch": 0.4756606397774687,
-      "grad_norm": 3.889265775680542,
-      "learning_rate": 1.2565908683833337e-05,
-      "loss": 0.8248,
-      "step": 10260
-    },
-    {
-      "epoch": 0.4758924432081595,
-      "grad_norm": 3.9012787342071533,
-      "learning_rate": 1.2558087279930859e-05,
-      "loss": 0.9967,
-      "step": 10265
-    },
-    {
-      "epoch": 0.4761242466388503,
-      "grad_norm": 3.298150062561035,
-      "learning_rate": 1.2550264201202054e-05,
-      "loss": 0.9277,
-      "step": 10270
-    },
-    {
-      "epoch": 0.476356050069541,
-      "grad_norm": 3.909010887145996,
-      "learning_rate": 1.2542439452768833e-05,
-      "loss": 0.9993,
-      "step": 10275
-    },
-    {
-      "epoch": 0.4765878535002318,
-      "grad_norm": 3.329361915588379,
-      "learning_rate": 1.2534613039754201e-05,
-      "loss": 0.9402,
-      "step": 10280
-    },
-    {
-      "epoch": 0.4768196569309226,
-      "grad_norm": 3.2942123413085938,
-      "learning_rate": 1.2526784967282255e-05,
-      "loss": 0.8344,
-      "step": 10285
-    },
-    {
-      "epoch": 0.47705146036161333,
-      "grad_norm": 4.618292808532715,
-      "learning_rate": 1.251895524047817e-05,
-      "loss": 0.8381,
-      "step": 10290
-    },
-    {
-      "epoch": 0.4772832637923041,
-      "grad_norm": 3.3553268909454346,
-      "learning_rate": 1.2511123864468218e-05,
-      "loss": 0.844,
-      "step": 10295
-    },
-    {
-      "epoch": 0.4775150672229949,
-      "grad_norm": 4.03507137298584,
-      "learning_rate": 1.2503290844379739e-05,
-      "loss": 0.9175,
-      "step": 10300
-    },
-    {
-      "epoch": 0.4775150672229949,
-      "eval_loss": 0.9433912634849548,
-      "eval_runtime": 11.2657,
-      "eval_samples_per_second": 11.273,
-      "eval_steps_per_second": 11.273,
-      "step": 10300
-    },
-    {
-      "epoch": 0.47774687065368565,
-      "grad_norm": 3.0558102130889893,
-      "learning_rate": 1.2495456185341157e-05,
-      "loss": 0.9507,
-      "step": 10305
-    },
-    {
-      "epoch": 0.47797867408437644,
-      "grad_norm": 3.6690430641174316,
-      "learning_rate": 1.2487619892481966e-05,
-      "loss": 0.8518,
-      "step": 10310
-    },
-    {
-      "epoch": 0.47821047751506723,
-      "grad_norm": 4.014634132385254,
-      "learning_rate": 1.2479781970932727e-05,
-      "loss": 0.8619,
-      "step": 10315
-    },
-    {
-      "epoch": 0.47844228094575797,
-      "grad_norm": 3.3118784427642822,
-      "learning_rate": 1.247194242582507e-05,
-      "loss": 0.9213,
-      "step": 10320
-    },
-    {
-      "epoch": 0.47867408437644876,
-      "grad_norm": 3.962343454360962,
-      "learning_rate": 1.2464101262291688e-05,
-      "loss": 0.9235,
-      "step": 10325
-    },
-    {
-      "epoch": 0.47890588780713955,
-      "grad_norm": 3.672382354736328,
-      "learning_rate": 1.2456258485466339e-05,
-      "loss": 0.8238,
-      "step": 10330
-    },
-    {
-      "epoch": 0.47913769123783034,
-      "grad_norm": 4.043935298919678,
-      "learning_rate": 1.2448414100483826e-05,
-      "loss": 0.986,
-      "step": 10335
-    },
-    {
-      "epoch": 0.4793694946685211,
-      "grad_norm": 3.571040630340576,
-      "learning_rate": 1.2440568112480009e-05,
-      "loss": 0.8952,
-      "step": 10340
-    },
-    {
-      "epoch": 0.47960129809921187,
-      "grad_norm": 4.216655731201172,
-      "learning_rate": 1.2432720526591808e-05,
-      "loss": 1.0396,
-      "step": 10345
-    },
-    {
-      "epoch": 0.47983310152990266,
-      "grad_norm": 3.7450881004333496,
-      "learning_rate": 1.2424871347957173e-05,
-      "loss": 0.9586,
-      "step": 10350
-    },
-    {
-      "epoch": 0.4800649049605934,
-      "grad_norm": 3.8393282890319824,
-      "learning_rate": 1.2417020581715104e-05,
-      "loss": 0.8527,
-      "step": 10355
-    },
-    {
-      "epoch": 0.4802967083912842,
-      "grad_norm": 3.087303400039673,
-      "learning_rate": 1.2409168233005644e-05,
-      "loss": 0.8458,
-      "step": 10360
-    },
-    {
-      "epoch": 0.480528511821975,
-      "grad_norm": 3.759127378463745,
-      "learning_rate": 1.2401314306969869e-05,
-      "loss": 1.0019,
-      "step": 10365
-    },
-    {
-      "epoch": 0.4807603152526657,
-      "grad_norm": 3.2928617000579834,
-      "learning_rate": 1.2393458808749885e-05,
-      "loss": 0.8309,
-      "step": 10370
-    },
-    {
-      "epoch": 0.4809921186833565,
-      "grad_norm": 3.7181479930877686,
-      "learning_rate": 1.2385601743488831e-05,
-      "loss": 0.8989,
-      "step": 10375
-    },
-    {
-      "epoch": 0.4812239221140473,
-      "grad_norm": 3.783705711364746,
-      "learning_rate": 1.2377743116330869e-05,
-      "loss": 0.8045,
-      "step": 10380
-    },
-    {
-      "epoch": 0.4814557255447381,
-      "grad_norm": 3.634629487991333,
-      "learning_rate": 1.2369882932421184e-05,
-      "loss": 0.9103,
-      "step": 10385
-    },
-    {
-      "epoch": 0.4816875289754288,
-      "grad_norm": 3.8915700912475586,
-      "learning_rate": 1.236202119690599e-05,
-      "loss": 0.9318,
-      "step": 10390
-    },
-    {
-      "epoch": 0.4819193324061196,
-      "grad_norm": 4.209451675415039,
-      "learning_rate": 1.2354157914932501e-05,
-      "loss": 0.9598,
-      "step": 10395
-    },
-    {
-      "epoch": 0.4821511358368104,
-      "grad_norm": 4.348424911499023,
-      "learning_rate": 1.234629309164895e-05,
-      "loss": 0.915,
-      "step": 10400
-    },
-    {
-      "epoch": 0.4821511358368104,
-      "eval_loss": 0.9418513178825378,
-      "eval_runtime": 11.2768,
-      "eval_samples_per_second": 11.262,
-      "eval_steps_per_second": 11.262,
-      "step": 10400
-    },
-    {
-      "epoch": 0.48238293926750114,
-      "grad_norm": 3.9658772945404053,
-      "learning_rate": 1.2338426732204585e-05,
-      "loss": 0.9433,
-      "step": 10405
-    },
-    {
-      "epoch": 0.48261474269819193,
-      "grad_norm": 3.0501604080200195,
-      "learning_rate": 1.2330558841749649e-05,
-      "loss": 0.6786,
-      "step": 10410
-    },
-    {
-      "epoch": 0.4828465461288827,
-      "grad_norm": 3.564727306365967,
-      "learning_rate": 1.2322689425435401e-05,
-      "loss": 0.9322,
-      "step": 10415
-    },
-    {
-      "epoch": 0.48307834955957346,
-      "grad_norm": 5.660946846008301,
-      "learning_rate": 1.2314818488414083e-05,
-      "loss": 1.0264,
-      "step": 10420
-    },
-    {
-      "epoch": 0.48331015299026425,
-      "grad_norm": 4.12052583694458,
-      "learning_rate": 1.2306946035838946e-05,
-      "loss": 0.8962,
-      "step": 10425
-    },
-    {
-      "epoch": 0.48354195642095504,
-      "grad_norm": 4.541821479797363,
-      "learning_rate": 1.229907207286423e-05,
-      "loss": 1.1351,
-      "step": 10430
-    },
-    {
-      "epoch": 0.4837737598516458,
-      "grad_norm": 3.592571496963501,
-      "learning_rate": 1.2291196604645153e-05,
-      "loss": 0.839,
-      "step": 10435
-    },
-    {
-      "epoch": 0.48400556328233657,
-      "grad_norm": 3.7811429500579834,
-      "learning_rate": 1.2283319636337935e-05,
-      "loss": 0.9498,
-      "step": 10440
-    },
-    {
-      "epoch": 0.48423736671302736,
-      "grad_norm": 3.5306897163391113,
-      "learning_rate": 1.227544117309977e-05,
-      "loss": 0.7831,
-      "step": 10445
-    },
-    {
-      "epoch": 0.48446917014371815,
-      "grad_norm": 3.1777243614196777,
-      "learning_rate": 1.2267561220088829e-05,
-      "loss": 0.7042,
-      "step": 10450
-    },
-    {
-      "epoch": 0.4847009735744089,
-      "grad_norm": 3.2184152603149414,
-      "learning_rate": 1.225967978246426e-05,
-      "loss": 0.8678,
-      "step": 10455
-    },
-    {
-      "epoch": 0.4849327770050997,
-      "grad_norm": 4.353787422180176,
-      "learning_rate": 1.2251796865386185e-05,
-      "loss": 0.9821,
-      "step": 10460
-    },
-    {
-      "epoch": 0.48516458043579047,
-      "grad_norm": 4.005190372467041,
-      "learning_rate": 1.2243912474015698e-05,
-      "loss": 0.9116,
-      "step": 10465
-    },
-    {
-      "epoch": 0.4853963838664812,
-      "grad_norm": 3.815784454345703,
-      "learning_rate": 1.2236026613514846e-05,
-      "loss": 0.8607,
-      "step": 10470
-    },
-    {
-      "epoch": 0.485628187297172,
-      "grad_norm": 3.319671154022217,
-      "learning_rate": 1.222813928904665e-05,
-      "loss": 0.8951,
-      "step": 10475
-    },
-    {
-      "epoch": 0.4858599907278628,
-      "grad_norm": 3.4613735675811768,
-      "learning_rate": 1.2220250505775086e-05,
-      "loss": 0.8038,
-      "step": 10480
-    },
-    {
-      "epoch": 0.4860917941585535,
-      "grad_norm": 4.234585285186768,
-      "learning_rate": 1.2212360268865079e-05,
-      "loss": 0.9614,
-      "step": 10485
-    },
-    {
-      "epoch": 0.4863235975892443,
-      "grad_norm": 3.492586374282837,
-      "learning_rate": 1.2204468583482514e-05,
-      "loss": 0.816,
-      "step": 10490
-    },
-    {
-      "epoch": 0.4865554010199351,
-      "grad_norm": 3.3907053470611572,
-      "learning_rate": 1.2196575454794222e-05,
-      "loss": 1.0767,
-      "step": 10495
-    },
-    {
-      "epoch": 0.48678720445062584,
-      "grad_norm": 3.7171268463134766,
-      "learning_rate": 1.2188680887967977e-05,
-      "loss": 0.9849,
-      "step": 10500
-    },
-    {
-      "epoch": 0.48678720445062584,
-      "eval_loss": 0.9430806040763855,
-      "eval_runtime": 11.2641,
-      "eval_samples_per_second": 11.275,
-      "eval_steps_per_second": 11.275,
-      "step": 10500
-    },
-    {
-      "epoch": 0.48701900788131663,
-      "grad_norm": 3.030805826187134,
-      "learning_rate": 1.2180784888172494e-05,
-      "loss": 0.8955,
-      "step": 10505
-    },
-    {
-      "epoch": 0.4872508113120074,
-      "grad_norm": 4.06427001953125,
-      "learning_rate": 1.2172887460577432e-05,
-      "loss": 1.0308,
-      "step": 10510
-    },
-    {
-      "epoch": 0.4874826147426982,
-      "grad_norm": 3.5869393348693848,
-      "learning_rate": 1.2164988610353375e-05,
-      "loss": 0.9683,
-      "step": 10515
-    },
-    {
-      "epoch": 0.48771441817338895,
-      "grad_norm": 3.8638756275177,
-      "learning_rate": 1.215708834267185e-05,
-      "loss": 1.0335,
-      "step": 10520
-    },
-    {
-      "epoch": 0.48794622160407974,
-      "grad_norm": 3.455765724182129,
-      "learning_rate": 1.21491866627053e-05,
-      "loss": 0.7771,
-      "step": 10525
-    },
-    {
-      "epoch": 0.48817802503477054,
-      "grad_norm": 3.3024017810821533,
-      "learning_rate": 1.2141283575627105e-05,
-      "loss": 0.9882,
-      "step": 10530
-    },
-    {
-      "epoch": 0.48840982846546127,
-      "grad_norm": 3.9688782691955566,
-      "learning_rate": 1.2133379086611559e-05,
-      "loss": 0.9241,
-      "step": 10535
-    },
-    {
-      "epoch": 0.48864163189615206,
-      "grad_norm": 4.1364874839782715,
-      "learning_rate": 1.212547320083387e-05,
-      "loss": 0.838,
-      "step": 10540
-    },
-    {
-      "epoch": 0.48887343532684285,
-      "grad_norm": 3.659271717071533,
-      "learning_rate": 1.211756592347017e-05,
-      "loss": 0.883,
-      "step": 10545
-    },
-    {
-      "epoch": 0.4891052387575336,
-      "grad_norm": 3.7890548706054688,
-      "learning_rate": 1.2109657259697496e-05,
-      "loss": 0.9718,
-      "step": 10550
-    },
-    {
-      "epoch": 0.4893370421882244,
-      "grad_norm": 3.4658312797546387,
-      "learning_rate": 1.210174721469379e-05,
-      "loss": 0.916,
-      "step": 10555
-    },
-    {
-      "epoch": 0.48956884561891517,
-      "grad_norm": 3.9163401126861572,
-      "learning_rate": 1.2093835793637911e-05,
-      "loss": 0.8764,
-      "step": 10560
-    },
-    {
-      "epoch": 0.4898006490496059,
-      "grad_norm": 3.979099750518799,
-      "learning_rate": 1.2085923001709607e-05,
-      "loss": 1.046,
-      "step": 10565
-    },
-    {
-      "epoch": 0.4900324524802967,
-      "grad_norm": 4.127171039581299,
-      "learning_rate": 1.2078008844089523e-05,
-      "loss": 1.011,
-      "step": 10570
-    },
-    {
-      "epoch": 0.4902642559109875,
-      "grad_norm": 5.003758907318115,
-      "learning_rate": 1.2070093325959209e-05,
-      "loss": 0.9325,
-      "step": 10575
-    },
-    {
-      "epoch": 0.4904960593416783,
-      "grad_norm": 3.5403106212615967,
-      "learning_rate": 1.2062176452501093e-05,
-      "loss": 1.002,
-      "step": 10580
-    },
-    {
-      "epoch": 0.490727862772369,
-      "grad_norm": 3.0813117027282715,
-      "learning_rate": 1.2054258228898498e-05,
-      "loss": 0.7354,
-      "step": 10585
-    },
-    {
-      "epoch": 0.4909596662030598,
-      "grad_norm": 4.302268981933594,
-      "learning_rate": 1.204633866033563e-05,
-      "loss": 1.0057,
-      "step": 10590
-    },
-    {
-      "epoch": 0.4911914696337506,
-      "grad_norm": 3.079711437225342,
-      "learning_rate": 1.2038417751997575e-05,
-      "loss": 0.8065,
-      "step": 10595
-    },
-    {
-      "epoch": 0.49142327306444133,
-      "grad_norm": 4.152148246765137,
-      "learning_rate": 1.2030495509070295e-05,
-      "loss": 0.9679,
-      "step": 10600
-    },
-    {
-      "epoch": 0.49142327306444133,
-      "eval_loss": 0.9401566982269287,
-      "eval_runtime": 11.2741,
-      "eval_samples_per_second": 11.265,
-      "eval_steps_per_second": 11.265,
-      "step": 10600
-    },
-    {
-      "epoch": 0.4916550764951321,
-      "grad_norm": 3.9036941528320312,
-      "learning_rate": 1.2022571936740626e-05,
-      "loss": 0.8449,
-      "step": 10605
-    },
-    {
-      "epoch": 0.4918868799258229,
-      "grad_norm": 3.7571048736572266,
-      "learning_rate": 1.201464704019628e-05,
-      "loss": 1.0602,
-      "step": 10610
-    },
-    {
-      "epoch": 0.49211868335651365,
-      "grad_norm": 4.245012283325195,
-      "learning_rate": 1.200672082462582e-05,
-      "loss": 0.9533,
-      "step": 10615
-    },
-    {
-      "epoch": 0.49235048678720444,
-      "grad_norm": 4.420340061187744,
-      "learning_rate": 1.1998793295218696e-05,
-      "loss": 0.8259,
-      "step": 10620
-    },
-    {
-      "epoch": 0.49258229021789524,
-      "grad_norm": 3.532597064971924,
-      "learning_rate": 1.1990864457165198e-05,
-      "loss": 0.9432,
-      "step": 10625
-    },
-    {
-      "epoch": 0.49281409364858597,
-      "grad_norm": 3.4388935565948486,
-      "learning_rate": 1.1982934315656487e-05,
-      "loss": 0.7395,
-      "step": 10630
-    },
-    {
-      "epoch": 0.49304589707927676,
-      "grad_norm": 3.3850276470184326,
-      "learning_rate": 1.1975002875884565e-05,
-      "loss": 0.9985,
-      "step": 10635
-    },
-    {
-      "epoch": 0.49327770050996755,
-      "grad_norm": 4.3244099617004395,
-      "learning_rate": 1.1967070143042291e-05,
-      "loss": 0.9041,
-      "step": 10640
-    },
-    {
-      "epoch": 0.49350950394065835,
-      "grad_norm": 3.7923336029052734,
-      "learning_rate": 1.1959136122323376e-05,
-      "loss": 0.7961,
-      "step": 10645
-    },
-    {
-      "epoch": 0.4937413073713491,
-      "grad_norm": 4.479656219482422,
-      "learning_rate": 1.195120081892236e-05,
-      "loss": 1.0191,
-      "step": 10650
-    },
-    {
-      "epoch": 0.49397311080203987,
-      "grad_norm": 4.143479347229004,
-      "learning_rate": 1.1943264238034633e-05,
-      "loss": 0.9074,
-      "step": 10655
-    },
-    {
-      "epoch": 0.49420491423273066,
-      "grad_norm": 4.062614917755127,
-      "learning_rate": 1.193532638485642e-05,
-      "loss": 0.7747,
-      "step": 10660
-    },
-    {
-      "epoch": 0.4944367176634214,
-      "grad_norm": 3.4529573917388916,
-      "learning_rate": 1.1927387264584778e-05,
-      "loss": 0.9046,
-      "step": 10665
-    },
-    {
-      "epoch": 0.4946685210941122,
-      "grad_norm": 3.946471929550171,
-      "learning_rate": 1.1919446882417595e-05,
-      "loss": 1.1305,
-      "step": 10670
-    },
-    {
-      "epoch": 0.494900324524803,
-      "grad_norm": 4.636277675628662,
-      "learning_rate": 1.191150524355358e-05,
-      "loss": 0.8582,
-      "step": 10675
-    },
-    {
-      "epoch": 0.4951321279554937,
-      "grad_norm": 3.944178819656372,
-      "learning_rate": 1.1903562353192273e-05,
-      "loss": 0.9054,
-      "step": 10680
-    },
-    {
-      "epoch": 0.4953639313861845,
-      "grad_norm": 3.4163365364074707,
-      "learning_rate": 1.1895618216534026e-05,
-      "loss": 0.9238,
-      "step": 10685
-    },
-    {
-      "epoch": 0.4955957348168753,
-      "grad_norm": 3.241332769393921,
-      "learning_rate": 1.1887672838780009e-05,
-      "loss": 0.8977,
-      "step": 10690
-    },
-    {
-      "epoch": 0.4958275382475661,
-      "grad_norm": 3.3289830684661865,
-      "learning_rate": 1.1879726225132208e-05,
-      "loss": 0.8232,
-      "step": 10695
-    },
-    {
-      "epoch": 0.4960593416782568,
-      "grad_norm": 4.045788288116455,
-      "learning_rate": 1.1871778380793418e-05,
-      "loss": 0.9514,
-      "step": 10700
-    },
-    {
-      "epoch": 0.4960593416782568,
-      "eval_loss": 0.9392388463020325,
-      "eval_runtime": 11.2754,
-      "eval_samples_per_second": 11.263,
-      "eval_steps_per_second": 11.263,
-      "step": 10700
-    },
-    {
-      "epoch": 0.4962911451089476,
-      "grad_norm": 3.331512928009033,
-      "learning_rate": 1.1863829310967232e-05,
-      "loss": 0.9257,
-      "step": 10705
-    },
-    {
-      "epoch": 0.4965229485396384,
-      "grad_norm": 4.671146392822266,
-      "learning_rate": 1.1855879020858053e-05,
-      "loss": 0.8757,
-      "step": 10710
-    },
-    {
-      "epoch": 0.49675475197032914,
-      "grad_norm": 4.140007019042969,
-      "learning_rate": 1.1847927515671081e-05,
-      "loss": 0.9834,
-      "step": 10715
-    },
-    {
-      "epoch": 0.49698655540101994,
-      "grad_norm": 3.917147636413574,
-      "learning_rate": 1.183997480061231e-05,
-      "loss": 0.9667,
-      "step": 10720
-    },
-    {
-      "epoch": 0.4972183588317107,
-      "grad_norm": 3.7735893726348877,
-      "learning_rate": 1.1832020880888528e-05,
-      "loss": 0.9726,
-      "step": 10725
-    },
-    {
-      "epoch": 0.49745016226240146,
-      "grad_norm": 4.691269397735596,
-      "learning_rate": 1.1824065761707315e-05,
-      "loss": 1.0185,
-      "step": 10730
-    },
-    {
-      "epoch": 0.49768196569309225,
-      "grad_norm": 3.9918999671936035,
-      "learning_rate": 1.1816109448277029e-05,
-      "loss": 0.8937,
-      "step": 10735
-    },
-    {
-      "epoch": 0.49791376912378305,
-      "grad_norm": 3.6287760734558105,
-      "learning_rate": 1.1808151945806807e-05,
-      "loss": 0.9186,
-      "step": 10740
-    },
-    {
-      "epoch": 0.4981455725544738,
-      "grad_norm": 3.8710758686065674,
-      "learning_rate": 1.1800193259506585e-05,
-      "loss": 0.8035,
-      "step": 10745
-    },
-    {
-      "epoch": 0.4983773759851646,
-      "grad_norm": 4.250580787658691,
-      "learning_rate": 1.1792233394587047e-05,
-      "loss": 0.9368,
-      "step": 10750
-    },
-    {
-      "epoch": 0.49860917941585536,
-      "grad_norm": 4.388089656829834,
-      "learning_rate": 1.1784272356259669e-05,
-      "loss": 1.0037,
-      "step": 10755
-    },
-    {
-      "epoch": 0.49884098284654615,
-      "grad_norm": 3.610869884490967,
-      "learning_rate": 1.1776310149736682e-05,
-      "loss": 0.8208,
-      "step": 10760
-    },
-    {
-      "epoch": 0.4990727862772369,
-      "grad_norm": 4.281845569610596,
-      "learning_rate": 1.1768346780231092e-05,
-      "loss": 0.891,
-      "step": 10765
-    },
-    {
-      "epoch": 0.4993045897079277,
-      "grad_norm": 3.3298609256744385,
-      "learning_rate": 1.1760382252956663e-05,
-      "loss": 0.7928,
-      "step": 10770
-    },
-    {
-      "epoch": 0.4995363931386185,
-      "grad_norm": 3.43632173538208,
-      "learning_rate": 1.1752416573127912e-05,
-      "loss": 0.9587,
-      "step": 10775
-    },
-    {
-      "epoch": 0.4997681965693092,
-      "grad_norm": 3.8259291648864746,
-      "learning_rate": 1.1744449745960119e-05,
-      "loss": 0.8569,
-      "step": 10780
-    },
-    {
-      "epoch": 0.5,
-      "grad_norm": 4.613092422485352,
-      "learning_rate": 1.1736481776669307e-05,
-      "loss": 0.9346,
-      "step": 10785
-    },
-    {
-      "epoch": 0.5002318034306907,
-      "grad_norm": 3.216547966003418,
-      "learning_rate": 1.1728512670472249e-05,
-      "loss": 0.8897,
-      "step": 10790
-    },
-    {
-      "epoch": 0.5004636068613816,
-      "grad_norm": 3.9909191131591797,
-      "learning_rate": 1.172054243258647e-05,
-      "loss": 1.0385,
-      "step": 10795
-    },
-    {
-      "epoch": 0.5006954102920723,
-      "grad_norm": 3.6476809978485107,
-      "learning_rate": 1.1712571068230228e-05,
-      "loss": 0.8202,
-      "step": 10800
-    },
-    {
-      "epoch": 0.5006954102920723,
-      "eval_loss": 0.9371610879898071,
-      "eval_runtime": 11.2585,
-      "eval_samples_per_second": 11.28,
-      "eval_steps_per_second": 11.28,
-      "step": 10800
-    },
-    {
-      "epoch": 0.500927213722763,
-      "grad_norm": 4.106610298156738,
-      "learning_rate": 1.1704598582622517e-05,
-      "loss": 0.9352,
-      "step": 10805
-    },
-    {
-      "epoch": 0.5011590171534539,
-      "grad_norm": 3.3235268592834473,
-      "learning_rate": 1.169662498098307e-05,
-      "loss": 0.7574,
-      "step": 10810
-    },
-    {
-      "epoch": 0.5013908205841446,
-      "grad_norm": 3.4809799194335938,
-      "learning_rate": 1.1688650268532353e-05,
-      "loss": 0.8316,
-      "step": 10815
-    },
-    {
-      "epoch": 0.5016226240148354,
-      "grad_norm": 4.205314636230469,
-      "learning_rate": 1.168067445049155e-05,
-      "loss": 0.9359,
-      "step": 10820
-    },
-    {
-      "epoch": 0.5018544274455262,
-      "grad_norm": 3.303571939468384,
-      "learning_rate": 1.1672697532082578e-05,
-      "loss": 0.8145,
-      "step": 10825
-    },
-    {
-      "epoch": 0.502086230876217,
-      "grad_norm": 4.021134853363037,
-      "learning_rate": 1.1664719518528068e-05,
-      "loss": 0.8672,
-      "step": 10830
-    },
-    {
-      "epoch": 0.5023180343069077,
-      "grad_norm": 3.5206027030944824,
-      "learning_rate": 1.1656740415051373e-05,
-      "loss": 0.9901,
-      "step": 10835
-    },
-    {
-      "epoch": 0.5025498377375985,
-      "grad_norm": 3.848743438720703,
-      "learning_rate": 1.1648760226876559e-05,
-      "loss": 0.87,
-      "step": 10840
-    },
-    {
-      "epoch": 0.5027816411682893,
-      "grad_norm": 4.378417015075684,
-      "learning_rate": 1.1640778959228396e-05,
-      "loss": 1.0345,
-      "step": 10845
-    },
-    {
-      "epoch": 0.5030134445989801,
-      "grad_norm": 3.704587697982788,
-      "learning_rate": 1.1632796617332366e-05,
-      "loss": 1.0267,
-      "step": 10850
-    },
-    {
-      "epoch": 0.5032452480296709,
-      "grad_norm": 4.409409046173096,
-      "learning_rate": 1.1624813206414658e-05,
-      "loss": 1.0559,
-      "step": 10855
-    },
-    {
-      "epoch": 0.5034770514603616,
-      "grad_norm": 3.70389986038208,
-      "learning_rate": 1.1616828731702153e-05,
-      "loss": 0.9175,
-      "step": 10860
-    },
-    {
-      "epoch": 0.5037088548910524,
-      "grad_norm": 3.591284990310669,
-      "learning_rate": 1.1608843198422432e-05,
-      "loss": 0.9618,
-      "step": 10865
-    },
-    {
-      "epoch": 0.5039406583217432,
-      "grad_norm": 4.028764247894287,
-      "learning_rate": 1.1600856611803769e-05,
-      "loss": 1.0365,
-      "step": 10870
-    },
-    {
-      "epoch": 0.5041724617524339,
-      "grad_norm": 3.5418801307678223,
-      "learning_rate": 1.1592868977075132e-05,
-      "loss": 0.9485,
-      "step": 10875
-    },
-    {
-      "epoch": 0.5044042651831248,
-      "grad_norm": 4.499399662017822,
-      "learning_rate": 1.158488029946616e-05,
-      "loss": 1.0452,
-      "step": 10880
-    },
-    {
-      "epoch": 0.5046360686138155,
-      "grad_norm": 3.686279058456421,
-      "learning_rate": 1.15768905842072e-05,
-      "loss": 0.7959,
-      "step": 10885
-    },
-    {
-      "epoch": 0.5048678720445062,
-      "grad_norm": 3.655107021331787,
-      "learning_rate": 1.1568899836529255e-05,
-      "loss": 0.8616,
-      "step": 10890
-    },
-    {
-      "epoch": 0.5050996754751971,
-      "grad_norm": 3.7078745365142822,
-      "learning_rate": 1.1560908061664014e-05,
-      "loss": 0.925,
-      "step": 10895
-    },
-    {
-      "epoch": 0.5053314789058878,
-      "grad_norm": 3.763909101486206,
-      "learning_rate": 1.1552915264843838e-05,
-      "loss": 0.8868,
-      "step": 10900
-    },
-    {
-      "epoch": 0.5053314789058878,
-      "eval_loss": 0.9361811280250549,
-      "eval_runtime": 11.2635,
-      "eval_samples_per_second": 11.275,
-      "eval_steps_per_second": 11.275,
-      "step": 10900
-    },
-    {
-      "epoch": 0.5055632823365785,
-      "grad_norm": 4.104317665100098,
-      "learning_rate": 1.1544921451301759e-05,
-      "loss": 0.8438,
-      "step": 10905
-    },
-    {
-      "epoch": 0.5057950857672694,
-      "grad_norm": 3.793743848800659,
-      "learning_rate": 1.153692662627147e-05,
-      "loss": 1.0021,
-      "step": 10910
-    },
-    {
-      "epoch": 0.5060268891979601,
-      "grad_norm": 3.819781541824341,
-      "learning_rate": 1.1528930794987329e-05,
-      "loss": 1.0518,
-      "step": 10915
-    },
-    {
-      "epoch": 0.5062586926286509,
-      "grad_norm": 3.5014266967773438,
-      "learning_rate": 1.1520933962684353e-05,
-      "loss": 0.7188,
-      "step": 10920
-    },
-    {
-      "epoch": 0.5064904960593417,
-      "grad_norm": 4.050322532653809,
-      "learning_rate": 1.1512936134598214e-05,
-      "loss": 0.9117,
-      "step": 10925
-    },
-    {
-      "epoch": 0.5067222994900324,
-      "grad_norm": 4.244816303253174,
-      "learning_rate": 1.1504937315965233e-05,
-      "loss": 0.9438,
-      "step": 10930
-    },
-    {
-      "epoch": 0.5069541029207232,
-      "grad_norm": 4.170376300811768,
-      "learning_rate": 1.1496937512022388e-05,
-      "loss": 1.0061,
-      "step": 10935
-    },
-    {
-      "epoch": 0.507185906351414,
-      "grad_norm": 3.4303226470947266,
-      "learning_rate": 1.148893672800729e-05,
-      "loss": 0.8894,
-      "step": 10940
-    },
-    {
-      "epoch": 0.5074177097821048,
-      "grad_norm": 4.357320308685303,
-      "learning_rate": 1.1480934969158197e-05,
-      "loss": 1.0696,
-      "step": 10945
-    },
-    {
-      "epoch": 0.5076495132127955,
-      "grad_norm": 3.495048999786377,
-      "learning_rate": 1.1472932240714017e-05,
-      "loss": 0.8701,
-      "step": 10950
-    },
-    {
-      "epoch": 0.5078813166434863,
-      "grad_norm": 3.7668962478637695,
-      "learning_rate": 1.1464928547914267e-05,
-      "loss": 0.8798,
-      "step": 10955
-    },
-    {
-      "epoch": 0.5081131200741771,
-      "grad_norm": 3.035940647125244,
-      "learning_rate": 1.1456923895999124e-05,
-      "loss": 0.8343,
-      "step": 10960
-    },
-    {
-      "epoch": 0.5083449235048678,
-      "grad_norm": 4.260811805725098,
-      "learning_rate": 1.144891829020937e-05,
-      "loss": 0.855,
-      "step": 10965
-    },
-    {
-      "epoch": 0.5085767269355587,
-      "grad_norm": 3.065490484237671,
-      "learning_rate": 1.1440911735786426e-05,
-      "loss": 1.0546,
-      "step": 10970
-    },
-    {
-      "epoch": 0.5088085303662494,
-      "grad_norm": 3.277255058288574,
-      "learning_rate": 1.1432904237972329e-05,
-      "loss": 0.9556,
-      "step": 10975
-    },
-    {
-      "epoch": 0.5090403337969402,
-      "grad_norm": 3.772494077682495,
-      "learning_rate": 1.142489580200973e-05,
-      "loss": 0.9838,
-      "step": 10980
-    },
-    {
-      "epoch": 0.509272137227631,
-      "grad_norm": 3.4952893257141113,
-      "learning_rate": 1.1416886433141903e-05,
-      "loss": 0.9328,
-      "step": 10985
-    },
-    {
-      "epoch": 0.5095039406583217,
-      "grad_norm": 4.249672889709473,
-      "learning_rate": 1.1408876136612722e-05,
-      "loss": 0.8854,
-      "step": 10990
-    },
-    {
-      "epoch": 0.5097357440890126,
-      "grad_norm": 3.2859761714935303,
-      "learning_rate": 1.1400864917666679e-05,
-      "loss": 0.8774,
-      "step": 10995
-    },
-    {
-      "epoch": 0.5099675475197033,
-      "grad_norm": 3.6365582942962646,
-      "learning_rate": 1.139285278154886e-05,
-      "loss": 0.8254,
-      "step": 11000
-    },
-    {
-      "epoch": 0.5099675475197033,
-      "eval_loss": 0.9347610473632812,
-      "eval_runtime": 11.2667,
-      "eval_samples_per_second": 11.272,
-      "eval_steps_per_second": 11.272,
-      "step": 11000
-    },
-    {
-      "epoch": 0.510199350950394,
-      "grad_norm": 3.846269369125366,
-      "learning_rate": 1.138483973350496e-05,
-      "loss": 0.9616,
-      "step": 11005
-    },
-    {
-      "epoch": 0.5104311543810849,
-      "grad_norm": 3.8627161979675293,
-      "learning_rate": 1.1376825778781266e-05,
-      "loss": 1.0072,
-      "step": 11010
-    },
-    {
-      "epoch": 0.5106629578117756,
-      "grad_norm": 3.971733808517456,
-      "learning_rate": 1.136881092262466e-05,
-      "loss": 0.9459,
-      "step": 11015
-    },
-    {
-      "epoch": 0.5108947612424664,
-      "grad_norm": 3.493520498275757,
-      "learning_rate": 1.1360795170282619e-05,
-      "loss": 0.8961,
-      "step": 11020
-    },
-    {
-      "epoch": 0.5111265646731572,
-      "grad_norm": 3.7606539726257324,
-      "learning_rate": 1.1352778527003194e-05,
-      "loss": 0.9448,
-      "step": 11025
-    },
-    {
-      "epoch": 0.5113583681038479,
-      "grad_norm": 11.092731475830078,
-      "learning_rate": 1.1344760998035031e-05,
-      "loss": 1.0258,
-      "step": 11030
-    },
-    {
-      "epoch": 0.5115901715345387,
-      "grad_norm": 4.570170879364014,
-      "learning_rate": 1.1336742588627358e-05,
-      "loss": 0.8798,
-      "step": 11035
-    },
-    {
-      "epoch": 0.5118219749652295,
-      "grad_norm": 3.612354278564453,
-      "learning_rate": 1.1328723304029962e-05,
-      "loss": 0.8494,
-      "step": 11040
-    },
-    {
-      "epoch": 0.5120537783959203,
-      "grad_norm": 3.9332404136657715,
-      "learning_rate": 1.1320703149493224e-05,
-      "loss": 0.9113,
-      "step": 11045
-    },
-    {
-      "epoch": 0.512285581826611,
-      "grad_norm": 3.531489133834839,
-      "learning_rate": 1.1312682130268082e-05,
-      "loss": 0.9557,
-      "step": 11050
-    },
-    {
-      "epoch": 0.5125173852573018,
-      "grad_norm": 3.8228137493133545,
-      "learning_rate": 1.1304660251606044e-05,
-      "loss": 0.7984,
-      "step": 11055
-    },
-    {
-      "epoch": 0.5127491886879926,
-      "grad_norm": 4.060273170471191,
-      "learning_rate": 1.129663751875918e-05,
-      "loss": 0.8869,
-      "step": 11060
-    },
-    {
-      "epoch": 0.5129809921186833,
-      "grad_norm": 3.425370931625366,
-      "learning_rate": 1.1288613936980116e-05,
-      "loss": 0.6703,
-      "step": 11065
-    },
-    {
-      "epoch": 0.5132127955493742,
-      "grad_norm": 3.5954689979553223,
-      "learning_rate": 1.1280589511522041e-05,
-      "loss": 0.7522,
-      "step": 11070
-    },
-    {
-      "epoch": 0.5134445989800649,
-      "grad_norm": 3.8219289779663086,
-      "learning_rate": 1.127256424763869e-05,
-      "loss": 1.0325,
-      "step": 11075
-    },
-    {
-      "epoch": 0.5136764024107556,
-      "grad_norm": 5.733139991760254,
-      "learning_rate": 1.1264538150584356e-05,
-      "loss": 1.0223,
-      "step": 11080
-    },
-    {
-      "epoch": 0.5139082058414465,
-      "grad_norm": 4.2980170249938965,
-      "learning_rate": 1.1256511225613863e-05,
-      "loss": 0.9425,
-      "step": 11085
-    },
-    {
-      "epoch": 0.5141400092721372,
-      "grad_norm": 5.1117024421691895,
-      "learning_rate": 1.1248483477982584e-05,
-      "loss": 0.962,
-      "step": 11090
-    },
-    {
-      "epoch": 0.5143718127028281,
-      "grad_norm": 3.2566781044006348,
-      "learning_rate": 1.1240454912946442e-05,
-      "loss": 0.9778,
-      "step": 11095
-    },
-    {
-      "epoch": 0.5146036161335188,
-      "grad_norm": 4.791970252990723,
-      "learning_rate": 1.1232425535761872e-05,
-      "loss": 0.9794,
-      "step": 11100
-    },
-    {
-      "epoch": 0.5146036161335188,
-      "eval_loss": 0.9343388676643372,
-      "eval_runtime": 11.267,
-      "eval_samples_per_second": 11.272,
-      "eval_steps_per_second": 11.272,
-      "step": 11100
-    },
-    {
-      "epoch": 0.5148354195642095,
-      "grad_norm": 3.983933687210083,
-      "learning_rate": 1.1224395351685862e-05,
-      "loss": 1.025,
-      "step": 11105
-    },
-    {
-      "epoch": 0.5150672229949004,
-      "grad_norm": 3.460850954055786,
-      "learning_rate": 1.1216364365975918e-05,
-      "loss": 0.7973,
-      "step": 11110
-    },
-    {
-      "epoch": 0.5152990264255911,
-      "grad_norm": 3.478369951248169,
-      "learning_rate": 1.1208332583890072e-05,
-      "loss": 0.8024,
-      "step": 11115
-    },
-    {
-      "epoch": 0.5155308298562818,
-      "grad_norm": 3.2050466537475586,
-      "learning_rate": 1.1200300010686878e-05,
-      "loss": 0.926,
-      "step": 11120
-    },
-    {
-      "epoch": 0.5157626332869727,
-      "grad_norm": 3.6445305347442627,
-      "learning_rate": 1.1192266651625406e-05,
-      "loss": 1.006,
-      "step": 11125
-    },
-    {
-      "epoch": 0.5159944367176634,
-      "grad_norm": 3.9382009506225586,
-      "learning_rate": 1.1184232511965244e-05,
-      "loss": 0.8803,
-      "step": 11130
-    },
-    {
-      "epoch": 0.5162262401483542,
-      "grad_norm": 4.49378776550293,
-      "learning_rate": 1.117619759696649e-05,
-      "loss": 1.0766,
-      "step": 11135
-    },
-    {
-      "epoch": 0.516458043579045,
-      "grad_norm": 3.932445764541626,
-      "learning_rate": 1.1168161911889753e-05,
-      "loss": 0.9086,
-      "step": 11140
-    },
-    {
-      "epoch": 0.5166898470097357,
-      "grad_norm": 3.898226022720337,
-      "learning_rate": 1.1160125461996137e-05,
-      "loss": 0.982,
-      "step": 11145
-    },
-    {
-      "epoch": 0.5169216504404265,
-      "grad_norm": 3.8211286067962646,
-      "learning_rate": 1.1152088252547249e-05,
-      "loss": 1.0433,
-      "step": 11150
-    },
-    {
-      "epoch": 0.5171534538711173,
-      "grad_norm": 3.4513728618621826,
-      "learning_rate": 1.1144050288805206e-05,
-      "loss": 0.8361,
-      "step": 11155
-    },
-    {
-      "epoch": 0.5173852573018081,
-      "grad_norm": 3.6396796703338623,
-      "learning_rate": 1.11360115760326e-05,
-      "loss": 0.8701,
-      "step": 11160
-    },
-    {
-      "epoch": 0.5176170607324988,
-      "grad_norm": 3.4153990745544434,
-      "learning_rate": 1.112797211949253e-05,
-      "loss": 0.8483,
-      "step": 11165
-    },
-    {
-      "epoch": 0.5178488641631896,
-      "grad_norm": 3.43082857131958,
-      "learning_rate": 1.1119931924448569e-05,
-      "loss": 0.7802,
-      "step": 11170
-    },
-    {
-      "epoch": 0.5180806675938804,
-      "grad_norm": 3.3968653678894043,
-      "learning_rate": 1.1111890996164782e-05,
-      "loss": 1.055,
-      "step": 11175
-    },
-    {
-      "epoch": 0.5183124710245711,
-      "grad_norm": 3.6078455448150635,
-      "learning_rate": 1.1103849339905712e-05,
-      "loss": 0.9291,
-      "step": 11180
-    },
-    {
-      "epoch": 0.518544274455262,
-      "grad_norm": 4.045047283172607,
-      "learning_rate": 1.1095806960936378e-05,
-      "loss": 1.0033,
-      "step": 11185
-    },
-    {
-      "epoch": 0.5187760778859527,
-      "grad_norm": 3.687140703201294,
-      "learning_rate": 1.1087763864522268e-05,
-      "loss": 0.7571,
-      "step": 11190
-    },
-    {
-      "epoch": 0.5190078813166434,
-      "grad_norm": 3.655649185180664,
-      "learning_rate": 1.1079720055929346e-05,
-      "loss": 0.8994,
-      "step": 11195
-    },
-    {
-      "epoch": 0.5192396847473343,
-      "grad_norm": 4.835019111633301,
-      "learning_rate": 1.107167554042404e-05,
-      "loss": 0.8423,
-      "step": 11200
-    },
-    {
-      "epoch": 0.5192396847473343,
-      "eval_loss": 0.9321886897087097,
-      "eval_runtime": 11.2611,
-      "eval_samples_per_second": 11.278,
-      "eval_steps_per_second": 11.278,
-      "step": 11200
-    },
-    {
-      "epoch": 0.519471488178025,
-      "grad_norm": 3.382221221923828,
-      "learning_rate": 1.1063630323273242e-05,
-      "loss": 0.9049,
-      "step": 11205
-    },
-    {
-      "epoch": 0.5197032916087158,
-      "grad_norm": 3.561631679534912,
-      "learning_rate": 1.10555844097443e-05,
-      "loss": 1.004,
-      "step": 11210
-    },
-    {
-      "epoch": 0.5199350950394066,
-      "grad_norm": 3.8593273162841797,
-      "learning_rate": 1.1047537805105025e-05,
-      "loss": 0.942,
-      "step": 11215
-    },
-    {
-      "epoch": 0.5201668984700973,
-      "grad_norm": 3.7144737243652344,
-      "learning_rate": 1.1039490514623665e-05,
-      "loss": 0.8805,
-      "step": 11220
-    },
-    {
-      "epoch": 0.5203987019007882,
-      "grad_norm": 4.249030590057373,
-      "learning_rate": 1.1031442543568943e-05,
-      "loss": 0.791,
-      "step": 11225
-    },
-    {
-      "epoch": 0.5206305053314789,
-      "grad_norm": 3.3096988201141357,
-      "learning_rate": 1.1023393897209996e-05,
-      "loss": 0.8479,
-      "step": 11230
-    },
-    {
-      "epoch": 0.5208623087621697,
-      "grad_norm": 4.292798042297363,
-      "learning_rate": 1.101534458081643e-05,
-      "loss": 1.0426,
-      "step": 11235
-    },
-    {
-      "epoch": 0.5210941121928605,
-      "grad_norm": 4.003423690795898,
-      "learning_rate": 1.1007294599658278e-05,
-      "loss": 0.97,
-      "step": 11240
-    },
-    {
-      "epoch": 0.5213259156235512,
-      "grad_norm": 3.7791945934295654,
-      "learning_rate": 1.0999243959006002e-05,
-      "loss": 0.8429,
-      "step": 11245
-    },
-    {
-      "epoch": 0.521557719054242,
-      "grad_norm": 4.104649543762207,
-      "learning_rate": 1.0991192664130513e-05,
-      "loss": 0.8678,
-      "step": 11250
-    },
-    {
-      "epoch": 0.5217895224849328,
-      "grad_norm": 4.052036285400391,
-      "learning_rate": 1.0983140720303136e-05,
-      "loss": 0.8715,
-      "step": 11255
-    },
-    {
-      "epoch": 0.5220213259156236,
-      "grad_norm": 3.6360881328582764,
-      "learning_rate": 1.0975088132795623e-05,
-      "loss": 0.9221,
-      "step": 11260
-    },
-    {
-      "epoch": 0.5222531293463143,
-      "grad_norm": 3.4455959796905518,
-      "learning_rate": 1.0967034906880151e-05,
-      "loss": 0.8846,
-      "step": 11265
-    },
-    {
-      "epoch": 0.5224849327770051,
-      "grad_norm": 4.016299724578857,
-      "learning_rate": 1.0958981047829317e-05,
-      "loss": 0.8542,
-      "step": 11270
-    },
-    {
-      "epoch": 0.5227167362076959,
-      "grad_norm": 3.5622754096984863,
-      "learning_rate": 1.0950926560916126e-05,
-      "loss": 0.7955,
-      "step": 11275
-    },
-    {
-      "epoch": 0.5229485396383866,
-      "grad_norm": 4.078085422515869,
-      "learning_rate": 1.0942871451413994e-05,
-      "loss": 0.9445,
-      "step": 11280
-    },
-    {
-      "epoch": 0.5231803430690775,
-      "grad_norm": 3.5976459980010986,
-      "learning_rate": 1.0934815724596759e-05,
-      "loss": 0.8949,
-      "step": 11285
-    },
-    {
-      "epoch": 0.5234121464997682,
-      "grad_norm": 4.441613674163818,
-      "learning_rate": 1.092675938573864e-05,
-      "loss": 0.8814,
-      "step": 11290
-    },
-    {
-      "epoch": 0.5236439499304589,
-      "grad_norm": 4.4883904457092285,
-      "learning_rate": 1.0918702440114272e-05,
-      "loss": 0.8916,
-      "step": 11295
-    },
-    {
-      "epoch": 0.5238757533611498,
-      "grad_norm": 3.8522751331329346,
-      "learning_rate": 1.091064489299869e-05,
-      "loss": 0.8884,
-      "step": 11300
-    },
-    {
-      "epoch": 0.5238757533611498,
-      "eval_loss": 0.9312301874160767,
-      "eval_runtime": 11.2698,
-      "eval_samples_per_second": 11.269,
-      "eval_steps_per_second": 11.269,
-      "step": 11300
-    },
-    {
-      "epoch": 0.5241075567918405,
-      "grad_norm": 3.6490557193756104,
-      "learning_rate": 1.0902586749667308e-05,
-      "loss": 0.8504,
-      "step": 11305
-    },
-    {
-      "epoch": 0.5243393602225312,
-      "grad_norm": 3.075930595397949,
-      "learning_rate": 1.0894528015395946e-05,
-      "loss": 0.7918,
-      "step": 11310
-    },
-    {
-      "epoch": 0.5245711636532221,
-      "grad_norm": 3.6393370628356934,
-      "learning_rate": 1.0886468695460802e-05,
-      "loss": 1.0254,
-      "step": 11315
-    },
-    {
-      "epoch": 0.5248029670839128,
-      "grad_norm": 4.083436012268066,
-      "learning_rate": 1.0878408795138458e-05,
-      "loss": 0.7521,
-      "step": 11320
-    },
-    {
-      "epoch": 0.5250347705146036,
-      "grad_norm": 4.432519912719727,
-      "learning_rate": 1.087034831970588e-05,
-      "loss": 0.948,
-      "step": 11325
-    },
-    {
-      "epoch": 0.5252665739452944,
-      "grad_norm": 3.2188823223114014,
-      "learning_rate": 1.0862287274440407e-05,
-      "loss": 0.9685,
-      "step": 11330
-    },
-    {
-      "epoch": 0.5254983773759851,
-      "grad_norm": 3.748762607574463,
-      "learning_rate": 1.085422566461975e-05,
-      "loss": 0.9177,
-      "step": 11335
-    },
-    {
-      "epoch": 0.5257301808066759,
-      "grad_norm": 3.7646892070770264,
-      "learning_rate": 1.0846163495521992e-05,
-      "loss": 0.9478,
-      "step": 11340
-    },
-    {
-      "epoch": 0.5259619842373667,
-      "grad_norm": 3.686171531677246,
-      "learning_rate": 1.0838100772425591e-05,
-      "loss": 0.975,
-      "step": 11345
-    },
-    {
-      "epoch": 0.5261937876680575,
-      "grad_norm": 5.668167591094971,
-      "learning_rate": 1.0830037500609348e-05,
-      "loss": 0.9199,
-      "step": 11350
-    },
-    {
-      "epoch": 0.5264255910987483,
-      "grad_norm": 3.6817402839660645,
-      "learning_rate": 1.0821973685352432e-05,
-      "loss": 0.9086,
-      "step": 11355
-    },
-    {
-      "epoch": 0.526657394529439,
-      "grad_norm": 3.8931546211242676,
-      "learning_rate": 1.0813909331934383e-05,
-      "loss": 0.879,
-      "step": 11360
-    },
-    {
-      "epoch": 0.5268891979601298,
-      "grad_norm": 3.524632215499878,
-      "learning_rate": 1.0805844445635065e-05,
-      "loss": 0.7434,
-      "step": 11365
-    },
-    {
-      "epoch": 0.5271210013908206,
-      "grad_norm": 3.135315418243408,
-      "learning_rate": 1.0797779031734716e-05,
-      "loss": 0.785,
-      "step": 11370
-    },
-    {
-      "epoch": 0.5273528048215114,
-      "grad_norm": 4.795115947723389,
-      "learning_rate": 1.0789713095513905e-05,
-      "loss": 0.9284,
-      "step": 11375
-    },
-    {
-      "epoch": 0.5275846082522021,
-      "grad_norm": 3.631051540374756,
-      "learning_rate": 1.078164664225355e-05,
-      "loss": 0.8551,
-      "step": 11380
-    },
-    {
-      "epoch": 0.527816411682893,
-      "grad_norm": 3.488373279571533,
-      "learning_rate": 1.0773579677234904e-05,
-      "loss": 0.8433,
-      "step": 11385
-    },
-    {
-      "epoch": 0.5280482151135837,
-      "grad_norm": 3.9285202026367188,
-      "learning_rate": 1.0765512205739554e-05,
-      "loss": 0.8897,
-      "step": 11390
-    },
-    {
-      "epoch": 0.5282800185442744,
-      "grad_norm": 3.3522801399230957,
-      "learning_rate": 1.0757444233049421e-05,
-      "loss": 0.8434,
-      "step": 11395
-    },
-    {
-      "epoch": 0.5285118219749653,
-      "grad_norm": 3.6392805576324463,
-      "learning_rate": 1.0749375764446756e-05,
-      "loss": 0.8998,
-      "step": 11400
-    },
-    {
-      "epoch": 0.5285118219749653,
-      "eval_loss": 0.9297853112220764,
-      "eval_runtime": 11.2781,
-      "eval_samples_per_second": 11.261,
-      "eval_steps_per_second": 11.261,
-      "step": 11400
-    },
-    {
-      "epoch": 0.528743625405656,
-      "grad_norm": 3.962904930114746,
-      "learning_rate": 1.074130680521413e-05,
-      "loss": 0.8843,
-      "step": 11405
-    },
-    {
-      "epoch": 0.5289754288363467,
-      "grad_norm": 4.593536853790283,
-      "learning_rate": 1.073323736063444e-05,
-      "loss": 0.8648,
-      "step": 11410
-    },
-    {
-      "epoch": 0.5292072322670376,
-      "grad_norm": 4.164706230163574,
-      "learning_rate": 1.0725167435990894e-05,
-      "loss": 0.8713,
-      "step": 11415
-    },
-    {
-      "epoch": 0.5294390356977283,
-      "grad_norm": 4.113262176513672,
-      "learning_rate": 1.0717097036567025e-05,
-      "loss": 0.8647,
-      "step": 11420
-    },
-    {
-      "epoch": 0.5296708391284191,
-      "grad_norm": 4.011596202850342,
-      "learning_rate": 1.0709026167646661e-05,
-      "loss": 0.9141,
-      "step": 11425
-    },
-    {
-      "epoch": 0.5299026425591099,
-      "grad_norm": 4.122656345367432,
-      "learning_rate": 1.0700954834513957e-05,
-      "loss": 0.8079,
-      "step": 11430
-    },
-    {
-      "epoch": 0.5301344459898006,
-      "grad_norm": 3.779600143432617,
-      "learning_rate": 1.069288304245336e-05,
-      "loss": 0.8531,
-      "step": 11435
-    },
-    {
-      "epoch": 0.5303662494204914,
-      "grad_norm": 3.971891164779663,
-      "learning_rate": 1.0684810796749611e-05,
-      "loss": 0.909,
-      "step": 11440
-    },
-    {
-      "epoch": 0.5305980528511822,
-      "grad_norm": 8.932957649230957,
-      "learning_rate": 1.0676738102687767e-05,
-      "loss": 1.0013,
-      "step": 11445
-    },
-    {
-      "epoch": 0.530829856281873,
-      "grad_norm": 3.974130153656006,
-      "learning_rate": 1.0668664965553156e-05,
-      "loss": 1.0049,
-      "step": 11450
-    },
-    {
-      "epoch": 0.5310616597125637,
-      "grad_norm": 3.069875478744507,
-      "learning_rate": 1.066059139063142e-05,
-      "loss": 0.9278,
-      "step": 11455
-    },
-    {
-      "epoch": 0.5312934631432545,
-      "grad_norm": 3.6674485206604004,
-      "learning_rate": 1.065251738320847e-05,
-      "loss": 0.707,
-      "step": 11460
-    },
-    {
-      "epoch": 0.5315252665739453,
-      "grad_norm": 3.577423334121704,
-      "learning_rate": 1.0644442948570506e-05,
-      "loss": 0.8667,
-      "step": 11465
-    },
-    {
-      "epoch": 0.5317570700046361,
-      "grad_norm": 4.023151874542236,
-      "learning_rate": 1.0636368092004005e-05,
-      "loss": 1.0108,
-      "step": 11470
-    },
-    {
-      "epoch": 0.5319888734353269,
-      "grad_norm": 3.7723028659820557,
-      "learning_rate": 1.0628292818795726e-05,
-      "loss": 0.9597,
-      "step": 11475
-    },
-    {
-      "epoch": 0.5322206768660176,
-      "grad_norm": 3.417649984359741,
-      "learning_rate": 1.0620217134232698e-05,
-      "loss": 0.8539,
-      "step": 11480
-    },
-    {
-      "epoch": 0.5324524802967084,
-      "grad_norm": 4.218043804168701,
-      "learning_rate": 1.0612141043602215e-05,
-      "loss": 0.8892,
-      "step": 11485
-    },
-    {
-      "epoch": 0.5326842837273992,
-      "grad_norm": 4.164142608642578,
-      "learning_rate": 1.0604064552191846e-05,
-      "loss": 0.9493,
-      "step": 11490
-    },
-    {
-      "epoch": 0.5329160871580899,
-      "grad_norm": 5.829555511474609,
-      "learning_rate": 1.059598766528941e-05,
-      "loss": 0.908,
-      "step": 11495
-    },
-    {
-      "epoch": 0.5331478905887808,
-      "grad_norm": 4.448476791381836,
-      "learning_rate": 1.0587910388182995e-05,
-      "loss": 0.8947,
-      "step": 11500
-    },
-    {
-      "epoch": 0.5331478905887808,
-      "eval_loss": 0.9298225045204163,
-      "eval_runtime": 11.2793,
-      "eval_samples_per_second": 11.26,
-      "eval_steps_per_second": 11.26,
-      "step": 11500
-    },
-    {
-      "epoch": 0.5333796940194715,
-      "grad_norm": 3.3360562324523926,
-      "learning_rate": 1.0579832726160948e-05,
-      "loss": 0.7823,
-      "step": 11505
-    },
-    {
-      "epoch": 0.5336114974501622,
-      "grad_norm": 3.199880599975586,
-      "learning_rate": 1.057175468451185e-05,
-      "loss": 0.8111,
-      "step": 11510
-    },
-    {
-      "epoch": 0.5338433008808531,
-      "grad_norm": 3.1434779167175293,
-      "learning_rate": 1.0563676268524551e-05,
-      "loss": 0.6705,
-      "step": 11515
-    },
-    {
-      "epoch": 0.5340751043115438,
-      "grad_norm": 3.9120616912841797,
-      "learning_rate": 1.0555597483488132e-05,
-      "loss": 1.1042,
-      "step": 11520
-    },
-    {
-      "epoch": 0.5343069077422345,
-      "grad_norm": 3.9502615928649902,
-      "learning_rate": 1.0547518334691924e-05,
-      "loss": 0.9414,
-      "step": 11525
-    },
-    {
-      "epoch": 0.5345387111729254,
-      "grad_norm": 3.2604763507843018,
-      "learning_rate": 1.053943882742549e-05,
-      "loss": 0.7735,
-      "step": 11530
-    },
-    {
-      "epoch": 0.5347705146036161,
-      "grad_norm": 3.5365970134735107,
-      "learning_rate": 1.053135896697863e-05,
-      "loss": 0.7848,
-      "step": 11535
-    },
-    {
-      "epoch": 0.5350023180343069,
-      "grad_norm": 3.263577461242676,
-      "learning_rate": 1.052327875864138e-05,
-      "loss": 0.7864,
-      "step": 11540
-    },
-    {
-      "epoch": 0.5352341214649977,
-      "grad_norm": 3.6061646938323975,
-      "learning_rate": 1.0515198207703997e-05,
-      "loss": 0.8842,
-      "step": 11545
-    },
-    {
-      "epoch": 0.5354659248956884,
-      "grad_norm": 3.667140245437622,
-      "learning_rate": 1.0507117319456965e-05,
-      "loss": 0.9251,
-      "step": 11550
-    },
-    {
-      "epoch": 0.5356977283263792,
-      "grad_norm": 3.3890786170959473,
-      "learning_rate": 1.0499036099190986e-05,
-      "loss": 0.8562,
-      "step": 11555
-    },
-    {
-      "epoch": 0.53592953175707,
-      "grad_norm": 3.4232475757598877,
-      "learning_rate": 1.0490954552196986e-05,
-      "loss": 0.7744,
-      "step": 11560
-    },
-    {
-      "epoch": 0.5361613351877608,
-      "grad_norm": 4.001134395599365,
-      "learning_rate": 1.0482872683766102e-05,
-      "loss": 0.7914,
-      "step": 11565
-    },
-    {
-      "epoch": 0.5363931386184515,
-      "grad_norm": 3.9395151138305664,
-      "learning_rate": 1.0474790499189675e-05,
-      "loss": 0.9072,
-      "step": 11570
-    },
-    {
-      "epoch": 0.5366249420491424,
-      "grad_norm": 4.937548637390137,
-      "learning_rate": 1.0466708003759262e-05,
-      "loss": 0.8502,
-      "step": 11575
-    },
-    {
-      "epoch": 0.5368567454798331,
-      "grad_norm": 4.098718166351318,
-      "learning_rate": 1.0458625202766621e-05,
-      "loss": 0.8955,
-      "step": 11580
-    },
-    {
-      "epoch": 0.5370885489105238,
-      "grad_norm": 6.9161906242370605,
-      "learning_rate": 1.0450542101503706e-05,
-      "loss": 0.9945,
-      "step": 11585
-    },
-    {
-      "epoch": 0.5373203523412147,
-      "grad_norm": 3.356679677963257,
-      "learning_rate": 1.0442458705262675e-05,
-      "loss": 0.8751,
-      "step": 11590
-    },
-    {
-      "epoch": 0.5375521557719054,
-      "grad_norm": 3.306637763977051,
-      "learning_rate": 1.043437501933587e-05,
-      "loss": 0.7993,
-      "step": 11595
-    },
-    {
-      "epoch": 0.5377839592025963,
-      "grad_norm": 4.0915961265563965,
-      "learning_rate": 1.0426291049015829e-05,
-      "loss": 0.6809,
-      "step": 11600
-    },
-    {
-      "epoch": 0.5377839592025963,
-      "eval_loss": 0.9296634793281555,
-      "eval_runtime": 11.2604,
-      "eval_samples_per_second": 11.278,
-      "eval_steps_per_second": 11.278,
-      "step": 11600
-    },
-    {
-      "epoch": 0.538015762633287,
-      "grad_norm": 3.1307010650634766,
-      "learning_rate": 1.0418206799595275e-05,
-      "loss": 0.7849,
-      "step": 11605
-    },
-    {
-      "epoch": 0.5382475660639777,
-      "grad_norm": 3.243574380874634,
-      "learning_rate": 1.0410122276367114e-05,
-      "loss": 0.7722,
-      "step": 11610
-    },
-    {
-      "epoch": 0.5384793694946686,
-      "grad_norm": 3.598766326904297,
-      "learning_rate": 1.0402037484624428e-05,
-      "loss": 0.9738,
-      "step": 11615
-    },
-    {
-      "epoch": 0.5387111729253593,
-      "grad_norm": 3.394896984100342,
-      "learning_rate": 1.0393952429660479e-05,
-      "loss": 0.9443,
-      "step": 11620
-    },
-    {
-      "epoch": 0.53894297635605,
-      "grad_norm": 3.3461711406707764,
-      "learning_rate": 1.0385867116768702e-05,
-      "loss": 0.7564,
-      "step": 11625
-    },
-    {
-      "epoch": 0.5391747797867409,
-      "grad_norm": 3.3700802326202393,
-      "learning_rate": 1.0377781551242693e-05,
-      "loss": 0.9093,
-      "step": 11630
-    },
-    {
-      "epoch": 0.5394065832174316,
-      "grad_norm": 3.464348554611206,
-      "learning_rate": 1.0369695738376226e-05,
-      "loss": 0.929,
-      "step": 11635
-    },
-    {
-      "epoch": 0.5396383866481224,
-      "grad_norm": 4.16630744934082,
-      "learning_rate": 1.0361609683463227e-05,
-      "loss": 1.0442,
-      "step": 11640
-    },
-    {
-      "epoch": 0.5398701900788132,
-      "grad_norm": 3.8371846675872803,
-      "learning_rate": 1.035352339179778e-05,
-      "loss": 0.8529,
-      "step": 11645
-    },
-    {
-      "epoch": 0.5401019935095039,
-      "grad_norm": 3.6530447006225586,
-      "learning_rate": 1.034543686867413e-05,
-      "loss": 0.942,
-      "step": 11650
-    },
-    {
-      "epoch": 0.5403337969401947,
-      "grad_norm": 3.9036331176757812,
-      "learning_rate": 1.0337350119386673e-05,
-      "loss": 0.8973,
-      "step": 11655
-    },
-    {
-      "epoch": 0.5405656003708855,
-      "grad_norm": 3.5922040939331055,
-      "learning_rate": 1.032926314922995e-05,
-      "loss": 0.9248,
-      "step": 11660
-    },
-    {
-      "epoch": 0.5407974038015763,
-      "grad_norm": 4.615749835968018,
-      "learning_rate": 1.0321175963498646e-05,
-      "loss": 0.97,
-      "step": 11665
-    },
-    {
-      "epoch": 0.541029207232267,
-      "grad_norm": 3.7986438274383545,
-      "learning_rate": 1.0313088567487589e-05,
-      "loss": 0.8226,
-      "step": 11670
-    },
-    {
-      "epoch": 0.5412610106629578,
-      "grad_norm": 3.754406452178955,
-      "learning_rate": 1.0305000966491746e-05,
-      "loss": 0.976,
-      "step": 11675
-    },
-    {
-      "epoch": 0.5414928140936486,
-      "grad_norm": 3.355897903442383,
-      "learning_rate": 1.0296913165806215e-05,
-      "loss": 0.8337,
-      "step": 11680
-    },
-    {
-      "epoch": 0.5417246175243393,
-      "grad_norm": 3.912876844406128,
-      "learning_rate": 1.0288825170726227e-05,
-      "loss": 0.9295,
-      "step": 11685
-    },
-    {
-      "epoch": 0.5419564209550302,
-      "grad_norm": 3.975902557373047,
-      "learning_rate": 1.0280736986547137e-05,
-      "loss": 1.0297,
-      "step": 11690
-    },
-    {
-      "epoch": 0.5421882243857209,
-      "grad_norm": 3.9908456802368164,
-      "learning_rate": 1.0272648618564436e-05,
-      "loss": 0.9178,
-      "step": 11695
-    },
-    {
-      "epoch": 0.5424200278164116,
-      "grad_norm": 3.790532350540161,
-      "learning_rate": 1.0264560072073716e-05,
-      "loss": 0.7329,
-      "step": 11700
-    },
-    {
-      "epoch": 0.5424200278164116,
-      "eval_loss": 0.926837682723999,
-      "eval_runtime": 11.2701,
-      "eval_samples_per_second": 11.269,
-      "eval_steps_per_second": 11.269,
-      "step": 11700
-    },
-    {
-      "epoch": 0.5426518312471025,
-      "grad_norm": 3.8717010021209717,
-      "learning_rate": 1.0256471352370693e-05,
-      "loss": 1.0964,
-      "step": 11705
-    },
-    {
-      "epoch": 0.5428836346777932,
-      "grad_norm": 3.893839120864868,
-      "learning_rate": 1.0248382464751214e-05,
-      "loss": 0.9117,
-      "step": 11710
-    },
-    {
-      "epoch": 0.5431154381084841,
-      "grad_norm": 3.5871944427490234,
-      "learning_rate": 1.0240293414511207e-05,
-      "loss": 0.9653,
-      "step": 11715
-    },
-    {
-      "epoch": 0.5433472415391748,
-      "grad_norm": 3.3438005447387695,
-      "learning_rate": 1.0232204206946725e-05,
-      "loss": 0.7898,
-      "step": 11720
-    },
-    {
-      "epoch": 0.5435790449698655,
-      "grad_norm": 3.8526430130004883,
-      "learning_rate": 1.0224114847353921e-05,
-      "loss": 0.9564,
-      "step": 11725
-    },
-    {
-      "epoch": 0.5438108484005564,
-      "grad_norm": 4.089172840118408,
-      "learning_rate": 1.0216025341029046e-05,
-      "loss": 0.9941,
-      "step": 11730
-    },
-    {
-      "epoch": 0.5440426518312471,
-      "grad_norm": 4.1344451904296875,
-      "learning_rate": 1.0207935693268444e-05,
-      "loss": 0.8658,
-      "step": 11735
-    },
-    {
-      "epoch": 0.5442744552619379,
-      "grad_norm": 4.014649868011475,
-      "learning_rate": 1.019984590936856e-05,
-      "loss": 1.0357,
-      "step": 11740
-    },
-    {
-      "epoch": 0.5445062586926287,
-      "grad_norm": 3.6267828941345215,
-      "learning_rate": 1.0191755994625917e-05,
-      "loss": 0.787,
-      "step": 11745
-    },
-    {
-      "epoch": 0.5447380621233194,
-      "grad_norm": 3.1016199588775635,
-      "learning_rate": 1.0183665954337134e-05,
-      "loss": 0.8991,
-      "step": 11750
-    },
-    {
-      "epoch": 0.5449698655540102,
-      "grad_norm": 4.051199913024902,
-      "learning_rate": 1.0175575793798906e-05,
-      "loss": 0.9813,
-      "step": 11755
-    },
-    {
-      "epoch": 0.545201668984701,
-      "grad_norm": 3.9017701148986816,
-      "learning_rate": 1.0167485518308006e-05,
-      "loss": 0.9734,
-      "step": 11760
-    },
-    {
-      "epoch": 0.5454334724153918,
-      "grad_norm": 4.230993747711182,
-      "learning_rate": 1.0159395133161287e-05,
-      "loss": 1.1052,
-      "step": 11765
-    },
-    {
-      "epoch": 0.5456652758460825,
-      "grad_norm": 3.363537549972534,
-      "learning_rate": 1.0151304643655674e-05,
-      "loss": 0.884,
-      "step": 11770
-    },
-    {
-      "epoch": 0.5458970792767733,
-      "grad_norm": 3.54296612739563,
-      "learning_rate": 1.0143214055088153e-05,
-      "loss": 0.8351,
-      "step": 11775
-    },
-    {
-      "epoch": 0.5461288827074641,
-      "grad_norm": 3.821948766708374,
-      "learning_rate": 1.0135123372755785e-05,
-      "loss": 0.9147,
-      "step": 11780
-    },
-    {
-      "epoch": 0.5463606861381548,
-      "grad_norm": 4.290584087371826,
-      "learning_rate": 1.0127032601955682e-05,
-      "loss": 0.9966,
-      "step": 11785
-    },
-    {
-      "epoch": 0.5465924895688457,
-      "grad_norm": 4.043396949768066,
-      "learning_rate": 1.011894174798502e-05,
-      "loss": 1.0158,
-      "step": 11790
-    },
-    {
-      "epoch": 0.5468242929995364,
-      "grad_norm": 3.4050636291503906,
-      "learning_rate": 1.011085081614103e-05,
-      "loss": 0.9174,
-      "step": 11795
-    },
-    {
-      "epoch": 0.5470560964302271,
-      "grad_norm": 3.734877109527588,
-      "learning_rate": 1.0102759811720995e-05,
-      "loss": 0.9592,
-      "step": 11800
-    },
-    {
-      "epoch": 0.5470560964302271,
-      "eval_loss": 0.9257230758666992,
-      "eval_runtime": 11.2689,
-      "eval_samples_per_second": 11.27,
-      "eval_steps_per_second": 11.27,
-      "step": 11800
-    },
-    {
-      "epoch": 0.547287899860918,
-      "grad_norm": 4.463562488555908,
-      "learning_rate": 1.0094668740022236e-05,
-      "loss": 0.7343,
-      "step": 11805
-    },
-    {
-      "epoch": 0.5475197032916087,
-      "grad_norm": 3.9908478260040283,
-      "learning_rate": 1.0086577606342131e-05,
-      "loss": 0.9073,
-      "step": 11810
-    },
-    {
-      "epoch": 0.5477515067222994,
-      "grad_norm": 3.766899824142456,
-      "learning_rate": 1.007848641597809e-05,
-      "loss": 0.8451,
-      "step": 11815
-    },
-    {
-      "epoch": 0.5479833101529903,
-      "grad_norm": 3.727322578430176,
-      "learning_rate": 1.0070395174227563e-05,
-      "loss": 0.7368,
-      "step": 11820
-    },
-    {
-      "epoch": 0.548215113583681,
-      "grad_norm": 3.233952522277832,
-      "learning_rate": 1.0062303886388031e-05,
-      "loss": 0.8511,
-      "step": 11825
-    },
-    {
-      "epoch": 0.5484469170143718,
-      "grad_norm": 3.7264208793640137,
-      "learning_rate": 1.0054212557757015e-05,
-      "loss": 0.7904,
-      "step": 11830
-    },
-    {
-      "epoch": 0.5486787204450626,
-      "grad_norm": 3.164869785308838,
-      "learning_rate": 1.0046121193632046e-05,
-      "loss": 0.8442,
-      "step": 11835
-    },
-    {
-      "epoch": 0.5489105238757533,
-      "grad_norm": 3.574117660522461,
-      "learning_rate": 1.0038029799310691e-05,
-      "loss": 1.0224,
-      "step": 11840
-    },
-    {
-      "epoch": 0.5491423273064442,
-      "grad_norm": 3.944805145263672,
-      "learning_rate": 1.002993838009054e-05,
-      "loss": 0.8456,
-      "step": 11845
-    },
-    {
-      "epoch": 0.5493741307371349,
-      "grad_norm": 3.691657543182373,
-      "learning_rate": 1.002184694126918e-05,
-      "loss": 0.988,
-      "step": 11850
-    },
-    {
-      "epoch": 0.5496059341678257,
-      "grad_norm": 3.9201769828796387,
-      "learning_rate": 1.0013755488144233e-05,
-      "loss": 0.8129,
-      "step": 11855
-    },
-    {
-      "epoch": 0.5498377375985165,
-      "grad_norm": 3.153733015060425,
-      "learning_rate": 1.0005664026013315e-05,
-      "loss": 0.8401,
-      "step": 11860
-    },
-    {
-      "epoch": 0.5500695410292072,
-      "grad_norm": 3.8798153400421143,
-      "learning_rate": 9.99757256017406e-06,
-      "loss": 0.8851,
-      "step": 11865
-    },
-    {
-      "epoch": 0.550301344459898,
-      "grad_norm": 2.954413414001465,
-      "learning_rate": 9.989481095924089e-06,
-      "loss": 0.9418,
-      "step": 11870
-    },
-    {
-      "epoch": 0.5505331478905888,
-      "grad_norm": 4.20355749130249,
-      "learning_rate": 9.981389638561038e-06,
-      "loss": 1.0055,
-      "step": 11875
-    },
-    {
-      "epoch": 0.5507649513212796,
-      "grad_norm": 3.426891565322876,
-      "learning_rate": 9.97329819338253e-06,
-      "loss": 0.8393,
-      "step": 11880
-    },
-    {
-      "epoch": 0.5509967547519703,
-      "grad_norm": 3.606313705444336,
-      "learning_rate": 9.96520676568618e-06,
-      "loss": 0.7815,
-      "step": 11885
-    },
-    {
-      "epoch": 0.5512285581826611,
-      "grad_norm": 4.329439640045166,
-      "learning_rate": 9.957115360769595e-06,
-      "loss": 0.9814,
-      "step": 11890
-    },
-    {
-      "epoch": 0.5514603616133519,
-      "grad_norm": 3.6122066974639893,
-      "learning_rate": 9.949023983930365e-06,
-      "loss": 0.9594,
-      "step": 11895
-    },
-    {
-      "epoch": 0.5516921650440426,
-      "grad_norm": 4.180810451507568,
-      "learning_rate": 9.940932640466063e-06,
-      "loss": 0.8931,
-      "step": 11900
-    },
-    {
-      "epoch": 0.5516921650440426,
-      "eval_loss": 0.9250540137290955,
-      "eval_runtime": 11.2655,
-      "eval_samples_per_second": 11.273,
-      "eval_steps_per_second": 11.273,
-      "step": 11900
-    },
-    {
-      "epoch": 0.5519239684747335,
-      "grad_norm": 4.088957786560059,
-      "learning_rate": 9.932841335674234e-06,
-      "loss": 1.0179,
-      "step": 11905
-    },
-    {
-      "epoch": 0.5521557719054242,
-      "grad_norm": 3.0207998752593994,
-      "learning_rate": 9.924750074852403e-06,
-      "loss": 0.7174,
-      "step": 11910
-    },
-    {
-      "epoch": 0.5523875753361149,
-      "grad_norm": 3.4280521869659424,
-      "learning_rate": 9.916658863298072e-06,
-      "loss": 0.9469,
-      "step": 11915
-    },
-    {
-      "epoch": 0.5526193787668058,
-      "grad_norm": 3.119837522506714,
-      "learning_rate": 9.908567706308701e-06,
-      "loss": 0.7976,
-      "step": 11920
-    },
-    {
-      "epoch": 0.5528511821974965,
-      "grad_norm": 3.94305419921875,
-      "learning_rate": 9.900476609181719e-06,
-      "loss": 1.062,
-      "step": 11925
-    },
-    {
-      "epoch": 0.5530829856281873,
-      "grad_norm": 3.694404363632202,
-      "learning_rate": 9.89238557721451e-06,
-      "loss": 0.9182,
-      "step": 11930
-    },
-    {
-      "epoch": 0.5533147890588781,
-      "grad_norm": 4.300608158111572,
-      "learning_rate": 9.884294615704426e-06,
-      "loss": 0.9331,
-      "step": 11935
-    },
-    {
-      "epoch": 0.5535465924895688,
-      "grad_norm": 3.950634479522705,
-      "learning_rate": 9.876203729948762e-06,
-      "loss": 0.8561,
-      "step": 11940
-    },
-    {
-      "epoch": 0.5537783959202596,
-      "grad_norm": 3.2765071392059326,
-      "learning_rate": 9.86811292524477e-06,
-      "loss": 0.7449,
-      "step": 11945
-    },
-    {
-      "epoch": 0.5540101993509504,
-      "grad_norm": 3.229647636413574,
-      "learning_rate": 9.86002220688965e-06,
-      "loss": 0.8359,
-      "step": 11950
-    },
-    {
-      "epoch": 0.5542420027816412,
-      "grad_norm": 4.144287109375,
-      "learning_rate": 9.851931580180543e-06,
-      "loss": 0.7521,
-      "step": 11955
-    },
-    {
-      "epoch": 0.5544738062123319,
-      "grad_norm": 3.954238176345825,
-      "learning_rate": 9.843841050414523e-06,
-      "loss": 0.8984,
-      "step": 11960
-    },
-    {
-      "epoch": 0.5547056096430227,
-      "grad_norm": 4.321442127227783,
-      "learning_rate": 9.835750622888612e-06,
-      "loss": 0.9637,
-      "step": 11965
-    },
-    {
-      "epoch": 0.5549374130737135,
-      "grad_norm": 3.347836971282959,
-      "learning_rate": 9.827660302899758e-06,
-      "loss": 0.8953,
-      "step": 11970
-    },
-    {
-      "epoch": 0.5551692165044043,
-      "grad_norm": 5.19721794128418,
-      "learning_rate": 9.819570095744843e-06,
-      "loss": 0.9844,
-      "step": 11975
-    },
-    {
-      "epoch": 0.555401019935095,
-      "grad_norm": 3.556494951248169,
-      "learning_rate": 9.811480006720674e-06,
-      "loss": 0.8301,
-      "step": 11980
-    },
-    {
-      "epoch": 0.5556328233657858,
-      "grad_norm": 3.1306145191192627,
-      "learning_rate": 9.80339004112398e-06,
-      "loss": 0.8594,
-      "step": 11985
-    },
-    {
-      "epoch": 0.5558646267964766,
-      "grad_norm": 3.2993454933166504,
-      "learning_rate": 9.795300204251401e-06,
-      "loss": 0.9242,
-      "step": 11990
-    },
-    {
-      "epoch": 0.5560964302271674,
-      "grad_norm": 3.9028289318084717,
-      "learning_rate": 9.787210501399507e-06,
-      "loss": 0.9108,
-      "step": 11995
-    },
-    {
-      "epoch": 0.5563282336578581,
-      "grad_norm": 3.538569927215576,
-      "learning_rate": 9.779120937864775e-06,
-      "loss": 0.8735,
-      "step": 12000
-    },
-    {
-      "epoch": 0.5563282336578581,
-      "eval_loss": 0.9230943322181702,
-      "eval_runtime": 11.2638,
-      "eval_samples_per_second": 11.275,
-      "eval_steps_per_second": 11.275,
-      "step": 12000
-    },
-    {
-      "epoch": 0.556560037088549,
-      "grad_norm": 3.572889566421509,
-      "learning_rate": 9.771031518943588e-06,
-      "loss": 0.9114,
-      "step": 12005
-    },
-    {
-      "epoch": 0.5567918405192397,
-      "grad_norm": 3.8642523288726807,
-      "learning_rate": 9.762942249932232e-06,
-      "loss": 0.9442,
-      "step": 12010
-    },
-    {
-      "epoch": 0.5570236439499304,
-      "grad_norm": 4.011014461517334,
-      "learning_rate": 9.754853136126907e-06,
-      "loss": 0.8789,
-      "step": 12015
-    },
-    {
-      "epoch": 0.5572554473806213,
-      "grad_norm": 3.090644359588623,
-      "learning_rate": 9.746764182823696e-06,
-      "loss": 0.9078,
-      "step": 12020
-    },
-    {
-      "epoch": 0.557487250811312,
-      "grad_norm": 4.063168048858643,
-      "learning_rate": 9.738675395318585e-06,
-      "loss": 0.9359,
-      "step": 12025
-    },
-    {
-      "epoch": 0.5577190542420027,
-      "grad_norm": 3.925661325454712,
-      "learning_rate": 9.73058677890745e-06,
-      "loss": 0.9003,
-      "step": 12030
-    },
-    {
-      "epoch": 0.5579508576726936,
-      "grad_norm": 4.365866184234619,
-      "learning_rate": 9.722498338886061e-06,
-      "loss": 1.0072,
-      "step": 12035
-    },
-    {
-      "epoch": 0.5581826611033843,
-      "grad_norm": 3.7354307174682617,
-      "learning_rate": 9.714410080550066e-06,
-      "loss": 1.0974,
-      "step": 12040
-    },
-    {
-      "epoch": 0.5584144645340751,
-      "grad_norm": 3.8771347999572754,
-      "learning_rate": 9.706322009194989e-06,
-      "loss": 0.9791,
-      "step": 12045
-    },
-    {
-      "epoch": 0.5586462679647659,
-      "grad_norm": 3.3021087646484375,
-      "learning_rate": 9.69823413011624e-06,
-      "loss": 0.8397,
-      "step": 12050
-    },
-    {
-      "epoch": 0.5588780713954566,
-      "grad_norm": 4.110974311828613,
-      "learning_rate": 9.6901464486091e-06,
-      "loss": 0.905,
-      "step": 12055
-    },
-    {
-      "epoch": 0.5591098748261474,
-      "grad_norm": 3.842860221862793,
-      "learning_rate": 9.682058969968724e-06,
-      "loss": 0.964,
-      "step": 12060
-    },
-    {
-      "epoch": 0.5593416782568382,
-      "grad_norm": 4.045979976654053,
-      "learning_rate": 9.673971699490133e-06,
-      "loss": 0.9184,
-      "step": 12065
-    },
-    {
-      "epoch": 0.559573481687529,
-      "grad_norm": 4.390701770782471,
-      "learning_rate": 9.665884642468207e-06,
-      "loss": 0.9431,
-      "step": 12070
-    },
-    {
-      "epoch": 0.5598052851182197,
-      "grad_norm": 3.036095380783081,
-      "learning_rate": 9.657797804197688e-06,
-      "loss": 0.8122,
-      "step": 12075
-    },
-    {
-      "epoch": 0.5600370885489105,
-      "grad_norm": 3.978038787841797,
-      "learning_rate": 9.64971118997318e-06,
-      "loss": 1.0041,
-      "step": 12080
-    },
-    {
-      "epoch": 0.5602688919796013,
-      "grad_norm": 3.818854570388794,
-      "learning_rate": 9.641624805089136e-06,
-      "loss": 1.0957,
-      "step": 12085
-    },
-    {
-      "epoch": 0.5605006954102921,
-      "grad_norm": 4.990757942199707,
-      "learning_rate": 9.633538654839856e-06,
-      "loss": 1.0599,
-      "step": 12090
-    },
-    {
-      "epoch": 0.5607324988409829,
-      "grad_norm": 3.926579236984253,
-      "learning_rate": 9.625452744519494e-06,
-      "loss": 0.8846,
-      "step": 12095
-    },
-    {
-      "epoch": 0.5609643022716736,
-      "grad_norm": 4.215625762939453,
-      "learning_rate": 9.617367079422044e-06,
-      "loss": 1.0451,
-      "step": 12100
-    },
-    {
-      "epoch": 0.5609643022716736,
-      "eval_loss": 0.9230334758758545,
-      "eval_runtime": 11.2713,
-      "eval_samples_per_second": 11.268,
-      "eval_steps_per_second": 11.268,
-      "step": 12100
-    },
-    {
-      "epoch": 0.5611961057023644,
-      "grad_norm": 3.8236374855041504,
-      "learning_rate": 9.609281664841333e-06,
-      "loss": 0.9031,
-      "step": 12105
-    },
-    {
-      "epoch": 0.5614279091330552,
-      "grad_norm": 4.061149597167969,
-      "learning_rate": 9.601196506071032e-06,
-      "loss": 0.7433,
-      "step": 12110
-    },
-    {
-      "epoch": 0.5616597125637459,
-      "grad_norm": 4.054765701293945,
-      "learning_rate": 9.59311160840464e-06,
-      "loss": 0.9256,
-      "step": 12115
-    },
-    {
-      "epoch": 0.5618915159944368,
-      "grad_norm": 3.500136137008667,
-      "learning_rate": 9.585026977135487e-06,
-      "loss": 0.8239,
-      "step": 12120
-    },
-    {
-      "epoch": 0.5621233194251275,
-      "grad_norm": 4.4179229736328125,
-      "learning_rate": 9.576942617556732e-06,
-      "loss": 0.9409,
-      "step": 12125
-    },
-    {
-      "epoch": 0.5623551228558182,
-      "grad_norm": 3.1021149158477783,
-      "learning_rate": 9.568858534961352e-06,
-      "loss": 0.9298,
-      "step": 12130
-    },
-    {
-      "epoch": 0.5625869262865091,
-      "grad_norm": 3.5426769256591797,
-      "learning_rate": 9.560774734642138e-06,
-      "loss": 0.9236,
-      "step": 12135
-    },
-    {
-      "epoch": 0.5628187297171998,
-      "grad_norm": 3.6595637798309326,
-      "learning_rate": 9.552691221891702e-06,
-      "loss": 0.9955,
-      "step": 12140
-    },
-    {
-      "epoch": 0.5630505331478906,
-      "grad_norm": 3.189774751663208,
-      "learning_rate": 9.544608002002471e-06,
-      "loss": 0.8035,
-      "step": 12145
-    },
-    {
-      "epoch": 0.5632823365785814,
-      "grad_norm": 4.558648586273193,
-      "learning_rate": 9.536525080266674e-06,
-      "loss": 0.9646,
-      "step": 12150
-    },
-    {
-      "epoch": 0.5635141400092721,
-      "grad_norm": 4.213756084442139,
-      "learning_rate": 9.528442461976347e-06,
-      "loss": 0.9128,
-      "step": 12155
-    },
-    {
-      "epoch": 0.5637459434399629,
-      "grad_norm": 3.8861300945281982,
-      "learning_rate": 9.52036015242333e-06,
-      "loss": 0.9403,
-      "step": 12160
-    },
-    {
-      "epoch": 0.5639777468706537,
-      "grad_norm": 3.3439478874206543,
-      "learning_rate": 9.512278156899255e-06,
-      "loss": 0.8125,
-      "step": 12165
-    },
-    {
-      "epoch": 0.5642095503013445,
-      "grad_norm": 4.4666748046875,
-      "learning_rate": 9.504196480695553e-06,
-      "loss": 0.9626,
-      "step": 12170
-    },
-    {
-      "epoch": 0.5644413537320352,
-      "grad_norm": 3.6612839698791504,
-      "learning_rate": 9.496115129103442e-06,
-      "loss": 0.8229,
-      "step": 12175
-    },
-    {
-      "epoch": 0.564673157162726,
-      "grad_norm": 3.3860251903533936,
-      "learning_rate": 9.488034107413935e-06,
-      "loss": 0.9714,
-      "step": 12180
-    },
-    {
-      "epoch": 0.5649049605934168,
-      "grad_norm": 4.411099433898926,
-      "learning_rate": 9.479953420917821e-06,
-      "loss": 0.9219,
-      "step": 12185
-    },
-    {
-      "epoch": 0.5651367640241075,
-      "grad_norm": 4.0682454109191895,
-      "learning_rate": 9.471873074905676e-06,
-      "loss": 0.9739,
-      "step": 12190
-    },
-    {
-      "epoch": 0.5653685674547984,
-      "grad_norm": 4.332896709442139,
-      "learning_rate": 9.463793074667845e-06,
-      "loss": 0.9527,
-      "step": 12195
-    },
-    {
-      "epoch": 0.5656003708854891,
-      "grad_norm": 3.3936986923217773,
-      "learning_rate": 9.455713425494449e-06,
-      "loss": 0.8179,
-      "step": 12200
-    },
-    {
-      "epoch": 0.5656003708854891,
-      "eval_loss": 0.9221216440200806,
-      "eval_runtime": 11.2783,
-      "eval_samples_per_second": 11.261,
-      "eval_steps_per_second": 11.261,
-      "step": 12200
-    },
-    {
-      "epoch": 0.5658321743161798,
-      "grad_norm": 3.9779462814331055,
-      "learning_rate": 9.447634132675389e-06,
-      "loss": 1.0487,
-      "step": 12205
-    },
-    {
-      "epoch": 0.5660639777468707,
-      "grad_norm": 3.2726247310638428,
-      "learning_rate": 9.439555201500321e-06,
-      "loss": 0.8528,
-      "step": 12210
-    },
-    {
-      "epoch": 0.5662957811775614,
-      "grad_norm": 4.14183235168457,
-      "learning_rate": 9.431476637258669e-06,
-      "loss": 0.9682,
-      "step": 12215
-    },
-    {
-      "epoch": 0.5665275846082523,
-      "grad_norm": 4.068934440612793,
-      "learning_rate": 9.423398445239618e-06,
-      "loss": 0.9826,
-      "step": 12220
-    },
-    {
-      "epoch": 0.566759388038943,
-      "grad_norm": 3.1168320178985596,
-      "learning_rate": 9.415320630732104e-06,
-      "loss": 0.9555,
-      "step": 12225
-    },
-    {
-      "epoch": 0.5669911914696337,
-      "grad_norm": 4.245713233947754,
-      "learning_rate": 9.407243199024822e-06,
-      "loss": 1.0328,
-      "step": 12230
-    },
-    {
-      "epoch": 0.5672229949003246,
-      "grad_norm": 3.5633294582366943,
-      "learning_rate": 9.399166155406209e-06,
-      "loss": 1.0491,
-      "step": 12235
-    },
-    {
-      "epoch": 0.5674547983310153,
-      "grad_norm": 3.3050804138183594,
-      "learning_rate": 9.39108950516446e-06,
-      "loss": 0.7442,
-      "step": 12240
-    },
-    {
-      "epoch": 0.567686601761706,
-      "grad_norm": 3.608316421508789,
-      "learning_rate": 9.383013253587499e-06,
-      "loss": 0.9549,
-      "step": 12245
-    },
-    {
-      "epoch": 0.5679184051923969,
-      "grad_norm": 3.2159860134124756,
-      "learning_rate": 9.374937405963002e-06,
-      "loss": 0.8389,
-      "step": 12250
-    },
-    {
-      "epoch": 0.5681502086230876,
-      "grad_norm": 4.187016010284424,
-      "learning_rate": 9.366861967578365e-06,
-      "loss": 0.8149,
-      "step": 12255
-    },
-    {
-      "epoch": 0.5683820120537784,
-      "grad_norm": 3.3455278873443604,
-      "learning_rate": 9.358786943720724e-06,
-      "loss": 0.8868,
-      "step": 12260
-    },
-    {
-      "epoch": 0.5686138154844692,
-      "grad_norm": 3.9041545391082764,
-      "learning_rate": 9.350712339676952e-06,
-      "loss": 0.9097,
-      "step": 12265
-    },
-    {
-      "epoch": 0.56884561891516,
-      "grad_norm": 3.5602474212646484,
-      "learning_rate": 9.342638160733638e-06,
-      "loss": 0.8567,
-      "step": 12270
-    },
-    {
-      "epoch": 0.5690774223458507,
-      "grad_norm": 3.781757354736328,
-      "learning_rate": 9.334564412177091e-06,
-      "loss": 1.0214,
-      "step": 12275
-    },
-    {
-      "epoch": 0.5693092257765415,
-      "grad_norm": 4.233317852020264,
-      "learning_rate": 9.32649109929334e-06,
-      "loss": 0.9249,
-      "step": 12280
-    },
-    {
-      "epoch": 0.5695410292072323,
-      "grad_norm": 3.0850915908813477,
-      "learning_rate": 9.31841822736813e-06,
-      "loss": 0.6689,
-      "step": 12285
-    },
-    {
-      "epoch": 0.569772832637923,
-      "grad_norm": 4.0126237869262695,
-      "learning_rate": 9.310345801686923e-06,
-      "loss": 0.9143,
-      "step": 12290
-    },
-    {
-      "epoch": 0.5700046360686138,
-      "grad_norm": 3.6224422454833984,
-      "learning_rate": 9.302273827534874e-06,
-      "loss": 0.7326,
-      "step": 12295
-    },
-    {
-      "epoch": 0.5702364394993046,
-      "grad_norm": 3.5666017532348633,
-      "learning_rate": 9.29420231019686e-06,
-      "loss": 0.9441,
-      "step": 12300
-    },
-    {
-      "epoch": 0.5702364394993046,
-      "eval_loss": 0.9209733009338379,
-      "eval_runtime": 11.2829,
-      "eval_samples_per_second": 11.256,
-      "eval_steps_per_second": 11.256,
-      "step": 12300
-    },
-    {
-      "epoch": 0.5704682429299953,
-      "grad_norm": 3.6667540073394775,
-      "learning_rate": 9.286131254957451e-06,
-      "loss": 1.0631,
-      "step": 12305
-    },
-    {
-      "epoch": 0.5707000463606862,
-      "grad_norm": 4.890436172485352,
-      "learning_rate": 9.278060667100907e-06,
-      "loss": 0.8892,
-      "step": 12310
-    },
-    {
-      "epoch": 0.5709318497913769,
-      "grad_norm": 4.06868839263916,
-      "learning_rate": 9.26999055191119e-06,
-      "loss": 0.9446,
-      "step": 12315
-    },
-    {
-      "epoch": 0.5711636532220676,
-      "grad_norm": 3.7493627071380615,
-      "learning_rate": 9.261920914671952e-06,
-      "loss": 0.8918,
-      "step": 12320
-    },
-    {
-      "epoch": 0.5713954566527585,
-      "grad_norm": 3.5982697010040283,
-      "learning_rate": 9.253851760666536e-06,
-      "loss": 0.8466,
-      "step": 12325
-    },
-    {
-      "epoch": 0.5716272600834492,
-      "grad_norm": 4.513077259063721,
-      "learning_rate": 9.245783095177959e-06,
-      "loss": 0.9269,
-      "step": 12330
-    },
-    {
-      "epoch": 0.5718590635141401,
-      "grad_norm": 3.5556976795196533,
-      "learning_rate": 9.237714923488929e-06,
-      "loss": 1.0331,
-      "step": 12335
-    },
-    {
-      "epoch": 0.5720908669448308,
-      "grad_norm": 3.3687074184417725,
-      "learning_rate": 9.229647250881817e-06,
-      "loss": 0.6496,
-      "step": 12340
-    },
-    {
-      "epoch": 0.5723226703755215,
-      "grad_norm": 3.337689161300659,
-      "learning_rate": 9.221580082638678e-06,
-      "loss": 0.8649,
-      "step": 12345
-    },
-    {
-      "epoch": 0.5725544738062124,
-      "grad_norm": 4.0120530128479,
-      "learning_rate": 9.213513424041236e-06,
-      "loss": 1.0376,
-      "step": 12350
-    },
-    {
-      "epoch": 0.5727862772369031,
-      "grad_norm": 4.060422897338867,
-      "learning_rate": 9.205447280370879e-06,
-      "loss": 0.9098,
-      "step": 12355
-    },
-    {
-      "epoch": 0.5730180806675939,
-      "grad_norm": 9.308208465576172,
-      "learning_rate": 9.197381656908654e-06,
-      "loss": 0.8756,
-      "step": 12360
-    },
-    {
-      "epoch": 0.5732498840982847,
-      "grad_norm": 3.8436498641967773,
-      "learning_rate": 9.189316558935282e-06,
-      "loss": 0.8995,
-      "step": 12365
-    },
-    {
-      "epoch": 0.5734816875289754,
-      "grad_norm": 3.3367080688476562,
-      "learning_rate": 9.181251991731119e-06,
-      "loss": 0.9903,
-      "step": 12370
-    },
-    {
-      "epoch": 0.5737134909596662,
-      "grad_norm": 3.9300827980041504,
-      "learning_rate": 9.173187960576189e-06,
-      "loss": 0.9093,
-      "step": 12375
-    },
-    {
-      "epoch": 0.573945294390357,
-      "grad_norm": 4.047109127044678,
-      "learning_rate": 9.165124470750153e-06,
-      "loss": 0.9035,
-      "step": 12380
-    },
-    {
-      "epoch": 0.5741770978210478,
-      "grad_norm": 3.9684131145477295,
-      "learning_rate": 9.157061527532335e-06,
-      "loss": 0.7974,
-      "step": 12385
-    },
-    {
-      "epoch": 0.5744089012517385,
-      "grad_norm": 3.2714669704437256,
-      "learning_rate": 9.148999136201686e-06,
-      "loss": 0.9637,
-      "step": 12390
-    },
-    {
-      "epoch": 0.5746407046824293,
-      "grad_norm": 4.41608190536499,
-      "learning_rate": 9.140937302036803e-06,
-      "loss": 0.8028,
-      "step": 12395
-    },
-    {
-      "epoch": 0.5748725081131201,
-      "grad_norm": 4.140214920043945,
-      "learning_rate": 9.13287603031591e-06,
-      "loss": 0.958,
-      "step": 12400
-    },
-    {
-      "epoch": 0.5748725081131201,
-      "eval_loss": 0.9194921851158142,
-      "eval_runtime": 11.2653,
-      "eval_samples_per_second": 11.274,
-      "eval_steps_per_second": 11.274,
-      "step": 12400
-    },
-    {
-      "epoch": 0.5751043115438108,
-      "grad_norm": 3.3030049800872803,
-      "learning_rate": 9.124815326316868e-06,
-      "loss": 0.9745,
-      "step": 12405
-    },
-    {
-      "epoch": 0.5753361149745017,
-      "grad_norm": 3.330787181854248,
-      "learning_rate": 9.116755195317176e-06,
-      "loss": 0.8783,
-      "step": 12410
-    },
-    {
-      "epoch": 0.5755679184051924,
-      "grad_norm": 3.8961637020111084,
-      "learning_rate": 9.10869564259394e-06,
-      "loss": 0.8882,
-      "step": 12415
-    },
-    {
-      "epoch": 0.5757997218358831,
-      "grad_norm": 3.4725751876831055,
-      "learning_rate": 9.100636673423898e-06,
-      "loss": 1.0014,
-      "step": 12420
-    },
-    {
-      "epoch": 0.576031525266574,
-      "grad_norm": 3.758121967315674,
-      "learning_rate": 9.092578293083408e-06,
-      "loss": 0.9907,
-      "step": 12425
-    },
-    {
-      "epoch": 0.5762633286972647,
-      "grad_norm": 3.889312267303467,
-      "learning_rate": 9.084520506848433e-06,
-      "loss": 0.8723,
-      "step": 12430
-    },
-    {
-      "epoch": 0.5764951321279554,
-      "grad_norm": 3.362896203994751,
-      "learning_rate": 9.076463319994554e-06,
-      "loss": 0.7634,
-      "step": 12435
-    },
-    {
-      "epoch": 0.5767269355586463,
-      "grad_norm": 3.958284854888916,
-      "learning_rate": 9.068406737796957e-06,
-      "loss": 0.804,
-      "step": 12440
-    },
-    {
-      "epoch": 0.576958738989337,
-      "grad_norm": 3.408536911010742,
-      "learning_rate": 9.060350765530436e-06,
-      "loss": 0.8166,
-      "step": 12445
-    },
-    {
-      "epoch": 0.5771905424200278,
-      "grad_norm": 3.5486555099487305,
-      "learning_rate": 9.052295408469382e-06,
-      "loss": 0.6871,
-      "step": 12450
-    },
-    {
-      "epoch": 0.5774223458507186,
-      "grad_norm": 3.405040740966797,
-      "learning_rate": 9.044240671887784e-06,
-      "loss": 0.7373,
-      "step": 12455
-    },
-    {
-      "epoch": 0.5776541492814093,
-      "grad_norm": 3.590054988861084,
-      "learning_rate": 9.036186561059222e-06,
-      "loss": 0.8471,
-      "step": 12460
-    },
-    {
-      "epoch": 0.5778859527121002,
-      "grad_norm": 3.6733016967773438,
-      "learning_rate": 9.02813308125687e-06,
-      "loss": 0.8568,
-      "step": 12465
-    },
-    {
-      "epoch": 0.5781177561427909,
-      "grad_norm": 3.4835875034332275,
-      "learning_rate": 9.020080237753489e-06,
-      "loss": 0.8211,
-      "step": 12470
-    },
-    {
-      "epoch": 0.5783495595734817,
-      "grad_norm": 3.9289448261260986,
-      "learning_rate": 9.012028035821423e-06,
-      "loss": 0.8948,
-      "step": 12475
-    },
-    {
-      "epoch": 0.5785813630041725,
-      "grad_norm": 6.000369071960449,
-      "learning_rate": 9.003976480732592e-06,
-      "loss": 0.9534,
-      "step": 12480
-    },
-    {
-      "epoch": 0.5788131664348632,
-      "grad_norm": 3.9709556102752686,
-      "learning_rate": 8.995925577758505e-06,
-      "loss": 0.8551,
-      "step": 12485
-    },
-    {
-      "epoch": 0.579044969865554,
-      "grad_norm": 3.99529767036438,
-      "learning_rate": 8.98787533217022e-06,
-      "loss": 0.732,
-      "step": 12490
-    },
-    {
-      "epoch": 0.5792767732962448,
-      "grad_norm": 3.711063861846924,
-      "learning_rate": 8.97982574923839e-06,
-      "loss": 0.9214,
-      "step": 12495
-    },
-    {
-      "epoch": 0.5795085767269356,
-      "grad_norm": 3.617279529571533,
-      "learning_rate": 8.971776834233223e-06,
-      "loss": 0.7603,
-      "step": 12500
-    },
-    {
-      "epoch": 0.5795085767269356,
-      "eval_loss": 0.9186437726020813,
-      "eval_runtime": 11.2712,
-      "eval_samples_per_second": 11.268,
-      "eval_steps_per_second": 11.268,
-      "step": 12500
-    },
-    {
-      "epoch": 0.5797403801576263,
-      "grad_norm": 3.9697864055633545,
-      "learning_rate": 8.963728592424483e-06,
-      "loss": 0.8395,
-      "step": 12505
-    },
-    {
-      "epoch": 0.5799721835883171,
-      "grad_norm": 3.5097768306732178,
-      "learning_rate": 8.955681029081512e-06,
-      "loss": 0.8685,
-      "step": 12510
-    },
-    {
-      "epoch": 0.5802039870190079,
-      "grad_norm": 3.709843635559082,
-      "learning_rate": 8.947634149473187e-06,
-      "loss": 1.0262,
-      "step": 12515
-    },
-    {
-      "epoch": 0.5804357904496986,
-      "grad_norm": 2.8888819217681885,
-      "learning_rate": 8.939587958867948e-06,
-      "loss": 0.7361,
-      "step": 12520
-    },
-    {
-      "epoch": 0.5806675938803895,
-      "grad_norm": 4.29230260848999,
-      "learning_rate": 8.931542462533783e-06,
-      "loss": 0.8786,
-      "step": 12525
-    },
-    {
-      "epoch": 0.5808993973110802,
-      "grad_norm": 4.221256256103516,
-      "learning_rate": 8.923497665738227e-06,
-      "loss": 0.9499,
-      "step": 12530
-    },
-    {
-      "epoch": 0.5811312007417709,
-      "grad_norm": 3.4252729415893555,
-      "learning_rate": 8.915453573748356e-06,
-      "loss": 0.7506,
-      "step": 12535
-    },
-    {
-      "epoch": 0.5813630041724618,
-      "grad_norm": 4.157860279083252,
-      "learning_rate": 8.907410191830782e-06,
-      "loss": 1.0024,
-      "step": 12540
-    },
-    {
-      "epoch": 0.5815948076031525,
-      "grad_norm": 4.022140979766846,
-      "learning_rate": 8.899367525251652e-06,
-      "loss": 1.0174,
-      "step": 12545
-    },
-    {
-      "epoch": 0.5818266110338433,
-      "grad_norm": 4.003377437591553,
-      "learning_rate": 8.891325579276645e-06,
-      "loss": 0.8262,
-      "step": 12550
-    },
-    {
-      "epoch": 0.5820584144645341,
-      "grad_norm": 3.017552614212036,
-      "learning_rate": 8.883284359170973e-06,
-      "loss": 0.9769,
-      "step": 12555
-    },
-    {
-      "epoch": 0.5822902178952248,
-      "grad_norm": 4.014686584472656,
-      "learning_rate": 8.875243870199368e-06,
-      "loss": 1.0247,
-      "step": 12560
-    },
-    {
-      "epoch": 0.5825220213259156,
-      "grad_norm": 3.7590744495391846,
-      "learning_rate": 8.867204117626083e-06,
-      "loss": 0.8188,
-      "step": 12565
-    },
-    {
-      "epoch": 0.5827538247566064,
-      "grad_norm": 4.151451110839844,
-      "learning_rate": 8.859165106714895e-06,
-      "loss": 0.9248,
-      "step": 12570
-    },
-    {
-      "epoch": 0.5829856281872972,
-      "grad_norm": 3.3136253356933594,
-      "learning_rate": 8.851126842729088e-06,
-      "loss": 0.7535,
-      "step": 12575
-    },
-    {
-      "epoch": 0.5832174316179879,
-      "grad_norm": 3.7408132553100586,
-      "learning_rate": 8.843089330931458e-06,
-      "loss": 0.7754,
-      "step": 12580
-    },
-    {
-      "epoch": 0.5834492350486787,
-      "grad_norm": 4.146431922912598,
-      "learning_rate": 8.83505257658431e-06,
-      "loss": 0.7483,
-      "step": 12585
-    },
-    {
-      "epoch": 0.5836810384793695,
-      "grad_norm": 4.425289630889893,
-      "learning_rate": 8.827016584949458e-06,
-      "loss": 0.9464,
-      "step": 12590
-    },
-    {
-      "epoch": 0.5839128419100603,
-      "grad_norm": 3.187209367752075,
-      "learning_rate": 8.81898136128821e-06,
-      "loss": 0.8334,
-      "step": 12595
-    },
-    {
-      "epoch": 0.5841446453407511,
-      "grad_norm": 2.9438233375549316,
-      "learning_rate": 8.810946910861374e-06,
-      "loss": 0.7319,
-      "step": 12600
-    },
-    {
-      "epoch": 0.5841446453407511,
-      "eval_loss": 0.9165540933609009,
-      "eval_runtime": 11.2682,
-      "eval_samples_per_second": 11.271,
-      "eval_steps_per_second": 11.271,
-      "step": 12600
-    },
-    {
-      "epoch": 0.5843764487714418,
-      "grad_norm": 4.2117767333984375,
-      "learning_rate": 8.802913238929248e-06,
-      "loss": 0.8086,
-      "step": 12605
-    },
-    {
-      "epoch": 0.5846082522021326,
-      "grad_norm": 3.517542839050293,
-      "learning_rate": 8.794880350751624e-06,
-      "loss": 0.8586,
-      "step": 12610
-    },
-    {
-      "epoch": 0.5848400556328234,
-      "grad_norm": 3.9067094326019287,
-      "learning_rate": 8.786848251587783e-06,
-      "loss": 0.8945,
-      "step": 12615
-    },
-    {
-      "epoch": 0.5850718590635141,
-      "grad_norm": 3.6777069568634033,
-      "learning_rate": 8.778816946696484e-06,
-      "loss": 0.865,
-      "step": 12620
-    },
-    {
-      "epoch": 0.585303662494205,
-      "grad_norm": 4.279821395874023,
-      "learning_rate": 8.770786441335967e-06,
-      "loss": 0.8967,
-      "step": 12625
-    },
-    {
-      "epoch": 0.5855354659248957,
-      "grad_norm": 3.570180892944336,
-      "learning_rate": 8.76275674076396e-06,
-      "loss": 0.9695,
-      "step": 12630
-    },
-    {
-      "epoch": 0.5857672693555864,
-      "grad_norm": 3.2852587699890137,
-      "learning_rate": 8.75472785023764e-06,
-      "loss": 0.943,
-      "step": 12635
-    },
-    {
-      "epoch": 0.5859990727862773,
-      "grad_norm": 5.42115592956543,
-      "learning_rate": 8.746699775013679e-06,
-      "loss": 0.838,
-      "step": 12640
-    },
-    {
-      "epoch": 0.586230876216968,
-      "grad_norm": 3.4503657817840576,
-      "learning_rate": 8.738672520348194e-06,
-      "loss": 0.7896,
-      "step": 12645
-    },
-    {
-      "epoch": 0.5864626796476587,
-      "grad_norm": 3.632561445236206,
-      "learning_rate": 8.730646091496783e-06,
-      "loss": 0.8426,
-      "step": 12650
-    },
-    {
-      "epoch": 0.5866944830783496,
-      "grad_norm": 3.065523624420166,
-      "learning_rate": 8.722620493714495e-06,
-      "loss": 0.7696,
-      "step": 12655
-    },
-    {
-      "epoch": 0.5869262865090403,
-      "grad_norm": 3.866663694381714,
-      "learning_rate": 8.714595732255833e-06,
-      "loss": 0.7732,
-      "step": 12660
-    },
-    {
-      "epoch": 0.5871580899397311,
-      "grad_norm": 4.122859001159668,
-      "learning_rate": 8.706571812374754e-06,
-      "loss": 0.8873,
-      "step": 12665
-    },
-    {
-      "epoch": 0.5873898933704219,
-      "grad_norm": 4.008296966552734,
-      "learning_rate": 8.69854873932466e-06,
-      "loss": 0.8638,
-      "step": 12670
-    },
-    {
-      "epoch": 0.5876216968011126,
-      "grad_norm": 3.9733011722564697,
-      "learning_rate": 8.690526518358411e-06,
-      "loss": 1.0573,
-      "step": 12675
-    },
-    {
-      "epoch": 0.5878535002318034,
-      "grad_norm": 4.02682638168335,
-      "learning_rate": 8.682505154728299e-06,
-      "loss": 0.8987,
-      "step": 12680
-    },
-    {
-      "epoch": 0.5880853036624942,
-      "grad_norm": 4.271615028381348,
-      "learning_rate": 8.674484653686052e-06,
-      "loss": 1.0168,
-      "step": 12685
-    },
-    {
-      "epoch": 0.588317107093185,
-      "grad_norm": 4.35434627532959,
-      "learning_rate": 8.666465020482847e-06,
-      "loss": 0.8627,
-      "step": 12690
-    },
-    {
-      "epoch": 0.5885489105238757,
-      "grad_norm": 3.5957367420196533,
-      "learning_rate": 8.658446260369274e-06,
-      "loss": 0.7889,
-      "step": 12695
-    },
-    {
-      "epoch": 0.5887807139545665,
-      "grad_norm": 3.276003360748291,
-      "learning_rate": 8.650428378595368e-06,
-      "loss": 0.6797,
-      "step": 12700
-    },
-    {
-      "epoch": 0.5887807139545665,
-      "eval_loss": 0.9164103865623474,
-      "eval_runtime": 11.2762,
-      "eval_samples_per_second": 11.263,
-      "eval_steps_per_second": 11.263,
-      "step": 12700
-    },
-    {
-      "epoch": 0.5890125173852573,
-      "grad_norm": 3.588099718093872,
-      "learning_rate": 8.64241138041058e-06,
-      "loss": 0.9398,
-      "step": 12705
-    },
-    {
-      "epoch": 0.5892443208159481,
-      "grad_norm": 3.816437244415283,
-      "learning_rate": 8.634395271063782e-06,
-      "loss": 0.9051,
-      "step": 12710
-    },
-    {
-      "epoch": 0.5894761242466389,
-      "grad_norm": 3.784233331680298,
-      "learning_rate": 8.626380055803272e-06,
-      "loss": 0.8581,
-      "step": 12715
-    },
-    {
-      "epoch": 0.5897079276773296,
-      "grad_norm": 3.619258165359497,
-      "learning_rate": 8.618365739876759e-06,
-      "loss": 0.8342,
-      "step": 12720
-    },
-    {
-      "epoch": 0.5899397311080204,
-      "grad_norm": 3.3081767559051514,
-      "learning_rate": 8.610352328531353e-06,
-      "loss": 0.6449,
-      "step": 12725
-    },
-    {
-      "epoch": 0.5901715345387112,
-      "grad_norm": 3.5461251735687256,
-      "learning_rate": 8.602339827013583e-06,
-      "loss": 0.8356,
-      "step": 12730
-    },
-    {
-      "epoch": 0.5904033379694019,
-      "grad_norm": 3.2143607139587402,
-      "learning_rate": 8.594328240569387e-06,
-      "loss": 0.8567,
-      "step": 12735
-    },
-    {
-      "epoch": 0.5906351414000928,
-      "grad_norm": 4.731443405151367,
-      "learning_rate": 8.586317574444088e-06,
-      "loss": 0.9381,
-      "step": 12740
-    },
-    {
-      "epoch": 0.5908669448307835,
-      "grad_norm": 3.678004741668701,
-      "learning_rate": 8.578307833882422e-06,
-      "loss": 0.7766,
-      "step": 12745
-    },
-    {
-      "epoch": 0.5910987482614742,
-      "grad_norm": 4.3722147941589355,
-      "learning_rate": 8.570299024128505e-06,
-      "loss": 0.9503,
-      "step": 12750
-    },
-    {
-      "epoch": 0.5913305516921651,
-      "grad_norm": 3.4704501628875732,
-      "learning_rate": 8.562291150425851e-06,
-      "loss": 0.7958,
-      "step": 12755
-    },
-    {
-      "epoch": 0.5915623551228558,
-      "grad_norm": 3.9162561893463135,
-      "learning_rate": 8.554284218017366e-06,
-      "loss": 0.8784,
-      "step": 12760
-    },
-    {
-      "epoch": 0.5917941585535466,
-      "grad_norm": 3.3849542140960693,
-      "learning_rate": 8.546278232145334e-06,
-      "loss": 0.9433,
-      "step": 12765
-    },
-    {
-      "epoch": 0.5920259619842374,
-      "grad_norm": 3.4642109870910645,
-      "learning_rate": 8.538273198051413e-06,
-      "loss": 0.9221,
-      "step": 12770
-    },
-    {
-      "epoch": 0.5922577654149281,
-      "grad_norm": 3.538447380065918,
-      "learning_rate": 8.530269120976654e-06,
-      "loss": 0.9911,
-      "step": 12775
-    },
-    {
-      "epoch": 0.5924895688456189,
-      "grad_norm": 3.6921002864837646,
-      "learning_rate": 8.522266006161467e-06,
-      "loss": 1.0559,
-      "step": 12780
-    },
-    {
-      "epoch": 0.5927213722763097,
-      "grad_norm": 4.147615432739258,
-      "learning_rate": 8.514263858845638e-06,
-      "loss": 0.9022,
-      "step": 12785
-    },
-    {
-      "epoch": 0.5929531757070005,
-      "grad_norm": 4.016373634338379,
-      "learning_rate": 8.506262684268312e-06,
-      "loss": 0.9778,
-      "step": 12790
-    },
-    {
-      "epoch": 0.5931849791376912,
-      "grad_norm": 3.8568859100341797,
-      "learning_rate": 8.498262487668015e-06,
-      "loss": 0.9409,
-      "step": 12795
-    },
-    {
-      "epoch": 0.593416782568382,
-      "grad_norm": 3.5835936069488525,
-      "learning_rate": 8.490263274282615e-06,
-      "loss": 0.9539,
-      "step": 12800
-    },
-    {
-      "epoch": 0.593416782568382,
-      "eval_loss": 0.9164746999740601,
-      "eval_runtime": 11.2684,
-      "eval_samples_per_second": 11.27,
-      "eval_steps_per_second": 11.27,
-      "step": 12800
-    },
-    {
-      "epoch": 0.5936485859990728,
-      "grad_norm": 4.477583885192871,
-      "learning_rate": 8.482265049349347e-06,
-      "loss": 0.9882,
-      "step": 12805
-    },
-    {
-      "epoch": 0.5938803894297635,
-      "grad_norm": 4.2472076416015625,
-      "learning_rate": 8.474267818104788e-06,
-      "loss": 1.0207,
-      "step": 12810
-    },
-    {
-      "epoch": 0.5941121928604544,
-      "grad_norm": 4.307184219360352,
-      "learning_rate": 8.466271585784872e-06,
-      "loss": 0.7898,
-      "step": 12815
-    },
-    {
-      "epoch": 0.5943439962911451,
-      "grad_norm": 4.007270336151123,
-      "learning_rate": 8.458276357624882e-06,
-      "loss": 0.8249,
-      "step": 12820
-    },
-    {
-      "epoch": 0.5945757997218358,
-      "grad_norm": 3.8119564056396484,
-      "learning_rate": 8.450282138859439e-06,
-      "loss": 0.8163,
-      "step": 12825
-    },
-    {
-      "epoch": 0.5948076031525267,
-      "grad_norm": 4.081138610839844,
-      "learning_rate": 8.4422889347225e-06,
-      "loss": 0.8888,
-      "step": 12830
-    },
-    {
-      "epoch": 0.5950394065832174,
-      "grad_norm": 3.449310302734375,
-      "learning_rate": 8.43429675044737e-06,
-      "loss": 0.856,
-      "step": 12835
-    },
-    {
-      "epoch": 0.5952712100139083,
-      "grad_norm": 3.682424783706665,
-      "learning_rate": 8.426305591266664e-06,
-      "loss": 1.0172,
-      "step": 12840
-    },
-    {
-      "epoch": 0.595503013444599,
-      "grad_norm": 4.203467845916748,
-      "learning_rate": 8.418315462412348e-06,
-      "loss": 0.7828,
-      "step": 12845
-    },
-    {
-      "epoch": 0.5957348168752897,
-      "grad_norm": 2.9372358322143555,
-      "learning_rate": 8.410326369115706e-06,
-      "loss": 0.8506,
-      "step": 12850
-    },
-    {
-      "epoch": 0.5959666203059806,
-      "grad_norm": 3.798149347305298,
-      "learning_rate": 8.402338316607336e-06,
-      "loss": 0.8294,
-      "step": 12855
-    },
-    {
-      "epoch": 0.5961984237366713,
-      "grad_norm": 3.2341666221618652,
-      "learning_rate": 8.394351310117168e-06,
-      "loss": 0.9562,
-      "step": 12860
-    },
-    {
-      "epoch": 0.596430227167362,
-      "grad_norm": 3.3644139766693115,
-      "learning_rate": 8.386365354874442e-06,
-      "loss": 0.782,
-      "step": 12865
-    },
-    {
-      "epoch": 0.5966620305980529,
-      "grad_norm": 5.455288410186768,
-      "learning_rate": 8.378380456107702e-06,
-      "loss": 0.9533,
-      "step": 12870
-    },
-    {
-      "epoch": 0.5968938340287436,
-      "grad_norm": 3.402225971221924,
-      "learning_rate": 8.370396619044806e-06,
-      "loss": 0.8046,
-      "step": 12875
-    },
-    {
-      "epoch": 0.5971256374594344,
-      "grad_norm": 3.9545254707336426,
-      "learning_rate": 8.362413848912925e-06,
-      "loss": 0.9077,
-      "step": 12880
-    },
-    {
-      "epoch": 0.5973574408901252,
-      "grad_norm": 3.285334587097168,
-      "learning_rate": 8.35443215093852e-06,
-      "loss": 0.8095,
-      "step": 12885
-    },
-    {
-      "epoch": 0.597589244320816,
-      "grad_norm": 3.7525312900543213,
-      "learning_rate": 8.34645153034735e-06,
-      "loss": 0.9235,
-      "step": 12890
-    },
-    {
-      "epoch": 0.5978210477515067,
-      "grad_norm": 3.530529022216797,
-      "learning_rate": 8.338471992364484e-06,
-      "loss": 0.9493,
-      "step": 12895
-    },
-    {
-      "epoch": 0.5980528511821975,
-      "grad_norm": 3.5442280769348145,
-      "learning_rate": 8.330493542214257e-06,
-      "loss": 0.8373,
-      "step": 12900
-    },
-    {
-      "epoch": 0.5980528511821975,
-      "eval_loss": 0.9144797921180725,
-      "eval_runtime": 11.2681,
-      "eval_samples_per_second": 11.271,
-      "eval_steps_per_second": 11.271,
-      "step": 12900
-    },
-    {
-      "epoch": 0.5982846546128883,
-      "grad_norm": 3.829141855239868,
-      "learning_rate": 8.322516185120314e-06,
-      "loss": 0.9114,
-      "step": 12905
-    },
-    {
-      "epoch": 0.598516458043579,
-      "grad_norm": 3.869734764099121,
-      "learning_rate": 8.314539926305573e-06,
-      "loss": 0.8432,
-      "step": 12910
-    },
-    {
-      "epoch": 0.5987482614742698,
-      "grad_norm": 3.453571081161499,
-      "learning_rate": 8.306564770992236e-06,
-      "loss": 0.7266,
-      "step": 12915
-    },
-    {
-      "epoch": 0.5989800649049606,
-      "grad_norm": 3.4968366622924805,
-      "learning_rate": 8.298590724401785e-06,
-      "loss": 0.8843,
-      "step": 12920
-    },
-    {
-      "epoch": 0.5992118683356513,
-      "grad_norm": 3.363662004470825,
-      "learning_rate": 8.290617791754975e-06,
-      "loss": 0.9189,
-      "step": 12925
-    },
-    {
-      "epoch": 0.5994436717663422,
-      "grad_norm": 6.320411205291748,
-      "learning_rate": 8.282645978271825e-06,
-      "loss": 0.9253,
-      "step": 12930
-    },
-    {
-      "epoch": 0.5996754751970329,
-      "grad_norm": 3.4955732822418213,
-      "learning_rate": 8.274675289171623e-06,
-      "loss": 0.8373,
-      "step": 12935
-    },
-    {
-      "epoch": 0.5999072786277236,
-      "grad_norm": 3.8414204120635986,
-      "learning_rate": 8.266705729672935e-06,
-      "loss": 0.9313,
-      "step": 12940
-    },
-    {
-      "epoch": 0.6001390820584145,
-      "grad_norm": 3.388800859451294,
-      "learning_rate": 8.25873730499357e-06,
-      "loss": 0.9714,
-      "step": 12945
-    },
-    {
-      "epoch": 0.6003708854891052,
-      "grad_norm": 3.6213204860687256,
-      "learning_rate": 8.250770020350602e-06,
-      "loss": 0.8529,
-      "step": 12950
-    },
-    {
-      "epoch": 0.6006026889197961,
-      "grad_norm": 4.339563846588135,
-      "learning_rate": 8.242803880960362e-06,
-      "loss": 0.9313,
-      "step": 12955
-    },
-    {
-      "epoch": 0.6008344923504868,
-      "grad_norm": 3.9930691719055176,
-      "learning_rate": 8.234838892038417e-06,
-      "loss": 0.8653,
-      "step": 12960
-    },
-    {
-      "epoch": 0.6010662957811775,
-      "grad_norm": 4.320202350616455,
-      "learning_rate": 8.226875058799596e-06,
-      "loss": 0.8899,
-      "step": 12965
-    },
-    {
-      "epoch": 0.6012980992118684,
-      "grad_norm": 4.060983180999756,
-      "learning_rate": 8.218912386457967e-06,
-      "loss": 0.8887,
-      "step": 12970
-    },
-    {
-      "epoch": 0.6015299026425591,
-      "grad_norm": 3.39972186088562,
-      "learning_rate": 8.210950880226833e-06,
-      "loss": 0.8836,
-      "step": 12975
-    },
-    {
-      "epoch": 0.6017617060732499,
-      "grad_norm": 4.337881565093994,
-      "learning_rate": 8.202990545318744e-06,
-      "loss": 0.8655,
-      "step": 12980
-    },
-    {
-      "epoch": 0.6019935095039407,
-      "grad_norm": 4.529641151428223,
-      "learning_rate": 8.195031386945465e-06,
-      "loss": 0.9368,
-      "step": 12985
-    },
-    {
-      "epoch": 0.6022253129346314,
-      "grad_norm": 4.606254577636719,
-      "learning_rate": 8.18707341031801e-06,
-      "loss": 0.8114,
-      "step": 12990
-    },
-    {
-      "epoch": 0.6024571163653222,
-      "grad_norm": 3.1904773712158203,
-      "learning_rate": 8.179116620646607e-06,
-      "loss": 0.9546,
-      "step": 12995
-    },
-    {
-      "epoch": 0.602688919796013,
-      "grad_norm": 3.775242805480957,
-      "learning_rate": 8.171161023140717e-06,
-      "loss": 0.8536,
-      "step": 13000
-    },
-    {
-      "epoch": 0.602688919796013,
-      "eval_loss": 0.9141844511032104,
-      "eval_runtime": 11.2721,
-      "eval_samples_per_second": 11.267,
-      "eval_steps_per_second": 11.267,
-      "step": 13000
-    },
-    {
-      "epoch": 0.6029207232267038,
-      "grad_norm": 3.759544849395752,
-      "learning_rate": 8.163206623009009e-06,
-      "loss": 0.9482,
-      "step": 13005
-    },
-    {
-      "epoch": 0.6031525266573945,
-      "grad_norm": 3.72965145111084,
-      "learning_rate": 8.155253425459379e-06,
-      "loss": 0.9966,
-      "step": 13010
-    },
-    {
-      "epoch": 0.6033843300880853,
-      "grad_norm": 3.206998586654663,
-      "learning_rate": 8.147301435698923e-06,
-      "loss": 0.8948,
-      "step": 13015
-    },
-    {
-      "epoch": 0.6036161335187761,
-      "grad_norm": 3.263573169708252,
-      "learning_rate": 8.139350658933957e-06,
-      "loss": 0.9882,
-      "step": 13020
-    },
-    {
-      "epoch": 0.6038479369494668,
-      "grad_norm": 3.676816701889038,
-      "learning_rate": 8.131401100370002e-06,
-      "loss": 0.871,
-      "step": 13025
-    },
-    {
-      "epoch": 0.6040797403801577,
-      "grad_norm": 4.554088115692139,
-      "learning_rate": 8.123452765211775e-06,
-      "loss": 1.0203,
-      "step": 13030
-    },
-    {
-      "epoch": 0.6043115438108484,
-      "grad_norm": 4.035562992095947,
-      "learning_rate": 8.115505658663193e-06,
-      "loss": 0.9691,
-      "step": 13035
-    },
-    {
-      "epoch": 0.6045433472415391,
-      "grad_norm": 4.155014991760254,
-      "learning_rate": 8.107559785927385e-06,
-      "loss": 0.83,
-      "step": 13040
-    },
-    {
-      "epoch": 0.60477515067223,
-      "grad_norm": 3.3499128818511963,
-      "learning_rate": 8.099615152206644e-06,
-      "loss": 0.8616,
-      "step": 13045
-    },
-    {
-      "epoch": 0.6050069541029207,
-      "grad_norm": 4.116346836090088,
-      "learning_rate": 8.09167176270247e-06,
-      "loss": 0.8987,
-      "step": 13050
-    },
-    {
-      "epoch": 0.6052387575336114,
-      "grad_norm": 4.1573100090026855,
-      "learning_rate": 8.08372962261555e-06,
-      "loss": 0.9406,
-      "step": 13055
-    },
-    {
-      "epoch": 0.6054705609643023,
-      "grad_norm": 3.7353973388671875,
-      "learning_rate": 8.07578873714574e-06,
-      "loss": 0.7806,
-      "step": 13060
-    },
-    {
-      "epoch": 0.605702364394993,
-      "grad_norm": 3.6239640712738037,
-      "learning_rate": 8.06784911149209e-06,
-      "loss": 0.9823,
-      "step": 13065
-    },
-    {
-      "epoch": 0.6059341678256838,
-      "grad_norm": 3.2447075843811035,
-      "learning_rate": 8.059910750852815e-06,
-      "loss": 0.7974,
-      "step": 13070
-    },
-    {
-      "epoch": 0.6061659712563746,
-      "grad_norm": 3.9731087684631348,
-      "learning_rate": 8.0519736604253e-06,
-      "loss": 0.8716,
-      "step": 13075
-    },
-    {
-      "epoch": 0.6063977746870653,
-      "grad_norm": 3.3315420150756836,
-      "learning_rate": 8.044037845406104e-06,
-      "loss": 1.0138,
-      "step": 13080
-    },
-    {
-      "epoch": 0.6066295781177562,
-      "grad_norm": 2.920837163925171,
-      "learning_rate": 8.036103310990954e-06,
-      "loss": 0.9604,
-      "step": 13085
-    },
-    {
-      "epoch": 0.6068613815484469,
-      "grad_norm": 3.556678295135498,
-      "learning_rate": 8.028170062374734e-06,
-      "loss": 0.9994,
-      "step": 13090
-    },
-    {
-      "epoch": 0.6070931849791377,
-      "grad_norm": 3.7681658267974854,
-      "learning_rate": 8.02023810475148e-06,
-      "loss": 0.9512,
-      "step": 13095
-    },
-    {
-      "epoch": 0.6073249884098285,
-      "grad_norm": 3.459141492843628,
-      "learning_rate": 8.012307443314399e-06,
-      "loss": 0.866,
-      "step": 13100
-    },
-    {
-      "epoch": 0.6073249884098285,
-      "eval_loss": 0.9133861064910889,
-      "eval_runtime": 11.2835,
-      "eval_samples_per_second": 11.255,
-      "eval_steps_per_second": 11.255,
-      "step": 13100
-    },
-    {
-      "epoch": 0.6075567918405193,
-      "grad_norm": 3.4361283779144287,
-      "learning_rate": 8.004378083255828e-06,
-      "loss": 0.9585,
-      "step": 13105
-    },
-    {
-      "epoch": 0.60778859527121,
-      "grad_norm": 4.018921375274658,
-      "learning_rate": 7.996450029767269e-06,
-      "loss": 1.0037,
-      "step": 13110
-    },
-    {
-      "epoch": 0.6080203987019008,
-      "grad_norm": 4.195085525512695,
-      "learning_rate": 7.988523288039365e-06,
-      "loss": 0.9713,
-      "step": 13115
-    },
-    {
-      "epoch": 0.6082522021325916,
-      "grad_norm": 3.3644561767578125,
-      "learning_rate": 7.980597863261895e-06,
-      "loss": 0.8191,
-      "step": 13120
-    },
-    {
-      "epoch": 0.6084840055632823,
-      "grad_norm": 4.499396800994873,
-      "learning_rate": 7.972673760623781e-06,
-      "loss": 0.8071,
-      "step": 13125
-    },
-    {
-      "epoch": 0.6087158089939732,
-      "grad_norm": 3.8352174758911133,
-      "learning_rate": 7.964750985313079e-06,
-      "loss": 1.066,
-      "step": 13130
-    },
-    {
-      "epoch": 0.6089476124246639,
-      "grad_norm": 3.380789041519165,
-      "learning_rate": 7.956829542516967e-06,
-      "loss": 0.8846,
-      "step": 13135
-    },
-    {
-      "epoch": 0.6091794158553546,
-      "grad_norm": 4.380323886871338,
-      "learning_rate": 7.948909437421764e-06,
-      "loss": 0.872,
-      "step": 13140
-    },
-    {
-      "epoch": 0.6094112192860455,
-      "grad_norm": 4.038541793823242,
-      "learning_rate": 7.940990675212905e-06,
-      "loss": 0.9652,
-      "step": 13145
-    },
-    {
-      "epoch": 0.6096430227167362,
-      "grad_norm": 3.7945806980133057,
-      "learning_rate": 7.933073261074951e-06,
-      "loss": 0.9835,
-      "step": 13150
-    },
-    {
-      "epoch": 0.6098748261474269,
-      "grad_norm": 4.038233757019043,
-      "learning_rate": 7.92515720019157e-06,
-      "loss": 0.8533,
-      "step": 13155
-    },
-    {
-      "epoch": 0.6101066295781178,
-      "grad_norm": 3.5028693675994873,
-      "learning_rate": 7.917242497745563e-06,
-      "loss": 0.7992,
-      "step": 13160
-    },
-    {
-      "epoch": 0.6103384330088085,
-      "grad_norm": 3.19927978515625,
-      "learning_rate": 7.909329158918817e-06,
-      "loss": 0.7886,
-      "step": 13165
-    },
-    {
-      "epoch": 0.6105702364394993,
-      "grad_norm": 3.3780574798583984,
-      "learning_rate": 7.901417188892347e-06,
-      "loss": 0.8571,
-      "step": 13170
-    },
-    {
-      "epoch": 0.6108020398701901,
-      "grad_norm": 3.4062836170196533,
-      "learning_rate": 7.893506592846262e-06,
-      "loss": 0.8582,
-      "step": 13175
-    },
-    {
-      "epoch": 0.6110338433008808,
-      "grad_norm": 3.2859694957733154,
-      "learning_rate": 7.88559737595977e-06,
-      "loss": 0.7769,
-      "step": 13180
-    },
-    {
-      "epoch": 0.6112656467315716,
-      "grad_norm": 3.8368353843688965,
-      "learning_rate": 7.877689543411184e-06,
-      "loss": 0.9078,
-      "step": 13185
-    },
-    {
-      "epoch": 0.6114974501622624,
-      "grad_norm": 3.6319220066070557,
-      "learning_rate": 7.869783100377906e-06,
-      "loss": 0.7553,
-      "step": 13190
-    },
-    {
-      "epoch": 0.6117292535929532,
-      "grad_norm": 3.9984467029571533,
-      "learning_rate": 7.861878052036425e-06,
-      "loss": 0.8557,
-      "step": 13195
-    },
-    {
-      "epoch": 0.6119610570236439,
-      "grad_norm": 3.637826681137085,
-      "learning_rate": 7.853974403562324e-06,
-      "loss": 0.8803,
-      "step": 13200
-    },
-    {
-      "epoch": 0.6119610570236439,
-      "eval_loss": 0.9121905565261841,
-      "eval_runtime": 11.273,
-      "eval_samples_per_second": 11.266,
-      "eval_steps_per_second": 11.266,
-      "step": 13200
-    },
-    {
-      "epoch": 0.6121928604543347,
-      "grad_norm": 3.8848507404327393,
-      "learning_rate": 7.846072160130258e-06,
-      "loss": 0.9607,
-      "step": 13205
-    },
-    {
-      "epoch": 0.6124246638850255,
-      "grad_norm": 3.7693490982055664,
-      "learning_rate": 7.838171326913977e-06,
-      "loss": 0.9417,
-      "step": 13210
-    },
-    {
-      "epoch": 0.6126564673157163,
-      "grad_norm": 3.1103503704071045,
-      "learning_rate": 7.8302719090863e-06,
-      "loss": 0.6954,
-      "step": 13215
-    },
-    {
-      "epoch": 0.6128882707464071,
-      "grad_norm": 3.501575469970703,
-      "learning_rate": 7.822373911819119e-06,
-      "loss": 0.8014,
-      "step": 13220
-    },
-    {
-      "epoch": 0.6131200741770978,
-      "grad_norm": 4.255192756652832,
-      "learning_rate": 7.81447734028339e-06,
-      "loss": 1.0034,
-      "step": 13225
-    },
-    {
-      "epoch": 0.6133518776077886,
-      "grad_norm": 3.4730281829833984,
-      "learning_rate": 7.806582199649152e-06,
-      "loss": 1.0145,
-      "step": 13230
-    },
-    {
-      "epoch": 0.6135836810384794,
-      "grad_norm": 3.566481828689575,
-      "learning_rate": 7.798688495085492e-06,
-      "loss": 0.9147,
-      "step": 13235
-    },
-    {
-      "epoch": 0.6138154844691701,
-      "grad_norm": 4.164937973022461,
-      "learning_rate": 7.790796231760563e-06,
-      "loss": 1.0266,
-      "step": 13240
-    },
-    {
-      "epoch": 0.614047287899861,
-      "grad_norm": 3.8256633281707764,
-      "learning_rate": 7.782905414841579e-06,
-      "loss": 0.9039,
-      "step": 13245
-    },
-    {
-      "epoch": 0.6142790913305517,
-      "grad_norm": 4.075922012329102,
-      "learning_rate": 7.77501604949479e-06,
-      "loss": 0.9964,
-      "step": 13250
-    },
-    {
-      "epoch": 0.6145108947612424,
-      "grad_norm": 3.1503143310546875,
-      "learning_rate": 7.767128140885515e-06,
-      "loss": 0.8551,
-      "step": 13255
-    },
-    {
-      "epoch": 0.6147426981919333,
-      "grad_norm": 3.707214832305908,
-      "learning_rate": 7.75924169417811e-06,
-      "loss": 0.8868,
-      "step": 13260
-    },
-    {
-      "epoch": 0.614974501622624,
-      "grad_norm": 4.248465538024902,
-      "learning_rate": 7.751356714535974e-06,
-      "loss": 0.9955,
-      "step": 13265
-    },
-    {
-      "epoch": 0.6152063050533148,
-      "grad_norm": 4.484118938446045,
-      "learning_rate": 7.743473207121549e-06,
-      "loss": 1.0392,
-      "step": 13270
-    },
-    {
-      "epoch": 0.6154381084840056,
-      "grad_norm": 3.491764783859253,
-      "learning_rate": 7.735591177096313e-06,
-      "loss": 0.9588,
-      "step": 13275
-    },
-    {
-      "epoch": 0.6156699119146963,
-      "grad_norm": 3.8887925148010254,
-      "learning_rate": 7.727710629620768e-06,
-      "loss": 0.8629,
-      "step": 13280
-    },
-    {
-      "epoch": 0.6159017153453871,
-      "grad_norm": 3.9234886169433594,
-      "learning_rate": 7.719831569854454e-06,
-      "loss": 0.7192,
-      "step": 13285
-    },
-    {
-      "epoch": 0.6161335187760779,
-      "grad_norm": 3.4256324768066406,
-      "learning_rate": 7.71195400295594e-06,
-      "loss": 0.8865,
-      "step": 13290
-    },
-    {
-      "epoch": 0.6163653222067687,
-      "grad_norm": 3.125434160232544,
-      "learning_rate": 7.704077934082808e-06,
-      "loss": 0.7298,
-      "step": 13295
-    },
-    {
-      "epoch": 0.6165971256374594,
-      "grad_norm": 3.844914436340332,
-      "learning_rate": 7.696203368391662e-06,
-      "loss": 0.8286,
-      "step": 13300
-    },
-    {
-      "epoch": 0.6165971256374594,
-      "eval_loss": 0.9110766649246216,
-      "eval_runtime": 11.2753,
-      "eval_samples_per_second": 11.264,
-      "eval_steps_per_second": 11.264,
-      "step": 13300
-    },
-    {
-      "epoch": 0.6168289290681502,
-      "grad_norm": 4.225092887878418,
-      "learning_rate": 7.688330311038133e-06,
-      "loss": 0.8003,
-      "step": 13305
-    },
-    {
-      "epoch": 0.617060732498841,
-      "grad_norm": 3.7063608169555664,
-      "learning_rate": 7.68045876717684e-06,
-      "loss": 0.8791,
-      "step": 13310
-    },
-    {
-      "epoch": 0.6172925359295317,
-      "grad_norm": 4.041941165924072,
-      "learning_rate": 7.672588741961438e-06,
-      "loss": 0.7651,
-      "step": 13315
-    },
-    {
-      "epoch": 0.6175243393602226,
-      "grad_norm": 3.5181355476379395,
-      "learning_rate": 7.664720240544571e-06,
-      "loss": 0.7582,
-      "step": 13320
-    },
-    {
-      "epoch": 0.6177561427909133,
-      "grad_norm": 3.7614009380340576,
-      "learning_rate": 7.65685326807789e-06,
-      "loss": 0.8476,
-      "step": 13325
-    },
-    {
-      "epoch": 0.6179879462216041,
-      "grad_norm": 3.4191246032714844,
-      "learning_rate": 7.648987829712049e-06,
-      "loss": 0.7782,
-      "step": 13330
-    },
-    {
-      "epoch": 0.6182197496522949,
-      "grad_norm": 3.7803595066070557,
-      "learning_rate": 7.641123930596693e-06,
-      "loss": 0.9768,
-      "step": 13335
-    },
-    {
-      "epoch": 0.6184515530829856,
-      "grad_norm": 4.002791881561279,
-      "learning_rate": 7.633261575880455e-06,
-      "loss": 0.9629,
-      "step": 13340
-    },
-    {
-      "epoch": 0.6186833565136765,
-      "grad_norm": 3.3021740913391113,
-      "learning_rate": 7.625400770710965e-06,
-      "loss": 0.8824,
-      "step": 13345
-    },
-    {
-      "epoch": 0.6189151599443672,
-      "grad_norm": 3.8734776973724365,
-      "learning_rate": 7.617541520234836e-06,
-      "loss": 0.9005,
-      "step": 13350
-    },
-    {
-      "epoch": 0.6191469633750579,
-      "grad_norm": 3.9729230403900146,
-      "learning_rate": 7.609683829597661e-06,
-      "loss": 0.7664,
-      "step": 13355
-    },
-    {
-      "epoch": 0.6193787668057488,
-      "grad_norm": 3.6947438716888428,
-      "learning_rate": 7.601827703944014e-06,
-      "loss": 0.8933,
-      "step": 13360
-    },
-    {
-      "epoch": 0.6196105702364395,
-      "grad_norm": 3.5670571327209473,
-      "learning_rate": 7.5939731484174475e-06,
-      "loss": 0.8265,
-      "step": 13365
-    },
-    {
-      "epoch": 0.6198423736671302,
-      "grad_norm": 4.319498538970947,
-      "learning_rate": 7.586120168160472e-06,
-      "loss": 0.9813,
-      "step": 13370
-    },
-    {
-      "epoch": 0.6200741770978211,
-      "grad_norm": 3.1896026134490967,
-      "learning_rate": 7.578268768314585e-06,
-      "loss": 0.8378,
-      "step": 13375
-    },
-    {
-      "epoch": 0.6203059805285118,
-      "grad_norm": 2.9694342613220215,
-      "learning_rate": 7.570418954020238e-06,
-      "loss": 0.828,
-      "step": 13380
-    },
-    {
-      "epoch": 0.6205377839592026,
-      "grad_norm": 4.684661388397217,
-      "learning_rate": 7.5625707304168475e-06,
-      "loss": 0.9626,
-      "step": 13385
-    },
-    {
-      "epoch": 0.6207695873898934,
-      "grad_norm": 3.447885036468506,
-      "learning_rate": 7.554724102642789e-06,
-      "loss": 0.8045,
-      "step": 13390
-    },
-    {
-      "epoch": 0.6210013908205841,
-      "grad_norm": 3.6924703121185303,
-      "learning_rate": 7.546879075835395e-06,
-      "loss": 0.9149,
-      "step": 13395
-    },
-    {
-      "epoch": 0.6212331942512749,
-      "grad_norm": 3.3220014572143555,
-      "learning_rate": 7.539035655130944e-06,
-      "loss": 0.7473,
-      "step": 13400
-    },
-    {
-      "epoch": 0.6212331942512749,
-      "eval_loss": 0.9087728261947632,
-      "eval_runtime": 11.2696,
-      "eval_samples_per_second": 11.269,
-      "eval_steps_per_second": 11.269,
-      "step": 13400
-    },
-    {
-      "epoch": 0.6214649976819657,
-      "grad_norm": 4.143969535827637,
-      "learning_rate": 7.531193845664666e-06,
-      "loss": 0.8948,
-      "step": 13405
-    },
-    {
-      "epoch": 0.6216968011126565,
-      "grad_norm": 3.8429365158081055,
-      "learning_rate": 7.523353652570735e-06,
-      "loss": 1.0161,
-      "step": 13410
-    },
-    {
-      "epoch": 0.6219286045433472,
-      "grad_norm": 5.148918151855469,
-      "learning_rate": 7.515515080982274e-06,
-      "loss": 0.841,
-      "step": 13415
-    },
-    {
-      "epoch": 0.622160407974038,
-      "grad_norm": 4.205752849578857,
-      "learning_rate": 7.507678136031336e-06,
-      "loss": 0.8718,
-      "step": 13420
-    },
-    {
-      "epoch": 0.6223922114047288,
-      "grad_norm": 3.2777302265167236,
-      "learning_rate": 7.499842822848913e-06,
-      "loss": 0.9281,
-      "step": 13425
-    },
-    {
-      "epoch": 0.6226240148354195,
-      "grad_norm": 3.4595041275024414,
-      "learning_rate": 7.4920091465649205e-06,
-      "loss": 0.8329,
-      "step": 13430
-    },
-    {
-      "epoch": 0.6228558182661104,
-      "grad_norm": 4.039912700653076,
-      "learning_rate": 7.484177112308216e-06,
-      "loss": 0.9747,
-      "step": 13435
-    },
-    {
-      "epoch": 0.6230876216968011,
-      "grad_norm": 3.4145519733428955,
-      "learning_rate": 7.476346725206573e-06,
-      "loss": 0.8099,
-      "step": 13440
-    },
-    {
-      "epoch": 0.6233194251274918,
-      "grad_norm": 3.8764584064483643,
-      "learning_rate": 7.468517990386687e-06,
-      "loss": 0.951,
-      "step": 13445
-    },
-    {
-      "epoch": 0.6235512285581827,
-      "grad_norm": 3.5151326656341553,
-      "learning_rate": 7.46069091297418e-06,
-      "loss": 0.8281,
-      "step": 13450
-    },
-    {
-      "epoch": 0.6237830319888734,
-      "grad_norm": 3.4390511512756348,
-      "learning_rate": 7.452865498093571e-06,
-      "loss": 1.0219,
-      "step": 13455
-    },
-    {
-      "epoch": 0.6240148354195643,
-      "grad_norm": 2.892617702484131,
-      "learning_rate": 7.445041750868308e-06,
-      "loss": 0.7701,
-      "step": 13460
-    },
-    {
-      "epoch": 0.624246638850255,
-      "grad_norm": 4.317920684814453,
-      "learning_rate": 7.437219676420742e-06,
-      "loss": 0.8255,
-      "step": 13465
-    },
-    {
-      "epoch": 0.6244784422809457,
-      "grad_norm": 4.20429801940918,
-      "learning_rate": 7.429399279872123e-06,
-      "loss": 0.9242,
-      "step": 13470
-    },
-    {
-      "epoch": 0.6247102457116366,
-      "grad_norm": 3.5195364952087402,
-      "learning_rate": 7.42158056634261e-06,
-      "loss": 0.8838,
-      "step": 13475
-    },
-    {
-      "epoch": 0.6249420491423273,
-      "grad_norm": 3.793839931488037,
-      "learning_rate": 7.413763540951262e-06,
-      "loss": 0.7726,
-      "step": 13480
-    },
-    {
-      "epoch": 0.625173852573018,
-      "grad_norm": 4.641863822937012,
-      "learning_rate": 7.405948208816018e-06,
-      "loss": 1.0465,
-      "step": 13485
-    },
-    {
-      "epoch": 0.6254056560037089,
-      "grad_norm": 3.6203348636627197,
-      "learning_rate": 7.398134575053724e-06,
-      "loss": 0.878,
-      "step": 13490
-    },
-    {
-      "epoch": 0.6256374594343996,
-      "grad_norm": 3.926771640777588,
-      "learning_rate": 7.390322644780107e-06,
-      "loss": 1.0517,
-      "step": 13495
-    },
-    {
-      "epoch": 0.6258692628650904,
-      "grad_norm": 3.5244922637939453,
-      "learning_rate": 7.382512423109781e-06,
-      "loss": 0.8058,
-      "step": 13500
-    },
-    {
-      "epoch": 0.6258692628650904,
-      "eval_loss": 0.9085682034492493,
-      "eval_runtime": 11.2676,
-      "eval_samples_per_second": 11.271,
-      "eval_steps_per_second": 11.271,
-      "step": 13500
-    },
-    {
-      "epoch": 0.6261010662957812,
-      "grad_norm": 4.184100151062012,
-      "learning_rate": 7.374703915156238e-06,
-      "loss": 0.937,
-      "step": 13505
-    },
-    {
-      "epoch": 0.626332869726472,
-      "grad_norm": 3.873671054840088,
-      "learning_rate": 7.3668971260318574e-06,
-      "loss": 0.9048,
-      "step": 13510
-    },
-    {
-      "epoch": 0.6265646731571627,
-      "grad_norm": 4.3199849128723145,
-      "learning_rate": 7.3590920608478766e-06,
-      "loss": 0.9245,
-      "step": 13515
-    },
-    {
-      "epoch": 0.6267964765878535,
-      "grad_norm": 4.238365173339844,
-      "learning_rate": 7.3512887247144185e-06,
-      "loss": 0.8239,
-      "step": 13520
-    },
-    {
-      "epoch": 0.6270282800185443,
-      "grad_norm": 3.413118839263916,
-      "learning_rate": 7.34348712274047e-06,
-      "loss": 0.8431,
-      "step": 13525
-    },
-    {
-      "epoch": 0.627260083449235,
-      "grad_norm": 3.717684268951416,
-      "learning_rate": 7.3356872600338795e-06,
-      "loss": 0.6615,
-      "step": 13530
-    },
-    {
-      "epoch": 0.6274918868799259,
-      "grad_norm": 3.8231186866760254,
-      "learning_rate": 7.327889141701362e-06,
-      "loss": 0.891,
-      "step": 13535
-    },
-    {
-      "epoch": 0.6277236903106166,
-      "grad_norm": 4.464356422424316,
-      "learning_rate": 7.320092772848488e-06,
-      "loss": 0.889,
-      "step": 13540
-    },
-    {
-      "epoch": 0.6279554937413073,
-      "grad_norm": 3.6922123432159424,
-      "learning_rate": 7.31229815857968e-06,
-      "loss": 0.8991,
-      "step": 13545
-    },
-    {
-      "epoch": 0.6281872971719982,
-      "grad_norm": 3.411736488342285,
-      "learning_rate": 7.3045053039982135e-06,
-      "loss": 0.821,
-      "step": 13550
-    },
-    {
-      "epoch": 0.6284191006026889,
-      "grad_norm": 3.353180408477783,
-      "learning_rate": 7.296714214206211e-06,
-      "loss": 0.8047,
-      "step": 13555
-    },
-    {
-      "epoch": 0.6286509040333796,
-      "grad_norm": 3.580899477005005,
-      "learning_rate": 7.288924894304646e-06,
-      "loss": 1.0161,
-      "step": 13560
-    },
-    {
-      "epoch": 0.6288827074640705,
-      "grad_norm": 4.194135665893555,
-      "learning_rate": 7.281137349393326e-06,
-      "loss": 0.8003,
-      "step": 13565
-    },
-    {
-      "epoch": 0.6291145108947612,
-      "grad_norm": 3.760871171951294,
-      "learning_rate": 7.273351584570898e-06,
-      "loss": 1.0035,
-      "step": 13570
-    },
-    {
-      "epoch": 0.6293463143254521,
-      "grad_norm": 3.420745611190796,
-      "learning_rate": 7.26556760493484e-06,
-      "loss": 0.7368,
-      "step": 13575
-    },
-    {
-      "epoch": 0.6295781177561428,
-      "grad_norm": 4.322468280792236,
-      "learning_rate": 7.257785415581471e-06,
-      "loss": 0.9997,
-      "step": 13580
-    },
-    {
-      "epoch": 0.6298099211868335,
-      "grad_norm": 4.65591287612915,
-      "learning_rate": 7.2500050216059305e-06,
-      "loss": 0.8858,
-      "step": 13585
-    },
-    {
-      "epoch": 0.6300417246175244,
-      "grad_norm": 3.4827702045440674,
-      "learning_rate": 7.2422264281021814e-06,
-      "loss": 0.7796,
-      "step": 13590
-    },
-    {
-      "epoch": 0.6302735280482151,
-      "grad_norm": 3.829669952392578,
-      "learning_rate": 7.234449640163015e-06,
-      "loss": 0.8724,
-      "step": 13595
-    },
-    {
-      "epoch": 0.6305053314789059,
-      "grad_norm": 3.7582693099975586,
-      "learning_rate": 7.2266746628800354e-06,
-      "loss": 0.7856,
-      "step": 13600
-    },
-    {
-      "epoch": 0.6305053314789059,
-      "eval_loss": 0.9077388048171997,
-      "eval_runtime": 11.2719,
-      "eval_samples_per_second": 11.267,
-      "eval_steps_per_second": 11.267,
-      "step": 13600
-    },
-    {
-      "epoch": 0.6307371349095967,
-      "grad_norm": 3.175899028778076,
-      "learning_rate": 7.218901501343658e-06,
-      "loss": 0.8942,
-      "step": 13605
-    },
-    {
-      "epoch": 0.6309689383402874,
-      "grad_norm": 4.218183994293213,
-      "learning_rate": 7.211130160643116e-06,
-      "loss": 0.9087,
-      "step": 13610
-    },
-    {
-      "epoch": 0.6312007417709782,
-      "grad_norm": 3.943713903427124,
-      "learning_rate": 7.203360645866446e-06,
-      "loss": 0.8711,
-      "step": 13615
-    },
-    {
-      "epoch": 0.631432545201669,
-      "grad_norm": 4.030193328857422,
-      "learning_rate": 7.195592962100493e-06,
-      "loss": 0.9171,
-      "step": 13620
-    },
-    {
-      "epoch": 0.6316643486323598,
-      "grad_norm": 3.495069742202759,
-      "learning_rate": 7.1878271144309e-06,
-      "loss": 0.7155,
-      "step": 13625
-    },
-    {
-      "epoch": 0.6318961520630505,
-      "grad_norm": 3.8679513931274414,
-      "learning_rate": 7.18006310794211e-06,
-      "loss": 0.9703,
-      "step": 13630
-    },
-    {
-      "epoch": 0.6321279554937413,
-      "grad_norm": 4.323716640472412,
-      "learning_rate": 7.1723009477173565e-06,
-      "loss": 0.9051,
-      "step": 13635
-    },
-    {
-      "epoch": 0.6323597589244321,
-      "grad_norm": 3.993604898452759,
-      "learning_rate": 7.1645406388386685e-06,
-      "loss": 0.8736,
-      "step": 13640
-    },
-    {
-      "epoch": 0.6325915623551228,
-      "grad_norm": 3.447650194168091,
-      "learning_rate": 7.156782186386864e-06,
-      "loss": 1.0247,
-      "step": 13645
-    },
-    {
-      "epoch": 0.6328233657858137,
-      "grad_norm": 3.7543070316314697,
-      "learning_rate": 7.149025595441537e-06,
-      "loss": 0.7316,
-      "step": 13650
-    },
-    {
-      "epoch": 0.6330551692165044,
-      "grad_norm": 3.229228973388672,
-      "learning_rate": 7.141270871081076e-06,
-      "loss": 0.8748,
-      "step": 13655
-    },
-    {
-      "epoch": 0.6332869726471951,
-      "grad_norm": 2.9329261779785156,
-      "learning_rate": 7.13351801838264e-06,
-      "loss": 0.7545,
-      "step": 13660
-    },
-    {
-      "epoch": 0.633518776077886,
-      "grad_norm": 3.830026865005493,
-      "learning_rate": 7.125767042422158e-06,
-      "loss": 0.9572,
-      "step": 13665
-    },
-    {
-      "epoch": 0.6337505795085767,
-      "grad_norm": 4.036183834075928,
-      "learning_rate": 7.118017948274336e-06,
-      "loss": 0.8909,
-      "step": 13670
-    },
-    {
-      "epoch": 0.6339823829392675,
-      "grad_norm": 3.7986652851104736,
-      "learning_rate": 7.110270741012648e-06,
-      "loss": 0.923,
-      "step": 13675
-    },
-    {
-      "epoch": 0.6342141863699583,
-      "grad_norm": 3.986788749694824,
-      "learning_rate": 7.102525425709332e-06,
-      "loss": 0.8906,
-      "step": 13680
-    },
-    {
-      "epoch": 0.634445989800649,
-      "grad_norm": 3.509570598602295,
-      "learning_rate": 7.09478200743539e-06,
-      "loss": 0.9297,
-      "step": 13685
-    },
-    {
-      "epoch": 0.6346777932313398,
-      "grad_norm": 4.8821797370910645,
-      "learning_rate": 7.087040491260575e-06,
-      "loss": 0.8174,
-      "step": 13690
-    },
-    {
-      "epoch": 0.6349095966620306,
-      "grad_norm": 3.297886371612549,
-      "learning_rate": 7.079300882253397e-06,
-      "loss": 0.8757,
-      "step": 13695
-    },
-    {
-      "epoch": 0.6351414000927214,
-      "grad_norm": 3.7602527141571045,
-      "learning_rate": 7.071563185481121e-06,
-      "loss": 0.8544,
-      "step": 13700
-    },
-    {
-      "epoch": 0.6351414000927214,
-      "eval_loss": 0.9052226543426514,
-      "eval_runtime": 11.2739,
-      "eval_samples_per_second": 11.265,
-      "eval_steps_per_second": 11.265,
-      "step": 13700
-    },
-    {
-      "epoch": 0.6353732035234122,
-      "grad_norm": 3.7896673679351807,
-      "learning_rate": 7.06382740600976e-06,
-      "loss": 1.032,
-      "step": 13705
-    },
-    {
-      "epoch": 0.6356050069541029,
-      "grad_norm": 4.140953540802002,
-      "learning_rate": 7.056093548904066e-06,
-      "loss": 0.893,
-      "step": 13710
-    },
-    {
-      "epoch": 0.6358368103847937,
-      "grad_norm": 3.7599258422851562,
-      "learning_rate": 7.048361619227546e-06,
-      "loss": 0.8545,
-      "step": 13715
-    },
-    {
-      "epoch": 0.6360686138154845,
-      "grad_norm": 3.4784529209136963,
-      "learning_rate": 7.04063162204242e-06,
-      "loss": 1.0141,
-      "step": 13720
-    },
-    {
-      "epoch": 0.6363004172461753,
-      "grad_norm": 3.0331788063049316,
-      "learning_rate": 7.0329035624096675e-06,
-      "loss": 0.902,
-      "step": 13725
-    },
-    {
-      "epoch": 0.636532220676866,
-      "grad_norm": 3.8510243892669678,
-      "learning_rate": 7.025177445388987e-06,
-      "loss": 0.734,
-      "step": 13730
-    },
-    {
-      "epoch": 0.6367640241075568,
-      "grad_norm": 3.2795825004577637,
-      "learning_rate": 7.017453276038808e-06,
-      "loss": 0.926,
-      "step": 13735
-    },
-    {
-      "epoch": 0.6369958275382476,
-      "grad_norm": 3.572727680206299,
-      "learning_rate": 7.009731059416287e-06,
-      "loss": 0.8886,
-      "step": 13740
-    },
-    {
-      "epoch": 0.6372276309689383,
-      "grad_norm": 3.5403172969818115,
-      "learning_rate": 7.0020108005773e-06,
-      "loss": 0.8549,
-      "step": 13745
-    },
-    {
-      "epoch": 0.6374594343996292,
-      "grad_norm": 4.365976333618164,
-      "learning_rate": 6.994292504576439e-06,
-      "loss": 0.8625,
-      "step": 13750
-    },
-    {
-      "epoch": 0.6376912378303199,
-      "grad_norm": 3.202730417251587,
-      "learning_rate": 6.986576176467013e-06,
-      "loss": 0.8318,
-      "step": 13755
-    },
-    {
-      "epoch": 0.6379230412610106,
-      "grad_norm": 3.6069822311401367,
-      "learning_rate": 6.9788618213010394e-06,
-      "loss": 0.9083,
-      "step": 13760
-    },
-    {
-      "epoch": 0.6381548446917015,
-      "grad_norm": 4.6335768699646,
-      "learning_rate": 6.971149444129254e-06,
-      "loss": 0.8498,
-      "step": 13765
-    },
-    {
-      "epoch": 0.6383866481223922,
-      "grad_norm": 3.619800329208374,
-      "learning_rate": 6.963439050001086e-06,
-      "loss": 0.9175,
-      "step": 13770
-    },
-    {
-      "epoch": 0.6386184515530829,
-      "grad_norm": 4.0841965675354,
-      "learning_rate": 6.955730643964675e-06,
-      "loss": 0.9527,
-      "step": 13775
-    },
-    {
-      "epoch": 0.6388502549837738,
-      "grad_norm": 2.9466323852539062,
-      "learning_rate": 6.948024231066849e-06,
-      "loss": 0.799,
-      "step": 13780
-    },
-    {
-      "epoch": 0.6390820584144645,
-      "grad_norm": 3.827254295349121,
-      "learning_rate": 6.94031981635314e-06,
-      "loss": 0.8248,
-      "step": 13785
-    },
-    {
-      "epoch": 0.6393138618451553,
-      "grad_norm": 4.052628040313721,
-      "learning_rate": 6.9326174048677695e-06,
-      "loss": 0.9989,
-      "step": 13790
-    },
-    {
-      "epoch": 0.6395456652758461,
-      "grad_norm": 3.57578182220459,
-      "learning_rate": 6.924917001653642e-06,
-      "loss": 0.823,
-      "step": 13795
-    },
-    {
-      "epoch": 0.6397774687065368,
-      "grad_norm": 3.9761085510253906,
-      "learning_rate": 6.917218611752359e-06,
-      "loss": 0.8732,
-      "step": 13800
-    },
-    {
-      "epoch": 0.6397774687065368,
-      "eval_loss": 0.9052022695541382,
-      "eval_runtime": 11.2744,
-      "eval_samples_per_second": 11.264,
-      "eval_steps_per_second": 11.264,
-      "step": 13800
-    },
-    {
-      "epoch": 0.6400092721372276,
-      "grad_norm": 3.8727567195892334,
-      "learning_rate": 6.909522240204194e-06,
-      "loss": 1.022,
-      "step": 13805
-    },
-    {
-      "epoch": 0.6402410755679184,
-      "grad_norm": 3.445838212966919,
-      "learning_rate": 6.9018278920480984e-06,
-      "loss": 0.7732,
-      "step": 13810
-    },
-    {
-      "epoch": 0.6404728789986092,
-      "grad_norm": 3.655327558517456,
-      "learning_rate": 6.894135572321703e-06,
-      "loss": 0.9275,
-      "step": 13815
-    },
-    {
-      "epoch": 0.6407046824292999,
-      "grad_norm": 3.729952096939087,
-      "learning_rate": 6.88644528606131e-06,
-      "loss": 0.8728,
-      "step": 13820
-    },
-    {
-      "epoch": 0.6409364858599907,
-      "grad_norm": 3.2580454349517822,
-      "learning_rate": 6.8787570383018905e-06,
-      "loss": 0.8933,
-      "step": 13825
-    },
-    {
-      "epoch": 0.6411682892906815,
-      "grad_norm": 3.6652071475982666,
-      "learning_rate": 6.8710708340770826e-06,
-      "loss": 0.7551,
-      "step": 13830
-    },
-    {
-      "epoch": 0.6414000927213723,
-      "grad_norm": 3.6860642433166504,
-      "learning_rate": 6.8633866784191834e-06,
-      "loss": 0.91,
-      "step": 13835
-    },
-    {
-      "epoch": 0.6416318961520631,
-      "grad_norm": 3.2946152687072754,
-      "learning_rate": 6.855704576359144e-06,
-      "loss": 0.855,
-      "step": 13840
-    },
-    {
-      "epoch": 0.6418636995827538,
-      "grad_norm": 2.9450178146362305,
-      "learning_rate": 6.848024532926581e-06,
-      "loss": 0.785,
-      "step": 13845
-    },
-    {
-      "epoch": 0.6420955030134446,
-      "grad_norm": 5.349577903747559,
-      "learning_rate": 6.840346553149761e-06,
-      "loss": 0.9715,
-      "step": 13850
-    },
-    {
-      "epoch": 0.6423273064441354,
-      "grad_norm": 3.259477376937866,
-      "learning_rate": 6.83267064205559e-06,
-      "loss": 0.8353,
-      "step": 13855
-    },
-    {
-      "epoch": 0.6425591098748261,
-      "grad_norm": 3.9054973125457764,
-      "learning_rate": 6.824996804669634e-06,
-      "loss": 1.0056,
-      "step": 13860
-    },
-    {
-      "epoch": 0.642790913305517,
-      "grad_norm": 4.273472785949707,
-      "learning_rate": 6.817325046016095e-06,
-      "loss": 0.9855,
-      "step": 13865
-    },
-    {
-      "epoch": 0.6430227167362077,
-      "grad_norm": 3.450132131576538,
-      "learning_rate": 6.809655371117805e-06,
-      "loss": 0.9567,
-      "step": 13870
-    },
-    {
-      "epoch": 0.6432545201668984,
-      "grad_norm": 4.0070624351501465,
-      "learning_rate": 6.8019877849962446e-06,
-      "loss": 0.9257,
-      "step": 13875
-    },
-    {
-      "epoch": 0.6434863235975893,
-      "grad_norm": 3.8173329830169678,
-      "learning_rate": 6.79432229267152e-06,
-      "loss": 0.882,
-      "step": 13880
-    },
-    {
-      "epoch": 0.64371812702828,
-      "grad_norm": 3.4225456714630127,
-      "learning_rate": 6.786658899162373e-06,
-      "loss": 0.7999,
-      "step": 13885
-    },
-    {
-      "epoch": 0.6439499304589708,
-      "grad_norm": 4.218738079071045,
-      "learning_rate": 6.778997609486166e-06,
-      "loss": 0.904,
-      "step": 13890
-    },
-    {
-      "epoch": 0.6441817338896616,
-      "grad_norm": 4.1808671951293945,
-      "learning_rate": 6.77133842865888e-06,
-      "loss": 0.999,
-      "step": 13895
-    },
-    {
-      "epoch": 0.6444135373203523,
-      "grad_norm": 3.5177528858184814,
-      "learning_rate": 6.763681361695125e-06,
-      "loss": 0.9198,
-      "step": 13900
-    },
-    {
-      "epoch": 0.6444135373203523,
-      "eval_loss": 0.9041063189506531,
-      "eval_runtime": 11.2757,
-      "eval_samples_per_second": 11.263,
-      "eval_steps_per_second": 11.263,
-      "step": 13900
-    },
-    {
-      "epoch": 0.6446453407510431,
-      "grad_norm": 3.941197156906128,
-      "learning_rate": 6.756026413608117e-06,
-      "loss": 0.805,
-      "step": 13905
-    },
-    {
-      "epoch": 0.6448771441817339,
-      "grad_norm": 3.7070484161376953,
-      "learning_rate": 6.748373589409696e-06,
-      "loss": 0.9117,
-      "step": 13910
-    },
-    {
-      "epoch": 0.6451089476124247,
-      "grad_norm": 5.314979553222656,
-      "learning_rate": 6.740722894110303e-06,
-      "loss": 0.8005,
-      "step": 13915
-    },
-    {
-      "epoch": 0.6453407510431154,
-      "grad_norm": 3.6489126682281494,
-      "learning_rate": 6.73307433271899e-06,
-      "loss": 0.7596,
-      "step": 13920
-    },
-    {
-      "epoch": 0.6455725544738062,
-      "grad_norm": 3.1416983604431152,
-      "learning_rate": 6.7254279102434015e-06,
-      "loss": 0.9408,
-      "step": 13925
-    },
-    {
-      "epoch": 0.645804357904497,
-      "grad_norm": 3.881897449493408,
-      "learning_rate": 6.717783631689797e-06,
-      "loss": 0.7941,
-      "step": 13930
-    },
-    {
-      "epoch": 0.6460361613351877,
-      "grad_norm": 4.629847049713135,
-      "learning_rate": 6.710141502063023e-06,
-      "loss": 0.8509,
-      "step": 13935
-    },
-    {
-      "epoch": 0.6462679647658786,
-      "grad_norm": 3.130946159362793,
-      "learning_rate": 6.702501526366518e-06,
-      "loss": 0.8722,
-      "step": 13940
-    },
-    {
-      "epoch": 0.6464997681965693,
-      "grad_norm": 3.541328191757202,
-      "learning_rate": 6.69486370960232e-06,
-      "loss": 0.9263,
-      "step": 13945
-    },
-    {
-      "epoch": 0.6467315716272601,
-      "grad_norm": 4.076601505279541,
-      "learning_rate": 6.6872280567710425e-06,
-      "loss": 0.8394,
-      "step": 13950
-    },
-    {
-      "epoch": 0.6469633750579509,
-      "grad_norm": 3.7371718883514404,
-      "learning_rate": 6.679594572871884e-06,
-      "loss": 0.9085,
-      "step": 13955
-    },
-    {
-      "epoch": 0.6471951784886416,
-      "grad_norm": 3.3321588039398193,
-      "learning_rate": 6.6719632629026285e-06,
-      "loss": 0.7783,
-      "step": 13960
-    },
-    {
-      "epoch": 0.6474269819193325,
-      "grad_norm": 3.7667531967163086,
-      "learning_rate": 6.664334131859631e-06,
-      "loss": 0.8365,
-      "step": 13965
-    },
-    {
-      "epoch": 0.6476587853500232,
-      "grad_norm": 3.918442487716675,
-      "learning_rate": 6.6567071847378275e-06,
-      "loss": 0.7859,
-      "step": 13970
-    },
-    {
-      "epoch": 0.6478905887807139,
-      "grad_norm": 3.935377836227417,
-      "learning_rate": 6.649082426530715e-06,
-      "loss": 0.8009,
-      "step": 13975
-    },
-    {
-      "epoch": 0.6481223922114048,
-      "grad_norm": 4.223479270935059,
-      "learning_rate": 6.641459862230366e-06,
-      "loss": 0.9154,
-      "step": 13980
-    },
-    {
-      "epoch": 0.6483541956420955,
-      "grad_norm": 3.7977757453918457,
-      "learning_rate": 6.6338394968274055e-06,
-      "loss": 0.9013,
-      "step": 13985
-    },
-    {
-      "epoch": 0.6485859990727862,
-      "grad_norm": 4.023576736450195,
-      "learning_rate": 6.626221335311029e-06,
-      "loss": 0.986,
-      "step": 13990
-    },
-    {
-      "epoch": 0.6488178025034771,
-      "grad_norm": 4.236316204071045,
-      "learning_rate": 6.618605382668988e-06,
-      "loss": 1.0194,
-      "step": 13995
-    },
-    {
-      "epoch": 0.6490496059341678,
-      "grad_norm": 3.5158629417419434,
-      "learning_rate": 6.610991643887579e-06,
-      "loss": 0.8858,
-      "step": 14000
-    },
-    {
-      "epoch": 0.6490496059341678,
-      "eval_loss": 0.9039557576179504,
-      "eval_runtime": 11.2818,
-      "eval_samples_per_second": 11.257,
-      "eval_steps_per_second": 11.257,
-      "step": 14000
-    },
-    {
-      "epoch": 0.6492814093648586,
-      "grad_norm": 4.082876205444336,
-      "learning_rate": 6.603380123951663e-06,
-      "loss": 0.9116,
-      "step": 14005
-    },
-    {
-      "epoch": 0.6495132127955494,
-      "grad_norm": 3.2537522315979004,
-      "learning_rate": 6.595770827844637e-06,
-      "loss": 0.8609,
-      "step": 14010
-    },
-    {
-      "epoch": 0.6497450162262401,
-      "grad_norm": 3.975658893585205,
-      "learning_rate": 6.588163760548446e-06,
-      "loss": 1.0489,
-      "step": 14015
-    },
-    {
-      "epoch": 0.6499768196569309,
-      "grad_norm": 3.7245466709136963,
-      "learning_rate": 6.580558927043575e-06,
-      "loss": 0.799,
-      "step": 14020
-    },
-    {
-      "epoch": 0.6502086230876217,
-      "grad_norm": 2.8868415355682373,
-      "learning_rate": 6.572956332309044e-06,
-      "loss": 0.8372,
-      "step": 14025
-    },
-    {
-      "epoch": 0.6504404265183125,
-      "grad_norm": 3.727860689163208,
-      "learning_rate": 6.5653559813224165e-06,
-      "loss": 0.7826,
-      "step": 14030
-    },
-    {
-      "epoch": 0.6506722299490032,
-      "grad_norm": 3.944013833999634,
-      "learning_rate": 6.557757879059779e-06,
-      "loss": 0.8501,
-      "step": 14035
-    },
-    {
-      "epoch": 0.650904033379694,
-      "grad_norm": 3.464038610458374,
-      "learning_rate": 6.550162030495746e-06,
-      "loss": 0.995,
-      "step": 14040
-    },
-    {
-      "epoch": 0.6511358368103848,
-      "grad_norm": 4.141759872436523,
-      "learning_rate": 6.5425684406034565e-06,
-      "loss": 0.8872,
-      "step": 14045
-    },
-    {
-      "epoch": 0.6513676402410755,
-      "grad_norm": 3.4732260704040527,
-      "learning_rate": 6.534977114354573e-06,
-      "loss": 1.038,
-      "step": 14050
-    },
-    {
-      "epoch": 0.6515994436717664,
-      "grad_norm": 3.323402166366577,
-      "learning_rate": 6.527388056719277e-06,
-      "loss": 0.7998,
-      "step": 14055
-    },
-    {
-      "epoch": 0.6518312471024571,
-      "grad_norm": 3.9810140132904053,
-      "learning_rate": 6.519801272666259e-06,
-      "loss": 0.9836,
-      "step": 14060
-    },
-    {
-      "epoch": 0.6520630505331478,
-      "grad_norm": 3.6072423458099365,
-      "learning_rate": 6.512216767162729e-06,
-      "loss": 0.8642,
-      "step": 14065
-    },
-    {
-      "epoch": 0.6522948539638387,
-      "grad_norm": 4.4568610191345215,
-      "learning_rate": 6.504634545174402e-06,
-      "loss": 0.888,
-      "step": 14070
-    },
-    {
-      "epoch": 0.6525266573945294,
-      "grad_norm": 3.5873544216156006,
-      "learning_rate": 6.497054611665492e-06,
-      "loss": 0.795,
-      "step": 14075
-    },
-    {
-      "epoch": 0.6527584608252203,
-      "grad_norm": 3.70351505279541,
-      "learning_rate": 6.489476971598721e-06,
-      "loss": 0.9346,
-      "step": 14080
-    },
-    {
-      "epoch": 0.652990264255911,
-      "grad_norm": 3.1498491764068604,
-      "learning_rate": 6.481901629935305e-06,
-      "loss": 0.8551,
-      "step": 14085
-    },
-    {
-      "epoch": 0.6532220676866017,
-      "grad_norm": 4.010249137878418,
-      "learning_rate": 6.4743285916349664e-06,
-      "loss": 0.8376,
-      "step": 14090
-    },
-    {
-      "epoch": 0.6534538711172926,
-      "grad_norm": 4.265456676483154,
-      "learning_rate": 6.466757861655903e-06,
-      "loss": 0.9528,
-      "step": 14095
-    },
-    {
-      "epoch": 0.6536856745479833,
-      "grad_norm": 3.2373321056365967,
-      "learning_rate": 6.459189444954817e-06,
-      "loss": 0.9482,
-      "step": 14100
-    },
-    {
-      "epoch": 0.6536856745479833,
-      "eval_loss": 0.9028105139732361,
-      "eval_runtime": 11.283,
-      "eval_samples_per_second": 11.256,
-      "eval_steps_per_second": 11.256,
-      "step": 14100
-    },
-    {
-      "epoch": 0.653917477978674,
-      "grad_norm": 3.4183366298675537,
-      "learning_rate": 6.451623346486883e-06,
-      "loss": 0.9387,
-      "step": 14105
-    },
-    {
-      "epoch": 0.6541492814093649,
-      "grad_norm": 3.709038019180298,
-      "learning_rate": 6.444059571205762e-06,
-      "loss": 0.8929,
-      "step": 14110
-    },
-    {
-      "epoch": 0.6543810848400556,
-      "grad_norm": 3.6893160343170166,
-      "learning_rate": 6.436498124063601e-06,
-      "loss": 0.8333,
-      "step": 14115
-    },
-    {
-      "epoch": 0.6546128882707464,
-      "grad_norm": 4.0160346031188965,
-      "learning_rate": 6.428939010011014e-06,
-      "loss": 0.963,
-      "step": 14120
-    },
-    {
-      "epoch": 0.6548446917014372,
-      "grad_norm": 3.783250570297241,
-      "learning_rate": 6.421382233997094e-06,
-      "loss": 0.7792,
-      "step": 14125
-    },
-    {
-      "epoch": 0.655076495132128,
-      "grad_norm": 4.110098838806152,
-      "learning_rate": 6.413827800969392e-06,
-      "loss": 0.8569,
-      "step": 14130
-    },
-    {
-      "epoch": 0.6553082985628187,
-      "grad_norm": 3.531893730163574,
-      "learning_rate": 6.406275715873941e-06,
-      "loss": 0.8379,
-      "step": 14135
-    },
-    {
-      "epoch": 0.6555401019935095,
-      "grad_norm": 3.7876014709472656,
-      "learning_rate": 6.398725983655225e-06,
-      "loss": 0.8394,
-      "step": 14140
-    },
-    {
-      "epoch": 0.6557719054242003,
-      "grad_norm": 3.642598867416382,
-      "learning_rate": 6.39117860925619e-06,
-      "loss": 0.8898,
-      "step": 14145
-    },
-    {
-      "epoch": 0.656003708854891,
-      "grad_norm": 3.1661229133605957,
-      "learning_rate": 6.383633597618244e-06,
-      "loss": 0.8901,
-      "step": 14150
-    },
-    {
-      "epoch": 0.6562355122855819,
-      "grad_norm": 3.3539295196533203,
-      "learning_rate": 6.376090953681242e-06,
-      "loss": 0.757,
-      "step": 14155
-    },
-    {
-      "epoch": 0.6564673157162726,
-      "grad_norm": 3.8540477752685547,
-      "learning_rate": 6.368550682383487e-06,
-      "loss": 0.8076,
-      "step": 14160
-    },
-    {
-      "epoch": 0.6566991191469633,
-      "grad_norm": 3.6921424865722656,
-      "learning_rate": 6.361012788661736e-06,
-      "loss": 0.9312,
-      "step": 14165
-    },
-    {
-      "epoch": 0.6569309225776542,
-      "grad_norm": 4.235689163208008,
-      "learning_rate": 6.353477277451179e-06,
-      "loss": 0.9182,
-      "step": 14170
-    },
-    {
-      "epoch": 0.6571627260083449,
-      "grad_norm": 3.6121301651000977,
-      "learning_rate": 6.345944153685461e-06,
-      "loss": 0.8698,
-      "step": 14175
-    },
-    {
-      "epoch": 0.6573945294390356,
-      "grad_norm": 3.3108012676239014,
-      "learning_rate": 6.3384134222966495e-06,
-      "loss": 0.8356,
-      "step": 14180
-    },
-    {
-      "epoch": 0.6576263328697265,
-      "grad_norm": 4.003962993621826,
-      "learning_rate": 6.330885088215255e-06,
-      "loss": 0.8549,
-      "step": 14185
-    },
-    {
-      "epoch": 0.6578581363004172,
-      "grad_norm": 3.306988000869751,
-      "learning_rate": 6.32335915637021e-06,
-      "loss": 0.7874,
-      "step": 14190
-    },
-    {
-      "epoch": 0.6580899397311081,
-      "grad_norm": 3.650463342666626,
-      "learning_rate": 6.315835631688885e-06,
-      "loss": 0.8735,
-      "step": 14195
-    },
-    {
-      "epoch": 0.6583217431617988,
-      "grad_norm": 3.5259885787963867,
-      "learning_rate": 6.308314519097063e-06,
-      "loss": 0.8242,
-      "step": 14200
-    },
-    {
-      "epoch": 0.6583217431617988,
-      "eval_loss": 0.9013914465904236,
-      "eval_runtime": 11.2736,
-      "eval_samples_per_second": 11.265,
-      "eval_steps_per_second": 11.265,
-      "step": 14200
-    },
-    {
-      "epoch": 0.6585535465924895,
-      "grad_norm": 3.500894784927368,
-      "learning_rate": 6.300795823518957e-06,
-      "loss": 0.8101,
-      "step": 14205
-    },
-    {
-      "epoch": 0.6587853500231804,
-      "grad_norm": 6.990668773651123,
-      "learning_rate": 6.293279549877194e-06,
-      "loss": 0.9567,
-      "step": 14210
-    },
-    {
-      "epoch": 0.6590171534538711,
-      "grad_norm": 3.8730952739715576,
-      "learning_rate": 6.285765703092817e-06,
-      "loss": 0.8529,
-      "step": 14215
-    },
-    {
-      "epoch": 0.6592489568845619,
-      "grad_norm": 3.886561393737793,
-      "learning_rate": 6.278254288085272e-06,
-      "loss": 0.9505,
-      "step": 14220
-    },
-    {
-      "epoch": 0.6594807603152527,
-      "grad_norm": 3.6404316425323486,
-      "learning_rate": 6.270745309772424e-06,
-      "loss": 0.7549,
-      "step": 14225
-    },
-    {
-      "epoch": 0.6597125637459434,
-      "grad_norm": 4.1507182121276855,
-      "learning_rate": 6.2632387730705345e-06,
-      "loss": 0.9244,
-      "step": 14230
-    },
-    {
-      "epoch": 0.6599443671766342,
-      "grad_norm": 4.353921890258789,
-      "learning_rate": 6.255734682894275e-06,
-      "loss": 0.9279,
-      "step": 14235
-    },
-    {
-      "epoch": 0.660176170607325,
-      "grad_norm": 3.7515358924865723,
-      "learning_rate": 6.248233044156706e-06,
-      "loss": 0.8284,
-      "step": 14240
-    },
-    {
-      "epoch": 0.6604079740380158,
-      "grad_norm": 3.5080502033233643,
-      "learning_rate": 6.240733861769291e-06,
-      "loss": 0.7591,
-      "step": 14245
-    },
-    {
-      "epoch": 0.6606397774687065,
-      "grad_norm": 3.134357452392578,
-      "learning_rate": 6.233237140641874e-06,
-      "loss": 0.7692,
-      "step": 14250
-    },
-    {
-      "epoch": 0.6608715808993973,
-      "grad_norm": 3.8307859897613525,
-      "learning_rate": 6.225742885682699e-06,
-      "loss": 0.7785,
-      "step": 14255
-    },
-    {
-      "epoch": 0.6611033843300881,
-      "grad_norm": 3.3515188694000244,
-      "learning_rate": 6.2182511017983936e-06,
-      "loss": 0.7713,
-      "step": 14260
-    },
-    {
-      "epoch": 0.6613351877607788,
-      "grad_norm": 3.2930147647857666,
-      "learning_rate": 6.2107617938939625e-06,
-      "loss": 0.7633,
-      "step": 14265
-    },
-    {
-      "epoch": 0.6615669911914697,
-      "grad_norm": 3.6122865676879883,
-      "learning_rate": 6.20327496687279e-06,
-      "loss": 0.8302,
-      "step": 14270
-    },
-    {
-      "epoch": 0.6617987946221604,
-      "grad_norm": 2.9412949085235596,
-      "learning_rate": 6.195790625636644e-06,
-      "loss": 0.7882,
-      "step": 14275
-    },
-    {
-      "epoch": 0.6620305980528511,
-      "grad_norm": 3.8114194869995117,
-      "learning_rate": 6.188308775085654e-06,
-      "loss": 0.9221,
-      "step": 14280
-    },
-    {
-      "epoch": 0.662262401483542,
-      "grad_norm": 3.629974842071533,
-      "learning_rate": 6.180829420118326e-06,
-      "loss": 0.7898,
-      "step": 14285
-    },
-    {
-      "epoch": 0.6624942049142327,
-      "grad_norm": 3.917085886001587,
-      "learning_rate": 6.1733525656315256e-06,
-      "loss": 0.8577,
-      "step": 14290
-    },
-    {
-      "epoch": 0.6627260083449235,
-      "grad_norm": 3.4944143295288086,
-      "learning_rate": 6.165878216520492e-06,
-      "loss": 0.8692,
-      "step": 14295
-    },
-    {
-      "epoch": 0.6629578117756143,
-      "grad_norm": 3.9613192081451416,
-      "learning_rate": 6.158406377678816e-06,
-      "loss": 0.8492,
-      "step": 14300
-    },
-    {
-      "epoch": 0.6629578117756143,
-      "eval_loss": 0.9019988179206848,
-      "eval_runtime": 11.2571,
-      "eval_samples_per_second": 11.282,
-      "eval_steps_per_second": 11.282,
-      "step": 14300
-    },
-    {
-      "epoch": 0.663189615206305,
-      "grad_norm": 3.555412769317627,
-      "learning_rate": 6.1509370539984455e-06,
-      "loss": 0.9138,
-      "step": 14305
-    },
-    {
-      "epoch": 0.6634214186369958,
-      "grad_norm": 3.3379549980163574,
-      "learning_rate": 6.143470250369682e-06,
-      "loss": 0.8428,
-      "step": 14310
-    },
-    {
-      "epoch": 0.6636532220676866,
-      "grad_norm": 3.9386861324310303,
-      "learning_rate": 6.136005971681176e-06,
-      "loss": 0.9693,
-      "step": 14315
-    },
-    {
-      "epoch": 0.6638850254983774,
-      "grad_norm": 3.368528366088867,
-      "learning_rate": 6.1285442228199325e-06,
-      "loss": 0.7739,
-      "step": 14320
-    },
-    {
-      "epoch": 0.6641168289290682,
-      "grad_norm": 3.3850362300872803,
-      "learning_rate": 6.121085008671288e-06,
-      "loss": 0.965,
-      "step": 14325
-    },
-    {
-      "epoch": 0.6643486323597589,
-      "grad_norm": 3.772986888885498,
-      "learning_rate": 6.113628334118929e-06,
-      "loss": 0.9258,
-      "step": 14330
-    },
-    {
-      "epoch": 0.6645804357904497,
-      "grad_norm": 3.1928980350494385,
-      "learning_rate": 6.10617420404488e-06,
-      "loss": 0.7586,
-      "step": 14335
-    },
-    {
-      "epoch": 0.6648122392211405,
-      "grad_norm": 3.8362560272216797,
-      "learning_rate": 6.098722623329489e-06,
-      "loss": 0.9274,
-      "step": 14340
-    },
-    {
-      "epoch": 0.6650440426518313,
-      "grad_norm": 3.8659775257110596,
-      "learning_rate": 6.091273596851446e-06,
-      "loss": 0.7597,
-      "step": 14345
-    },
-    {
-      "epoch": 0.665275846082522,
-      "grad_norm": 3.232468605041504,
-      "learning_rate": 6.0838271294877586e-06,
-      "loss": 0.8773,
-      "step": 14350
-    },
-    {
-      "epoch": 0.6655076495132128,
-      "grad_norm": 3.664357900619507,
-      "learning_rate": 6.076383226113771e-06,
-      "loss": 0.9295,
-      "step": 14355
-    },
-    {
-      "epoch": 0.6657394529439036,
-      "grad_norm": 3.904632091522217,
-      "learning_rate": 6.0689418916031426e-06,
-      "loss": 0.9202,
-      "step": 14360
-    },
-    {
-      "epoch": 0.6659712563745943,
-      "grad_norm": 3.8306760787963867,
-      "learning_rate": 6.061503130827846e-06,
-      "loss": 0.9047,
-      "step": 14365
-    },
-    {
-      "epoch": 0.6662030598052852,
-      "grad_norm": 4.240020751953125,
-      "learning_rate": 6.054066948658175e-06,
-      "loss": 0.9071,
-      "step": 14370
-    },
-    {
-      "epoch": 0.6664348632359759,
-      "grad_norm": 3.161769390106201,
-      "learning_rate": 6.046633349962731e-06,
-      "loss": 0.8829,
-      "step": 14375
-    },
-    {
-      "epoch": 0.6666666666666666,
-      "grad_norm": 3.986294984817505,
-      "learning_rate": 6.039202339608432e-06,
-      "loss": 0.8351,
-      "step": 14380
-    },
-    {
-      "epoch": 0.6668984700973575,
-      "grad_norm": 3.352684497833252,
-      "learning_rate": 6.031773922460493e-06,
-      "loss": 0.8495,
-      "step": 14385
-    },
-    {
-      "epoch": 0.6671302735280482,
-      "grad_norm": 4.166463851928711,
-      "learning_rate": 6.024348103382436e-06,
-      "loss": 1.0153,
-      "step": 14390
-    },
-    {
-      "epoch": 0.667362076958739,
-      "grad_norm": 4.052969932556152,
-      "learning_rate": 6.016924887236077e-06,
-      "loss": 0.7314,
-      "step": 14395
-    },
-    {
-      "epoch": 0.6675938803894298,
-      "grad_norm": 3.7992799282073975,
-      "learning_rate": 6.009504278881529e-06,
-      "loss": 0.8469,
-      "step": 14400
-    },
-    {
-      "epoch": 0.6675938803894298,
-      "eval_loss": 0.9014151096343994,
-      "eval_runtime": 11.2535,
-      "eval_samples_per_second": 11.285,
-      "eval_steps_per_second": 11.285,
-      "step": 14400
-    },
-    {
-      "epoch": 0.6678256838201205,
-      "grad_norm": 4.891984462738037,
-      "learning_rate": 6.002086283177203e-06,
-      "loss": 0.9517,
-      "step": 14405
-    },
-    {
-      "epoch": 0.6680574872508113,
-      "grad_norm": 3.334757089614868,
-      "learning_rate": 5.994670904979793e-06,
-      "loss": 0.8671,
-      "step": 14410
-    },
-    {
-      "epoch": 0.6682892906815021,
-      "grad_norm": 3.49668288230896,
-      "learning_rate": 5.987258149144289e-06,
-      "loss": 0.8352,
-      "step": 14415
-    },
-    {
-      "epoch": 0.6685210941121928,
-      "grad_norm": 4.175472259521484,
-      "learning_rate": 5.979848020523953e-06,
-      "loss": 1.0171,
-      "step": 14420
-    },
-    {
-      "epoch": 0.6687528975428836,
-      "grad_norm": 3.3329074382781982,
-      "learning_rate": 5.972440523970329e-06,
-      "loss": 0.9542,
-      "step": 14425
-    },
-    {
-      "epoch": 0.6689847009735744,
-      "grad_norm": 3.4944443702697754,
-      "learning_rate": 5.965035664333241e-06,
-      "loss": 0.7861,
-      "step": 14430
-    },
-    {
-      "epoch": 0.6692165044042652,
-      "grad_norm": 5.090800762176514,
-      "learning_rate": 5.957633446460783e-06,
-      "loss": 1.0441,
-      "step": 14435
-    },
-    {
-      "epoch": 0.6694483078349559,
-      "grad_norm": 3.310312032699585,
-      "learning_rate": 5.950233875199328e-06,
-      "loss": 0.7745,
-      "step": 14440
-    },
-    {
-      "epoch": 0.6696801112656467,
-      "grad_norm": 3.937577962875366,
-      "learning_rate": 5.942836955393507e-06,
-      "loss": 1.1439,
-      "step": 14445
-    },
-    {
-      "epoch": 0.6699119146963375,
-      "grad_norm": 4.166878700256348,
-      "learning_rate": 5.93544269188622e-06,
-      "loss": 0.7785,
-      "step": 14450
-    },
-    {
-      "epoch": 0.6701437181270283,
-      "grad_norm": 3.7805533409118652,
-      "learning_rate": 5.928051089518623e-06,
-      "loss": 0.7695,
-      "step": 14455
-    },
-    {
-      "epoch": 0.6703755215577191,
-      "grad_norm": 4.2814249992370605,
-      "learning_rate": 5.92066215313013e-06,
-      "loss": 0.9514,
-      "step": 14460
-    },
-    {
-      "epoch": 0.6706073249884098,
-      "grad_norm": 3.207496166229248,
-      "learning_rate": 5.913275887558418e-06,
-      "loss": 0.7387,
-      "step": 14465
-    },
-    {
-      "epoch": 0.6708391284191006,
-      "grad_norm": 3.688148021697998,
-      "learning_rate": 5.905892297639408e-06,
-      "loss": 0.9069,
-      "step": 14470
-    },
-    {
-      "epoch": 0.6710709318497914,
-      "grad_norm": 3.606431245803833,
-      "learning_rate": 5.898511388207269e-06,
-      "loss": 0.8267,
-      "step": 14475
-    },
-    {
-      "epoch": 0.6713027352804821,
-      "grad_norm": 4.078047752380371,
-      "learning_rate": 5.89113316409442e-06,
-      "loss": 1.0075,
-      "step": 14480
-    },
-    {
-      "epoch": 0.671534538711173,
-      "grad_norm": 3.6268293857574463,
-      "learning_rate": 5.883757630131517e-06,
-      "loss": 0.8306,
-      "step": 14485
-    },
-    {
-      "epoch": 0.6717663421418637,
-      "grad_norm": 3.7984752655029297,
-      "learning_rate": 5.8763847911474555e-06,
-      "loss": 0.9554,
-      "step": 14490
-    },
-    {
-      "epoch": 0.6719981455725544,
-      "grad_norm": 3.4812657833099365,
-      "learning_rate": 5.869014651969366e-06,
-      "loss": 0.9154,
-      "step": 14495
-    },
-    {
-      "epoch": 0.6722299490032453,
-      "grad_norm": 3.199309825897217,
-      "learning_rate": 5.8616472174226165e-06,
-      "loss": 0.5889,
-      "step": 14500
-    },
-    {
-      "epoch": 0.6722299490032453,
-      "eval_loss": 0.9003604054450989,
-      "eval_runtime": 11.3032,
-      "eval_samples_per_second": 11.236,
-      "eval_steps_per_second": 11.236,
-      "step": 14500
-    },
-    {
-      "epoch": 0.672461752433936,
-      "grad_norm": 4.093687057495117,
-      "learning_rate": 5.8542824923308e-06,
-      "loss": 0.8131,
-      "step": 14505
-    },
-    {
-      "epoch": 0.6726935558646268,
-      "grad_norm": 4.430140018463135,
-      "learning_rate": 5.846920481515736e-06,
-      "loss": 0.9598,
-      "step": 14510
-    },
-    {
-      "epoch": 0.6729253592953176,
-      "grad_norm": 3.939739465713501,
-      "learning_rate": 5.839561189797464e-06,
-      "loss": 0.8837,
-      "step": 14515
-    },
-    {
-      "epoch": 0.6731571627260083,
-      "grad_norm": 3.9287960529327393,
-      "learning_rate": 5.832204621994251e-06,
-      "loss": 0.6834,
-      "step": 14520
-    },
-    {
-      "epoch": 0.6733889661566991,
-      "grad_norm": 4.755311012268066,
-      "learning_rate": 5.82485078292257e-06,
-      "loss": 0.8228,
-      "step": 14525
-    },
-    {
-      "epoch": 0.6736207695873899,
-      "grad_norm": 3.7588210105895996,
-      "learning_rate": 5.817499677397116e-06,
-      "loss": 0.8636,
-      "step": 14530
-    },
-    {
-      "epoch": 0.6738525730180807,
-      "grad_norm": 3.697885751724243,
-      "learning_rate": 5.810151310230798e-06,
-      "loss": 0.783,
-      "step": 14535
-    },
-    {
-      "epoch": 0.6740843764487714,
-      "grad_norm": 3.8682544231414795,
-      "learning_rate": 5.802805686234721e-06,
-      "loss": 0.9349,
-      "step": 14540
-    },
-    {
-      "epoch": 0.6743161798794622,
-      "grad_norm": 3.55385684967041,
-      "learning_rate": 5.795462810218191e-06,
-      "loss": 0.9123,
-      "step": 14545
-    },
-    {
-      "epoch": 0.674547983310153,
-      "grad_norm": 3.61362624168396,
-      "learning_rate": 5.788122686988732e-06,
-      "loss": 0.8943,
-      "step": 14550
-    },
-    {
-      "epoch": 0.6747797867408437,
-      "grad_norm": 4.1845245361328125,
-      "learning_rate": 5.7807853213520584e-06,
-      "loss": 0.9365,
-      "step": 14555
-    },
-    {
-      "epoch": 0.6750115901715346,
-      "grad_norm": 3.7930006980895996,
-      "learning_rate": 5.773450718112068e-06,
-      "loss": 0.7716,
-      "step": 14560
-    },
-    {
-      "epoch": 0.6752433936022253,
-      "grad_norm": 3.108151435852051,
-      "learning_rate": 5.766118882070864e-06,
-      "loss": 0.7694,
-      "step": 14565
-    },
-    {
-      "epoch": 0.6754751970329161,
-      "grad_norm": 4.250346660614014,
-      "learning_rate": 5.758789818028741e-06,
-      "loss": 0.9389,
-      "step": 14570
-    },
-    {
-      "epoch": 0.6757070004636069,
-      "grad_norm": 3.8213279247283936,
-      "learning_rate": 5.751463530784154e-06,
-      "loss": 0.9436,
-      "step": 14575
-    },
-    {
-      "epoch": 0.6759388038942976,
-      "grad_norm": 3.573568105697632,
-      "learning_rate": 5.7441400251337656e-06,
-      "loss": 0.8428,
-      "step": 14580
-    },
-    {
-      "epoch": 0.6761706073249885,
-      "grad_norm": 3.095245122909546,
-      "learning_rate": 5.736819305872407e-06,
-      "loss": 0.9342,
-      "step": 14585
-    },
-    {
-      "epoch": 0.6764024107556792,
-      "grad_norm": 3.4886996746063232,
-      "learning_rate": 5.729501377793094e-06,
-      "loss": 0.7714,
-      "step": 14590
-    },
-    {
-      "epoch": 0.6766342141863699,
-      "grad_norm": 3.144732713699341,
-      "learning_rate": 5.722186245687e-06,
-      "loss": 0.8221,
-      "step": 14595
-    },
-    {
-      "epoch": 0.6768660176170608,
-      "grad_norm": 3.8869521617889404,
-      "learning_rate": 5.71487391434347e-06,
-      "loss": 0.831,
-      "step": 14600
-    },
-    {
-      "epoch": 0.6768660176170608,
-      "eval_loss": 0.8991519808769226,
-      "eval_runtime": 11.2746,
-      "eval_samples_per_second": 11.264,
-      "eval_steps_per_second": 11.264,
-      "step": 14600
-    },
-    {
-      "epoch": 0.6770978210477515,
-      "grad_norm": 3.6212191581726074,
-      "learning_rate": 5.707564388550033e-06,
-      "loss": 0.8467,
-      "step": 14605
-    },
-    {
-      "epoch": 0.6773296244784422,
-      "grad_norm": 3.0634665489196777,
-      "learning_rate": 5.7002576730923575e-06,
-      "loss": 0.8487,
-      "step": 14610
-    },
-    {
-      "epoch": 0.6775614279091331,
-      "grad_norm": 3.5993947982788086,
-      "learning_rate": 5.692953772754287e-06,
-      "loss": 0.7163,
-      "step": 14615
-    },
-    {
-      "epoch": 0.6777932313398238,
-      "grad_norm": 3.701159954071045,
-      "learning_rate": 5.685652692317825e-06,
-      "loss": 0.8995,
-      "step": 14620
-    },
-    {
-      "epoch": 0.6780250347705146,
-      "grad_norm": 3.198442220687866,
-      "learning_rate": 5.6783544365631135e-06,
-      "loss": 0.6766,
-      "step": 14625
-    },
-    {
-      "epoch": 0.6782568382012054,
-      "grad_norm": 3.342956304550171,
-      "learning_rate": 5.671059010268452e-06,
-      "loss": 0.6306,
-      "step": 14630
-    },
-    {
-      "epoch": 0.6784886416318961,
-      "grad_norm": 4.532590389251709,
-      "learning_rate": 5.663766418210293e-06,
-      "loss": 0.9384,
-      "step": 14635
-    },
-    {
-      "epoch": 0.6787204450625869,
-      "grad_norm": 3.206352949142456,
-      "learning_rate": 5.656476665163233e-06,
-      "loss": 0.7051,
-      "step": 14640
-    },
-    {
-      "epoch": 0.6789522484932777,
-      "grad_norm": 3.7517313957214355,
-      "learning_rate": 5.649189755899999e-06,
-      "loss": 0.9279,
-      "step": 14645
-    },
-    {
-      "epoch": 0.6791840519239685,
-      "grad_norm": 3.682551383972168,
-      "learning_rate": 5.641905695191465e-06,
-      "loss": 0.7678,
-      "step": 14650
-    },
-    {
-      "epoch": 0.6794158553546592,
-      "grad_norm": 3.4281675815582275,
-      "learning_rate": 5.634624487806644e-06,
-      "loss": 0.8488,
-      "step": 14655
-    },
-    {
-      "epoch": 0.67964765878535,
-      "grad_norm": 3.929577112197876,
-      "learning_rate": 5.627346138512671e-06,
-      "loss": 0.7632,
-      "step": 14660
-    },
-    {
-      "epoch": 0.6798794622160408,
-      "grad_norm": 3.8639705181121826,
-      "learning_rate": 5.620070652074808e-06,
-      "loss": 0.9047,
-      "step": 14665
-    },
-    {
-      "epoch": 0.6801112656467315,
-      "grad_norm": 3.5569334030151367,
-      "learning_rate": 5.612798033256454e-06,
-      "loss": 0.8764,
-      "step": 14670
-    },
-    {
-      "epoch": 0.6803430690774224,
-      "grad_norm": 3.9777040481567383,
-      "learning_rate": 5.605528286819131e-06,
-      "loss": 0.9697,
-      "step": 14675
-    },
-    {
-      "epoch": 0.6805748725081131,
-      "grad_norm": 4.220929145812988,
-      "learning_rate": 5.598261417522462e-06,
-      "loss": 0.8711,
-      "step": 14680
-    },
-    {
-      "epoch": 0.6808066759388038,
-      "grad_norm": 4.128261089324951,
-      "learning_rate": 5.5909974301242095e-06,
-      "loss": 0.8907,
-      "step": 14685
-    },
-    {
-      "epoch": 0.6810384793694947,
-      "grad_norm": 3.572897434234619,
-      "learning_rate": 5.5837363293802335e-06,
-      "loss": 0.7123,
-      "step": 14690
-    },
-    {
-      "epoch": 0.6812702828001854,
-      "grad_norm": 4.047999382019043,
-      "learning_rate": 5.576478120044506e-06,
-      "loss": 0.8807,
-      "step": 14695
-    },
-    {
-      "epoch": 0.6815020862308763,
-      "grad_norm": 4.205711364746094,
-      "learning_rate": 5.5692228068691125e-06,
-      "loss": 0.9255,
-      "step": 14700
-    },
-    {
-      "epoch": 0.6815020862308763,
-      "eval_loss": 0.8986382484436035,
-      "eval_runtime": 11.2621,
-      "eval_samples_per_second": 11.277,
-      "eval_steps_per_second": 11.277,
-      "step": 14700
-    },
-    {
-      "epoch": 0.681733889661567,
-      "grad_norm": 3.3091366291046143,
-      "learning_rate": 5.5619703946042405e-06,
-      "loss": 0.6813,
-      "step": 14705
-    },
-    {
-      "epoch": 0.6819656930922577,
-      "grad_norm": 3.9704837799072266,
-      "learning_rate": 5.55472088799817e-06,
-      "loss": 0.8628,
-      "step": 14710
-    },
-    {
-      "epoch": 0.6821974965229486,
-      "grad_norm": 3.0079431533813477,
-      "learning_rate": 5.547474291797293e-06,
-      "loss": 0.8371,
-      "step": 14715
-    },
-    {
-      "epoch": 0.6824292999536393,
-      "grad_norm": 4.201705455780029,
-      "learning_rate": 5.540230610746082e-06,
-      "loss": 0.8795,
-      "step": 14720
-    },
-    {
-      "epoch": 0.6826611033843301,
-      "grad_norm": 3.8901288509368896,
-      "learning_rate": 5.53298984958711e-06,
-      "loss": 1.0424,
-      "step": 14725
-    },
-    {
-      "epoch": 0.6828929068150209,
-      "grad_norm": 3.0045812129974365,
-      "learning_rate": 5.525752013061032e-06,
-      "loss": 0.6785,
-      "step": 14730
-    },
-    {
-      "epoch": 0.6831247102457116,
-      "grad_norm": 4.07247257232666,
-      "learning_rate": 5.518517105906593e-06,
-      "loss": 0.903,
-      "step": 14735
-    },
-    {
-      "epoch": 0.6833565136764024,
-      "grad_norm": 4.029781341552734,
-      "learning_rate": 5.5112851328606235e-06,
-      "loss": 0.9391,
-      "step": 14740
-    },
-    {
-      "epoch": 0.6835883171070932,
-      "grad_norm": 3.749738931655884,
-      "learning_rate": 5.504056098658022e-06,
-      "loss": 0.8587,
-      "step": 14745
-    },
-    {
-      "epoch": 0.683820120537784,
-      "grad_norm": 3.17695689201355,
-      "learning_rate": 5.496830008031768e-06,
-      "loss": 0.7638,
-      "step": 14750
-    },
-    {
-      "epoch": 0.6840519239684747,
-      "grad_norm": 3.8042185306549072,
-      "learning_rate": 5.489606865712915e-06,
-      "loss": 0.8975,
-      "step": 14755
-    },
-    {
-      "epoch": 0.6842837273991655,
-      "grad_norm": 3.345367670059204,
-      "learning_rate": 5.482386676430593e-06,
-      "loss": 0.8314,
-      "step": 14760
-    },
-    {
-      "epoch": 0.6845155308298563,
-      "grad_norm": 3.19871187210083,
-      "learning_rate": 5.475169444911982e-06,
-      "loss": 0.7643,
-      "step": 14765
-    },
-    {
-      "epoch": 0.684747334260547,
-      "grad_norm": 3.753889799118042,
-      "learning_rate": 5.467955175882338e-06,
-      "loss": 0.7962,
-      "step": 14770
-    },
-    {
-      "epoch": 0.6849791376912379,
-      "grad_norm": 4.27974271774292,
-      "learning_rate": 5.460743874064984e-06,
-      "loss": 0.7481,
-      "step": 14775
-    },
-    {
-      "epoch": 0.6852109411219286,
-      "grad_norm": 3.5102105140686035,
-      "learning_rate": 5.453535544181274e-06,
-      "loss": 0.7998,
-      "step": 14780
-    },
-    {
-      "epoch": 0.6854427445526193,
-      "grad_norm": 3.4035680294036865,
-      "learning_rate": 5.4463301909506415e-06,
-      "loss": 0.8603,
-      "step": 14785
-    },
-    {
-      "epoch": 0.6856745479833102,
-      "grad_norm": 3.453497886657715,
-      "learning_rate": 5.439127819090566e-06,
-      "loss": 0.7311,
-      "step": 14790
-    },
-    {
-      "epoch": 0.6859063514140009,
-      "grad_norm": 4.089419841766357,
-      "learning_rate": 5.431928433316562e-06,
-      "loss": 0.7608,
-      "step": 14795
-    },
-    {
-      "epoch": 0.6861381548446917,
-      "grad_norm": 3.4193642139434814,
-      "learning_rate": 5.424732038342204e-06,
-      "loss": 0.9064,
-      "step": 14800
-    },
-    {
-      "epoch": 0.6861381548446917,
-      "eval_loss": 0.8975695371627808,
-      "eval_runtime": 11.2774,
-      "eval_samples_per_second": 11.262,
-      "eval_steps_per_second": 11.262,
-      "step": 14800
-    },
-    {
-      "epoch": 0.6863699582753825,
-      "grad_norm": 4.701401710510254,
-      "learning_rate": 5.4175386388791075e-06,
-      "loss": 0.8915,
-      "step": 14805
-    },
-    {
-      "epoch": 0.6866017617060732,
-      "grad_norm": 3.8055918216705322,
-      "learning_rate": 5.410348239636918e-06,
-      "loss": 0.987,
-      "step": 14810
-    },
-    {
-      "epoch": 0.6868335651367641,
-      "grad_norm": 3.4142491817474365,
-      "learning_rate": 5.4031608453233145e-06,
-      "loss": 0.7781,
-      "step": 14815
-    },
-    {
-      "epoch": 0.6870653685674548,
-      "grad_norm": 3.7377047538757324,
-      "learning_rate": 5.39597646064402e-06,
-      "loss": 0.8501,
-      "step": 14820
-    },
-    {
-      "epoch": 0.6872971719981456,
-      "grad_norm": 3.767622470855713,
-      "learning_rate": 5.388795090302788e-06,
-      "loss": 0.804,
-      "step": 14825
-    },
-    {
-      "epoch": 0.6875289754288364,
-      "grad_norm": 3.2815511226654053,
-      "learning_rate": 5.381616739001386e-06,
-      "loss": 0.8258,
-      "step": 14830
-    },
-    {
-      "epoch": 0.6877607788595271,
-      "grad_norm": 3.6691720485687256,
-      "learning_rate": 5.374441411439609e-06,
-      "loss": 0.8332,
-      "step": 14835
-    },
-    {
-      "epoch": 0.6879925822902179,
-      "grad_norm": 3.6556670665740967,
-      "learning_rate": 5.367269112315278e-06,
-      "loss": 0.8806,
-      "step": 14840
-    },
-    {
-      "epoch": 0.6882243857209087,
-      "grad_norm": 3.983771324157715,
-      "learning_rate": 5.360099846324232e-06,
-      "loss": 0.9924,
-      "step": 14845
-    },
-    {
-      "epoch": 0.6884561891515995,
-      "grad_norm": 3.640467405319214,
-      "learning_rate": 5.352933618160311e-06,
-      "loss": 0.9716,
-      "step": 14850
-    },
-    {
-      "epoch": 0.6886879925822902,
-      "grad_norm": 3.7547876834869385,
-      "learning_rate": 5.345770432515381e-06,
-      "loss": 0.9835,
-      "step": 14855
-    },
-    {
-      "epoch": 0.688919796012981,
-      "grad_norm": 3.748774528503418,
-      "learning_rate": 5.338610294079311e-06,
-      "loss": 0.7202,
-      "step": 14860
-    },
-    {
-      "epoch": 0.6891515994436718,
-      "grad_norm": 3.7663495540618896,
-      "learning_rate": 5.331453207539974e-06,
-      "loss": 0.8268,
-      "step": 14865
-    },
-    {
-      "epoch": 0.6893834028743625,
-      "grad_norm": 3.5820882320404053,
-      "learning_rate": 5.324299177583239e-06,
-      "loss": 0.7453,
-      "step": 14870
-    },
-    {
-      "epoch": 0.6896152063050534,
-      "grad_norm": 4.511739253997803,
-      "learning_rate": 5.317148208892982e-06,
-      "loss": 0.8281,
-      "step": 14875
-    },
-    {
-      "epoch": 0.6898470097357441,
-      "grad_norm": 3.408353567123413,
-      "learning_rate": 5.3100003061510785e-06,
-      "loss": 0.8948,
-      "step": 14880
-    },
-    {
-      "epoch": 0.6900788131664348,
-      "grad_norm": 3.573390483856201,
-      "learning_rate": 5.302855474037383e-06,
-      "loss": 0.8971,
-      "step": 14885
-    },
-    {
-      "epoch": 0.6903106165971257,
-      "grad_norm": 4.010354518890381,
-      "learning_rate": 5.295713717229753e-06,
-      "loss": 0.8969,
-      "step": 14890
-    },
-    {
-      "epoch": 0.6905424200278164,
-      "grad_norm": 3.8016562461853027,
-      "learning_rate": 5.288575040404025e-06,
-      "loss": 0.719,
-      "step": 14895
-    },
-    {
-      "epoch": 0.6907742234585071,
-      "grad_norm": 3.699970245361328,
-      "learning_rate": 5.281439448234016e-06,
-      "loss": 0.8287,
-      "step": 14900
-    },
-    {
-      "epoch": 0.6907742234585071,
-      "eval_loss": 0.8960445523262024,
-      "eval_runtime": 11.2607,
-      "eval_samples_per_second": 11.278,
-      "eval_steps_per_second": 11.278,
-      "step": 14900
-    },
-    {
-      "epoch": 0.691006026889198,
-      "grad_norm": 4.231236934661865,
-      "learning_rate": 5.274306945391532e-06,
-      "loss": 1.005,
-      "step": 14905
-    },
-    {
-      "epoch": 0.6912378303198887,
-      "grad_norm": 4.223328590393066,
-      "learning_rate": 5.267177536546355e-06,
-      "loss": 0.9441,
-      "step": 14910
-    },
-    {
-      "epoch": 0.6914696337505795,
-      "grad_norm": 3.7817633152008057,
-      "learning_rate": 5.260051226366234e-06,
-      "loss": 0.8287,
-      "step": 14915
-    },
-    {
-      "epoch": 0.6917014371812703,
-      "grad_norm": 3.6774466037750244,
-      "learning_rate": 5.2529280195168996e-06,
-      "loss": 0.8102,
-      "step": 14920
-    },
-    {
-      "epoch": 0.691933240611961,
-      "grad_norm": 3.3471529483795166,
-      "learning_rate": 5.245807920662038e-06,
-      "loss": 0.7598,
-      "step": 14925
-    },
-    {
-      "epoch": 0.6921650440426518,
-      "grad_norm": 3.2454981803894043,
-      "learning_rate": 5.238690934463315e-06,
-      "loss": 0.7236,
-      "step": 14930
-    },
-    {
-      "epoch": 0.6923968474733426,
-      "grad_norm": 4.209705829620361,
-      "learning_rate": 5.231577065580345e-06,
-      "loss": 0.976,
-      "step": 14935
-    },
-    {
-      "epoch": 0.6926286509040334,
-      "grad_norm": 3.823845386505127,
-      "learning_rate": 5.2244663186707075e-06,
-      "loss": 0.9594,
-      "step": 14940
-    },
-    {
-      "epoch": 0.6928604543347242,
-      "grad_norm": 3.8407392501831055,
-      "learning_rate": 5.217358698389945e-06,
-      "loss": 0.8413,
-      "step": 14945
-    },
-    {
-      "epoch": 0.6930922577654149,
-      "grad_norm": 3.701613664627075,
-      "learning_rate": 5.210254209391543e-06,
-      "loss": 0.7563,
-      "step": 14950
-    },
-    {
-      "epoch": 0.6933240611961057,
-      "grad_norm": 3.6802921295166016,
-      "learning_rate": 5.203152856326933e-06,
-      "loss": 0.8716,
-      "step": 14955
-    },
-    {
-      "epoch": 0.6935558646267965,
-      "grad_norm": 3.6844983100891113,
-      "learning_rate": 5.196054643845505e-06,
-      "loss": 0.9297,
-      "step": 14960
-    },
-    {
-      "epoch": 0.6937876680574873,
-      "grad_norm": 3.931682825088501,
-      "learning_rate": 5.188959576594593e-06,
-      "loss": 0.8128,
-      "step": 14965
-    },
-    {
-      "epoch": 0.694019471488178,
-      "grad_norm": 3.934149980545044,
-      "learning_rate": 5.181867659219457e-06,
-      "loss": 0.8199,
-      "step": 14970
-    },
-    {
-      "epoch": 0.6942512749188688,
-      "grad_norm": 3.853133201599121,
-      "learning_rate": 5.174778896363312e-06,
-      "loss": 0.9809,
-      "step": 14975
-    },
-    {
-      "epoch": 0.6944830783495596,
-      "grad_norm": 4.392022609710693,
-      "learning_rate": 5.167693292667304e-06,
-      "loss": 1.0305,
-      "step": 14980
-    },
-    {
-      "epoch": 0.6947148817802503,
-      "grad_norm": 4.075222492218018,
-      "learning_rate": 5.160610852770493e-06,
-      "loss": 0.9731,
-      "step": 14985
-    },
-    {
-      "epoch": 0.6949466852109412,
-      "grad_norm": 4.099298000335693,
-      "learning_rate": 5.153531581309887e-06,
-      "loss": 0.9576,
-      "step": 14990
-    },
-    {
-      "epoch": 0.6951784886416319,
-      "grad_norm": 4.013690948486328,
-      "learning_rate": 5.146455482920423e-06,
-      "loss": 0.7349,
-      "step": 14995
-    },
-    {
-      "epoch": 0.6954102920723226,
-      "grad_norm": 3.4314472675323486,
-      "learning_rate": 5.139382562234938e-06,
-      "loss": 0.8236,
-      "step": 15000
-    },
-    {
-      "epoch": 0.6954102920723226,
-      "eval_loss": 0.8956095576286316,
-      "eval_runtime": 11.2867,
-      "eval_samples_per_second": 11.252,
-      "eval_steps_per_second": 11.252,
-      "step": 15000
-    },
-    {
-      "epoch": 0.6956420955030135,
-      "grad_norm": 2.756856679916382,
-      "learning_rate": 5.132312823884209e-06,
-      "loss": 0.8249,
-      "step": 15005
-    },
-    {
-      "epoch": 0.6958738989337042,
-      "grad_norm": 3.730117082595825,
-      "learning_rate": 5.125246272496927e-06,
-      "loss": 0.8329,
-      "step": 15010
-    },
-    {
-      "epoch": 0.696105702364395,
-      "grad_norm": 3.479635715484619,
-      "learning_rate": 5.118182912699684e-06,
-      "loss": 0.8444,
-      "step": 15015
-    },
-    {
-      "epoch": 0.6963375057950858,
-      "grad_norm": 3.8835439682006836,
-      "learning_rate": 5.11112274911699e-06,
-      "loss": 0.8925,
-      "step": 15020
-    },
-    {
-      "epoch": 0.6965693092257765,
-      "grad_norm": 4.538135051727295,
-      "learning_rate": 5.1040657863712666e-06,
-      "loss": 0.911,
-      "step": 15025
-    },
-    {
-      "epoch": 0.6968011126564673,
-      "grad_norm": 3.9026095867156982,
-      "learning_rate": 5.097012029082837e-06,
-      "loss": 0.9225,
-      "step": 15030
-    },
-    {
-      "epoch": 0.6970329160871581,
-      "grad_norm": 4.518922805786133,
-      "learning_rate": 5.0899614818699205e-06,
-      "loss": 1.0216,
-      "step": 15035
-    },
-    {
-      "epoch": 0.6972647195178489,
-      "grad_norm": 4.434970855712891,
-      "learning_rate": 5.0829141493486435e-06,
-      "loss": 0.7916,
-      "step": 15040
-    },
-    {
-      "epoch": 0.6974965229485396,
-      "grad_norm": 3.540708541870117,
-      "learning_rate": 5.075870036133015e-06,
-      "loss": 0.7286,
-      "step": 15045
-    },
-    {
-      "epoch": 0.6977283263792304,
-      "grad_norm": 3.813382387161255,
-      "learning_rate": 5.068829146834955e-06,
-      "loss": 0.6732,
-      "step": 15050
-    },
-    {
-      "epoch": 0.6979601298099212,
-      "grad_norm": 3.9284863471984863,
-      "learning_rate": 5.061791486064251e-06,
-      "loss": 0.7924,
-      "step": 15055
-    },
-    {
-      "epoch": 0.6981919332406119,
-      "grad_norm": 3.7859926223754883,
-      "learning_rate": 5.054757058428595e-06,
-      "loss": 1.0506,
-      "step": 15060
-    },
-    {
-      "epoch": 0.6984237366713028,
-      "grad_norm": 4.451103687286377,
-      "learning_rate": 5.047725868533557e-06,
-      "loss": 0.789,
-      "step": 15065
-    },
-    {
-      "epoch": 0.6986555401019935,
-      "grad_norm": 4.454600811004639,
-      "learning_rate": 5.040697920982576e-06,
-      "loss": 0.8248,
-      "step": 15070
-    },
-    {
-      "epoch": 0.6988873435326843,
-      "grad_norm": 3.176328659057617,
-      "learning_rate": 5.03367322037698e-06,
-      "loss": 0.8367,
-      "step": 15075
-    },
-    {
-      "epoch": 0.6991191469633751,
-      "grad_norm": 3.6452035903930664,
-      "learning_rate": 5.026651771315972e-06,
-      "loss": 0.9312,
-      "step": 15080
-    },
-    {
-      "epoch": 0.6993509503940658,
-      "grad_norm": 3.4964804649353027,
-      "learning_rate": 5.0196335783966255e-06,
-      "loss": 0.9832,
-      "step": 15085
-    },
-    {
-      "epoch": 0.6995827538247567,
-      "grad_norm": 3.5905039310455322,
-      "learning_rate": 5.012618646213869e-06,
-      "loss": 0.8185,
-      "step": 15090
-    },
-    {
-      "epoch": 0.6998145572554474,
-      "grad_norm": 4.217673301696777,
-      "learning_rate": 5.0056069793605175e-06,
-      "loss": 0.8456,
-      "step": 15095
-    },
-    {
-      "epoch": 0.7000463606861381,
-      "grad_norm": 3.272371292114258,
-      "learning_rate": 4.998598582427231e-06,
-      "loss": 0.8687,
-      "step": 15100
-    },
-    {
-      "epoch": 0.7000463606861381,
-      "eval_loss": 0.8956237435340881,
-      "eval_runtime": 11.2682,
-      "eval_samples_per_second": 11.271,
-      "eval_steps_per_second": 11.271,
-      "step": 15100
-    },
-    {
-      "epoch": 0.700278164116829,
-      "grad_norm": 3.845468044281006,
-      "learning_rate": 4.991593460002532e-06,
-      "loss": 0.8925,
-      "step": 15105
-    },
-    {
-      "epoch": 0.7005099675475197,
-      "grad_norm": 4.131564617156982,
-      "learning_rate": 4.984591616672805e-06,
-      "loss": 0.8506,
-      "step": 15110
-    },
-    {
-      "epoch": 0.7007417709782104,
-      "grad_norm": 3.886279344558716,
-      "learning_rate": 4.977593057022286e-06,
-      "loss": 0.8359,
-      "step": 15115
-    },
-    {
-      "epoch": 0.7009735744089013,
-      "grad_norm": 3.745542049407959,
-      "learning_rate": 4.9705977856330566e-06,
-      "loss": 0.6869,
-      "step": 15120
-    },
-    {
-      "epoch": 0.701205377839592,
-      "grad_norm": 4.212728977203369,
-      "learning_rate": 4.963605807085052e-06,
-      "loss": 0.8538,
-      "step": 15125
-    },
-    {
-      "epoch": 0.7014371812702828,
-      "grad_norm": 4.162337779998779,
-      "learning_rate": 4.956617125956042e-06,
-      "loss": 0.8974,
-      "step": 15130
-    },
-    {
-      "epoch": 0.7016689847009736,
-      "grad_norm": 4.083733558654785,
-      "learning_rate": 4.9496317468216505e-06,
-      "loss": 0.8506,
-      "step": 15135
-    },
-    {
-      "epoch": 0.7019007881316643,
-      "grad_norm": 3.7863776683807373,
-      "learning_rate": 4.942649674255326e-06,
-      "loss": 0.8284,
-      "step": 15140
-    },
-    {
-      "epoch": 0.7021325915623551,
-      "grad_norm": 3.9084062576293945,
-      "learning_rate": 4.93567091282836e-06,
-      "loss": 0.7824,
-      "step": 15145
-    },
-    {
-      "epoch": 0.7023643949930459,
-      "grad_norm": 4.295900344848633,
-      "learning_rate": 4.928695467109879e-06,
-      "loss": 0.8566,
-      "step": 15150
-    },
-    {
-      "epoch": 0.7025961984237367,
-      "grad_norm": 3.5533487796783447,
-      "learning_rate": 4.921723341666833e-06,
-      "loss": 0.9455,
-      "step": 15155
-    },
-    {
-      "epoch": 0.7028280018544274,
-      "grad_norm": 3.6774909496307373,
-      "learning_rate": 4.914754541063994e-06,
-      "loss": 0.8561,
-      "step": 15160
-    },
-    {
-      "epoch": 0.7030598052851182,
-      "grad_norm": 3.6324589252471924,
-      "learning_rate": 4.907789069863965e-06,
-      "loss": 0.662,
-      "step": 15165
-    },
-    {
-      "epoch": 0.703291608715809,
-      "grad_norm": 4.190369129180908,
-      "learning_rate": 4.900826932627174e-06,
-      "loss": 0.9028,
-      "step": 15170
-    },
-    {
-      "epoch": 0.7035234121464997,
-      "grad_norm": 4.633210182189941,
-      "learning_rate": 4.893868133911849e-06,
-      "loss": 0.9034,
-      "step": 15175
-    },
-    {
-      "epoch": 0.7037552155771906,
-      "grad_norm": 3.308100700378418,
-      "learning_rate": 4.886912678274046e-06,
-      "loss": 0.8148,
-      "step": 15180
-    },
-    {
-      "epoch": 0.7039870190078813,
-      "grad_norm": 3.5198941230773926,
-      "learning_rate": 4.879960570267638e-06,
-      "loss": 0.8963,
-      "step": 15185
-    },
-    {
-      "epoch": 0.7042188224385721,
-      "grad_norm": 3.3143296241760254,
-      "learning_rate": 4.873011814444279e-06,
-      "loss": 0.7713,
-      "step": 15190
-    },
-    {
-      "epoch": 0.7044506258692629,
-      "grad_norm": 3.4104104042053223,
-      "learning_rate": 4.8660664153534555e-06,
-      "loss": 0.9218,
-      "step": 15195
-    },
-    {
-      "epoch": 0.7046824292999536,
-      "grad_norm": 3.8021419048309326,
-      "learning_rate": 4.8591243775424505e-06,
-      "loss": 0.7878,
-      "step": 15200
-    },
-    {
-      "epoch": 0.7046824292999536,
-      "eval_loss": 0.8943904638290405,
-      "eval_runtime": 11.268,
-      "eval_samples_per_second": 11.271,
-      "eval_steps_per_second": 11.271,
-      "step": 15200
-    },
-    {
-      "epoch": 0.7049142327306445,
-      "grad_norm": 3.752955913543701,
-      "learning_rate": 4.852185705556334e-06,
-      "loss": 0.8282,
-      "step": 15205
-    },
-    {
-      "epoch": 0.7051460361613352,
-      "grad_norm": 4.037057399749756,
-      "learning_rate": 4.845250403937985e-06,
-      "loss": 0.9621,
-      "step": 15210
-    },
-    {
-      "epoch": 0.7053778395920259,
-      "grad_norm": 3.99474835395813,
-      "learning_rate": 4.8383184772280745e-06,
-      "loss": 0.8936,
-      "step": 15215
-    },
-    {
-      "epoch": 0.7056096430227168,
-      "grad_norm": 4.396583080291748,
-      "learning_rate": 4.831389929965059e-06,
-      "loss": 0.9539,
-      "step": 15220
-    },
-    {
-      "epoch": 0.7058414464534075,
-      "grad_norm": 4.322121620178223,
-      "learning_rate": 4.8244647666851815e-06,
-      "loss": 0.7406,
-      "step": 15225
-    },
-    {
-      "epoch": 0.7060732498840983,
-      "grad_norm": 3.7498438358306885,
-      "learning_rate": 4.817542991922475e-06,
-      "loss": 0.7946,
-      "step": 15230
-    },
-    {
-      "epoch": 0.7063050533147891,
-      "grad_norm": 3.8820183277130127,
-      "learning_rate": 4.810624610208754e-06,
-      "loss": 0.7112,
-      "step": 15235
-    },
-    {
-      "epoch": 0.7065368567454798,
-      "grad_norm": 3.285693407058716,
-      "learning_rate": 4.803709626073604e-06,
-      "loss": 0.7001,
-      "step": 15240
-    },
-    {
-      "epoch": 0.7067686601761706,
-      "grad_norm": 4.146936416625977,
-      "learning_rate": 4.796798044044398e-06,
-      "loss": 0.7498,
-      "step": 15245
-    },
-    {
-      "epoch": 0.7070004636068614,
-      "grad_norm": 3.7342891693115234,
-      "learning_rate": 4.789889868646265e-06,
-      "loss": 0.8432,
-      "step": 15250
-    },
-    {
-      "epoch": 0.7072322670375522,
-      "grad_norm": 3.5254108905792236,
-      "learning_rate": 4.782985104402123e-06,
-      "loss": 0.8564,
-      "step": 15255
-    },
-    {
-      "epoch": 0.7074640704682429,
-      "grad_norm": 3.4690604209899902,
-      "learning_rate": 4.776083755832638e-06,
-      "loss": 0.8843,
-      "step": 15260
-    },
-    {
-      "epoch": 0.7076958738989337,
-      "grad_norm": 3.0940325260162354,
-      "learning_rate": 4.769185827456254e-06,
-      "loss": 0.8507,
-      "step": 15265
-    },
-    {
-      "epoch": 0.7079276773296245,
-      "grad_norm": 3.7676939964294434,
-      "learning_rate": 4.762291323789172e-06,
-      "loss": 0.9011,
-      "step": 15270
-    },
-    {
-      "epoch": 0.7081594807603152,
-      "grad_norm": 4.3535308837890625,
-      "learning_rate": 4.755400249345348e-06,
-      "loss": 0.9,
-      "step": 15275
-    },
-    {
-      "epoch": 0.708391284191006,
-      "grad_norm": 4.5304083824157715,
-      "learning_rate": 4.748512608636487e-06,
-      "loss": 0.8402,
-      "step": 15280
-    },
-    {
-      "epoch": 0.7086230876216968,
-      "grad_norm": 3.6635782718658447,
-      "learning_rate": 4.741628406172063e-06,
-      "loss": 0.7709,
-      "step": 15285
-    },
-    {
-      "epoch": 0.7088548910523875,
-      "grad_norm": 3.2135374546051025,
-      "learning_rate": 4.73474764645928e-06,
-      "loss": 0.9065,
-      "step": 15290
-    },
-    {
-      "epoch": 0.7090866944830784,
-      "grad_norm": 4.172916412353516,
-      "learning_rate": 4.7278703340031e-06,
-      "loss": 0.9198,
-      "step": 15295
-    },
-    {
-      "epoch": 0.7093184979137691,
-      "grad_norm": 3.894068717956543,
-      "learning_rate": 4.720996473306231e-06,
-      "loss": 0.7857,
-      "step": 15300
-    },
-    {
-      "epoch": 0.7093184979137691,
-      "eval_loss": 0.8935646414756775,
-      "eval_runtime": 11.2772,
-      "eval_samples_per_second": 11.262,
-      "eval_steps_per_second": 11.262,
-      "step": 15300
-    },
-    {
-      "epoch": 0.7095503013444598,
-      "grad_norm": 3.9130654335021973,
-      "learning_rate": 4.714126068869107e-06,
-      "loss": 0.9399,
-      "step": 15305
-    },
-    {
-      "epoch": 0.7097821047751507,
-      "grad_norm": 3.7862632274627686,
-      "learning_rate": 4.707259125189906e-06,
-      "loss": 0.8922,
-      "step": 15310
-    },
-    {
-      "epoch": 0.7100139082058414,
-      "grad_norm": 4.150571823120117,
-      "learning_rate": 4.700395646764544e-06,
-      "loss": 0.9313,
-      "step": 15315
-    },
-    {
-      "epoch": 0.7102457116365323,
-      "grad_norm": 4.1176862716674805,
-      "learning_rate": 4.693535638086669e-06,
-      "loss": 0.8162,
-      "step": 15320
-    },
-    {
-      "epoch": 0.710477515067223,
-      "grad_norm": 3.9354631900787354,
-      "learning_rate": 4.686679103647647e-06,
-      "loss": 0.9237,
-      "step": 15325
-    },
-    {
-      "epoch": 0.7107093184979137,
-      "grad_norm": 3.2577991485595703,
-      "learning_rate": 4.679826047936582e-06,
-      "loss": 0.7905,
-      "step": 15330
-    },
-    {
-      "epoch": 0.7109411219286046,
-      "grad_norm": 3.445051670074463,
-      "learning_rate": 4.672976475440288e-06,
-      "loss": 0.8439,
-      "step": 15335
-    },
-    {
-      "epoch": 0.7111729253592953,
-      "grad_norm": 3.5186078548431396,
-      "learning_rate": 4.666130390643312e-06,
-      "loss": 0.8311,
-      "step": 15340
-    },
-    {
-      "epoch": 0.7114047287899861,
-      "grad_norm": 3.8396527767181396,
-      "learning_rate": 4.6592877980279025e-06,
-      "loss": 0.8022,
-      "step": 15345
-    },
-    {
-      "epoch": 0.7116365322206769,
-      "grad_norm": 3.7967803478240967,
-      "learning_rate": 4.652448702074034e-06,
-      "loss": 0.8061,
-      "step": 15350
-    },
-    {
-      "epoch": 0.7118683356513676,
-      "grad_norm": 3.51462459564209,
-      "learning_rate": 4.6456131072593905e-06,
-      "loss": 0.8804,
-      "step": 15355
-    },
-    {
-      "epoch": 0.7121001390820584,
-      "grad_norm": 3.5005767345428467,
-      "learning_rate": 4.638781018059357e-06,
-      "loss": 0.7588,
-      "step": 15360
-    },
-    {
-      "epoch": 0.7123319425127492,
-      "grad_norm": 2.943761110305786,
-      "learning_rate": 4.631952438947021e-06,
-      "loss": 0.8174,
-      "step": 15365
-    },
-    {
-      "epoch": 0.71256374594344,
-      "grad_norm": 4.694367408752441,
-      "learning_rate": 4.625127374393185e-06,
-      "loss": 0.9243,
-      "step": 15370
-    },
-    {
-      "epoch": 0.7127955493741307,
-      "grad_norm": 4.095063209533691,
-      "learning_rate": 4.618305828866344e-06,
-      "loss": 0.8334,
-      "step": 15375
-    },
-    {
-      "epoch": 0.7130273528048215,
-      "grad_norm": 3.5834052562713623,
-      "learning_rate": 4.611487806832681e-06,
-      "loss": 0.7467,
-      "step": 15380
-    },
-    {
-      "epoch": 0.7132591562355123,
-      "grad_norm": 3.4568352699279785,
-      "learning_rate": 4.6046733127560835e-06,
-      "loss": 0.768,
-      "step": 15385
-    },
-    {
-      "epoch": 0.713490959666203,
-      "grad_norm": 4.342470645904541,
-      "learning_rate": 4.597862351098132e-06,
-      "loss": 0.9905,
-      "step": 15390
-    },
-    {
-      "epoch": 0.7137227630968939,
-      "grad_norm": 3.9144160747528076,
-      "learning_rate": 4.591054926318074e-06,
-      "loss": 0.8793,
-      "step": 15395
-    },
-    {
-      "epoch": 0.7139545665275846,
-      "grad_norm": 3.455382823944092,
-      "learning_rate": 4.584251042872859e-06,
-      "loss": 0.7232,
-      "step": 15400
-    },
-    {
-      "epoch": 0.7139545665275846,
-      "eval_loss": 0.8931986093521118,
-      "eval_runtime": 11.2766,
-      "eval_samples_per_second": 11.262,
-      "eval_steps_per_second": 11.262,
-      "step": 15400
-    },
-    {
-      "epoch": 0.7141863699582753,
-      "grad_norm": 4.558581829071045,
-      "learning_rate": 4.577450705217118e-06,
-      "loss": 0.8319,
-      "step": 15405
-    },
-    {
-      "epoch": 0.7144181733889662,
-      "grad_norm": 3.382648468017578,
-      "learning_rate": 4.5706539178031504e-06,
-      "loss": 0.8465,
-      "step": 15410
-    },
-    {
-      "epoch": 0.7146499768196569,
-      "grad_norm": 4.303096771240234,
-      "learning_rate": 4.5638606850809374e-06,
-      "loss": 0.9637,
-      "step": 15415
-    },
-    {
-      "epoch": 0.7148817802503477,
-      "grad_norm": 3.2249696254730225,
-      "learning_rate": 4.557071011498138e-06,
-      "loss": 0.8946,
-      "step": 15420
-    },
-    {
-      "epoch": 0.7151135836810385,
-      "grad_norm": 4.005526065826416,
-      "learning_rate": 4.550284901500072e-06,
-      "loss": 0.842,
-      "step": 15425
-    },
-    {
-      "epoch": 0.7153453871117292,
-      "grad_norm": 4.499578952789307,
-      "learning_rate": 4.543502359529724e-06,
-      "loss": 0.9361,
-      "step": 15430
-    },
-    {
-      "epoch": 0.7155771905424201,
-      "grad_norm": 3.953172206878662,
-      "learning_rate": 4.536723390027751e-06,
-      "loss": 0.9228,
-      "step": 15435
-    },
-    {
-      "epoch": 0.7158089939731108,
-      "grad_norm": 3.8294663429260254,
-      "learning_rate": 4.529947997432472e-06,
-      "loss": 0.7677,
-      "step": 15440
-    },
-    {
-      "epoch": 0.7160407974038016,
-      "grad_norm": 3.6534347534179688,
-      "learning_rate": 4.523176186179853e-06,
-      "loss": 0.8623,
-      "step": 15445
-    },
-    {
-      "epoch": 0.7162726008344924,
-      "grad_norm": 3.6950314044952393,
-      "learning_rate": 4.516407960703529e-06,
-      "loss": 0.8982,
-      "step": 15450
-    },
-    {
-      "epoch": 0.7165044042651831,
-      "grad_norm": 4.337798595428467,
-      "learning_rate": 4.509643325434775e-06,
-      "loss": 0.9184,
-      "step": 15455
-    },
-    {
-      "epoch": 0.7167362076958739,
-      "grad_norm": 3.7928266525268555,
-      "learning_rate": 4.502882284802526e-06,
-      "loss": 0.9644,
-      "step": 15460
-    },
-    {
-      "epoch": 0.7169680111265647,
-      "grad_norm": 3.2648489475250244,
-      "learning_rate": 4.496124843233351e-06,
-      "loss": 0.7637,
-      "step": 15465
-    },
-    {
-      "epoch": 0.7171998145572555,
-      "grad_norm": 3.6675827503204346,
-      "learning_rate": 4.489371005151476e-06,
-      "loss": 0.7412,
-      "step": 15470
-    },
-    {
-      "epoch": 0.7174316179879462,
-      "grad_norm": 3.5607333183288574,
-      "learning_rate": 4.482620774978763e-06,
-      "loss": 0.889,
-      "step": 15475
-    },
-    {
-      "epoch": 0.717663421418637,
-      "grad_norm": 3.4801552295684814,
-      "learning_rate": 4.475874157134709e-06,
-      "loss": 0.8214,
-      "step": 15480
-    },
-    {
-      "epoch": 0.7178952248493278,
-      "grad_norm": 3.724231004714966,
-      "learning_rate": 4.469131156036441e-06,
-      "loss": 0.673,
-      "step": 15485
-    },
-    {
-      "epoch": 0.7181270282800185,
-      "grad_norm": 4.352002143859863,
-      "learning_rate": 4.462391776098736e-06,
-      "loss": 0.9104,
-      "step": 15490
-    },
-    {
-      "epoch": 0.7183588317107094,
-      "grad_norm": 3.5941615104675293,
-      "learning_rate": 4.455656021733977e-06,
-      "loss": 0.8045,
-      "step": 15495
-    },
-    {
-      "epoch": 0.7185906351414001,
-      "grad_norm": 4.01120662689209,
-      "learning_rate": 4.448923897352193e-06,
-      "loss": 0.8604,
-      "step": 15500
-    },
-    {
-      "epoch": 0.7185906351414001,
-      "eval_loss": 0.8920189738273621,
-      "eval_runtime": 11.2594,
-      "eval_samples_per_second": 11.279,
-      "eval_steps_per_second": 11.279,
-      "step": 15500
-    },
-    {
-      "epoch": 0.7188224385720908,
-      "grad_norm": 3.968578815460205,
-      "learning_rate": 4.442195407361029e-06,
-      "loss": 0.8634,
-      "step": 15505
-    },
-    {
-      "epoch": 0.7190542420027817,
-      "grad_norm": 3.571563243865967,
-      "learning_rate": 4.435470556165747e-06,
-      "loss": 0.9375,
-      "step": 15510
-    },
-    {
-      "epoch": 0.7192860454334724,
-      "grad_norm": 3.326970100402832,
-      "learning_rate": 4.428749348169228e-06,
-      "loss": 0.7336,
-      "step": 15515
-    },
-    {
-      "epoch": 0.7195178488641631,
-      "grad_norm": 4.140428066253662,
-      "learning_rate": 4.4220317877719684e-06,
-      "loss": 0.8382,
-      "step": 15520
-    },
-    {
-      "epoch": 0.719749652294854,
-      "grad_norm": 3.8773670196533203,
-      "learning_rate": 4.415317879372086e-06,
-      "loss": 0.8804,
-      "step": 15525
-    },
-    {
-      "epoch": 0.7199814557255447,
-      "grad_norm": 4.370347023010254,
-      "learning_rate": 4.408607627365289e-06,
-      "loss": 0.9379,
-      "step": 15530
-    },
-    {
-      "epoch": 0.7202132591562355,
-      "grad_norm": 3.609670400619507,
-      "learning_rate": 4.401901036144909e-06,
-      "loss": 0.7642,
-      "step": 15535
-    },
-    {
-      "epoch": 0.7204450625869263,
-      "grad_norm": 4.242509841918945,
-      "learning_rate": 4.395198110101866e-06,
-      "loss": 0.8219,
-      "step": 15540
-    },
-    {
-      "epoch": 0.720676866017617,
-      "grad_norm": 3.3745124340057373,
-      "learning_rate": 4.388498853624696e-06,
-      "loss": 0.8442,
-      "step": 15545
-    },
-    {
-      "epoch": 0.7209086694483078,
-      "grad_norm": 3.047438859939575,
-      "learning_rate": 4.381803271099515e-06,
-      "loss": 0.6754,
-      "step": 15550
-    },
-    {
-      "epoch": 0.7211404728789986,
-      "grad_norm": 4.263495922088623,
-      "learning_rate": 4.375111366910049e-06,
-      "loss": 0.873,
-      "step": 15555
-    },
-    {
-      "epoch": 0.7213722763096894,
-      "grad_norm": 3.8155529499053955,
-      "learning_rate": 4.368423145437611e-06,
-      "loss": 0.9339,
-      "step": 15560
-    },
-    {
-      "epoch": 0.7216040797403802,
-      "grad_norm": 3.9193153381347656,
-      "learning_rate": 4.3617386110611e-06,
-      "loss": 0.9423,
-      "step": 15565
-    },
-    {
-      "epoch": 0.721835883171071,
-      "grad_norm": 3.6092798709869385,
-      "learning_rate": 4.355057768156997e-06,
-      "loss": 0.8167,
-      "step": 15570
-    },
-    {
-      "epoch": 0.7220676866017617,
-      "grad_norm": 3.869521141052246,
-      "learning_rate": 4.348380621099376e-06,
-      "loss": 0.9408,
-      "step": 15575
-    },
-    {
-      "epoch": 0.7222994900324525,
-      "grad_norm": 4.508119583129883,
-      "learning_rate": 4.3417071742598895e-06,
-      "loss": 0.8851,
-      "step": 15580
-    },
-    {
-      "epoch": 0.7225312934631433,
-      "grad_norm": 3.347935199737549,
-      "learning_rate": 4.335037432007758e-06,
-      "loss": 0.6706,
-      "step": 15585
-    },
-    {
-      "epoch": 0.722763096893834,
-      "grad_norm": 3.8962862491607666,
-      "learning_rate": 4.328371398709786e-06,
-      "loss": 0.8095,
-      "step": 15590
-    },
-    {
-      "epoch": 0.7229949003245248,
-      "grad_norm": 3.721971273422241,
-      "learning_rate": 4.3217090787303564e-06,
-      "loss": 0.8029,
-      "step": 15595
-    },
-    {
-      "epoch": 0.7232267037552156,
-      "grad_norm": 3.0758113861083984,
-      "learning_rate": 4.315050476431395e-06,
-      "loss": 0.7492,
-      "step": 15600
-    },
-    {
-      "epoch": 0.7232267037552156,
-      "eval_loss": 0.8914928436279297,
-      "eval_runtime": 11.2606,
-      "eval_samples_per_second": 11.278,
-      "eval_steps_per_second": 11.278,
-      "step": 15600
-    },
-    {
-      "epoch": 0.7234585071859063,
-      "grad_norm": 3.8786678314208984,
-      "learning_rate": 4.308395596172418e-06,
-      "loss": 0.9624,
-      "step": 15605
-    },
-    {
-      "epoch": 0.7236903106165972,
-      "grad_norm": 3.3441412448883057,
-      "learning_rate": 4.301744442310499e-06,
-      "loss": 0.7183,
-      "step": 15610
-    },
-    {
-      "epoch": 0.7239221140472879,
-      "grad_norm": 4.816915988922119,
-      "learning_rate": 4.295097019200264e-06,
-      "loss": 0.8229,
-      "step": 15615
-    },
-    {
-      "epoch": 0.7241539174779786,
-      "grad_norm": 3.833200216293335,
-      "learning_rate": 4.288453331193904e-06,
-      "loss": 0.7384,
-      "step": 15620
-    },
-    {
-      "epoch": 0.7243857209086695,
-      "grad_norm": 3.196971893310547,
-      "learning_rate": 4.281813382641169e-06,
-      "loss": 0.8744,
-      "step": 15625
-    },
-    {
-      "epoch": 0.7246175243393602,
-      "grad_norm": 3.8103346824645996,
-      "learning_rate": 4.275177177889347e-06,
-      "loss": 0.8335,
-      "step": 15630
-    },
-    {
-      "epoch": 0.724849327770051,
-      "grad_norm": 3.861053228378296,
-      "learning_rate": 4.268544721283281e-06,
-      "loss": 0.744,
-      "step": 15635
-    },
-    {
-      "epoch": 0.7250811312007418,
-      "grad_norm": 3.6687607765197754,
-      "learning_rate": 4.261916017165364e-06,
-      "loss": 0.9155,
-      "step": 15640
-    },
-    {
-      "epoch": 0.7253129346314325,
-      "grad_norm": 3.7486867904663086,
-      "learning_rate": 4.2552910698755325e-06,
-      "loss": 0.862,
-      "step": 15645
-    },
-    {
-      "epoch": 0.7255447380621233,
-      "grad_norm": 3.7146894931793213,
-      "learning_rate": 4.2486698837512545e-06,
-      "loss": 0.8686,
-      "step": 15650
-    },
-    {
-      "epoch": 0.7257765414928141,
-      "grad_norm": 4.370089054107666,
-      "learning_rate": 4.242052463127545e-06,
-      "loss": 0.8455,
-      "step": 15655
-    },
-    {
-      "epoch": 0.7260083449235049,
-      "grad_norm": 3.910057306289673,
-      "learning_rate": 4.235438812336946e-06,
-      "loss": 0.9133,
-      "step": 15660
-    },
-    {
-      "epoch": 0.7262401483541956,
-      "grad_norm": 3.6310927867889404,
-      "learning_rate": 4.228828935709541e-06,
-      "loss": 0.874,
-      "step": 15665
-    },
-    {
-      "epoch": 0.7264719517848864,
-      "grad_norm": 3.224461317062378,
-      "learning_rate": 4.222222837572929e-06,
-      "loss": 0.6786,
-      "step": 15670
-    },
-    {
-      "epoch": 0.7267037552155772,
-      "grad_norm": 3.6823716163635254,
-      "learning_rate": 4.215620522252247e-06,
-      "loss": 0.815,
-      "step": 15675
-    },
-    {
-      "epoch": 0.7269355586462679,
-      "grad_norm": 3.2620420455932617,
-      "learning_rate": 4.209021994070155e-06,
-      "loss": 0.7819,
-      "step": 15680
-    },
-    {
-      "epoch": 0.7271673620769588,
-      "grad_norm": 3.7991809844970703,
-      "learning_rate": 4.202427257346825e-06,
-      "loss": 0.8341,
-      "step": 15685
-    },
-    {
-      "epoch": 0.7273991655076495,
-      "grad_norm": 3.8811824321746826,
-      "learning_rate": 4.195836316399948e-06,
-      "loss": 0.8063,
-      "step": 15690
-    },
-    {
-      "epoch": 0.7276309689383403,
-      "grad_norm": 3.89102840423584,
-      "learning_rate": 4.18924917554474e-06,
-      "loss": 0.8389,
-      "step": 15695
-    },
-    {
-      "epoch": 0.7278627723690311,
-      "grad_norm": 3.7965166568756104,
-      "learning_rate": 4.182665839093916e-06,
-      "loss": 0.7412,
-      "step": 15700
-    },
-    {
-      "epoch": 0.7278627723690311,
-      "eval_loss": 0.8913096785545349,
-      "eval_runtime": 11.2647,
-      "eval_samples_per_second": 11.274,
-      "eval_steps_per_second": 11.274,
-      "step": 15700
-    },
-    {
-      "epoch": 0.7280945757997218,
-      "grad_norm": 3.3673911094665527,
-      "learning_rate": 4.1760863113577085e-06,
-      "loss": 0.8514,
-      "step": 15705
-    },
-    {
-      "epoch": 0.7283263792304127,
-      "grad_norm": 3.435807704925537,
-      "learning_rate": 4.169510596643859e-06,
-      "loss": 0.861,
-      "step": 15710
-    },
-    {
-      "epoch": 0.7285581826611034,
-      "grad_norm": 4.1871137619018555,
-      "learning_rate": 4.1629386992576025e-06,
-      "loss": 0.9678,
-      "step": 15715
-    },
-    {
-      "epoch": 0.7287899860917941,
-      "grad_norm": 4.157339096069336,
-      "learning_rate": 4.156370623501678e-06,
-      "loss": 0.8808,
-      "step": 15720
-    },
-    {
-      "epoch": 0.729021789522485,
-      "grad_norm": 3.5089452266693115,
-      "learning_rate": 4.149806373676328e-06,
-      "loss": 0.8607,
-      "step": 15725
-    },
-    {
-      "epoch": 0.7292535929531757,
-      "grad_norm": 3.601043462753296,
-      "learning_rate": 4.1432459540792866e-06,
-      "loss": 0.872,
-      "step": 15730
-    },
-    {
-      "epoch": 0.7294853963838664,
-      "grad_norm": 3.9285526275634766,
-      "learning_rate": 4.136689369005776e-06,
-      "loss": 0.8477,
-      "step": 15735
-    },
-    {
-      "epoch": 0.7297171998145573,
-      "grad_norm": 3.5868418216705322,
-      "learning_rate": 4.130136622748516e-06,
-      "loss": 0.8735,
-      "step": 15740
-    },
-    {
-      "epoch": 0.729949003245248,
-      "grad_norm": 3.829298496246338,
-      "learning_rate": 4.12358771959771e-06,
-      "loss": 0.7792,
-      "step": 15745
-    },
-    {
-      "epoch": 0.7301808066759388,
-      "grad_norm": 2.9413342475891113,
-      "learning_rate": 4.1170426638410436e-06,
-      "loss": 0.7645,
-      "step": 15750
-    },
-    {
-      "epoch": 0.7304126101066296,
-      "grad_norm": 3.4420745372772217,
-      "learning_rate": 4.1105014597636775e-06,
-      "loss": 0.9818,
-      "step": 15755
-    },
-    {
-      "epoch": 0.7306444135373203,
-      "grad_norm": 3.5504422187805176,
-      "learning_rate": 4.1039641116482605e-06,
-      "loss": 0.7684,
-      "step": 15760
-    },
-    {
-      "epoch": 0.7308762169680111,
-      "grad_norm": 3.317941665649414,
-      "learning_rate": 4.0974306237749205e-06,
-      "loss": 0.6928,
-      "step": 15765
-    },
-    {
-      "epoch": 0.7311080203987019,
-      "grad_norm": 3.58900785446167,
-      "learning_rate": 4.090901000421244e-06,
-      "loss": 0.7373,
-      "step": 15770
-    },
-    {
-      "epoch": 0.7313398238293927,
-      "grad_norm": 3.978079080581665,
-      "learning_rate": 4.084375245862293e-06,
-      "loss": 0.7678,
-      "step": 15775
-    },
-    {
-      "epoch": 0.7315716272600834,
-      "grad_norm": 3.7968006134033203,
-      "learning_rate": 4.0778533643706e-06,
-      "loss": 0.6788,
-      "step": 15780
-    },
-    {
-      "epoch": 0.7318034306907742,
-      "grad_norm": 3.2984938621520996,
-      "learning_rate": 4.071335360216164e-06,
-      "loss": 0.749,
-      "step": 15785
-    },
-    {
-      "epoch": 0.732035234121465,
-      "grad_norm": 3.3127284049987793,
-      "learning_rate": 4.064821237666433e-06,
-      "loss": 0.8255,
-      "step": 15790
-    },
-    {
-      "epoch": 0.7322670375521557,
-      "grad_norm": 3.829406499862671,
-      "learning_rate": 4.058311000986326e-06,
-      "loss": 0.8941,
-      "step": 15795
-    },
-    {
-      "epoch": 0.7324988409828466,
-      "grad_norm": 3.6632189750671387,
-      "learning_rate": 4.05180465443822e-06,
-      "loss": 0.9017,
-      "step": 15800
-    },
-    {
-      "epoch": 0.7324988409828466,
-      "eval_loss": 0.890110969543457,
-      "eval_runtime": 11.284,
-      "eval_samples_per_second": 11.255,
-      "eval_steps_per_second": 11.255,
-      "step": 15800
-    },
-    {
-      "epoch": 0.7327306444135373,
-      "grad_norm": 3.796661138534546,
-      "learning_rate": 4.045302202281926e-06,
-      "loss": 0.8491,
-      "step": 15805
-    },
-    {
-      "epoch": 0.7329624478442281,
-      "grad_norm": 4.159597396850586,
-      "learning_rate": 4.0388036487747225e-06,
-      "loss": 0.8807,
-      "step": 15810
-    },
-    {
-      "epoch": 0.7331942512749189,
-      "grad_norm": 4.575572490692139,
-      "learning_rate": 4.032308998171336e-06,
-      "loss": 0.8274,
-      "step": 15815
-    },
-    {
-      "epoch": 0.7334260547056096,
-      "grad_norm": 3.7240734100341797,
-      "learning_rate": 4.025818254723925e-06,
-      "loss": 1.0875,
-      "step": 15820
-    },
-    {
-      "epoch": 0.7336578581363005,
-      "grad_norm": 3.9850006103515625,
-      "learning_rate": 4.019331422682101e-06,
-      "loss": 1.015,
-      "step": 15825
-    },
-    {
-      "epoch": 0.7338896615669912,
-      "grad_norm": 3.4893457889556885,
-      "learning_rate": 4.0128485062929145e-06,
-      "loss": 0.794,
-      "step": 15830
-    },
-    {
-      "epoch": 0.7341214649976819,
-      "grad_norm": 3.174949884414673,
-      "learning_rate": 4.006369509800846e-06,
-      "loss": 0.7296,
-      "step": 15835
-    },
-    {
-      "epoch": 0.7343532684283728,
-      "grad_norm": 3.194807529449463,
-      "learning_rate": 3.999894437447807e-06,
-      "loss": 0.8823,
-      "step": 15840
-    },
-    {
-      "epoch": 0.7345850718590635,
-      "grad_norm": 4.552024841308594,
-      "learning_rate": 3.993423293473152e-06,
-      "loss": 0.9614,
-      "step": 15845
-    },
-    {
-      "epoch": 0.7348168752897543,
-      "grad_norm": 3.7308084964752197,
-      "learning_rate": 3.986956082113658e-06,
-      "loss": 0.9514,
-      "step": 15850
-    },
-    {
-      "epoch": 0.7350486787204451,
-      "grad_norm": 3.8906126022338867,
-      "learning_rate": 3.980492807603519e-06,
-      "loss": 0.8865,
-      "step": 15855
-    },
-    {
-      "epoch": 0.7352804821511358,
-      "grad_norm": 3.130915641784668,
-      "learning_rate": 3.974033474174367e-06,
-      "loss": 0.7193,
-      "step": 15860
-    },
-    {
-      "epoch": 0.7355122855818266,
-      "grad_norm": 4.078449726104736,
-      "learning_rate": 3.967578086055237e-06,
-      "loss": 0.7484,
-      "step": 15865
-    },
-    {
-      "epoch": 0.7357440890125174,
-      "grad_norm": 3.706937074661255,
-      "learning_rate": 3.961126647472596e-06,
-      "loss": 0.8345,
-      "step": 15870
-    },
-    {
-      "epoch": 0.7359758924432082,
-      "grad_norm": 3.131194591522217,
-      "learning_rate": 3.954679162650313e-06,
-      "loss": 0.6977,
-      "step": 15875
-    },
-    {
-      "epoch": 0.7362076958738989,
-      "grad_norm": 3.336714029312134,
-      "learning_rate": 3.948235635809675e-06,
-      "loss": 0.7781,
-      "step": 15880
-    },
-    {
-      "epoch": 0.7364394993045897,
-      "grad_norm": 3.725440502166748,
-      "learning_rate": 3.941796071169383e-06,
-      "loss": 0.7968,
-      "step": 15885
-    },
-    {
-      "epoch": 0.7366713027352805,
-      "grad_norm": 3.469977378845215,
-      "learning_rate": 3.9353604729455296e-06,
-      "loss": 0.9139,
-      "step": 15890
-    },
-    {
-      "epoch": 0.7369031061659712,
-      "grad_norm": 4.227980136871338,
-      "learning_rate": 3.928928845351617e-06,
-      "loss": 0.8296,
-      "step": 15895
-    },
-    {
-      "epoch": 0.7371349095966621,
-      "grad_norm": 3.965374708175659,
-      "learning_rate": 3.922501192598557e-06,
-      "loss": 0.8702,
-      "step": 15900
-    },
-    {
-      "epoch": 0.7371349095966621,
-      "eval_loss": 0.8892138004302979,
-      "eval_runtime": 11.2674,
-      "eval_samples_per_second": 11.271,
-      "eval_steps_per_second": 11.271,
-      "step": 15900
-    },
-    {
-      "epoch": 0.7373667130273528,
-      "grad_norm": 3.8569962978363037,
-      "learning_rate": 3.916077518894642e-06,
-      "loss": 0.8873,
-      "step": 15905
-    },
-    {
-      "epoch": 0.7375985164580435,
-      "grad_norm": 3.320935010910034,
-      "learning_rate": 3.909657828445572e-06,
-      "loss": 0.8956,
-      "step": 15910
-    },
-    {
-      "epoch": 0.7378303198887344,
-      "grad_norm": 3.896636962890625,
-      "learning_rate": 3.903242125454439e-06,
-      "loss": 1.0005,
-      "step": 15915
-    },
-    {
-      "epoch": 0.7380621233194251,
-      "grad_norm": 3.972095489501953,
-      "learning_rate": 3.896830414121717e-06,
-      "loss": 0.9519,
-      "step": 15920
-    },
-    {
-      "epoch": 0.7382939267501158,
-      "grad_norm": 3.7143542766571045,
-      "learning_rate": 3.890422698645267e-06,
-      "loss": 0.8063,
-      "step": 15925
-    },
-    {
-      "epoch": 0.7385257301808067,
-      "grad_norm": 3.7164463996887207,
-      "learning_rate": 3.88401898322034e-06,
-      "loss": 0.7474,
-      "step": 15930
-    },
-    {
-      "epoch": 0.7387575336114974,
-      "grad_norm": 4.113378047943115,
-      "learning_rate": 3.87761927203957e-06,
-      "loss": 0.9298,
-      "step": 15935
-    },
-    {
-      "epoch": 0.7389893370421883,
-      "grad_norm": 3.508679151535034,
-      "learning_rate": 3.871223569292955e-06,
-      "loss": 0.9423,
-      "step": 15940
-    },
-    {
-      "epoch": 0.739221140472879,
-      "grad_norm": 3.7724037170410156,
-      "learning_rate": 3.8648318791678844e-06,
-      "loss": 0.9195,
-      "step": 15945
-    },
-    {
-      "epoch": 0.7394529439035697,
-      "grad_norm": 4.135608673095703,
-      "learning_rate": 3.8584442058491145e-06,
-      "loss": 0.7762,
-      "step": 15950
-    },
-    {
-      "epoch": 0.7396847473342606,
-      "grad_norm": 4.278160095214844,
-      "learning_rate": 3.852060553518771e-06,
-      "loss": 0.8267,
-      "step": 15955
-    },
-    {
-      "epoch": 0.7399165507649513,
-      "grad_norm": 4.063554286956787,
-      "learning_rate": 3.8456809263563425e-06,
-      "loss": 0.941,
-      "step": 15960
-    },
-    {
-      "epoch": 0.7401483541956421,
-      "grad_norm": 4.410874843597412,
-      "learning_rate": 3.839305328538688e-06,
-      "loss": 0.8817,
-      "step": 15965
-    },
-    {
-      "epoch": 0.7403801576263329,
-      "grad_norm": 4.527900218963623,
-      "learning_rate": 3.832933764240036e-06,
-      "loss": 1.025,
-      "step": 15970
-    },
-    {
-      "epoch": 0.7406119610570236,
-      "grad_norm": 3.8349719047546387,
-      "learning_rate": 3.8265662376319545e-06,
-      "loss": 0.6803,
-      "step": 15975
-    },
-    {
-      "epoch": 0.7408437644877144,
-      "grad_norm": 4.201470375061035,
-      "learning_rate": 3.820202752883389e-06,
-      "loss": 0.7962,
-      "step": 15980
-    },
-    {
-      "epoch": 0.7410755679184052,
-      "grad_norm": 4.27456521987915,
-      "learning_rate": 3.8138433141606233e-06,
-      "loss": 0.83,
-      "step": 15985
-    },
-    {
-      "epoch": 0.741307371349096,
-      "grad_norm": 3.525296926498413,
-      "learning_rate": 3.8074879256272958e-06,
-      "loss": 0.8171,
-      "step": 15990
-    },
-    {
-      "epoch": 0.7415391747797867,
-      "grad_norm": 3.365690231323242,
-      "learning_rate": 3.8011365914443977e-06,
-      "loss": 0.7481,
-      "step": 15995
-    },
-    {
-      "epoch": 0.7417709782104775,
-      "grad_norm": 3.8495423793792725,
-      "learning_rate": 3.7947893157702663e-06,
-      "loss": 0.8634,
-      "step": 16000
-    },
-    {
-      "epoch": 0.7417709782104775,
-      "eval_loss": 0.8883469700813293,
-      "eval_runtime": 11.2669,
-      "eval_samples_per_second": 11.272,
-      "eval_steps_per_second": 11.272,
-      "step": 16000
-    },
-    {
-      "epoch": 0.7420027816411683,
-      "grad_norm": 3.318895101547241,
-      "learning_rate": 3.7884461027605756e-06,
-      "loss": 0.7934,
-      "step": 16005
-    },
-    {
-      "epoch": 0.742234585071859,
-      "grad_norm": 4.090661525726318,
-      "learning_rate": 3.7821069565683388e-06,
-      "loss": 0.7891,
-      "step": 16010
-    },
-    {
-      "epoch": 0.7424663885025499,
-      "grad_norm": 3.940528631210327,
-      "learning_rate": 3.7757718813439147e-06,
-      "loss": 0.8408,
-      "step": 16015
-    },
-    {
-      "epoch": 0.7426981919332406,
-      "grad_norm": 3.4445431232452393,
-      "learning_rate": 3.769440881234996e-06,
-      "loss": 0.7104,
-      "step": 16020
-    },
-    {
-      "epoch": 0.7429299953639313,
-      "grad_norm": 3.6801605224609375,
-      "learning_rate": 3.763113960386595e-06,
-      "loss": 0.8792,
-      "step": 16025
-    },
-    {
-      "epoch": 0.7431617987946222,
-      "grad_norm": 3.427544116973877,
-      "learning_rate": 3.756791122941068e-06,
-      "loss": 0.754,
-      "step": 16030
-    },
-    {
-      "epoch": 0.7433936022253129,
-      "grad_norm": 4.70984411239624,
-      "learning_rate": 3.750472373038093e-06,
-      "loss": 0.868,
-      "step": 16035
-    },
-    {
-      "epoch": 0.7436254056560037,
-      "grad_norm": 3.9222209453582764,
-      "learning_rate": 3.744157714814669e-06,
-      "loss": 0.8452,
-      "step": 16040
-    },
-    {
-      "epoch": 0.7438572090866945,
-      "grad_norm": 3.9630048274993896,
-      "learning_rate": 3.737847152405113e-06,
-      "loss": 0.8538,
-      "step": 16045
-    },
-    {
-      "epoch": 0.7440890125173852,
-      "grad_norm": 4.065320014953613,
-      "learning_rate": 3.73154068994107e-06,
-      "loss": 0.8741,
-      "step": 16050
-    },
-    {
-      "epoch": 0.7443208159480761,
-      "grad_norm": 3.2823739051818848,
-      "learning_rate": 3.7252383315514974e-06,
-      "loss": 0.7499,
-      "step": 16055
-    },
-    {
-      "epoch": 0.7445526193787668,
-      "grad_norm": 4.034680366516113,
-      "learning_rate": 3.7189400813626577e-06,
-      "loss": 1.0117,
-      "step": 16060
-    },
-    {
-      "epoch": 0.7447844228094576,
-      "grad_norm": 3.311304807662964,
-      "learning_rate": 3.7126459434981377e-06,
-      "loss": 0.6851,
-      "step": 16065
-    },
-    {
-      "epoch": 0.7450162262401484,
-      "grad_norm": 3.439509391784668,
-      "learning_rate": 3.706355922078817e-06,
-      "loss": 0.948,
-      "step": 16070
-    },
-    {
-      "epoch": 0.7452480296708391,
-      "grad_norm": 3.8146088123321533,
-      "learning_rate": 3.7000700212228937e-06,
-      "loss": 0.814,
-      "step": 16075
-    },
-    {
-      "epoch": 0.7454798331015299,
-      "grad_norm": 3.7652645111083984,
-      "learning_rate": 3.693788245045854e-06,
-      "loss": 0.917,
-      "step": 16080
-    },
-    {
-      "epoch": 0.7457116365322207,
-      "grad_norm": 3.8624465465545654,
-      "learning_rate": 3.687510597660495e-06,
-      "loss": 0.8091,
-      "step": 16085
-    },
-    {
-      "epoch": 0.7459434399629115,
-      "grad_norm": 3.9942777156829834,
-      "learning_rate": 3.6812370831769117e-06,
-      "loss": 0.7945,
-      "step": 16090
-    },
-    {
-      "epoch": 0.7461752433936022,
-      "grad_norm": 3.569929361343384,
-      "learning_rate": 3.6749677057024834e-06,
-      "loss": 0.8493,
-      "step": 16095
-    },
-    {
-      "epoch": 0.746407046824293,
-      "grad_norm": 3.523303508758545,
-      "learning_rate": 3.6687024693418838e-06,
-      "loss": 0.803,
-      "step": 16100
-    },
-    {
-      "epoch": 0.746407046824293,
-      "eval_loss": 0.8879801034927368,
-      "eval_runtime": 11.2743,
-      "eval_samples_per_second": 11.265,
-      "eval_steps_per_second": 11.265,
-      "step": 16100
-    },
-    {
-      "epoch": 0.7466388502549838,
-      "grad_norm": 3.59873366355896,
-      "learning_rate": 3.6624413781970815e-06,
-      "loss": 0.8011,
-      "step": 16105
-    },
-    {
-      "epoch": 0.7468706536856745,
-      "grad_norm": 3.186870813369751,
-      "learning_rate": 3.6561844363673215e-06,
-      "loss": 0.6644,
-      "step": 16110
-    },
-    {
-      "epoch": 0.7471024571163654,
-      "grad_norm": 4.058816432952881,
-      "learning_rate": 3.6499316479491408e-06,
-      "loss": 0.9662,
-      "step": 16115
-    },
-    {
-      "epoch": 0.7473342605470561,
-      "grad_norm": 4.230071067810059,
-      "learning_rate": 3.643683017036358e-06,
-      "loss": 0.9362,
-      "step": 16120
-    },
-    {
-      "epoch": 0.7475660639777468,
-      "grad_norm": 4.078927516937256,
-      "learning_rate": 3.6374385477200613e-06,
-      "loss": 0.9531,
-      "step": 16125
-    },
-    {
-      "epoch": 0.7477978674084377,
-      "grad_norm": 3.8070027828216553,
-      "learning_rate": 3.631198244088615e-06,
-      "loss": 0.864,
-      "step": 16130
-    },
-    {
-      "epoch": 0.7480296708391284,
-      "grad_norm": 4.256518840789795,
-      "learning_rate": 3.6249621102276643e-06,
-      "loss": 0.816,
-      "step": 16135
-    },
-    {
-      "epoch": 0.7482614742698191,
-      "grad_norm": 4.053735256195068,
-      "learning_rate": 3.618730150220121e-06,
-      "loss": 0.8013,
-      "step": 16140
-    },
-    {
-      "epoch": 0.74849327770051,
-      "grad_norm": 4.204226970672607,
-      "learning_rate": 3.6125023681461566e-06,
-      "loss": 0.8703,
-      "step": 16145
-    },
-    {
-      "epoch": 0.7487250811312007,
-      "grad_norm": 3.6390607357025146,
-      "learning_rate": 3.6062787680832178e-06,
-      "loss": 0.7165,
-      "step": 16150
-    },
-    {
-      "epoch": 0.7489568845618915,
-      "grad_norm": 3.5119662284851074,
-      "learning_rate": 3.6000593541060123e-06,
-      "loss": 0.8844,
-      "step": 16155
-    },
-    {
-      "epoch": 0.7491886879925823,
-      "grad_norm": 3.674628496170044,
-      "learning_rate": 3.593844130286499e-06,
-      "loss": 0.817,
-      "step": 16160
-    },
-    {
-      "epoch": 0.749420491423273,
-      "grad_norm": 3.410306215286255,
-      "learning_rate": 3.5876331006938957e-06,
-      "loss": 0.8219,
-      "step": 16165
-    },
-    {
-      "epoch": 0.7496522948539638,
-      "grad_norm": 3.05379056930542,
-      "learning_rate": 3.581426269394679e-06,
-      "loss": 0.8094,
-      "step": 16170
-    },
-    {
-      "epoch": 0.7498840982846546,
-      "grad_norm": 3.567718267440796,
-      "learning_rate": 3.5752236404525797e-06,
-      "loss": 0.9139,
-      "step": 16175
-    },
-    {
-      "epoch": 0.7501159017153454,
-      "grad_norm": 3.88759708404541,
-      "learning_rate": 3.569025217928563e-06,
-      "loss": 0.9061,
-      "step": 16180
-    },
-    {
-      "epoch": 0.7503477051460362,
-      "grad_norm": 4.398759841918945,
-      "learning_rate": 3.562831005880857e-06,
-      "loss": 1.0038,
-      "step": 16185
-    },
-    {
-      "epoch": 0.750579508576727,
-      "grad_norm": 4.027905464172363,
-      "learning_rate": 3.5566410083649218e-06,
-      "loss": 0.9197,
-      "step": 16190
-    },
-    {
-      "epoch": 0.7508113120074177,
-      "grad_norm": 3.7599446773529053,
-      "learning_rate": 3.550455229433457e-06,
-      "loss": 0.9728,
-      "step": 16195
-    },
-    {
-      "epoch": 0.7510431154381085,
-      "grad_norm": 3.042257070541382,
-      "learning_rate": 3.544273673136409e-06,
-      "loss": 0.7215,
-      "step": 16200
-    },
-    {
-      "epoch": 0.7510431154381085,
-      "eval_loss": 0.8876115083694458,
-      "eval_runtime": 11.2651,
-      "eval_samples_per_second": 11.274,
-      "eval_steps_per_second": 11.274,
-      "step": 16200
-    },
-    {
-      "epoch": 0.7512749188687993,
-      "grad_norm": 3.53696870803833,
-      "learning_rate": 3.538096343520959e-06,
-      "loss": 0.7741,
-      "step": 16205
-    },
-    {
-      "epoch": 0.75150672229949,
-      "grad_norm": 3.3951306343078613,
-      "learning_rate": 3.53192324463151e-06,
-      "loss": 0.9269,
-      "step": 16210
-    },
-    {
-      "epoch": 0.7517385257301809,
-      "grad_norm": 3.8003907203674316,
-      "learning_rate": 3.5257543805097096e-06,
-      "loss": 0.807,
-      "step": 16215
-    },
-    {
-      "epoch": 0.7519703291608716,
-      "grad_norm": 3.8604257106781006,
-      "learning_rate": 3.519589755194417e-06,
-      "loss": 0.8091,
-      "step": 16220
-    },
-    {
-      "epoch": 0.7522021325915623,
-      "grad_norm": 4.3903398513793945,
-      "learning_rate": 3.513429372721733e-06,
-      "loss": 0.8131,
-      "step": 16225
-    },
-    {
-      "epoch": 0.7524339360222532,
-      "grad_norm": 3.688931703567505,
-      "learning_rate": 3.507273237124965e-06,
-      "loss": 0.9296,
-      "step": 16230
-    },
-    {
-      "epoch": 0.7526657394529439,
-      "grad_norm": 3.84499454498291,
-      "learning_rate": 3.501121352434651e-06,
-      "loss": 1.002,
-      "step": 16235
-    },
-    {
-      "epoch": 0.7528975428836346,
-      "grad_norm": 3.5966436862945557,
-      "learning_rate": 3.494973722678544e-06,
-      "loss": 0.8786,
-      "step": 16240
-    },
-    {
-      "epoch": 0.7531293463143255,
-      "grad_norm": 4.628900051116943,
-      "learning_rate": 3.4888303518816068e-06,
-      "loss": 0.9503,
-      "step": 16245
-    },
-    {
-      "epoch": 0.7533611497450162,
-      "grad_norm": 4.280179500579834,
-      "learning_rate": 3.482691244066012e-06,
-      "loss": 0.8784,
-      "step": 16250
-    },
-    {
-      "epoch": 0.753592953175707,
-      "grad_norm": 4.282437801361084,
-      "learning_rate": 3.47655640325115e-06,
-      "loss": 0.8438,
-      "step": 16255
-    },
-    {
-      "epoch": 0.7538247566063978,
-      "grad_norm": 4.042704105377197,
-      "learning_rate": 3.4704258334536156e-06,
-      "loss": 0.8948,
-      "step": 16260
-    },
-    {
-      "epoch": 0.7540565600370885,
-      "grad_norm": 4.1212239265441895,
-      "learning_rate": 3.464299538687197e-06,
-      "loss": 0.8409,
-      "step": 16265
-    },
-    {
-      "epoch": 0.7542883634677793,
-      "grad_norm": 4.124472141265869,
-      "learning_rate": 3.458177522962898e-06,
-      "loss": 0.9342,
-      "step": 16270
-    },
-    {
-      "epoch": 0.7545201668984701,
-      "grad_norm": 3.704339027404785,
-      "learning_rate": 3.4520597902889077e-06,
-      "loss": 0.8526,
-      "step": 16275
-    },
-    {
-      "epoch": 0.7547519703291609,
-      "grad_norm": 3.406463384628296,
-      "learning_rate": 3.445946344670623e-06,
-      "loss": 0.9123,
-      "step": 16280
-    },
-    {
-      "epoch": 0.7549837737598516,
-      "grad_norm": 4.044938087463379,
-      "learning_rate": 3.439837190110622e-06,
-      "loss": 1.0306,
-      "step": 16285
-    },
-    {
-      "epoch": 0.7552155771905424,
-      "grad_norm": 3.2226245403289795,
-      "learning_rate": 3.433732330608682e-06,
-      "loss": 0.7938,
-      "step": 16290
-    },
-    {
-      "epoch": 0.7554473806212332,
-      "grad_norm": 3.456693649291992,
-      "learning_rate": 3.42763177016177e-06,
-      "loss": 0.8515,
-      "step": 16295
-    },
-    {
-      "epoch": 0.7556791840519239,
-      "grad_norm": 4.75590181350708,
-      "learning_rate": 3.4215355127640303e-06,
-      "loss": 1.0084,
-      "step": 16300
-    },
-    {
-      "epoch": 0.7556791840519239,
-      "eval_loss": 0.8864190578460693,
-      "eval_runtime": 11.2633,
-      "eval_samples_per_second": 11.276,
-      "eval_steps_per_second": 11.276,
-      "step": 16300
-    },
-    {
-      "epoch": 0.7559109874826148,
-      "grad_norm": 3.5325708389282227,
-      "learning_rate": 3.4154435624067907e-06,
-      "loss": 0.8667,
-      "step": 16305
-    },
-    {
-      "epoch": 0.7561427909133055,
-      "grad_norm": 4.027612686157227,
-      "learning_rate": 3.4093559230785676e-06,
-      "loss": 0.9299,
-      "step": 16310
-    },
-    {
-      "epoch": 0.7563745943439963,
-      "grad_norm": 3.76712965965271,
-      "learning_rate": 3.403272598765044e-06,
-      "loss": 1.0016,
-      "step": 16315
-    },
-    {
-      "epoch": 0.7566063977746871,
-      "grad_norm": 3.77254581451416,
-      "learning_rate": 3.3971935934490852e-06,
-      "loss": 1.0005,
-      "step": 16320
-    },
-    {
-      "epoch": 0.7568382012053778,
-      "grad_norm": 3.5619187355041504,
-      "learning_rate": 3.39111891111073e-06,
-      "loss": 0.8961,
-      "step": 16325
-    },
-    {
-      "epoch": 0.7570700046360687,
-      "grad_norm": 3.8188278675079346,
-      "learning_rate": 3.385048555727182e-06,
-      "loss": 0.8626,
-      "step": 16330
-    },
-    {
-      "epoch": 0.7573018080667594,
-      "grad_norm": 3.299271583557129,
-      "learning_rate": 3.3789825312728086e-06,
-      "loss": 0.8992,
-      "step": 16335
-    },
-    {
-      "epoch": 0.7575336114974501,
-      "grad_norm": 4.4377360343933105,
-      "learning_rate": 3.3729208417191505e-06,
-      "loss": 0.8702,
-      "step": 16340
-    },
-    {
-      "epoch": 0.757765414928141,
-      "grad_norm": 3.683051347732544,
-      "learning_rate": 3.3668634910349085e-06,
-      "loss": 0.9598,
-      "step": 16345
-    },
-    {
-      "epoch": 0.7579972183588317,
-      "grad_norm": 4.029439926147461,
-      "learning_rate": 3.3608104831859344e-06,
-      "loss": 0.9332,
-      "step": 16350
-    },
-    {
-      "epoch": 0.7582290217895225,
-      "grad_norm": 3.681857109069824,
-      "learning_rate": 3.3547618221352473e-06,
-      "loss": 0.8892,
-      "step": 16355
-    },
-    {
-      "epoch": 0.7584608252202133,
-      "grad_norm": 3.72904896736145,
-      "learning_rate": 3.348717511843016e-06,
-      "loss": 0.8558,
-      "step": 16360
-    },
-    {
-      "epoch": 0.758692628650904,
-      "grad_norm": 4.226546764373779,
-      "learning_rate": 3.3426775562665605e-06,
-      "loss": 0.8188,
-      "step": 16365
-    },
-    {
-      "epoch": 0.7589244320815948,
-      "grad_norm": 4.173439025878906,
-      "learning_rate": 3.336641959360345e-06,
-      "loss": 0.9299,
-      "step": 16370
-    },
-    {
-      "epoch": 0.7591562355122856,
-      "grad_norm": 2.8630075454711914,
-      "learning_rate": 3.3306107250759867e-06,
-      "loss": 0.761,
-      "step": 16375
-    },
-    {
-      "epoch": 0.7593880389429764,
-      "grad_norm": 3.6998066902160645,
-      "learning_rate": 3.3245838573622503e-06,
-      "loss": 0.74,
-      "step": 16380
-    },
-    {
-      "epoch": 0.7596198423736671,
-      "grad_norm": 3.66335129737854,
-      "learning_rate": 3.318561360165027e-06,
-      "loss": 0.758,
-      "step": 16385
-    },
-    {
-      "epoch": 0.7598516458043579,
-      "grad_norm": 4.026821136474609,
-      "learning_rate": 3.3125432374273638e-06,
-      "loss": 0.8289,
-      "step": 16390
-    },
-    {
-      "epoch": 0.7600834492350487,
-      "grad_norm": 3.9597976207733154,
-      "learning_rate": 3.3065294930894308e-06,
-      "loss": 0.8404,
-      "step": 16395
-    },
-    {
-      "epoch": 0.7603152526657394,
-      "grad_norm": 3.870403528213501,
-      "learning_rate": 3.3005201310885326e-06,
-      "loss": 0.8141,
-      "step": 16400
-    },
-    {
-      "epoch": 0.7603152526657394,
-      "eval_loss": 0.8859652280807495,
-      "eval_runtime": 11.2773,
-      "eval_samples_per_second": 11.262,
-      "eval_steps_per_second": 11.262,
-      "step": 16400
-    },
-    {
-      "epoch": 0.7605470560964303,
-      "grad_norm": 3.7531723976135254,
-      "learning_rate": 3.2945151553591127e-06,
-      "loss": 0.9562,
-      "step": 16405
-    },
-    {
-      "epoch": 0.760778859527121,
-      "grad_norm": 3.4687469005584717,
-      "learning_rate": 3.288514569832739e-06,
-      "loss": 0.9494,
-      "step": 16410
-    },
-    {
-      "epoch": 0.7610106629578117,
-      "grad_norm": 4.315207004547119,
-      "learning_rate": 3.2825183784380997e-06,
-      "loss": 0.9185,
-      "step": 16415
-    },
-    {
-      "epoch": 0.7612424663885026,
-      "grad_norm": 5.365420818328857,
-      "learning_rate": 3.2765265851010164e-06,
-      "loss": 0.9603,
-      "step": 16420
-    },
-    {
-      "epoch": 0.7614742698191933,
-      "grad_norm": 3.6974503993988037,
-      "learning_rate": 3.2705391937444188e-06,
-      "loss": 0.7098,
-      "step": 16425
-    },
-    {
-      "epoch": 0.7617060732498842,
-      "grad_norm": 3.763038396835327,
-      "learning_rate": 3.2645562082883665e-06,
-      "loss": 0.7566,
-      "step": 16430
-    },
-    {
-      "epoch": 0.7619378766805749,
-      "grad_norm": 3.7737200260162354,
-      "learning_rate": 3.258577632650024e-06,
-      "loss": 0.7182,
-      "step": 16435
-    },
-    {
-      "epoch": 0.7621696801112656,
-      "grad_norm": 3.475022554397583,
-      "learning_rate": 3.252603470743676e-06,
-      "loss": 0.7762,
-      "step": 16440
-    },
-    {
-      "epoch": 0.7624014835419565,
-      "grad_norm": 3.917085647583008,
-      "learning_rate": 3.246633726480719e-06,
-      "loss": 0.807,
-      "step": 16445
-    },
-    {
-      "epoch": 0.7626332869726472,
-      "grad_norm": 3.9613709449768066,
-      "learning_rate": 3.240668403769649e-06,
-      "loss": 0.9799,
-      "step": 16450
-    },
-    {
-      "epoch": 0.7628650904033379,
-      "grad_norm": 3.5182406902313232,
-      "learning_rate": 3.2347075065160693e-06,
-      "loss": 0.7412,
-      "step": 16455
-    },
-    {
-      "epoch": 0.7630968938340288,
-      "grad_norm": 3.8503899574279785,
-      "learning_rate": 3.2287510386226916e-06,
-      "loss": 0.9763,
-      "step": 16460
-    },
-    {
-      "epoch": 0.7633286972647195,
-      "grad_norm": 3.6205215454101562,
-      "learning_rate": 3.2227990039893253e-06,
-      "loss": 0.932,
-      "step": 16465
-    },
-    {
-      "epoch": 0.7635605006954103,
-      "grad_norm": 4.192770481109619,
-      "learning_rate": 3.2168514065128697e-06,
-      "loss": 0.8804,
-      "step": 16470
-    },
-    {
-      "epoch": 0.7637923041261011,
-      "grad_norm": 3.577810049057007,
-      "learning_rate": 3.2109082500873334e-06,
-      "loss": 0.7342,
-      "step": 16475
-    },
-    {
-      "epoch": 0.7640241075567918,
-      "grad_norm": 3.62418532371521,
-      "learning_rate": 3.204969538603805e-06,
-      "loss": 0.8044,
-      "step": 16480
-    },
-    {
-      "epoch": 0.7642559109874826,
-      "grad_norm": 3.3772356510162354,
-      "learning_rate": 3.199035275950463e-06,
-      "loss": 0.8625,
-      "step": 16485
-    },
-    {
-      "epoch": 0.7644877144181734,
-      "grad_norm": 3.894425868988037,
-      "learning_rate": 3.1931054660125802e-06,
-      "loss": 0.9266,
-      "step": 16490
-    },
-    {
-      "epoch": 0.7647195178488642,
-      "grad_norm": 3.513700485229492,
-      "learning_rate": 3.1871801126725133e-06,
-      "loss": 0.8347,
-      "step": 16495
-    },
-    {
-      "epoch": 0.7649513212795549,
-      "grad_norm": 3.5783920288085938,
-      "learning_rate": 3.1812592198097016e-06,
-      "loss": 0.7945,
-      "step": 16500
-    },
-    {
-      "epoch": 0.7649513212795549,
-      "eval_loss": 0.8850834965705872,
-      "eval_runtime": 11.2607,
-      "eval_samples_per_second": 11.278,
-      "eval_steps_per_second": 11.278,
-      "step": 16500
-    },
-    {
-      "epoch": 0.7651831247102457,
-      "grad_norm": 4.362255573272705,
-      "learning_rate": 3.1753427913006563e-06,
-      "loss": 0.8344,
-      "step": 16505
-    },
-    {
-      "epoch": 0.7654149281409365,
-      "grad_norm": 3.7136001586914062,
-      "learning_rate": 3.169430831018969e-06,
-      "loss": 0.722,
-      "step": 16510
-    },
-    {
-      "epoch": 0.7656467315716272,
-      "grad_norm": 3.7046449184417725,
-      "learning_rate": 3.163523342835315e-06,
-      "loss": 0.7368,
-      "step": 16515
-    },
-    {
-      "epoch": 0.7658785350023181,
-      "grad_norm": 3.598536729812622,
-      "learning_rate": 3.1576203306174268e-06,
-      "loss": 0.9693,
-      "step": 16520
-    },
-    {
-      "epoch": 0.7661103384330088,
-      "grad_norm": 3.1772992610931396,
-      "learning_rate": 3.1517217982301184e-06,
-      "loss": 0.7942,
-      "step": 16525
-    },
-    {
-      "epoch": 0.7663421418636995,
-      "grad_norm": 4.3848395347595215,
-      "learning_rate": 3.1458277495352687e-06,
-      "loss": 0.9413,
-      "step": 16530
-    },
-    {
-      "epoch": 0.7665739452943904,
-      "grad_norm": 3.851550340652466,
-      "learning_rate": 3.1399381883918144e-06,
-      "loss": 1.0156,
-      "step": 16535
-    },
-    {
-      "epoch": 0.7668057487250811,
-      "grad_norm": 3.9418606758117676,
-      "learning_rate": 3.134053118655758e-06,
-      "loss": 0.9909,
-      "step": 16540
-    },
-    {
-      "epoch": 0.7670375521557719,
-      "grad_norm": 3.4677114486694336,
-      "learning_rate": 3.1281725441801625e-06,
-      "loss": 0.7909,
-      "step": 16545
-    },
-    {
-      "epoch": 0.7672693555864627,
-      "grad_norm": 4.036655902862549,
-      "learning_rate": 3.1222964688151515e-06,
-      "loss": 0.8848,
-      "step": 16550
-    },
-    {
-      "epoch": 0.7675011590171534,
-      "grad_norm": 4.098992824554443,
-      "learning_rate": 3.1164248964078913e-06,
-      "loss": 0.8305,
-      "step": 16555
-    },
-    {
-      "epoch": 0.7677329624478443,
-      "grad_norm": 3.2788567543029785,
-      "learning_rate": 3.110557830802612e-06,
-      "loss": 0.9747,
-      "step": 16560
-    },
-    {
-      "epoch": 0.767964765878535,
-      "grad_norm": 3.436088800430298,
-      "learning_rate": 3.1046952758405913e-06,
-      "loss": 0.73,
-      "step": 16565
-    },
-    {
-      "epoch": 0.7681965693092258,
-      "grad_norm": 3.365654468536377,
-      "learning_rate": 3.098837235360146e-06,
-      "loss": 0.7866,
-      "step": 16570
-    },
-    {
-      "epoch": 0.7684283727399166,
-      "grad_norm": 3.64009428024292,
-      "learning_rate": 3.092983713196639e-06,
-      "loss": 0.9828,
-      "step": 16575
-    },
-    {
-      "epoch": 0.7686601761706073,
-      "grad_norm": 4.071859359741211,
-      "learning_rate": 3.0871347131824823e-06,
-      "loss": 0.8831,
-      "step": 16580
-    },
-    {
-      "epoch": 0.7688919796012981,
-      "grad_norm": 3.6441633701324463,
-      "learning_rate": 3.0812902391471244e-06,
-      "loss": 0.8001,
-      "step": 16585
-    },
-    {
-      "epoch": 0.7691237830319889,
-      "grad_norm": 3.392503023147583,
-      "learning_rate": 3.075450294917044e-06,
-      "loss": 0.772,
-      "step": 16590
-    },
-    {
-      "epoch": 0.7693555864626797,
-      "grad_norm": 3.7266502380371094,
-      "learning_rate": 3.0696148843157646e-06,
-      "loss": 0.8001,
-      "step": 16595
-    },
-    {
-      "epoch": 0.7695873898933704,
-      "grad_norm": 3.742562770843506,
-      "learning_rate": 3.063784011163833e-06,
-      "loss": 0.9027,
-      "step": 16600
-    },
-    {
-      "epoch": 0.7695873898933704,
-      "eval_loss": 0.8843169808387756,
-      "eval_runtime": 11.2638,
-      "eval_samples_per_second": 11.275,
-      "eval_steps_per_second": 11.275,
-      "step": 16600
-    },
-    {
-      "epoch": 0.7698191933240612,
-      "grad_norm": 3.8710248470306396,
-      "learning_rate": 3.0579576792788234e-06,
-      "loss": 0.8078,
-      "step": 16605
-    },
-    {
-      "epoch": 0.770050996754752,
-      "grad_norm": 2.7604763507843018,
-      "learning_rate": 3.0521358924753466e-06,
-      "loss": 0.7083,
-      "step": 16610
-    },
-    {
-      "epoch": 0.7702828001854427,
-      "grad_norm": 3.607058525085449,
-      "learning_rate": 3.0463186545650346e-06,
-      "loss": 0.8634,
-      "step": 16615
-    },
-    {
-      "epoch": 0.7705146036161336,
-      "grad_norm": 2.935387372970581,
-      "learning_rate": 3.040505969356532e-06,
-      "loss": 0.7493,
-      "step": 16620
-    },
-    {
-      "epoch": 0.7707464070468243,
-      "grad_norm": 3.4781177043914795,
-      "learning_rate": 3.0346978406555172e-06,
-      "loss": 0.8851,
-      "step": 16625
-    },
-    {
-      "epoch": 0.770978210477515,
-      "grad_norm": 3.8345789909362793,
-      "learning_rate": 3.028894272264671e-06,
-      "loss": 0.7353,
-      "step": 16630
-    },
-    {
-      "epoch": 0.7712100139082059,
-      "grad_norm": 3.4135794639587402,
-      "learning_rate": 3.0230952679837023e-06,
-      "loss": 0.7298,
-      "step": 16635
-    },
-    {
-      "epoch": 0.7714418173388966,
-      "grad_norm": 3.7546191215515137,
-      "learning_rate": 3.0173008316093166e-06,
-      "loss": 0.9568,
-      "step": 16640
-    },
-    {
-      "epoch": 0.7716736207695873,
-      "grad_norm": 4.952125072479248,
-      "learning_rate": 3.0115109669352426e-06,
-      "loss": 0.9374,
-      "step": 16645
-    },
-    {
-      "epoch": 0.7719054242002782,
-      "grad_norm": 4.161298751831055,
-      "learning_rate": 3.0057256777522116e-06,
-      "loss": 0.8104,
-      "step": 16650
-    },
-    {
-      "epoch": 0.7721372276309689,
-      "grad_norm": 3.4128384590148926,
-      "learning_rate": 2.999944967847954e-06,
-      "loss": 0.7943,
-      "step": 16655
-    },
-    {
-      "epoch": 0.7723690310616597,
-      "grad_norm": 3.6398515701293945,
-      "learning_rate": 2.9941688410072057e-06,
-      "loss": 0.7745,
-      "step": 16660
-    },
-    {
-      "epoch": 0.7726008344923505,
-      "grad_norm": 4.075809478759766,
-      "learning_rate": 2.9883973010117017e-06,
-      "loss": 0.7606,
-      "step": 16665
-    },
-    {
-      "epoch": 0.7728326379230412,
-      "grad_norm": 3.837707281112671,
-      "learning_rate": 2.98263035164018e-06,
-      "loss": 0.7089,
-      "step": 16670
-    },
-    {
-      "epoch": 0.7730644413537321,
-      "grad_norm": 3.544987440109253,
-      "learning_rate": 2.97686799666836e-06,
-      "loss": 0.8661,
-      "step": 16675
-    },
-    {
-      "epoch": 0.7732962447844228,
-      "grad_norm": 4.513082027435303,
-      "learning_rate": 2.971110239868964e-06,
-      "loss": 0.8233,
-      "step": 16680
-    },
-    {
-      "epoch": 0.7735280482151136,
-      "grad_norm": 3.475184917449951,
-      "learning_rate": 2.9653570850117065e-06,
-      "loss": 0.9276,
-      "step": 16685
-    },
-    {
-      "epoch": 0.7737598516458044,
-      "grad_norm": 3.9753215312957764,
-      "learning_rate": 2.95960853586327e-06,
-      "loss": 0.795,
-      "step": 16690
-    },
-    {
-      "epoch": 0.7739916550764951,
-      "grad_norm": 3.1066231727600098,
-      "learning_rate": 2.95386459618734e-06,
-      "loss": 0.85,
-      "step": 16695
-    },
-    {
-      "epoch": 0.7742234585071859,
-      "grad_norm": 3.46185564994812,
-      "learning_rate": 2.948125269744583e-06,
-      "loss": 0.7633,
-      "step": 16700
-    },
-    {
-      "epoch": 0.7742234585071859,
-      "eval_loss": 0.8845146298408508,
-      "eval_runtime": 11.2686,
-      "eval_samples_per_second": 11.27,
-      "eval_steps_per_second": 11.27,
-      "step": 16700
-    },
-    {
-      "epoch": 0.7744552619378767,
-      "grad_norm": 3.0039331912994385,
-      "learning_rate": 2.942390560292633e-06,
-      "loss": 0.8481,
-      "step": 16705
-    },
-    {
-      "epoch": 0.7746870653685675,
-      "grad_norm": 3.653935432434082,
-      "learning_rate": 2.936660471586116e-06,
-      "loss": 0.9245,
-      "step": 16710
-    },
-    {
-      "epoch": 0.7749188687992582,
-      "grad_norm": 3.9792356491088867,
-      "learning_rate": 2.9309350073766173e-06,
-      "loss": 0.8936,
-      "step": 16715
-    },
-    {
-      "epoch": 0.775150672229949,
-      "grad_norm": 3.7885968685150146,
-      "learning_rate": 2.9252141714127115e-06,
-      "loss": 0.8169,
-      "step": 16720
-    },
-    {
-      "epoch": 0.7753824756606398,
-      "grad_norm": 4.647716045379639,
-      "learning_rate": 2.919497967439926e-06,
-      "loss": 0.9129,
-      "step": 16725
-    },
-    {
-      "epoch": 0.7756142790913305,
-      "grad_norm": 4.3390398025512695,
-      "learning_rate": 2.9137863992007666e-06,
-      "loss": 0.8678,
-      "step": 16730
-    },
-    {
-      "epoch": 0.7758460825220214,
-      "grad_norm": 3.9097447395324707,
-      "learning_rate": 2.908079470434706e-06,
-      "loss": 0.9015,
-      "step": 16735
-    },
-    {
-      "epoch": 0.7760778859527121,
-      "grad_norm": 3.679063320159912,
-      "learning_rate": 2.902377184878169e-06,
-      "loss": 0.8466,
-      "step": 16740
-    },
-    {
-      "epoch": 0.7763096893834028,
-      "grad_norm": 4.423166275024414,
-      "learning_rate": 2.8966795462645448e-06,
-      "loss": 0.8707,
-      "step": 16745
-    },
-    {
-      "epoch": 0.7765414928140937,
-      "grad_norm": 3.802765130996704,
-      "learning_rate": 2.8909865583241825e-06,
-      "loss": 0.8169,
-      "step": 16750
-    },
-    {
-      "epoch": 0.7767732962447844,
-      "grad_norm": 4.170059680938721,
-      "learning_rate": 2.8852982247843886e-06,
-      "loss": 0.9063,
-      "step": 16755
-    },
-    {
-      "epoch": 0.7770050996754752,
-      "grad_norm": 4.481255054473877,
-      "learning_rate": 2.8796145493694127e-06,
-      "loss": 0.9702,
-      "step": 16760
-    },
-    {
-      "epoch": 0.777236903106166,
-      "grad_norm": 3.3113086223602295,
-      "learning_rate": 2.873935535800465e-06,
-      "loss": 0.861,
-      "step": 16765
-    },
-    {
-      "epoch": 0.7774687065368567,
-      "grad_norm": 4.658487796783447,
-      "learning_rate": 2.8682611877957003e-06,
-      "loss": 0.8002,
-      "step": 16770
-    },
-    {
-      "epoch": 0.7777005099675475,
-      "grad_norm": 3.762125015258789,
-      "learning_rate": 2.862591509070214e-06,
-      "loss": 0.8989,
-      "step": 16775
-    },
-    {
-      "epoch": 0.7779323133982383,
-      "grad_norm": 3.8254103660583496,
-      "learning_rate": 2.856926503336047e-06,
-      "loss": 0.8452,
-      "step": 16780
-    },
-    {
-      "epoch": 0.778164116828929,
-      "grad_norm": 3.9982810020446777,
-      "learning_rate": 2.851266174302183e-06,
-      "loss": 0.9509,
-      "step": 16785
-    },
-    {
-      "epoch": 0.7783959202596198,
-      "grad_norm": 3.761439561843872,
-      "learning_rate": 2.845610525674547e-06,
-      "loss": 0.7328,
-      "step": 16790
-    },
-    {
-      "epoch": 0.7786277236903106,
-      "grad_norm": 3.8584911823272705,
-      "learning_rate": 2.8399595611559892e-06,
-      "loss": 1.0582,
-      "step": 16795
-    },
-    {
-      "epoch": 0.7788595271210014,
-      "grad_norm": 4.529983997344971,
-      "learning_rate": 2.834313284446303e-06,
-      "loss": 0.7333,
-      "step": 16800
-    },
-    {
-      "epoch": 0.7788595271210014,
-      "eval_loss": 0.883889377117157,
-      "eval_runtime": 11.2662,
-      "eval_samples_per_second": 11.273,
-      "eval_steps_per_second": 11.273,
-      "step": 16800
-    },
-    {
-      "epoch": 0.7790913305516922,
-      "grad_norm": 3.4581003189086914,
-      "learning_rate": 2.8286716992422083e-06,
-      "loss": 0.7859,
-      "step": 16805
-    },
-    {
-      "epoch": 0.779323133982383,
-      "grad_norm": 3.182137966156006,
-      "learning_rate": 2.8230348092373474e-06,
-      "loss": 0.8461,
-      "step": 16810
-    },
-    {
-      "epoch": 0.7795549374130737,
-      "grad_norm": 4.005838394165039,
-      "learning_rate": 2.8174026181223e-06,
-      "loss": 0.8896,
-      "step": 16815
-    },
-    {
-      "epoch": 0.7797867408437645,
-      "grad_norm": 3.5771877765655518,
-      "learning_rate": 2.8117751295845673e-06,
-      "loss": 0.8756,
-      "step": 16820
-    },
-    {
-      "epoch": 0.7800185442744553,
-      "grad_norm": 3.877868890762329,
-      "learning_rate": 2.8061523473085626e-06,
-      "loss": 0.7942,
-      "step": 16825
-    },
-    {
-      "epoch": 0.780250347705146,
-      "grad_norm": 3.636414051055908,
-      "learning_rate": 2.8005342749756292e-06,
-      "loss": 0.7895,
-      "step": 16830
-    },
-    {
-      "epoch": 0.7804821511358369,
-      "grad_norm": 3.0762102603912354,
-      "learning_rate": 2.7949209162640144e-06,
-      "loss": 0.7918,
-      "step": 16835
-    },
-    {
-      "epoch": 0.7807139545665276,
-      "grad_norm": 4.274107933044434,
-      "learning_rate": 2.7893122748488943e-06,
-      "loss": 0.9119,
-      "step": 16840
-    },
-    {
-      "epoch": 0.7809457579972183,
-      "grad_norm": 4.0760602951049805,
-      "learning_rate": 2.7837083544023425e-06,
-      "loss": 0.9507,
-      "step": 16845
-    },
-    {
-      "epoch": 0.7811775614279092,
-      "grad_norm": 3.145632028579712,
-      "learning_rate": 2.7781091585933484e-06,
-      "loss": 0.7351,
-      "step": 16850
-    },
-    {
-      "epoch": 0.7814093648585999,
-      "grad_norm": 4.766406536102295,
-      "learning_rate": 2.772514691087813e-06,
-      "loss": 0.9983,
-      "step": 16855
-    },
-    {
-      "epoch": 0.7816411682892906,
-      "grad_norm": 2.734017848968506,
-      "learning_rate": 2.7669249555485323e-06,
-      "loss": 0.7706,
-      "step": 16860
-    },
-    {
-      "epoch": 0.7818729717199815,
-      "grad_norm": 3.6416406631469727,
-      "learning_rate": 2.761339955635205e-06,
-      "loss": 0.8816,
-      "step": 16865
-    },
-    {
-      "epoch": 0.7821047751506722,
-      "grad_norm": 3.407658576965332,
-      "learning_rate": 2.755759695004434e-06,
-      "loss": 0.8622,
-      "step": 16870
-    },
-    {
-      "epoch": 0.782336578581363,
-      "grad_norm": 3.256417989730835,
-      "learning_rate": 2.7501841773097227e-06,
-      "loss": 0.684,
-      "step": 16875
-    },
-    {
-      "epoch": 0.7825683820120538,
-      "grad_norm": 4.0953826904296875,
-      "learning_rate": 2.7446134062014563e-06,
-      "loss": 0.8825,
-      "step": 16880
-    },
-    {
-      "epoch": 0.7828001854427445,
-      "grad_norm": 3.547276258468628,
-      "learning_rate": 2.739047385326924e-06,
-      "loss": 0.6165,
-      "step": 16885
-    },
-    {
-      "epoch": 0.7830319888734353,
-      "grad_norm": 3.827667474746704,
-      "learning_rate": 2.7334861183303064e-06,
-      "loss": 0.8814,
-      "step": 16890
-    },
-    {
-      "epoch": 0.7832637923041261,
-      "grad_norm": 4.139094352722168,
-      "learning_rate": 2.727929608852655e-06,
-      "loss": 0.9706,
-      "step": 16895
-    },
-    {
-      "epoch": 0.7834955957348169,
-      "grad_norm": 4.088076591491699,
-      "learning_rate": 2.722377860531924e-06,
-      "loss": 1.0389,
-      "step": 16900
-    },
-    {
-      "epoch": 0.7834955957348169,
-      "eval_loss": 0.8829403519630432,
-      "eval_runtime": 11.2748,
-      "eval_samples_per_second": 11.264,
-      "eval_steps_per_second": 11.264,
-      "step": 16900
-    },
-    {
-      "epoch": 0.7837273991655076,
-      "grad_norm": 3.537174940109253,
-      "learning_rate": 2.716830877002946e-06,
-      "loss": 0.9587,
-      "step": 16905
-    },
-    {
-      "epoch": 0.7839592025961984,
-      "grad_norm": 4.048293113708496,
-      "learning_rate": 2.7112886618974264e-06,
-      "loss": 0.9287,
-      "step": 16910
-    },
-    {
-      "epoch": 0.7841910060268892,
-      "grad_norm": 4.15169095993042,
-      "learning_rate": 2.7057512188439615e-06,
-      "loss": 0.8293,
-      "step": 16915
-    },
-    {
-      "epoch": 0.7844228094575799,
-      "grad_norm": 3.689326763153076,
-      "learning_rate": 2.700218551468009e-06,
-      "loss": 0.7086,
-      "step": 16920
-    },
-    {
-      "epoch": 0.7846546128882708,
-      "grad_norm": 3.1314663887023926,
-      "learning_rate": 2.6946906633919136e-06,
-      "loss": 0.9349,
-      "step": 16925
-    },
-    {
-      "epoch": 0.7848864163189615,
-      "grad_norm": 4.037022113800049,
-      "learning_rate": 2.6891675582348774e-06,
-      "loss": 0.9397,
-      "step": 16930
-    },
-    {
-      "epoch": 0.7851182197496523,
-      "grad_norm": 3.389939785003662,
-      "learning_rate": 2.683649239612982e-06,
-      "loss": 0.8126,
-      "step": 16935
-    },
-    {
-      "epoch": 0.7853500231803431,
-      "grad_norm": 4.172562599182129,
-      "learning_rate": 2.6781357111391725e-06,
-      "loss": 0.9107,
-      "step": 16940
-    },
-    {
-      "epoch": 0.7855818266110338,
-      "grad_norm": 3.2886476516723633,
-      "learning_rate": 2.672626976423256e-06,
-      "loss": 0.6793,
-      "step": 16945
-    },
-    {
-      "epoch": 0.7858136300417247,
-      "grad_norm": 3.162684679031372,
-      "learning_rate": 2.6671230390718962e-06,
-      "loss": 0.8093,
-      "step": 16950
-    },
-    {
-      "epoch": 0.7860454334724154,
-      "grad_norm": 4.041141033172607,
-      "learning_rate": 2.661623902688625e-06,
-      "loss": 1.0565,
-      "step": 16955
-    },
-    {
-      "epoch": 0.7862772369031061,
-      "grad_norm": 3.6500637531280518,
-      "learning_rate": 2.65612957087383e-06,
-      "loss": 0.7986,
-      "step": 16960
-    },
-    {
-      "epoch": 0.786509040333797,
-      "grad_norm": 3.767634868621826,
-      "learning_rate": 2.650640047224744e-06,
-      "loss": 0.8563,
-      "step": 16965
-    },
-    {
-      "epoch": 0.7867408437644877,
-      "grad_norm": 3.3875231742858887,
-      "learning_rate": 2.645155335335461e-06,
-      "loss": 0.7401,
-      "step": 16970
-    },
-    {
-      "epoch": 0.7869726471951785,
-      "grad_norm": 3.8744959831237793,
-      "learning_rate": 2.639675438796926e-06,
-      "loss": 0.7653,
-      "step": 16975
-    },
-    {
-      "epoch": 0.7872044506258693,
-      "grad_norm": 4.187319278717041,
-      "learning_rate": 2.6342003611969225e-06,
-      "loss": 0.6953,
-      "step": 16980
-    },
-    {
-      "epoch": 0.78743625405656,
-      "grad_norm": 3.9528913497924805,
-      "learning_rate": 2.6287301061200822e-06,
-      "loss": 0.7537,
-      "step": 16985
-    },
-    {
-      "epoch": 0.7876680574872508,
-      "grad_norm": 3.991596221923828,
-      "learning_rate": 2.623264677147882e-06,
-      "loss": 0.9745,
-      "step": 16990
-    },
-    {
-      "epoch": 0.7878998609179416,
-      "grad_norm": 3.6371634006500244,
-      "learning_rate": 2.6178040778586424e-06,
-      "loss": 0.921,
-      "step": 16995
-    },
-    {
-      "epoch": 0.7881316643486324,
-      "grad_norm": 4.425895690917969,
-      "learning_rate": 2.6123483118275097e-06,
-      "loss": 0.8142,
-      "step": 17000
-    },
-    {
-      "epoch": 0.7881316643486324,
-      "eval_loss": 0.8824594020843506,
-      "eval_runtime": 11.268,
-      "eval_samples_per_second": 11.271,
-      "eval_steps_per_second": 11.271,
-      "step": 17000
-    },
-    {
-      "epoch": 0.7883634677793231,
-      "grad_norm": 3.481616735458374,
-      "learning_rate": 2.60689738262648e-06,
-      "loss": 0.7512,
-      "step": 17005
-    },
-    {
-      "epoch": 0.7885952712100139,
-      "grad_norm": 3.797632932662964,
-      "learning_rate": 2.601451293824372e-06,
-      "loss": 0.7542,
-      "step": 17010
-    },
-    {
-      "epoch": 0.7888270746407047,
-      "grad_norm": 4.2333173751831055,
-      "learning_rate": 2.5960100489868377e-06,
-      "loss": 0.9087,
-      "step": 17015
-    },
-    {
-      "epoch": 0.7890588780713954,
-      "grad_norm": 4.05795955657959,
-      "learning_rate": 2.5905736516763626e-06,
-      "loss": 0.8283,
-      "step": 17020
-    },
-    {
-      "epoch": 0.7892906815020863,
-      "grad_norm": 3.3479185104370117,
-      "learning_rate": 2.5851421054522552e-06,
-      "loss": 0.697,
-      "step": 17025
-    },
-    {
-      "epoch": 0.789522484932777,
-      "grad_norm": 4.0138139724731445,
-      "learning_rate": 2.579715413870645e-06,
-      "loss": 0.7351,
-      "step": 17030
-    },
-    {
-      "epoch": 0.7897542883634677,
-      "grad_norm": 3.8981614112854004,
-      "learning_rate": 2.5742935804844894e-06,
-      "loss": 0.9307,
-      "step": 17035
-    },
-    {
-      "epoch": 0.7899860917941586,
-      "grad_norm": 4.282942295074463,
-      "learning_rate": 2.568876608843556e-06,
-      "loss": 0.941,
-      "step": 17040
-    },
-    {
-      "epoch": 0.7902178952248493,
-      "grad_norm": 4.372350215911865,
-      "learning_rate": 2.5634645024944414e-06,
-      "loss": 1.047,
-      "step": 17045
-    },
-    {
-      "epoch": 0.7904496986555402,
-      "grad_norm": 4.30950927734375,
-      "learning_rate": 2.5580572649805445e-06,
-      "loss": 0.9546,
-      "step": 17050
-    },
-    {
-      "epoch": 0.7906815020862309,
-      "grad_norm": 3.705162286758423,
-      "learning_rate": 2.5526548998420843e-06,
-      "loss": 0.8446,
-      "step": 17055
-    },
-    {
-      "epoch": 0.7909133055169216,
-      "grad_norm": 3.317826271057129,
-      "learning_rate": 2.5472574106160907e-06,
-      "loss": 0.9309,
-      "step": 17060
-    },
-    {
-      "epoch": 0.7911451089476125,
-      "grad_norm": 4.324466705322266,
-      "learning_rate": 2.5418648008363955e-06,
-      "loss": 0.835,
-      "step": 17065
-    },
-    {
-      "epoch": 0.7913769123783032,
-      "grad_norm": 3.8251116275787354,
-      "learning_rate": 2.5364770740336364e-06,
-      "loss": 0.8074,
-      "step": 17070
-    },
-    {
-      "epoch": 0.791608715808994,
-      "grad_norm": 3.2614023685455322,
-      "learning_rate": 2.531094233735257e-06,
-      "loss": 0.8442,
-      "step": 17075
-    },
-    {
-      "epoch": 0.7918405192396848,
-      "grad_norm": 3.841665267944336,
-      "learning_rate": 2.5257162834655057e-06,
-      "loss": 0.7189,
-      "step": 17080
-    },
-    {
-      "epoch": 0.7920723226703755,
-      "grad_norm": 3.531162977218628,
-      "learning_rate": 2.5203432267454188e-06,
-      "loss": 0.9309,
-      "step": 17085
-    },
-    {
-      "epoch": 0.7923041261010663,
-      "grad_norm": 4.039501667022705,
-      "learning_rate": 2.514975067092835e-06,
-      "loss": 0.8324,
-      "step": 17090
-    },
-    {
-      "epoch": 0.7925359295317571,
-      "grad_norm": 3.6314001083374023,
-      "learning_rate": 2.509611808022394e-06,
-      "loss": 0.8618,
-      "step": 17095
-    },
-    {
-      "epoch": 0.7927677329624478,
-      "grad_norm": 4.1385178565979,
-      "learning_rate": 2.5042534530455076e-06,
-      "loss": 0.8456,
-      "step": 17100
-    },
-    {
-      "epoch": 0.7927677329624478,
-      "eval_loss": 0.8818289637565613,
-      "eval_runtime": 11.2697,
-      "eval_samples_per_second": 11.269,
-      "eval_steps_per_second": 11.269,
-      "step": 17100
-    },
-    {
-      "epoch": 0.7929995363931386,
-      "grad_norm": 3.2137322425842285,
-      "learning_rate": 2.498900005670394e-06,
-      "loss": 0.7141,
-      "step": 17105
-    },
-    {
-      "epoch": 0.7932313398238294,
-      "grad_norm": 4.110595226287842,
-      "learning_rate": 2.4935514694020545e-06,
-      "loss": 0.867,
-      "step": 17110
-    },
-    {
-      "epoch": 0.7934631432545202,
-      "grad_norm": 3.7244606018066406,
-      "learning_rate": 2.488207847742269e-06,
-      "loss": 0.9114,
-      "step": 17115
-    },
-    {
-      "epoch": 0.7936949466852109,
-      "grad_norm": 2.8758745193481445,
-      "learning_rate": 2.4828691441896056e-06,
-      "loss": 0.7327,
-      "step": 17120
-    },
-    {
-      "epoch": 0.7939267501159017,
-      "grad_norm": 3.3576953411102295,
-      "learning_rate": 2.477535362239414e-06,
-      "loss": 0.9152,
-      "step": 17125
-    },
-    {
-      "epoch": 0.7941585535465925,
-      "grad_norm": 3.668461561203003,
-      "learning_rate": 2.472206505383815e-06,
-      "loss": 0.9391,
-      "step": 17130
-    },
-    {
-      "epoch": 0.7943903569772832,
-      "grad_norm": 3.574362277984619,
-      "learning_rate": 2.4668825771117065e-06,
-      "loss": 0.8111,
-      "step": 17135
-    },
-    {
-      "epoch": 0.7946221604079741,
-      "grad_norm": 3.9505579471588135,
-      "learning_rate": 2.4615635809087646e-06,
-      "loss": 0.9073,
-      "step": 17140
-    },
-    {
-      "epoch": 0.7948539638386648,
-      "grad_norm": 3.6357901096343994,
-      "learning_rate": 2.456249520257433e-06,
-      "loss": 0.8859,
-      "step": 17145
-    },
-    {
-      "epoch": 0.7950857672693555,
-      "grad_norm": 4.1312127113342285,
-      "learning_rate": 2.4509403986369242e-06,
-      "loss": 1.0306,
-      "step": 17150
-    },
-    {
-      "epoch": 0.7953175707000464,
-      "grad_norm": 3.519533634185791,
-      "learning_rate": 2.445636219523212e-06,
-      "loss": 0.8652,
-      "step": 17155
-    },
-    {
-      "epoch": 0.7955493741307371,
-      "grad_norm": 3.3949859142303467,
-      "learning_rate": 2.4403369863890425e-06,
-      "loss": 0.7245,
-      "step": 17160
-    },
-    {
-      "epoch": 0.7957811775614279,
-      "grad_norm": 3.7756998538970947,
-      "learning_rate": 2.4350427027039234e-06,
-      "loss": 0.8728,
-      "step": 17165
-    },
-    {
-      "epoch": 0.7960129809921187,
-      "grad_norm": 3.843841075897217,
-      "learning_rate": 2.429753371934114e-06,
-      "loss": 0.856,
-      "step": 17170
-    },
-    {
-      "epoch": 0.7962447844228094,
-      "grad_norm": 3.760075330734253,
-      "learning_rate": 2.424468997542635e-06,
-      "loss": 1.0026,
-      "step": 17175
-    },
-    {
-      "epoch": 0.7964765878535003,
-      "grad_norm": 3.853175163269043,
-      "learning_rate": 2.4191895829892707e-06,
-      "loss": 0.7359,
-      "step": 17180
-    },
-    {
-      "epoch": 0.796708391284191,
-      "grad_norm": 3.496713638305664,
-      "learning_rate": 2.4139151317305386e-06,
-      "loss": 0.7392,
-      "step": 17185
-    },
-    {
-      "epoch": 0.7969401947148818,
-      "grad_norm": 3.5978150367736816,
-      "learning_rate": 2.4086456472197227e-06,
-      "loss": 0.9091,
-      "step": 17190
-    },
-    {
-      "epoch": 0.7971719981455726,
-      "grad_norm": 3.749814033508301,
-      "learning_rate": 2.403381132906851e-06,
-      "loss": 0.794,
-      "step": 17195
-    },
-    {
-      "epoch": 0.7974038015762633,
-      "grad_norm": 4.097865581512451,
-      "learning_rate": 2.3981215922386993e-06,
-      "loss": 0.9018,
-      "step": 17200
-    },
-    {
-      "epoch": 0.7974038015762633,
-      "eval_loss": 0.8813273906707764,
-      "eval_runtime": 11.2722,
-      "eval_samples_per_second": 11.267,
-      "eval_steps_per_second": 11.267,
-      "step": 17200
-    },
-    {
-      "epoch": 0.7976356050069541,
-      "grad_norm": 4.155472755432129,
-      "learning_rate": 2.3928670286587797e-06,
-      "loss": 0.8676,
-      "step": 17205
-    },
-    {
-      "epoch": 0.7978674084376449,
-      "grad_norm": 3.5118024349212646,
-      "learning_rate": 2.3876174456073564e-06,
-      "loss": 0.8646,
-      "step": 17210
-    },
-    {
-      "epoch": 0.7980992118683357,
-      "grad_norm": 3.5498898029327393,
-      "learning_rate": 2.3823728465214235e-06,
-      "loss": 0.85,
-      "step": 17215
-    },
-    {
-      "epoch": 0.7983310152990264,
-      "grad_norm": 4.195731163024902,
-      "learning_rate": 2.377133234834713e-06,
-      "loss": 0.8185,
-      "step": 17220
-    },
-    {
-      "epoch": 0.7985628187297172,
-      "grad_norm": 4.20908260345459,
-      "learning_rate": 2.3718986139776967e-06,
-      "loss": 0.8682,
-      "step": 17225
-    },
-    {
-      "epoch": 0.798794622160408,
-      "grad_norm": 4.149493217468262,
-      "learning_rate": 2.3666689873775793e-06,
-      "loss": 1.0926,
-      "step": 17230
-    },
-    {
-      "epoch": 0.7990264255910987,
-      "grad_norm": 4.024598598480225,
-      "learning_rate": 2.361444358458288e-06,
-      "loss": 0.9183,
-      "step": 17235
-    },
-    {
-      "epoch": 0.7992582290217896,
-      "grad_norm": 3.8815884590148926,
-      "learning_rate": 2.356224730640486e-06,
-      "loss": 0.8183,
-      "step": 17240
-    },
-    {
-      "epoch": 0.7994900324524803,
-      "grad_norm": 4.009396553039551,
-      "learning_rate": 2.3510101073415537e-06,
-      "loss": 0.7657,
-      "step": 17245
-    },
-    {
-      "epoch": 0.799721835883171,
-      "grad_norm": 3.6904358863830566,
-      "learning_rate": 2.345800491975606e-06,
-      "loss": 0.9904,
-      "step": 17250
-    },
-    {
-      "epoch": 0.7999536393138619,
-      "grad_norm": 4.349132537841797,
-      "learning_rate": 2.3405958879534687e-06,
-      "loss": 0.7577,
-      "step": 17255
-    },
-    {
-      "epoch": 0.8001854427445526,
-      "grad_norm": 3.6113533973693848,
-      "learning_rate": 2.3353962986826904e-06,
-      "loss": 0.8239,
-      "step": 17260
-    },
-    {
-      "epoch": 0.8004172461752433,
-      "grad_norm": 3.7421815395355225,
-      "learning_rate": 2.3302017275675427e-06,
-      "loss": 0.8305,
-      "step": 17265
-    },
-    {
-      "epoch": 0.8006490496059342,
-      "grad_norm": 3.947587013244629,
-      "learning_rate": 2.325012178009002e-06,
-      "loss": 0.9255,
-      "step": 17270
-    },
-    {
-      "epoch": 0.8008808530366249,
-      "grad_norm": 4.150144577026367,
-      "learning_rate": 2.319827653404757e-06,
-      "loss": 0.8958,
-      "step": 17275
-    },
-    {
-      "epoch": 0.8011126564673157,
-      "grad_norm": 3.0124239921569824,
-      "learning_rate": 2.314648157149214e-06,
-      "loss": 0.6897,
-      "step": 17280
-    },
-    {
-      "epoch": 0.8013444598980065,
-      "grad_norm": 3.2395570278167725,
-      "learning_rate": 2.3094736926334883e-06,
-      "loss": 0.8505,
-      "step": 17285
-    },
-    {
-      "epoch": 0.8015762633286972,
-      "grad_norm": 3.67964768409729,
-      "learning_rate": 2.304304263245387e-06,
-      "loss": 0.7883,
-      "step": 17290
-    },
-    {
-      "epoch": 0.8018080667593881,
-      "grad_norm": 3.6921372413635254,
-      "learning_rate": 2.2991398723694346e-06,
-      "loss": 0.7569,
-      "step": 17295
-    },
-    {
-      "epoch": 0.8020398701900788,
-      "grad_norm": 3.7016260623931885,
-      "learning_rate": 2.2939805233868575e-06,
-      "loss": 0.7654,
-      "step": 17300
-    },
-    {
-      "epoch": 0.8020398701900788,
-      "eval_loss": 0.8810060620307922,
-      "eval_runtime": 11.2796,
-      "eval_samples_per_second": 11.259,
-      "eval_steps_per_second": 11.259,
-      "step": 17300
-    },
-    {
-      "epoch": 0.8022716736207696,
-      "grad_norm": 3.850266933441162,
-      "learning_rate": 2.288826219675563e-06,
-      "loss": 0.8341,
-      "step": 17305
-    },
-    {
-      "epoch": 0.8025034770514604,
-      "grad_norm": 3.525607109069824,
-      "learning_rate": 2.2836769646101743e-06,
-      "loss": 0.8223,
-      "step": 17310
-    },
-    {
-      "epoch": 0.8027352804821511,
-      "grad_norm": 4.057685375213623,
-      "learning_rate": 2.2785327615620055e-06,
-      "loss": 0.8506,
-      "step": 17315
-    },
-    {
-      "epoch": 0.8029670839128419,
-      "grad_norm": 4.184149742126465,
-      "learning_rate": 2.273393613899053e-06,
-      "loss": 1.0103,
-      "step": 17320
-    },
-    {
-      "epoch": 0.8031988873435327,
-      "grad_norm": 3.58420729637146,
-      "learning_rate": 2.2682595249860138e-06,
-      "loss": 0.828,
-      "step": 17325
-    },
-    {
-      "epoch": 0.8034306907742235,
-      "grad_norm": 3.695439577102661,
-      "learning_rate": 2.263130498184274e-06,
-      "loss": 0.8649,
-      "step": 17330
-    },
-    {
-      "epoch": 0.8036624942049142,
-      "grad_norm": 3.5146336555480957,
-      "learning_rate": 2.2580065368518943e-06,
-      "loss": 0.834,
-      "step": 17335
-    },
-    {
-      "epoch": 0.803894297635605,
-      "grad_norm": 3.4850876331329346,
-      "learning_rate": 2.2528876443436266e-06,
-      "loss": 0.6973,
-      "step": 17340
-    },
-    {
-      "epoch": 0.8041261010662958,
-      "grad_norm": 3.5119123458862305,
-      "learning_rate": 2.247773824010903e-06,
-      "loss": 0.6997,
-      "step": 17345
-    },
-    {
-      "epoch": 0.8043579044969865,
-      "grad_norm": 4.371307373046875,
-      "learning_rate": 2.242665079201839e-06,
-      "loss": 0.8097,
-      "step": 17350
-    },
-    {
-      "epoch": 0.8045897079276774,
-      "grad_norm": 3.5886342525482178,
-      "learning_rate": 2.237561413261218e-06,
-      "loss": 0.8238,
-      "step": 17355
-    },
-    {
-      "epoch": 0.8048215113583681,
-      "grad_norm": 3.4891817569732666,
-      "learning_rate": 2.232462829530506e-06,
-      "loss": 0.7724,
-      "step": 17360
-    },
-    {
-      "epoch": 0.8050533147890588,
-      "grad_norm": 3.905630588531494,
-      "learning_rate": 2.2273693313478352e-06,
-      "loss": 0.708,
-      "step": 17365
-    },
-    {
-      "epoch": 0.8052851182197497,
-      "grad_norm": 3.9939043521881104,
-      "learning_rate": 2.222280922048017e-06,
-      "loss": 0.8429,
-      "step": 17370
-    },
-    {
-      "epoch": 0.8055169216504404,
-      "grad_norm": 3.5375664234161377,
-      "learning_rate": 2.2171976049625186e-06,
-      "loss": 0.9396,
-      "step": 17375
-    },
-    {
-      "epoch": 0.8057487250811312,
-      "grad_norm": 3.547438144683838,
-      "learning_rate": 2.2121193834194855e-06,
-      "loss": 0.7913,
-      "step": 17380
-    },
-    {
-      "epoch": 0.805980528511822,
-      "grad_norm": 4.114473342895508,
-      "learning_rate": 2.207046260743726e-06,
-      "loss": 0.8347,
-      "step": 17385
-    },
-    {
-      "epoch": 0.8062123319425127,
-      "grad_norm": 3.792337656021118,
-      "learning_rate": 2.2019782402566934e-06,
-      "loss": 0.8177,
-      "step": 17390
-    },
-    {
-      "epoch": 0.8064441353732035,
-      "grad_norm": 3.6505558490753174,
-      "learning_rate": 2.196915325276522e-06,
-      "loss": 0.8452,
-      "step": 17395
-    },
-    {
-      "epoch": 0.8066759388038943,
-      "grad_norm": 4.075436592102051,
-      "learning_rate": 2.1918575191179937e-06,
-      "loss": 0.7933,
-      "step": 17400
-    },
-    {
-      "epoch": 0.8066759388038943,
-      "eval_loss": 0.8806984424591064,
-      "eval_runtime": 11.2762,
-      "eval_samples_per_second": 11.263,
-      "eval_steps_per_second": 11.263,
-      "step": 17400
-    },
-    {
-      "epoch": 0.8069077422345851,
-      "grad_norm": 4.127367973327637,
-      "learning_rate": 2.186804825092542e-06,
-      "loss": 0.8318,
-      "step": 17405
-    },
-    {
-      "epoch": 0.8071395456652758,
-      "grad_norm": 3.078382968902588,
-      "learning_rate": 2.1817572465082627e-06,
-      "loss": 0.7779,
-      "step": 17410
-    },
-    {
-      "epoch": 0.8073713490959666,
-      "grad_norm": 3.11124587059021,
-      "learning_rate": 2.1767147866698967e-06,
-      "loss": 0.8903,
-      "step": 17415
-    },
-    {
-      "epoch": 0.8076031525266574,
-      "grad_norm": 4.386844635009766,
-      "learning_rate": 2.171677448878833e-06,
-      "loss": 0.9116,
-      "step": 17420
-    },
-    {
-      "epoch": 0.8078349559573482,
-      "grad_norm": 3.5533015727996826,
-      "learning_rate": 2.166645236433106e-06,
-      "loss": 0.8415,
-      "step": 17425
-    },
-    {
-      "epoch": 0.808066759388039,
-      "grad_norm": 3.453176736831665,
-      "learning_rate": 2.1616181526273994e-06,
-      "loss": 0.8109,
-      "step": 17430
-    },
-    {
-      "epoch": 0.8082985628187297,
-      "grad_norm": 3.496314525604248,
-      "learning_rate": 2.156596200753038e-06,
-      "loss": 0.976,
-      "step": 17435
-    },
-    {
-      "epoch": 0.8085303662494205,
-      "grad_norm": 3.6104514598846436,
-      "learning_rate": 2.1515793840979827e-06,
-      "loss": 0.7953,
-      "step": 17440
-    },
-    {
-      "epoch": 0.8087621696801113,
-      "grad_norm": 3.6141369342803955,
-      "learning_rate": 2.146567705946837e-06,
-      "loss": 0.7809,
-      "step": 17445
-    },
-    {
-      "epoch": 0.808993973110802,
-      "grad_norm": 4.1143293380737305,
-      "learning_rate": 2.1415611695808334e-06,
-      "loss": 0.9777,
-      "step": 17450
-    },
-    {
-      "epoch": 0.8092257765414929,
-      "grad_norm": 3.966641426086426,
-      "learning_rate": 2.1365597782778493e-06,
-      "loss": 0.9165,
-      "step": 17455
-    },
-    {
-      "epoch": 0.8094575799721836,
-      "grad_norm": 3.243546962738037,
-      "learning_rate": 2.13156353531238e-06,
-      "loss": 0.856,
-      "step": 17460
-    },
-    {
-      "epoch": 0.8096893834028743,
-      "grad_norm": 3.5764763355255127,
-      "learning_rate": 2.1265724439555592e-06,
-      "loss": 0.7839,
-      "step": 17465
-    },
-    {
-      "epoch": 0.8099211868335652,
-      "grad_norm": 3.726209878921509,
-      "learning_rate": 2.1215865074751487e-06,
-      "loss": 0.7766,
-      "step": 17470
-    },
-    {
-      "epoch": 0.8101529902642559,
-      "grad_norm": 4.007201671600342,
-      "learning_rate": 2.116605729135528e-06,
-      "loss": 0.6973,
-      "step": 17475
-    },
-    {
-      "epoch": 0.8103847936949466,
-      "grad_norm": 3.5548746585845947,
-      "learning_rate": 2.1116301121977024e-06,
-      "loss": 0.9117,
-      "step": 17480
-    },
-    {
-      "epoch": 0.8106165971256375,
-      "grad_norm": 4.227999210357666,
-      "learning_rate": 2.1066596599193e-06,
-      "loss": 0.9328,
-      "step": 17485
-    },
-    {
-      "epoch": 0.8108484005563282,
-      "grad_norm": 4.0675201416015625,
-      "learning_rate": 2.1016943755545704e-06,
-      "loss": 1.0801,
-      "step": 17490
-    },
-    {
-      "epoch": 0.811080203987019,
-      "grad_norm": 4.169510364532471,
-      "learning_rate": 2.0967342623543695e-06,
-      "loss": 0.8303,
-      "step": 17495
-    },
-    {
-      "epoch": 0.8113120074177098,
-      "grad_norm": 3.53352952003479,
-      "learning_rate": 2.0917793235661777e-06,
-      "loss": 0.7763,
-      "step": 17500
-    },
-    {
-      "epoch": 0.8113120074177098,
-      "eval_loss": 0.880153477191925,
-      "eval_runtime": 11.2652,
-      "eval_samples_per_second": 11.274,
-      "eval_steps_per_second": 11.274,
-      "step": 17500
-    },
-    {
-      "epoch": 0.8115438108484005,
-      "grad_norm": 3.3654439449310303,
-      "learning_rate": 2.0868295624340874e-06,
-      "loss": 0.932,
-      "step": 17505
-    },
-    {
-      "epoch": 0.8117756142790913,
-      "grad_norm": 4.071439266204834,
-      "learning_rate": 2.0818849821987875e-06,
-      "loss": 0.9639,
-      "step": 17510
-    },
-    {
-      "epoch": 0.8120074177097821,
-      "grad_norm": 3.366882085800171,
-      "learning_rate": 2.0769455860975896e-06,
-      "loss": 0.9948,
-      "step": 17515
-    },
-    {
-      "epoch": 0.8122392211404729,
-      "grad_norm": 4.544328212738037,
-      "learning_rate": 2.07201137736441e-06,
-      "loss": 0.9075,
-      "step": 17520
-    },
-    {
-      "epoch": 0.8124710245711636,
-      "grad_norm": 4.452793598175049,
-      "learning_rate": 2.0670823592297583e-06,
-      "loss": 0.8756,
-      "step": 17525
-    },
-    {
-      "epoch": 0.8127028280018544,
-      "grad_norm": 3.932635545730591,
-      "learning_rate": 2.0621585349207562e-06,
-      "loss": 0.9399,
-      "step": 17530
-    },
-    {
-      "epoch": 0.8129346314325452,
-      "grad_norm": 3.453287363052368,
-      "learning_rate": 2.0572399076611228e-06,
-      "loss": 0.7789,
-      "step": 17535
-    },
-    {
-      "epoch": 0.8131664348632359,
-      "grad_norm": 3.538092613220215,
-      "learning_rate": 2.052326480671171e-06,
-      "loss": 0.8716,
-      "step": 17540
-    },
-    {
-      "epoch": 0.8133982382939268,
-      "grad_norm": 4.095542907714844,
-      "learning_rate": 2.047418257167808e-06,
-      "loss": 0.8383,
-      "step": 17545
-    },
-    {
-      "epoch": 0.8136300417246175,
-      "grad_norm": 3.764967918395996,
-      "learning_rate": 2.04251524036454e-06,
-      "loss": 0.8201,
-      "step": 17550
-    },
-    {
-      "epoch": 0.8138618451553083,
-      "grad_norm": 3.590008497238159,
-      "learning_rate": 2.037617433471464e-06,
-      "loss": 0.6739,
-      "step": 17555
-    },
-    {
-      "epoch": 0.8140936485859991,
-      "grad_norm": 3.986955404281616,
-      "learning_rate": 2.0327248396952582e-06,
-      "loss": 0.8388,
-      "step": 17560
-    },
-    {
-      "epoch": 0.8143254520166898,
-      "grad_norm": 3.728408098220825,
-      "learning_rate": 2.0278374622391983e-06,
-      "loss": 0.9149,
-      "step": 17565
-    },
-    {
-      "epoch": 0.8145572554473807,
-      "grad_norm": 3.5650875568389893,
-      "learning_rate": 2.0229553043031324e-06,
-      "loss": 0.8783,
-      "step": 17570
-    },
-    {
-      "epoch": 0.8147890588780714,
-      "grad_norm": 4.118309497833252,
-      "learning_rate": 2.018078369083506e-06,
-      "loss": 0.7925,
-      "step": 17575
-    },
-    {
-      "epoch": 0.8150208623087621,
-      "grad_norm": 4.11269998550415,
-      "learning_rate": 2.0132066597733302e-06,
-      "loss": 0.9272,
-      "step": 17580
-    },
-    {
-      "epoch": 0.815252665739453,
-      "grad_norm": 3.2680740356445312,
-      "learning_rate": 2.0083401795622046e-06,
-      "loss": 0.8929,
-      "step": 17585
-    },
-    {
-      "epoch": 0.8154844691701437,
-      "grad_norm": 3.561012029647827,
-      "learning_rate": 2.0034789316363067e-06,
-      "loss": 0.7216,
-      "step": 17590
-    },
-    {
-      "epoch": 0.8157162726008345,
-      "grad_norm": 3.8684680461883545,
-      "learning_rate": 1.99862291917838e-06,
-      "loss": 0.8087,
-      "step": 17595
-    },
-    {
-      "epoch": 0.8159480760315253,
-      "grad_norm": 3.5848188400268555,
-      "learning_rate": 1.993772145367742e-06,
-      "loss": 0.9077,
-      "step": 17600
-    },
-    {
-      "epoch": 0.8159480760315253,
-      "eval_loss": 0.8796182870864868,
-      "eval_runtime": 11.273,
-      "eval_samples_per_second": 11.266,
-      "eval_steps_per_second": 11.266,
-      "step": 17600
-    },
-    {
-      "epoch": 0.816179879462216,
-      "grad_norm": 3.892946243286133,
-      "learning_rate": 1.9889266133802877e-06,
-      "loss": 0.8927,
-      "step": 17605
-    },
-    {
-      "epoch": 0.8164116828929068,
-      "grad_norm": 3.616286277770996,
-      "learning_rate": 1.98408632638847e-06,
-      "loss": 0.838,
-      "step": 17610
-    },
-    {
-      "epoch": 0.8166434863235976,
-      "grad_norm": 3.7281265258789062,
-      "learning_rate": 1.979251287561316e-06,
-      "loss": 0.7584,
-      "step": 17615
-    },
-    {
-      "epoch": 0.8168752897542884,
-      "grad_norm": 3.5958240032196045,
-      "learning_rate": 1.9744215000644153e-06,
-      "loss": 0.8676,
-      "step": 17620
-    },
-    {
-      "epoch": 0.8171070931849791,
-      "grad_norm": 3.500471591949463,
-      "learning_rate": 1.9695969670599156e-06,
-      "loss": 0.7494,
-      "step": 17625
-    },
-    {
-      "epoch": 0.8173388966156699,
-      "grad_norm": 3.6047792434692383,
-      "learning_rate": 1.9647776917065253e-06,
-      "loss": 0.8516,
-      "step": 17630
-    },
-    {
-      "epoch": 0.8175707000463607,
-      "grad_norm": 3.827798366546631,
-      "learning_rate": 1.959963677159512e-06,
-      "loss": 0.9751,
-      "step": 17635
-    },
-    {
-      "epoch": 0.8178025034770514,
-      "grad_norm": 3.7230641841888428,
-      "learning_rate": 1.9551549265707025e-06,
-      "loss": 0.8542,
-      "step": 17640
-    },
-    {
-      "epoch": 0.8180343069077423,
-      "grad_norm": 3.4928386211395264,
-      "learning_rate": 1.9503514430884685e-06,
-      "loss": 0.7689,
-      "step": 17645
-    },
-    {
-      "epoch": 0.818266110338433,
-      "grad_norm": 3.538079261779785,
-      "learning_rate": 1.9455532298577417e-06,
-      "loss": 0.6581,
-      "step": 17650
-    },
-    {
-      "epoch": 0.8184979137691237,
-      "grad_norm": 3.569852590560913,
-      "learning_rate": 1.9407602900199975e-06,
-      "loss": 0.9138,
-      "step": 17655
-    },
-    {
-      "epoch": 0.8187297171998146,
-      "grad_norm": 4.027120590209961,
-      "learning_rate": 1.935972626713263e-06,
-      "loss": 0.9222,
-      "step": 17660
-    },
-    {
-      "epoch": 0.8189615206305053,
-      "grad_norm": 3.9751973152160645,
-      "learning_rate": 1.931190243072105e-06,
-      "loss": 0.8335,
-      "step": 17665
-    },
-    {
-      "epoch": 0.8191933240611962,
-      "grad_norm": 3.7292730808258057,
-      "learning_rate": 1.9264131422276387e-06,
-      "loss": 0.8459,
-      "step": 17670
-    },
-    {
-      "epoch": 0.8194251274918869,
-      "grad_norm": 4.147140979766846,
-      "learning_rate": 1.921641327307523e-06,
-      "loss": 0.7935,
-      "step": 17675
-    },
-    {
-      "epoch": 0.8196569309225776,
-      "grad_norm": 4.169584274291992,
-      "learning_rate": 1.9168748014359494e-06,
-      "loss": 0.8263,
-      "step": 17680
-    },
-    {
-      "epoch": 0.8198887343532685,
-      "grad_norm": 3.447660446166992,
-      "learning_rate": 1.9121135677336445e-06,
-      "loss": 1.0089,
-      "step": 17685
-    },
-    {
-      "epoch": 0.8201205377839592,
-      "grad_norm": 3.7443501949310303,
-      "learning_rate": 1.9073576293178797e-06,
-      "loss": 0.8516,
-      "step": 17690
-    },
-    {
-      "epoch": 0.82035234121465,
-      "grad_norm": 3.317669630050659,
-      "learning_rate": 1.9026069893024556e-06,
-      "loss": 0.7569,
-      "step": 17695
-    },
-    {
-      "epoch": 0.8205841446453408,
-      "grad_norm": 3.989811897277832,
-      "learning_rate": 1.8978616507976989e-06,
-      "loss": 0.6587,
-      "step": 17700
-    },
-    {
-      "epoch": 0.8205841446453408,
-      "eval_loss": 0.8793162703514099,
-      "eval_runtime": 11.2717,
-      "eval_samples_per_second": 11.267,
-      "eval_steps_per_second": 11.267,
-      "step": 17700
-    },
-    {
-      "epoch": 0.8208159480760315,
-      "grad_norm": 3.4570274353027344,
-      "learning_rate": 1.8931216169104704e-06,
-      "loss": 0.8214,
-      "step": 17705
-    },
-    {
-      "epoch": 0.8210477515067223,
-      "grad_norm": 3.902951240539551,
-      "learning_rate": 1.8883868907441615e-06,
-      "loss": 0.7536,
-      "step": 17710
-    },
-    {
-      "epoch": 0.8212795549374131,
-      "grad_norm": 3.0927488803863525,
-      "learning_rate": 1.8836574753986758e-06,
-      "loss": 0.7152,
-      "step": 17715
-    },
-    {
-      "epoch": 0.8215113583681039,
-      "grad_norm": 4.087670803070068,
-      "learning_rate": 1.8789333739704496e-06,
-      "loss": 0.9579,
-      "step": 17720
-    },
-    {
-      "epoch": 0.8217431617987946,
-      "grad_norm": 3.9503560066223145,
-      "learning_rate": 1.8742145895524432e-06,
-      "loss": 0.7369,
-      "step": 17725
-    },
-    {
-      "epoch": 0.8219749652294854,
-      "grad_norm": 2.7508480548858643,
-      "learning_rate": 1.8695011252341244e-06,
-      "loss": 0.7558,
-      "step": 17730
-    },
-    {
-      "epoch": 0.8222067686601762,
-      "grad_norm": 7.382952690124512,
-      "learning_rate": 1.8647929841014867e-06,
-      "loss": 0.9415,
-      "step": 17735
-    },
-    {
-      "epoch": 0.8224385720908669,
-      "grad_norm": 3.3182077407836914,
-      "learning_rate": 1.8600901692370377e-06,
-      "loss": 0.8364,
-      "step": 17740
-    },
-    {
-      "epoch": 0.8226703755215578,
-      "grad_norm": 4.234379291534424,
-      "learning_rate": 1.8553926837197956e-06,
-      "loss": 1.0117,
-      "step": 17745
-    },
-    {
-      "epoch": 0.8229021789522485,
-      "grad_norm": 3.2763216495513916,
-      "learning_rate": 1.8507005306252856e-06,
-      "loss": 0.8022,
-      "step": 17750
-    },
-    {
-      "epoch": 0.8231339823829392,
-      "grad_norm": 4.160790920257568,
-      "learning_rate": 1.8460137130255472e-06,
-      "loss": 0.9042,
-      "step": 17755
-    },
-    {
-      "epoch": 0.8233657858136301,
-      "grad_norm": 3.7351162433624268,
-      "learning_rate": 1.8413322339891303e-06,
-      "loss": 0.8661,
-      "step": 17760
-    },
-    {
-      "epoch": 0.8235975892443208,
-      "grad_norm": 3.6955184936523438,
-      "learning_rate": 1.8366560965810787e-06,
-      "loss": 0.8112,
-      "step": 17765
-    },
-    {
-      "epoch": 0.8238293926750115,
-      "grad_norm": 3.1400656700134277,
-      "learning_rate": 1.8319853038629486e-06,
-      "loss": 0.8455,
-      "step": 17770
-    },
-    {
-      "epoch": 0.8240611961057024,
-      "grad_norm": 3.6449131965637207,
-      "learning_rate": 1.8273198588927887e-06,
-      "loss": 0.7387,
-      "step": 17775
-    },
-    {
-      "epoch": 0.8242929995363931,
-      "grad_norm": 3.8718135356903076,
-      "learning_rate": 1.8226597647251564e-06,
-      "loss": 0.8486,
-      "step": 17780
-    },
-    {
-      "epoch": 0.8245248029670839,
-      "grad_norm": 3.834111452102661,
-      "learning_rate": 1.8180050244110947e-06,
-      "loss": 0.7541,
-      "step": 17785
-    },
-    {
-      "epoch": 0.8247566063977747,
-      "grad_norm": 3.8849096298217773,
-      "learning_rate": 1.813355640998149e-06,
-      "loss": 0.9113,
-      "step": 17790
-    },
-    {
-      "epoch": 0.8249884098284654,
-      "grad_norm": 4.433188438415527,
-      "learning_rate": 1.8087116175303577e-06,
-      "loss": 0.8947,
-      "step": 17795
-    },
-    {
-      "epoch": 0.8252202132591563,
-      "grad_norm": 3.6504225730895996,
-      "learning_rate": 1.8040729570482452e-06,
-      "loss": 0.8461,
-      "step": 17800
-    },
-    {
-      "epoch": 0.8252202132591563,
-      "eval_loss": 0.8788275718688965,
-      "eval_runtime": 11.2707,
-      "eval_samples_per_second": 11.268,
-      "eval_steps_per_second": 11.268,
-      "step": 17800
-    },
-    {
-      "epoch": 0.825452016689847,
-      "grad_norm": 3.6747710704803467,
-      "learning_rate": 1.7994396625888255e-06,
-      "loss": 0.865,
-      "step": 17805
-    },
-    {
-      "epoch": 0.8256838201205378,
-      "grad_norm": 3.45422101020813,
-      "learning_rate": 1.7948117371856045e-06,
-      "loss": 0.9433,
-      "step": 17810
-    },
-    {
-      "epoch": 0.8259156235512286,
-      "grad_norm": 3.293832302093506,
-      "learning_rate": 1.7901891838685659e-06,
-      "loss": 0.8554,
-      "step": 17815
-    },
-    {
-      "epoch": 0.8261474269819193,
-      "grad_norm": 3.420264959335327,
-      "learning_rate": 1.78557200566418e-06,
-      "loss": 0.8656,
-      "step": 17820
-    },
-    {
-      "epoch": 0.8263792304126101,
-      "grad_norm": 4.223310470581055,
-      "learning_rate": 1.7809602055954012e-06,
-      "loss": 0.8573,
-      "step": 17825
-    },
-    {
-      "epoch": 0.8266110338433009,
-      "grad_norm": 3.8550260066986084,
-      "learning_rate": 1.776353786681656e-06,
-      "loss": 0.8986,
-      "step": 17830
-    },
-    {
-      "epoch": 0.8268428372739917,
-      "grad_norm": 3.8833189010620117,
-      "learning_rate": 1.771752751938849e-06,
-      "loss": 0.7821,
-      "step": 17835
-    },
-    {
-      "epoch": 0.8270746407046824,
-      "grad_norm": 3.930795669555664,
-      "learning_rate": 1.7671571043793634e-06,
-      "loss": 0.8693,
-      "step": 17840
-    },
-    {
-      "epoch": 0.8273064441353732,
-      "grad_norm": 4.167591571807861,
-      "learning_rate": 1.762566847012055e-06,
-      "loss": 0.7985,
-      "step": 17845
-    },
-    {
-      "epoch": 0.827538247566064,
-      "grad_norm": 3.4307236671447754,
-      "learning_rate": 1.7579819828422451e-06,
-      "loss": 0.6935,
-      "step": 17850
-    },
-    {
-      "epoch": 0.8277700509967547,
-      "grad_norm": 4.414237976074219,
-      "learning_rate": 1.7534025148717327e-06,
-      "loss": 0.7815,
-      "step": 17855
-    },
-    {
-      "epoch": 0.8280018544274456,
-      "grad_norm": 3.4899954795837402,
-      "learning_rate": 1.7488284460987737e-06,
-      "loss": 0.9855,
-      "step": 17860
-    },
-    {
-      "epoch": 0.8282336578581363,
-      "grad_norm": 3.925497055053711,
-      "learning_rate": 1.7442597795180994e-06,
-      "loss": 0.8605,
-      "step": 17865
-    },
-    {
-      "epoch": 0.828465461288827,
-      "grad_norm": 4.437492370605469,
-      "learning_rate": 1.7396965181208935e-06,
-      "loss": 0.9315,
-      "step": 17870
-    },
-    {
-      "epoch": 0.8286972647195179,
-      "grad_norm": 3.858232021331787,
-      "learning_rate": 1.7351386648948088e-06,
-      "loss": 0.7953,
-      "step": 17875
-    },
-    {
-      "epoch": 0.8289290681502086,
-      "grad_norm": 3.927126884460449,
-      "learning_rate": 1.7305862228239579e-06,
-      "loss": 0.9262,
-      "step": 17880
-    },
-    {
-      "epoch": 0.8291608715808994,
-      "grad_norm": 3.5855329036712646,
-      "learning_rate": 1.7260391948889043e-06,
-      "loss": 0.8873,
-      "step": 17885
-    },
-    {
-      "epoch": 0.8293926750115902,
-      "grad_norm": 4.403600215911865,
-      "learning_rate": 1.7214975840666681e-06,
-      "loss": 0.8922,
-      "step": 17890
-    },
-    {
-      "epoch": 0.8296244784422809,
-      "grad_norm": 3.798386812210083,
-      "learning_rate": 1.7169613933307294e-06,
-      "loss": 0.934,
-      "step": 17895
-    },
-    {
-      "epoch": 0.8298562818729717,
-      "grad_norm": 3.3800482749938965,
-      "learning_rate": 1.7124306256510104e-06,
-      "loss": 0.873,
-      "step": 17900
-    },
-    {
-      "epoch": 0.8298562818729717,
-      "eval_loss": 0.8784139752388,
-      "eval_runtime": 11.2612,
-      "eval_samples_per_second": 11.278,
-      "eval_steps_per_second": 11.278,
-      "step": 17900
-    },
-    {
-      "epoch": 0.8300880853036625,
-      "grad_norm": 3.2892932891845703,
-      "learning_rate": 1.7079052839938881e-06,
-      "loss": 0.8444,
-      "step": 17905
-    },
-    {
-      "epoch": 0.8303198887343533,
-      "grad_norm": 4.069420337677002,
-      "learning_rate": 1.703385371322188e-06,
-      "loss": 0.737,
-      "step": 17910
-    },
-    {
-      "epoch": 0.8305516921650441,
-      "grad_norm": 3.6989574432373047,
-      "learning_rate": 1.6988708905951834e-06,
-      "loss": 0.8348,
-      "step": 17915
-    },
-    {
-      "epoch": 0.8307834955957348,
-      "grad_norm": 3.9831721782684326,
-      "learning_rate": 1.6943618447685773e-06,
-      "loss": 0.7892,
-      "step": 17920
-    },
-    {
-      "epoch": 0.8310152990264256,
-      "grad_norm": 3.6940042972564697,
-      "learning_rate": 1.6898582367945292e-06,
-      "loss": 0.9604,
-      "step": 17925
-    },
-    {
-      "epoch": 0.8312471024571164,
-      "grad_norm": 3.789471387863159,
-      "learning_rate": 1.6853600696216344e-06,
-      "loss": 0.9228,
-      "step": 17930
-    },
-    {
-      "epoch": 0.8314789058878072,
-      "grad_norm": 3.592510461807251,
-      "learning_rate": 1.6808673461949222e-06,
-      "loss": 0.8276,
-      "step": 17935
-    },
-    {
-      "epoch": 0.8317107093184979,
-      "grad_norm": 3.432971715927124,
-      "learning_rate": 1.6763800694558608e-06,
-      "loss": 0.8355,
-      "step": 17940
-    },
-    {
-      "epoch": 0.8319425127491887,
-      "grad_norm": 3.7236011028289795,
-      "learning_rate": 1.6718982423423557e-06,
-      "loss": 0.8081,
-      "step": 17945
-    },
-    {
-      "epoch": 0.8321743161798795,
-      "grad_norm": 3.5183401107788086,
-      "learning_rate": 1.6674218677887388e-06,
-      "loss": 0.7581,
-      "step": 17950
-    },
-    {
-      "epoch": 0.8324061196105702,
-      "grad_norm": 3.4557225704193115,
-      "learning_rate": 1.66295094872577e-06,
-      "loss": 0.8102,
-      "step": 17955
-    },
-    {
-      "epoch": 0.832637923041261,
-      "grad_norm": 3.7359535694122314,
-      "learning_rate": 1.6584854880806445e-06,
-      "loss": 0.969,
-      "step": 17960
-    },
-    {
-      "epoch": 0.8328697264719518,
-      "grad_norm": 3.518160343170166,
-      "learning_rate": 1.6540254887769824e-06,
-      "loss": 0.7597,
-      "step": 17965
-    },
-    {
-      "epoch": 0.8331015299026425,
-      "grad_norm": 3.740760087966919,
-      "learning_rate": 1.6495709537348238e-06,
-      "loss": 0.7805,
-      "step": 17970
-    },
-    {
-      "epoch": 0.8333333333333334,
-      "grad_norm": 3.5426766872406006,
-      "learning_rate": 1.6451218858706374e-06,
-      "loss": 0.807,
-      "step": 17975
-    },
-    {
-      "epoch": 0.8335651367640241,
-      "grad_norm": 3.917860746383667,
-      "learning_rate": 1.6406782880973038e-06,
-      "loss": 0.8023,
-      "step": 17980
-    },
-    {
-      "epoch": 0.8337969401947148,
-      "grad_norm": 3.597020149230957,
-      "learning_rate": 1.636240163324132e-06,
-      "loss": 0.8634,
-      "step": 17985
-    },
-    {
-      "epoch": 0.8340287436254057,
-      "grad_norm": 3.77585506439209,
-      "learning_rate": 1.631807514456839e-06,
-      "loss": 0.8499,
-      "step": 17990
-    },
-    {
-      "epoch": 0.8342605470560964,
-      "grad_norm": 3.657649278640747,
-      "learning_rate": 1.6273803443975633e-06,
-      "loss": 0.8618,
-      "step": 17995
-    },
-    {
-      "epoch": 0.8344923504867872,
-      "grad_norm": 4.247122287750244,
-      "learning_rate": 1.6229586560448562e-06,
-      "loss": 0.908,
-      "step": 18000
-    },
-    {
-      "epoch": 0.8344923504867872,
-      "eval_loss": 0.8780138492584229,
-      "eval_runtime": 11.2702,
-      "eval_samples_per_second": 11.269,
-      "eval_steps_per_second": 11.269,
-      "step": 18000
-    },
-    {
-      "epoch": 0.834724153917478,
-      "grad_norm": 3.2867918014526367,
-      "learning_rate": 1.6185424522936743e-06,
-      "loss": 0.8368,
-      "step": 18005
-    },
-    {
-      "epoch": 0.8349559573481687,
-      "grad_norm": 3.2922523021698,
-      "learning_rate": 1.6141317360353847e-06,
-      "loss": 0.741,
-      "step": 18010
-    },
-    {
-      "epoch": 0.8351877607788595,
-      "grad_norm": 3.9418952465057373,
-      "learning_rate": 1.6097265101577697e-06,
-      "loss": 0.7602,
-      "step": 18015
-    },
-    {
-      "epoch": 0.8354195642095503,
-      "grad_norm": 3.6202547550201416,
-      "learning_rate": 1.6053267775450032e-06,
-      "loss": 0.8347,
-      "step": 18020
-    },
-    {
-      "epoch": 0.8356513676402411,
-      "grad_norm": 3.360456705093384,
-      "learning_rate": 1.6009325410776754e-06,
-      "loss": 0.7795,
-      "step": 18025
-    },
-    {
-      "epoch": 0.8358831710709318,
-      "grad_norm": 3.2623724937438965,
-      "learning_rate": 1.5965438036327741e-06,
-      "loss": 0.89,
-      "step": 18030
-    },
-    {
-      "epoch": 0.8361149745016226,
-      "grad_norm": 4.737086772918701,
-      "learning_rate": 1.5921605680836838e-06,
-      "loss": 0.96,
-      "step": 18035
-    },
-    {
-      "epoch": 0.8363467779323134,
-      "grad_norm": 3.349141836166382,
-      "learning_rate": 1.5877828373001857e-06,
-      "loss": 0.8303,
-      "step": 18040
-    },
-    {
-      "epoch": 0.8365785813630042,
-      "grad_norm": 4.2403950691223145,
-      "learning_rate": 1.5834106141484628e-06,
-      "loss": 0.9327,
-      "step": 18045
-    },
-    {
-      "epoch": 0.836810384793695,
-      "grad_norm": 3.949309825897217,
-      "learning_rate": 1.5790439014910908e-06,
-      "loss": 0.7971,
-      "step": 18050
-    },
-    {
-      "epoch": 0.8370421882243857,
-      "grad_norm": 3.6255528926849365,
-      "learning_rate": 1.5746827021870315e-06,
-      "loss": 0.8399,
-      "step": 18055
-    },
-    {
-      "epoch": 0.8372739916550765,
-      "grad_norm": 3.5447537899017334,
-      "learning_rate": 1.5703270190916453e-06,
-      "loss": 0.721,
-      "step": 18060
-    },
-    {
-      "epoch": 0.8375057950857673,
-      "grad_norm": 4.385354042053223,
-      "learning_rate": 1.5659768550566768e-06,
-      "loss": 0.8497,
-      "step": 18065
-    },
-    {
-      "epoch": 0.837737598516458,
-      "grad_norm": 3.67063307762146,
-      "learning_rate": 1.5616322129302586e-06,
-      "loss": 0.9161,
-      "step": 18070
-    },
-    {
-      "epoch": 0.8379694019471489,
-      "grad_norm": 3.8790194988250732,
-      "learning_rate": 1.557293095556901e-06,
-      "loss": 0.9169,
-      "step": 18075
-    },
-    {
-      "epoch": 0.8382012053778396,
-      "grad_norm": 3.567552089691162,
-      "learning_rate": 1.5529595057775093e-06,
-      "loss": 0.7768,
-      "step": 18080
-    },
-    {
-      "epoch": 0.8384330088085303,
-      "grad_norm": 3.9901583194732666,
-      "learning_rate": 1.5486314464293627e-06,
-      "loss": 0.8621,
-      "step": 18085
-    },
-    {
-      "epoch": 0.8386648122392212,
-      "grad_norm": 3.9711081981658936,
-      "learning_rate": 1.5443089203461203e-06,
-      "loss": 0.8611,
-      "step": 18090
-    },
-    {
-      "epoch": 0.8388966156699119,
-      "grad_norm": 3.800506353378296,
-      "learning_rate": 1.539991930357816e-06,
-      "loss": 0.8638,
-      "step": 18095
-    },
-    {
-      "epoch": 0.8391284191006027,
-      "grad_norm": 3.651494026184082,
-      "learning_rate": 1.5356804792908652e-06,
-      "loss": 0.9013,
-      "step": 18100
-    },
-    {
-      "epoch": 0.8391284191006027,
-      "eval_loss": 0.8775888681411743,
-      "eval_runtime": 11.2697,
-      "eval_samples_per_second": 11.269,
-      "eval_steps_per_second": 11.269,
-      "step": 18100
-    },
-    {
-      "epoch": 0.8393602225312935,
-      "grad_norm": 4.302680969238281,
-      "learning_rate": 1.5313745699680505e-06,
-      "loss": 0.9629,
-      "step": 18105
-    },
-    {
-      "epoch": 0.8395920259619842,
-      "grad_norm": 3.702319860458374,
-      "learning_rate": 1.5270742052085296e-06,
-      "loss": 0.8363,
-      "step": 18110
-    },
-    {
-      "epoch": 0.839823829392675,
-      "grad_norm": 3.423970937728882,
-      "learning_rate": 1.522779387827833e-06,
-      "loss": 0.7962,
-      "step": 18115
-    },
-    {
-      "epoch": 0.8400556328233658,
-      "grad_norm": 3.711111068725586,
-      "learning_rate": 1.5184901206378545e-06,
-      "loss": 0.838,
-      "step": 18120
-    },
-    {
-      "epoch": 0.8402874362540566,
-      "grad_norm": 3.613584041595459,
-      "learning_rate": 1.514206406446851e-06,
-      "loss": 0.7964,
-      "step": 18125
-    },
-    {
-      "epoch": 0.8405192396847473,
-      "grad_norm": 3.8854293823242188,
-      "learning_rate": 1.5099282480594523e-06,
-      "loss": 0.7295,
-      "step": 18130
-    },
-    {
-      "epoch": 0.8407510431154381,
-      "grad_norm": 3.463650941848755,
-      "learning_rate": 1.5056556482766483e-06,
-      "loss": 0.7444,
-      "step": 18135
-    },
-    {
-      "epoch": 0.8409828465461289,
-      "grad_norm": 3.2122750282287598,
-      "learning_rate": 1.501388609895782e-06,
-      "loss": 0.7282,
-      "step": 18140
-    },
-    {
-      "epoch": 0.8412146499768196,
-      "grad_norm": 4.111193656921387,
-      "learning_rate": 1.4971271357105655e-06,
-      "loss": 0.9369,
-      "step": 18145
-    },
-    {
-      "epoch": 0.8414464534075105,
-      "grad_norm": 3.4979441165924072,
-      "learning_rate": 1.4928712285110647e-06,
-      "loss": 0.8797,
-      "step": 18150
-    },
-    {
-      "epoch": 0.8416782568382012,
-      "grad_norm": 4.0167341232299805,
-      "learning_rate": 1.4886208910836986e-06,
-      "loss": 0.7936,
-      "step": 18155
-    },
-    {
-      "epoch": 0.8419100602688919,
-      "grad_norm": 3.7691385746002197,
-      "learning_rate": 1.4843761262112355e-06,
-      "loss": 0.7724,
-      "step": 18160
-    },
-    {
-      "epoch": 0.8421418636995828,
-      "grad_norm": 4.039342880249023,
-      "learning_rate": 1.4801369366728059e-06,
-      "loss": 0.9082,
-      "step": 18165
-    },
-    {
-      "epoch": 0.8423736671302735,
-      "grad_norm": 4.11240816116333,
-      "learning_rate": 1.4759033252438837e-06,
-      "loss": 0.9212,
-      "step": 18170
-    },
-    {
-      "epoch": 0.8426054705609644,
-      "grad_norm": 3.6150293350219727,
-      "learning_rate": 1.4716752946962886e-06,
-      "loss": 1.0067,
-      "step": 18175
-    },
-    {
-      "epoch": 0.8428372739916551,
-      "grad_norm": 3.085090398788452,
-      "learning_rate": 1.4674528477981931e-06,
-      "loss": 0.7283,
-      "step": 18180
-    },
-    {
-      "epoch": 0.8430690774223458,
-      "grad_norm": 3.6002490520477295,
-      "learning_rate": 1.4632359873141055e-06,
-      "loss": 0.7917,
-      "step": 18185
-    },
-    {
-      "epoch": 0.8433008808530367,
-      "grad_norm": 3.630927324295044,
-      "learning_rate": 1.4590247160048865e-06,
-      "loss": 0.8492,
-      "step": 18190
-    },
-    {
-      "epoch": 0.8435326842837274,
-      "grad_norm": 3.3591744899749756,
-      "learning_rate": 1.454819036627727e-06,
-      "loss": 0.8766,
-      "step": 18195
-    },
-    {
-      "epoch": 0.8437644877144181,
-      "grad_norm": 3.287750005722046,
-      "learning_rate": 1.4506189519361636e-06,
-      "loss": 0.7674,
-      "step": 18200
-    },
-    {
-      "epoch": 0.8437644877144181,
-      "eval_loss": 0.8770650029182434,
-      "eval_runtime": 11.2617,
-      "eval_samples_per_second": 11.277,
-      "eval_steps_per_second": 11.277,
-      "step": 18200
-    },
-    {
-      "epoch": 0.843996291145109,
-      "grad_norm": 3.973710536956787,
-      "learning_rate": 1.4464244646800718e-06,
-      "loss": 0.7988,
-      "step": 18205
-    },
-    {
-      "epoch": 0.8442280945757997,
-      "grad_norm": 4.015002250671387,
-      "learning_rate": 1.4422355776056552e-06,
-      "loss": 1.0665,
-      "step": 18210
-    },
-    {
-      "epoch": 0.8444598980064905,
-      "grad_norm": 3.847734212875366,
-      "learning_rate": 1.438052293455452e-06,
-      "loss": 0.8656,
-      "step": 18215
-    },
-    {
-      "epoch": 0.8446917014371813,
-      "grad_norm": 4.58140230178833,
-      "learning_rate": 1.4338746149683414e-06,
-      "loss": 0.9376,
-      "step": 18220
-    },
-    {
-      "epoch": 0.844923504867872,
-      "grad_norm": 3.9086010456085205,
-      "learning_rate": 1.4297025448795187e-06,
-      "loss": 0.789,
-      "step": 18225
-    },
-    {
-      "epoch": 0.8451553082985628,
-      "grad_norm": 4.11086368560791,
-      "learning_rate": 1.425536085920517e-06,
-      "loss": 0.7864,
-      "step": 18230
-    },
-    {
-      "epoch": 0.8453871117292536,
-      "grad_norm": 3.6161892414093018,
-      "learning_rate": 1.421375240819196e-06,
-      "loss": 0.9303,
-      "step": 18235
-    },
-    {
-      "epoch": 0.8456189151599444,
-      "grad_norm": 3.4047515392303467,
-      "learning_rate": 1.4172200122997327e-06,
-      "loss": 0.7881,
-      "step": 18240
-    },
-    {
-      "epoch": 0.8458507185906351,
-      "grad_norm": 4.340506553649902,
-      "learning_rate": 1.4130704030826304e-06,
-      "loss": 0.9115,
-      "step": 18245
-    },
-    {
-      "epoch": 0.846082522021326,
-      "grad_norm": 3.772585391998291,
-      "learning_rate": 1.408926415884716e-06,
-      "loss": 0.9117,
-      "step": 18250
-    },
-    {
-      "epoch": 0.8463143254520167,
-      "grad_norm": 4.270310878753662,
-      "learning_rate": 1.4047880534191338e-06,
-      "loss": 0.927,
-      "step": 18255
-    },
-    {
-      "epoch": 0.8465461288827074,
-      "grad_norm": 3.5970466136932373,
-      "learning_rate": 1.400655318395343e-06,
-      "loss": 0.8635,
-      "step": 18260
-    },
-    {
-      "epoch": 0.8467779323133983,
-      "grad_norm": 2.8994784355163574,
-      "learning_rate": 1.3965282135191205e-06,
-      "loss": 0.8529,
-      "step": 18265
-    },
-    {
-      "epoch": 0.847009735744089,
-      "grad_norm": 3.8191888332366943,
-      "learning_rate": 1.3924067414925613e-06,
-      "loss": 0.719,
-      "step": 18270
-    },
-    {
-      "epoch": 0.8472415391747797,
-      "grad_norm": 3.4908547401428223,
-      "learning_rate": 1.3882909050140646e-06,
-      "loss": 0.8734,
-      "step": 18275
-    },
-    {
-      "epoch": 0.8474733426054706,
-      "grad_norm": 4.079155921936035,
-      "learning_rate": 1.3841807067783419e-06,
-      "loss": 0.9573,
-      "step": 18280
-    },
-    {
-      "epoch": 0.8477051460361613,
-      "grad_norm": 4.186222553253174,
-      "learning_rate": 1.380076149476417e-06,
-      "loss": 0.8626,
-      "step": 18285
-    },
-    {
-      "epoch": 0.8479369494668522,
-      "grad_norm": 3.282341718673706,
-      "learning_rate": 1.375977235795619e-06,
-      "loss": 0.7254,
-      "step": 18290
-    },
-    {
-      "epoch": 0.8481687528975429,
-      "grad_norm": 4.016537666320801,
-      "learning_rate": 1.3718839684195795e-06,
-      "loss": 0.8116,
-      "step": 18295
-    },
-    {
-      "epoch": 0.8484005563282336,
-      "grad_norm": 3.811028480529785,
-      "learning_rate": 1.3677963500282377e-06,
-      "loss": 0.876,
-      "step": 18300
-    },
-    {
-      "epoch": 0.8484005563282336,
-      "eval_loss": 0.8769089579582214,
-      "eval_runtime": 11.2618,
-      "eval_samples_per_second": 11.277,
-      "eval_steps_per_second": 11.277,
-      "step": 18300
-    },
-    {
-      "epoch": 0.8486323597589245,
-      "grad_norm": 3.929299831390381,
-      "learning_rate": 1.3637143832978305e-06,
-      "loss": 0.9984,
-      "step": 18305
-    },
-    {
-      "epoch": 0.8488641631896152,
-      "grad_norm": 3.347557544708252,
-      "learning_rate": 1.3596380709008938e-06,
-      "loss": 0.7783,
-      "step": 18310
-    },
-    {
-      "epoch": 0.849095966620306,
-      "grad_norm": 3.877534866333008,
-      "learning_rate": 1.355567415506266e-06,
-      "loss": 0.8746,
-      "step": 18315
-    },
-    {
-      "epoch": 0.8493277700509968,
-      "grad_norm": 3.770537853240967,
-      "learning_rate": 1.3515024197790794e-06,
-      "loss": 0.9068,
-      "step": 18320
-    },
-    {
-      "epoch": 0.8495595734816875,
-      "grad_norm": 4.099272727966309,
-      "learning_rate": 1.3474430863807608e-06,
-      "loss": 0.8611,
-      "step": 18325
-    },
-    {
-      "epoch": 0.8497913769123783,
-      "grad_norm": 4.282232284545898,
-      "learning_rate": 1.3433894179690266e-06,
-      "loss": 0.8107,
-      "step": 18330
-    },
-    {
-      "epoch": 0.8500231803430691,
-      "grad_norm": 3.5612881183624268,
-      "learning_rate": 1.3393414171978892e-06,
-      "loss": 0.8929,
-      "step": 18335
-    },
-    {
-      "epoch": 0.8502549837737599,
-      "grad_norm": 3.6753652095794678,
-      "learning_rate": 1.3352990867176519e-06,
-      "loss": 0.7928,
-      "step": 18340
-    },
-    {
-      "epoch": 0.8504867872044506,
-      "grad_norm": 3.9664688110351562,
-      "learning_rate": 1.3312624291748965e-06,
-      "loss": 0.9438,
-      "step": 18345
-    },
-    {
-      "epoch": 0.8507185906351414,
-      "grad_norm": 4.463581085205078,
-      "learning_rate": 1.3272314472124992e-06,
-      "loss": 0.8554,
-      "step": 18350
-    },
-    {
-      "epoch": 0.8509503940658322,
-      "grad_norm": 4.024453163146973,
-      "learning_rate": 1.3232061434696186e-06,
-      "loss": 0.8932,
-      "step": 18355
-    },
-    {
-      "epoch": 0.8511821974965229,
-      "grad_norm": 3.6871092319488525,
-      "learning_rate": 1.3191865205816923e-06,
-      "loss": 0.8562,
-      "step": 18360
-    },
-    {
-      "epoch": 0.8514140009272138,
-      "grad_norm": 3.467263698577881,
-      "learning_rate": 1.3151725811804395e-06,
-      "loss": 0.8576,
-      "step": 18365
-    },
-    {
-      "epoch": 0.8516458043579045,
-      "grad_norm": 3.4049699306488037,
-      "learning_rate": 1.3111643278938613e-06,
-      "loss": 0.7381,
-      "step": 18370
-    },
-    {
-      "epoch": 0.8518776077885952,
-      "grad_norm": 3.9009854793548584,
-      "learning_rate": 1.3071617633462351e-06,
-      "loss": 0.8384,
-      "step": 18375
-    },
-    {
-      "epoch": 0.8521094112192861,
-      "grad_norm": 4.61009407043457,
-      "learning_rate": 1.3031648901581084e-06,
-      "loss": 0.7304,
-      "step": 18380
-    },
-    {
-      "epoch": 0.8523412146499768,
-      "grad_norm": 3.888120412826538,
-      "learning_rate": 1.2991737109463132e-06,
-      "loss": 0.983,
-      "step": 18385
-    },
-    {
-      "epoch": 0.8525730180806675,
-      "grad_norm": 3.6879374980926514,
-      "learning_rate": 1.2951882283239414e-06,
-      "loss": 0.9002,
-      "step": 18390
-    },
-    {
-      "epoch": 0.8528048215113584,
-      "grad_norm": 3.665576457977295,
-      "learning_rate": 1.291208444900366e-06,
-      "loss": 0.7605,
-      "step": 18395
-    },
-    {
-      "epoch": 0.8530366249420491,
-      "grad_norm": 3.582505226135254,
-      "learning_rate": 1.287234363281219e-06,
-      "loss": 0.8616,
-      "step": 18400
-    },
-    {
-      "epoch": 0.8530366249420491,
-      "eval_loss": 0.8765570521354675,
-      "eval_runtime": 11.2681,
-      "eval_samples_per_second": 11.271,
-      "eval_steps_per_second": 11.271,
-      "step": 18400
-    },
-    {
-      "epoch": 0.8532684283727399,
-      "grad_norm": 3.894113063812256,
-      "learning_rate": 1.283265986068406e-06,
-      "loss": 0.7478,
-      "step": 18405
-    },
-    {
-      "epoch": 0.8535002318034307,
-      "grad_norm": 3.673455238342285,
-      "learning_rate": 1.2793033158600987e-06,
-      "loss": 0.7809,
-      "step": 18410
-    },
-    {
-      "epoch": 0.8537320352341214,
-      "grad_norm": 3.5018301010131836,
-      "learning_rate": 1.2753463552507262e-06,
-      "loss": 0.971,
-      "step": 18415
-    },
-    {
-      "epoch": 0.8539638386648123,
-      "grad_norm": 3.3875620365142822,
-      "learning_rate": 1.2713951068309815e-06,
-      "loss": 0.9165,
-      "step": 18420
-    },
-    {
-      "epoch": 0.854195642095503,
-      "grad_norm": 3.3227803707122803,
-      "learning_rate": 1.2674495731878233e-06,
-      "loss": 0.7961,
-      "step": 18425
-    },
-    {
-      "epoch": 0.8544274455261938,
-      "grad_norm": 4.07461404800415,
-      "learning_rate": 1.2635097569044585e-06,
-      "loss": 0.7933,
-      "step": 18430
-    },
-    {
-      "epoch": 0.8546592489568846,
-      "grad_norm": 4.2566914558410645,
-      "learning_rate": 1.2595756605603604e-06,
-      "loss": 0.9242,
-      "step": 18435
-    },
-    {
-      "epoch": 0.8548910523875753,
-      "grad_norm": 3.8578896522521973,
-      "learning_rate": 1.2556472867312553e-06,
-      "loss": 0.8824,
-      "step": 18440
-    },
-    {
-      "epoch": 0.8551228558182661,
-      "grad_norm": 3.371788501739502,
-      "learning_rate": 1.251724637989118e-06,
-      "loss": 0.8345,
-      "step": 18445
-    },
-    {
-      "epoch": 0.8553546592489569,
-      "grad_norm": 3.8980839252471924,
-      "learning_rate": 1.2478077169021773e-06,
-      "loss": 0.8757,
-      "step": 18450
-    },
-    {
-      "epoch": 0.8555864626796477,
-      "grad_norm": 3.4333198070526123,
-      "learning_rate": 1.2438965260349134e-06,
-      "loss": 0.7938,
-      "step": 18455
-    },
-    {
-      "epoch": 0.8558182661103384,
-      "grad_norm": 3.738865852355957,
-      "learning_rate": 1.2399910679480576e-06,
-      "loss": 0.7182,
-      "step": 18460
-    },
-    {
-      "epoch": 0.8560500695410292,
-      "grad_norm": 4.250709533691406,
-      "learning_rate": 1.23609134519858e-06,
-      "loss": 0.8313,
-      "step": 18465
-    },
-    {
-      "epoch": 0.85628187297172,
-      "grad_norm": 5.0909743309021,
-      "learning_rate": 1.232197360339702e-06,
-      "loss": 1.0371,
-      "step": 18470
-    },
-    {
-      "epoch": 0.8565136764024107,
-      "grad_norm": 3.838352680206299,
-      "learning_rate": 1.2283091159208882e-06,
-      "loss": 0.9044,
-      "step": 18475
-    },
-    {
-      "epoch": 0.8567454798331016,
-      "grad_norm": 3.0013608932495117,
-      "learning_rate": 1.2244266144878415e-06,
-      "loss": 0.7732,
-      "step": 18480
-    },
-    {
-      "epoch": 0.8569772832637923,
-      "grad_norm": 3.4243435859680176,
-      "learning_rate": 1.2205498585825037e-06,
-      "loss": 0.6783,
-      "step": 18485
-    },
-    {
-      "epoch": 0.857209086694483,
-      "grad_norm": 3.4078502655029297,
-      "learning_rate": 1.2166788507430593e-06,
-      "loss": 0.7527,
-      "step": 18490
-    },
-    {
-      "epoch": 0.8574408901251739,
-      "grad_norm": 3.5167553424835205,
-      "learning_rate": 1.212813593503931e-06,
-      "loss": 0.8503,
-      "step": 18495
-    },
-    {
-      "epoch": 0.8576726935558646,
-      "grad_norm": 4.191205978393555,
-      "learning_rate": 1.2089540893957675e-06,
-      "loss": 0.7204,
-      "step": 18500
-    },
-    {
-      "epoch": 0.8576726935558646,
-      "eval_loss": 0.8762512803077698,
-      "eval_runtime": 11.2696,
-      "eval_samples_per_second": 11.269,
-      "eval_steps_per_second": 11.269,
-      "step": 18500
-    },
-    {
-      "epoch": 0.8579044969865554,
-      "grad_norm": 3.3614768981933594,
-      "learning_rate": 1.205100340945462e-06,
-      "loss": 0.6965,
-      "step": 18505
-    },
-    {
-      "epoch": 0.8581363004172462,
-      "grad_norm": 3.390779733657837,
-      "learning_rate": 1.2012523506761299e-06,
-      "loss": 0.768,
-      "step": 18510
-    },
-    {
-      "epoch": 0.8583681038479369,
-      "grad_norm": 4.055974960327148,
-      "learning_rate": 1.1974101211071197e-06,
-      "loss": 0.937,
-      "step": 18515
-    },
-    {
-      "epoch": 0.8585999072786277,
-      "grad_norm": 3.6975340843200684,
-      "learning_rate": 1.1935736547540122e-06,
-      "loss": 0.7743,
-      "step": 18520
-    },
-    {
-      "epoch": 0.8588317107093185,
-      "grad_norm": 4.171230792999268,
-      "learning_rate": 1.189742954128611e-06,
-      "loss": 0.9207,
-      "step": 18525
-    },
-    {
-      "epoch": 0.8590635141400093,
-      "grad_norm": 4.331693649291992,
-      "learning_rate": 1.1859180217389433e-06,
-      "loss": 0.821,
-      "step": 18530
-    },
-    {
-      "epoch": 0.8592953175707001,
-      "grad_norm": 4.08601713180542,
-      "learning_rate": 1.1820988600892669e-06,
-      "loss": 0.9393,
-      "step": 18535
-    },
-    {
-      "epoch": 0.8595271210013908,
-      "grad_norm": 3.40236234664917,
-      "learning_rate": 1.1782854716800496e-06,
-      "loss": 0.8371,
-      "step": 18540
-    },
-    {
-      "epoch": 0.8597589244320816,
-      "grad_norm": 3.1915442943573,
-      "learning_rate": 1.174477859007993e-06,
-      "loss": 0.8073,
-      "step": 18545
-    },
-    {
-      "epoch": 0.8599907278627724,
-      "grad_norm": 3.60099720954895,
-      "learning_rate": 1.1706760245660053e-06,
-      "loss": 0.8328,
-      "step": 18550
-    },
-    {
-      "epoch": 0.8602225312934632,
-      "grad_norm": 3.87766170501709,
-      "learning_rate": 1.16687997084322e-06,
-      "loss": 0.8903,
-      "step": 18555
-    },
-    {
-      "epoch": 0.8604543347241539,
-      "grad_norm": 4.071938991546631,
-      "learning_rate": 1.163089700324982e-06,
-      "loss": 0.815,
-      "step": 18560
-    },
-    {
-      "epoch": 0.8606861381548447,
-      "grad_norm": 4.054932117462158,
-      "learning_rate": 1.1593052154928508e-06,
-      "loss": 0.9528,
-      "step": 18565
-    },
-    {
-      "epoch": 0.8609179415855355,
-      "grad_norm": 3.4975640773773193,
-      "learning_rate": 1.1555265188245934e-06,
-      "loss": 0.8449,
-      "step": 18570
-    },
-    {
-      "epoch": 0.8611497450162262,
-      "grad_norm": 4.184132099151611,
-      "learning_rate": 1.1517536127941952e-06,
-      "loss": 0.9859,
-      "step": 18575
-    },
-    {
-      "epoch": 0.8613815484469171,
-      "grad_norm": 4.251339435577393,
-      "learning_rate": 1.1479864998718482e-06,
-      "loss": 0.9542,
-      "step": 18580
-    },
-    {
-      "epoch": 0.8616133518776078,
-      "grad_norm": 4.361698627471924,
-      "learning_rate": 1.1442251825239459e-06,
-      "loss": 0.9167,
-      "step": 18585
-    },
-    {
-      "epoch": 0.8618451553082985,
-      "grad_norm": 3.473511219024658,
-      "learning_rate": 1.140469663213094e-06,
-      "loss": 0.902,
-      "step": 18590
-    },
-    {
-      "epoch": 0.8620769587389894,
-      "grad_norm": 3.937403440475464,
-      "learning_rate": 1.1367199443980991e-06,
-      "loss": 0.9195,
-      "step": 18595
-    },
-    {
-      "epoch": 0.8623087621696801,
-      "grad_norm": 3.817412853240967,
-      "learning_rate": 1.1329760285339675e-06,
-      "loss": 0.7739,
-      "step": 18600
-    },
-    {
-      "epoch": 0.8623087621696801,
-      "eval_loss": 0.8760548233985901,
-      "eval_runtime": 11.2847,
-      "eval_samples_per_second": 11.254,
-      "eval_steps_per_second": 11.254,
-      "step": 18600
-    },
-    {
-      "epoch": 0.8625405656003708,
-      "grad_norm": 3.9234488010406494,
-      "learning_rate": 1.1292379180719104e-06,
-      "loss": 0.7825,
-      "step": 18605
-    },
-    {
-      "epoch": 0.8627723690310617,
-      "grad_norm": 3.362407684326172,
-      "learning_rate": 1.125505615459338e-06,
-      "loss": 0.5777,
-      "step": 18610
-    },
-    {
-      "epoch": 0.8630041724617524,
-      "grad_norm": 3.790205717086792,
-      "learning_rate": 1.1217791231398578e-06,
-      "loss": 0.8194,
-      "step": 18615
-    },
-    {
-      "epoch": 0.8632359758924432,
-      "grad_norm": 3.6676652431488037,
-      "learning_rate": 1.1180584435532716e-06,
-      "loss": 0.8417,
-      "step": 18620
-    },
-    {
-      "epoch": 0.863467779323134,
-      "grad_norm": 4.089963436126709,
-      "learning_rate": 1.1143435791355728e-06,
-      "loss": 0.9652,
-      "step": 18625
-    },
-    {
-      "epoch": 0.8636995827538247,
-      "grad_norm": 3.81626033782959,
-      "learning_rate": 1.1106345323189548e-06,
-      "loss": 0.8223,
-      "step": 18630
-    },
-    {
-      "epoch": 0.8639313861845155,
-      "grad_norm": 4.048824310302734,
-      "learning_rate": 1.1069313055317953e-06,
-      "loss": 0.9674,
-      "step": 18635
-    },
-    {
-      "epoch": 0.8641631896152063,
-      "grad_norm": 4.039358139038086,
-      "learning_rate": 1.103233901198666e-06,
-      "loss": 0.8702,
-      "step": 18640
-    },
-    {
-      "epoch": 0.8643949930458971,
-      "grad_norm": 3.317441463470459,
-      "learning_rate": 1.0995423217403245e-06,
-      "loss": 0.7957,
-      "step": 18645
-    },
-    {
-      "epoch": 0.8646267964765878,
-      "grad_norm": 4.214531421661377,
-      "learning_rate": 1.0958565695737167e-06,
-      "loss": 0.8281,
-      "step": 18650
-    },
-    {
-      "epoch": 0.8648585999072786,
-      "grad_norm": 3.687647581100464,
-      "learning_rate": 1.0921766471119665e-06,
-      "loss": 0.877,
-      "step": 18655
-    },
-    {
-      "epoch": 0.8650904033379694,
-      "grad_norm": 3.635422945022583,
-      "learning_rate": 1.0885025567643913e-06,
-      "loss": 0.8062,
-      "step": 18660
-    },
-    {
-      "epoch": 0.8653222067686602,
-      "grad_norm": 3.2454099655151367,
-      "learning_rate": 1.0848343009364837e-06,
-      "loss": 0.7843,
-      "step": 18665
-    },
-    {
-      "epoch": 0.865554010199351,
-      "grad_norm": 4.3716139793396,
-      "learning_rate": 1.0811718820299155e-06,
-      "loss": 0.7872,
-      "step": 18670
-    },
-    {
-      "epoch": 0.8657858136300417,
-      "grad_norm": 4.972971439361572,
-      "learning_rate": 1.0775153024425411e-06,
-      "loss": 1.0013,
-      "step": 18675
-    },
-    {
-      "epoch": 0.8660176170607325,
-      "grad_norm": 3.9282686710357666,
-      "learning_rate": 1.0738645645683921e-06,
-      "loss": 0.8474,
-      "step": 18680
-    },
-    {
-      "epoch": 0.8662494204914233,
-      "grad_norm": 4.049283027648926,
-      "learning_rate": 1.0702196707976686e-06,
-      "loss": 0.8647,
-      "step": 18685
-    },
-    {
-      "epoch": 0.866481223922114,
-      "grad_norm": 3.5909154415130615,
-      "learning_rate": 1.0665806235167498e-06,
-      "loss": 0.8355,
-      "step": 18690
-    },
-    {
-      "epoch": 0.8667130273528049,
-      "grad_norm": 4.021440029144287,
-      "learning_rate": 1.0629474251081873e-06,
-      "loss": 0.8835,
-      "step": 18695
-    },
-    {
-      "epoch": 0.8669448307834956,
-      "grad_norm": 3.9168529510498047,
-      "learning_rate": 1.0593200779507028e-06,
-      "loss": 0.801,
-      "step": 18700
-    },
-    {
-      "epoch": 0.8669448307834956,
-      "eval_loss": 0.8755388855934143,
-      "eval_runtime": 11.2609,
-      "eval_samples_per_second": 11.278,
-      "eval_steps_per_second": 11.278,
-      "step": 18700
-    },
-    {
-      "epoch": 0.8671766342141863,
-      "grad_norm": 4.553606986999512,
-      "learning_rate": 1.0556985844191848e-06,
-      "loss": 0.8309,
-      "step": 18705
-    },
-    {
-      "epoch": 0.8674084376448772,
-      "grad_norm": 4.2659478187561035,
-      "learning_rate": 1.0520829468846938e-06,
-      "loss": 0.9314,
-      "step": 18710
-    },
-    {
-      "epoch": 0.8676402410755679,
-      "grad_norm": 3.4218521118164062,
-      "learning_rate": 1.0484731677144521e-06,
-      "loss": 0.8183,
-      "step": 18715
-    },
-    {
-      "epoch": 0.8678720445062587,
-      "grad_norm": 4.528285026550293,
-      "learning_rate": 1.044869249271845e-06,
-      "loss": 0.8522,
-      "step": 18720
-    },
-    {
-      "epoch": 0.8681038479369495,
-      "grad_norm": 3.9115729331970215,
-      "learning_rate": 1.0412711939164266e-06,
-      "loss": 0.7733,
-      "step": 18725
-    },
-    {
-      "epoch": 0.8683356513676402,
-      "grad_norm": 3.4151837825775146,
-      "learning_rate": 1.0376790040039098e-06,
-      "loss": 0.9614,
-      "step": 18730
-    },
-    {
-      "epoch": 0.868567454798331,
-      "grad_norm": 4.008971214294434,
-      "learning_rate": 1.0340926818861642e-06,
-      "loss": 0.8216,
-      "step": 18735
-    },
-    {
-      "epoch": 0.8687992582290218,
-      "grad_norm": 3.121033191680908,
-      "learning_rate": 1.0305122299112246e-06,
-      "loss": 0.8541,
-      "step": 18740
-    },
-    {
-      "epoch": 0.8690310616597126,
-      "grad_norm": 3.4723434448242188,
-      "learning_rate": 1.026937650423272e-06,
-      "loss": 0.7798,
-      "step": 18745
-    },
-    {
-      "epoch": 0.8692628650904033,
-      "grad_norm": 4.1530961990356445,
-      "learning_rate": 1.023368945762655e-06,
-      "loss": 0.9794,
-      "step": 18750
-    },
-    {
-      "epoch": 0.8694946685210941,
-      "grad_norm": 4.090208053588867,
-      "learning_rate": 1.0198061182658637e-06,
-      "loss": 0.7676,
-      "step": 18755
-    },
-    {
-      "epoch": 0.8697264719517849,
-      "grad_norm": 3.8618550300598145,
-      "learning_rate": 1.0162491702655486e-06,
-      "loss": 0.7873,
-      "step": 18760
-    },
-    {
-      "epoch": 0.8699582753824756,
-      "grad_norm": 3.4474642276763916,
-      "learning_rate": 1.0126981040905116e-06,
-      "loss": 0.9073,
-      "step": 18765
-    },
-    {
-      "epoch": 0.8701900788131665,
-      "grad_norm": 3.3244881629943848,
-      "learning_rate": 1.009152922065697e-06,
-      "loss": 0.8019,
-      "step": 18770
-    },
-    {
-      "epoch": 0.8704218822438572,
-      "grad_norm": 3.623584747314453,
-      "learning_rate": 1.0056136265121985e-06,
-      "loss": 0.8823,
-      "step": 18775
-    },
-    {
-      "epoch": 0.8706536856745479,
-      "grad_norm": 3.2864067554473877,
-      "learning_rate": 1.002080219747259e-06,
-      "loss": 0.7796,
-      "step": 18780
-    },
-    {
-      "epoch": 0.8708854891052388,
-      "grad_norm": 3.941572666168213,
-      "learning_rate": 9.985527040842669e-07,
-      "loss": 0.7145,
-      "step": 18785
-    },
-    {
-      "epoch": 0.8711172925359295,
-      "grad_norm": 4.011714935302734,
-      "learning_rate": 9.950310818327469e-07,
-      "loss": 0.8252,
-      "step": 18790
-    },
-    {
-      "epoch": 0.8713490959666204,
-      "grad_norm": 3.8345000743865967,
-      "learning_rate": 9.915153552983725e-07,
-      "loss": 0.9874,
-      "step": 18795
-    },
-    {
-      "epoch": 0.8715808993973111,
-      "grad_norm": 3.2887487411499023,
-      "learning_rate": 9.880055267829524e-07,
-      "loss": 0.8881,
-      "step": 18800
-    },
-    {
-      "epoch": 0.8715808993973111,
-      "eval_loss": 0.8754202723503113,
-      "eval_runtime": 11.2653,
-      "eval_samples_per_second": 11.274,
-      "eval_steps_per_second": 11.274,
-      "step": 18800
-    },
-    {
-      "epoch": 0.8718127028280018,
-      "grad_norm": 4.1776652336120605,
-      "learning_rate": 9.84501598584433e-07,
-      "loss": 0.9445,
-      "step": 18805
-    },
-    {
-      "epoch": 0.8720445062586927,
-      "grad_norm": 4.137359619140625,
-      "learning_rate": 9.810035729969026e-07,
-      "loss": 0.8771,
-      "step": 18810
-    },
-    {
-      "epoch": 0.8722763096893834,
-      "grad_norm": 4.066986083984375,
-      "learning_rate": 9.775114523105833e-07,
-      "loss": 1.0272,
-      "step": 18815
-    },
-    {
-      "epoch": 0.8725081131200741,
-      "grad_norm": 3.746918201446533,
-      "learning_rate": 9.740252388118266e-07,
-      "loss": 0.7733,
-      "step": 18820
-    },
-    {
-      "epoch": 0.872739916550765,
-      "grad_norm": 3.638258457183838,
-      "learning_rate": 9.70544934783123e-07,
-      "loss": 0.7531,
-      "step": 18825
-    },
-    {
-      "epoch": 0.8729717199814557,
-      "grad_norm": 3.6451056003570557,
-      "learning_rate": 9.67070542503089e-07,
-      "loss": 0.9688,
-      "step": 18830
-    },
-    {
-      "epoch": 0.8732035234121465,
-      "grad_norm": 3.232944965362549,
-      "learning_rate": 9.63602064246475e-07,
-      "loss": 0.9892,
-      "step": 18835
-    },
-    {
-      "epoch": 0.8734353268428373,
-      "grad_norm": 3.4214277267456055,
-      "learning_rate": 9.601395022841542e-07,
-      "loss": 0.8214,
-      "step": 18840
-    },
-    {
-      "epoch": 0.873667130273528,
-      "grad_norm": 4.174017906188965,
-      "learning_rate": 9.56682858883129e-07,
-      "loss": 0.8695,
-      "step": 18845
-    },
-    {
-      "epoch": 0.8738989337042188,
-      "grad_norm": 3.725480556488037,
-      "learning_rate": 9.5323213630653e-07,
-      "loss": 0.815,
-      "step": 18850
-    },
-    {
-      "epoch": 0.8741307371349096,
-      "grad_norm": 3.5639865398406982,
-      "learning_rate": 9.497873368136056e-07,
-      "loss": 0.7668,
-      "step": 18855
-    },
-    {
-      "epoch": 0.8743625405656004,
-      "grad_norm": 4.303888320922852,
-      "learning_rate": 9.463484626597275e-07,
-      "loss": 0.9614,
-      "step": 18860
-    },
-    {
-      "epoch": 0.8745943439962911,
-      "grad_norm": 3.6800920963287354,
-      "learning_rate": 9.429155160963899e-07,
-      "loss": 0.7031,
-      "step": 18865
-    },
-    {
-      "epoch": 0.874826147426982,
-      "grad_norm": 3.8146042823791504,
-      "learning_rate": 9.394884993712095e-07,
-      "loss": 0.8,
-      "step": 18870
-    },
-    {
-      "epoch": 0.8750579508576727,
-      "grad_norm": 4.3860368728637695,
-      "learning_rate": 9.360674147279103e-07,
-      "loss": 0.9592,
-      "step": 18875
-    },
-    {
-      "epoch": 0.8752897542883634,
-      "grad_norm": 3.643028736114502,
-      "learning_rate": 9.326522644063429e-07,
-      "loss": 0.7462,
-      "step": 18880
-    },
-    {
-      "epoch": 0.8755215577190543,
-      "grad_norm": 3.8522374629974365,
-      "learning_rate": 9.292430506424688e-07,
-      "loss": 0.8639,
-      "step": 18885
-    },
-    {
-      "epoch": 0.875753361149745,
-      "grad_norm": 4.47439432144165,
-      "learning_rate": 9.258397756683601e-07,
-      "loss": 0.7049,
-      "step": 18890
-    },
-    {
-      "epoch": 0.8759851645804357,
-      "grad_norm": 3.729813814163208,
-      "learning_rate": 9.224424417122036e-07,
-      "loss": 0.7833,
-      "step": 18895
-    },
-    {
-      "epoch": 0.8762169680111266,
-      "grad_norm": 3.2106189727783203,
-      "learning_rate": 9.190510509982942e-07,
-      "loss": 0.8653,
-      "step": 18900
-    },
-    {
-      "epoch": 0.8762169680111266,
-      "eval_loss": 0.8750000596046448,
-      "eval_runtime": 11.2681,
-      "eval_samples_per_second": 11.271,
-      "eval_steps_per_second": 11.271,
-      "step": 18900
-    },
-    {
-      "epoch": 0.8764487714418173,
-      "grad_norm": 3.6238574981689453,
-      "learning_rate": 9.156656057470415e-07,
-      "loss": 0.8184,
-      "step": 18905
-    },
-    {
-      "epoch": 0.8766805748725082,
-      "grad_norm": 4.165507793426514,
-      "learning_rate": 9.122861081749524e-07,
-      "loss": 0.802,
-      "step": 18910
-    },
-    {
-      "epoch": 0.8769123783031989,
-      "grad_norm": 3.3286075592041016,
-      "learning_rate": 9.089125604946514e-07,
-      "loss": 0.7304,
-      "step": 18915
-    },
-    {
-      "epoch": 0.8771441817338896,
-      "grad_norm": 3.1397790908813477,
-      "learning_rate": 9.055449649148585e-07,
-      "loss": 0.7384,
-      "step": 18920
-    },
-    {
-      "epoch": 0.8773759851645805,
-      "grad_norm": 3.556872844696045,
-      "learning_rate": 9.021833236403977e-07,
-      "loss": 0.7844,
-      "step": 18925
-    },
-    {
-      "epoch": 0.8776077885952712,
-      "grad_norm": 3.5317461490631104,
-      "learning_rate": 8.988276388721995e-07,
-      "loss": 0.9778,
-      "step": 18930
-    },
-    {
-      "epoch": 0.877839592025962,
-      "grad_norm": 3.8792641162872314,
-      "learning_rate": 8.954779128072932e-07,
-      "loss": 0.8529,
-      "step": 18935
-    },
-    {
-      "epoch": 0.8780713954566528,
-      "grad_norm": 3.7412779331207275,
-      "learning_rate": 8.921341476388024e-07,
-      "loss": 0.8977,
-      "step": 18940
-    },
-    {
-      "epoch": 0.8783031988873435,
-      "grad_norm": 3.8396799564361572,
-      "learning_rate": 8.887963455559534e-07,
-      "loss": 0.7698,
-      "step": 18945
-    },
-    {
-      "epoch": 0.8785350023180343,
-      "grad_norm": 3.3994853496551514,
-      "learning_rate": 8.854645087440628e-07,
-      "loss": 0.973,
-      "step": 18950
-    },
-    {
-      "epoch": 0.8787668057487251,
-      "grad_norm": 3.9767489433288574,
-      "learning_rate": 8.8213863938455e-07,
-      "loss": 1.01,
-      "step": 18955
-    },
-    {
-      "epoch": 0.8789986091794159,
-      "grad_norm": 3.4473788738250732,
-      "learning_rate": 8.788187396549164e-07,
-      "loss": 0.7616,
-      "step": 18960
-    },
-    {
-      "epoch": 0.8792304126101066,
-      "grad_norm": 3.0410356521606445,
-      "learning_rate": 8.755048117287646e-07,
-      "loss": 0.8133,
-      "step": 18965
-    },
-    {
-      "epoch": 0.8794622160407974,
-      "grad_norm": 4.893867015838623,
-      "learning_rate": 8.721968577757833e-07,
-      "loss": 0.8119,
-      "step": 18970
-    },
-    {
-      "epoch": 0.8796940194714882,
-      "grad_norm": 3.6503591537475586,
-      "learning_rate": 8.688948799617502e-07,
-      "loss": 0.8597,
-      "step": 18975
-    },
-    {
-      "epoch": 0.8799258229021789,
-      "grad_norm": 3.970219612121582,
-      "learning_rate": 8.655988804485283e-07,
-      "loss": 0.8174,
-      "step": 18980
-    },
-    {
-      "epoch": 0.8801576263328698,
-      "grad_norm": 3.8891892433166504,
-      "learning_rate": 8.6230886139407e-07,
-      "loss": 0.8486,
-      "step": 18985
-    },
-    {
-      "epoch": 0.8803894297635605,
-      "grad_norm": 3.3773927688598633,
-      "learning_rate": 8.590248249524114e-07,
-      "loss": 0.9104,
-      "step": 18990
-    },
-    {
-      "epoch": 0.8806212331942512,
-      "grad_norm": 4.802641868591309,
-      "learning_rate": 8.557467732736679e-07,
-      "loss": 0.797,
-      "step": 18995
-    },
-    {
-      "epoch": 0.8808530366249421,
-      "grad_norm": 3.5772879123687744,
-      "learning_rate": 8.52474708504043e-07,
-      "loss": 0.8144,
-      "step": 19000
-    },
-    {
-      "epoch": 0.8808530366249421,
-      "eval_loss": 0.8745938539505005,
-      "eval_runtime": 11.2692,
-      "eval_samples_per_second": 11.27,
-      "eval_steps_per_second": 11.27,
-      "step": 19000
-    },
-    {
-      "epoch": 0.8810848400556328,
-      "grad_norm": 3.7284469604492188,
-      "learning_rate": 8.492086327858184e-07,
-      "loss": 0.905,
-      "step": 19005
-    },
-    {
-      "epoch": 0.8813166434863235,
-      "grad_norm": 3.670050621032715,
-      "learning_rate": 8.459485482573481e-07,
-      "loss": 0.8371,
-      "step": 19010
-    },
-    {
-      "epoch": 0.8815484469170144,
-      "grad_norm": 3.598022222518921,
-      "learning_rate": 8.426944570530715e-07,
-      "loss": 0.8005,
-      "step": 19015
-    },
-    {
-      "epoch": 0.8817802503477051,
-      "grad_norm": 4.089553356170654,
-      "learning_rate": 8.394463613035031e-07,
-      "loss": 0.9309,
-      "step": 19020
-    },
-    {
-      "epoch": 0.8820120537783959,
-      "grad_norm": 3.748443365097046,
-      "learning_rate": 8.362042631352274e-07,
-      "loss": 0.9283,
-      "step": 19025
-    },
-    {
-      "epoch": 0.8822438572090867,
-      "grad_norm": 3.325829267501831,
-      "learning_rate": 8.329681646709076e-07,
-      "loss": 0.8398,
-      "step": 19030
-    },
-    {
-      "epoch": 0.8824756606397774,
-      "grad_norm": 4.052063465118408,
-      "learning_rate": 8.297380680292732e-07,
-      "loss": 0.8329,
-      "step": 19035
-    },
-    {
-      "epoch": 0.8827074640704683,
-      "grad_norm": 4.558441162109375,
-      "learning_rate": 8.265139753251306e-07,
-      "loss": 0.9717,
-      "step": 19040
-    },
-    {
-      "epoch": 0.882939267501159,
-      "grad_norm": 3.689945697784424,
-      "learning_rate": 8.232958886693498e-07,
-      "loss": 0.7946,
-      "step": 19045
-    },
-    {
-      "epoch": 0.8831710709318498,
-      "grad_norm": 3.7369065284729004,
-      "learning_rate": 8.200838101688713e-07,
-      "loss": 1.0788,
-      "step": 19050
-    },
-    {
-      "epoch": 0.8834028743625406,
-      "grad_norm": 4.071127891540527,
-      "learning_rate": 8.168777419267026e-07,
-      "loss": 0.9555,
-      "step": 19055
-    },
-    {
-      "epoch": 0.8836346777932313,
-      "grad_norm": 3.3217930793762207,
-      "learning_rate": 8.136776860419149e-07,
-      "loss": 0.8766,
-      "step": 19060
-    },
-    {
-      "epoch": 0.8838664812239221,
-      "grad_norm": 3.262667179107666,
-      "learning_rate": 8.104836446096409e-07,
-      "loss": 0.7637,
-      "step": 19065
-    },
-    {
-      "epoch": 0.8840982846546129,
-      "grad_norm": 3.5461394786834717,
-      "learning_rate": 8.07295619721078e-07,
-      "loss": 0.8512,
-      "step": 19070
-    },
-    {
-      "epoch": 0.8843300880853037,
-      "grad_norm": 3.659705638885498,
-      "learning_rate": 8.041136134634864e-07,
-      "loss": 0.9289,
-      "step": 19075
-    },
-    {
-      "epoch": 0.8845618915159944,
-      "grad_norm": 3.627200126647949,
-      "learning_rate": 8.009376279201819e-07,
-      "loss": 0.8783,
-      "step": 19080
-    },
-    {
-      "epoch": 0.8847936949466852,
-      "grad_norm": 3.535076856613159,
-      "learning_rate": 7.977676651705391e-07,
-      "loss": 0.8446,
-      "step": 19085
-    },
-    {
-      "epoch": 0.885025498377376,
-      "grad_norm": 3.2673180103302,
-      "learning_rate": 7.946037272899965e-07,
-      "loss": 0.9139,
-      "step": 19090
-    },
-    {
-      "epoch": 0.8852573018080667,
-      "grad_norm": 3.439281702041626,
-      "learning_rate": 7.914458163500327e-07,
-      "loss": 0.8172,
-      "step": 19095
-    },
-    {
-      "epoch": 0.8854891052387576,
-      "grad_norm": 3.935387134552002,
-      "learning_rate": 7.882939344181939e-07,
-      "loss": 0.7661,
-      "step": 19100
-    },
-    {
-      "epoch": 0.8854891052387576,
-      "eval_loss": 0.8745239973068237,
-      "eval_runtime": 11.2651,
-      "eval_samples_per_second": 11.274,
-      "eval_steps_per_second": 11.274,
-      "step": 19100
-    },
-    {
-      "epoch": 0.8857209086694483,
-      "grad_norm": 4.236382484436035,
-      "learning_rate": 7.851480835580749e-07,
-      "loss": 0.9492,
-      "step": 19105
-    },
-    {
-      "epoch": 0.885952712100139,
-      "grad_norm": 3.854919672012329,
-      "learning_rate": 7.820082658293227e-07,
-      "loss": 0.8794,
-      "step": 19110
-    },
-    {
-      "epoch": 0.8861845155308299,
-      "grad_norm": 4.203670024871826,
-      "learning_rate": 7.788744832876294e-07,
-      "loss": 0.9396,
-      "step": 19115
-    },
-    {
-      "epoch": 0.8864163189615206,
-      "grad_norm": 3.6348228454589844,
-      "learning_rate": 7.757467379847438e-07,
-      "loss": 0.8706,
-      "step": 19120
-    },
-    {
-      "epoch": 0.8866481223922114,
-      "grad_norm": 3.4574408531188965,
-      "learning_rate": 7.726250319684558e-07,
-      "loss": 0.8792,
-      "step": 19125
-    },
-    {
-      "epoch": 0.8868799258229022,
-      "grad_norm": 3.5718040466308594,
-      "learning_rate": 7.695093672826015e-07,
-      "loss": 0.7707,
-      "step": 19130
-    },
-    {
-      "epoch": 0.8871117292535929,
-      "grad_norm": 3.444611072540283,
-      "learning_rate": 7.663997459670647e-07,
-      "loss": 0.8391,
-      "step": 19135
-    },
-    {
-      "epoch": 0.8873435326842837,
-      "grad_norm": 4.593543529510498,
-      "learning_rate": 7.632961700577723e-07,
-      "loss": 0.882,
-      "step": 19140
-    },
-    {
-      "epoch": 0.8875753361149745,
-      "grad_norm": 4.171180248260498,
-      "learning_rate": 7.601986415866903e-07,
-      "loss": 0.9814,
-      "step": 19145
-    },
-    {
-      "epoch": 0.8878071395456653,
-      "grad_norm": 3.3400700092315674,
-      "learning_rate": 7.571071625818283e-07,
-      "loss": 0.592,
-      "step": 19150
-    },
-    {
-      "epoch": 0.8880389429763561,
-      "grad_norm": 3.698866128921509,
-      "learning_rate": 7.540217350672318e-07,
-      "loss": 0.9165,
-      "step": 19155
-    },
-    {
-      "epoch": 0.8882707464070468,
-      "grad_norm": 4.187894821166992,
-      "learning_rate": 7.509423610629896e-07,
-      "loss": 0.8477,
-      "step": 19160
-    },
-    {
-      "epoch": 0.8885025498377376,
-      "grad_norm": 3.720407724380493,
-      "learning_rate": 7.478690425852197e-07,
-      "loss": 0.9191,
-      "step": 19165
-    },
-    {
-      "epoch": 0.8887343532684284,
-      "grad_norm": 4.027552604675293,
-      "learning_rate": 7.448017816460829e-07,
-      "loss": 0.8655,
-      "step": 19170
-    },
-    {
-      "epoch": 0.8889661566991192,
-      "grad_norm": 3.9079010486602783,
-      "learning_rate": 7.417405802537725e-07,
-      "loss": 0.7091,
-      "step": 19175
-    },
-    {
-      "epoch": 0.8891979601298099,
-      "grad_norm": 3.516188144683838,
-      "learning_rate": 7.386854404125088e-07,
-      "loss": 0.7183,
-      "step": 19180
-    },
-    {
-      "epoch": 0.8894297635605007,
-      "grad_norm": 2.9960403442382812,
-      "learning_rate": 7.356363641225484e-07,
-      "loss": 0.7832,
-      "step": 19185
-    },
-    {
-      "epoch": 0.8896615669911915,
-      "grad_norm": 4.731221675872803,
-      "learning_rate": 7.325933533801766e-07,
-      "loss": 0.8783,
-      "step": 19190
-    },
-    {
-      "epoch": 0.8898933704218822,
-      "grad_norm": 3.7551865577697754,
-      "learning_rate": 7.29556410177712e-07,
-      "loss": 0.9412,
-      "step": 19195
-    },
-    {
-      "epoch": 0.8901251738525731,
-      "grad_norm": 3.833738327026367,
-      "learning_rate": 7.265255365034907e-07,
-      "loss": 0.7348,
-      "step": 19200
-    },
-    {
-      "epoch": 0.8901251738525731,
-      "eval_loss": 0.8742764592170715,
-      "eval_runtime": 11.2535,
-      "eval_samples_per_second": 11.285,
-      "eval_steps_per_second": 11.285,
-      "step": 19200
-    },
-    {
-      "epoch": 0.8903569772832638,
-      "grad_norm": 3.7898571491241455,
-      "learning_rate": 7.235007343418843e-07,
-      "loss": 0.8018,
-      "step": 19205
-    },
-    {
-      "epoch": 0.8905887807139545,
-      "grad_norm": 3.9264822006225586,
-      "learning_rate": 7.204820056732897e-07,
-      "loss": 0.9964,
-      "step": 19210
-    },
-    {
-      "epoch": 0.8908205841446454,
-      "grad_norm": 3.960806369781494,
-      "learning_rate": 7.17469352474115e-07,
-      "loss": 0.7303,
-      "step": 19215
-    },
-    {
-      "epoch": 0.8910523875753361,
-      "grad_norm": 3.6674251556396484,
-      "learning_rate": 7.144627767168033e-07,
-      "loss": 1.031,
-      "step": 19220
-    },
-    {
-      "epoch": 0.8912841910060268,
-      "grad_norm": 3.619192123413086,
-      "learning_rate": 7.114622803698157e-07,
-      "loss": 0.8463,
-      "step": 19225
-    },
-    {
-      "epoch": 0.8915159944367177,
-      "grad_norm": 3.470325469970703,
-      "learning_rate": 7.084678653976295e-07,
-      "loss": 0.8261,
-      "step": 19230
-    },
-    {
-      "epoch": 0.8917477978674084,
-      "grad_norm": 4.503883361816406,
-      "learning_rate": 7.054795337607445e-07,
-      "loss": 0.9888,
-      "step": 19235
-    },
-    {
-      "epoch": 0.8919796012980992,
-      "grad_norm": 3.7686662673950195,
-      "learning_rate": 7.024972874156766e-07,
-      "loss": 0.9952,
-      "step": 19240
-    },
-    {
-      "epoch": 0.89221140472879,
-      "grad_norm": 2.989572286605835,
-      "learning_rate": 6.99521128314955e-07,
-      "loss": 0.821,
-      "step": 19245
-    },
-    {
-      "epoch": 0.8924432081594807,
-      "grad_norm": 3.647934913635254,
-      "learning_rate": 6.965510584071234e-07,
-      "loss": 1.0117,
-      "step": 19250
-    },
-    {
-      "epoch": 0.8926750115901715,
-      "grad_norm": 4.123228549957275,
-      "learning_rate": 6.935870796367438e-07,
-      "loss": 0.811,
-      "step": 19255
-    },
-    {
-      "epoch": 0.8929068150208623,
-      "grad_norm": 3.645017385482788,
-      "learning_rate": 6.906291939443877e-07,
-      "loss": 0.8398,
-      "step": 19260
-    },
-    {
-      "epoch": 0.8931386184515531,
-      "grad_norm": 3.1021242141723633,
-      "learning_rate": 6.876774032666356e-07,
-      "loss": 0.6563,
-      "step": 19265
-    },
-    {
-      "epoch": 0.8933704218822438,
-      "grad_norm": 3.449209451675415,
-      "learning_rate": 6.847317095360762e-07,
-      "loss": 0.6527,
-      "step": 19270
-    },
-    {
-      "epoch": 0.8936022253129347,
-      "grad_norm": 3.8125498294830322,
-      "learning_rate": 6.817921146813111e-07,
-      "loss": 0.8687,
-      "step": 19275
-    },
-    {
-      "epoch": 0.8938340287436254,
-      "grad_norm": 4.294896125793457,
-      "learning_rate": 6.788586206269476e-07,
-      "loss": 0.7868,
-      "step": 19280
-    },
-    {
-      "epoch": 0.8940658321743162,
-      "grad_norm": 4.346077919006348,
-      "learning_rate": 6.759312292935949e-07,
-      "loss": 0.9154,
-      "step": 19285
-    },
-    {
-      "epoch": 0.894297635605007,
-      "grad_norm": 3.7401599884033203,
-      "learning_rate": 6.730099425978709e-07,
-      "loss": 0.9543,
-      "step": 19290
-    },
-    {
-      "epoch": 0.8945294390356977,
-      "grad_norm": 3.5999746322631836,
-      "learning_rate": 6.70094762452399e-07,
-      "loss": 0.7718,
-      "step": 19295
-    },
-    {
-      "epoch": 0.8947612424663886,
-      "grad_norm": 3.3646798133850098,
-      "learning_rate": 6.671856907657936e-07,
-      "loss": 0.9172,
-      "step": 19300
-    },
-    {
-      "epoch": 0.8947612424663886,
-      "eval_loss": 0.8741331100463867,
-      "eval_runtime": 11.2872,
-      "eval_samples_per_second": 11.252,
-      "eval_steps_per_second": 11.252,
-      "step": 19300
-    },
-    {
-      "epoch": 0.8949930458970793,
-      "grad_norm": 3.7011406421661377,
-      "learning_rate": 6.642827294426801e-07,
-      "loss": 0.7227,
-      "step": 19305
-    },
-    {
-      "epoch": 0.89522484932777,
-      "grad_norm": 3.1988883018493652,
-      "learning_rate": 6.613858803836814e-07,
-      "loss": 0.6719,
-      "step": 19310
-    },
-    {
-      "epoch": 0.8954566527584609,
-      "grad_norm": 3.0587880611419678,
-      "learning_rate": 6.58495145485416e-07,
-      "loss": 0.7688,
-      "step": 19315
-    },
-    {
-      "epoch": 0.8956884561891516,
-      "grad_norm": 4.00978422164917,
-      "learning_rate": 6.556105266404999e-07,
-      "loss": 0.7579,
-      "step": 19320
-    },
-    {
-      "epoch": 0.8959202596198423,
-      "grad_norm": 4.0303802490234375,
-      "learning_rate": 6.527320257375491e-07,
-      "loss": 0.8285,
-      "step": 19325
-    },
-    {
-      "epoch": 0.8961520630505332,
-      "grad_norm": 3.5370168685913086,
-      "learning_rate": 6.498596446611671e-07,
-      "loss": 1.005,
-      "step": 19330
-    },
-    {
-      "epoch": 0.8963838664812239,
-      "grad_norm": 3.032008647918701,
-      "learning_rate": 6.469933852919541e-07,
-      "loss": 0.7647,
-      "step": 19335
-    },
-    {
-      "epoch": 0.8966156699119147,
-      "grad_norm": 3.647174596786499,
-      "learning_rate": 6.441332495065034e-07,
-      "loss": 0.8656,
-      "step": 19340
-    },
-    {
-      "epoch": 0.8968474733426055,
-      "grad_norm": 3.3170273303985596,
-      "learning_rate": 6.412792391773992e-07,
-      "loss": 1.0113,
-      "step": 19345
-    },
-    {
-      "epoch": 0.8970792767732962,
-      "grad_norm": 3.2985119819641113,
-      "learning_rate": 6.384313561732114e-07,
-      "loss": 0.9544,
-      "step": 19350
-    },
-    {
-      "epoch": 0.897311080203987,
-      "grad_norm": 3.6995089054107666,
-      "learning_rate": 6.35589602358504e-07,
-      "loss": 1.0398,
-      "step": 19355
-    },
-    {
-      "epoch": 0.8975428836346778,
-      "grad_norm": 3.910785436630249,
-      "learning_rate": 6.32753979593821e-07,
-      "loss": 0.9002,
-      "step": 19360
-    },
-    {
-      "epoch": 0.8977746870653686,
-      "grad_norm": 3.369471788406372,
-      "learning_rate": 6.299244897357004e-07,
-      "loss": 0.8634,
-      "step": 19365
-    },
-    {
-      "epoch": 0.8980064904960593,
-      "grad_norm": 4.074563980102539,
-      "learning_rate": 6.271011346366574e-07,
-      "loss": 0.9008,
-      "step": 19370
-    },
-    {
-      "epoch": 0.8982382939267501,
-      "grad_norm": 3.610403299331665,
-      "learning_rate": 6.242839161451952e-07,
-      "loss": 0.8019,
-      "step": 19375
-    },
-    {
-      "epoch": 0.8984700973574409,
-      "grad_norm": 3.7776875495910645,
-      "learning_rate": 6.214728361057997e-07,
-      "loss": 0.9037,
-      "step": 19380
-    },
-    {
-      "epoch": 0.8987019007881316,
-      "grad_norm": 4.211784839630127,
-      "learning_rate": 6.186678963589354e-07,
-      "loss": 0.7532,
-      "step": 19385
-    },
-    {
-      "epoch": 0.8989337042188225,
-      "grad_norm": 3.3378918170928955,
-      "learning_rate": 6.158690987410443e-07,
-      "loss": 0.8697,
-      "step": 19390
-    },
-    {
-      "epoch": 0.8991655076495132,
-      "grad_norm": 3.1902554035186768,
-      "learning_rate": 6.130764450845539e-07,
-      "loss": 0.7719,
-      "step": 19395
-    },
-    {
-      "epoch": 0.8993973110802039,
-      "grad_norm": 4.117619514465332,
-      "learning_rate": 6.102899372178639e-07,
-      "loss": 1.0541,
-      "step": 19400
-    },
-    {
-      "epoch": 0.8993973110802039,
-      "eval_loss": 0.8739257454872131,
-      "eval_runtime": 11.2619,
-      "eval_samples_per_second": 11.277,
-      "eval_steps_per_second": 11.277,
-      "step": 19400
-    },
-    {
-      "epoch": 0.8996291145108948,
-      "grad_norm": 3.4233851432800293,
-      "learning_rate": 6.075095769653516e-07,
-      "loss": 0.8865,
-      "step": 19405
-    },
-    {
-      "epoch": 0.8998609179415855,
-      "grad_norm": 3.4738340377807617,
-      "learning_rate": 6.047353661473698e-07,
-      "loss": 0.7926,
-      "step": 19410
-    },
-    {
-      "epoch": 0.9000927213722764,
-      "grad_norm": 3.7461628913879395,
-      "learning_rate": 6.019673065802479e-07,
-      "loss": 0.8197,
-      "step": 19415
-    },
-    {
-      "epoch": 0.9003245248029671,
-      "grad_norm": 3.637523889541626,
-      "learning_rate": 5.992054000762793e-07,
-      "loss": 0.8103,
-      "step": 19420
-    },
-    {
-      "epoch": 0.9005563282336578,
-      "grad_norm": 4.224597930908203,
-      "learning_rate": 5.964496484437365e-07,
-      "loss": 0.8117,
-      "step": 19425
-    },
-    {
-      "epoch": 0.9007881316643487,
-      "grad_norm": 3.7092161178588867,
-      "learning_rate": 5.937000534868631e-07,
-      "loss": 0.8398,
-      "step": 19430
-    },
-    {
-      "epoch": 0.9010199350950394,
-      "grad_norm": 3.013432502746582,
-      "learning_rate": 5.909566170058644e-07,
-      "loss": 0.8795,
-      "step": 19435
-    },
-    {
-      "epoch": 0.9012517385257302,
-      "grad_norm": 3.9475176334381104,
-      "learning_rate": 5.882193407969216e-07,
-      "loss": 0.7751,
-      "step": 19440
-    },
-    {
-      "epoch": 0.901483541956421,
-      "grad_norm": 3.6993296146392822,
-      "learning_rate": 5.854882266521788e-07,
-      "loss": 0.8293,
-      "step": 19445
-    },
-    {
-      "epoch": 0.9017153453871117,
-      "grad_norm": 3.6457648277282715,
-      "learning_rate": 5.827632763597468e-07,
-      "loss": 0.8728,
-      "step": 19450
-    },
-    {
-      "epoch": 0.9019471488178025,
-      "grad_norm": 3.7621054649353027,
-      "learning_rate": 5.800444917036962e-07,
-      "loss": 0.8707,
-      "step": 19455
-    },
-    {
-      "epoch": 0.9021789522484933,
-      "grad_norm": 4.108449935913086,
-      "learning_rate": 5.773318744640688e-07,
-      "loss": 0.8549,
-      "step": 19460
-    },
-    {
-      "epoch": 0.902410755679184,
-      "grad_norm": 3.445681095123291,
-      "learning_rate": 5.746254264168649e-07,
-      "loss": 0.6915,
-      "step": 19465
-    },
-    {
-      "epoch": 0.9026425591098748,
-      "grad_norm": 3.283756732940674,
-      "learning_rate": 5.719251493340417e-07,
-      "loss": 0.8681,
-      "step": 19470
-    },
-    {
-      "epoch": 0.9028743625405656,
-      "grad_norm": 3.5738472938537598,
-      "learning_rate": 5.692310449835237e-07,
-      "loss": 0.9259,
-      "step": 19475
-    },
-    {
-      "epoch": 0.9031061659712564,
-      "grad_norm": 3.6439244747161865,
-      "learning_rate": 5.665431151291867e-07,
-      "loss": 0.7761,
-      "step": 19480
-    },
-    {
-      "epoch": 0.9033379694019471,
-      "grad_norm": 3.921151876449585,
-      "learning_rate": 5.638613615308708e-07,
-      "loss": 0.8634,
-      "step": 19485
-    },
-    {
-      "epoch": 0.903569772832638,
-      "grad_norm": 3.847851514816284,
-      "learning_rate": 5.61185785944366e-07,
-      "loss": 0.7767,
-      "step": 19490
-    },
-    {
-      "epoch": 0.9038015762633287,
-      "grad_norm": 4.24960994720459,
-      "learning_rate": 5.585163901214197e-07,
-      "loss": 0.7588,
-      "step": 19495
-    },
-    {
-      "epoch": 0.9040333796940194,
-      "grad_norm": 3.390794515609741,
-      "learning_rate": 5.558531758097396e-07,
-      "loss": 0.9196,
-      "step": 19500
-    },
-    {
-      "epoch": 0.9040333796940194,
-      "eval_loss": 0.8735858201980591,
-      "eval_runtime": 11.2655,
-      "eval_samples_per_second": 11.273,
-      "eval_steps_per_second": 11.273,
-      "step": 19500
-    },
-    {
-      "epoch": 0.9042651831247103,
-      "grad_norm": 3.261413097381592,
-      "learning_rate": 5.531961447529743e-07,
-      "loss": 0.7741,
-      "step": 19505
-    },
-    {
-      "epoch": 0.904496986555401,
-      "grad_norm": 3.9807848930358887,
-      "learning_rate": 5.505452986907322e-07,
-      "loss": 0.709,
-      "step": 19510
-    },
-    {
-      "epoch": 0.9047287899860917,
-      "grad_norm": 4.0768632888793945,
-      "learning_rate": 5.479006393585728e-07,
-      "loss": 0.8462,
-      "step": 19515
-    },
-    {
-      "epoch": 0.9049605934167826,
-      "grad_norm": 3.7398440837860107,
-      "learning_rate": 5.452621684879988e-07,
-      "loss": 0.8092,
-      "step": 19520
-    },
-    {
-      "epoch": 0.9051923968474733,
-      "grad_norm": 3.3626298904418945,
-      "learning_rate": 5.426298878064684e-07,
-      "loss": 0.7902,
-      "step": 19525
-    },
-    {
-      "epoch": 0.9054242002781642,
-      "grad_norm": 3.6051082611083984,
-      "learning_rate": 5.40003799037383e-07,
-      "loss": 0.8525,
-      "step": 19530
-    },
-    {
-      "epoch": 0.9056560037088549,
-      "grad_norm": 3.7399423122406006,
-      "learning_rate": 5.373839039000905e-07,
-      "loss": 0.8411,
-      "step": 19535
-    },
-    {
-      "epoch": 0.9058878071395456,
-      "grad_norm": 3.000722646713257,
-      "learning_rate": 5.347702041098834e-07,
-      "loss": 0.6594,
-      "step": 19540
-    },
-    {
-      "epoch": 0.9061196105702365,
-      "grad_norm": 3.077620029449463,
-      "learning_rate": 5.321627013779984e-07,
-      "loss": 0.7542,
-      "step": 19545
-    },
-    {
-      "epoch": 0.9063514140009272,
-      "grad_norm": 3.0676279067993164,
-      "learning_rate": 5.295613974116165e-07,
-      "loss": 0.6199,
-      "step": 19550
-    },
-    {
-      "epoch": 0.906583217431618,
-      "grad_norm": 3.53662371635437,
-      "learning_rate": 5.269662939138576e-07,
-      "loss": 0.6298,
-      "step": 19555
-    },
-    {
-      "epoch": 0.9068150208623088,
-      "grad_norm": 3.4631190299987793,
-      "learning_rate": 5.243773925837847e-07,
-      "loss": 0.8718,
-      "step": 19560
-    },
-    {
-      "epoch": 0.9070468242929995,
-      "grad_norm": 3.2487220764160156,
-      "learning_rate": 5.217946951163955e-07,
-      "loss": 0.7724,
-      "step": 19565
-    },
-    {
-      "epoch": 0.9072786277236903,
-      "grad_norm": 4.004088401794434,
-      "learning_rate": 5.192182032026327e-07,
-      "loss": 0.9907,
-      "step": 19570
-    },
-    {
-      "epoch": 0.9075104311543811,
-      "grad_norm": 3.633103132247925,
-      "learning_rate": 5.166479185293694e-07,
-      "loss": 0.9067,
-      "step": 19575
-    },
-    {
-      "epoch": 0.9077422345850719,
-      "grad_norm": 4.3104095458984375,
-      "learning_rate": 5.140838427794192e-07,
-      "loss": 0.8647,
-      "step": 19580
-    },
-    {
-      "epoch": 0.9079740380157626,
-      "grad_norm": 4.145888805389404,
-      "learning_rate": 5.115259776315307e-07,
-      "loss": 0.817,
-      "step": 19585
-    },
-    {
-      "epoch": 0.9082058414464534,
-      "grad_norm": 3.9163827896118164,
-      "learning_rate": 5.089743247603829e-07,
-      "loss": 0.8899,
-      "step": 19590
-    },
-    {
-      "epoch": 0.9084376448771442,
-      "grad_norm": 3.5132017135620117,
-      "learning_rate": 5.064288858365885e-07,
-      "loss": 0.747,
-      "step": 19595
-    },
-    {
-      "epoch": 0.9086694483078349,
-      "grad_norm": 3.710824966430664,
-      "learning_rate": 5.038896625266942e-07,
-      "loss": 0.9545,
-      "step": 19600
-    },
-    {
-      "epoch": 0.9086694483078349,
-      "eval_loss": 0.8735297918319702,
-      "eval_runtime": 11.2733,
-      "eval_samples_per_second": 11.266,
-      "eval_steps_per_second": 11.266,
-      "step": 19600
-    },
-    {
-      "epoch": 0.9089012517385258,
-      "grad_norm": 3.6324193477630615,
-      "learning_rate": 5.013566564931771e-07,
-      "loss": 0.8615,
-      "step": 19605
-    },
-    {
-      "epoch": 0.9091330551692165,
-      "grad_norm": 3.605224370956421,
-      "learning_rate": 4.988298693944382e-07,
-      "loss": 0.7686,
-      "step": 19610
-    },
-    {
-      "epoch": 0.9093648585999072,
-      "grad_norm": 4.282202243804932,
-      "learning_rate": 4.963093028848154e-07,
-      "loss": 0.8672,
-      "step": 19615
-    },
-    {
-      "epoch": 0.9095966620305981,
-      "grad_norm": 3.4218647480010986,
-      "learning_rate": 4.937949586145696e-07,
-      "loss": 0.7751,
-      "step": 19620
-    },
-    {
-      "epoch": 0.9098284654612888,
-      "grad_norm": 4.3748345375061035,
-      "learning_rate": 4.91286838229883e-07,
-      "loss": 1.0064,
-      "step": 19625
-    },
-    {
-      "epoch": 0.9100602688919796,
-      "grad_norm": 3.9138522148132324,
-      "learning_rate": 4.887849433728709e-07,
-      "loss": 0.7735,
-      "step": 19630
-    },
-    {
-      "epoch": 0.9102920723226704,
-      "grad_norm": 3.642190933227539,
-      "learning_rate": 4.862892756815707e-07,
-      "loss": 0.9095,
-      "step": 19635
-    },
-    {
-      "epoch": 0.9105238757533611,
-      "grad_norm": 3.8553028106689453,
-      "learning_rate": 4.837998367899388e-07,
-      "loss": 0.7353,
-      "step": 19640
-    },
-    {
-      "epoch": 0.9107556791840519,
-      "grad_norm": 3.8063559532165527,
-      "learning_rate": 4.813166283278569e-07,
-      "loss": 0.995,
-      "step": 19645
-    },
-    {
-      "epoch": 0.9109874826147427,
-      "grad_norm": 3.4844539165496826,
-      "learning_rate": 4.78839651921128e-07,
-      "loss": 0.7777,
-      "step": 19650
-    },
-    {
-      "epoch": 0.9112192860454335,
-      "grad_norm": 3.6754894256591797,
-      "learning_rate": 4.763689091914736e-07,
-      "loss": 0.8534,
-      "step": 19655
-    },
-    {
-      "epoch": 0.9114510894761243,
-      "grad_norm": 3.5442609786987305,
-      "learning_rate": 4.73904401756532e-07,
-      "loss": 0.7036,
-      "step": 19660
-    },
-    {
-      "epoch": 0.911682892906815,
-      "grad_norm": 3.768289804458618,
-      "learning_rate": 4.714461312298624e-07,
-      "loss": 0.8011,
-      "step": 19665
-    },
-    {
-      "epoch": 0.9119146963375058,
-      "grad_norm": 3.958264112472534,
-      "learning_rate": 4.689940992209407e-07,
-      "loss": 0.9242,
-      "step": 19670
-    },
-    {
-      "epoch": 0.9121464997681966,
-      "grad_norm": 3.9602081775665283,
-      "learning_rate": 4.66548307335154e-07,
-      "loss": 1.0384,
-      "step": 19675
-    },
-    {
-      "epoch": 0.9123783031988874,
-      "grad_norm": 3.245852470397949,
-      "learning_rate": 4.6410875717380897e-07,
-      "loss": 0.7956,
-      "step": 19680
-    },
-    {
-      "epoch": 0.9126101066295781,
-      "grad_norm": 3.358444929122925,
-      "learning_rate": 4.616754503341214e-07,
-      "loss": 0.7481,
-      "step": 19685
-    },
-    {
-      "epoch": 0.9128419100602689,
-      "grad_norm": 3.517886161804199,
-      "learning_rate": 4.592483884092247e-07,
-      "loss": 0.872,
-      "step": 19690
-    },
-    {
-      "epoch": 0.9130737134909597,
-      "grad_norm": 3.7229044437408447,
-      "learning_rate": 4.568275729881577e-07,
-      "loss": 0.755,
-      "step": 19695
-    },
-    {
-      "epoch": 0.9133055169216504,
-      "grad_norm": 3.6528728008270264,
-      "learning_rate": 4.5441300565587264e-07,
-      "loss": 0.7834,
-      "step": 19700
-    },
-    {
-      "epoch": 0.9133055169216504,
-      "eval_loss": 0.8733378052711487,
-      "eval_runtime": 11.2608,
-      "eval_samples_per_second": 11.278,
-      "eval_steps_per_second": 11.278,
-      "step": 19700
-    },
-    {
-      "epoch": 0.9135373203523413,
-      "grad_norm": 3.802203893661499,
-      "learning_rate": 4.5200468799323383e-07,
-      "loss": 0.9832,
-      "step": 19705
-    },
-    {
-      "epoch": 0.913769123783032,
-      "grad_norm": 4.232209205627441,
-      "learning_rate": 4.496026215770077e-07,
-      "loss": 0.7035,
-      "step": 19710
-    },
-    {
-      "epoch": 0.9140009272137227,
-      "grad_norm": 4.124170780181885,
-      "learning_rate": 4.472068079798708e-07,
-      "loss": 1.0385,
-      "step": 19715
-    },
-    {
-      "epoch": 0.9142327306444136,
-      "grad_norm": 3.3067452907562256,
-      "learning_rate": 4.4481724877040834e-07,
-      "loss": 0.9504,
-      "step": 19720
-    },
-    {
-      "epoch": 0.9144645340751043,
-      "grad_norm": 3.844074249267578,
-      "learning_rate": 4.424339455131044e-07,
-      "loss": 0.9187,
-      "step": 19725
-    },
-    {
-      "epoch": 0.914696337505795,
-      "grad_norm": 3.3239314556121826,
-      "learning_rate": 4.4005689976835297e-07,
-      "loss": 0.8161,
-      "step": 19730
-    },
-    {
-      "epoch": 0.9149281409364859,
-      "grad_norm": 3.4764716625213623,
-      "learning_rate": 4.3768611309245145e-07,
-      "loss": 0.7984,
-      "step": 19735
-    },
-    {
-      "epoch": 0.9151599443671766,
-      "grad_norm": 3.828855276107788,
-      "learning_rate": 4.3532158703759487e-07,
-      "loss": 0.7868,
-      "step": 19740
-    },
-    {
-      "epoch": 0.9153917477978674,
-      "grad_norm": 3.310119390487671,
-      "learning_rate": 4.329633231518804e-07,
-      "loss": 0.8396,
-      "step": 19745
-    },
-    {
-      "epoch": 0.9156235512285582,
-      "grad_norm": 3.7126951217651367,
-      "learning_rate": 4.306113229793074e-07,
-      "loss": 0.8955,
-      "step": 19750
-    },
-    {
-      "epoch": 0.9158553546592489,
-      "grad_norm": 3.6136951446533203,
-      "learning_rate": 4.2826558805977527e-07,
-      "loss": 0.775,
-      "step": 19755
-    },
-    {
-      "epoch": 0.9160871580899397,
-      "grad_norm": 3.726264715194702,
-      "learning_rate": 4.259261199290765e-07,
-      "loss": 0.7419,
-      "step": 19760
-    },
-    {
-      "epoch": 0.9163189615206305,
-      "grad_norm": 3.7436885833740234,
-      "learning_rate": 4.2359292011890484e-07,
-      "loss": 0.7245,
-      "step": 19765
-    },
-    {
-      "epoch": 0.9165507649513213,
-      "grad_norm": 3.374852180480957,
-      "learning_rate": 4.2126599015684723e-07,
-      "loss": 0.7779,
-      "step": 19770
-    },
-    {
-      "epoch": 0.9167825683820121,
-      "grad_norm": 3.1694605350494385,
-      "learning_rate": 4.189453315663905e-07,
-      "loss": 0.942,
-      "step": 19775
-    },
-    {
-      "epoch": 0.9170143718127028,
-      "grad_norm": 4.185464382171631,
-      "learning_rate": 4.1663094586690824e-07,
-      "loss": 0.7565,
-      "step": 19780
-    },
-    {
-      "epoch": 0.9172461752433936,
-      "grad_norm": 4.132409572601318,
-      "learning_rate": 4.143228345736716e-07,
-      "loss": 0.8305,
-      "step": 19785
-    },
-    {
-      "epoch": 0.9174779786740844,
-      "grad_norm": 3.8933820724487305,
-      "learning_rate": 4.1202099919784633e-07,
-      "loss": 0.9198,
-      "step": 19790
-    },
-    {
-      "epoch": 0.9177097821047752,
-      "grad_norm": 3.0863237380981445,
-      "learning_rate": 4.097254412464824e-07,
-      "loss": 0.7378,
-      "step": 19795
-    },
-    {
-      "epoch": 0.9179415855354659,
-      "grad_norm": 3.9306046962738037,
-      "learning_rate": 4.0743616222252424e-07,
-      "loss": 0.8401,
-      "step": 19800
-    },
-    {
-      "epoch": 0.9179415855354659,
-      "eval_loss": 0.8732080459594727,
-      "eval_runtime": 11.2654,
-      "eval_samples_per_second": 11.273,
-      "eval_steps_per_second": 11.273,
-      "step": 19800
-    },
-    {
-      "epoch": 0.9181733889661567,
-      "grad_norm": 3.6145589351654053,
-      "learning_rate": 4.0515316362480516e-07,
-      "loss": 0.8457,
-      "step": 19805
-    },
-    {
-      "epoch": 0.9184051923968475,
-      "grad_norm": 3.4796619415283203,
-      "learning_rate": 4.028764469480473e-07,
-      "loss": 0.8131,
-      "step": 19810
-    },
-    {
-      "epoch": 0.9186369958275382,
-      "grad_norm": 4.206612586975098,
-      "learning_rate": 4.006060136828549e-07,
-      "loss": 0.7882,
-      "step": 19815
-    },
-    {
-      "epoch": 0.9188687992582291,
-      "grad_norm": 3.936595916748047,
-      "learning_rate": 3.983418653157245e-07,
-      "loss": 0.7965,
-      "step": 19820
-    },
-    {
-      "epoch": 0.9191006026889198,
-      "grad_norm": 4.071559906005859,
-      "learning_rate": 3.960840033290358e-07,
-      "loss": 0.9554,
-      "step": 19825
-    },
-    {
-      "epoch": 0.9193324061196105,
-      "grad_norm": 3.226321220397949,
-      "learning_rate": 3.938324292010498e-07,
-      "loss": 0.7675,
-      "step": 19830
-    },
-    {
-      "epoch": 0.9195642095503014,
-      "grad_norm": 4.5919036865234375,
-      "learning_rate": 3.9158714440591496e-07,
-      "loss": 0.7775,
-      "step": 19835
-    },
-    {
-      "epoch": 0.9197960129809921,
-      "grad_norm": 3.887747049331665,
-      "learning_rate": 3.893481504136598e-07,
-      "loss": 0.7748,
-      "step": 19840
-    },
-    {
-      "epoch": 0.9200278164116829,
-      "grad_norm": 3.5264813899993896,
-      "learning_rate": 3.8711544869019404e-07,
-      "loss": 0.6458,
-      "step": 19845
-    },
-    {
-      "epoch": 0.9202596198423737,
-      "grad_norm": 4.041652679443359,
-      "learning_rate": 3.8488904069730713e-07,
-      "loss": 0.7981,
-      "step": 19850
-    },
-    {
-      "epoch": 0.9204914232730644,
-      "grad_norm": 3.4928979873657227,
-      "learning_rate": 3.826689278926721e-07,
-      "loss": 0.79,
-      "step": 19855
-    },
-    {
-      "epoch": 0.9207232267037552,
-      "grad_norm": 3.8515238761901855,
-      "learning_rate": 3.8045511172983493e-07,
-      "loss": 0.8458,
-      "step": 19860
-    },
-    {
-      "epoch": 0.920955030134446,
-      "grad_norm": 4.285694122314453,
-      "learning_rate": 3.782475936582208e-07,
-      "loss": 0.8627,
-      "step": 19865
-    },
-    {
-      "epoch": 0.9211868335651368,
-      "grad_norm": 3.648986577987671,
-      "learning_rate": 3.7604637512313245e-07,
-      "loss": 0.881,
-      "step": 19870
-    },
-    {
-      "epoch": 0.9214186369958275,
-      "grad_norm": 3.3540894985198975,
-      "learning_rate": 3.7385145756575037e-07,
-      "loss": 0.9356,
-      "step": 19875
-    },
-    {
-      "epoch": 0.9216504404265183,
-      "grad_norm": 4.117437362670898,
-      "learning_rate": 3.716628424231228e-07,
-      "loss": 0.9502,
-      "step": 19880
-    },
-    {
-      "epoch": 0.9218822438572091,
-      "grad_norm": 3.526902675628662,
-      "learning_rate": 3.694805311281802e-07,
-      "loss": 0.7778,
-      "step": 19885
-    },
-    {
-      "epoch": 0.9221140472878998,
-      "grad_norm": 3.7025606632232666,
-      "learning_rate": 3.6730452510971736e-07,
-      "loss": 0.7941,
-      "step": 19890
-    },
-    {
-      "epoch": 0.9223458507185907,
-      "grad_norm": 3.383410930633545,
-      "learning_rate": 3.6513482579240813e-07,
-      "loss": 0.6289,
-      "step": 19895
-    },
-    {
-      "epoch": 0.9225776541492814,
-      "grad_norm": 4.023029327392578,
-      "learning_rate": 3.629714345967916e-07,
-      "loss": 0.9294,
-      "step": 19900
-    },
-    {
-      "epoch": 0.9225776541492814,
-      "eval_loss": 0.8729880452156067,
-      "eval_runtime": 11.2731,
-      "eval_samples_per_second": 11.266,
-      "eval_steps_per_second": 11.266,
-      "step": 19900
-    },
-    {
-      "epoch": 0.9228094575799722,
-      "grad_norm": 3.646693229675293,
-      "learning_rate": 3.608143529392816e-07,
-      "loss": 0.8774,
-      "step": 19905
-    },
-    {
-      "epoch": 0.923041261010663,
-      "grad_norm": 3.4620437622070312,
-      "learning_rate": 3.586635822321594e-07,
-      "loss": 0.8828,
-      "step": 19910
-    },
-    {
-      "epoch": 0.9232730644413537,
-      "grad_norm": 3.016693353652954,
-      "learning_rate": 3.565191238835719e-07,
-      "loss": 0.8007,
-      "step": 19915
-    },
-    {
-      "epoch": 0.9235048678720446,
-      "grad_norm": 3.731323003768921,
-      "learning_rate": 3.54380979297535e-07,
-      "loss": 0.94,
-      "step": 19920
-    },
-    {
-      "epoch": 0.9237366713027353,
-      "grad_norm": 3.6027114391326904,
-      "learning_rate": 3.522491498739311e-07,
-      "loss": 0.9447,
-      "step": 19925
-    },
-    {
-      "epoch": 0.923968474733426,
-      "grad_norm": 3.1686642169952393,
-      "learning_rate": 3.5012363700850815e-07,
-      "loss": 0.8824,
-      "step": 19930
-    },
-    {
-      "epoch": 0.9242002781641169,
-      "grad_norm": 3.9111225605010986,
-      "learning_rate": 3.4800444209287744e-07,
-      "loss": 0.7697,
-      "step": 19935
-    },
-    {
-      "epoch": 0.9244320815948076,
-      "grad_norm": 3.5328216552734375,
-      "learning_rate": 3.458915665145168e-07,
-      "loss": 0.8081,
-      "step": 19940
-    },
-    {
-      "epoch": 0.9246638850254983,
-      "grad_norm": 4.168152332305908,
-      "learning_rate": 3.4378501165676313e-07,
-      "loss": 0.7272,
-      "step": 19945
-    },
-    {
-      "epoch": 0.9248956884561892,
-      "grad_norm": 3.5316083431243896,
-      "learning_rate": 3.416847788988131e-07,
-      "loss": 0.8657,
-      "step": 19950
-    },
-    {
-      "epoch": 0.9251274918868799,
-      "grad_norm": 3.775172472000122,
-      "learning_rate": 3.395908696157313e-07,
-      "loss": 0.908,
-      "step": 19955
-    },
-    {
-      "epoch": 0.9253592953175707,
-      "grad_norm": 3.273514986038208,
-      "learning_rate": 3.3750328517843766e-07,
-      "loss": 0.9044,
-      "step": 19960
-    },
-    {
-      "epoch": 0.9255910987482615,
-      "grad_norm": 3.712826728820801,
-      "learning_rate": 3.354220269537101e-07,
-      "loss": 0.9362,
-      "step": 19965
-    },
-    {
-      "epoch": 0.9258229021789522,
-      "grad_norm": 3.922048568725586,
-      "learning_rate": 3.333470963041874e-07,
-      "loss": 0.8773,
-      "step": 19970
-    },
-    {
-      "epoch": 0.926054705609643,
-      "grad_norm": 4.05618953704834,
-      "learning_rate": 3.3127849458836404e-07,
-      "loss": 0.8257,
-      "step": 19975
-    },
-    {
-      "epoch": 0.9262865090403338,
-      "grad_norm": 3.3972959518432617,
-      "learning_rate": 3.292162231605922e-07,
-      "loss": 0.836,
-      "step": 19980
-    },
-    {
-      "epoch": 0.9265183124710246,
-      "grad_norm": 4.221245765686035,
-      "learning_rate": 3.271602833710774e-07,
-      "loss": 0.8406,
-      "step": 19985
-    },
-    {
-      "epoch": 0.9267501159017153,
-      "grad_norm": 2.942352056503296,
-      "learning_rate": 3.251106765658807e-07,
-      "loss": 0.7356,
-      "step": 19990
-    },
-    {
-      "epoch": 0.9269819193324061,
-      "grad_norm": 4.593667507171631,
-      "learning_rate": 3.230674040869186e-07,
-      "loss": 0.9098,
-      "step": 19995
-    },
-    {
-      "epoch": 0.9272137227630969,
-      "grad_norm": 3.9805760383605957,
-      "learning_rate": 3.210304672719577e-07,
-      "loss": 0.8497,
-      "step": 20000
-    },
-    {
-      "epoch": 0.9272137227630969,
-      "eval_loss": 0.8728947639465332,
-      "eval_runtime": 11.2692,
-      "eval_samples_per_second": 11.27,
-      "eval_steps_per_second": 11.27,
-      "step": 20000
-    },
-    {
-      "epoch": 0.9274455261937876,
-      "grad_norm": 3.8057963848114014,
-      "learning_rate": 3.1899986745461685e-07,
-      "loss": 0.7294,
-      "step": 20005
-    },
-    {
-      "epoch": 0.9276773296244785,
-      "grad_norm": 4.51212739944458,
-      "learning_rate": 3.1697560596436915e-07,
-      "loss": 0.8166,
-      "step": 20010
-    },
-    {
-      "epoch": 0.9279091330551692,
-      "grad_norm": 4.221890449523926,
-      "learning_rate": 3.149576841265323e-07,
-      "loss": 0.826,
-      "step": 20015
-    },
-    {
-      "epoch": 0.9281409364858599,
-      "grad_norm": 3.759281873703003,
-      "learning_rate": 3.1294610326227825e-07,
-      "loss": 0.7127,
-      "step": 20020
-    },
-    {
-      "epoch": 0.9283727399165508,
-      "grad_norm": 3.804107189178467,
-      "learning_rate": 3.109408646886247e-07,
-      "loss": 0.8837,
-      "step": 20025
-    },
-    {
-      "epoch": 0.9286045433472415,
-      "grad_norm": 3.8112025260925293,
-      "learning_rate": 3.089419697184415e-07,
-      "loss": 0.7981,
-      "step": 20030
-    },
-    {
-      "epoch": 0.9288363467779324,
-      "grad_norm": 3.5084173679351807,
-      "learning_rate": 3.0694941966043613e-07,
-      "loss": 0.7923,
-      "step": 20035
-    },
-    {
-      "epoch": 0.9290681502086231,
-      "grad_norm": 4.406299114227295,
-      "learning_rate": 3.0496321581917066e-07,
-      "loss": 0.8866,
-      "step": 20040
-    },
-    {
-      "epoch": 0.9292999536393138,
-      "grad_norm": 4.219856262207031,
-      "learning_rate": 3.029833594950493e-07,
-      "loss": 0.6969,
-      "step": 20045
-    },
-    {
-      "epoch": 0.9295317570700047,
-      "grad_norm": 4.252161979675293,
-      "learning_rate": 3.010098519843174e-07,
-      "loss": 0.8416,
-      "step": 20050
-    },
-    {
-      "epoch": 0.9297635605006954,
-      "grad_norm": 4.353281497955322,
-      "learning_rate": 2.990426945790681e-07,
-      "loss": 0.9764,
-      "step": 20055
-    },
-    {
-      "epoch": 0.9299953639313862,
-      "grad_norm": 3.8172295093536377,
-      "learning_rate": 2.970818885672355e-07,
-      "loss": 0.9966,
-      "step": 20060
-    },
-    {
-      "epoch": 0.930227167362077,
-      "grad_norm": 3.732015609741211,
-      "learning_rate": 2.95127435232595e-07,
-      "loss": 0.9422,
-      "step": 20065
-    },
-    {
-      "epoch": 0.9304589707927677,
-      "grad_norm": 3.7781100273132324,
-      "learning_rate": 2.9317933585476075e-07,
-      "loss": 0.8624,
-      "step": 20070
-    },
-    {
-      "epoch": 0.9306907742234585,
-      "grad_norm": 3.4985451698303223,
-      "learning_rate": 2.912375917091903e-07,
-      "loss": 0.6331,
-      "step": 20075
-    },
-    {
-      "epoch": 0.9309225776541493,
-      "grad_norm": 3.749589204788208,
-      "learning_rate": 2.8930220406718e-07,
-      "loss": 0.7681,
-      "step": 20080
-    },
-    {
-      "epoch": 0.93115438108484,
-      "grad_norm": 4.194093704223633,
-      "learning_rate": 2.8737317419586185e-07,
-      "loss": 0.8769,
-      "step": 20085
-    },
-    {
-      "epoch": 0.9313861845155308,
-      "grad_norm": 3.5070695877075195,
-      "learning_rate": 2.854505033582078e-07,
-      "loss": 0.9185,
-      "step": 20090
-    },
-    {
-      "epoch": 0.9316179879462216,
-      "grad_norm": 3.497375011444092,
-      "learning_rate": 2.835341928130242e-07,
-      "loss": 0.9454,
-      "step": 20095
-    },
-    {
-      "epoch": 0.9318497913769124,
-      "grad_norm": 3.9200212955474854,
-      "learning_rate": 2.8162424381495633e-07,
-      "loss": 0.8779,
-      "step": 20100
-    },
-    {
-      "epoch": 0.9318497913769124,
-      "eval_loss": 0.8727548718452454,
-      "eval_runtime": 11.2653,
-      "eval_samples_per_second": 11.274,
-      "eval_steps_per_second": 11.274,
-      "step": 20100
-    },
-    {
-      "epoch": 0.9320815948076031,
-      "grad_norm": 3.960589647293091,
-      "learning_rate": 2.7972065761448065e-07,
-      "loss": 0.8635,
-      "step": 20105
-    },
-    {
-      "epoch": 0.932313398238294,
-      "grad_norm": 3.295830011367798,
-      "learning_rate": 2.7782343545791014e-07,
-      "loss": 0.7369,
-      "step": 20110
-    },
-    {
-      "epoch": 0.9325452016689847,
-      "grad_norm": 3.6758456230163574,
-      "learning_rate": 2.759325785873923e-07,
-      "loss": 0.7372,
-      "step": 20115
-    },
-    {
-      "epoch": 0.9327770050996754,
-      "grad_norm": 4.0003180503845215,
-      "learning_rate": 2.740480882409047e-07,
-      "loss": 0.8826,
-      "step": 20120
-    },
-    {
-      "epoch": 0.9330088085303663,
-      "grad_norm": 3.0322558879852295,
-      "learning_rate": 2.7216996565225583e-07,
-      "loss": 0.7696,
-      "step": 20125
-    },
-    {
-      "epoch": 0.933240611961057,
-      "grad_norm": 3.4727134704589844,
-      "learning_rate": 2.7029821205108885e-07,
-      "loss": 0.9406,
-      "step": 20130
-    },
-    {
-      "epoch": 0.9334724153917477,
-      "grad_norm": 4.067834377288818,
-      "learning_rate": 2.6843282866287344e-07,
-      "loss": 0.8498,
-      "step": 20135
-    },
-    {
-      "epoch": 0.9337042188224386,
-      "grad_norm": 3.396549701690674,
-      "learning_rate": 2.6657381670891047e-07,
-      "loss": 0.9788,
-      "step": 20140
-    },
-    {
-      "epoch": 0.9339360222531293,
-      "grad_norm": 3.457141399383545,
-      "learning_rate": 2.6472117740633074e-07,
-      "loss": 0.8151,
-      "step": 20145
-    },
-    {
-      "epoch": 0.9341678256838202,
-      "grad_norm": 4.950340747833252,
-      "learning_rate": 2.6287491196808957e-07,
-      "loss": 0.9633,
-      "step": 20150
-    },
-    {
-      "epoch": 0.9343996291145109,
-      "grad_norm": 3.4806978702545166,
-      "learning_rate": 2.610350216029689e-07,
-      "loss": 0.8388,
-      "step": 20155
-    },
-    {
-      "epoch": 0.9346314325452016,
-      "grad_norm": 4.318567276000977,
-      "learning_rate": 2.592015075155807e-07,
-      "loss": 0.8218,
-      "step": 20160
-    },
-    {
-      "epoch": 0.9348632359758925,
-      "grad_norm": 3.5549590587615967,
-      "learning_rate": 2.573743709063603e-07,
-      "loss": 0.9195,
-      "step": 20165
-    },
-    {
-      "epoch": 0.9350950394065832,
-      "grad_norm": 3.809382915496826,
-      "learning_rate": 2.555536129715652e-07,
-      "loss": 0.7518,
-      "step": 20170
-    },
-    {
-      "epoch": 0.935326842837274,
-      "grad_norm": 3.2903544902801514,
-      "learning_rate": 2.537392349032808e-07,
-      "loss": 0.8851,
-      "step": 20175
-    },
-    {
-      "epoch": 0.9355586462679648,
-      "grad_norm": 3.7616612911224365,
-      "learning_rate": 2.519312378894112e-07,
-      "loss": 0.9809,
-      "step": 20180
-    },
-    {
-      "epoch": 0.9357904496986555,
-      "grad_norm": 3.5184905529022217,
-      "learning_rate": 2.5012962311368737e-07,
-      "loss": 0.8735,
-      "step": 20185
-    },
-    {
-      "epoch": 0.9360222531293463,
-      "grad_norm": 4.265666961669922,
-      "learning_rate": 2.48334391755658e-07,
-      "loss": 0.8784,
-      "step": 20190
-    },
-    {
-      "epoch": 0.9362540565600371,
-      "grad_norm": 3.5565555095672607,
-      "learning_rate": 2.4654554499069395e-07,
-      "loss": 0.769,
-      "step": 20195
-    },
-    {
-      "epoch": 0.9364858599907279,
-      "grad_norm": 3.7593798637390137,
-      "learning_rate": 2.4476308398998727e-07,
-      "loss": 0.6453,
-      "step": 20200
-    },
-    {
-      "epoch": 0.9364858599907279,
-      "eval_loss": 0.8727805614471436,
-      "eval_runtime": 11.2633,
-      "eval_samples_per_second": 11.276,
-      "eval_steps_per_second": 11.276,
-      "step": 20200
-    },
-    {
-      "epoch": 0.9367176634214186,
-      "grad_norm": 4.56554651260376,
-      "learning_rate": 2.429870099205456e-07,
-      "loss": 0.8942,
-      "step": 20205
-    },
-    {
-      "epoch": 0.9369494668521094,
-      "grad_norm": 3.816410541534424,
-      "learning_rate": 2.4121732394519646e-07,
-      "loss": 0.919,
-      "step": 20210
-    },
-    {
-      "epoch": 0.9371812702828002,
-      "grad_norm": 3.6307473182678223,
-      "learning_rate": 2.394540272225887e-07,
-      "loss": 0.7905,
-      "step": 20215
-    },
-    {
-      "epoch": 0.9374130737134909,
-      "grad_norm": 4.1224684715271,
-      "learning_rate": 2.37697120907181e-07,
-      "loss": 0.7759,
-      "step": 20220
-    },
-    {
-      "epoch": 0.9376448771441818,
-      "grad_norm": 4.303830623626709,
-      "learning_rate": 2.3594660614925326e-07,
-      "loss": 0.9309,
-      "step": 20225
-    },
-    {
-      "epoch": 0.9378766805748725,
-      "grad_norm": 3.734203815460205,
-      "learning_rate": 2.3420248409490086e-07,
-      "loss": 1.0117,
-      "step": 20230
-    },
-    {
-      "epoch": 0.9381084840055632,
-      "grad_norm": 3.7188565731048584,
-      "learning_rate": 2.3246475588603267e-07,
-      "loss": 0.8211,
-      "step": 20235
-    },
-    {
-      "epoch": 0.9383402874362541,
-      "grad_norm": 4.2501983642578125,
-      "learning_rate": 2.3073342266036748e-07,
-      "loss": 0.8411,
-      "step": 20240
-    },
-    {
-      "epoch": 0.9385720908669448,
-      "grad_norm": 4.016298294067383,
-      "learning_rate": 2.2900848555144295e-07,
-      "loss": 0.8139,
-      "step": 20245
-    },
-    {
-      "epoch": 0.9388038942976356,
-      "grad_norm": 3.4987635612487793,
-      "learning_rate": 2.2728994568860685e-07,
-      "loss": 0.8471,
-      "step": 20250
-    },
-    {
-      "epoch": 0.9390356977283264,
-      "grad_norm": 4.0096049308776855,
-      "learning_rate": 2.2557780419701802e-07,
-      "loss": 0.7444,
-      "step": 20255
-    },
-    {
-      "epoch": 0.9392675011590171,
-      "grad_norm": 3.374089002609253,
-      "learning_rate": 2.2387206219764645e-07,
-      "loss": 0.8296,
-      "step": 20260
-    },
-    {
-      "epoch": 0.9394993045897079,
-      "grad_norm": 3.474876642227173,
-      "learning_rate": 2.2217272080727436e-07,
-      "loss": 0.9066,
-      "step": 20265
-    },
-    {
-      "epoch": 0.9397311080203987,
-      "grad_norm": 3.869378089904785,
-      "learning_rate": 2.2047978113849068e-07,
-      "loss": 0.7102,
-      "step": 20270
-    },
-    {
-      "epoch": 0.9399629114510895,
-      "grad_norm": 3.837383270263672,
-      "learning_rate": 2.1879324429969095e-07,
-      "loss": 0.8952,
-      "step": 20275
-    },
-    {
-      "epoch": 0.9401947148817803,
-      "grad_norm": 3.9950919151306152,
-      "learning_rate": 2.1711311139508528e-07,
-      "loss": 0.9188,
-      "step": 20280
-    },
-    {
-      "epoch": 0.940426518312471,
-      "grad_norm": 3.3735032081604004,
-      "learning_rate": 2.1543938352468597e-07,
-      "loss": 0.8943,
-      "step": 20285
-    },
-    {
-      "epoch": 0.9406583217431618,
-      "grad_norm": 3.7731034755706787,
-      "learning_rate": 2.137720617843131e-07,
-      "loss": 0.8451,
-      "step": 20290
-    },
-    {
-      "epoch": 0.9408901251738526,
-      "grad_norm": 3.9715733528137207,
-      "learning_rate": 2.1211114726559233e-07,
-      "loss": 0.8222,
-      "step": 20295
-    },
-    {
-      "epoch": 0.9411219286045434,
-      "grad_norm": 3.0871033668518066,
-      "learning_rate": 2.1045664105595387e-07,
-      "loss": 0.8157,
-      "step": 20300
-    },
-    {
-      "epoch": 0.9411219286045434,
-      "eval_loss": 0.8726610541343689,
-      "eval_runtime": 11.2667,
-      "eval_samples_per_second": 11.272,
-      "eval_steps_per_second": 11.272,
-      "step": 20300
-    },
-    {
-      "epoch": 0.9413537320352341,
-      "grad_norm": 4.362879276275635,
-      "learning_rate": 2.088085442386356e-07,
-      "loss": 0.9497,
-      "step": 20305
-    },
-    {
-      "epoch": 0.9415855354659249,
-      "grad_norm": 3.8125417232513428,
-      "learning_rate": 2.0716685789267445e-07,
-      "loss": 0.9322,
-      "step": 20310
-    },
-    {
-      "epoch": 0.9418173388966157,
-      "grad_norm": 3.584050416946411,
-      "learning_rate": 2.0553158309291167e-07,
-      "loss": 0.7188,
-      "step": 20315
-    },
-    {
-      "epoch": 0.9420491423273064,
-      "grad_norm": 3.3716397285461426,
-      "learning_rate": 2.039027209099942e-07,
-      "loss": 0.8026,
-      "step": 20320
-    },
-    {
-      "epoch": 0.9422809457579973,
-      "grad_norm": 3.3109800815582275,
-      "learning_rate": 2.0228027241036675e-07,
-      "loss": 0.8779,
-      "step": 20325
-    },
-    {
-      "epoch": 0.942512749188688,
-      "grad_norm": 3.862812042236328,
-      "learning_rate": 2.0066423865627405e-07,
-      "loss": 0.9238,
-      "step": 20330
-    },
-    {
-      "epoch": 0.9427445526193787,
-      "grad_norm": 3.545383930206299,
-      "learning_rate": 1.9905462070576532e-07,
-      "loss": 0.9785,
-      "step": 20335
-    },
-    {
-      "epoch": 0.9429763560500696,
-      "grad_norm": 3.8723387718200684,
-      "learning_rate": 1.974514196126842e-07,
-      "loss": 0.8293,
-      "step": 20340
-    },
-    {
-      "epoch": 0.9432081594807603,
-      "grad_norm": 3.7321619987487793,
-      "learning_rate": 1.9585463642667668e-07,
-      "loss": 0.8364,
-      "step": 20345
-    },
-    {
-      "epoch": 0.943439962911451,
-      "grad_norm": 3.676527261734009,
-      "learning_rate": 1.9426427219318755e-07,
-      "loss": 0.8345,
-      "step": 20350
-    },
-    {
-      "epoch": 0.9436717663421419,
-      "grad_norm": 3.3420934677124023,
-      "learning_rate": 1.9268032795345615e-07,
-      "loss": 0.7551,
-      "step": 20355
-    },
-    {
-      "epoch": 0.9439035697728326,
-      "grad_norm": 3.7380433082580566,
-      "learning_rate": 1.9110280474451847e-07,
-      "loss": 0.8608,
-      "step": 20360
-    },
-    {
-      "epoch": 0.9441353732035234,
-      "grad_norm": 3.892584800720215,
-      "learning_rate": 1.8953170359920835e-07,
-      "loss": 0.9361,
-      "step": 20365
-    },
-    {
-      "epoch": 0.9443671766342142,
-      "grad_norm": 3.387708902359009,
-      "learning_rate": 1.8796702554615521e-07,
-      "loss": 0.6998,
-      "step": 20370
-    },
-    {
-      "epoch": 0.944598980064905,
-      "grad_norm": 4.322428226470947,
-      "learning_rate": 1.8640877160978066e-07,
-      "loss": 0.8976,
-      "step": 20375
-    },
-    {
-      "epoch": 0.9448307834955957,
-      "grad_norm": 3.898622989654541,
-      "learning_rate": 1.8485694281030197e-07,
-      "loss": 0.8889,
-      "step": 20380
-    },
-    {
-      "epoch": 0.9450625869262865,
-      "grad_norm": 4.003245830535889,
-      "learning_rate": 1.833115401637331e-07,
-      "loss": 0.8091,
-      "step": 20385
-    },
-    {
-      "epoch": 0.9452943903569773,
-      "grad_norm": 3.791826009750366,
-      "learning_rate": 1.8177256468187354e-07,
-      "loss": 0.8516,
-      "step": 20390
-    },
-    {
-      "epoch": 0.9455261937876681,
-      "grad_norm": 4.986959457397461,
-      "learning_rate": 1.8024001737231957e-07,
-      "loss": 0.9104,
-      "step": 20395
-    },
-    {
-      "epoch": 0.9457579972183588,
-      "grad_norm": 3.9102840423583984,
-      "learning_rate": 1.7871389923845738e-07,
-      "loss": 0.8164,
-      "step": 20400
-    },
-    {
-      "epoch": 0.9457579972183588,
-      "eval_loss": 0.8726429343223572,
-      "eval_runtime": 11.267,
-      "eval_samples_per_second": 11.272,
-      "eval_steps_per_second": 11.272,
-      "step": 20400
-    },
-    {
-      "epoch": 0.9459898006490496,
-      "grad_norm": 3.6218159198760986,
-      "learning_rate": 1.7719421127946668e-07,
-      "loss": 0.7913,
-      "step": 20405
-    },
-    {
-      "epoch": 0.9462216040797404,
-      "grad_norm": 3.935687780380249,
-      "learning_rate": 1.7568095449031263e-07,
-      "loss": 0.8897,
-      "step": 20410
-    },
-    {
-      "epoch": 0.9464534075104312,
-      "grad_norm": 4.482246398925781,
-      "learning_rate": 1.7417412986175163e-07,
-      "loss": 0.7946,
-      "step": 20415
-    },
-    {
-      "epoch": 0.9466852109411219,
-      "grad_norm": 3.267892599105835,
-      "learning_rate": 1.726737383803312e-07,
-      "loss": 0.8567,
-      "step": 20420
-    },
-    {
-      "epoch": 0.9469170143718127,
-      "grad_norm": 4.164371013641357,
-      "learning_rate": 1.7117978102838328e-07,
-      "loss": 0.8608,
-      "step": 20425
-    },
-    {
-      "epoch": 0.9471488178025035,
-      "grad_norm": 3.210048198699951,
-      "learning_rate": 1.6969225878402883e-07,
-      "loss": 0.8283,
-      "step": 20430
-    },
-    {
-      "epoch": 0.9473806212331942,
-      "grad_norm": 4.08260440826416,
-      "learning_rate": 1.682111726211777e-07,
-      "loss": 0.8282,
-      "step": 20435
-    },
-    {
-      "epoch": 0.9476124246638851,
-      "grad_norm": 3.357116222381592,
-      "learning_rate": 1.6673652350952307e-07,
-      "loss": 0.7393,
-      "step": 20440
-    },
-    {
-      "epoch": 0.9478442280945758,
-      "grad_norm": 4.292599678039551,
-      "learning_rate": 1.6526831241454268e-07,
-      "loss": 0.8278,
-      "step": 20445
-    },
-    {
-      "epoch": 0.9480760315252665,
-      "grad_norm": 4.124458312988281,
-      "learning_rate": 1.638065402975031e-07,
-      "loss": 0.8579,
-      "step": 20450
-    },
-    {
-      "epoch": 0.9483078349559574,
-      "grad_norm": 3.8831825256347656,
-      "learning_rate": 1.6235120811545325e-07,
-      "loss": 0.834,
-      "step": 20455
-    },
-    {
-      "epoch": 0.9485396383866481,
-      "grad_norm": 4.089710235595703,
-      "learning_rate": 1.6090231682122536e-07,
-      "loss": 0.8033,
-      "step": 20460
-    },
-    {
-      "epoch": 0.9487714418173389,
-      "grad_norm": 4.102946758270264,
-      "learning_rate": 1.5945986736343287e-07,
-      "loss": 0.7532,
-      "step": 20465
-    },
-    {
-      "epoch": 0.9490032452480297,
-      "grad_norm": 3.067202091217041,
-      "learning_rate": 1.58023860686477e-07,
-      "loss": 0.8644,
-      "step": 20470
-    },
-    {
-      "epoch": 0.9492350486787204,
-      "grad_norm": 3.8552489280700684,
-      "learning_rate": 1.565942977305368e-07,
-      "loss": 0.9556,
-      "step": 20475
-    },
-    {
-      "epoch": 0.9494668521094112,
-      "grad_norm": 4.418707370758057,
-      "learning_rate": 1.5517117943157022e-07,
-      "loss": 0.8926,
-      "step": 20480
-    },
-    {
-      "epoch": 0.949698655540102,
-      "grad_norm": 3.372945785522461,
-      "learning_rate": 1.5375450672132198e-07,
-      "loss": 0.886,
-      "step": 20485
-    },
-    {
-      "epoch": 0.9499304589707928,
-      "grad_norm": 3.378730297088623,
-      "learning_rate": 1.5234428052731233e-07,
-      "loss": 0.8458,
-      "step": 20490
-    },
-    {
-      "epoch": 0.9501622624014835,
-      "grad_norm": 3.5803749561309814,
-      "learning_rate": 1.509405017728416e-07,
-      "loss": 0.7429,
-      "step": 20495
-    },
-    {
-      "epoch": 0.9503940658321743,
-      "grad_norm": 3.7786428928375244,
-      "learning_rate": 1.4954317137699016e-07,
-      "loss": 0.755,
-      "step": 20500
-    },
-    {
-      "epoch": 0.9503940658321743,
-      "eval_loss": 0.8725816011428833,
-      "eval_runtime": 11.2693,
-      "eval_samples_per_second": 11.27,
-      "eval_steps_per_second": 11.27,
-      "step": 20500
-    },
-    {
-      "epoch": 0.9506258692628651,
-      "grad_norm": 4.169706344604492,
-      "learning_rate": 1.481522902546151e-07,
-      "loss": 0.943,
-      "step": 20505
-    },
-    {
-      "epoch": 0.9508576726935558,
-      "grad_norm": 3.189598560333252,
-      "learning_rate": 1.467678593163524e-07,
-      "loss": 0.8133,
-      "step": 20510
-    },
-    {
-      "epoch": 0.9510894761242467,
-      "grad_norm": 4.087679386138916,
-      "learning_rate": 1.453898794686126e-07,
-      "loss": 0.7238,
-      "step": 20515
-    },
-    {
-      "epoch": 0.9513212795549374,
-      "grad_norm": 4.687533378601074,
-      "learning_rate": 1.4401835161358512e-07,
-      "loss": 0.8274,
-      "step": 20520
-    },
-    {
-      "epoch": 0.9515530829856282,
-      "grad_norm": 4.421091079711914,
-      "learning_rate": 1.42653276649235e-07,
-      "loss": 1.0243,
-      "step": 20525
-    },
-    {
-      "epoch": 0.951784886416319,
-      "grad_norm": 3.2969958782196045,
-      "learning_rate": 1.4129465546930177e-07,
-      "loss": 0.7684,
-      "step": 20530
-    },
-    {
-      "epoch": 0.9520166898470097,
-      "grad_norm": 3.84316086769104,
-      "learning_rate": 1.399424889632972e-07,
-      "loss": 0.8361,
-      "step": 20535
-    },
-    {
-      "epoch": 0.9522484932777006,
-      "grad_norm": 3.4402241706848145,
-      "learning_rate": 1.3859677801651094e-07,
-      "loss": 0.7908,
-      "step": 20540
-    },
-    {
-      "epoch": 0.9524802967083913,
-      "grad_norm": 3.5798709392547607,
-      "learning_rate": 1.3725752351000488e-07,
-      "loss": 0.6813,
-      "step": 20545
-    },
-    {
-      "epoch": 0.952712100139082,
-      "grad_norm": 3.9807188510894775,
-      "learning_rate": 1.3592472632061204e-07,
-      "loss": 0.7419,
-      "step": 20550
-    },
-    {
-      "epoch": 0.9529439035697729,
-      "grad_norm": 3.7711613178253174,
-      "learning_rate": 1.345983873209411e-07,
-      "loss": 0.9607,
-      "step": 20555
-    },
-    {
-      "epoch": 0.9531757070004636,
-      "grad_norm": 3.3518753051757812,
-      "learning_rate": 1.3327850737936853e-07,
-      "loss": 0.7509,
-      "step": 20560
-    },
-    {
-      "epoch": 0.9534075104311543,
-      "grad_norm": 5.417386531829834,
-      "learning_rate": 1.319650873600431e-07,
-      "loss": 0.9085,
-      "step": 20565
-    },
-    {
-      "epoch": 0.9536393138618452,
-      "grad_norm": 5.1365790367126465,
-      "learning_rate": 1.3065812812288692e-07,
-      "loss": 0.8834,
-      "step": 20570
-    },
-    {
-      "epoch": 0.9538711172925359,
-      "grad_norm": 3.5701076984405518,
-      "learning_rate": 1.2935763052358886e-07,
-      "loss": 0.8912,
-      "step": 20575
-    },
-    {
-      "epoch": 0.9541029207232267,
-      "grad_norm": 4.482658863067627,
-      "learning_rate": 1.280635954136089e-07,
-      "loss": 0.932,
-      "step": 20580
-    },
-    {
-      "epoch": 0.9543347241539175,
-      "grad_norm": 4.094351291656494,
-      "learning_rate": 1.2677602364017382e-07,
-      "loss": 0.9277,
-      "step": 20585
-    },
-    {
-      "epoch": 0.9545665275846082,
-      "grad_norm": 3.1283581256866455,
-      "learning_rate": 1.254949160462826e-07,
-      "loss": 0.7529,
-      "step": 20590
-    },
-    {
-      "epoch": 0.954798331015299,
-      "grad_norm": 3.5420258045196533,
-      "learning_rate": 1.2422027347069877e-07,
-      "loss": 0.7797,
-      "step": 20595
-    },
-    {
-      "epoch": 0.9550301344459898,
-      "grad_norm": 3.7095041275024414,
-      "learning_rate": 1.229520967479536e-07,
-      "loss": 0.9017,
-      "step": 20600
-    },
-    {
-      "epoch": 0.9550301344459898,
-      "eval_loss": 0.872434139251709,
-      "eval_runtime": 11.2645,
-      "eval_samples_per_second": 11.274,
-      "eval_steps_per_second": 11.274,
-      "step": 20600
-    },
-    {
-      "epoch": 0.9552619378766806,
-      "grad_norm": 4.508840084075928,
-      "learning_rate": 1.2169038670834522e-07,
-      "loss": 0.9815,
-      "step": 20605
-    },
-    {
-      "epoch": 0.9554937413073713,
-      "grad_norm": 3.8109028339385986,
-      "learning_rate": 1.2043514417794055e-07,
-      "loss": 0.8162,
-      "step": 20610
-    },
-    {
-      "epoch": 0.9557255447380621,
-      "grad_norm": 3.4969725608825684,
-      "learning_rate": 1.1918636997856669e-07,
-      "loss": 0.6752,
-      "step": 20615
-    },
-    {
-      "epoch": 0.9559573481687529,
-      "grad_norm": 3.203023910522461,
-      "learning_rate": 1.1794406492781962e-07,
-      "loss": 0.9055,
-      "step": 20620
-    },
-    {
-      "epoch": 0.9561891515994436,
-      "grad_norm": 4.043225288391113,
-      "learning_rate": 1.1670822983906094e-07,
-      "loss": 0.9176,
-      "step": 20625
-    },
-    {
-      "epoch": 0.9564209550301345,
-      "grad_norm": 3.5347437858581543,
-      "learning_rate": 1.1547886552141229e-07,
-      "loss": 0.8326,
-      "step": 20630
-    },
-    {
-      "epoch": 0.9566527584608252,
-      "grad_norm": 4.313322067260742,
-      "learning_rate": 1.1425597277976097e-07,
-      "loss": 1.0108,
-      "step": 20635
-    },
-    {
-      "epoch": 0.9568845618915159,
-      "grad_norm": 2.926300525665283,
-      "learning_rate": 1.1303955241475872e-07,
-      "loss": 0.6825,
-      "step": 20640
-    },
-    {
-      "epoch": 0.9571163653222068,
-      "grad_norm": 3.7507739067077637,
-      "learning_rate": 1.1182960522281739e-07,
-      "loss": 0.6986,
-      "step": 20645
-    },
-    {
-      "epoch": 0.9573481687528975,
-      "grad_norm": 3.813779354095459,
-      "learning_rate": 1.1062613199610994e-07,
-      "loss": 0.9632,
-      "step": 20650
-    },
-    {
-      "epoch": 0.9575799721835884,
-      "grad_norm": 3.9066879749298096,
-      "learning_rate": 1.0942913352257389e-07,
-      "loss": 0.7497,
-      "step": 20655
-    },
-    {
-      "epoch": 0.9578117756142791,
-      "grad_norm": 3.8484296798706055,
-      "learning_rate": 1.0823861058590568e-07,
-      "loss": 0.8,
-      "step": 20660
-    },
-    {
-      "epoch": 0.9580435790449698,
-      "grad_norm": 3.3053674697875977,
-      "learning_rate": 1.0705456396556179e-07,
-      "loss": 0.775,
-      "step": 20665
-    },
-    {
-      "epoch": 0.9582753824756607,
-      "grad_norm": 3.2667057514190674,
-      "learning_rate": 1.0587699443675992e-07,
-      "loss": 0.7743,
-      "step": 20670
-    },
-    {
-      "epoch": 0.9585071859063514,
-      "grad_norm": 4.044334411621094,
-      "learning_rate": 1.0470590277047554e-07,
-      "loss": 0.9362,
-      "step": 20675
-    },
-    {
-      "epoch": 0.9587389893370422,
-      "grad_norm": 3.7937705516815186,
-      "learning_rate": 1.0354128973344536e-07,
-      "loss": 0.9165,
-      "step": 20680
-    },
-    {
-      "epoch": 0.958970792767733,
-      "grad_norm": 4.389861106872559,
-      "learning_rate": 1.0238315608815941e-07,
-      "loss": 1.0114,
-      "step": 20685
-    },
-    {
-      "epoch": 0.9592025961984237,
-      "grad_norm": 3.973416805267334,
-      "learning_rate": 1.0123150259287229e-07,
-      "loss": 0.8243,
-      "step": 20690
-    },
-    {
-      "epoch": 0.9594343996291145,
-      "grad_norm": 4.305580139160156,
-      "learning_rate": 1.0008633000159085e-07,
-      "loss": 0.9289,
-      "step": 20695
-    },
-    {
-      "epoch": 0.9596662030598053,
-      "grad_norm": 4.440845489501953,
-      "learning_rate": 9.894763906407978e-08,
-      "loss": 1.033,
-      "step": 20700
-    },
-    {
-      "epoch": 0.9596662030598053,
-      "eval_loss": 0.8724677562713623,
-      "eval_runtime": 11.2771,
-      "eval_samples_per_second": 11.262,
-      "eval_steps_per_second": 11.262,
-      "step": 20700
-    },
-    {
-      "epoch": 0.9598980064904961,
-      "grad_norm": 4.005496025085449,
-      "learning_rate": 9.781543052586273e-08,
-      "loss": 0.798,
-      "step": 20705
-    },
-    {
-      "epoch": 0.9601298099211868,
-      "grad_norm": 3.747995615005493,
-      "learning_rate": 9.668970512821674e-08,
-      "loss": 0.8338,
-      "step": 20710
-    },
-    {
-      "epoch": 0.9603616133518776,
-      "grad_norm": 3.5021567344665527,
-      "learning_rate": 9.557046360817335e-08,
-      "loss": 0.8593,
-      "step": 20715
-    },
-    {
-      "epoch": 0.9605934167825684,
-      "grad_norm": 2.960855484008789,
-      "learning_rate": 9.445770669852084e-08,
-      "loss": 0.7712,
-      "step": 20720
-    },
-    {
-      "epoch": 0.9608252202132591,
-      "grad_norm": 3.840736150741577,
-      "learning_rate": 9.335143512780198e-08,
-      "loss": 0.8052,
-      "step": 20725
-    },
-    {
-      "epoch": 0.96105702364395,
-      "grad_norm": 3.2714478969573975,
-      "learning_rate": 9.225164962031297e-08,
-      "loss": 0.8372,
-      "step": 20730
-    },
-    {
-      "epoch": 0.9612888270746407,
-      "grad_norm": 3.7272298336029053,
-      "learning_rate": 9.115835089610337e-08,
-      "loss": 0.8361,
-      "step": 20735
-    },
-    {
-      "epoch": 0.9615206305053314,
-      "grad_norm": 3.4043381214141846,
-      "learning_rate": 9.007153967097504e-08,
-      "loss": 0.8576,
-      "step": 20740
-    },
-    {
-      "epoch": 0.9617524339360223,
-      "grad_norm": 3.917623281478882,
-      "learning_rate": 8.899121665648436e-08,
-      "loss": 0.7097,
-      "step": 20745
-    },
-    {
-      "epoch": 0.961984237366713,
-      "grad_norm": 3.294656276702881,
-      "learning_rate": 8.791738255993665e-08,
-      "loss": 0.8041,
-      "step": 20750
-    },
-    {
-      "epoch": 0.9622160407974037,
-      "grad_norm": 4.912676811218262,
-      "learning_rate": 8.685003808439174e-08,
-      "loss": 0.9048,
-      "step": 20755
-    },
-    {
-      "epoch": 0.9624478442280946,
-      "grad_norm": 3.2419793605804443,
-      "learning_rate": 8.578918392866064e-08,
-      "loss": 0.7305,
-      "step": 20760
-    },
-    {
-      "epoch": 0.9626796476587853,
-      "grad_norm": 3.819314956665039,
-      "learning_rate": 8.473482078730221e-08,
-      "loss": 1.0409,
-      "step": 20765
-    },
-    {
-      "epoch": 0.9629114510894762,
-      "grad_norm": 2.945345640182495,
-      "learning_rate": 8.36869493506276e-08,
-      "loss": 0.7376,
-      "step": 20770
-    },
-    {
-      "epoch": 0.9631432545201669,
-      "grad_norm": 3.5658223628997803,
-      "learning_rate": 8.264557030469689e-08,
-      "loss": 0.8783,
-      "step": 20775
-    },
-    {
-      "epoch": 0.9633750579508576,
-      "grad_norm": 4.4376444816589355,
-      "learning_rate": 8.161068433132135e-08,
-      "loss": 0.8467,
-      "step": 20780
-    },
-    {
-      "epoch": 0.9636068613815485,
-      "grad_norm": 3.707390069961548,
-      "learning_rate": 8.058229210805901e-08,
-      "loss": 0.8506,
-      "step": 20785
-    },
-    {
-      "epoch": 0.9638386648122392,
-      "grad_norm": 3.9231960773468018,
-      "learning_rate": 7.956039430821572e-08,
-      "loss": 0.9538,
-      "step": 20790
-    },
-    {
-      "epoch": 0.96407046824293,
-      "grad_norm": 3.662692070007324,
-      "learning_rate": 7.85449916008485e-08,
-      "loss": 0.8462,
-      "step": 20795
-    },
-    {
-      "epoch": 0.9643022716736208,
-      "grad_norm": 3.5197982788085938,
-      "learning_rate": 7.75360846507589e-08,
-      "loss": 0.8184,
-      "step": 20800
-    },
-    {
-      "epoch": 0.9643022716736208,
-      "eval_loss": 0.8724554181098938,
-      "eval_runtime": 11.2675,
-      "eval_samples_per_second": 11.271,
-      "eval_steps_per_second": 11.271,
-      "step": 20800
-    },
-    {
-      "epoch": 0.9645340751043116,
-      "grad_norm": 3.512238025665283,
-      "learning_rate": 7.653367411849632e-08,
-      "loss": 0.8036,
-      "step": 20805
-    },
-    {
-      "epoch": 0.9647658785350023,
-      "grad_norm": 3.1053659915924072,
-      "learning_rate": 7.5537760660358e-08,
-      "loss": 0.7613,
-      "step": 20810
-    },
-    {
-      "epoch": 0.9649976819656931,
-      "grad_norm": 4.269003391265869,
-      "learning_rate": 7.454834492838681e-08,
-      "loss": 1.027,
-      "step": 20815
-    },
-    {
-      "epoch": 0.9652294853963839,
-      "grad_norm": 4.014410018920898,
-      "learning_rate": 7.356542757037011e-08,
-      "loss": 0.8457,
-      "step": 20820
-    },
-    {
-      "epoch": 0.9654612888270746,
-      "grad_norm": 4.1713738441467285,
-      "learning_rate": 7.258900922984313e-08,
-      "loss": 0.8088,
-      "step": 20825
-    },
-    {
-      "epoch": 0.9656930922577655,
-      "grad_norm": 4.075442790985107,
-      "learning_rate": 7.161909054608451e-08,
-      "loss": 0.9568,
-      "step": 20830
-    },
-    {
-      "epoch": 0.9659248956884562,
-      "grad_norm": 3.7142698764801025,
-      "learning_rate": 7.065567215411629e-08,
-      "loss": 0.9301,
-      "step": 20835
-    },
-    {
-      "epoch": 0.9661566991191469,
-      "grad_norm": 3.9417366981506348,
-      "learning_rate": 6.969875468470721e-08,
-      "loss": 0.7984,
-      "step": 20840
-    },
-    {
-      "epoch": 0.9663885025498378,
-      "grad_norm": 4.075379371643066,
-      "learning_rate": 6.874833876436948e-08,
-      "loss": 0.838,
-      "step": 20845
-    },
-    {
-      "epoch": 0.9666203059805285,
-      "grad_norm": 3.971217393875122,
-      "learning_rate": 6.780442501535534e-08,
-      "loss": 0.7235,
-      "step": 20850
-    },
-    {
-      "epoch": 0.9668521094112192,
-      "grad_norm": 3.244826078414917,
-      "learning_rate": 6.686701405566376e-08,
-      "loss": 0.7689,
-      "step": 20855
-    },
-    {
-      "epoch": 0.9670839128419101,
-      "grad_norm": 3.7413337230682373,
-      "learning_rate": 6.593610649903493e-08,
-      "loss": 0.8759,
-      "step": 20860
-    },
-    {
-      "epoch": 0.9673157162726008,
-      "grad_norm": 3.7818310260772705,
-      "learning_rate": 6.501170295495019e-08,
-      "loss": 0.8278,
-      "step": 20865
-    },
-    {
-      "epoch": 0.9675475197032916,
-      "grad_norm": 3.9411723613739014,
-      "learning_rate": 6.40938040286343e-08,
-      "loss": 0.783,
-      "step": 20870
-    },
-    {
-      "epoch": 0.9677793231339824,
-      "grad_norm": 3.7038698196411133,
-      "learning_rate": 6.318241032105099e-08,
-      "loss": 0.8504,
-      "step": 20875
-    },
-    {
-      "epoch": 0.9680111265646731,
-      "grad_norm": 3.660369396209717,
-      "learning_rate": 6.22775224289085e-08,
-      "loss": 0.9106,
-      "step": 20880
-    },
-    {
-      "epoch": 0.9682429299953639,
-      "grad_norm": 3.6254231929779053,
-      "learning_rate": 6.13791409446518e-08,
-      "loss": 0.8556,
-      "step": 20885
-    },
-    {
-      "epoch": 0.9684747334260547,
-      "grad_norm": 3.7178094387054443,
-      "learning_rate": 6.048726645646819e-08,
-      "loss": 0.7655,
-      "step": 20890
-    },
-    {
-      "epoch": 0.9687065368567455,
-      "grad_norm": 3.7253708839416504,
-      "learning_rate": 5.960189954828277e-08,
-      "loss": 0.9205,
-      "step": 20895
-    },
-    {
-      "epoch": 0.9689383402874363,
-      "grad_norm": 5.241983413696289,
-      "learning_rate": 5.87230407997641e-08,
-      "loss": 0.8271,
-      "step": 20900
-    },
-    {
-      "epoch": 0.9689383402874363,
-      "eval_loss": 0.8723733425140381,
-      "eval_runtime": 11.274,
-      "eval_samples_per_second": 11.265,
-      "eval_steps_per_second": 11.265,
-      "step": 20900
-    },
-    {
-      "epoch": 0.969170143718127,
-      "grad_norm": 3.9776298999786377,
-      "learning_rate": 5.785069078631411e-08,
-      "loss": 1.0213,
-      "step": 20905
-    },
-    {
-      "epoch": 0.9694019471488178,
-      "grad_norm": 4.242099285125732,
-      "learning_rate": 5.698485007907817e-08,
-      "loss": 0.7948,
-      "step": 20910
-    },
-    {
-      "epoch": 0.9696337505795086,
-      "grad_norm": 3.6130335330963135,
-      "learning_rate": 5.612551924493836e-08,
-      "loss": 0.8865,
-      "step": 20915
-    },
-    {
-      "epoch": 0.9698655540101994,
-      "grad_norm": 3.73136305809021,
-      "learning_rate": 5.527269884651132e-08,
-      "loss": 0.8146,
-      "step": 20920
-    },
-    {
-      "epoch": 0.9700973574408901,
-      "grad_norm": 3.5985543727874756,
-      "learning_rate": 5.442638944215706e-08,
-      "loss": 0.7532,
-      "step": 20925
-    },
-    {
-      "epoch": 0.9703291608715809,
-      "grad_norm": 3.584609031677246,
-      "learning_rate": 5.3586591585969016e-08,
-      "loss": 0.8114,
-      "step": 20930
-    },
-    {
-      "epoch": 0.9705609643022717,
-      "grad_norm": 3.8732564449310303,
-      "learning_rate": 5.2753305827776266e-08,
-      "loss": 0.8789,
-      "step": 20935
-    },
-    {
-      "epoch": 0.9707927677329624,
-      "grad_norm": 4.232132434844971,
-      "learning_rate": 5.192653271314907e-08,
-      "loss": 0.8396,
-      "step": 20940
-    },
-    {
-      "epoch": 0.9710245711636533,
-      "grad_norm": 3.9535317420959473,
-      "learning_rate": 5.1106272783388865e-08,
-      "loss": 0.7687,
-      "step": 20945
-    },
-    {
-      "epoch": 0.971256374594344,
-      "grad_norm": 3.3350656032562256,
-      "learning_rate": 5.029252657553607e-08,
-      "loss": 0.7718,
-      "step": 20950
-    },
-    {
-      "epoch": 0.9714881780250347,
-      "grad_norm": 3.9200267791748047,
-      "learning_rate": 4.948529462236451e-08,
-      "loss": 0.8927,
-      "step": 20955
-    },
-    {
-      "epoch": 0.9717199814557256,
-      "grad_norm": 4.135104656219482,
-      "learning_rate": 4.868457745238253e-08,
-      "loss": 0.7815,
-      "step": 20960
-    },
-    {
-      "epoch": 0.9719517848864163,
-      "grad_norm": 3.703777551651001,
-      "learning_rate": 4.7890375589836334e-08,
-      "loss": 0.8333,
-      "step": 20965
-    },
-    {
-      "epoch": 0.972183588317107,
-      "grad_norm": 4.566959381103516,
-      "learning_rate": 4.710268955470332e-08,
-      "loss": 1.0059,
-      "step": 20970
-    },
-    {
-      "epoch": 0.9724153917477979,
-      "grad_norm": 3.848745107650757,
-      "learning_rate": 4.63215198626954e-08,
-      "loss": 0.8187,
-      "step": 20975
-    },
-    {
-      "epoch": 0.9726471951784886,
-      "grad_norm": 4.270590305328369,
-      "learning_rate": 4.554686702525901e-08,
-      "loss": 0.9922,
-      "step": 20980
-    },
-    {
-      "epoch": 0.9728789986091794,
-      "grad_norm": 4.593664646148682,
-      "learning_rate": 4.477873154957402e-08,
-      "loss": 0.8601,
-      "step": 20985
-    },
-    {
-      "epoch": 0.9731108020398702,
-      "grad_norm": 4.01714563369751,
-      "learning_rate": 4.401711393855257e-08,
-      "loss": 0.9009,
-      "step": 20990
-    },
-    {
-      "epoch": 0.973342605470561,
-      "grad_norm": 3.4283859729766846,
-      "learning_rate": 4.326201469083802e-08,
-      "loss": 0.7851,
-      "step": 20995
-    },
-    {
-      "epoch": 0.9735744089012517,
-      "grad_norm": 3.7679734230041504,
-      "learning_rate": 4.251343430080934e-08,
-      "loss": 0.8801,
-      "step": 21000
-    },
-    {
-      "epoch": 0.9735744089012517,
-      "eval_loss": 0.8723198771476746,
-      "eval_runtime": 11.2701,
-      "eval_samples_per_second": 11.269,
-      "eval_steps_per_second": 11.269,
-      "step": 21000
-    },
-    {
-      "epoch": 0.9738062123319425,
-      "grad_norm": 3.2241997718811035,
-      "learning_rate": 4.177137325857672e-08,
-      "loss": 0.8128,
-      "step": 21005
-    },
-    {
-      "epoch": 0.9740380157626333,
-      "grad_norm": 4.312016010284424,
-      "learning_rate": 4.103583204997708e-08,
-      "loss": 0.8834,
-      "step": 21010
-    },
-    {
-      "epoch": 0.9742698191933241,
-      "grad_norm": 4.0609941482543945,
-      "learning_rate": 4.03068111565863e-08,
-      "loss": 0.9945,
-      "step": 21015
-    },
-    {
-      "epoch": 0.9745016226240149,
-      "grad_norm": 4.250798225402832,
-      "learning_rate": 3.9584311055707036e-08,
-      "loss": 0.9721,
-      "step": 21020
-    },
-    {
-      "epoch": 0.9747334260547056,
-      "grad_norm": 3.8703458309173584,
-      "learning_rate": 3.886833222037201e-08,
-      "loss": 0.7808,
-      "step": 21025
-    },
-    {
-      "epoch": 0.9749652294853964,
-      "grad_norm": 4.010364055633545,
-      "learning_rate": 3.815887511934735e-08,
-      "loss": 0.7971,
-      "step": 21030
-    },
-    {
-      "epoch": 0.9751970329160872,
-      "grad_norm": 3.7243025302886963,
-      "learning_rate": 3.745594021712484e-08,
-      "loss": 0.8468,
-      "step": 21035
-    },
-    {
-      "epoch": 0.9754288363467779,
-      "grad_norm": 3.7182421684265137,
-      "learning_rate": 3.6759527973931894e-08,
-      "loss": 0.8744,
-      "step": 21040
-    },
-    {
-      "epoch": 0.9756606397774688,
-      "grad_norm": 3.662433624267578,
-      "learning_rate": 3.6069638845719347e-08,
-      "loss": 0.7838,
-      "step": 21045
-    },
-    {
-      "epoch": 0.9758924432081595,
-      "grad_norm": 3.870577096939087,
-      "learning_rate": 3.538627328417255e-08,
-      "loss": 0.7402,
-      "step": 21050
-    },
-    {
-      "epoch": 0.9761242466388502,
-      "grad_norm": 4.221757888793945,
-      "learning_rate": 3.470943173670249e-08,
-      "loss": 0.7746,
-      "step": 21055
-    },
-    {
-      "epoch": 0.9763560500695411,
-      "grad_norm": 3.4447455406188965,
-      "learning_rate": 3.4039114646449154e-08,
-      "loss": 0.6891,
-      "step": 21060
-    },
-    {
-      "epoch": 0.9765878535002318,
-      "grad_norm": 3.708122491836548,
-      "learning_rate": 3.337532245228148e-08,
-      "loss": 0.7792,
-      "step": 21065
-    },
-    {
-      "epoch": 0.9768196569309225,
-      "grad_norm": 4.2107744216918945,
-      "learning_rate": 3.271805558879737e-08,
-      "loss": 0.8349,
-      "step": 21070
-    },
-    {
-      "epoch": 0.9770514603616134,
-      "grad_norm": 4.240828037261963,
-      "learning_rate": 3.2067314486320386e-08,
-      "loss": 0.8192,
-      "step": 21075
-    },
-    {
-      "epoch": 0.9772832637923041,
-      "grad_norm": 3.749157428741455,
-      "learning_rate": 3.1423099570903057e-08,
-      "loss": 0.9146,
-      "step": 21080
-    },
-    {
-      "epoch": 0.9775150672229949,
-      "grad_norm": 5.120777130126953,
-      "learning_rate": 3.078541126432466e-08,
-      "loss": 0.9231,
-      "step": 21085
-    },
-    {
-      "epoch": 0.9777468706536857,
-      "grad_norm": 3.7648401260375977,
-      "learning_rate": 3.0154249984090114e-08,
-      "loss": 0.8833,
-      "step": 21090
-    },
-    {
-      "epoch": 0.9779786740843764,
-      "grad_norm": 3.9840023517608643,
-      "learning_rate": 2.9529616143434416e-08,
-      "loss": 0.7592,
-      "step": 21095
-    },
-    {
-      "epoch": 0.9782104775150672,
-      "grad_norm": 3.708495855331421,
-      "learning_rate": 2.8911510151314883e-08,
-      "loss": 0.7202,
-      "step": 21100
-    },
-    {
-      "epoch": 0.9782104775150672,
-      "eval_loss": 0.8723410367965698,
-      "eval_runtime": 11.2623,
-      "eval_samples_per_second": 11.277,
-      "eval_steps_per_second": 11.277,
-      "step": 21100
-    },
-    {
-      "epoch": 0.978442280945758,
-      "grad_norm": 3.607025384902954,
-      "learning_rate": 2.8299932412416686e-08,
-      "loss": 0.8083,
-      "step": 21105
-    },
-    {
-      "epoch": 0.9786740843764488,
-      "grad_norm": 3.52121901512146,
-      "learning_rate": 2.7694883327152867e-08,
-      "loss": 0.8049,
-      "step": 21110
-    },
-    {
-      "epoch": 0.9789058878071395,
-      "grad_norm": 3.9384076595306396,
-      "learning_rate": 2.7096363291657657e-08,
-      "loss": 0.9872,
-      "step": 21115
-    },
-    {
-      "epoch": 0.9791376912378303,
-      "grad_norm": 3.609402656555176,
-      "learning_rate": 2.6504372697795377e-08,
-      "loss": 0.8411,
-      "step": 21120
-    },
-    {
-      "epoch": 0.9793694946685211,
-      "grad_norm": 4.334535598754883,
-      "learning_rate": 2.5918911933151548e-08,
-      "loss": 0.8178,
-      "step": 21125
-    },
-    {
-      "epoch": 0.9796012980992118,
-      "grad_norm": 3.8995320796966553,
-      "learning_rate": 2.5339981381037326e-08,
-      "loss": 0.935,
-      "step": 21130
-    },
-    {
-      "epoch": 0.9798331015299027,
-      "grad_norm": 4.084055423736572,
-      "learning_rate": 2.4767581420491738e-08,
-      "loss": 0.96,
-      "step": 21135
-    },
-    {
-      "epoch": 0.9800649049605934,
-      "grad_norm": 4.670334815979004,
-      "learning_rate": 2.4201712426272783e-08,
-      "loss": 0.7796,
-      "step": 21140
-    },
-    {
-      "epoch": 0.9802967083912842,
-      "grad_norm": 3.5102531909942627,
-      "learning_rate": 2.3642374768866326e-08,
-      "loss": 0.8136,
-      "step": 21145
-    },
-    {
-      "epoch": 0.980528511821975,
-      "grad_norm": 4.083783149719238,
-      "learning_rate": 2.3089568814480546e-08,
-      "loss": 0.9431,
-      "step": 21150
-    },
-    {
-      "epoch": 0.9807603152526657,
-      "grad_norm": 4.43557596206665,
-      "learning_rate": 2.2543294925048142e-08,
-      "loss": 0.9378,
-      "step": 21155
-    },
-    {
-      "epoch": 0.9809921186833566,
-      "grad_norm": 3.0960304737091064,
-      "learning_rate": 2.2003553458224135e-08,
-      "loss": 0.8348,
-      "step": 21160
-    },
-    {
-      "epoch": 0.9812239221140473,
-      "grad_norm": 4.227958679199219,
-      "learning_rate": 2.1470344767386962e-08,
-      "loss": 0.849,
-      "step": 21165
-    },
-    {
-      "epoch": 0.981455725544738,
-      "grad_norm": 3.6003923416137695,
-      "learning_rate": 2.0943669201638482e-08,
-      "loss": 0.8085,
-      "step": 21170
-    },
-    {
-      "epoch": 0.9816875289754289,
-      "grad_norm": 3.7972798347473145,
-      "learning_rate": 2.0423527105802865e-08,
-      "loss": 0.8498,
-      "step": 21175
-    },
-    {
-      "epoch": 0.9819193324061196,
-      "grad_norm": 3.423555850982666,
-      "learning_rate": 1.9909918820425476e-08,
-      "loss": 0.8486,
-      "step": 21180
-    },
-    {
-      "epoch": 0.9821511358368104,
-      "grad_norm": 3.361279010772705,
-      "learning_rate": 1.9402844681777334e-08,
-      "loss": 0.819,
-      "step": 21185
-    },
-    {
-      "epoch": 0.9823829392675012,
-      "grad_norm": 3.7809674739837646,
-      "learning_rate": 1.89023050218462e-08,
-      "loss": 0.8435,
-      "step": 21190
-    },
-    {
-      "epoch": 0.9826147426981919,
-      "grad_norm": 3.679137945175171,
-      "learning_rate": 1.8408300168346605e-08,
-      "loss": 0.8949,
-      "step": 21195
-    },
-    {
-      "epoch": 0.9828465461288827,
-      "grad_norm": 3.5145680904388428,
-      "learning_rate": 1.7920830444712045e-08,
-      "loss": 0.7488,
-      "step": 21200
-    },
-    {
-      "epoch": 0.9828465461288827,
-      "eval_loss": 0.872364342212677,
-      "eval_runtime": 11.2794,
-      "eval_samples_per_second": 11.259,
-      "eval_steps_per_second": 11.259,
-      "step": 21200
-    },
-    {
-      "epoch": 0.9830783495595735,
-      "grad_norm": 3.847689628601074,
-      "learning_rate": 1.7439896170098337e-08,
-      "loss": 0.7922,
-      "step": 21205
-    },
-    {
-      "epoch": 0.9833101529902643,
-      "grad_norm": 3.7140440940856934,
-      "learning_rate": 1.696549765938027e-08,
-      "loss": 0.7843,
-      "step": 21210
-    },
-    {
-      "epoch": 0.983541956420955,
-      "grad_norm": 3.18188738822937,
-      "learning_rate": 1.649763522315606e-08,
-      "loss": 0.861,
-      "step": 21215
-    },
-    {
-      "epoch": 0.9837737598516458,
-      "grad_norm": 3.7658166885375977,
-      "learning_rate": 1.6036309167745122e-08,
-      "loss": 0.6235,
-      "step": 21220
-    },
-    {
-      "epoch": 0.9840055632823366,
-      "grad_norm": 3.599701404571533,
-      "learning_rate": 1.558151979518363e-08,
-      "loss": 0.7665,
-      "step": 21225
-    },
-    {
-      "epoch": 0.9842373667130273,
-      "grad_norm": 4.016284942626953,
-      "learning_rate": 1.5133267403232287e-08,
-      "loss": 0.9517,
-      "step": 21230
-    },
-    {
-      "epoch": 0.9844691701437182,
-      "grad_norm": 3.475090265274048,
-      "learning_rate": 1.469155228536856e-08,
-      "loss": 0.8615,
-      "step": 21235
-    },
-    {
-      "epoch": 0.9847009735744089,
-      "grad_norm": 4.081587791442871,
-      "learning_rate": 1.4256374730793333e-08,
-      "loss": 0.8005,
-      "step": 21240
-    },
-    {
-      "epoch": 0.9849327770050996,
-      "grad_norm": 3.612630605697632,
-      "learning_rate": 1.382773502442314e-08,
-      "loss": 0.8847,
-      "step": 21245
-    },
-    {
-      "epoch": 0.9851645804357905,
-      "grad_norm": 3.211310863494873,
-      "learning_rate": 1.3405633446897936e-08,
-      "loss": 0.7803,
-      "step": 21250
-    },
-    {
-      "epoch": 0.9853963838664812,
-      "grad_norm": 3.2925093173980713,
-      "learning_rate": 1.299007027457444e-08,
-      "loss": 0.8834,
-      "step": 21255
-    },
-    {
-      "epoch": 0.9856281872971719,
-      "grad_norm": 3.815549850463867,
-      "learning_rate": 1.2581045779529455e-08,
-      "loss": 0.8248,
-      "step": 21260
-    },
-    {
-      "epoch": 0.9858599907278628,
-      "grad_norm": 3.699127435684204,
-      "learning_rate": 1.2178560229558766e-08,
-      "loss": 0.9212,
-      "step": 21265
-    },
-    {
-      "epoch": 0.9860917941585535,
-      "grad_norm": 3.7043263912200928,
-      "learning_rate": 1.1782613888177142e-08,
-      "loss": 0.9181,
-      "step": 21270
-    },
-    {
-      "epoch": 0.9863235975892444,
-      "grad_norm": 4.019752502441406,
-      "learning_rate": 1.1393207014618324e-08,
-      "loss": 0.9589,
-      "step": 21275
-    },
-    {
-      "epoch": 0.9865554010199351,
-      "grad_norm": 4.226020812988281,
-      "learning_rate": 1.101033986383393e-08,
-      "loss": 0.9858,
-      "step": 21280
-    },
-    {
-      "epoch": 0.9867872044506258,
-      "grad_norm": 3.518124580383301,
-      "learning_rate": 1.0634012686493444e-08,
-      "loss": 0.8142,
-      "step": 21285
-    },
-    {
-      "epoch": 0.9870190078813167,
-      "grad_norm": 4.09403657913208,
-      "learning_rate": 1.0264225728985333e-08,
-      "loss": 0.9373,
-      "step": 21290
-    },
-    {
-      "epoch": 0.9872508113120074,
-      "grad_norm": 3.8443639278411865,
-      "learning_rate": 9.90097923341593e-09,
-      "loss": 0.8152,
-      "step": 21295
-    },
-    {
-      "epoch": 0.9874826147426982,
-      "grad_norm": 4.340132713317871,
-      "learning_rate": 9.544273437609441e-09,
-      "loss": 0.7602,
-      "step": 21300
-    },
-    {
-      "epoch": 0.9874826147426982,
-      "eval_loss": 0.8722533583641052,
-      "eval_runtime": 11.2748,
-      "eval_samples_per_second": 11.264,
-      "eval_steps_per_second": 11.264,
-      "step": 21300
-    },
-    {
-      "epoch": 0.987714418173389,
-      "grad_norm": 3.7367873191833496,
-      "learning_rate": 9.194108575106831e-09,
-      "loss": 0.8509,
-      "step": 21305
-    },
-    {
-      "epoch": 0.9879462216040797,
-      "grad_norm": 3.9204158782958984,
-      "learning_rate": 8.850484875169152e-09,
-      "loss": 0.9756,
-      "step": 21310
-    },
-    {
-      "epoch": 0.9881780250347705,
-      "grad_norm": 4.005619049072266,
-      "learning_rate": 8.513402562772e-09,
-      "loss": 0.7471,
-      "step": 21315
-    },
-    {
-      "epoch": 0.9884098284654613,
-      "grad_norm": 3.505582332611084,
-      "learning_rate": 8.182861858608837e-09,
-      "loss": 0.802,
-      "step": 21320
-    },
-    {
-      "epoch": 0.9886416318961521,
-      "grad_norm": 3.2862749099731445,
-      "learning_rate": 7.858862979092107e-09,
-      "loss": 0.8035,
-      "step": 21325
-    },
-    {
-      "epoch": 0.9888734353268428,
-      "grad_norm": 3.761301040649414,
-      "learning_rate": 7.541406136348794e-09,
-      "loss": 0.8301,
-      "step": 21330
-    },
-    {
-      "epoch": 0.9891052387575336,
-      "grad_norm": 2.987476348876953,
-      "learning_rate": 7.230491538222639e-09,
-      "loss": 0.7306,
-      "step": 21335
-    },
-    {
-      "epoch": 0.9893370421882244,
-      "grad_norm": 3.9521970748901367,
-      "learning_rate": 6.9261193882774746e-09,
-      "loss": 0.8284,
-      "step": 21340
-    },
-    {
-      "epoch": 0.9895688456189151,
-      "grad_norm": 3.9067275524139404,
-      "learning_rate": 6.628289885789452e-09,
-      "loss": 0.8549,
-      "step": 21345
-    },
-    {
-      "epoch": 0.989800649049606,
-      "grad_norm": 3.8883721828460693,
-      "learning_rate": 6.337003225753702e-09,
-      "loss": 0.8476,
-      "step": 21350
-    },
-    {
-      "epoch": 0.9900324524802967,
-      "grad_norm": 4.2398295402526855,
-      "learning_rate": 6.052259598881005e-09,
-      "loss": 0.9362,
-      "step": 21355
-    },
-    {
-      "epoch": 0.9902642559109874,
-      "grad_norm": 3.8248987197875977,
-      "learning_rate": 5.774059191597792e-09,
-      "loss": 0.8736,
-      "step": 21360
-    },
-    {
-      "epoch": 0.9904960593416783,
-      "grad_norm": 4.027286052703857,
-      "learning_rate": 5.502402186047251e-09,
-      "loss": 0.8286,
-      "step": 21365
-    },
-    {
-      "epoch": 0.990727862772369,
-      "grad_norm": 3.1771819591522217,
-      "learning_rate": 5.237288760087111e-09,
-      "loss": 0.8632,
-      "step": 21370
-    },
-    {
-      "epoch": 0.9909596662030598,
-      "grad_norm": 3.260469675064087,
-      "learning_rate": 4.9787190872940815e-09,
-      "loss": 0.769,
-      "step": 21375
-    },
-    {
-      "epoch": 0.9911914696337506,
-      "grad_norm": 3.7272191047668457,
-      "learning_rate": 4.726693336957189e-09,
-      "loss": 0.9642,
-      "step": 21380
-    },
-    {
-      "epoch": 0.9914232730644413,
-      "grad_norm": 3.7813467979431152,
-      "learning_rate": 4.48121167408111e-09,
-      "loss": 0.8887,
-      "step": 21385
-    },
-    {
-      "epoch": 0.9916550764951322,
-      "grad_norm": 3.121617078781128,
-      "learning_rate": 4.242274259389501e-09,
-      "loss": 0.775,
-      "step": 21390
-    },
-    {
-      "epoch": 0.9918868799258229,
-      "grad_norm": 3.8441171646118164,
-      "learning_rate": 4.009881249318337e-09,
-      "loss": 0.8256,
-      "step": 21395
-    },
-    {
-      "epoch": 0.9921186833565137,
-      "grad_norm": 4.062152862548828,
-      "learning_rate": 3.7840327960181336e-09,
-      "loss": 0.744,
-      "step": 21400
-    },
-    {
-      "epoch": 0.9921186833565137,
-      "eval_loss": 0.8722884654998779,
-      "eval_runtime": 11.2657,
-      "eval_samples_per_second": 11.273,
-      "eval_steps_per_second": 11.273,
-      "step": 21400
-    },
-    {
-      "epoch": 0.9923504867872045,
-      "grad_norm": 3.2572579383850098,
-      "learning_rate": 3.564729047357274e-09,
-      "loss": 0.7409,
-      "step": 21405
-    },
-    {
-      "epoch": 0.9925822902178952,
-      "grad_norm": 4.003219127655029,
-      "learning_rate": 3.3519701469175714e-09,
-      "loss": 0.8981,
-      "step": 21410
-    },
-    {
-      "epoch": 0.992814093648586,
-      "grad_norm": 3.633725881576538,
-      "learning_rate": 3.145756233996489e-09,
-      "loss": 0.7598,
-      "step": 21415
-    },
-    {
-      "epoch": 0.9930458970792768,
-      "grad_norm": 3.757777214050293,
-      "learning_rate": 2.946087443606027e-09,
-      "loss": 0.9012,
-      "step": 21420
-    },
-    {
-      "epoch": 0.9932777005099676,
-      "grad_norm": 3.0937254428863525,
-      "learning_rate": 2.7529639064716175e-09,
-      "loss": 0.7436,
-      "step": 21425
-    },
-    {
-      "epoch": 0.9935095039406583,
-      "grad_norm": 4.3624677658081055,
-      "learning_rate": 2.5663857490365597e-09,
-      "loss": 0.8341,
-      "step": 21430
-    },
-    {
-      "epoch": 0.9937413073713491,
-      "grad_norm": 3.44085693359375,
-      "learning_rate": 2.3863530934564727e-09,
-      "loss": 0.6942,
-      "step": 21435
-    },
-    {
-      "epoch": 0.9939731108020399,
-      "grad_norm": 4.770618438720703,
-      "learning_rate": 2.212866057601515e-09,
-      "loss": 0.8691,
-      "step": 21440
-    },
-    {
-      "epoch": 0.9942049142327306,
-      "grad_norm": 4.3806915283203125,
-      "learning_rate": 2.0459247550574933e-09,
-      "loss": 0.8095,
-      "step": 21445
-    },
-    {
-      "epoch": 0.9944367176634215,
-      "grad_norm": 3.6676011085510254,
-      "learning_rate": 1.885529295123645e-09,
-      "loss": 0.8438,
-      "step": 21450
-    },
-    {
-      "epoch": 0.9946685210941122,
-      "grad_norm": 4.355186939239502,
-      "learning_rate": 1.7316797828126341e-09,
-      "loss": 0.9678,
-      "step": 21455
-    },
-    {
-      "epoch": 0.9949003245248029,
-      "grad_norm": 3.5051472187042236,
-      "learning_rate": 1.5843763188538863e-09,
-      "loss": 0.8384,
-      "step": 21460
-    },
-    {
-      "epoch": 0.9951321279554938,
-      "grad_norm": 3.705413579940796,
-      "learning_rate": 1.4436189996902549e-09,
-      "loss": 0.9325,
-      "step": 21465
-    },
-    {
-      "epoch": 0.9953639313861845,
-      "grad_norm": 3.9275357723236084,
-      "learning_rate": 1.3094079174769126e-09,
-      "loss": 0.9495,
-      "step": 21470
-    },
-    {
-      "epoch": 0.9955957348168752,
-      "grad_norm": 3.814878225326538,
-      "learning_rate": 1.1817431600835706e-09,
-      "loss": 0.8951,
-      "step": 21475
-    },
-    {
-      "epoch": 0.9958275382475661,
-      "grad_norm": 3.2377772331237793,
-      "learning_rate": 1.0606248110967e-09,
-      "loss": 0.7856,
-      "step": 21480
-    },
-    {
-      "epoch": 0.9960593416782568,
-      "grad_norm": 3.7217166423797607,
-      "learning_rate": 9.460529498139803e-10,
-      "loss": 0.9289,
-      "step": 21485
-    },
-    {
-      "epoch": 0.9962911451089476,
-      "grad_norm": 3.9231810569763184,
-      "learning_rate": 8.380276512465202e-10,
-      "loss": 0.8453,
-      "step": 21490
-    },
-    {
-      "epoch": 0.9965229485396384,
-      "grad_norm": 3.294024705886841,
-      "learning_rate": 7.365489861221875e-10,
-      "loss": 0.8054,
-      "step": 21495
-    },
-    {
-      "epoch": 0.9967547519703291,
-      "grad_norm": 3.8653006553649902,
-      "learning_rate": 6.416170208789485e-10,
-      "loss": 0.8285,
-      "step": 21500
-    },
-    {
-      "epoch": 0.9967547519703291,
-      "eval_loss": 0.8723641037940979,
-      "eval_runtime": 11.2669,
-      "eval_samples_per_second": 11.272,
-      "eval_steps_per_second": 11.272,
-      "step": 21500
-    },
-    {
-      "epoch": 0.9969865554010199,
-      "grad_norm": 3.1132254600524902,
-      "learning_rate": 5.532318176726393e-10,
-      "loss": 0.8889,
-      "step": 21505
-    },
-    {
-      "epoch": 0.9972183588317107,
-      "grad_norm": 3.57822322845459,
-      "learning_rate": 4.713934343691939e-10,
-      "loss": 0.8632,
-      "step": 21510
-    },
-    {
-      "epoch": 0.9974501622624015,
-      "grad_norm": 4.240692615509033,
-      "learning_rate": 3.9610192455019624e-10,
-      "loss": 0.8483,
-      "step": 21515
-    },
-    {
-      "epoch": 0.9976819656930923,
-      "grad_norm": 4.356186866760254,
-      "learning_rate": 3.2735733751065866e-10,
-      "loss": 0.9656,
-      "step": 21520
-    },
-    {
-      "epoch": 0.997913769123783,
-      "grad_norm": 3.6947021484375,
-      "learning_rate": 2.6515971825791243e-10,
-      "loss": 0.681,
-      "step": 21525
-    },
-    {
-      "epoch": 0.9981455725544738,
-      "grad_norm": 3.716294288635254,
-      "learning_rate": 2.0950910751493802e-10,
-      "loss": 0.7692,
-      "step": 21530
-    },
-    {
-      "epoch": 0.9983773759851646,
-      "grad_norm": 4.128090858459473,
-      "learning_rate": 1.604055417170347e-10,
-      "loss": 0.8413,
-      "step": 21535
-    },
-    {
-      "epoch": 0.9986091794158554,
-      "grad_norm": 4.070624828338623,
-      "learning_rate": 1.1784905301182036e-10,
-      "loss": 0.8496,
-      "step": 21540
-    },
-    {
-      "epoch": 0.9988409828465461,
-      "grad_norm": 3.773017406463623,
-      "learning_rate": 8.183966926367249e-11,
-      "loss": 0.8795,
-      "step": 21545
-    },
-    {
-      "epoch": 0.999072786277237,
-      "grad_norm": 4.658735275268555,
-      "learning_rate": 5.237741404817698e-11,
-      "loss": 0.9189,
-      "step": 21550
-    },
-    {
-      "epoch": 0.9993045897079277,
-      "grad_norm": 3.159343957901001,
-      "learning_rate": 2.946230665434868e-11,
-      "loss": 0.8484,
-      "step": 21555
-    },
-    {
-      "epoch": 0.9995363931386184,
-      "grad_norm": 3.9690911769866943,
-      "learning_rate": 1.3094362085741553e-11,
-      "loss": 0.9825,
-      "step": 21560
-    },
-    {
-      "epoch": 0.9997681965693093,
-      "grad_norm": 3.8402626514434814,
-      "learning_rate": 3.2735910571179976e-12,
-      "loss": 0.7572,
-      "step": 21565
-    },
-    {
-      "epoch": 1.0,
-      "grad_norm": 3.908137321472168,
-      "learning_rate": 0.0,
-      "loss": 0.787,
-      "step": 21570
-    },
-    {
-      "epoch": 1.0,
-      "step": 21570,
-      "total_flos": 37930024304640.0,
-      "train_loss": 0.9211775431878377,
-      "train_runtime": 29016.4089,
-      "train_samples_per_second": 0.743,
-      "train_steps_per_second": 0.743
+      "epoch": 0.9999069680900549,
+      "step": 2687,
+      "total_flos": 8.247315480402985e+17,
+      "train_loss": 0.638783668202896,
+      "train_runtime": 14784.0018,
+      "train_samples_per_second": 1.454,
+      "train_steps_per_second": 0.182
     }
   ],
   "logging_steps": 5,
-  "max_steps": 21570,
+  "max_steps": 2687,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 1,
   "save_steps": 500,
@@ -31953,7 +4002,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 37930024304640.0,
+  "total_flos": 8.247315480402985e+17,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null