diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,14315 +1,10481 @@
 {
-  "best_metric": 0.006774452514946461,
-  "best_model_checkpoint": "/home/paperspace/Data/models/akash_unifo_757/llm3br256/checkpoint-1500",
-  "epoch": 4.996477745757285,
+  "best_metric": 0.0018192220013588667,
+  "best_model_checkpoint": "/home/paperspace/Data/models/akash_unifo_757/llm3br256/checkpoint-1300",
+  "epoch": 4.606060606060606,
   "eval_steps": 25,
-  "global_step": 1950,
+  "global_step": 1425,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.002561639449247518,
-      "grad_norm": 0.23406565189361572,
-      "learning_rate": 5.128205128205128e-07,
-      "loss": 0.066,
+      "epoch": 0.0032323232323232323,
+      "grad_norm": 0.12207423150539398,
+      "learning_rate": 1.2936610608020701e-07,
+      "loss": 0.0263,
       "step": 1
     },
     {
-      "epoch": 0.005123278898495036,
-      "grad_norm": 0.27100902795791626,
-      "learning_rate": 1.0256410256410257e-06,
-      "loss": 0.0991,
+      "epoch": 0.006464646464646465,
+      "grad_norm": 0.14607124030590057,
+      "learning_rate": 2.5873221216041403e-07,
+      "loss": 0.0254,
       "step": 2
     },
     {
-      "epoch": 0.007684918347742555,
-      "grad_norm": 0.3291971683502197,
-      "learning_rate": 1.5384615384615387e-06,
-      "loss": 0.0805,
+      "epoch": 0.009696969696969697,
+      "grad_norm": 0.14543259143829346,
+      "learning_rate": 3.8809831824062096e-07,
+      "loss": 0.0244,
       "step": 3
     },
     {
-      "epoch": 0.010246557796990073,
-      "grad_norm": 0.17330855131149292,
-      "learning_rate": 2.0512820512820513e-06,
-      "loss": 0.0754,
+      "epoch": 0.01292929292929293,
+      "grad_norm": 0.10753531754016876,
+      "learning_rate": 5.174644243208281e-07,
+      "loss": 0.0261,
       "step": 4
     },
     {
-      "epoch": 0.012808197246237591,
-      "grad_norm": 0.2650797665119171,
-      "learning_rate": 2.564102564102564e-06,
-      "loss": 0.0837,
+      "epoch": 0.01616161616161616,
+      "grad_norm": 0.15195326507091522,
+      "learning_rate": 6.468305304010349e-07,
+      "loss": 0.0293,
       "step": 5
     },
     {
-      "epoch": 0.01536983669548511,
-      "grad_norm": 0.3484267294406891,
-      "learning_rate": 3.0769230769230774e-06,
-      "loss": 0.0938,
+      "epoch": 0.019393939393939394,
+      "grad_norm": 0.11548787355422974,
+      "learning_rate": 7.761966364812419e-07,
+      "loss": 0.0254,
       "step": 6
     },
     {
-      "epoch": 0.01793147614473263,
-      "grad_norm": 0.3011722266674042,
-      "learning_rate": 3.5897435897435896e-06,
-      "loss": 0.0777,
+      "epoch": 0.022626262626262626,
+      "grad_norm": 0.1295611560344696,
+      "learning_rate": 9.055627425614489e-07,
+      "loss": 0.0258,
       "step": 7
     },
     {
-      "epoch": 0.020493115593980146,
-      "grad_norm": 0.184189110994339,
-      "learning_rate": 4.102564102564103e-06,
-      "loss": 0.0612,
+      "epoch": 0.02585858585858586,
+      "grad_norm": 0.1256393939256668,
+      "learning_rate": 1.0349288486416561e-06,
+      "loss": 0.0287,
       "step": 8
     },
     {
-      "epoch": 0.023054755043227664,
-      "grad_norm": 0.18665608763694763,
-      "learning_rate": 4.615384615384616e-06,
-      "loss": 0.0629,
+      "epoch": 0.02909090909090909,
+      "grad_norm": 0.14576725661754608,
+      "learning_rate": 1.164294954721863e-06,
+      "loss": 0.0336,
       "step": 9
     },
     {
-      "epoch": 0.025616394492475183,
-      "grad_norm": 0.20304617285728455,
-      "learning_rate": 5.128205128205128e-06,
-      "loss": 0.0628,
+      "epoch": 0.03232323232323232,
+      "grad_norm": 0.1189284399151802,
+      "learning_rate": 1.2936610608020699e-06,
+      "loss": 0.0258,
       "step": 10
     },
     {
-      "epoch": 0.0281780339417227,
-      "grad_norm": 0.2860122323036194,
-      "learning_rate": 5.641025641025641e-06,
-      "loss": 0.0792,
+      "epoch": 0.035555555555555556,
+      "grad_norm": 0.11903569847345352,
+      "learning_rate": 1.423027166882277e-06,
+      "loss": 0.0265,
       "step": 11
     },
     {
-      "epoch": 0.03073967339097022,
-      "grad_norm": 0.21296048164367676,
-      "learning_rate": 6.153846153846155e-06,
-      "loss": 0.0575,
+      "epoch": 0.03878787878787879,
+      "grad_norm": 0.11668527871370316,
+      "learning_rate": 1.5523932729624839e-06,
+      "loss": 0.0262,
       "step": 12
     },
     {
-      "epoch": 0.03330131284021774,
-      "grad_norm": 0.17642973363399506,
-      "learning_rate": 6.666666666666667e-06,
-      "loss": 0.0563,
+      "epoch": 0.04202020202020202,
+      "grad_norm": 0.11931510269641876,
+      "learning_rate": 1.6817593790426907e-06,
+      "loss": 0.0253,
       "step": 13
     },
     {
-      "epoch": 0.03586295228946526,
-      "grad_norm": 0.17329099774360657,
-      "learning_rate": 7.179487179487179e-06,
-      "loss": 0.0592,
+      "epoch": 0.04525252525252525,
+      "grad_norm": 0.10579048842191696,
+      "learning_rate": 1.8111254851228978e-06,
+      "loss": 0.025,
       "step": 14
     },
     {
-      "epoch": 0.03842459173871278,
-      "grad_norm": 0.11658388376235962,
-      "learning_rate": 7.692307692307694e-06,
-      "loss": 0.0448,
+      "epoch": 0.048484848484848485,
+      "grad_norm": 0.12630702555179596,
+      "learning_rate": 1.9404915912031045e-06,
+      "loss": 0.0248,
       "step": 15
     },
     {
-      "epoch": 0.04098623118796029,
-      "grad_norm": 0.11522124707698822,
-      "learning_rate": 8.205128205128205e-06,
-      "loss": 0.0469,
+      "epoch": 0.05171717171717172,
+      "grad_norm": 0.10258150845766068,
+      "learning_rate": 2.0698576972833122e-06,
+      "loss": 0.0195,
       "step": 16
     },
     {
-      "epoch": 0.04354787063720781,
-      "grad_norm": 0.15479238331317902,
-      "learning_rate": 8.717948717948717e-06,
-      "loss": 0.0648,
+      "epoch": 0.05494949494949495,
+      "grad_norm": 0.11002039164304733,
+      "learning_rate": 2.199223803363519e-06,
+      "loss": 0.0239,
       "step": 17
     },
     {
-      "epoch": 0.04610951008645533,
-      "grad_norm": 0.11629645526409149,
-      "learning_rate": 9.230769230769232e-06,
-      "loss": 0.061,
+      "epoch": 0.05818181818181818,
+      "grad_norm": 0.09065742790699005,
+      "learning_rate": 2.328589909443726e-06,
+      "loss": 0.0245,
       "step": 18
     },
     {
-      "epoch": 0.04867114953570285,
-      "grad_norm": 0.09216923266649246,
-      "learning_rate": 9.743589743589744e-06,
-      "loss": 0.033,
+      "epoch": 0.061414141414141414,
+      "grad_norm": 0.10896775871515274,
+      "learning_rate": 2.457956015523933e-06,
+      "loss": 0.0205,
       "step": 19
     },
     {
-      "epoch": 0.051232788984950366,
-      "grad_norm": 0.10224215686321259,
-      "learning_rate": 1.0256410256410256e-05,
-      "loss": 0.0329,
+      "epoch": 0.06464646464646465,
+      "grad_norm": 0.08943977952003479,
+      "learning_rate": 2.5873221216041398e-06,
+      "loss": 0.0191,
       "step": 20
     },
     {
-      "epoch": 0.053794428434197884,
-      "grad_norm": 0.06882911175489426,
-      "learning_rate": 1.0769230769230771e-05,
-      "loss": 0.0377,
+      "epoch": 0.06787878787878789,
+      "grad_norm": 0.08949627727270126,
+      "learning_rate": 2.716688227684347e-06,
+      "loss": 0.0183,
       "step": 21
     },
     {
-      "epoch": 0.0563560678834454,
-      "grad_norm": 0.08286888152360916,
-      "learning_rate": 1.1282051282051283e-05,
-      "loss": 0.0454,
+      "epoch": 0.07111111111111111,
+      "grad_norm": 0.07559475302696228,
+      "learning_rate": 2.846054333764554e-06,
+      "loss": 0.0198,
       "step": 22
     },
     {
-      "epoch": 0.05891770733269292,
-      "grad_norm": 0.08083554357290268,
-      "learning_rate": 1.1794871794871795e-05,
-      "loss": 0.032,
+      "epoch": 0.07434343434343435,
+      "grad_norm": 0.07494395971298218,
+      "learning_rate": 2.975420439844761e-06,
+      "loss": 0.0156,
       "step": 23
     },
     {
-      "epoch": 0.06147934678194044,
-      "grad_norm": 0.08587988466024399,
-      "learning_rate": 1.230769230769231e-05,
-      "loss": 0.0455,
+      "epoch": 0.07757575757575758,
+      "grad_norm": 0.06786634773015976,
+      "learning_rate": 3.1047865459249677e-06,
+      "loss": 0.0174,
       "step": 24
     },
     {
-      "epoch": 0.06404098623118797,
-      "grad_norm": 0.08212395757436752,
-      "learning_rate": 1.282051282051282e-05,
-      "loss": 0.0269,
+      "epoch": 0.08080808080808081,
+      "grad_norm": 0.05896899104118347,
+      "learning_rate": 3.234152652005175e-06,
+      "loss": 0.0124,
       "step": 25
     },
     {
-      "epoch": 0.06404098623118797,
-      "eval_loss": 0.03533320128917694,
-      "eval_runtime": 19.4814,
-      "eval_samples_per_second": 5.133,
-      "eval_steps_per_second": 1.283,
+      "epoch": 0.08080808080808081,
+      "eval_loss": 0.012615734711289406,
+      "eval_runtime": 18.1786,
+      "eval_samples_per_second": 5.501,
+      "eval_steps_per_second": 1.375,
       "step": 25
     },
     {
-      "epoch": 0.06660262568043548,
-      "grad_norm": 0.07568451762199402,
-      "learning_rate": 1.3333333333333333e-05,
-      "loss": 0.0326,
+      "epoch": 0.08404040404040404,
+      "grad_norm": 0.0574406273663044,
+      "learning_rate": 3.3635187580853815e-06,
+      "loss": 0.0188,
       "step": 26
     },
     {
-      "epoch": 0.069164265129683,
-      "grad_norm": 0.0810084268450737,
-      "learning_rate": 1.3846153846153847e-05,
-      "loss": 0.0367,
+      "epoch": 0.08727272727272728,
+      "grad_norm": 0.05586415156722069,
+      "learning_rate": 3.492884864165589e-06,
+      "loss": 0.012,
       "step": 27
     },
     {
-      "epoch": 0.07172590457893052,
-      "grad_norm": 0.06835253536701202,
-      "learning_rate": 1.4358974358974359e-05,
-      "loss": 0.0344,
+      "epoch": 0.0905050505050505,
+      "grad_norm": 0.05215555801987648,
+      "learning_rate": 3.6222509702457957e-06,
+      "loss": 0.0103,
       "step": 28
     },
     {
-      "epoch": 0.07428754402817804,
-      "grad_norm": 0.06612852215766907,
-      "learning_rate": 1.4871794871794872e-05,
-      "loss": 0.0267,
+      "epoch": 0.09373737373737374,
+      "grad_norm": 0.046766623854637146,
+      "learning_rate": 3.751617076326003e-06,
+      "loss": 0.0154,
       "step": 29
     },
     {
-      "epoch": 0.07684918347742556,
-      "grad_norm": 0.07679713517427444,
-      "learning_rate": 1.5384615384615387e-05,
-      "loss": 0.0293,
+      "epoch": 0.09696969696969697,
+      "grad_norm": 0.045624759048223495,
+      "learning_rate": 3.880983182406209e-06,
+      "loss": 0.0084,
       "step": 30
     },
     {
-      "epoch": 0.07941082292667308,
-      "grad_norm": 0.06691370159387589,
-      "learning_rate": 1.5897435897435897e-05,
-      "loss": 0.0349,
+      "epoch": 0.10020202020202021,
+      "grad_norm": 0.04809311404824257,
+      "learning_rate": 4.010349288486417e-06,
+      "loss": 0.0171,
       "step": 31
     },
     {
-      "epoch": 0.08197246237592058,
-      "grad_norm": 0.06954924762248993,
-      "learning_rate": 1.641025641025641e-05,
-      "loss": 0.0285,
+      "epoch": 0.10343434343434343,
+      "grad_norm": 0.03344085440039635,
+      "learning_rate": 4.1397153945666245e-06,
+      "loss": 0.0127,
       "step": 32
     },
     {
-      "epoch": 0.0845341018251681,
-      "grad_norm": 0.06169690564274788,
-      "learning_rate": 1.6923076923076924e-05,
-      "loss": 0.0289,
+      "epoch": 0.10666666666666667,
+      "grad_norm": 0.034952469170093536,
+      "learning_rate": 4.2690815006468305e-06,
+      "loss": 0.0124,
       "step": 33
     },
     {
-      "epoch": 0.08709574127441562,
-      "grad_norm": 0.05233864113688469,
-      "learning_rate": 1.7435897435897434e-05,
-      "loss": 0.0259,
+      "epoch": 0.1098989898989899,
+      "grad_norm": 0.03240083158016205,
+      "learning_rate": 4.398447606727038e-06,
+      "loss": 0.01,
       "step": 34
     },
     {
-      "epoch": 0.08965738072366314,
-      "grad_norm": 0.06114787235856056,
-      "learning_rate": 1.794871794871795e-05,
-      "loss": 0.0313,
+      "epoch": 0.11313131313131314,
+      "grad_norm": 0.04099609702825546,
+      "learning_rate": 4.527813712807244e-06,
+      "loss": 0.0127,
       "step": 35
     },
     {
-      "epoch": 0.09221902017291066,
-      "grad_norm": 0.06372269988059998,
-      "learning_rate": 1.8461538461538465e-05,
-      "loss": 0.0246,
+      "epoch": 0.11636363636363636,
+      "grad_norm": 0.030916305258870125,
+      "learning_rate": 4.657179818887452e-06,
+      "loss": 0.0113,
       "step": 36
     },
     {
-      "epoch": 0.09478065962215818,
-      "grad_norm": 0.061186905950307846,
-      "learning_rate": 1.8974358974358975e-05,
-      "loss": 0.0264,
+      "epoch": 0.1195959595959596,
+      "grad_norm": 0.025972744449973106,
+      "learning_rate": 4.786545924967659e-06,
+      "loss": 0.0113,
       "step": 37
     },
     {
-      "epoch": 0.0973422990714057,
-      "grad_norm": 0.06332549452781677,
-      "learning_rate": 1.9487179487179488e-05,
-      "loss": 0.0396,
+      "epoch": 0.12282828282828283,
+      "grad_norm": 0.03363306075334549,
+      "learning_rate": 4.915912031047866e-06,
+      "loss": 0.0138,
       "step": 38
     },
     {
-      "epoch": 0.09990393852065321,
-      "grad_norm": 0.0470692403614521,
-      "learning_rate": 2e-05,
-      "loss": 0.0274,
+      "epoch": 0.12606060606060607,
+      "grad_norm": 0.023121491074562073,
+      "learning_rate": 5.045278137128073e-06,
+      "loss": 0.0066,
       "step": 39
     },
     {
-      "epoch": 0.10246557796990073,
-      "grad_norm": 0.0600397065281868,
-      "learning_rate": 2.0512820512820512e-05,
-      "loss": 0.0373,
+      "epoch": 0.1292929292929293,
+      "grad_norm": 0.030786685645580292,
+      "learning_rate": 5.1746442432082795e-06,
+      "loss": 0.0102,
       "step": 40
     },
     {
-      "epoch": 0.10502721741914825,
-      "grad_norm": 0.05703675001859665,
-      "learning_rate": 2.102564102564103e-05,
-      "loss": 0.0223,
+      "epoch": 0.13252525252525252,
+      "grad_norm": 0.033025529235601425,
+      "learning_rate": 5.304010349288486e-06,
+      "loss": 0.01,
       "step": 41
     },
     {
-      "epoch": 0.10758885686839577,
-      "grad_norm": 0.07255468517541885,
-      "learning_rate": 2.1538461538461542e-05,
-      "loss": 0.0424,
+      "epoch": 0.13575757575757577,
+      "grad_norm": 0.023587767034769058,
+      "learning_rate": 5.433376455368694e-06,
+      "loss": 0.0083,
       "step": 42
     },
     {
-      "epoch": 0.11015049631764329,
-      "grad_norm": 0.06241394206881523,
-      "learning_rate": 2.2051282051282052e-05,
-      "loss": 0.0313,
+      "epoch": 0.138989898989899,
+      "grad_norm": 0.027042975649237633,
+      "learning_rate": 5.5627425614489e-06,
+      "loss": 0.0111,
       "step": 43
     },
     {
-      "epoch": 0.1127121357668908,
-      "grad_norm": 0.056400399655103683,
-      "learning_rate": 2.2564102564102566e-05,
-      "loss": 0.0196,
+      "epoch": 0.14222222222222222,
+      "grad_norm": 0.031232647597789764,
+      "learning_rate": 5.692108667529108e-06,
+      "loss": 0.0127,
       "step": 44
     },
     {
-      "epoch": 0.11527377521613832,
-      "grad_norm": 0.04130598530173302,
-      "learning_rate": 2.307692307692308e-05,
-      "loss": 0.0243,
+      "epoch": 0.14545454545454545,
+      "grad_norm": 0.025501983240246773,
+      "learning_rate": 5.821474773609315e-06,
+      "loss": 0.0082,
       "step": 45
     },
     {
-      "epoch": 0.11783541466538584,
-      "grad_norm": 0.05306409299373627,
-      "learning_rate": 2.358974358974359e-05,
-      "loss": 0.0274,
+      "epoch": 0.1486868686868687,
+      "grad_norm": 0.024088839069008827,
+      "learning_rate": 5.950840879689522e-06,
+      "loss": 0.0086,
       "step": 46
     },
     {
-      "epoch": 0.12039705411463336,
-      "grad_norm": 0.04285677894949913,
-      "learning_rate": 2.4102564102564103e-05,
-      "loss": 0.0161,
+      "epoch": 0.15191919191919193,
+      "grad_norm": 0.02032627910375595,
+      "learning_rate": 6.0802069857697286e-06,
+      "loss": 0.0067,
       "step": 47
     },
     {
-      "epoch": 0.12295869356388088,
-      "grad_norm": 0.05328772962093353,
-      "learning_rate": 2.461538461538462e-05,
-      "loss": 0.0243,
+      "epoch": 0.15515151515151515,
+      "grad_norm": 0.018834169954061508,
+      "learning_rate": 6.2095730918499354e-06,
+      "loss": 0.0066,
       "step": 48
     },
     {
-      "epoch": 0.1255203330131284,
-      "grad_norm": 0.05073712766170502,
-      "learning_rate": 2.512820512820513e-05,
-      "loss": 0.0206,
+      "epoch": 0.15838383838383838,
+      "grad_norm": 0.0212345439940691,
+      "learning_rate": 6.338939197930142e-06,
+      "loss": 0.0079,
       "step": 49
     },
     {
-      "epoch": 0.12808197246237593,
-      "grad_norm": 0.05035106837749481,
-      "learning_rate": 2.564102564102564e-05,
-      "loss": 0.0361,
+      "epoch": 0.16161616161616163,
+      "grad_norm": 0.020874127745628357,
+      "learning_rate": 6.46830530401035e-06,
+      "loss": 0.0042,
       "step": 50
     },
     {
-      "epoch": 0.12808197246237593,
-      "eval_loss": 0.02137046866118908,
-      "eval_runtime": 15.8283,
-      "eval_samples_per_second": 6.318,
-      "eval_steps_per_second": 1.579,
+      "epoch": 0.16161616161616163,
+      "eval_loss": 0.006814153399318457,
+      "eval_runtime": 15.4553,
+      "eval_samples_per_second": 6.47,
+      "eval_steps_per_second": 1.618,
       "step": 50
     },
     {
-      "epoch": 0.13064361191162344,
-      "grad_norm": 0.0659724548459053,
-      "learning_rate": 2.6153846153846157e-05,
-      "loss": 0.031,
+      "epoch": 0.16484848484848486,
+      "grad_norm": 0.021333854645490646,
+      "learning_rate": 6.597671410090557e-06,
+      "loss": 0.0075,
       "step": 51
     },
     {
-      "epoch": 0.13320525136087097,
-      "grad_norm": 0.04782676696777344,
-      "learning_rate": 2.6666666666666667e-05,
-      "loss": 0.0238,
+      "epoch": 0.16808080808080808,
+      "grad_norm": 0.02766190655529499,
+      "learning_rate": 6.727037516170763e-06,
+      "loss": 0.0125,
       "step": 52
     },
     {
-      "epoch": 0.13576689081011847,
-      "grad_norm": 0.05571295693516731,
-      "learning_rate": 2.717948717948718e-05,
-      "loss": 0.0258,
+      "epoch": 0.1713131313131313,
+      "grad_norm": 0.022314516827464104,
+      "learning_rate": 6.856403622250971e-06,
+      "loss": 0.0094,
       "step": 53
     },
     {
-      "epoch": 0.138328530259366,
-      "grad_norm": 0.05261208489537239,
-      "learning_rate": 2.7692307692307694e-05,
-      "loss": 0.0221,
+      "epoch": 0.17454545454545456,
+      "grad_norm": 0.02085626684129238,
+      "learning_rate": 6.985769728331178e-06,
+      "loss": 0.0061,
       "step": 54
     },
     {
-      "epoch": 0.1408901697086135,
-      "grad_norm": 0.053560130298137665,
-      "learning_rate": 2.8205128205128207e-05,
-      "loss": 0.032,
+      "epoch": 0.17777777777777778,
+      "grad_norm": 0.02437152899801731,
+      "learning_rate": 7.115135834411385e-06,
+      "loss": 0.0096,
       "step": 55
     },
     {
-      "epoch": 0.14345180915786104,
-      "grad_norm": 0.0561230406165123,
-      "learning_rate": 2.8717948717948717e-05,
-      "loss": 0.0165,
+      "epoch": 0.181010101010101,
+      "grad_norm": 0.023625781759619713,
+      "learning_rate": 7.244501940491591e-06,
+      "loss": 0.0115,
       "step": 56
     },
     {
-      "epoch": 0.14601344860710855,
-      "grad_norm": 0.05372655764222145,
-      "learning_rate": 2.9230769230769234e-05,
-      "loss": 0.0295,
+      "epoch": 0.18424242424242424,
+      "grad_norm": 0.019835669547319412,
+      "learning_rate": 7.373868046571798e-06,
+      "loss": 0.0053,
       "step": 57
     },
     {
-      "epoch": 0.14857508805635608,
-      "grad_norm": 0.05659542977809906,
-      "learning_rate": 2.9743589743589744e-05,
-      "loss": 0.0229,
+      "epoch": 0.1874747474747475,
+      "grad_norm": 0.020804336294531822,
+      "learning_rate": 7.503234152652006e-06,
+      "loss": 0.0106,
       "step": 58
     },
     {
-      "epoch": 0.15113672750560359,
-      "grad_norm": 0.051467038691043854,
-      "learning_rate": 3.0256410256410257e-05,
-      "loss": 0.0192,
+      "epoch": 0.1907070707070707,
+      "grad_norm": 0.02094445377588272,
+      "learning_rate": 7.632600258732213e-06,
+      "loss": 0.0097,
       "step": 59
     },
     {
-      "epoch": 0.15369836695485112,
-      "grad_norm": 0.0462658628821373,
-      "learning_rate": 3.0769230769230774e-05,
-      "loss": 0.0236,
+      "epoch": 0.19393939393939394,
+      "grad_norm": 0.01580234244465828,
+      "learning_rate": 7.761966364812418e-06,
+      "loss": 0.0086,
       "step": 60
     },
     {
-      "epoch": 0.15626000640409862,
-      "grad_norm": 0.0534682422876358,
-      "learning_rate": 3.128205128205128e-05,
-      "loss": 0.027,
+      "epoch": 0.19717171717171716,
+      "grad_norm": 0.016111252829432487,
+      "learning_rate": 7.891332470892627e-06,
+      "loss": 0.0079,
       "step": 61
     },
     {
-      "epoch": 0.15882164585334616,
-      "grad_norm": 0.045332759618759155,
-      "learning_rate": 3.1794871794871795e-05,
-      "loss": 0.0291,
+      "epoch": 0.20040404040404042,
+      "grad_norm": 0.017682602629065514,
+      "learning_rate": 8.020698576972833e-06,
+      "loss": 0.0065,
       "step": 62
     },
     {
-      "epoch": 0.16138328530259366,
-      "grad_norm": 0.04076462239027023,
-      "learning_rate": 3.230769230769231e-05,
-      "loss": 0.0245,
+      "epoch": 0.20363636363636364,
+      "grad_norm": 0.014977304264903069,
+      "learning_rate": 8.15006468305304e-06,
+      "loss": 0.0062,
       "step": 63
     },
     {
-      "epoch": 0.16394492475184116,
-      "grad_norm": 0.036102596670389175,
-      "learning_rate": 3.282051282051282e-05,
-      "loss": 0.0153,
+      "epoch": 0.20686868686868687,
+      "grad_norm": 0.027255840599536896,
+      "learning_rate": 8.279430789133249e-06,
+      "loss": 0.0178,
       "step": 64
     },
     {
-      "epoch": 0.1665065642010887,
-      "grad_norm": 0.06842701137065887,
-      "learning_rate": 3.3333333333333335e-05,
-      "loss": 0.0506,
+      "epoch": 0.2101010101010101,
+      "grad_norm": 0.01817743107676506,
+      "learning_rate": 8.408796895213454e-06,
+      "loss": 0.0075,
       "step": 65
     },
     {
-      "epoch": 0.1690682036503362,
-      "grad_norm": 0.04626695066690445,
-      "learning_rate": 3.384615384615385e-05,
-      "loss": 0.0331,
+      "epoch": 0.21333333333333335,
+      "grad_norm": 0.014269635081291199,
+      "learning_rate": 8.538163001293661e-06,
+      "loss": 0.0044,
       "step": 66
     },
     {
-      "epoch": 0.17162984309958373,
-      "grad_norm": 0.058188967406749725,
-      "learning_rate": 3.435897435897436e-05,
-      "loss": 0.0171,
+      "epoch": 0.21656565656565657,
+      "grad_norm": 0.022783976048231125,
+      "learning_rate": 8.66752910737387e-06,
+      "loss": 0.0147,
       "step": 67
     },
     {
-      "epoch": 0.17419148254883124,
-      "grad_norm": 0.03610702231526375,
-      "learning_rate": 3.487179487179487e-05,
-      "loss": 0.0246,
+      "epoch": 0.2197979797979798,
+      "grad_norm": 0.02248130738735199,
+      "learning_rate": 8.796895213454076e-06,
+      "loss": 0.0121,
       "step": 68
     },
     {
-      "epoch": 0.17675312199807877,
-      "grad_norm": 0.043155767023563385,
-      "learning_rate": 3.538461538461539e-05,
-      "loss": 0.0157,
+      "epoch": 0.22303030303030302,
+      "grad_norm": 0.014569518156349659,
+      "learning_rate": 8.926261319534282e-06,
+      "loss": 0.0042,
       "step": 69
     },
     {
-      "epoch": 0.17931476144732628,
-      "grad_norm": 0.04548976942896843,
-      "learning_rate": 3.58974358974359e-05,
-      "loss": 0.0298,
+      "epoch": 0.22626262626262628,
+      "grad_norm": 0.014306961558759212,
+      "learning_rate": 9.055627425614489e-06,
+      "loss": 0.0075,
       "step": 70
     },
     {
-      "epoch": 0.1818764008965738,
-      "grad_norm": 0.04875415563583374,
-      "learning_rate": 3.641025641025641e-05,
-      "loss": 0.0217,
+      "epoch": 0.2294949494949495,
+      "grad_norm": 0.021996863186359406,
+      "learning_rate": 9.184993531694697e-06,
+      "loss": 0.0155,
       "step": 71
     },
     {
-      "epoch": 0.1844380403458213,
-      "grad_norm": 0.053107842803001404,
-      "learning_rate": 3.692307692307693e-05,
-      "loss": 0.0184,
+      "epoch": 0.23272727272727273,
+      "grad_norm": 0.015790125355124474,
+      "learning_rate": 9.314359637774904e-06,
+      "loss": 0.0063,
       "step": 72
     },
     {
-      "epoch": 0.18699967979506885,
-      "grad_norm": 0.04690933972597122,
-      "learning_rate": 3.7435897435897436e-05,
-      "loss": 0.0162,
+      "epoch": 0.23595959595959595,
+      "grad_norm": 0.017747273668646812,
+      "learning_rate": 9.44372574385511e-06,
+      "loss": 0.0097,
       "step": 73
     },
     {
-      "epoch": 0.18956131924431635,
-      "grad_norm": 0.04177139699459076,
-      "learning_rate": 3.794871794871795e-05,
-      "loss": 0.0201,
+      "epoch": 0.2391919191919192,
+      "grad_norm": 0.019176697358489037,
+      "learning_rate": 9.573091849935318e-06,
+      "loss": 0.0105,
       "step": 74
     },
     {
-      "epoch": 0.19212295869356388,
-      "grad_norm": 0.07623627781867981,
-      "learning_rate": 3.846153846153846e-05,
-      "loss": 0.0301,
+      "epoch": 0.24242424242424243,
+      "grad_norm": 0.016284696757793427,
+      "learning_rate": 9.702457956015525e-06,
+      "loss": 0.0074,
       "step": 75
     },
     {
-      "epoch": 0.19212295869356388,
-      "eval_loss": 0.01890488900244236,
-      "eval_runtime": 15.7788,
-      "eval_samples_per_second": 6.338,
-      "eval_steps_per_second": 1.584,
+      "epoch": 0.24242424242424243,
+      "eval_loss": 0.005784325767308474,
+      "eval_runtime": 15.5689,
+      "eval_samples_per_second": 6.423,
+      "eval_steps_per_second": 1.606,
       "step": 75
     },
     {
-      "epoch": 0.1946845981428114,
-      "grad_norm": 0.05075330659747124,
-      "learning_rate": 3.8974358974358976e-05,
-      "loss": 0.0267,
+      "epoch": 0.24565656565656566,
+      "grad_norm": 0.017876651138067245,
+      "learning_rate": 9.831824062095732e-06,
+      "loss": 0.0073,
       "step": 76
     },
     {
-      "epoch": 0.19724623759205892,
-      "grad_norm": 0.06402429938316345,
-      "learning_rate": 3.948717948717949e-05,
-      "loss": 0.0324,
+      "epoch": 0.24888888888888888,
+      "grad_norm": 0.016466056928038597,
+      "learning_rate": 9.961190168175938e-06,
+      "loss": 0.0096,
       "step": 77
     },
     {
-      "epoch": 0.19980787704130643,
-      "grad_norm": 0.041263386607170105,
-      "learning_rate": 4e-05,
-      "loss": 0.0187,
+      "epoch": 0.25212121212121213,
+      "grad_norm": 0.019224438816308975,
+      "learning_rate": 1.0090556274256145e-05,
+      "loss": 0.0136,
       "step": 78
     },
     {
-      "epoch": 0.20236951649055396,
-      "grad_norm": 0.051044903695583344,
-      "learning_rate": 4.051282051282052e-05,
-      "loss": 0.0289,
+      "epoch": 0.25535353535353533,
+      "grad_norm": 0.014781850390136242,
+      "learning_rate": 1.0219922380336352e-05,
+      "loss": 0.0067,
       "step": 79
     },
     {
-      "epoch": 0.20493115593980146,
-      "grad_norm": 0.06309404224157333,
-      "learning_rate": 4.1025641025641023e-05,
-      "loss": 0.0323,
+      "epoch": 0.2585858585858586,
+      "grad_norm": 0.01763291470706463,
+      "learning_rate": 1.0349288486416559e-05,
+      "loss": 0.0088,
       "step": 80
     },
     {
-      "epoch": 0.207492795389049,
-      "grad_norm": 0.04254679009318352,
-      "learning_rate": 4.1538461538461544e-05,
-      "loss": 0.0212,
+      "epoch": 0.26181818181818184,
+      "grad_norm": 0.019103335216641426,
+      "learning_rate": 1.0478654592496766e-05,
+      "loss": 0.0097,
       "step": 81
     },
     {
-      "epoch": 0.2100544348382965,
-      "grad_norm": 0.03591303154826164,
-      "learning_rate": 4.205128205128206e-05,
-      "loss": 0.0204,
+      "epoch": 0.26505050505050504,
+      "grad_norm": 0.016623804345726967,
+      "learning_rate": 1.0608020698576973e-05,
+      "loss": 0.0087,
       "step": 82
     },
     {
-      "epoch": 0.21261607428754403,
-      "grad_norm": 0.047834377735853195,
-      "learning_rate": 4.2564102564102564e-05,
-      "loss": 0.0201,
+      "epoch": 0.2682828282828283,
+      "grad_norm": 0.015380220487713814,
+      "learning_rate": 1.073738680465718e-05,
+      "loss": 0.0043,
       "step": 83
     },
     {
-      "epoch": 0.21517771373679154,
-      "grad_norm": 0.04303661361336708,
-      "learning_rate": 4.3076923076923084e-05,
-      "loss": 0.0207,
+      "epoch": 0.27151515151515154,
+      "grad_norm": 0.015298287384212017,
+      "learning_rate": 1.0866752910737388e-05,
+      "loss": 0.0063,
       "step": 84
     },
     {
-      "epoch": 0.21773935318603907,
-      "grad_norm": 0.03778654336929321,
-      "learning_rate": 4.358974358974359e-05,
-      "loss": 0.0138,
+      "epoch": 0.27474747474747474,
+      "grad_norm": 0.01960132084786892,
+      "learning_rate": 1.0996119016817593e-05,
+      "loss": 0.0096,
       "step": 85
     },
     {
-      "epoch": 0.22030099263528657,
-      "grad_norm": 0.059220101684331894,
-      "learning_rate": 4.4102564102564104e-05,
-      "loss": 0.0294,
+      "epoch": 0.277979797979798,
+      "grad_norm": 0.01212010532617569,
+      "learning_rate": 1.11254851228978e-05,
+      "loss": 0.0034,
       "step": 86
     },
     {
-      "epoch": 0.2228626320845341,
-      "grad_norm": 0.06273186206817627,
-      "learning_rate": 4.461538461538462e-05,
-      "loss": 0.0194,
+      "epoch": 0.2812121212121212,
+      "grad_norm": 0.0169272031635046,
+      "learning_rate": 1.1254851228978009e-05,
+      "loss": 0.0052,
       "step": 87
     },
     {
-      "epoch": 0.2254242715337816,
-      "grad_norm": 0.05138327553868294,
-      "learning_rate": 4.512820512820513e-05,
-      "loss": 0.0257,
+      "epoch": 0.28444444444444444,
+      "grad_norm": 0.017383215948939323,
+      "learning_rate": 1.1384217335058216e-05,
+      "loss": 0.0077,
       "step": 88
     },
     {
-      "epoch": 0.22798591098302914,
-      "grad_norm": 0.05410440266132355,
-      "learning_rate": 4.5641025641025645e-05,
-      "loss": 0.0203,
+      "epoch": 0.2876767676767677,
+      "grad_norm": 0.01304047554731369,
+      "learning_rate": 1.1513583441138421e-05,
+      "loss": 0.0064,
       "step": 89
     },
     {
-      "epoch": 0.23054755043227665,
-      "grad_norm": 0.04224589839577675,
-      "learning_rate": 4.615384615384616e-05,
-      "loss": 0.0181,
+      "epoch": 0.2909090909090909,
+      "grad_norm": 0.012733686715364456,
+      "learning_rate": 1.164294954721863e-05,
+      "loss": 0.0035,
       "step": 90
     },
     {
-      "epoch": 0.23310918988152418,
-      "grad_norm": 0.060652345418930054,
-      "learning_rate": 4.666666666666667e-05,
-      "loss": 0.0252,
+      "epoch": 0.29414141414141415,
+      "grad_norm": 0.011904511600732803,
+      "learning_rate": 1.1772315653298836e-05,
+      "loss": 0.0035,
       "step": 91
     },
     {
-      "epoch": 0.2356708293307717,
-      "grad_norm": 0.06052101030945778,
-      "learning_rate": 4.717948717948718e-05,
-      "loss": 0.0326,
+      "epoch": 0.2973737373737374,
+      "grad_norm": 0.017259540036320686,
+      "learning_rate": 1.1901681759379043e-05,
+      "loss": 0.007,
       "step": 92
     },
     {
-      "epoch": 0.23823246878001922,
-      "grad_norm": 0.050801996141672134,
-      "learning_rate": 4.76923076923077e-05,
-      "loss": 0.0282,
+      "epoch": 0.3006060606060606,
+      "grad_norm": 0.016907405108213425,
+      "learning_rate": 1.203104786545925e-05,
+      "loss": 0.0066,
       "step": 93
     },
     {
-      "epoch": 0.24079410822926672,
-      "grad_norm": 0.042453132569789886,
-      "learning_rate": 4.8205128205128205e-05,
-      "loss": 0.0166,
+      "epoch": 0.30383838383838385,
+      "grad_norm": 0.01367953885346651,
+      "learning_rate": 1.2160413971539457e-05,
+      "loss": 0.0048,
       "step": 94
     },
     {
-      "epoch": 0.24335574767851426,
-      "grad_norm": 0.037676140666007996,
-      "learning_rate": 4.871794871794872e-05,
-      "loss": 0.0188,
+      "epoch": 0.30707070707070705,
+      "grad_norm": 0.01934090442955494,
+      "learning_rate": 1.2289780077619664e-05,
+      "loss": 0.0099,
       "step": 95
     },
     {
-      "epoch": 0.24591738712776176,
-      "grad_norm": 0.037714943289756775,
-      "learning_rate": 4.923076923076924e-05,
-      "loss": 0.0141,
+      "epoch": 0.3103030303030303,
+      "grad_norm": 0.014079134911298752,
+      "learning_rate": 1.2419146183699871e-05,
+      "loss": 0.0048,
       "step": 96
     },
     {
-      "epoch": 0.2484790265770093,
-      "grad_norm": 0.04738244786858559,
-      "learning_rate": 4.9743589743589746e-05,
-      "loss": 0.0348,
+      "epoch": 0.31353535353535356,
+      "grad_norm": 0.01526681799441576,
+      "learning_rate": 1.254851228978008e-05,
+      "loss": 0.0055,
       "step": 97
     },
     {
-      "epoch": 0.2510406660262568,
-      "grad_norm": 0.039117615669965744,
-      "learning_rate": 5.025641025641026e-05,
-      "loss": 0.0184,
+      "epoch": 0.31676767676767675,
+      "grad_norm": 0.015970032662153244,
+      "learning_rate": 1.2677878395860285e-05,
+      "loss": 0.0053,
       "step": 98
     },
     {
-      "epoch": 0.25360230547550433,
-      "grad_norm": 0.03788371384143829,
-      "learning_rate": 5.0769230769230766e-05,
-      "loss": 0.0192,
+      "epoch": 0.32,
+      "grad_norm": 0.014750220812857151,
+      "learning_rate": 1.2807244501940493e-05,
+      "loss": 0.0048,
       "step": 99
     },
     {
-      "epoch": 0.25616394492475186,
-      "grad_norm": 0.07863015681505203,
-      "learning_rate": 5.128205128205128e-05,
-      "loss": 0.0217,
+      "epoch": 0.32323232323232326,
+      "grad_norm": 0.014153756201267242,
+      "learning_rate": 1.29366106080207e-05,
+      "loss": 0.0052,
       "step": 100
     },
     {
-      "epoch": 0.25616394492475186,
-      "eval_loss": 0.017253482714295387,
-      "eval_runtime": 15.7742,
-      "eval_samples_per_second": 6.339,
-      "eval_steps_per_second": 1.585,
+      "epoch": 0.32323232323232326,
+      "eval_loss": 0.005152097903192043,
+      "eval_runtime": 15.5449,
+      "eval_samples_per_second": 6.433,
+      "eval_steps_per_second": 1.608,
       "step": 100
     },
     {
-      "epoch": 0.25872558437399934,
-      "grad_norm": 0.04783935844898224,
-      "learning_rate": 5.17948717948718e-05,
-      "loss": 0.0193,
+      "epoch": 0.32646464646464646,
+      "grad_norm": 0.02274342067539692,
+      "learning_rate": 1.3065976714100905e-05,
+      "loss": 0.0073,
       "step": 101
     },
     {
-      "epoch": 0.2612872238232469,
-      "grad_norm": 0.04175376892089844,
-      "learning_rate": 5.230769230769231e-05,
-      "loss": 0.0155,
+      "epoch": 0.3296969696969697,
+      "grad_norm": 0.014152687042951584,
+      "learning_rate": 1.3195342820181114e-05,
+      "loss": 0.0047,
       "step": 102
     },
     {
-      "epoch": 0.2638488632724944,
-      "grad_norm": 0.05416353419423103,
-      "learning_rate": 5.2820512820512826e-05,
-      "loss": 0.0317,
+      "epoch": 0.3329292929292929,
+      "grad_norm": 0.02202792838215828,
+      "learning_rate": 1.332470892626132e-05,
+      "loss": 0.0091,
       "step": 103
     },
     {
-      "epoch": 0.26641050272174194,
-      "grad_norm": 0.028790106996893883,
-      "learning_rate": 5.333333333333333e-05,
-      "loss": 0.0131,
+      "epoch": 0.33616161616161616,
+      "grad_norm": 0.013995049521327019,
+      "learning_rate": 1.3454075032341526e-05,
+      "loss": 0.0047,
       "step": 104
     },
     {
-      "epoch": 0.2689721421709894,
-      "grad_norm": 0.057330165058374405,
-      "learning_rate": 5.384615384615385e-05,
-      "loss": 0.0176,
+      "epoch": 0.3393939393939394,
+      "grad_norm": 0.016882814466953278,
+      "learning_rate": 1.3583441138421735e-05,
+      "loss": 0.0077,
       "step": 105
     },
     {
-      "epoch": 0.27153378162023695,
-      "grad_norm": 0.07159876078367233,
-      "learning_rate": 5.435897435897436e-05,
-      "loss": 0.0228,
+      "epoch": 0.3426262626262626,
+      "grad_norm": 0.015208212658762932,
+      "learning_rate": 1.3712807244501941e-05,
+      "loss": 0.007,
       "step": 106
     },
     {
-      "epoch": 0.2740954210694845,
-      "grad_norm": 0.04449112340807915,
-      "learning_rate": 5.487179487179488e-05,
-      "loss": 0.0202,
+      "epoch": 0.34585858585858587,
+      "grad_norm": 0.011098811402916908,
+      "learning_rate": 1.384217335058215e-05,
+      "loss": 0.0053,
       "step": 107
     },
     {
-      "epoch": 0.276657060518732,
-      "grad_norm": 0.06222032755613327,
-      "learning_rate": 5.538461538461539e-05,
-      "loss": 0.0298,
+      "epoch": 0.3490909090909091,
+      "grad_norm": 0.01402447558939457,
+      "learning_rate": 1.3971539456662355e-05,
+      "loss": 0.0048,
       "step": 108
     },
     {
-      "epoch": 0.2792186999679795,
-      "grad_norm": 0.04552404209971428,
-      "learning_rate": 5.58974358974359e-05,
-      "loss": 0.0149,
+      "epoch": 0.3523232323232323,
+      "grad_norm": 0.01941692642867565,
+      "learning_rate": 1.4100905562742562e-05,
+      "loss": 0.0075,
       "step": 109
     },
     {
-      "epoch": 0.281780339417227,
-      "grad_norm": 0.03565036877989769,
-      "learning_rate": 5.6410256410256414e-05,
-      "loss": 0.0162,
+      "epoch": 0.35555555555555557,
+      "grad_norm": 0.012594842351973057,
+      "learning_rate": 1.423027166882277e-05,
+      "loss": 0.004,
       "step": 110
     },
     {
-      "epoch": 0.28434197886647455,
-      "grad_norm": 0.04282074794173241,
-      "learning_rate": 5.692307692307692e-05,
-      "loss": 0.0191,
+      "epoch": 0.35878787878787877,
+      "grad_norm": 0.016337120905518532,
+      "learning_rate": 1.4359637774902976e-05,
+      "loss": 0.0095,
       "step": 111
     },
     {
-      "epoch": 0.2869036183157221,
-      "grad_norm": 0.0391264371573925,
-      "learning_rate": 5.7435897435897434e-05,
-      "loss": 0.0144,
+      "epoch": 0.362020202020202,
+      "grad_norm": 0.01168188638985157,
+      "learning_rate": 1.4489003880983183e-05,
+      "loss": 0.0032,
       "step": 112
     },
     {
-      "epoch": 0.28946525776496956,
-      "grad_norm": 0.048470303416252136,
-      "learning_rate": 5.7948717948717954e-05,
-      "loss": 0.0305,
+      "epoch": 0.3652525252525253,
+      "grad_norm": 0.013965807855129242,
+      "learning_rate": 1.4618369987063391e-05,
+      "loss": 0.003,
       "step": 113
     },
     {
-      "epoch": 0.2920268972142171,
-      "grad_norm": 0.04939977452158928,
-      "learning_rate": 5.846153846153847e-05,
-      "loss": 0.0239,
+      "epoch": 0.36848484848484847,
+      "grad_norm": 0.013490543700754642,
+      "learning_rate": 1.4747736093143596e-05,
+      "loss": 0.0055,
       "step": 114
     },
     {
-      "epoch": 0.29458853666346463,
-      "grad_norm": 0.0445936843752861,
-      "learning_rate": 5.897435897435898e-05,
-      "loss": 0.0155,
+      "epoch": 0.3717171717171717,
+      "grad_norm": 0.012756850570440292,
+      "learning_rate": 1.4877102199223805e-05,
+      "loss": 0.0042,
       "step": 115
     },
     {
-      "epoch": 0.29715017611271216,
-      "grad_norm": 0.05109110474586487,
-      "learning_rate": 5.948717948717949e-05,
-      "loss": 0.0251,
+      "epoch": 0.374949494949495,
+      "grad_norm": 0.014388620853424072,
+      "learning_rate": 1.5006468305304012e-05,
+      "loss": 0.0056,
       "step": 116
     },
     {
-      "epoch": 0.29971181556195964,
-      "grad_norm": 0.044525668025016785,
-      "learning_rate": 6e-05,
-      "loss": 0.0274,
+      "epoch": 0.3781818181818182,
+      "grad_norm": 0.019619980826973915,
+      "learning_rate": 1.5135834411384217e-05,
+      "loss": 0.0097,
       "step": 117
     },
     {
-      "epoch": 0.30227345501120717,
-      "grad_norm": 0.051556315273046494,
-      "learning_rate": 6.0512820512820515e-05,
-      "loss": 0.0171,
+      "epoch": 0.3814141414141414,
+      "grad_norm": 0.012313769198954105,
+      "learning_rate": 1.5265200517464426e-05,
+      "loss": 0.0038,
       "step": 118
     },
     {
-      "epoch": 0.3048350944604547,
-      "grad_norm": 0.06288602948188782,
-      "learning_rate": 6.1025641025641035e-05,
-      "loss": 0.0277,
+      "epoch": 0.3846464646464646,
+      "grad_norm": 0.015437609516084194,
+      "learning_rate": 1.5394566623544633e-05,
+      "loss": 0.0052,
       "step": 119
     },
     {
-      "epoch": 0.30739673390970224,
-      "grad_norm": 0.03961167111992836,
-      "learning_rate": 6.153846153846155e-05,
-      "loss": 0.0166,
+      "epoch": 0.3878787878787879,
+      "grad_norm": 0.014744014479219913,
+      "learning_rate": 1.5523932729624836e-05,
+      "loss": 0.0075,
       "step": 120
     },
     {
-      "epoch": 0.3099583733589497,
-      "grad_norm": 0.05251983925700188,
-      "learning_rate": 6.205128205128206e-05,
-      "loss": 0.0294,
+      "epoch": 0.39111111111111113,
+      "grad_norm": 0.01886366680264473,
+      "learning_rate": 1.5653298835705046e-05,
+      "loss": 0.0082,
       "step": 121
     },
     {
-      "epoch": 0.31252001280819725,
-      "grad_norm": 0.03636966273188591,
-      "learning_rate": 6.256410256410256e-05,
-      "loss": 0.0126,
+      "epoch": 0.39434343434343433,
+      "grad_norm": 0.01593458466231823,
+      "learning_rate": 1.5782664941785253e-05,
+      "loss": 0.0048,
       "step": 122
     },
     {
-      "epoch": 0.3150816522574448,
-      "grad_norm": 0.047448430210351944,
-      "learning_rate": 6.307692307692308e-05,
-      "loss": 0.0201,
+      "epoch": 0.3975757575757576,
+      "grad_norm": 0.015406963415443897,
+      "learning_rate": 1.591203104786546e-05,
+      "loss": 0.0057,
       "step": 123
     },
     {
-      "epoch": 0.3176432917066923,
-      "grad_norm": 0.05578154698014259,
-      "learning_rate": 6.358974358974359e-05,
-      "loss": 0.0227,
+      "epoch": 0.40080808080808084,
+      "grad_norm": 0.013901729136705399,
+      "learning_rate": 1.6041397153945667e-05,
+      "loss": 0.0056,
       "step": 124
     },
     {
-      "epoch": 0.3202049311559398,
-      "grad_norm": 0.08173694461584091,
-      "learning_rate": 6.410256410256412e-05,
-      "loss": 0.0186,
+      "epoch": 0.40404040404040403,
+      "grad_norm": 0.013464200310409069,
+      "learning_rate": 1.6170763260025874e-05,
+      "loss": 0.0041,
       "step": 125
     },
     {
-      "epoch": 0.3202049311559398,
-      "eval_loss": 0.016197798773646355,
-      "eval_runtime": 15.7682,
-      "eval_samples_per_second": 6.342,
-      "eval_steps_per_second": 1.585,
+      "epoch": 0.40404040404040403,
+      "eval_loss": 0.004761708900332451,
+      "eval_runtime": 15.4792,
+      "eval_samples_per_second": 6.46,
+      "eval_steps_per_second": 1.615,
       "step": 125
     },
     {
-      "epoch": 0.3227665706051873,
-      "grad_norm": 0.048432037234306335,
-      "learning_rate": 6.461538461538462e-05,
-      "loss": 0.0143,
+      "epoch": 0.4072727272727273,
+      "grad_norm": 0.013553191907703876,
+      "learning_rate": 1.630012936610608e-05,
+      "loss": 0.0036,
       "step": 126
     },
     {
-      "epoch": 0.32532821005443485,
-      "grad_norm": 0.03601989895105362,
-      "learning_rate": 6.512820512820513e-05,
-      "loss": 0.0183,
+      "epoch": 0.4105050505050505,
+      "grad_norm": 0.011703286319971085,
+      "learning_rate": 1.6429495472186288e-05,
+      "loss": 0.0034,
       "step": 127
     },
     {
-      "epoch": 0.32788984950368233,
-      "grad_norm": 0.045704398304224014,
-      "learning_rate": 6.564102564102564e-05,
-      "loss": 0.0204,
+      "epoch": 0.41373737373737374,
+      "grad_norm": 0.017381293699145317,
+      "learning_rate": 1.6558861578266498e-05,
+      "loss": 0.0053,
       "step": 128
     },
     {
-      "epoch": 0.33045148895292986,
-      "grad_norm": 0.04779214784502983,
-      "learning_rate": 6.615384615384616e-05,
-      "loss": 0.0157,
+      "epoch": 0.416969696969697,
+      "grad_norm": 0.01946461759507656,
+      "learning_rate": 1.66882276843467e-05,
+      "loss": 0.0129,
       "step": 129
     },
     {
-      "epoch": 0.3330131284021774,
-      "grad_norm": 0.03140333667397499,
-      "learning_rate": 6.666666666666667e-05,
-      "loss": 0.0149,
+      "epoch": 0.4202020202020202,
+      "grad_norm": 0.0168154239654541,
+      "learning_rate": 1.6817593790426908e-05,
+      "loss": 0.0044,
       "step": 130
     },
     {
-      "epoch": 0.3355747678514249,
-      "grad_norm": 0.03896024823188782,
-      "learning_rate": 6.717948717948718e-05,
-      "loss": 0.0251,
+      "epoch": 0.42343434343434344,
+      "grad_norm": 0.017521236091852188,
+      "learning_rate": 1.694695989650712e-05,
+      "loss": 0.0064,
       "step": 131
     },
     {
-      "epoch": 0.3381364073006724,
-      "grad_norm": 0.03296570107340813,
-      "learning_rate": 6.76923076923077e-05,
-      "loss": 0.0193,
+      "epoch": 0.4266666666666667,
+      "grad_norm": 0.01116028893738985,
+      "learning_rate": 1.7076326002587322e-05,
+      "loss": 0.0038,
       "step": 132
     },
     {
-      "epoch": 0.34069804674991994,
-      "grad_norm": 0.04064328223466873,
-      "learning_rate": 6.820512820512821e-05,
-      "loss": 0.0269,
+      "epoch": 0.4298989898989899,
+      "grad_norm": 0.019972849637269974,
+      "learning_rate": 1.720569210866753e-05,
+      "loss": 0.0131,
       "step": 133
     },
     {
-      "epoch": 0.34325968619916747,
-      "grad_norm": 0.04517941549420357,
-      "learning_rate": 6.871794871794872e-05,
-      "loss": 0.0152,
+      "epoch": 0.43313131313131314,
+      "grad_norm": 0.018875565379858017,
+      "learning_rate": 1.733505821474774e-05,
+      "loss": 0.0089,
       "step": 134
     },
     {
-      "epoch": 0.345821325648415,
-      "grad_norm": 0.037514880299568176,
-      "learning_rate": 6.923076923076924e-05,
-      "loss": 0.029,
+      "epoch": 0.43636363636363634,
+      "grad_norm": 0.0182089451700449,
+      "learning_rate": 1.7464424320827943e-05,
+      "loss": 0.008,
       "step": 135
     },
     {
-      "epoch": 0.3483829650976625,
-      "grad_norm": 0.04924914240837097,
-      "learning_rate": 6.974358974358974e-05,
-      "loss": 0.0156,
+      "epoch": 0.4395959595959596,
+      "grad_norm": 0.01558719016611576,
+      "learning_rate": 1.7593790426908153e-05,
+      "loss": 0.005,
       "step": 136
     },
     {
-      "epoch": 0.35094460454691,
-      "grad_norm": 0.0434134341776371,
-      "learning_rate": 7.025641025641025e-05,
-      "loss": 0.0118,
+      "epoch": 0.44282828282828285,
+      "grad_norm": 0.014480640180408955,
+      "learning_rate": 1.7723156532988356e-05,
+      "loss": 0.0074,
       "step": 137
     },
     {
-      "epoch": 0.35350624399615754,
-      "grad_norm": 0.046721406280994415,
-      "learning_rate": 7.076923076923078e-05,
-      "loss": 0.0238,
+      "epoch": 0.44606060606060605,
+      "grad_norm": 0.014394573867321014,
+      "learning_rate": 1.7852522639068563e-05,
+      "loss": 0.0085,
       "step": 138
     },
     {
-      "epoch": 0.3560678834454051,
-      "grad_norm": 0.04098832234740257,
-      "learning_rate": 7.128205128205129e-05,
-      "loss": 0.0188,
+      "epoch": 0.4492929292929293,
+      "grad_norm": 0.012834247201681137,
+      "learning_rate": 1.7981888745148774e-05,
+      "loss": 0.0047,
       "step": 139
     },
     {
-      "epoch": 0.35862952289465255,
-      "grad_norm": 0.03906092047691345,
-      "learning_rate": 7.17948717948718e-05,
-      "loss": 0.016,
+      "epoch": 0.45252525252525255,
+      "grad_norm": 0.012524325400590897,
+      "learning_rate": 1.8111254851228977e-05,
+      "loss": 0.0039,
       "step": 140
     },
     {
-      "epoch": 0.3611911623439001,
-      "grad_norm": 0.046422287821769714,
-      "learning_rate": 7.23076923076923e-05,
-      "loss": 0.0296,
+      "epoch": 0.45575757575757575,
+      "grad_norm": 0.019671326503157616,
+      "learning_rate": 1.8240620957309184e-05,
+      "loss": 0.0056,
       "step": 141
     },
     {
-      "epoch": 0.3637528017931476,
-      "grad_norm": 0.05698518827557564,
-      "learning_rate": 7.282051282051282e-05,
-      "loss": 0.0191,
+      "epoch": 0.458989898989899,
+      "grad_norm": 0.014606963843107224,
+      "learning_rate": 1.8369987063389394e-05,
+      "loss": 0.0043,
       "step": 142
     },
     {
-      "epoch": 0.36631444124239515,
-      "grad_norm": 0.04666212946176529,
-      "learning_rate": 7.333333333333333e-05,
-      "loss": 0.0234,
+      "epoch": 0.4622222222222222,
+      "grad_norm": 0.012379477731883526,
+      "learning_rate": 1.8499353169469598e-05,
+      "loss": 0.006,
       "step": 143
     },
     {
-      "epoch": 0.3688760806916426,
-      "grad_norm": 0.03416457027196884,
-      "learning_rate": 7.384615384615386e-05,
-      "loss": 0.0197,
+      "epoch": 0.46545454545454545,
+      "grad_norm": 0.015988007187843323,
+      "learning_rate": 1.8628719275549808e-05,
+      "loss": 0.0061,
       "step": 144
     },
     {
-      "epoch": 0.37143772014089016,
-      "grad_norm": 0.04253166541457176,
-      "learning_rate": 7.435897435897436e-05,
-      "loss": 0.0189,
+      "epoch": 0.4686868686868687,
+      "grad_norm": 0.012083993293344975,
+      "learning_rate": 1.8758085381630015e-05,
+      "loss": 0.0053,
       "step": 145
     },
     {
-      "epoch": 0.3739993595901377,
-      "grad_norm": 0.03973923251032829,
-      "learning_rate": 7.487179487179487e-05,
-      "loss": 0.0182,
+      "epoch": 0.4719191919191919,
+      "grad_norm": 0.013649413362145424,
+      "learning_rate": 1.888745148771022e-05,
+      "loss": 0.0048,
       "step": 146
     },
     {
-      "epoch": 0.3765609990393852,
-      "grad_norm": 0.03506443649530411,
-      "learning_rate": 7.538461538461539e-05,
-      "loss": 0.0138,
+      "epoch": 0.47515151515151516,
+      "grad_norm": 0.015521523542702198,
+      "learning_rate": 1.901681759379043e-05,
+      "loss": 0.0089,
       "step": 147
     },
     {
-      "epoch": 0.3791226384886327,
-      "grad_norm": 0.054082706570625305,
-      "learning_rate": 7.58974358974359e-05,
-      "loss": 0.0284,
+      "epoch": 0.4783838383838384,
+      "grad_norm": 0.013323163613677025,
+      "learning_rate": 1.9146183699870636e-05,
+      "loss": 0.0049,
       "step": 148
     },
     {
-      "epoch": 0.38168427793788023,
-      "grad_norm": 0.057566624134778976,
-      "learning_rate": 7.641025641025641e-05,
-      "loss": 0.0162,
+      "epoch": 0.4816161616161616,
+      "grad_norm": 0.01375036034733057,
+      "learning_rate": 1.927554980595084e-05,
+      "loss": 0.0056,
       "step": 149
     },
     {
-      "epoch": 0.38424591738712777,
-      "grad_norm": 0.04910784214735031,
-      "learning_rate": 7.692307692307693e-05,
-      "loss": 0.0216,
+      "epoch": 0.48484848484848486,
+      "grad_norm": 0.013851992785930634,
+      "learning_rate": 1.940491591203105e-05,
+      "loss": 0.0079,
       "step": 150
     },
     {
-      "epoch": 0.38424591738712777,
-      "eval_loss": 0.015280881896615028,
-      "eval_runtime": 15.7668,
-      "eval_samples_per_second": 6.342,
-      "eval_steps_per_second": 1.586,
+      "epoch": 0.48484848484848486,
+      "eval_loss": 0.004405347164720297,
+      "eval_runtime": 15.5362,
+      "eval_samples_per_second": 6.437,
+      "eval_steps_per_second": 1.609,
       "step": 150
     },
     {
-      "epoch": 0.3868075568363753,
-      "grad_norm": 0.034558191895484924,
-      "learning_rate": 7.743589743589744e-05,
-      "loss": 0.0145,
+      "epoch": 0.48808080808080806,
+      "grad_norm": 0.013039483688771725,
+      "learning_rate": 1.9534282018111256e-05,
+      "loss": 0.0054,
       "step": 151
     },
     {
-      "epoch": 0.3893691962856228,
-      "grad_norm": 0.03240443766117096,
-      "learning_rate": 7.794871794871795e-05,
-      "loss": 0.0167,
+      "epoch": 0.4913131313131313,
+      "grad_norm": 0.01177810039371252,
+      "learning_rate": 1.9663648124191463e-05,
+      "loss": 0.0044,
       "step": 152
     },
     {
-      "epoch": 0.3919308357348703,
-      "grad_norm": 0.03908954933285713,
-      "learning_rate": 7.846153846153847e-05,
-      "loss": 0.0254,
+      "epoch": 0.49454545454545457,
+      "grad_norm": 0.018158679828047752,
+      "learning_rate": 1.979301423027167e-05,
+      "loss": 0.0111,
       "step": 153
     },
     {
-      "epoch": 0.39449247518411784,
-      "grad_norm": 0.04887024685740471,
-      "learning_rate": 7.897435897435898e-05,
-      "loss": 0.0174,
+      "epoch": 0.49777777777777776,
+      "grad_norm": 0.012318884022533894,
+      "learning_rate": 1.9922380336351877e-05,
+      "loss": 0.0034,
       "step": 154
     },
     {
-      "epoch": 0.3970541146333654,
-      "grad_norm": 0.035764534026384354,
-      "learning_rate": 7.948717948717948e-05,
-      "loss": 0.0193,
+      "epoch": 0.501010101010101,
+      "grad_norm": 0.014878074638545513,
+      "learning_rate": 2.0051746442432084e-05,
+      "loss": 0.0067,
       "step": 155
     },
     {
-      "epoch": 0.39961575408261285,
-      "grad_norm": 0.02639930136501789,
-      "learning_rate": 8e-05,
-      "loss": 0.0122,
+      "epoch": 0.5042424242424243,
+      "grad_norm": 0.01715923845767975,
+      "learning_rate": 2.018111254851229e-05,
+      "loss": 0.0039,
       "step": 156
     },
     {
-      "epoch": 0.4021773935318604,
-      "grad_norm": 0.029830757528543472,
-      "learning_rate": 8.051282051282052e-05,
-      "loss": 0.0121,
+      "epoch": 0.5074747474747475,
+      "grad_norm": 0.015519075095653534,
+      "learning_rate": 2.0310478654592497e-05,
+      "loss": 0.0054,
       "step": 157
     },
     {
-      "epoch": 0.4047390329811079,
-      "grad_norm": 0.054097048938274384,
-      "learning_rate": 8.102564102564103e-05,
-      "loss": 0.0357,
+      "epoch": 0.5107070707070707,
+      "grad_norm": 0.018458817154169083,
+      "learning_rate": 2.0439844760672704e-05,
+      "loss": 0.0069,
       "step": 158
     },
     {
-      "epoch": 0.40730067243035545,
-      "grad_norm": 0.03186127543449402,
-      "learning_rate": 8.153846153846155e-05,
-      "loss": 0.0133,
+      "epoch": 0.5139393939393939,
+      "grad_norm": 0.02576751634478569,
+      "learning_rate": 2.056921086675291e-05,
+      "loss": 0.0046,
       "step": 159
     },
     {
-      "epoch": 0.4098623118796029,
-      "grad_norm": 0.03369072452187538,
-      "learning_rate": 8.205128205128205e-05,
-      "loss": 0.0125,
+      "epoch": 0.5171717171717172,
+      "grad_norm": 0.018317924812436104,
+      "learning_rate": 2.0698576972833118e-05,
+      "loss": 0.0084,
       "step": 160
     },
     {
-      "epoch": 0.41242395132885046,
-      "grad_norm": 0.03541966527700424,
-      "learning_rate": 8.256410256410256e-05,
-      "loss": 0.0138,
+      "epoch": 0.5204040404040404,
+      "grad_norm": 0.01574290543794632,
+      "learning_rate": 2.0827943078913325e-05,
+      "loss": 0.0048,
       "step": 161
     },
     {
-      "epoch": 0.414985590778098,
-      "grad_norm": 0.02782193198800087,
-      "learning_rate": 8.307692307692309e-05,
-      "loss": 0.0118,
+      "epoch": 0.5236363636363637,
+      "grad_norm": 0.01320936344563961,
+      "learning_rate": 2.0957309184993532e-05,
+      "loss": 0.0031,
       "step": 162
     },
     {
-      "epoch": 0.4175472302273455,
-      "grad_norm": 0.06363299489021301,
-      "learning_rate": 8.35897435897436e-05,
-      "loss": 0.0339,
+      "epoch": 0.5268686868686868,
+      "grad_norm": 0.01142280362546444,
+      "learning_rate": 2.108667529107374e-05,
+      "loss": 0.0046,
       "step": 163
     },
     {
-      "epoch": 0.420108869676593,
-      "grad_norm": 0.03677621856331825,
-      "learning_rate": 8.410256410256411e-05,
-      "loss": 0.0148,
+      "epoch": 0.5301010101010101,
+      "grad_norm": 0.019800543785095215,
+      "learning_rate": 2.1216041397153946e-05,
+      "loss": 0.0088,
       "step": 164
     },
     {
-      "epoch": 0.42267050912584053,
-      "grad_norm": 0.039866507053375244,
-      "learning_rate": 8.461538461538461e-05,
-      "loss": 0.0122,
+      "epoch": 0.5333333333333333,
+      "grad_norm": 0.012508701533079147,
+      "learning_rate": 2.1345407503234156e-05,
+      "loss": 0.0047,
       "step": 165
     },
     {
-      "epoch": 0.42523214857508806,
-      "grad_norm": 0.04610941559076309,
-      "learning_rate": 8.512820512820513e-05,
-      "loss": 0.0307,
+      "epoch": 0.5365656565656566,
+      "grad_norm": 0.014040361158549786,
+      "learning_rate": 2.147477360931436e-05,
+      "loss": 0.0038,
       "step": 166
     },
     {
-      "epoch": 0.4277937880243356,
-      "grad_norm": 0.04283754900097847,
-      "learning_rate": 8.564102564102564e-05,
-      "loss": 0.0176,
+      "epoch": 0.5397979797979798,
+      "grad_norm": 0.011568131856620312,
+      "learning_rate": 2.1604139715394566e-05,
+      "loss": 0.0057,
       "step": 167
     },
     {
-      "epoch": 0.4303554274735831,
-      "grad_norm": 0.040971748530864716,
-      "learning_rate": 8.615384615384617e-05,
-      "loss": 0.0164,
+      "epoch": 0.5430303030303031,
+      "grad_norm": 0.012783310376107693,
+      "learning_rate": 2.1733505821474777e-05,
+      "loss": 0.006,
       "step": 168
     },
     {
-      "epoch": 0.4329170669228306,
-      "grad_norm": 0.03258652612566948,
-      "learning_rate": 8.666666666666667e-05,
-      "loss": 0.0152,
+      "epoch": 0.5462626262626262,
+      "grad_norm": 0.012915881350636482,
+      "learning_rate": 2.186287192755498e-05,
+      "loss": 0.0049,
       "step": 169
     },
     {
-      "epoch": 0.43547870637207814,
-      "grad_norm": 0.03341250121593475,
-      "learning_rate": 8.717948717948718e-05,
-      "loss": 0.0128,
+      "epoch": 0.5494949494949495,
+      "grad_norm": 0.014070299454033375,
+      "learning_rate": 2.1992238033635187e-05,
+      "loss": 0.0057,
       "step": 170
     },
     {
-      "epoch": 0.43804034582132567,
-      "grad_norm": 0.030181603506207466,
-      "learning_rate": 8.76923076923077e-05,
-      "loss": 0.0138,
+      "epoch": 0.5527272727272727,
+      "grad_norm": 0.012185858562588692,
+      "learning_rate": 2.2121604139715397e-05,
+      "loss": 0.0038,
       "step": 171
     },
     {
-      "epoch": 0.44060198527057315,
-      "grad_norm": 0.0757037103176117,
-      "learning_rate": 8.820512820512821e-05,
-      "loss": 0.018,
+      "epoch": 0.555959595959596,
+      "grad_norm": 0.010173555463552475,
+      "learning_rate": 2.22509702457956e-05,
+      "loss": 0.0028,
       "step": 172
     },
     {
-      "epoch": 0.4431636247198207,
-      "grad_norm": 0.035500697791576385,
-      "learning_rate": 8.871794871794872e-05,
-      "loss": 0.0206,
+      "epoch": 0.5591919191919192,
+      "grad_norm": 0.013600927777588367,
+      "learning_rate": 2.238033635187581e-05,
+      "loss": 0.004,
       "step": 173
     },
     {
-      "epoch": 0.4457252641690682,
-      "grad_norm": 0.02978915348649025,
-      "learning_rate": 8.923076923076924e-05,
-      "loss": 0.0141,
+      "epoch": 0.5624242424242424,
+      "grad_norm": 0.01405523344874382,
+      "learning_rate": 2.2509702457956018e-05,
+      "loss": 0.0045,
       "step": 174
     },
     {
-      "epoch": 0.44828690361831575,
-      "grad_norm": 0.04094715043902397,
-      "learning_rate": 8.974358974358975e-05,
-      "loss": 0.0138,
+      "epoch": 0.5656565656565656,
+      "grad_norm": 0.01713472791016102,
+      "learning_rate": 2.263906856403622e-05,
+      "loss": 0.0097,
       "step": 175
     },
     {
-      "epoch": 0.44828690361831575,
-      "eval_loss": 0.015311915427446365,
-      "eval_runtime": 15.7999,
-      "eval_samples_per_second": 6.329,
-      "eval_steps_per_second": 1.582,
+      "epoch": 0.5656565656565656,
+      "eval_loss": 0.004139286000281572,
+      "eval_runtime": 15.5054,
+      "eval_samples_per_second": 6.449,
+      "eval_steps_per_second": 1.612,
       "step": 175
     },
     {
-      "epoch": 0.4508485430675632,
-      "grad_norm": 0.03736531361937523,
-      "learning_rate": 9.025641025641026e-05,
-      "loss": 0.0133,
+      "epoch": 0.5688888888888889,
+      "grad_norm": 0.016253681853413582,
+      "learning_rate": 2.276843467011643e-05,
+      "loss": 0.0048,
       "step": 176
     },
     {
-      "epoch": 0.45341018251681076,
-      "grad_norm": 0.026462465524673462,
-      "learning_rate": 9.076923076923078e-05,
-      "loss": 0.0087,
+      "epoch": 0.5721212121212121,
+      "grad_norm": 0.013580011203885078,
+      "learning_rate": 2.289780077619664e-05,
+      "loss": 0.0054,
       "step": 177
     },
     {
-      "epoch": 0.4559718219660583,
-      "grad_norm": 0.044491320848464966,
-      "learning_rate": 9.128205128205129e-05,
-      "loss": 0.0168,
+      "epoch": 0.5753535353535354,
+      "grad_norm": 0.014574569649994373,
+      "learning_rate": 2.3027166882276842e-05,
+      "loss": 0.0036,
       "step": 178
     },
     {
-      "epoch": 0.4585334614153058,
-      "grad_norm": 0.040210917592048645,
-      "learning_rate": 9.179487179487179e-05,
-      "loss": 0.0217,
+      "epoch": 0.5785858585858585,
+      "grad_norm": 0.014509929344058037,
+      "learning_rate": 2.3156532988357052e-05,
+      "loss": 0.0042,
       "step": 179
     },
     {
-      "epoch": 0.4610951008645533,
-      "grad_norm": 0.067959725856781,
-      "learning_rate": 9.230769230769232e-05,
-      "loss": 0.0192,
+      "epoch": 0.5818181818181818,
+      "grad_norm": 0.015430472791194916,
+      "learning_rate": 2.328589909443726e-05,
+      "loss": 0.0054,
       "step": 180
     },
     {
-      "epoch": 0.46365674031380083,
-      "grad_norm": 0.03490135073661804,
-      "learning_rate": 9.282051282051283e-05,
-      "loss": 0.0118,
+      "epoch": 0.585050505050505,
+      "grad_norm": 0.007407560013234615,
+      "learning_rate": 2.3415265200517466e-05,
+      "loss": 0.0022,
       "step": 181
     },
     {
-      "epoch": 0.46621837976304836,
-      "grad_norm": 0.04227762669324875,
-      "learning_rate": 9.333333333333334e-05,
-      "loss": 0.0146,
+      "epoch": 0.5882828282828283,
+      "grad_norm": 0.01594449393451214,
+      "learning_rate": 2.3544631306597673e-05,
+      "loss": 0.0043,
       "step": 182
     },
     {
-      "epoch": 0.4687800192122959,
-      "grad_norm": 0.07856088131666183,
-      "learning_rate": 9.384615384615386e-05,
-      "loss": 0.0187,
+      "epoch": 0.5915151515151515,
+      "grad_norm": 0.01776193454861641,
+      "learning_rate": 2.367399741267788e-05,
+      "loss": 0.0061,
       "step": 183
     },
     {
-      "epoch": 0.4713416586615434,
-      "grad_norm": 0.039737801998853683,
-      "learning_rate": 9.435897435897436e-05,
-      "loss": 0.0187,
+      "epoch": 0.5947474747474748,
+      "grad_norm": 0.01671319641172886,
+      "learning_rate": 2.3803363518758087e-05,
+      "loss": 0.0049,
       "step": 184
     },
     {
-      "epoch": 0.4739032981107909,
-      "grad_norm": 0.04330054298043251,
-      "learning_rate": 9.487179487179487e-05,
-      "loss": 0.0188,
+      "epoch": 0.597979797979798,
+      "grad_norm": 0.01862150803208351,
+      "learning_rate": 2.3932729624838294e-05,
+      "loss": 0.0042,
       "step": 185
     },
     {
-      "epoch": 0.47646493756003844,
-      "grad_norm": 0.045348942279815674,
-      "learning_rate": 9.53846153846154e-05,
-      "loss": 0.0216,
+      "epoch": 0.6012121212121212,
+      "grad_norm": 0.012478983029723167,
+      "learning_rate": 2.40620957309185e-05,
+      "loss": 0.0036,
       "step": 186
     },
     {
-      "epoch": 0.47902657700928597,
-      "grad_norm": 0.03437650576233864,
-      "learning_rate": 9.589743589743591e-05,
-      "loss": 0.0184,
+      "epoch": 0.6044444444444445,
+      "grad_norm": 0.016773445531725883,
+      "learning_rate": 2.4191461836998707e-05,
+      "loss": 0.0053,
       "step": 187
     },
     {
-      "epoch": 0.48158821645853345,
-      "grad_norm": 0.03575362637639046,
-      "learning_rate": 9.641025641025641e-05,
-      "loss": 0.0223,
+      "epoch": 0.6076767676767677,
+      "grad_norm": 0.012951131910085678,
+      "learning_rate": 2.4320827943078914e-05,
+      "loss": 0.0036,
       "step": 188
     },
     {
-      "epoch": 0.484149855907781,
-      "grad_norm": 0.022593902423977852,
-      "learning_rate": 9.692307692307692e-05,
-      "loss": 0.0128,
+      "epoch": 0.610909090909091,
+      "grad_norm": 0.018362637609243393,
+      "learning_rate": 2.445019404915912e-05,
+      "loss": 0.0069,
       "step": 189
     },
     {
-      "epoch": 0.4867114953570285,
-      "grad_norm": 0.04331602156162262,
-      "learning_rate": 9.743589743589744e-05,
-      "loss": 0.0254,
+      "epoch": 0.6141414141414141,
+      "grad_norm": 0.014896437525749207,
+      "learning_rate": 2.4579560155239328e-05,
+      "loss": 0.0044,
       "step": 190
     },
     {
-      "epoch": 0.48927313480627604,
-      "grad_norm": 0.02803720161318779,
-      "learning_rate": 9.794871794871795e-05,
-      "loss": 0.0132,
+      "epoch": 0.6173737373737374,
+      "grad_norm": 0.014703130349516869,
+      "learning_rate": 2.4708926261319535e-05,
+      "loss": 0.0067,
       "step": 191
     },
     {
-      "epoch": 0.4918347742555235,
-      "grad_norm": 0.036241576075553894,
-      "learning_rate": 9.846153846153848e-05,
-      "loss": 0.012,
+      "epoch": 0.6206060606060606,
+      "grad_norm": 0.016904670745134354,
+      "learning_rate": 2.4838292367399742e-05,
+      "loss": 0.006,
       "step": 192
     },
     {
-      "epoch": 0.49439641370477105,
-      "grad_norm": 0.03179619088768959,
-      "learning_rate": 9.897435897435898e-05,
-      "loss": 0.0126,
+      "epoch": 0.6238383838383839,
+      "grad_norm": 0.015180383808910847,
+      "learning_rate": 2.496765847347995e-05,
+      "loss": 0.0047,
       "step": 193
     },
     {
-      "epoch": 0.4969580531540186,
-      "grad_norm": 0.03643405809998512,
-      "learning_rate": 9.948717948717949e-05,
-      "loss": 0.0184,
+      "epoch": 0.6270707070707071,
+      "grad_norm": 0.015023029409348965,
+      "learning_rate": 2.509702457956016e-05,
+      "loss": 0.0043,
       "step": 194
     },
     {
-      "epoch": 0.49951969260326606,
-      "grad_norm": 0.03236660733819008,
-      "learning_rate": 0.0001,
-      "loss": 0.0174,
+      "epoch": 0.6303030303030303,
+      "grad_norm": 0.01615927927196026,
+      "learning_rate": 2.5226390685640362e-05,
+      "loss": 0.0075,
       "step": 195
     },
     {
-      "epoch": 0.5020813320525136,
-      "grad_norm": 0.025117913261055946,
-      "learning_rate": 9.999991989024631e-05,
-      "loss": 0.0125,
+      "epoch": 0.6335353535353535,
+      "grad_norm": 0.01465796958655119,
+      "learning_rate": 2.535575679172057e-05,
+      "loss": 0.0038,
       "step": 196
     },
     {
-      "epoch": 0.5046429715017611,
-      "grad_norm": 0.026827501133084297,
-      "learning_rate": 9.999967956124192e-05,
-      "loss": 0.0125,
+      "epoch": 0.6367676767676768,
+      "grad_norm": 0.015334381721913815,
+      "learning_rate": 2.548512289780078e-05,
+      "loss": 0.0029,
       "step": 197
     },
     {
-      "epoch": 0.5072046109510087,
-      "grad_norm": 0.02300351858139038,
-      "learning_rate": 9.999927901375693e-05,
-      "loss": 0.0098,
+      "epoch": 0.64,
+      "grad_norm": 0.013329016044735909,
+      "learning_rate": 2.5614489003880986e-05,
+      "loss": 0.0041,
       "step": 198
     },
     {
-      "epoch": 0.5097662504002561,
-      "grad_norm": 0.03161875531077385,
-      "learning_rate": 9.999871824907485e-05,
-      "loss": 0.02,
+      "epoch": 0.6432323232323233,
+      "grad_norm": 0.010669299401342869,
+      "learning_rate": 2.574385510996119e-05,
+      "loss": 0.0032,
       "step": 199
     },
     {
-      "epoch": 0.5123278898495037,
-      "grad_norm": 0.040188804268836975,
-      "learning_rate": 9.999799726899262e-05,
-      "loss": 0.021,
+      "epoch": 0.6464646464646465,
+      "grad_norm": 0.016828523948788643,
+      "learning_rate": 2.58732212160414e-05,
+      "loss": 0.0049,
       "step": 200
     },
     {
-      "epoch": 0.5123278898495037,
-      "eval_loss": 0.014618256129324436,
-      "eval_runtime": 15.7792,
-      "eval_samples_per_second": 6.337,
-      "eval_steps_per_second": 1.584,
+      "epoch": 0.6464646464646465,
+      "eval_loss": 0.003835035488009453,
+      "eval_runtime": 15.4689,
+      "eval_samples_per_second": 6.465,
+      "eval_steps_per_second": 1.616,
       "step": 200
     },
     {
-      "epoch": 0.5148895292987512,
-      "grad_norm": 0.027994215488433838,
-      "learning_rate": 9.999711607582052e-05,
-      "loss": 0.0157,
+      "epoch": 0.6496969696969697,
+      "grad_norm": 0.013488763011991978,
+      "learning_rate": 2.6002587322121607e-05,
+      "loss": 0.0029,
       "step": 201
     },
     {
-      "epoch": 0.5174511687479987,
-      "grad_norm": 0.03944910690188408,
-      "learning_rate": 9.999607467238223e-05,
-      "loss": 0.0199,
+      "epoch": 0.6529292929292929,
+      "grad_norm": 0.014697612263262272,
+      "learning_rate": 2.613195342820181e-05,
+      "loss": 0.0027,
       "step": 202
     },
     {
-      "epoch": 0.5200128081972463,
-      "grad_norm": 0.03262855485081673,
-      "learning_rate": 9.999487306201483e-05,
-      "loss": 0.0162,
+      "epoch": 0.6561616161616162,
+      "grad_norm": 0.013086420483887196,
+      "learning_rate": 2.626131953428202e-05,
+      "loss": 0.0035,
       "step": 203
     },
     {
-      "epoch": 0.5225744476464937,
-      "grad_norm": 0.020725559443235397,
-      "learning_rate": 9.999351124856874e-05,
-      "loss": 0.0089,
+      "epoch": 0.6593939393939394,
+      "grad_norm": 0.015521716326475143,
+      "learning_rate": 2.6390685640362228e-05,
+      "loss": 0.0045,
       "step": 204
     },
     {
-      "epoch": 0.5251360870957412,
-      "grad_norm": 0.03486771881580353,
-      "learning_rate": 9.999198923640774e-05,
-      "loss": 0.0133,
+      "epoch": 0.6626262626262627,
+      "grad_norm": 0.015355012379586697,
+      "learning_rate": 2.652005174644243e-05,
+      "loss": 0.0043,
       "step": 205
     },
     {
-      "epoch": 0.5276977265449888,
-      "grad_norm": 0.03679514676332474,
-      "learning_rate": 9.999030703040895e-05,
-      "loss": 0.0231,
+      "epoch": 0.6658585858585858,
+      "grad_norm": 0.017829393967986107,
+      "learning_rate": 2.664941785252264e-05,
+      "loss": 0.0065,
       "step": 206
     },
     {
-      "epoch": 0.5302593659942363,
-      "grad_norm": 0.029446182772517204,
-      "learning_rate": 9.998846463596281e-05,
-      "loss": 0.013,
+      "epoch": 0.6690909090909091,
+      "grad_norm": 0.016501599922776222,
+      "learning_rate": 2.677878395860285e-05,
+      "loss": 0.0049,
       "step": 207
     },
     {
-      "epoch": 0.5328210054434839,
-      "grad_norm": 0.03434671834111214,
-      "learning_rate": 9.998646205897309e-05,
-      "loss": 0.0187,
+      "epoch": 0.6723232323232323,
+      "grad_norm": 0.016663990914821625,
+      "learning_rate": 2.6908150064683052e-05,
+      "loss": 0.0061,
       "step": 208
     },
     {
-      "epoch": 0.5353826448927314,
-      "grad_norm": 0.03190897777676582,
-      "learning_rate": 9.998429930585682e-05,
-      "loss": 0.0141,
+      "epoch": 0.6755555555555556,
+      "grad_norm": 0.013605273328721523,
+      "learning_rate": 2.7037516170763262e-05,
+      "loss": 0.0057,
       "step": 209
     },
     {
-      "epoch": 0.5379442843419788,
-      "grad_norm": 0.03730105608701706,
-      "learning_rate": 9.998197638354428e-05,
-      "loss": 0.0217,
+      "epoch": 0.6787878787878788,
+      "grad_norm": 0.02209271304309368,
+      "learning_rate": 2.716688227684347e-05,
+      "loss": 0.006,
       "step": 210
     },
     {
-      "epoch": 0.5405059237912264,
-      "grad_norm": 0.03427678719162941,
-      "learning_rate": 9.997949329947904e-05,
-      "loss": 0.0207,
+      "epoch": 0.682020202020202,
+      "grad_norm": 0.017327651381492615,
+      "learning_rate": 2.7296248382923673e-05,
+      "loss": 0.0072,
       "step": 211
     },
     {
-      "epoch": 0.5430675632404739,
-      "grad_norm": 0.03182192146778107,
-      "learning_rate": 9.997685006161788e-05,
-      "loss": 0.0272,
+      "epoch": 0.6852525252525252,
+      "grad_norm": 0.017740705981850624,
+      "learning_rate": 2.7425614489003883e-05,
+      "loss": 0.006,
       "step": 212
     },
     {
-      "epoch": 0.5456292026897214,
-      "grad_norm": 0.02717447653412819,
-      "learning_rate": 9.997404667843075e-05,
-      "loss": 0.0135,
+      "epoch": 0.6884848484848485,
+      "grad_norm": 0.012097117491066456,
+      "learning_rate": 2.755498059508409e-05,
+      "loss": 0.0041,
       "step": 213
     },
     {
-      "epoch": 0.548190842138969,
-      "grad_norm": 0.04445464909076691,
-      "learning_rate": 9.99710831589008e-05,
-      "loss": 0.0265,
+      "epoch": 0.6917171717171717,
+      "grad_norm": 0.020837504416704178,
+      "learning_rate": 2.76843467011643e-05,
+      "loss": 0.0049,
       "step": 214
     },
     {
-      "epoch": 0.5507524815882164,
-      "grad_norm": 0.03553137183189392,
-      "learning_rate": 9.996795951252427e-05,
-      "loss": 0.0174,
+      "epoch": 0.694949494949495,
+      "grad_norm": 0.011065317317843437,
+      "learning_rate": 2.7813712807244503e-05,
+      "loss": 0.0031,
       "step": 215
     },
     {
-      "epoch": 0.553314121037464,
-      "grad_norm": 0.034598637372255325,
-      "learning_rate": 9.996467574931057e-05,
-      "loss": 0.0178,
+      "epoch": 0.6981818181818182,
+      "grad_norm": 0.015117717906832695,
+      "learning_rate": 2.794307891332471e-05,
+      "loss": 0.0062,
       "step": 216
     },
     {
-      "epoch": 0.5558757604867115,
-      "grad_norm": 0.030017906799912453,
-      "learning_rate": 9.996123187978215e-05,
-      "loss": 0.0128,
+      "epoch": 0.7014141414141414,
+      "grad_norm": 0.023171257227659225,
+      "learning_rate": 2.807244501940492e-05,
+      "loss": 0.0108,
       "step": 217
     },
     {
-      "epoch": 0.558437399935959,
-      "grad_norm": 0.033853303641080856,
-      "learning_rate": 9.995762791497452e-05,
-      "loss": 0.0173,
+      "epoch": 0.7046464646464646,
+      "grad_norm": 0.017575595527887344,
+      "learning_rate": 2.8201811125485124e-05,
+      "loss": 0.0054,
       "step": 218
     },
     {
-      "epoch": 0.5609990393852066,
-      "grad_norm": 0.030432872474193573,
-      "learning_rate": 9.99538638664362e-05,
-      "loss": 0.0138,
+      "epoch": 0.7078787878787879,
+      "grad_norm": 0.015625666826963425,
+      "learning_rate": 2.833117723156533e-05,
+      "loss": 0.0039,
       "step": 219
     },
     {
-      "epoch": 0.563560678834454,
-      "grad_norm": 0.03020687773823738,
-      "learning_rate": 9.994993974622862e-05,
-      "loss": 0.0143,
+      "epoch": 0.7111111111111111,
+      "grad_norm": 0.01898195594549179,
+      "learning_rate": 2.846054333764554e-05,
+      "loss": 0.0043,
       "step": 220
     },
     {
-      "epoch": 0.5661223182837015,
-      "grad_norm": 0.035555846989154816,
-      "learning_rate": 9.994585556692624e-05,
-      "loss": 0.0228,
+      "epoch": 0.7143434343434344,
+      "grad_norm": 0.02059677056968212,
+      "learning_rate": 2.8589909443725745e-05,
+      "loss": 0.0103,
       "step": 221
     },
     {
-      "epoch": 0.5686839577329491,
-      "grad_norm": 0.036969736218452454,
-      "learning_rate": 9.994161134161634e-05,
-      "loss": 0.0241,
+      "epoch": 0.7175757575757575,
+      "grad_norm": 0.02094186656177044,
+      "learning_rate": 2.871927554980595e-05,
+      "loss": 0.0048,
       "step": 222
     },
     {
-      "epoch": 0.5712455971821966,
-      "grad_norm": 0.0440659373998642,
-      "learning_rate": 9.993720708389909e-05,
-      "loss": 0.0241,
+      "epoch": 0.7208080808080808,
+      "grad_norm": 0.029153166338801384,
+      "learning_rate": 2.8848641655886162e-05,
+      "loss": 0.0031,
       "step": 223
     },
     {
-      "epoch": 0.5738072366314442,
-      "grad_norm": 0.03319979086518288,
-      "learning_rate": 9.993264280788744e-05,
-      "loss": 0.0253,
+      "epoch": 0.724040404040404,
+      "grad_norm": 0.01566002517938614,
+      "learning_rate": 2.8978007761966365e-05,
+      "loss": 0.0056,
       "step": 224
     },
     {
-      "epoch": 0.5763688760806917,
-      "grad_norm": 0.026040690019726753,
-      "learning_rate": 9.992791852820709e-05,
-      "loss": 0.0166,
+      "epoch": 0.7272727272727273,
+      "grad_norm": 0.014786525629460812,
+      "learning_rate": 2.9107373868046572e-05,
+      "loss": 0.0053,
       "step": 225
     },
     {
-      "epoch": 0.5763688760806917,
-      "eval_loss": 0.013338856399059296,
-      "eval_runtime": 15.7439,
-      "eval_samples_per_second": 6.352,
-      "eval_steps_per_second": 1.588,
+      "epoch": 0.7272727272727273,
+      "eval_loss": 0.003789926180616021,
+      "eval_runtime": 15.5412,
+      "eval_samples_per_second": 6.434,
+      "eval_steps_per_second": 1.609,
       "step": 225
     },
     {
-      "epoch": 0.5789305155299391,
-      "grad_norm": 0.02438931353390217,
-      "learning_rate": 9.992303425999652e-05,
-      "loss": 0.0108,
+      "epoch": 0.7305050505050505,
+      "grad_norm": 0.02039331942796707,
+      "learning_rate": 2.9236739974126783e-05,
+      "loss": 0.0082,
       "step": 226
     },
     {
-      "epoch": 0.5814921549791867,
-      "grad_norm": 0.023932714015245438,
-      "learning_rate": 9.99179900189068e-05,
-      "loss": 0.011,
+      "epoch": 0.7337373737373737,
+      "grad_norm": 0.015376322902739048,
+      "learning_rate": 2.936610608020699e-05,
+      "loss": 0.0054,
       "step": 227
     },
     {
-      "epoch": 0.5840537944284342,
-      "grad_norm": 0.0349879153072834,
-      "learning_rate": 9.991278582110167e-05,
-      "loss": 0.0279,
+      "epoch": 0.7369696969696969,
+      "grad_norm": 0.012364913709461689,
+      "learning_rate": 2.9495472186287193e-05,
+      "loss": 0.004,
       "step": 228
     },
     {
-      "epoch": 0.5866154338776817,
-      "grad_norm": 0.037302933633327484,
-      "learning_rate": 9.99074216832574e-05,
-      "loss": 0.0177,
+      "epoch": 0.7402020202020202,
+      "grad_norm": 0.018084056675434113,
+      "learning_rate": 2.9624838292367403e-05,
+      "loss": 0.0075,
       "step": 229
     },
     {
-      "epoch": 0.5891770733269293,
-      "grad_norm": 0.033939432352781296,
-      "learning_rate": 9.990189762256276e-05,
-      "loss": 0.018,
+      "epoch": 0.7434343434343434,
+      "grad_norm": 0.013406695798039436,
+      "learning_rate": 2.975420439844761e-05,
+      "loss": 0.0031,
       "step": 230
     },
     {
-      "epoch": 0.5917387127761767,
-      "grad_norm": 0.025025606155395508,
-      "learning_rate": 9.989621365671902e-05,
-      "loss": 0.0151,
+      "epoch": 0.7466666666666667,
+      "grad_norm": 0.017664622515439987,
+      "learning_rate": 2.9883570504527814e-05,
+      "loss": 0.0055,
       "step": 231
     },
     {
-      "epoch": 0.5943003522254243,
-      "grad_norm": 0.031279440969228745,
-      "learning_rate": 9.98903698039398e-05,
-      "loss": 0.0119,
+      "epoch": 0.74989898989899,
+      "grad_norm": 0.010642223991453648,
+      "learning_rate": 3.0012936610608024e-05,
+      "loss": 0.0032,
       "step": 232
     },
     {
-      "epoch": 0.5968619916746718,
-      "grad_norm": 0.03417769819498062,
-      "learning_rate": 9.988436608295112e-05,
-      "loss": 0.0202,
+      "epoch": 0.7531313131313131,
+      "grad_norm": 0.011258029378950596,
+      "learning_rate": 3.014230271668823e-05,
+      "loss": 0.0029,
       "step": 233
     },
     {
-      "epoch": 0.5994236311239193,
-      "grad_norm": 0.048715490847826004,
-      "learning_rate": 9.987820251299122e-05,
-      "loss": 0.0227,
+      "epoch": 0.7563636363636363,
+      "grad_norm": 0.019109345972537994,
+      "learning_rate": 3.0271668822768434e-05,
+      "loss": 0.0036,
       "step": 234
     },
     {
-      "epoch": 0.6019852705731669,
-      "grad_norm": 0.03958699479699135,
-      "learning_rate": 9.987187911381059e-05,
-      "loss": 0.0211,
+      "epoch": 0.7595959595959596,
+      "grad_norm": 0.014539861120283604,
+      "learning_rate": 3.0401034928848644e-05,
+      "loss": 0.0061,
       "step": 235
     },
     {
-      "epoch": 0.6045469100224143,
-      "grad_norm": 0.034175582230091095,
-      "learning_rate": 9.986539590567186e-05,
-      "loss": 0.0116,
+      "epoch": 0.7628282828282829,
+      "grad_norm": 0.011992333456873894,
+      "learning_rate": 3.053040103492885e-05,
+      "loss": 0.0034,
       "step": 236
     },
     {
-      "epoch": 0.6071085494716618,
-      "grad_norm": 0.035210780799388885,
-      "learning_rate": 9.985875290934977e-05,
-      "loss": 0.0225,
+      "epoch": 0.7660606060606061,
+      "grad_norm": 0.017860975116491318,
+      "learning_rate": 3.0659767141009055e-05,
+      "loss": 0.0051,
       "step": 237
     },
     {
-      "epoch": 0.6096701889209094,
-      "grad_norm": 0.033485546708106995,
-      "learning_rate": 9.985195014613105e-05,
-      "loss": 0.0167,
+      "epoch": 0.7692929292929293,
+      "grad_norm": 0.01571412943303585,
+      "learning_rate": 3.0789133247089265e-05,
+      "loss": 0.0072,
       "step": 238
     },
     {
-      "epoch": 0.6122318283701569,
-      "grad_norm": 0.03267650306224823,
-      "learning_rate": 9.984498763781444e-05,
-      "loss": 0.0135,
+      "epoch": 0.7725252525252525,
+      "grad_norm": 0.017271071672439575,
+      "learning_rate": 3.0918499353169475e-05,
+      "loss": 0.0065,
       "step": 239
     },
     {
-      "epoch": 0.6147934678194045,
-      "grad_norm": 0.028668779879808426,
-      "learning_rate": 9.983786540671051e-05,
-      "loss": 0.0119,
+      "epoch": 0.7757575757575758,
+      "grad_norm": 0.03471958264708519,
+      "learning_rate": 3.104786545924967e-05,
+      "loss": 0.0064,
       "step": 240
     },
     {
-      "epoch": 0.617355107268652,
-      "grad_norm": 0.028682347387075424,
-      "learning_rate": 9.983058347564167e-05,
-      "loss": 0.0104,
+      "epoch": 0.778989898989899,
+      "grad_norm": 0.014726453460752964,
+      "learning_rate": 3.117723156532988e-05,
+      "loss": 0.0044,
       "step": 241
     },
     {
-      "epoch": 0.6199167467178994,
-      "grad_norm": 0.02997414395213127,
-      "learning_rate": 9.982314186794207e-05,
-      "loss": 0.0143,
+      "epoch": 0.7822222222222223,
+      "grad_norm": 0.0163166094571352,
+      "learning_rate": 3.130659767141009e-05,
+      "loss": 0.0062,
       "step": 242
     },
     {
-      "epoch": 0.622478386167147,
-      "grad_norm": 0.028074778616428375,
-      "learning_rate": 9.981554060745754e-05,
-      "loss": 0.0103,
+      "epoch": 0.7854545454545454,
+      "grad_norm": 0.013584435917437077,
+      "learning_rate": 3.14359637774903e-05,
+      "loss": 0.0058,
       "step": 243
     },
     {
-      "epoch": 0.6250400256163945,
-      "grad_norm": 0.035082463175058365,
-      "learning_rate": 9.980777971854545e-05,
-      "loss": 0.0208,
+      "epoch": 0.7886868686868687,
+      "grad_norm": 0.013390389271080494,
+      "learning_rate": 3.1565329883570506e-05,
+      "loss": 0.0039,
       "step": 244
     },
     {
-      "epoch": 0.627601665065642,
-      "grad_norm": 0.03993948549032211,
-      "learning_rate": 9.979985922607475e-05,
-      "loss": 0.0201,
+      "epoch": 0.7919191919191919,
+      "grad_norm": 0.01425924152135849,
+      "learning_rate": 3.169469598965072e-05,
+      "loss": 0.0042,
       "step": 245
     },
     {
-      "epoch": 0.6301633045148896,
-      "grad_norm": 0.027228379622101784,
-      "learning_rate": 9.979177915542577e-05,
-      "loss": 0.0145,
+      "epoch": 0.7951515151515152,
+      "grad_norm": 0.006553538143634796,
+      "learning_rate": 3.182406209573092e-05,
+      "loss": 0.0019,
       "step": 246
     },
     {
-      "epoch": 0.632724943964137,
-      "grad_norm": 0.02641998790204525,
-      "learning_rate": 9.978353953249022e-05,
-      "loss": 0.0116,
+      "epoch": 0.7983838383838384,
+      "grad_norm": 0.01564977318048477,
+      "learning_rate": 3.1953428201811124e-05,
+      "loss": 0.0064,
       "step": 247
     },
     {
-      "epoch": 0.6352865834133846,
-      "grad_norm": 0.04445129632949829,
-      "learning_rate": 9.977514038367106e-05,
-      "loss": 0.0161,
+      "epoch": 0.8016161616161617,
+      "grad_norm": 0.014616156928241253,
+      "learning_rate": 3.2082794307891334e-05,
+      "loss": 0.0049,
       "step": 248
     },
     {
-      "epoch": 0.6378482228626321,
-      "grad_norm": 0.041566576808691025,
-      "learning_rate": 9.976658173588244e-05,
-      "loss": 0.0344,
+      "epoch": 0.8048484848484848,
+      "grad_norm": 0.010781046003103256,
+      "learning_rate": 3.2212160413971544e-05,
+      "loss": 0.0024,
       "step": 249
     },
     {
-      "epoch": 0.6404098623118796,
-      "grad_norm": 0.031222566962242126,
-      "learning_rate": 9.97578636165496e-05,
-      "loss": 0.0109,
+      "epoch": 0.8080808080808081,
+      "grad_norm": 0.01528346911072731,
+      "learning_rate": 3.234152652005175e-05,
+      "loss": 0.0041,
       "step": 250
     },
     {
-      "epoch": 0.6404098623118796,
-      "eval_loss": 0.011403795331716537,
-      "eval_runtime": 15.8176,
-      "eval_samples_per_second": 6.322,
-      "eval_steps_per_second": 1.581,
+      "epoch": 0.8080808080808081,
+      "eval_loss": 0.003647405421361327,
+      "eval_runtime": 15.5593,
+      "eval_samples_per_second": 6.427,
+      "eval_steps_per_second": 1.607,
       "step": 250
     },
     {
-      "epoch": 0.6429715017611272,
-      "grad_norm": 0.03305068984627724,
-      "learning_rate": 9.974898605360881e-05,
-      "loss": 0.0161,
+      "epoch": 0.8113131313131313,
+      "grad_norm": 0.013240045867860317,
+      "learning_rate": 3.247089262613196e-05,
+      "loss": 0.0044,
       "step": 251
     },
     {
-      "epoch": 0.6455331412103746,
-      "grad_norm": 0.032358262687921524,
-      "learning_rate": 9.973994907550722e-05,
-      "loss": 0.0151,
+      "epoch": 0.8145454545454546,
+      "grad_norm": 0.020097479224205017,
+      "learning_rate": 3.260025873221216e-05,
+      "loss": 0.0114,
       "step": 252
     },
     {
-      "epoch": 0.6480947806596221,
-      "grad_norm": 0.022366944700479507,
-      "learning_rate": 9.973075271120287e-05,
-      "loss": 0.0133,
+      "epoch": 0.8177777777777778,
+      "grad_norm": 0.010310143232345581,
+      "learning_rate": 3.2729624838292365e-05,
+      "loss": 0.0024,
       "step": 253
     },
     {
-      "epoch": 0.6506564201088697,
-      "grad_norm": 0.034839093685150146,
-      "learning_rate": 9.972139699016446e-05,
-      "loss": 0.0114,
+      "epoch": 0.821010101010101,
+      "grad_norm": 0.015720993280410767,
+      "learning_rate": 3.2858990944372575e-05,
+      "loss": 0.005,
       "step": 254
     },
     {
-      "epoch": 0.6532180595581172,
-      "grad_norm": 0.03348885849118233,
-      "learning_rate": 9.97118819423714e-05,
-      "loss": 0.02,
+      "epoch": 0.8242424242424242,
+      "grad_norm": 0.020740246400237083,
+      "learning_rate": 3.2988357050452786e-05,
+      "loss": 0.0079,
       "step": 255
     },
     {
-      "epoch": 0.6557796990073647,
-      "grad_norm": 0.024363042786717415,
-      "learning_rate": 9.970220759831361e-05,
-      "loss": 0.0108,
+      "epoch": 0.8274747474747475,
+      "grad_norm": 0.014475876465439796,
+      "learning_rate": 3.3117723156532996e-05,
+      "loss": 0.0051,
       "step": 256
     },
     {
-      "epoch": 0.6583413384566122,
-      "grad_norm": 0.027448901906609535,
-      "learning_rate": 9.969237398899145e-05,
-      "loss": 0.0122,
+      "epoch": 0.8307070707070707,
+      "grad_norm": 0.013090684078633785,
+      "learning_rate": 3.324708926261319e-05,
+      "loss": 0.0033,
       "step": 257
     },
     {
-      "epoch": 0.6609029779058597,
-      "grad_norm": 0.04620316997170448,
-      "learning_rate": 9.968238114591566e-05,
-      "loss": 0.0232,
+      "epoch": 0.833939393939394,
+      "grad_norm": 0.011114723049104214,
+      "learning_rate": 3.33764553686934e-05,
+      "loss": 0.0041,
       "step": 258
     },
     {
-      "epoch": 0.6634646173551073,
-      "grad_norm": 0.033076781779527664,
-      "learning_rate": 9.967222910110719e-05,
-      "loss": 0.0196,
+      "epoch": 0.8371717171717171,
+      "grad_norm": 0.010789109393954277,
+      "learning_rate": 3.350582147477361e-05,
+      "loss": 0.003,
       "step": 259
     },
     {
-      "epoch": 0.6660262568043548,
-      "grad_norm": 0.021274225786328316,
-      "learning_rate": 9.966191788709716e-05,
-      "loss": 0.0097,
+      "epoch": 0.8404040404040404,
+      "grad_norm": 0.012872577644884586,
+      "learning_rate": 3.3635187580853817e-05,
+      "loss": 0.0032,
       "step": 260
     },
     {
-      "epoch": 0.6685878962536023,
-      "grad_norm": 0.046971239149570465,
-      "learning_rate": 9.965144753692672e-05,
-      "loss": 0.0232,
+      "epoch": 0.8436363636363636,
+      "grad_norm": 0.013579788617789745,
+      "learning_rate": 3.376455368693403e-05,
+      "loss": 0.0042,
       "step": 261
     },
     {
-      "epoch": 0.6711495357028499,
-      "grad_norm": 0.029293019324541092,
-      "learning_rate": 9.964081808414696e-05,
-      "loss": 0.0194,
+      "epoch": 0.8468686868686869,
+      "grad_norm": 0.010664577595889568,
+      "learning_rate": 3.389391979301424e-05,
+      "loss": 0.0026,
       "step": 262
     },
     {
-      "epoch": 0.6737111751520973,
-      "grad_norm": 0.027929246425628662,
-      "learning_rate": 9.963002956281879e-05,
-      "loss": 0.01,
+      "epoch": 0.8501010101010101,
+      "grad_norm": 0.017305690795183182,
+      "learning_rate": 3.4023285899094434e-05,
+      "loss": 0.0053,
       "step": 263
     },
     {
-      "epoch": 0.6762728146013448,
-      "grad_norm": 0.0408920980989933,
-      "learning_rate": 9.961908200751284e-05,
-      "loss": 0.0207,
+      "epoch": 0.8533333333333334,
+      "grad_norm": 0.014377721585333347,
+      "learning_rate": 3.4152652005174644e-05,
+      "loss": 0.0045,
       "step": 264
     },
     {
-      "epoch": 0.6788344540505924,
-      "grad_norm": 0.02565447986125946,
-      "learning_rate": 9.960797545330936e-05,
-      "loss": 0.0189,
+      "epoch": 0.8565656565656565,
+      "grad_norm": 0.012759726494550705,
+      "learning_rate": 3.4282018111254854e-05,
+      "loss": 0.0042,
       "step": 265
     },
     {
-      "epoch": 0.6813960934998399,
-      "grad_norm": 0.037796501070261,
-      "learning_rate": 9.959670993579808e-05,
-      "loss": 0.0164,
+      "epoch": 0.8597979797979798,
+      "grad_norm": 0.012683255597949028,
+      "learning_rate": 3.441138421733506e-05,
+      "loss": 0.004,
       "step": 266
     },
     {
-      "epoch": 0.6839577329490875,
-      "grad_norm": 0.025817422196269035,
-      "learning_rate": 9.95852854910781e-05,
-      "loss": 0.0095,
+      "epoch": 0.863030303030303,
+      "grad_norm": 0.02059134468436241,
+      "learning_rate": 3.454075032341527e-05,
+      "loss": 0.0078,
       "step": 267
     },
     {
-      "epoch": 0.6865193723983349,
-      "grad_norm": 0.027300599962472916,
-      "learning_rate": 9.95737021557578e-05,
-      "loss": 0.0124,
+      "epoch": 0.8662626262626263,
+      "grad_norm": 0.010593985207378864,
+      "learning_rate": 3.467011642949548e-05,
+      "loss": 0.0029,
       "step": 268
     },
     {
-      "epoch": 0.6890810118475824,
-      "grad_norm": 0.03461704030632973,
-      "learning_rate": 9.956195996695474e-05,
-      "loss": 0.0263,
+      "epoch": 0.8694949494949495,
+      "grad_norm": 0.01322435587644577,
+      "learning_rate": 3.4799482535575675e-05,
+      "loss": 0.0039,
       "step": 269
     },
     {
-      "epoch": 0.69164265129683,
-      "grad_norm": 0.03362032026052475,
-      "learning_rate": 9.955005896229543e-05,
-      "loss": 0.0176,
+      "epoch": 0.8727272727272727,
+      "grad_norm": 0.011545834131538868,
+      "learning_rate": 3.4928848641655885e-05,
+      "loss": 0.0031,
       "step": 270
     },
     {
-      "epoch": 0.6942042907460775,
-      "grad_norm": 0.027980942279100418,
-      "learning_rate": 9.953799917991535e-05,
-      "loss": 0.0148,
+      "epoch": 0.8759595959595959,
+      "grad_norm": 0.016838576644659042,
+      "learning_rate": 3.5058214747736096e-05,
+      "loss": 0.009,
       "step": 271
     },
     {
-      "epoch": 0.696765930195325,
-      "grad_norm": 0.030006052926182747,
-      "learning_rate": 9.952578065845874e-05,
-      "loss": 0.0158,
+      "epoch": 0.8791919191919192,
+      "grad_norm": 0.013376126997172832,
+      "learning_rate": 3.5187580853816306e-05,
+      "loss": 0.005,
       "step": 272
     },
     {
-      "epoch": 0.6993275696445725,
-      "grad_norm": 0.03503892198204994,
-      "learning_rate": 9.951340343707852e-05,
-      "loss": 0.0247,
+      "epoch": 0.8824242424242424,
+      "grad_norm": 0.012144695967435837,
+      "learning_rate": 3.531694695989651e-05,
+      "loss": 0.003,
       "step": 273
     },
     {
-      "epoch": 0.70188920909382,
-      "grad_norm": 0.025154704228043556,
-      "learning_rate": 9.950086755543613e-05,
-      "loss": 0.0125,
+      "epoch": 0.8856565656565657,
+      "grad_norm": 0.014813927002251148,
+      "learning_rate": 3.544631306597671e-05,
+      "loss": 0.0044,
       "step": 274
     },
     {
-      "epoch": 0.7044508485430676,
-      "grad_norm": 0.024434421211481094,
-      "learning_rate": 9.948817305370143e-05,
-      "loss": 0.0127,
+      "epoch": 0.8888888888888888,
+      "grad_norm": 0.011357988230884075,
+      "learning_rate": 3.557567917205692e-05,
+      "loss": 0.0025,
       "step": 275
     },
     {
-      "epoch": 0.7044508485430676,
-      "eval_loss": 0.011695047840476036,
-      "eval_runtime": 15.7735,
-      "eval_samples_per_second": 6.34,
-      "eval_steps_per_second": 1.585,
+      "epoch": 0.8888888888888888,
+      "eval_loss": 0.0033953615929931402,
+      "eval_runtime": 15.5672,
+      "eval_samples_per_second": 6.424,
+      "eval_steps_per_second": 1.606,
       "step": 275
     },
     {
-      "epoch": 0.7070124879923151,
-      "grad_norm": 0.02198687754571438,
-      "learning_rate": 9.947531997255256e-05,
-      "loss": 0.0111,
+      "epoch": 0.8921212121212121,
+      "grad_norm": 0.014036865904927254,
+      "learning_rate": 3.570504527813713e-05,
+      "loss": 0.0041,
       "step": 276
     },
     {
-      "epoch": 0.7095741274415626,
-      "grad_norm": 0.0338519886136055,
-      "learning_rate": 9.94623083531758e-05,
-      "loss": 0.0289,
+      "epoch": 0.8953535353535353,
+      "grad_norm": 0.01647082157433033,
+      "learning_rate": 3.583441138421734e-05,
+      "loss": 0.0042,
       "step": 277
     },
     {
-      "epoch": 0.7121357668908102,
-      "grad_norm": 0.02596919983625412,
-      "learning_rate": 9.944913823726546e-05,
-      "loss": 0.0113,
+      "epoch": 0.8985858585858586,
+      "grad_norm": 0.0176159106194973,
+      "learning_rate": 3.596377749029755e-05,
+      "loss": 0.0087,
       "step": 278
     },
     {
-      "epoch": 0.7146974063400576,
-      "grad_norm": 0.03738397732377052,
-      "learning_rate": 9.94358096670237e-05,
-      "loss": 0.0193,
+      "epoch": 0.9018181818181819,
+      "grad_norm": 0.011774410493671894,
+      "learning_rate": 3.609314359637775e-05,
+      "loss": 0.0028,
       "step": 279
     },
     {
-      "epoch": 0.7172590457893051,
-      "grad_norm": 0.020664948970079422,
-      "learning_rate": 9.94223226851605e-05,
-      "loss": 0.0136,
+      "epoch": 0.9050505050505051,
+      "grad_norm": 0.013519537635147572,
+      "learning_rate": 3.6222509702457954e-05,
+      "loss": 0.004,
       "step": 280
     },
     {
-      "epoch": 0.7198206852385527,
-      "grad_norm": 0.021599043160676956,
-      "learning_rate": 9.940867733489341e-05,
-      "loss": 0.0107,
+      "epoch": 0.9082828282828282,
+      "grad_norm": 0.014084594324231148,
+      "learning_rate": 3.6351875808538164e-05,
+      "loss": 0.0047,
       "step": 281
     },
     {
-      "epoch": 0.7223823246878002,
-      "grad_norm": 0.024695029482245445,
-      "learning_rate": 9.939487365994742e-05,
-      "loss": 0.0154,
+      "epoch": 0.9115151515151515,
+      "grad_norm": 0.017424482852220535,
+      "learning_rate": 3.648124191461837e-05,
+      "loss": 0.0072,
       "step": 282
     },
     {
-      "epoch": 0.7249439641370478,
-      "grad_norm": 0.028182033449411392,
-      "learning_rate": 9.938091170455494e-05,
-      "loss": 0.0141,
+      "epoch": 0.9147474747474748,
+      "grad_norm": 0.01249826792627573,
+      "learning_rate": 3.661060802069858e-05,
+      "loss": 0.0064,
       "step": 283
     },
     {
-      "epoch": 0.7275056035862952,
-      "grad_norm": 0.02947639487683773,
-      "learning_rate": 9.936679151345547e-05,
-      "loss": 0.021,
+      "epoch": 0.917979797979798,
+      "grad_norm": 0.012520184740424156,
+      "learning_rate": 3.673997412677879e-05,
+      "loss": 0.0053,
       "step": 284
     },
     {
-      "epoch": 0.7300672430355427,
-      "grad_norm": 0.025371797382831573,
-      "learning_rate": 9.935251313189564e-05,
-      "loss": 0.0152,
+      "epoch": 0.9212121212121213,
+      "grad_norm": 0.020676422864198685,
+      "learning_rate": 3.6869340232859e-05,
+      "loss": 0.0091,
       "step": 285
     },
     {
-      "epoch": 0.7326288824847903,
-      "grad_norm": 0.024596597999334335,
-      "learning_rate": 9.933807660562898e-05,
-      "loss": 0.0145,
+      "epoch": 0.9244444444444444,
+      "grad_norm": 0.012115418910980225,
+      "learning_rate": 3.6998706338939195e-05,
+      "loss": 0.0038,
       "step": 286
     },
     {
-      "epoch": 0.7351905219340378,
-      "grad_norm": 0.026564741507172585,
-      "learning_rate": 9.932348198091568e-05,
-      "loss": 0.0174,
+      "epoch": 0.9276767676767677,
+      "grad_norm": 0.014116480946540833,
+      "learning_rate": 3.7128072445019406e-05,
+      "loss": 0.0052,
       "step": 287
     },
     {
-      "epoch": 0.7377521613832853,
-      "grad_norm": 0.021402254700660706,
-      "learning_rate": 9.930872930452269e-05,
-      "loss": 0.0119,
+      "epoch": 0.9309090909090909,
+      "grad_norm": 0.01281814556568861,
+      "learning_rate": 3.7257438551099616e-05,
+      "loss": 0.0034,
       "step": 288
     },
     {
-      "epoch": 0.7403138008325328,
-      "grad_norm": 0.0332682728767395,
-      "learning_rate": 9.929381862372329e-05,
-      "loss": 0.0174,
+      "epoch": 0.9341414141414142,
+      "grad_norm": 0.011999557726085186,
+      "learning_rate": 3.738680465717982e-05,
+      "loss": 0.0033,
       "step": 289
     },
     {
-      "epoch": 0.7428754402817803,
-      "grad_norm": 0.027293076738715172,
-      "learning_rate": 9.927874998629714e-05,
-      "loss": 0.017,
+      "epoch": 0.9373737373737374,
+      "grad_norm": 0.019353436306118965,
+      "learning_rate": 3.751617076326003e-05,
+      "loss": 0.0112,
       "step": 290
     },
     {
-      "epoch": 0.7454370797310279,
-      "grad_norm": 0.023125778883695602,
-      "learning_rate": 9.926352344053002e-05,
-      "loss": 0.014,
+      "epoch": 0.9406060606060606,
+      "grad_norm": 0.014781869947910309,
+      "learning_rate": 3.764553686934023e-05,
+      "loss": 0.006,
       "step": 291
     },
     {
-      "epoch": 0.7479987191802754,
-      "grad_norm": 0.028066707774996758,
-      "learning_rate": 9.924813903521373e-05,
-      "loss": 0.0199,
+      "epoch": 0.9438383838383838,
+      "grad_norm": 0.010922358371317387,
+      "learning_rate": 3.777490297542044e-05,
+      "loss": 0.0022,
       "step": 292
     },
     {
-      "epoch": 0.7505603586295229,
-      "grad_norm": 0.025812916457653046,
-      "learning_rate": 9.923259681964592e-05,
-      "loss": 0.0143,
+      "epoch": 0.9470707070707071,
+      "grad_norm": 0.015073117800056934,
+      "learning_rate": 3.790426908150065e-05,
+      "loss": 0.0045,
       "step": 293
     },
     {
-      "epoch": 0.7531219980787704,
-      "grad_norm": 0.020010055974125862,
-      "learning_rate": 9.921689684362989e-05,
-      "loss": 0.0108,
+      "epoch": 0.9503030303030303,
+      "grad_norm": 0.013030101545155048,
+      "learning_rate": 3.803363518758086e-05,
+      "loss": 0.0033,
       "step": 294
     },
     {
-      "epoch": 0.7556836375280179,
-      "grad_norm": 0.027351340278983116,
-      "learning_rate": 9.92010391574745e-05,
-      "loss": 0.0142,
+      "epoch": 0.9535353535353536,
+      "grad_norm": 0.011426662094891071,
+      "learning_rate": 3.816300129366106e-05,
+      "loss": 0.0027,
       "step": 295
     },
     {
-      "epoch": 0.7582452769772654,
-      "grad_norm": 0.02779037319123745,
-      "learning_rate": 9.918502381199397e-05,
-      "loss": 0.0126,
+      "epoch": 0.9567676767676768,
+      "grad_norm": 0.014295799657702446,
+      "learning_rate": 3.829236739974127e-05,
+      "loss": 0.0028,
       "step": 296
     },
     {
-      "epoch": 0.760806916426513,
-      "grad_norm": 0.019299093633890152,
-      "learning_rate": 9.916885085850771e-05,
-      "loss": 0.0095,
+      "epoch": 0.96,
+      "grad_norm": 0.0171195175498724,
+      "learning_rate": 3.8421733505821475e-05,
+      "loss": 0.0047,
       "step": 297
     },
     {
-      "epoch": 0.7633685558757605,
-      "grad_norm": 0.02957143820822239,
-      "learning_rate": 9.915252034884018e-05,
-      "loss": 0.0244,
+      "epoch": 0.9632323232323232,
+      "grad_norm": 0.018932072445750237,
+      "learning_rate": 3.855109961190168e-05,
+      "loss": 0.0051,
       "step": 298
     },
     {
-      "epoch": 0.7659301953250081,
-      "grad_norm": 0.02845151349902153,
-      "learning_rate": 9.913603233532067e-05,
-      "loss": 0.0134,
+      "epoch": 0.9664646464646465,
+      "grad_norm": 0.016705717891454697,
+      "learning_rate": 3.868046571798189e-05,
+      "loss": 0.0038,
       "step": 299
     },
     {
-      "epoch": 0.7684918347742555,
-      "grad_norm": 0.019368290901184082,
-      "learning_rate": 9.911938687078324e-05,
-      "loss": 0.0095,
+      "epoch": 0.9696969696969697,
+      "grad_norm": 0.018960297107696533,
+      "learning_rate": 3.88098318240621e-05,
+      "loss": 0.0038,
       "step": 300
     },
     {
-      "epoch": 0.7684918347742555,
-      "eval_loss": 0.011247427202761173,
-      "eval_runtime": 15.7866,
-      "eval_samples_per_second": 6.334,
-      "eval_steps_per_second": 1.584,
+      "epoch": 0.9696969696969697,
+      "eval_loss": 0.0035091019235551357,
+      "eval_runtime": 15.4857,
+      "eval_samples_per_second": 6.458,
+      "eval_steps_per_second": 1.614,
       "step": 300
     },
     {
-      "epoch": 0.771053474223503,
-      "grad_norm": 0.027626069262623787,
-      "learning_rate": 9.910258400856646e-05,
-      "loss": 0.0131,
+      "epoch": 0.972929292929293,
+      "grad_norm": 0.015397065319120884,
+      "learning_rate": 3.893919793014231e-05,
+      "loss": 0.0043,
       "step": 301
     },
     {
-      "epoch": 0.7736151136727506,
-      "grad_norm": 0.018759949132800102,
-      "learning_rate": 9.908562380251323e-05,
-      "loss": 0.01,
+      "epoch": 0.9761616161616161,
+      "grad_norm": 0.012262605130672455,
+      "learning_rate": 3.906856403622251e-05,
+      "loss": 0.0035,
       "step": 302
     },
     {
-      "epoch": 0.7761767531219981,
-      "grad_norm": 0.0227984506636858,
-      "learning_rate": 9.906850630697068e-05,
-      "loss": 0.0103,
+      "epoch": 0.9793939393939394,
+      "grad_norm": 0.011876139789819717,
+      "learning_rate": 3.9197930142302716e-05,
+      "loss": 0.0029,
       "step": 303
     },
     {
-      "epoch": 0.7787383925712456,
-      "grad_norm": 0.022225648164749146,
-      "learning_rate": 9.905123157678995e-05,
-      "loss": 0.013,
+      "epoch": 0.9826262626262626,
+      "grad_norm": 0.02050217054784298,
+      "learning_rate": 3.9327296248382926e-05,
+      "loss": 0.0056,
       "step": 304
     },
     {
-      "epoch": 0.7813000320204931,
-      "grad_norm": 0.02687612920999527,
-      "learning_rate": 9.9033799667326e-05,
-      "loss": 0.0144,
+      "epoch": 0.9858585858585859,
+      "grad_norm": 0.01525296550244093,
+      "learning_rate": 3.945666235446313e-05,
+      "loss": 0.0057,
       "step": 305
     },
     {
-      "epoch": 0.7838616714697406,
-      "grad_norm": 0.027007579803466797,
-      "learning_rate": 9.901621063443747e-05,
-      "loss": 0.0244,
+      "epoch": 0.9890909090909091,
+      "grad_norm": 0.010231145657598972,
+      "learning_rate": 3.958602846054334e-05,
+      "loss": 0.0028,
       "step": 306
     },
     {
-      "epoch": 0.7864233109189882,
-      "grad_norm": 0.030599134042859077,
-      "learning_rate": 9.89984645344865e-05,
-      "loss": 0.0198,
+      "epoch": 0.9923232323232323,
+      "grad_norm": 0.018879607319831848,
+      "learning_rate": 3.971539456662355e-05,
+      "loss": 0.0086,
       "step": 307
     },
     {
-      "epoch": 0.7889849503682357,
-      "grad_norm": 0.025911178439855576,
-      "learning_rate": 9.898056142433851e-05,
-      "loss": 0.0113,
+      "epoch": 0.9955555555555555,
+      "grad_norm": 0.018934015184640884,
+      "learning_rate": 3.9844760672703754e-05,
+      "loss": 0.0082,
       "step": 308
     },
     {
-      "epoch": 0.7915465898174832,
-      "grad_norm": 0.028860868886113167,
-      "learning_rate": 9.896250136136205e-05,
-      "loss": 0.0146,
+      "epoch": 0.9987878787878788,
+      "grad_norm": 0.015406518243253231,
+      "learning_rate": 3.997412677878396e-05,
+      "loss": 0.0077,
       "step": 309
     },
     {
-      "epoch": 0.7941082292667307,
-      "grad_norm": 0.02513432502746582,
-      "learning_rate": 9.89442844034286e-05,
-      "loss": 0.0176,
+      "epoch": 1.002020202020202,
+      "grad_norm": 0.017992492765188217,
+      "learning_rate": 4.010349288486417e-05,
+      "loss": 0.0063,
       "step": 310
     },
     {
-      "epoch": 0.7966698687159782,
-      "grad_norm": 0.0275620985776186,
-      "learning_rate": 9.89259106089124e-05,
-      "loss": 0.0114,
+      "epoch": 1.0052525252525253,
+      "grad_norm": 0.0095031363889575,
+      "learning_rate": 4.023285899094437e-05,
+      "loss": 0.0034,
       "step": 311
     },
     {
-      "epoch": 0.7992315081652257,
-      "grad_norm": 0.01991882734000683,
-      "learning_rate": 9.890738003669029e-05,
-      "loss": 0.0108,
+      "epoch": 1.0084848484848485,
+      "grad_norm": 0.02079678885638714,
+      "learning_rate": 4.036222509702458e-05,
+      "loss": 0.0061,
       "step": 312
     },
     {
-      "epoch": 0.8017931476144733,
-      "grad_norm": 0.027305619791150093,
-      "learning_rate": 9.888869274614141e-05,
-      "loss": 0.0226,
+      "epoch": 1.0117171717171718,
+      "grad_norm": 0.016700437292456627,
+      "learning_rate": 4.049159120310479e-05,
+      "loss": 0.0036,
       "step": 313
     },
     {
-      "epoch": 0.8043547870637208,
-      "grad_norm": 0.025072338059544563,
-      "learning_rate": 9.886984879714713e-05,
-      "loss": 0.0122,
+      "epoch": 1.014949494949495,
+      "grad_norm": 0.01143790502101183,
+      "learning_rate": 4.0620957309184995e-05,
+      "loss": 0.0031,
       "step": 314
     },
     {
-      "epoch": 0.8069164265129684,
-      "grad_norm": 0.022108634933829308,
-      "learning_rate": 9.885084825009086e-05,
-      "loss": 0.0129,
+      "epoch": 1.018181818181818,
+      "grad_norm": 0.018342172726988792,
+      "learning_rate": 4.07503234152652e-05,
+      "loss": 0.0072,
       "step": 315
     },
     {
-      "epoch": 0.8094780659622158,
-      "grad_norm": 0.030420850962400436,
-      "learning_rate": 9.883169116585772e-05,
-      "loss": 0.0175,
+      "epoch": 1.0214141414141413,
+      "grad_norm": 0.012468880973756313,
+      "learning_rate": 4.087968952134541e-05,
+      "loss": 0.0032,
       "step": 316
     },
     {
-      "epoch": 0.8120397054114633,
-      "grad_norm": 0.03353031352162361,
-      "learning_rate": 9.88123776058345e-05,
-      "loss": 0.0144,
+      "epoch": 1.0246464646464646,
+      "grad_norm": 0.016855724155902863,
+      "learning_rate": 4.100905562742562e-05,
+      "loss": 0.0034,
       "step": 317
     },
     {
-      "epoch": 0.8146013448607109,
-      "grad_norm": 0.0281111728399992,
-      "learning_rate": 9.879290763190938e-05,
-      "loss": 0.0118,
+      "epoch": 1.0278787878787878,
+      "grad_norm": 0.0079115005210042,
+      "learning_rate": 4.113842173350582e-05,
+      "loss": 0.0025,
       "step": 318
     },
     {
-      "epoch": 0.8171629843099584,
-      "grad_norm": 0.02738821692764759,
-      "learning_rate": 9.877328130647175e-05,
-      "loss": 0.0114,
+      "epoch": 1.031111111111111,
+      "grad_norm": 0.010777114890515804,
+      "learning_rate": 4.126778783958603e-05,
+      "loss": 0.0027,
       "step": 319
     },
     {
-      "epoch": 0.8197246237592059,
-      "grad_norm": 0.026797035709023476,
-      "learning_rate": 9.875349869241201e-05,
-      "loss": 0.011,
+      "epoch": 1.0343434343434343,
+      "grad_norm": 0.01189594715833664,
+      "learning_rate": 4.1397153945666236e-05,
+      "loss": 0.0024,
       "step": 320
     },
     {
-      "epoch": 0.8222862632084534,
-      "grad_norm": 0.0358194038271904,
-      "learning_rate": 9.87335598531214e-05,
-      "loss": 0.0123,
+      "epoch": 1.0375757575757576,
+      "grad_norm": 0.009758715517818928,
+      "learning_rate": 4.152652005174644e-05,
+      "loss": 0.0025,
       "step": 321
     },
     {
-      "epoch": 0.8248479026577009,
-      "grad_norm": 0.022085469216108322,
-      "learning_rate": 9.87134648524917e-05,
-      "loss": 0.012,
+      "epoch": 1.0408080808080808,
+      "grad_norm": 0.015210709534585476,
+      "learning_rate": 4.165588615782665e-05,
+      "loss": 0.0072,
       "step": 322
     },
     {
-      "epoch": 0.8274095421069484,
-      "grad_norm": 0.02928551472723484,
-      "learning_rate": 9.869321375491516e-05,
-      "loss": 0.0203,
+      "epoch": 1.044040404040404,
+      "grad_norm": 0.01121676154434681,
+      "learning_rate": 4.178525226390686e-05,
+      "loss": 0.003,
       "step": 323
     },
     {
-      "epoch": 0.829971181556196,
-      "grad_norm": 0.025247866287827492,
-      "learning_rate": 9.867280662528419e-05,
-      "loss": 0.0123,
+      "epoch": 1.0472727272727274,
+      "grad_norm": 0.011481767520308495,
+      "learning_rate": 4.1914618369987064e-05,
+      "loss": 0.0037,
       "step": 324
     },
     {
-      "epoch": 0.8325328210054435,
-      "grad_norm": 0.026824558153748512,
-      "learning_rate": 9.865224352899119e-05,
-      "loss": 0.0142,
+      "epoch": 1.0505050505050506,
+      "grad_norm": 0.010206876322627068,
+      "learning_rate": 4.2043984476067274e-05,
+      "loss": 0.0025,
       "step": 325
     },
     {
-      "epoch": 0.8325328210054435,
-      "eval_loss": 0.011021601967513561,
-      "eval_runtime": 15.7711,
-      "eval_samples_per_second": 6.341,
-      "eval_steps_per_second": 1.585,
+      "epoch": 1.0505050505050506,
+      "eval_loss": 0.003410192672163248,
+      "eval_runtime": 15.5353,
+      "eval_samples_per_second": 6.437,
+      "eval_steps_per_second": 1.609,
       "step": 325
     },
     {
-      "epoch": 0.835094460454691,
-      "grad_norm": 0.028063297271728516,
-      "learning_rate": 9.863152453192836e-05,
-      "loss": 0.0123,
+      "epoch": 1.0537373737373736,
+      "grad_norm": 0.017532726749777794,
+      "learning_rate": 4.217335058214748e-05,
+      "loss": 0.006,
       "step": 326
     },
     {
-      "epoch": 0.8376560999039385,
-      "grad_norm": 0.026745524257421494,
-      "learning_rate": 9.861064970048744e-05,
-      "loss": 0.0121,
+      "epoch": 1.056969696969697,
+      "grad_norm": 0.010058831423521042,
+      "learning_rate": 4.230271668822768e-05,
+      "loss": 0.0028,
       "step": 327
     },
     {
-      "epoch": 0.840217739353186,
-      "grad_norm": 0.0259662177413702,
-      "learning_rate": 9.85896191015595e-05,
-      "loss": 0.0111,
+      "epoch": 1.0602020202020201,
+      "grad_norm": 0.019731098785996437,
+      "learning_rate": 4.243208279430789e-05,
+      "loss": 0.0036,
       "step": 328
     },
     {
-      "epoch": 0.8427793788024336,
-      "grad_norm": 0.03781949728727341,
-      "learning_rate": 9.856843280253485e-05,
-      "loss": 0.0228,
+      "epoch": 1.0634343434343434,
+      "grad_norm": 0.012425841763615608,
+      "learning_rate": 4.25614489003881e-05,
+      "loss": 0.0034,
       "step": 329
     },
     {
-      "epoch": 0.8453410182516811,
-      "grad_norm": 0.019805191084742546,
-      "learning_rate": 9.85470908713026e-05,
-      "loss": 0.0099,
+      "epoch": 1.0666666666666667,
+      "grad_norm": 0.014694919809699059,
+      "learning_rate": 4.269081500646831e-05,
+      "loss": 0.0039,
       "step": 330
     },
     {
-      "epoch": 0.8479026577009285,
-      "grad_norm": 0.03662043437361717,
-      "learning_rate": 9.852559337625066e-05,
-      "loss": 0.021,
+      "epoch": 1.06989898989899,
+      "grad_norm": 0.014344077557325363,
+      "learning_rate": 4.2820181112548515e-05,
+      "loss": 0.0045,
       "step": 331
     },
     {
-      "epoch": 0.8504642971501761,
-      "grad_norm": 0.021586574614048004,
-      "learning_rate": 9.850394038626539e-05,
-      "loss": 0.0111,
+      "epoch": 1.0731313131313132,
+      "grad_norm": 0.012936536222696304,
+      "learning_rate": 4.294954721862872e-05,
+      "loss": 0.0034,
       "step": 332
     },
     {
-      "epoch": 0.8530259365994236,
-      "grad_norm": 0.024759409949183464,
-      "learning_rate": 9.848213197073139e-05,
-      "loss": 0.0109,
+      "epoch": 1.0763636363636364,
+      "grad_norm": 0.013108815997838974,
+      "learning_rate": 4.307891332470893e-05,
+      "loss": 0.0039,
       "step": 333
     },
     {
-      "epoch": 0.8555875760486712,
-      "grad_norm": 0.027631640434265137,
-      "learning_rate": 9.846016819953136e-05,
-      "loss": 0.0106,
+      "epoch": 1.0795959595959597,
+      "grad_norm": 0.016730261966586113,
+      "learning_rate": 4.320827943078913e-05,
+      "loss": 0.0054,
       "step": 334
     },
     {
-      "epoch": 0.8581492154979187,
-      "grad_norm": 0.029770981520414352,
-      "learning_rate": 9.843804914304577e-05,
-      "loss": 0.0151,
+      "epoch": 1.082828282828283,
+      "grad_norm": 0.019490597769618034,
+      "learning_rate": 4.333764553686934e-05,
+      "loss": 0.0084,
       "step": 335
     },
     {
-      "epoch": 0.8607108549471661,
-      "grad_norm": 0.02814965881407261,
-      "learning_rate": 9.841577487215271e-05,
-      "loss": 0.0158,
+      "epoch": 1.086060606060606,
+      "grad_norm": 0.008431133814156055,
+      "learning_rate": 4.346701164294955e-05,
+      "loss": 0.0018,
       "step": 336
     },
     {
-      "epoch": 0.8632724943964137,
-      "grad_norm": 0.026514781638979912,
-      "learning_rate": 9.839334545822768e-05,
-      "loss": 0.013,
+      "epoch": 1.0892929292929292,
+      "grad_norm": 0.011448248289525509,
+      "learning_rate": 4.359637774902976e-05,
+      "loss": 0.0029,
       "step": 337
     },
     {
-      "epoch": 0.8658341338456612,
-      "grad_norm": 0.022735703736543655,
-      "learning_rate": 9.837076097314319e-05,
-      "loss": 0.0157,
+      "epoch": 1.0925252525252525,
+      "grad_norm": 0.016211332753300667,
+      "learning_rate": 4.372574385510996e-05,
+      "loss": 0.0072,
       "step": 338
     },
     {
-      "epoch": 0.8683957732949087,
-      "grad_norm": 0.026729697361588478,
-      "learning_rate": 9.834802148926882e-05,
-      "loss": 0.0189,
+      "epoch": 1.0957575757575757,
+      "grad_norm": 0.0187988318502903,
+      "learning_rate": 4.385510996119017e-05,
+      "loss": 0.006,
       "step": 339
     },
     {
-      "epoch": 0.8709574127441563,
-      "grad_norm": 0.025248195976018906,
-      "learning_rate": 9.83251270794707e-05,
-      "loss": 0.0121,
+      "epoch": 1.098989898989899,
+      "grad_norm": 0.009271546266973019,
+      "learning_rate": 4.3984476067270374e-05,
+      "loss": 0.003,
       "step": 340
     },
     {
-      "epoch": 0.8735190521934038,
-      "grad_norm": 0.025944946333765984,
-      "learning_rate": 9.830207781711147e-05,
-      "loss": 0.0135,
+      "epoch": 1.1022222222222222,
+      "grad_norm": 0.006935407407581806,
+      "learning_rate": 4.4113842173350584e-05,
+      "loss": 0.0018,
       "step": 341
     },
     {
-      "epoch": 0.8760806916426513,
-      "grad_norm": 0.025446701794862747,
-      "learning_rate": 9.827887377604995e-05,
-      "loss": 0.0189,
+      "epoch": 1.1054545454545455,
+      "grad_norm": 0.01265312172472477,
+      "learning_rate": 4.4243208279430794e-05,
+      "loss": 0.0058,
       "step": 342
     },
     {
-      "epoch": 0.8786423310918988,
-      "grad_norm": 0.0214429572224617,
-      "learning_rate": 9.825551503064094e-05,
-      "loss": 0.0118,
+      "epoch": 1.1086868686868687,
+      "grad_norm": 0.011098830960690975,
+      "learning_rate": 4.4372574385511e-05,
+      "loss": 0.0039,
       "step": 343
     },
     {
-      "epoch": 0.8812039705411463,
-      "grad_norm": 0.025962600484490395,
-      "learning_rate": 9.8232001655735e-05,
-      "loss": 0.0107,
+      "epoch": 1.111919191919192,
+      "grad_norm": 0.010020444169640541,
+      "learning_rate": 4.45019404915912e-05,
+      "loss": 0.0029,
       "step": 344
     },
     {
-      "epoch": 0.8837656099903939,
-      "grad_norm": 0.026138875633478165,
-      "learning_rate": 9.820833372667812e-05,
-      "loss": 0.0107,
+      "epoch": 1.1151515151515152,
+      "grad_norm": 0.01472175307571888,
+      "learning_rate": 4.463130659767141e-05,
+      "loss": 0.0068,
       "step": 345
     },
     {
-      "epoch": 0.8863272494396414,
-      "grad_norm": 0.0326177254319191,
-      "learning_rate": 9.81845113193116e-05,
-      "loss": 0.0136,
+      "epoch": 1.1183838383838385,
+      "grad_norm": 0.009227897971868515,
+      "learning_rate": 4.476067270375162e-05,
+      "loss": 0.0026,
       "step": 346
     },
     {
-      "epoch": 0.8888888888888888,
-      "grad_norm": 0.03477172553539276,
-      "learning_rate": 9.816053450997172e-05,
-      "loss": 0.0127,
+      "epoch": 1.1216161616161617,
+      "grad_norm": 0.017713405191898346,
+      "learning_rate": 4.4890038809831825e-05,
+      "loss": 0.0078,
       "step": 347
     },
     {
-      "epoch": 0.8914505283381364,
-      "grad_norm": 0.03032323718070984,
-      "learning_rate": 9.813640337548954e-05,
-      "loss": 0.025,
+      "epoch": 1.1248484848484848,
+      "grad_norm": 0.01242606621235609,
+      "learning_rate": 4.5019404915912036e-05,
+      "loss": 0.0045,
       "step": 348
     },
     {
-      "epoch": 0.8940121677873839,
-      "grad_norm": 0.02440054714679718,
-      "learning_rate": 9.811211799319062e-05,
-      "loss": 0.0143,
+      "epoch": 1.128080808080808,
+      "grad_norm": 0.021053126081824303,
+      "learning_rate": 4.514877102199224e-05,
+      "loss": 0.0138,
       "step": 349
     },
     {
-      "epoch": 0.8965738072366315,
-      "grad_norm": 0.02038225159049034,
-      "learning_rate": 9.80876784408948e-05,
-      "loss": 0.0134,
+      "epoch": 1.1313131313131313,
+      "grad_norm": 0.010382373817265034,
+      "learning_rate": 4.527813712807244e-05,
+      "loss": 0.0025,
       "step": 350
     },
     {
-      "epoch": 0.8965738072366315,
-      "eval_loss": 0.010415345430374146,
-      "eval_runtime": 15.6951,
-      "eval_samples_per_second": 6.371,
-      "eval_steps_per_second": 1.593,
+      "epoch": 1.1313131313131313,
+      "eval_loss": 0.00315338303335011,
+      "eval_runtime": 15.5746,
+      "eval_samples_per_second": 6.421,
+      "eval_steps_per_second": 1.605,
       "step": 350
     },
     {
-      "epoch": 0.899135446685879,
-      "grad_norm": 0.04363004118204117,
-      "learning_rate": 9.806308479691595e-05,
-      "loss": 0.0335,
+      "epoch": 1.1345454545454545,
+      "grad_norm": 0.011846295557916164,
+      "learning_rate": 4.540750323415265e-05,
+      "loss": 0.0025,
       "step": 351
     },
     {
-      "epoch": 0.9016970861351264,
-      "grad_norm": 0.029303282499313354,
-      "learning_rate": 9.803833714006169e-05,
-      "loss": 0.0184,
+      "epoch": 1.1377777777777778,
+      "grad_norm": 0.017388341948390007,
+      "learning_rate": 4.553686934023286e-05,
+      "loss": 0.0055,
       "step": 352
     },
     {
-      "epoch": 0.904258725584374,
-      "grad_norm": 0.02172860875725746,
-      "learning_rate": 9.801343554963317e-05,
-      "loss": 0.0105,
+      "epoch": 1.141010101010101,
+      "grad_norm": 0.010204081423580647,
+      "learning_rate": 4.566623544631307e-05,
+      "loss": 0.0034,
       "step": 353
     },
     {
-      "epoch": 0.9068203650336215,
-      "grad_norm": 0.023900095373392105,
-      "learning_rate": 9.798838010542482e-05,
-      "loss": 0.0104,
+      "epoch": 1.1442424242424243,
+      "grad_norm": 0.011596186086535454,
+      "learning_rate": 4.579560155239328e-05,
+      "loss": 0.0037,
       "step": 354
     },
     {
-      "epoch": 0.909382004482869,
-      "grad_norm": 0.02863960526883602,
-      "learning_rate": 9.796317088772403e-05,
-      "loss": 0.0191,
+      "epoch": 1.1474747474747475,
+      "grad_norm": 0.013310596346855164,
+      "learning_rate": 4.592496765847348e-05,
+      "loss": 0.0028,
       "step": 355
     },
     {
-      "epoch": 0.9119436439321166,
-      "grad_norm": 0.03499027341604233,
-      "learning_rate": 9.793780797731096e-05,
-      "loss": 0.0133,
+      "epoch": 1.1507070707070708,
+      "grad_norm": 0.010546430945396423,
+      "learning_rate": 4.6054333764553684e-05,
+      "loss": 0.0024,
       "step": 356
     },
     {
-      "epoch": 0.914505283381364,
-      "grad_norm": 0.025026291608810425,
-      "learning_rate": 9.791229145545831e-05,
-      "loss": 0.0105,
+      "epoch": 1.1539393939393938,
+      "grad_norm": 0.012264569289982319,
+      "learning_rate": 4.6183699870633894e-05,
+      "loss": 0.0028,
       "step": 357
     },
     {
-      "epoch": 0.9170669228306116,
-      "grad_norm": 0.019676843658089638,
-      "learning_rate": 9.788662140393096e-05,
-      "loss": 0.01,
+      "epoch": 1.157171717171717,
+      "grad_norm": 0.00728724617511034,
+      "learning_rate": 4.6313065976714105e-05,
+      "loss": 0.0018,
       "step": 358
     },
     {
-      "epoch": 0.9196285622798591,
-      "grad_norm": 0.029065057635307312,
-      "learning_rate": 9.786079790498575e-05,
-      "loss": 0.0164,
+      "epoch": 1.1604040404040403,
+      "grad_norm": 0.011520588770508766,
+      "learning_rate": 4.6442432082794315e-05,
+      "loss": 0.0038,
       "step": 359
     },
     {
-      "epoch": 0.9221902017291066,
-      "grad_norm": 0.028103569522500038,
-      "learning_rate": 9.783482104137127e-05,
-      "loss": 0.0104,
+      "epoch": 1.1636363636363636,
+      "grad_norm": 0.010877084918320179,
+      "learning_rate": 4.657179818887452e-05,
+      "loss": 0.0033,
       "step": 360
     },
     {
-      "epoch": 0.9247518411783542,
-      "grad_norm": 0.02853201888501644,
-      "learning_rate": 9.780869089632749e-05,
-      "loss": 0.021,
+      "epoch": 1.1668686868686868,
+      "grad_norm": 0.02235741913318634,
+      "learning_rate": 4.670116429495472e-05,
+      "loss": 0.0062,
       "step": 361
     },
     {
-      "epoch": 0.9273134806276017,
-      "grad_norm": 0.022803885862231255,
-      "learning_rate": 9.778240755358564e-05,
-      "loss": 0.011,
+      "epoch": 1.17010101010101,
+      "grad_norm": 0.009432629682123661,
+      "learning_rate": 4.683053040103493e-05,
+      "loss": 0.0022,
       "step": 362
     },
     {
-      "epoch": 0.9298751200768491,
-      "grad_norm": 0.024698354303836823,
-      "learning_rate": 9.775597109736776e-05,
-      "loss": 0.0126,
+      "epoch": 1.1733333333333333,
+      "grad_norm": 0.011103375814855099,
+      "learning_rate": 4.6959896507115136e-05,
+      "loss": 0.0028,
       "step": 363
     },
     {
-      "epoch": 0.9324367595260967,
-      "grad_norm": 0.027431121096014977,
-      "learning_rate": 9.77293816123866e-05,
-      "loss": 0.0199,
+      "epoch": 1.1765656565656566,
+      "grad_norm": 0.010600670240819454,
+      "learning_rate": 4.7089262613195346e-05,
+      "loss": 0.0034,
       "step": 364
     },
     {
-      "epoch": 0.9349983989753442,
-      "grad_norm": 0.02329852059483528,
-      "learning_rate": 9.770263918384524e-05,
-      "loss": 0.012,
+      "epoch": 1.1797979797979798,
+      "grad_norm": 0.010880649089813232,
+      "learning_rate": 4.7218628719275556e-05,
+      "loss": 0.0028,
       "step": 365
     },
     {
-      "epoch": 0.9375600384245918,
-      "grad_norm": 0.019519859924912453,
-      "learning_rate": 9.767574389743682e-05,
-      "loss": 0.0083,
+      "epoch": 1.183030303030303,
+      "grad_norm": 0.017898019403219223,
+      "learning_rate": 4.734799482535576e-05,
+      "loss": 0.0054,
       "step": 366
     },
     {
-      "epoch": 0.9401216778738393,
-      "grad_norm": 0.02185627445578575,
-      "learning_rate": 9.764869583934437e-05,
-      "loss": 0.0078,
+      "epoch": 1.1862626262626264,
+      "grad_norm": 0.014189201407134533,
+      "learning_rate": 4.747736093143596e-05,
+      "loss": 0.0052,
       "step": 367
     },
     {
-      "epoch": 0.9426833173230867,
-      "grad_norm": 0.03656851127743721,
-      "learning_rate": 9.76214950962404e-05,
-      "loss": 0.017,
+      "epoch": 1.1894949494949496,
+      "grad_norm": 0.012748179957270622,
+      "learning_rate": 4.760672703751617e-05,
+      "loss": 0.0045,
       "step": 368
     },
     {
-      "epoch": 0.9452449567723343,
-      "grad_norm": 0.025890717282891273,
-      "learning_rate": 9.759414175528672e-05,
-      "loss": 0.0171,
+      "epoch": 1.1927272727272726,
+      "grad_norm": 0.01102366577833891,
+      "learning_rate": 4.773609314359638e-05,
+      "loss": 0.0034,
       "step": 369
     },
     {
-      "epoch": 0.9478065962215818,
-      "grad_norm": 0.02669275552034378,
-      "learning_rate": 9.75666359041341e-05,
-      "loss": 0.0129,
+      "epoch": 1.195959595959596,
+      "grad_norm": 0.010250872932374477,
+      "learning_rate": 4.786545924967659e-05,
+      "loss": 0.003,
       "step": 370
     },
     {
-      "epoch": 0.9503682356708293,
-      "grad_norm": 0.0161345936357975,
-      "learning_rate": 9.753897763092202e-05,
-      "loss": 0.01,
+      "epoch": 1.1991919191919191,
+      "grad_norm": 0.015184353105723858,
+      "learning_rate": 4.79948253557568e-05,
+      "loss": 0.0077,
       "step": 371
     },
     {
-      "epoch": 0.9529298751200769,
-      "grad_norm": 0.023527219891548157,
-      "learning_rate": 9.751116702427836e-05,
-      "loss": 0.0107,
+      "epoch": 1.2024242424242424,
+      "grad_norm": 0.008492082357406616,
+      "learning_rate": 4.8124191461837e-05,
+      "loss": 0.0023,
       "step": 372
     },
     {
-      "epoch": 0.9554915145693244,
-      "grad_norm": 0.0295490100979805,
-      "learning_rate": 9.748320417331915e-05,
-      "loss": 0.0151,
+      "epoch": 1.2056565656565656,
+      "grad_norm": 0.01109049841761589,
+      "learning_rate": 4.8253557567917204e-05,
+      "loss": 0.0047,
       "step": 373
     },
     {
-      "epoch": 0.9580531540185719,
-      "grad_norm": 0.027163468301296234,
-      "learning_rate": 9.745508916764832e-05,
-      "loss": 0.0147,
+      "epoch": 1.208888888888889,
+      "grad_norm": 0.017075804993510246,
+      "learning_rate": 4.8382923673997415e-05,
+      "loss": 0.0047,
       "step": 374
     },
     {
-      "epoch": 0.9606147934678194,
-      "grad_norm": 0.025246214121580124,
-      "learning_rate": 9.742682209735727e-05,
-      "loss": 0.0124,
+      "epoch": 1.2121212121212122,
+      "grad_norm": 0.00920112431049347,
+      "learning_rate": 4.8512289780077625e-05,
+      "loss": 0.0028,
       "step": 375
     },
     {
-      "epoch": 0.9606147934678194,
-      "eval_loss": 0.009683902375400066,
-      "eval_runtime": 15.6724,
-      "eval_samples_per_second": 6.381,
-      "eval_steps_per_second": 1.595,
+      "epoch": 1.2121212121212122,
+      "eval_loss": 0.0030464103911072016,
+      "eval_runtime": 15.5314,
+      "eval_samples_per_second": 6.439,
+      "eval_steps_per_second": 1.61,
       "step": 375
     },
     {
-      "epoch": 0.9631764329170669,
-      "grad_norm": 0.02218458242714405,
-      "learning_rate": 9.739840305302477e-05,
-      "loss": 0.0111,
+      "epoch": 1.2153535353535354,
+      "grad_norm": 0.013430137187242508,
+      "learning_rate": 4.864165588615783e-05,
+      "loss": 0.0036,
       "step": 376
     },
     {
-      "epoch": 0.9657380723663145,
-      "grad_norm": 0.03978296369314194,
-      "learning_rate": 9.736983212571646e-05,
-      "loss": 0.0208,
+      "epoch": 1.2185858585858587,
+      "grad_norm": 0.009438606910407543,
+      "learning_rate": 4.877102199223804e-05,
+      "loss": 0.0027,
       "step": 377
     },
     {
-      "epoch": 0.968299711815562,
-      "grad_norm": 0.031306296586990356,
-      "learning_rate": 9.73411094069848e-05,
-      "loss": 0.0159,
+      "epoch": 1.221818181818182,
+      "grad_norm": 0.012487027794122696,
+      "learning_rate": 4.890038809831824e-05,
+      "loss": 0.0038,
       "step": 378
     },
     {
-      "epoch": 0.9708613512648094,
-      "grad_norm": 0.020451745018363,
-      "learning_rate": 9.731223498886855e-05,
-      "loss": 0.0095,
+      "epoch": 1.225050505050505,
+      "grad_norm": 0.012544943019747734,
+      "learning_rate": 4.9029754204398446e-05,
+      "loss": 0.0034,
       "step": 379
     },
     {
-      "epoch": 0.973422990714057,
-      "grad_norm": 0.030514508485794067,
-      "learning_rate": 9.728320896389263e-05,
-      "loss": 0.016,
+      "epoch": 1.2282828282828282,
+      "grad_norm": 0.010194891132414341,
+      "learning_rate": 4.9159120310478656e-05,
+      "loss": 0.0027,
       "step": 380
     },
     {
-      "epoch": 0.9759846301633045,
-      "grad_norm": 0.023473547771573067,
-      "learning_rate": 9.725403142506774e-05,
-      "loss": 0.0117,
+      "epoch": 1.2315151515151515,
+      "grad_norm": 0.013430687598884106,
+      "learning_rate": 4.9288486416558866e-05,
+      "loss": 0.0055,
       "step": 381
     },
     {
-      "epoch": 0.9785462696125521,
-      "grad_norm": 0.025907503440976143,
-      "learning_rate": 9.722470246589008e-05,
-      "loss": 0.0116,
+      "epoch": 1.2347474747474747,
+      "grad_norm": 0.01295363437384367,
+      "learning_rate": 4.941785252263907e-05,
+      "loss": 0.0028,
       "step": 382
     },
     {
-      "epoch": 0.9811079090617996,
-      "grad_norm": 0.02568023093044758,
-      "learning_rate": 9.719522218034112e-05,
-      "loss": 0.0122,
+      "epoch": 1.237979797979798,
+      "grad_norm": 0.009068100713193417,
+      "learning_rate": 4.954721862871928e-05,
+      "loss": 0.0023,
       "step": 383
     },
     {
-      "epoch": 0.983669548511047,
-      "grad_norm": 0.03493017703294754,
-      "learning_rate": 9.716559066288715e-05,
-      "loss": 0.0165,
+      "epoch": 1.2412121212121212,
+      "grad_norm": 0.011907044798135757,
+      "learning_rate": 4.9676584734799483e-05,
+      "loss": 0.0026,
       "step": 384
     },
     {
-      "epoch": 0.9862311879602946,
-      "grad_norm": 0.02209441363811493,
-      "learning_rate": 9.713580800847916e-05,
-      "loss": 0.0154,
+      "epoch": 1.2444444444444445,
+      "grad_norm": 0.022994644939899445,
+      "learning_rate": 4.980595084087969e-05,
+      "loss": 0.0104,
       "step": 385
     },
     {
-      "epoch": 0.9887928274095421,
-      "grad_norm": 0.03291688114404678,
-      "learning_rate": 9.710587431255236e-05,
-      "loss": 0.0087,
+      "epoch": 1.2476767676767677,
+      "grad_norm": 0.014316575601696968,
+      "learning_rate": 4.99353169469599e-05,
+      "loss": 0.0044,
       "step": 386
     },
     {
-      "epoch": 0.9913544668587896,
-      "grad_norm": 0.025987064465880394,
-      "learning_rate": 9.7075789671026e-05,
-      "loss": 0.0113,
+      "epoch": 1.250909090909091,
+      "grad_norm": 0.01257970929145813,
+      "learning_rate": 5.006468305304011e-05,
+      "loss": 0.0042,
       "step": 387
     },
     {
-      "epoch": 0.9939161063080372,
-      "grad_norm": 0.03901522606611252,
-      "learning_rate": 9.7045554180303e-05,
-      "loss": 0.013,
+      "epoch": 1.2541414141414142,
+      "grad_norm": 0.009610496461391449,
+      "learning_rate": 5.019404915912032e-05,
+      "loss": 0.0029,
       "step": 388
     },
     {
-      "epoch": 0.9964777457572846,
-      "grad_norm": 0.020256483927369118,
-      "learning_rate": 9.701516793726967e-05,
-      "loss": 0.0105,
+      "epoch": 1.2573737373737375,
+      "grad_norm": 0.011172009631991386,
+      "learning_rate": 5.032341526520052e-05,
+      "loss": 0.0034,
       "step": 389
     },
     {
-      "epoch": 0.9990393852065321,
-      "grad_norm": 0.03212558478116989,
-      "learning_rate": 9.698463103929542e-05,
-      "loss": 0.0211,
+      "epoch": 1.2606060606060607,
+      "grad_norm": 0.011792114935815334,
+      "learning_rate": 5.0452781371280725e-05,
+      "loss": 0.0034,
       "step": 390
     },
     {
-      "epoch": 1.0019212295869357,
-      "grad_norm": 0.04988866299390793,
-      "learning_rate": 9.695394358423236e-05,
-      "loss": 0.0198,
+      "epoch": 1.2638383838383838,
+      "grad_norm": 0.007270593661814928,
+      "learning_rate": 5.058214747736093e-05,
+      "loss": 0.0025,
       "step": 391
     },
     {
-      "epoch": 1.0044828690361831,
-      "grad_norm": 0.025478005409240723,
-      "learning_rate": 9.692310567041505e-05,
-      "loss": 0.0124,
+      "epoch": 1.267070707070707,
+      "grad_norm": 0.008094442076981068,
+      "learning_rate": 5.071151358344114e-05,
+      "loss": 0.003,
       "step": 392
     },
     {
-      "epoch": 1.0070445084854307,
-      "grad_norm": 0.01753000169992447,
-      "learning_rate": 9.689211739666023e-05,
-      "loss": 0.009,
+      "epoch": 1.2703030303030303,
+      "grad_norm": 0.012139970436692238,
+      "learning_rate": 5.084087968952135e-05,
+      "loss": 0.004,
       "step": 393
     },
     {
-      "epoch": 1.0096061479346783,
-      "grad_norm": 0.02424152009189129,
-      "learning_rate": 9.68609788622664e-05,
-      "loss": 0.0094,
+      "epoch": 1.2735353535353535,
+      "grad_norm": 0.014527830295264721,
+      "learning_rate": 5.097024579560156e-05,
+      "loss": 0.0072,
       "step": 394
     },
     {
-      "epoch": 1.0121677873839257,
-      "grad_norm": 0.02567497454583645,
-      "learning_rate": 9.682969016701358e-05,
-      "loss": 0.0162,
+      "epoch": 1.2767676767676768,
+      "grad_norm": 0.010034168139100075,
+      "learning_rate": 5.109961190168176e-05,
+      "loss": 0.0031,
       "step": 395
     },
     {
-      "epoch": 1.0147294268331732,
-      "grad_norm": 0.02308768406510353,
-      "learning_rate": 9.679825141116296e-05,
-      "loss": 0.0094,
+      "epoch": 1.28,
+      "grad_norm": 0.010067771188914776,
+      "learning_rate": 5.122897800776197e-05,
+      "loss": 0.0025,
       "step": 396
     },
     {
-      "epoch": 1.0172910662824208,
-      "grad_norm": 0.02406313829123974,
-      "learning_rate": 9.676666269545655e-05,
-      "loss": 0.0107,
+      "epoch": 1.2832323232323233,
+      "grad_norm": 0.010714643634855747,
+      "learning_rate": 5.135834411384217e-05,
+      "loss": 0.004,
       "step": 397
     },
     {
-      "epoch": 1.0198527057316682,
-      "grad_norm": 0.026956558227539062,
-      "learning_rate": 9.673492412111698e-05,
-      "loss": 0.0142,
+      "epoch": 1.2864646464646465,
+      "grad_norm": 0.017075076699256897,
+      "learning_rate": 5.148771021992238e-05,
+      "loss": 0.0052,
       "step": 398
     },
     {
-      "epoch": 1.0224143451809158,
-      "grad_norm": 0.0189517792314291,
-      "learning_rate": 9.670303578984697e-05,
-      "loss": 0.0093,
+      "epoch": 1.2896969696969696,
+      "grad_norm": 0.013271777890622616,
+      "learning_rate": 5.161707632600259e-05,
+      "loss": 0.0046,
       "step": 399
     },
     {
-      "epoch": 1.0249759846301634,
-      "grad_norm": 0.025387246161699295,
-      "learning_rate": 9.66709978038292e-05,
-      "loss": 0.0122,
+      "epoch": 1.2929292929292928,
+      "grad_norm": 0.02206435054540634,
+      "learning_rate": 5.17464424320828e-05,
+      "loss": 0.0065,
       "step": 400
     },
     {
-      "epoch": 1.0249759846301634,
-      "eval_loss": 0.009496361948549747,
-      "eval_runtime": 15.6763,
-      "eval_samples_per_second": 6.379,
-      "eval_steps_per_second": 1.595,
+      "epoch": 1.2929292929292928,
+      "eval_loss": 0.003116590902209282,
+      "eval_runtime": 15.6093,
+      "eval_samples_per_second": 6.406,
+      "eval_steps_per_second": 1.602,
       "step": 400
     },
     {
-      "epoch": 1.0275376240794107,
-      "grad_norm": 0.020283183082938194,
-      "learning_rate": 9.663881026572586e-05,
-      "loss": 0.0109,
+      "epoch": 1.296161616161616,
+      "grad_norm": 0.017453907057642937,
+      "learning_rate": 5.1875808538163004e-05,
+      "loss": 0.0054,
       "step": 401
     },
     {
-      "epoch": 1.0300992635286583,
-      "grad_norm": 0.02017321065068245,
-      "learning_rate": 9.66064732786784e-05,
-      "loss": 0.0142,
+      "epoch": 1.2993939393939393,
+      "grad_norm": 0.010734484530985355,
+      "learning_rate": 5.2005174644243214e-05,
+      "loss": 0.0026,
       "step": 402
     },
     {
-      "epoch": 1.032660902977906,
-      "grad_norm": 0.02185225673019886,
-      "learning_rate": 9.657398694630712e-05,
-      "loss": 0.0069,
+      "epoch": 1.3026262626262626,
+      "grad_norm": 0.008877218700945377,
+      "learning_rate": 5.213454075032341e-05,
+      "loss": 0.0025,
       "step": 403
     },
     {
-      "epoch": 1.0352225424271533,
-      "grad_norm": 0.02693125233054161,
-      "learning_rate": 9.654135137271093e-05,
-      "loss": 0.0121,
+      "epoch": 1.3058585858585858,
+      "grad_norm": 0.00901103112846613,
+      "learning_rate": 5.226390685640362e-05,
+      "loss": 0.0028,
       "step": 404
     },
     {
-      "epoch": 1.0377841818764009,
-      "grad_norm": 0.026309996843338013,
-      "learning_rate": 9.650856666246693e-05,
-      "loss": 0.0144,
+      "epoch": 1.309090909090909,
+      "grad_norm": 0.01145117450505495,
+      "learning_rate": 5.239327296248383e-05,
+      "loss": 0.0046,
       "step": 405
     },
     {
-      "epoch": 1.0403458213256485,
-      "grad_norm": 0.022313056513667107,
-      "learning_rate": 9.64756329206301e-05,
-      "loss": 0.0086,
+      "epoch": 1.3123232323232323,
+      "grad_norm": 0.010362985543906689,
+      "learning_rate": 5.252263906856404e-05,
+      "loss": 0.0028,
       "step": 406
     },
     {
-      "epoch": 1.042907460774896,
-      "grad_norm": 0.033154942095279694,
-      "learning_rate": 9.644255025273304e-05,
-      "loss": 0.0173,
+      "epoch": 1.3155555555555556,
+      "grad_norm": 0.008662184700369835,
+      "learning_rate": 5.2652005174644245e-05,
+      "loss": 0.0028,
       "step": 407
     },
     {
-      "epoch": 1.0454691002241434,
-      "grad_norm": 0.021595394238829613,
-      "learning_rate": 9.640931876478547e-05,
-      "loss": 0.0108,
+      "epoch": 1.3187878787878788,
+      "grad_norm": 0.015215790830552578,
+      "learning_rate": 5.2781371280724455e-05,
+      "loss": 0.0053,
       "step": 408
     },
     {
-      "epoch": 1.048030739673391,
-      "grad_norm": 0.02003701590001583,
-      "learning_rate": 9.637593856327411e-05,
-      "loss": 0.0098,
+      "epoch": 1.322020202020202,
+      "grad_norm": 0.010266825556755066,
+      "learning_rate": 5.2910737386804666e-05,
+      "loss": 0.0029,
       "step": 409
     },
     {
-      "epoch": 1.0505923791226386,
-      "grad_norm": 0.022105243057012558,
-      "learning_rate": 9.634240975516209e-05,
-      "loss": 0.0088,
+      "epoch": 1.3252525252525253,
+      "grad_norm": 0.010717876255512238,
+      "learning_rate": 5.304010349288486e-05,
+      "loss": 0.0031,
       "step": 410
     },
     {
-      "epoch": 1.053154018571886,
-      "grad_norm": 0.01878410391509533,
-      "learning_rate": 9.630873244788883e-05,
-      "loss": 0.0081,
+      "epoch": 1.3284848484848486,
+      "grad_norm": 0.011546500958502293,
+      "learning_rate": 5.316946959896507e-05,
+      "loss": 0.0032,
       "step": 411
     },
     {
-      "epoch": 1.0557156580211335,
-      "grad_norm": 0.02455531433224678,
-      "learning_rate": 9.627490674936954e-05,
-      "loss": 0.0132,
+      "epoch": 1.3317171717171719,
+      "grad_norm": 0.015700813382864,
+      "learning_rate": 5.329883570504528e-05,
+      "loss": 0.0038,
       "step": 412
     },
     {
-      "epoch": 1.0582772974703811,
-      "grad_norm": 0.02545606903731823,
-      "learning_rate": 9.624093276799498e-05,
-      "loss": 0.0106,
+      "epoch": 1.3349494949494949,
+      "grad_norm": 0.007087106816470623,
+      "learning_rate": 5.3428201811125486e-05,
+      "loss": 0.0021,
       "step": 413
     },
     {
-      "epoch": 1.0608389369196285,
-      "grad_norm": 0.019652288407087326,
-      "learning_rate": 9.6206810612631e-05,
-      "loss": 0.0105,
+      "epoch": 1.3381818181818181,
+      "grad_norm": 0.008797608315944672,
+      "learning_rate": 5.35575679172057e-05,
+      "loss": 0.0029,
       "step": 414
     },
     {
-      "epoch": 1.063400576368876,
-      "grad_norm": 0.022004250437021255,
-      "learning_rate": 9.617254039261834e-05,
-      "loss": 0.0136,
+      "epoch": 1.3414141414141414,
+      "grad_norm": 0.008964994922280312,
+      "learning_rate": 5.368693402328591e-05,
+      "loss": 0.002,
       "step": 415
     },
     {
-      "epoch": 1.0659622158181237,
-      "grad_norm": 0.02517751231789589,
-      "learning_rate": 9.613812221777212e-05,
-      "loss": 0.0133,
+      "epoch": 1.3446464646464646,
+      "grad_norm": 0.014129131101071835,
+      "learning_rate": 5.3816300129366104e-05,
+      "loss": 0.0048,
       "step": 416
     },
     {
-      "epoch": 1.068523855267371,
-      "grad_norm": 0.021614955738186836,
-      "learning_rate": 9.610355619838163e-05,
-      "loss": 0.0094,
+      "epoch": 1.347878787878788,
+      "grad_norm": 0.008821067400276661,
+      "learning_rate": 5.3945666235446314e-05,
+      "loss": 0.0031,
       "step": 417
     },
     {
-      "epoch": 1.0710854947166186,
-      "grad_norm": 0.038878973573446274,
-      "learning_rate": 9.606884244520987e-05,
-      "loss": 0.0084,
+      "epoch": 1.3511111111111112,
+      "grad_norm": 0.01535008754581213,
+      "learning_rate": 5.4075032341526524e-05,
+      "loss": 0.0082,
       "step": 418
     },
     {
-      "epoch": 1.0736471341658662,
-      "grad_norm": 0.020645610988140106,
-      "learning_rate": 9.603398106949326e-05,
-      "loss": 0.0109,
+      "epoch": 1.3543434343434344,
+      "grad_norm": 0.007779187057167292,
+      "learning_rate": 5.420439844760673e-05,
+      "loss": 0.0027,
       "step": 419
     },
     {
-      "epoch": 1.0762087736151136,
-      "grad_norm": 0.018204625695943832,
-      "learning_rate": 9.599897218294122e-05,
-      "loss": 0.0082,
+      "epoch": 1.3575757575757577,
+      "grad_norm": 0.010275239124894142,
+      "learning_rate": 5.433376455368694e-05,
+      "loss": 0.0039,
       "step": 420
     },
     {
-      "epoch": 1.0787704130643612,
-      "grad_norm": 0.02104598470032215,
-      "learning_rate": 9.59638158977359e-05,
-      "loss": 0.009,
+      "epoch": 1.3608080808080807,
+      "grad_norm": 0.010350524447858334,
+      "learning_rate": 5.446313065976715e-05,
+      "loss": 0.0036,
       "step": 421
     },
     {
-      "epoch": 1.0813320525136088,
-      "grad_norm": 0.03736334294080734,
-      "learning_rate": 9.592851232653175e-05,
-      "loss": 0.0177,
+      "epoch": 1.364040404040404,
+      "grad_norm": 0.009732598438858986,
+      "learning_rate": 5.4592496765847345e-05,
+      "loss": 0.0043,
       "step": 422
     },
     {
-      "epoch": 1.0838936919628561,
-      "grad_norm": 0.026802612468600273,
-      "learning_rate": 9.58930615824552e-05,
-      "loss": 0.0111,
+      "epoch": 1.3672727272727272,
+      "grad_norm": 0.013486373238265514,
+      "learning_rate": 5.4721862871927555e-05,
+      "loss": 0.0035,
       "step": 423
     },
     {
-      "epoch": 1.0864553314121037,
-      "grad_norm": 0.01986725442111492,
-      "learning_rate": 9.585746377910425e-05,
-      "loss": 0.0081,
+      "epoch": 1.3705050505050504,
+      "grad_norm": 0.013745920732617378,
+      "learning_rate": 5.4851228978007766e-05,
+      "loss": 0.0043,
       "step": 424
     },
     {
-      "epoch": 1.0890169708613513,
-      "grad_norm": 0.03875807672739029,
-      "learning_rate": 9.582171903054816e-05,
-      "loss": 0.0211,
+      "epoch": 1.3737373737373737,
+      "grad_norm": 0.007328703533858061,
+      "learning_rate": 5.498059508408797e-05,
+      "loss": 0.0021,
       "step": 425
     },
     {
-      "epoch": 1.0890169708613513,
-      "eval_loss": 0.009211279451847076,
-      "eval_runtime": 15.6813,
-      "eval_samples_per_second": 6.377,
-      "eval_steps_per_second": 1.594,
+      "epoch": 1.3737373737373737,
+      "eval_loss": 0.003028572304174304,
+      "eval_runtime": 15.523,
+      "eval_samples_per_second": 6.442,
+      "eval_steps_per_second": 1.611,
       "step": 425
     },
     {
-      "epoch": 1.0915786103105989,
-      "grad_norm": 0.019004806876182556,
-      "learning_rate": 9.578582745132704e-05,
-      "loss": 0.0069,
+      "epoch": 1.376969696969697,
+      "grad_norm": 0.014379718340933323,
+      "learning_rate": 5.510996119016818e-05,
+      "loss": 0.004,
       "step": 426
     },
     {
-      "epoch": 1.0941402497598463,
-      "grad_norm": 0.024430232122540474,
-      "learning_rate": 9.574978915645152e-05,
-      "loss": 0.0088,
+      "epoch": 1.3802020202020202,
+      "grad_norm": 0.010442657396197319,
+      "learning_rate": 5.523932729624839e-05,
+      "loss": 0.0032,
       "step": 427
     },
     {
-      "epoch": 1.0967018892090938,
-      "grad_norm": 0.02753954567015171,
-      "learning_rate": 9.571360426140235e-05,
-      "loss": 0.015,
+      "epoch": 1.3834343434343435,
+      "grad_norm": 0.01950971782207489,
+      "learning_rate": 5.53686934023286e-05,
+      "loss": 0.011,
       "step": 428
     },
     {
-      "epoch": 1.0992635286583414,
-      "grad_norm": 0.0206803847104311,
-      "learning_rate": 9.567727288213005e-05,
-      "loss": 0.0093,
+      "epoch": 1.3866666666666667,
+      "grad_norm": 0.012882012873888016,
+      "learning_rate": 5.5498059508408797e-05,
+      "loss": 0.0056,
       "step": 429
     },
     {
-      "epoch": 1.1018251681075888,
-      "grad_norm": 0.02357391081750393,
-      "learning_rate": 9.564079513505455e-05,
-      "loss": 0.0142,
+      "epoch": 1.38989898989899,
+      "grad_norm": 0.013975840061903,
+      "learning_rate": 5.562742561448901e-05,
+      "loss": 0.0041,
       "step": 430
     },
     {
-      "epoch": 1.1043868075568364,
-      "grad_norm": 0.0350952111184597,
-      "learning_rate": 9.560417113706477e-05,
-      "loss": 0.0139,
+      "epoch": 1.3931313131313132,
+      "grad_norm": 0.01314020436257124,
+      "learning_rate": 5.575679172056921e-05,
+      "loss": 0.0044,
       "step": 431
     },
     {
-      "epoch": 1.106948447006084,
-      "grad_norm": 0.023076610639691353,
-      "learning_rate": 9.556740100551829e-05,
-      "loss": 0.0101,
+      "epoch": 1.3963636363636365,
+      "grad_norm": 0.008707460016012192,
+      "learning_rate": 5.588615782664942e-05,
+      "loss": 0.0028,
       "step": 432
     },
     {
-      "epoch": 1.1095100864553313,
-      "grad_norm": 0.023970071226358414,
-      "learning_rate": 9.553048485824095e-05,
-      "loss": 0.0104,
+      "epoch": 1.3995959595959597,
+      "grad_norm": 0.013438042253255844,
+      "learning_rate": 5.601552393272963e-05,
+      "loss": 0.004,
       "step": 433
     },
     {
-      "epoch": 1.112071725904579,
-      "grad_norm": 0.02964206598699093,
-      "learning_rate": 9.549342281352649e-05,
-      "loss": 0.0231,
+      "epoch": 1.4028282828282828,
+      "grad_norm": 0.007997852750122547,
+      "learning_rate": 5.614489003880984e-05,
+      "loss": 0.0029,
       "step": 434
     },
     {
-      "epoch": 1.1146333653538265,
-      "grad_norm": 0.020167216658592224,
-      "learning_rate": 9.54562149901362e-05,
-      "loss": 0.0098,
+      "epoch": 1.406060606060606,
+      "grad_norm": 0.011055397801101208,
+      "learning_rate": 5.627425614489004e-05,
+      "loss": 0.0023,
       "step": 435
     },
     {
-      "epoch": 1.1171950048030739,
-      "grad_norm": 0.021479401737451553,
-      "learning_rate": 9.54188615072984e-05,
-      "loss": 0.0159,
+      "epoch": 1.4092929292929293,
+      "grad_norm": 0.005197928287088871,
+      "learning_rate": 5.640362225097025e-05,
+      "loss": 0.0015,
       "step": 436
     },
     {
-      "epoch": 1.1197566442523215,
-      "grad_norm": 0.015459496527910233,
-      "learning_rate": 9.538136248470827e-05,
-      "loss": 0.0062,
+      "epoch": 1.4125252525252525,
+      "grad_norm": 0.00932388473302126,
+      "learning_rate": 5.653298835705045e-05,
+      "loss": 0.0029,
       "step": 437
     },
     {
-      "epoch": 1.122318283701569,
-      "grad_norm": 0.02548816241323948,
-      "learning_rate": 9.534371804252728e-05,
-      "loss": 0.0133,
+      "epoch": 1.4157575757575758,
+      "grad_norm": 0.009899857454001904,
+      "learning_rate": 5.666235446313066e-05,
+      "loss": 0.0028,
       "step": 438
     },
     {
-      "epoch": 1.1248799231508164,
-      "grad_norm": 0.018775109201669693,
-      "learning_rate": 9.530592830138292e-05,
-      "loss": 0.0088,
+      "epoch": 1.418989898989899,
+      "grad_norm": 0.013995957560837269,
+      "learning_rate": 5.679172056921087e-05,
+      "loss": 0.0041,
       "step": 439
     },
     {
-      "epoch": 1.127441562600064,
-      "grad_norm": 0.031876884400844574,
-      "learning_rate": 9.526799338236827e-05,
-      "loss": 0.0124,
+      "epoch": 1.4222222222222223,
+      "grad_norm": 0.014002947136759758,
+      "learning_rate": 5.692108667529108e-05,
+      "loss": 0.005,
       "step": 440
     },
     {
-      "epoch": 1.1300032020493116,
-      "grad_norm": 0.03683663532137871,
-      "learning_rate": 9.522991340704162e-05,
-      "loss": 0.0174,
+      "epoch": 1.4254545454545455,
+      "grad_norm": 0.008107703179121017,
+      "learning_rate": 5.7050452781371286e-05,
+      "loss": 0.0029,
       "step": 441
     },
     {
-      "epoch": 1.1325648414985592,
-      "grad_norm": 0.02702777087688446,
-      "learning_rate": 9.519168849742604e-05,
-      "loss": 0.0102,
+      "epoch": 1.4286868686868686,
+      "grad_norm": 0.011317477561533451,
+      "learning_rate": 5.717981888745149e-05,
+      "loss": 0.0043,
       "step": 442
     },
     {
-      "epoch": 1.1351264809478065,
-      "grad_norm": 0.028959324583411217,
-      "learning_rate": 9.515331877600906e-05,
-      "loss": 0.0106,
+      "epoch": 1.4319191919191918,
+      "grad_norm": 0.011309563182294369,
+      "learning_rate": 5.730918499353169e-05,
+      "loss": 0.0025,
       "step": 443
     },
     {
-      "epoch": 1.1376881203970541,
-      "grad_norm": 0.029596347361803055,
-      "learning_rate": 9.511480436574225e-05,
-      "loss": 0.0096,
+      "epoch": 1.435151515151515,
+      "grad_norm": 0.009117142297327518,
+      "learning_rate": 5.74385510996119e-05,
+      "loss": 0.0028,
       "step": 444
     },
     {
-      "epoch": 1.1402497598463017,
-      "grad_norm": 0.025607751682400703,
-      "learning_rate": 9.507614539004082e-05,
-      "loss": 0.0131,
+      "epoch": 1.4383838383838383,
+      "grad_norm": 0.011336748488247395,
+      "learning_rate": 5.7567917205692113e-05,
+      "loss": 0.0052,
       "step": 445
     },
     {
-      "epoch": 1.142811399295549,
-      "grad_norm": 0.023415246978402138,
-      "learning_rate": 9.503734197278316e-05,
-      "loss": 0.0107,
+      "epoch": 1.4416161616161616,
+      "grad_norm": 0.018933039158582687,
+      "learning_rate": 5.7697283311772324e-05,
+      "loss": 0.006,
       "step": 446
     },
     {
-      "epoch": 1.1453730387447967,
-      "grad_norm": 0.029267001897096634,
-      "learning_rate": 9.49983942383106e-05,
-      "loss": 0.0116,
+      "epoch": 1.4448484848484848,
+      "grad_norm": 0.007407647557556629,
+      "learning_rate": 5.782664941785253e-05,
+      "loss": 0.0021,
       "step": 447
     },
     {
-      "epoch": 1.1479346781940443,
-      "grad_norm": 0.021906791254878044,
-      "learning_rate": 9.495930231142689e-05,
-      "loss": 0.0113,
+      "epoch": 1.448080808080808,
+      "grad_norm": 0.011548957787454128,
+      "learning_rate": 5.795601552393273e-05,
+      "loss": 0.0028,
       "step": 448
     },
     {
-      "epoch": 1.1504963176432916,
-      "grad_norm": 0.0230691060423851,
-      "learning_rate": 9.492006631739777e-05,
-      "loss": 0.0123,
+      "epoch": 1.4513131313131313,
+      "grad_norm": 0.006151426583528519,
+      "learning_rate": 5.8085381630012934e-05,
+      "loss": 0.0017,
       "step": 449
     },
     {
-      "epoch": 1.1530579570925392,
-      "grad_norm": 0.023082030937075615,
-      "learning_rate": 9.48806863819507e-05,
-      "loss": 0.01,
+      "epoch": 1.4545454545454546,
+      "grad_norm": 0.00826767273247242,
+      "learning_rate": 5.8214747736093145e-05,
+      "loss": 0.0026,
       "step": 450
     },
     {
-      "epoch": 1.1530579570925392,
-      "eval_loss": 0.009998048655688763,
-      "eval_runtime": 15.7731,
-      "eval_samples_per_second": 6.34,
-      "eval_steps_per_second": 1.585,
+      "epoch": 1.4545454545454546,
+      "eval_loss": 0.0030124636832624674,
+      "eval_runtime": 15.5538,
+      "eval_samples_per_second": 6.429,
+      "eval_steps_per_second": 1.607,
       "step": 450
     },
     {
-      "epoch": 1.1556195965417868,
-      "grad_norm": 0.026397304609417915,
-      "learning_rate": 9.484116263127436e-05,
-      "loss": 0.0123,
+      "epoch": 1.4577777777777778,
+      "grad_norm": 0.010314207524061203,
+      "learning_rate": 5.8344113842173355e-05,
+      "loss": 0.0031,
       "step": 451
     },
     {
-      "epoch": 1.1581812359910342,
-      "grad_norm": 0.02318529225885868,
-      "learning_rate": 9.480149519201825e-05,
-      "loss": 0.0107,
+      "epoch": 1.461010101010101,
+      "grad_norm": 0.010531079955399036,
+      "learning_rate": 5.8473479948253565e-05,
+      "loss": 0.0041,
       "step": 452
     },
     {
-      "epoch": 1.1607428754402818,
-      "grad_norm": 0.021503478288650513,
-      "learning_rate": 9.476168419129235e-05,
-      "loss": 0.0124,
+      "epoch": 1.4642424242424243,
+      "grad_norm": 0.00876904372125864,
+      "learning_rate": 5.860284605433377e-05,
+      "loss": 0.003,
       "step": 453
     },
     {
-      "epoch": 1.1633045148895294,
-      "grad_norm": 0.03301948681473732,
-      "learning_rate": 9.47217297566666e-05,
-      "loss": 0.0136,
+      "epoch": 1.4674747474747476,
+      "grad_norm": 0.008487108163535595,
+      "learning_rate": 5.873221216041398e-05,
+      "loss": 0.0027,
       "step": 454
     },
     {
-      "epoch": 1.1658661543387767,
-      "grad_norm": 0.021176926791667938,
-      "learning_rate": 9.468163201617062e-05,
-      "loss": 0.0084,
+      "epoch": 1.4707070707070706,
+      "grad_norm": 0.008451130241155624,
+      "learning_rate": 5.8861578266494176e-05,
+      "loss": 0.0024,
       "step": 455
     },
     {
-      "epoch": 1.1684277937880243,
-      "grad_norm": 0.021742813289165497,
-      "learning_rate": 9.464139109829321e-05,
-      "loss": 0.0112,
+      "epoch": 1.4739393939393939,
+      "grad_norm": 0.009484423324465752,
+      "learning_rate": 5.8990944372574386e-05,
+      "loss": 0.0024,
       "step": 456
     },
     {
-      "epoch": 1.170989433237272,
-      "grad_norm": 0.01652652770280838,
-      "learning_rate": 9.460100713198196e-05,
-      "loss": 0.0063,
+      "epoch": 1.4771717171717171,
+      "grad_norm": 0.008541501127183437,
+      "learning_rate": 5.9120310478654596e-05,
+      "loss": 0.0028,
       "step": 457
     },
     {
-      "epoch": 1.1735510726865193,
-      "grad_norm": 0.02279038541018963,
-      "learning_rate": 9.456048024664288e-05,
-      "loss": 0.0097,
+      "epoch": 1.4804040404040404,
+      "grad_norm": 0.009456522762775421,
+      "learning_rate": 5.9249676584734806e-05,
+      "loss": 0.0033,
       "step": 458
     },
     {
-      "epoch": 1.1761127121357668,
-      "grad_norm": 0.019902581349015236,
-      "learning_rate": 9.451981057213992e-05,
-      "loss": 0.0091,
+      "epoch": 1.4836363636363636,
+      "grad_norm": 0.007811453193426132,
+      "learning_rate": 5.937904269081501e-05,
+      "loss": 0.0021,
       "step": 459
     },
     {
-      "epoch": 1.1786743515850144,
-      "grad_norm": 0.03399882838129997,
-      "learning_rate": 9.447899823879456e-05,
-      "loss": 0.018,
+      "epoch": 1.486868686868687,
+      "grad_norm": 0.011969913728535175,
+      "learning_rate": 5.950840879689522e-05,
+      "loss": 0.0044,
       "step": 460
     },
     {
-      "epoch": 1.181235991034262,
-      "grad_norm": 0.022234100848436356,
-      "learning_rate": 9.443804337738544e-05,
-      "loss": 0.0068,
+      "epoch": 1.4901010101010101,
+      "grad_norm": 0.009493980556726456,
+      "learning_rate": 5.963777490297542e-05,
+      "loss": 0.0034,
       "step": 461
     },
     {
-      "epoch": 1.1837976304835094,
-      "grad_norm": 0.019693007692694664,
-      "learning_rate": 9.439694611914796e-05,
-      "loss": 0.0054,
+      "epoch": 1.4933333333333334,
+      "grad_norm": 0.012684321962296963,
+      "learning_rate": 5.976714100905563e-05,
+      "loss": 0.0037,
       "step": 462
     },
     {
-      "epoch": 1.186359269932757,
-      "grad_norm": 0.02781878225505352,
-      "learning_rate": 9.43557065957737e-05,
-      "loss": 0.0166,
+      "epoch": 1.4965656565656564,
+      "grad_norm": 0.013734826818108559,
+      "learning_rate": 5.989650711513584e-05,
+      "loss": 0.0053,
       "step": 463
     },
     {
-      "epoch": 1.1889209093820046,
-      "grad_norm": 0.018343890085816383,
-      "learning_rate": 9.431432493941022e-05,
-      "loss": 0.0085,
+      "epoch": 1.4997979797979797,
+      "grad_norm": 0.01171690970659256,
+      "learning_rate": 6.002587322121605e-05,
+      "loss": 0.003,
       "step": 464
     },
     {
-      "epoch": 1.191482548831252,
-      "grad_norm": 0.03791218250989914,
-      "learning_rate": 9.42728012826605e-05,
-      "loss": 0.014,
+      "epoch": 1.503030303030303,
+      "grad_norm": 0.009820333682000637,
+      "learning_rate": 6.015523932729625e-05,
+      "loss": 0.0023,
       "step": 465
     },
     {
-      "epoch": 1.1940441882804995,
-      "grad_norm": 0.025248322635889053,
-      "learning_rate": 9.423113575858252e-05,
-      "loss": 0.01,
+      "epoch": 1.5062626262626262,
+      "grad_norm": 0.0082180704921484,
+      "learning_rate": 6.028460543337646e-05,
+      "loss": 0.0022,
       "step": 466
     },
     {
-      "epoch": 1.196605827729747,
-      "grad_norm": 0.020284641534090042,
-      "learning_rate": 9.418932850068889e-05,
-      "loss": 0.0082,
+      "epoch": 1.5094949494949494,
+      "grad_norm": 0.0106166647747159,
+      "learning_rate": 6.041397153945667e-05,
+      "loss": 0.0031,
       "step": 467
     },
     {
-      "epoch": 1.1991674671789945,
-      "grad_norm": 0.02318977750837803,
-      "learning_rate": 9.414737964294636e-05,
-      "loss": 0.0104,
+      "epoch": 1.5127272727272727,
+      "grad_norm": 0.009280703961849213,
+      "learning_rate": 6.054333764553687e-05,
+      "loss": 0.0028,
       "step": 468
     },
     {
-      "epoch": 1.201729106628242,
-      "grad_norm": 0.020158514380455017,
-      "learning_rate": 9.410528931977542e-05,
-      "loss": 0.0087,
+      "epoch": 1.515959595959596,
+      "grad_norm": 0.005328455474227667,
+      "learning_rate": 6.067270375161708e-05,
+      "loss": 0.0017,
       "step": 469
     },
     {
-      "epoch": 1.2042907460774896,
-      "grad_norm": 0.028263043612241745,
-      "learning_rate": 9.406305766604995e-05,
-      "loss": 0.0124,
+      "epoch": 1.5191919191919192,
+      "grad_norm": 0.014295658096671104,
+      "learning_rate": 6.080206985769729e-05,
+      "loss": 0.0069,
       "step": 470
     },
     {
-      "epoch": 1.206852385526737,
-      "grad_norm": 0.03137621656060219,
-      "learning_rate": 9.402068481709658e-05,
-      "loss": 0.0206,
+      "epoch": 1.5224242424242425,
+      "grad_norm": 0.00929258018732071,
+      "learning_rate": 6.093143596377749e-05,
+      "loss": 0.0027,
       "step": 471
     },
     {
-      "epoch": 1.2094140249759846,
-      "grad_norm": 0.02518923208117485,
-      "learning_rate": 9.397817090869448e-05,
-      "loss": 0.0162,
+      "epoch": 1.5256565656565657,
+      "grad_norm": 0.009517964906990528,
+      "learning_rate": 6.10608020698577e-05,
+      "loss": 0.0015,
       "step": 472
     },
     {
-      "epoch": 1.2119756644252322,
-      "grad_norm": 0.021693777292966843,
-      "learning_rate": 9.393551607707478e-05,
-      "loss": 0.0111,
+      "epoch": 1.528888888888889,
+      "grad_norm": 0.010964248329401016,
+      "learning_rate": 6.119016817593791e-05,
+      "loss": 0.0047,
       "step": 473
     },
     {
-      "epoch": 1.2145373038744798,
-      "grad_norm": 0.023595202714204788,
-      "learning_rate": 9.389272045892024e-05,
-      "loss": 0.0086,
+      "epoch": 1.5321212121212122,
+      "grad_norm": 0.012098269537091255,
+      "learning_rate": 6.131953428201811e-05,
+      "loss": 0.0057,
       "step": 474
     },
     {
-      "epoch": 1.2170989433237271,
-      "grad_norm": 0.025104286149144173,
-      "learning_rate": 9.384978419136468e-05,
-      "loss": 0.0158,
+      "epoch": 1.5353535353535355,
+      "grad_norm": 0.011843489482998848,
+      "learning_rate": 6.144890038809832e-05,
+      "loss": 0.0035,
       "step": 475
     },
     {
-      "epoch": 1.2170989433237271,
-      "eval_loss": 0.008896440267562866,
-      "eval_runtime": 15.7543,
-      "eval_samples_per_second": 6.347,
-      "eval_steps_per_second": 1.587,
+      "epoch": 1.5353535353535355,
+      "eval_loss": 0.0028126242104917765,
+      "eval_runtime": 15.5424,
+      "eval_samples_per_second": 6.434,
+      "eval_steps_per_second": 1.609,
       "step": 475
     },
     {
-      "epoch": 1.2196605827729747,
-      "grad_norm": 0.016638245433568954,
-      "learning_rate": 9.380670741199266e-05,
-      "loss": 0.0079,
+      "epoch": 1.5385858585858587,
+      "grad_norm": 0.008462205529212952,
+      "learning_rate": 6.157826649417853e-05,
+      "loss": 0.0036,
       "step": 476
     },
     {
-      "epoch": 1.2222222222222223,
-      "grad_norm": 0.015983251854777336,
-      "learning_rate": 9.3763490258839e-05,
-      "loss": 0.0067,
+      "epoch": 1.541818181818182,
+      "grad_norm": 0.00864575058221817,
+      "learning_rate": 6.170763260025874e-05,
+      "loss": 0.0033,
       "step": 477
     },
     {
-      "epoch": 1.2247838616714697,
-      "grad_norm": 0.020340844988822937,
-      "learning_rate": 9.37201328703883e-05,
-      "loss": 0.0102,
+      "epoch": 1.545050505050505,
+      "grad_norm": 0.01465240865945816,
+      "learning_rate": 6.183699870633895e-05,
+      "loss": 0.0052,
       "step": 478
     },
     {
-      "epoch": 1.2273455011207173,
-      "grad_norm": 0.019760936498641968,
-      "learning_rate": 9.367663538557457e-05,
-      "loss": 0.0099,
+      "epoch": 1.5482828282828283,
+      "grad_norm": 0.015150130726397038,
+      "learning_rate": 6.196636481241915e-05,
+      "loss": 0.0035,
       "step": 479
     },
     {
-      "epoch": 1.2299071405699649,
-      "grad_norm": 0.027867259457707405,
-      "learning_rate": 9.363299794378073e-05,
-      "loss": 0.0183,
+      "epoch": 1.5515151515151515,
+      "grad_norm": 0.010665367357432842,
+      "learning_rate": 6.209573091849934e-05,
+      "loss": 0.0046,
       "step": 480
     },
     {
-      "epoch": 1.2324687800192122,
-      "grad_norm": 0.025274328887462616,
-      "learning_rate": 9.358922068483812e-05,
-      "loss": 0.0144,
+      "epoch": 1.5547474747474748,
+      "grad_norm": 0.011034867726266384,
+      "learning_rate": 6.222509702457955e-05,
+      "loss": 0.0043,
       "step": 481
     },
     {
-      "epoch": 1.2350304194684598,
-      "grad_norm": 0.025409767404198647,
-      "learning_rate": 9.354530374902622e-05,
-      "loss": 0.0127,
+      "epoch": 1.557979797979798,
+      "grad_norm": 0.015952058136463165,
+      "learning_rate": 6.235446313065976e-05,
+      "loss": 0.0042,
       "step": 482
     },
     {
-      "epoch": 1.2375920589177074,
-      "grad_norm": 0.0174652561545372,
-      "learning_rate": 9.350124727707197e-05,
-      "loss": 0.0081,
+      "epoch": 1.561212121212121,
+      "grad_norm": 0.007438727654516697,
+      "learning_rate": 6.248382923673998e-05,
+      "loss": 0.0022,
       "step": 483
     },
     {
-      "epoch": 1.2401536983669548,
-      "grad_norm": 0.020473089069128036,
-      "learning_rate": 9.345705141014953e-05,
-      "loss": 0.0107,
+      "epoch": 1.5644444444444443,
+      "grad_norm": 0.009822466410696507,
+      "learning_rate": 6.261319534282019e-05,
+      "loss": 0.003,
       "step": 484
     },
     {
-      "epoch": 1.2427153378162024,
-      "grad_norm": 0.02481193281710148,
-      "learning_rate": 9.341271628987968e-05,
-      "loss": 0.0162,
+      "epoch": 1.5676767676767676,
+      "grad_norm": 0.011421862058341503,
+      "learning_rate": 6.27425614489004e-05,
+      "loss": 0.0044,
       "step": 485
     },
     {
-      "epoch": 1.24527697726545,
-      "grad_norm": 0.018237514421343803,
-      "learning_rate": 9.336824205832947e-05,
-      "loss": 0.0094,
+      "epoch": 1.5709090909090908,
+      "grad_norm": 0.010750948451459408,
+      "learning_rate": 6.28719275549806e-05,
+      "loss": 0.0041,
       "step": 486
     },
     {
-      "epoch": 1.2478386167146973,
-      "grad_norm": 0.016897965222597122,
-      "learning_rate": 9.332362885801167e-05,
-      "loss": 0.0064,
+      "epoch": 1.574141414141414,
+      "grad_norm": 0.013682820834219456,
+      "learning_rate": 6.30012936610608e-05,
+      "loss": 0.0045,
       "step": 487
     },
     {
-      "epoch": 1.250400256163945,
-      "grad_norm": 0.01858009584248066,
-      "learning_rate": 9.327887683188436e-05,
-      "loss": 0.0065,
+      "epoch": 1.5773737373737373,
+      "grad_norm": 0.025229288265109062,
+      "learning_rate": 6.313065976714101e-05,
+      "loss": 0.0093,
       "step": 488
     },
     {
-      "epoch": 1.2529618956131925,
-      "grad_norm": 0.02213798649609089,
-      "learning_rate": 9.323398612335055e-05,
-      "loss": 0.009,
+      "epoch": 1.5806060606060606,
+      "grad_norm": 0.00909426063299179,
+      "learning_rate": 6.326002587322122e-05,
+      "loss": 0.002,
       "step": 489
     },
     {
-      "epoch": 1.2555235350624399,
-      "grad_norm": 0.03265499696135521,
-      "learning_rate": 9.318895687625751e-05,
-      "loss": 0.0128,
+      "epoch": 1.5838383838383838,
+      "grad_norm": 0.011271797120571136,
+      "learning_rate": 6.338939197930143e-05,
+      "loss": 0.003,
       "step": 490
     },
     {
-      "epoch": 1.2580851745116874,
-      "grad_norm": 0.01880866289138794,
-      "learning_rate": 9.314378923489658e-05,
-      "loss": 0.0092,
+      "epoch": 1.587070707070707,
+      "grad_norm": 0.015798212960362434,
+      "learning_rate": 6.351875808538163e-05,
+      "loss": 0.0043,
       "step": 491
     },
     {
-      "epoch": 1.260646813960935,
-      "grad_norm": 0.03112734854221344,
-      "learning_rate": 9.309848334400246e-05,
-      "loss": 0.011,
+      "epoch": 1.5903030303030303,
+      "grad_norm": 0.00970592349767685,
+      "learning_rate": 6.364812419146184e-05,
+      "loss": 0.0032,
       "step": 492
     },
     {
-      "epoch": 1.2632084534101824,
-      "grad_norm": 0.028772294521331787,
-      "learning_rate": 9.305303934875294e-05,
-      "loss": 0.0149,
+      "epoch": 1.5935353535353536,
+      "grad_norm": 0.011688323691487312,
+      "learning_rate": 6.377749029754204e-05,
+      "loss": 0.0033,
       "step": 493
     },
     {
-      "epoch": 1.26577009285943,
-      "grad_norm": 0.015158562920987606,
-      "learning_rate": 9.300745739476829e-05,
-      "loss": 0.0071,
+      "epoch": 1.5967676767676768,
+      "grad_norm": 0.011503425426781178,
+      "learning_rate": 6.390685640362225e-05,
+      "loss": 0.0035,
       "step": 494
     },
     {
-      "epoch": 1.2683317323086776,
-      "grad_norm": 0.02867531031370163,
-      "learning_rate": 9.296173762811085e-05,
-      "loss": 0.0203,
+      "epoch": 1.6,
+      "grad_norm": 0.005506394896656275,
+      "learning_rate": 6.403622250970246e-05,
+      "loss": 0.0017,
       "step": 495
     },
     {
-      "epoch": 1.270893371757925,
-      "grad_norm": 0.03381772339344025,
-      "learning_rate": 9.291588019528462e-05,
-      "loss": 0.021,
+      "epoch": 1.6032323232323233,
+      "grad_norm": 0.007110868580639362,
+      "learning_rate": 6.416558861578267e-05,
+      "loss": 0.0027,
       "step": 496
     },
     {
-      "epoch": 1.2734550112071725,
-      "grad_norm": 0.019410375505685806,
-      "learning_rate": 9.28698852432347e-05,
-      "loss": 0.0088,
+      "epoch": 1.6064646464646466,
+      "grad_norm": 0.021086519584059715,
+      "learning_rate": 6.429495472186288e-05,
+      "loss": 0.0055,
       "step": 497
     },
     {
-      "epoch": 1.2760166506564201,
-      "grad_norm": 0.025806061923503876,
-      "learning_rate": 9.282375291934687e-05,
-      "loss": 0.0097,
+      "epoch": 1.6096969696969698,
+      "grad_norm": 0.010809206403791904,
+      "learning_rate": 6.442432082794309e-05,
+      "loss": 0.0031,
       "step": 498
     },
     {
-      "epoch": 1.2785782901056677,
-      "grad_norm": 0.02822774648666382,
-      "learning_rate": 9.277748337144709e-05,
-      "loss": 0.016,
+      "epoch": 1.6129292929292929,
+      "grad_norm": 0.009927364997565746,
+      "learning_rate": 6.45536869340233e-05,
+      "loss": 0.0031,
       "step": 499
     },
     {
-      "epoch": 1.281139929554915,
-      "grad_norm": 0.025421006605029106,
-      "learning_rate": 9.273107674780102e-05,
-      "loss": 0.0176,
+      "epoch": 1.6161616161616161,
+      "grad_norm": 0.007981827482581139,
+      "learning_rate": 6.46830530401035e-05,
+      "loss": 0.0029,
       "step": 500
     },
     {
-      "epoch": 1.281139929554915,
-      "eval_loss": 0.008733320981264114,
-      "eval_runtime": 15.7696,
-      "eval_samples_per_second": 6.341,
-      "eval_steps_per_second": 1.585,
+      "epoch": 1.6161616161616161,
+      "eval_loss": 0.0027610582765191793,
+      "eval_runtime": 15.4879,
+      "eval_samples_per_second": 6.457,
+      "eval_steps_per_second": 1.614,
       "step": 500
     },
     {
-      "epoch": 1.2837015690041627,
-      "grad_norm": 0.022834056988358498,
-      "learning_rate": 9.268453319711363e-05,
-      "loss": 0.0102,
+      "epoch": 1.6193939393939394,
+      "grad_norm": 0.009775879792869091,
+      "learning_rate": 6.48124191461837e-05,
+      "loss": 0.0041,
       "step": 501
     },
     {
-      "epoch": 1.2862632084534102,
-      "grad_norm": 0.02453305944800377,
-      "learning_rate": 9.263785286852857e-05,
-      "loss": 0.0128,
+      "epoch": 1.6226262626262626,
+      "grad_norm": 0.01406659372150898,
+      "learning_rate": 6.494178525226392e-05,
+      "loss": 0.0044,
       "step": 502
     },
     {
-      "epoch": 1.2888248479026578,
-      "grad_norm": 0.020842405036091805,
-      "learning_rate": 9.259103591162785e-05,
-      "loss": 0.0061,
+      "epoch": 1.625858585858586,
+      "grad_norm": 0.010167881846427917,
+      "learning_rate": 6.507115135834411e-05,
+      "loss": 0.0046,
       "step": 503
     },
     {
-      "epoch": 1.2913864873519052,
-      "grad_norm": 0.033657580614089966,
-      "learning_rate": 9.254408247643126e-05,
-      "loss": 0.0231,
+      "epoch": 1.6290909090909091,
+      "grad_norm": 0.009142483584582806,
+      "learning_rate": 6.520051746442432e-05,
+      "loss": 0.0027,
       "step": 504
     },
     {
-      "epoch": 1.2939481268011528,
-      "grad_norm": 0.020036958158016205,
-      "learning_rate": 9.249699271339593e-05,
-      "loss": 0.0092,
+      "epoch": 1.6323232323232322,
+      "grad_norm": 0.013479731045663357,
+      "learning_rate": 6.532988357050453e-05,
+      "loss": 0.0048,
       "step": 505
     },
     {
-      "epoch": 1.2965097662504004,
-      "grad_norm": 0.02098897099494934,
-      "learning_rate": 9.244976677341583e-05,
-      "loss": 0.0092,
+      "epoch": 1.6355555555555554,
+      "grad_norm": 0.013354233466088772,
+      "learning_rate": 6.545924967658473e-05,
+      "loss": 0.0062,
       "step": 506
     },
     {
-      "epoch": 1.2990714056996477,
-      "grad_norm": 0.017251061275601387,
-      "learning_rate": 9.24024048078213e-05,
-      "loss": 0.0093,
+      "epoch": 1.6387878787878787,
+      "grad_norm": 0.00613620039075613,
+      "learning_rate": 6.558861578266494e-05,
+      "loss": 0.0025,
       "step": 507
     },
     {
-      "epoch": 1.3016330451488953,
-      "grad_norm": 0.02004789188504219,
-      "learning_rate": 9.235490696837855e-05,
-      "loss": 0.0117,
+      "epoch": 1.642020202020202,
+      "grad_norm": 0.011386699974536896,
+      "learning_rate": 6.571798188874515e-05,
+      "loss": 0.0058,
       "step": 508
     },
     {
-      "epoch": 1.304194684598143,
-      "grad_norm": 0.020399397239089012,
-      "learning_rate": 9.23072734072892e-05,
-      "loss": 0.0087,
+      "epoch": 1.6452525252525252,
+      "grad_norm": 0.012126428075134754,
+      "learning_rate": 6.584734799482536e-05,
+      "loss": 0.0058,
       "step": 509
     },
     {
-      "epoch": 1.3067563240473903,
-      "grad_norm": 0.025535760447382927,
-      "learning_rate": 9.225950427718975e-05,
-      "loss": 0.0104,
+      "epoch": 1.6484848484848484,
+      "grad_norm": 0.007690794765949249,
+      "learning_rate": 6.597671410090557e-05,
+      "loss": 0.0027,
       "step": 510
     },
     {
-      "epoch": 1.3093179634966379,
-      "grad_norm": 0.03579006716609001,
-      "learning_rate": 9.221159973115112e-05,
-      "loss": 0.0191,
+      "epoch": 1.6517171717171717,
+      "grad_norm": 0.01104055903851986,
+      "learning_rate": 6.610608020698578e-05,
+      "loss": 0.0053,
       "step": 511
     },
     {
-      "epoch": 1.3118796029458855,
-      "grad_norm": 0.029484592378139496,
-      "learning_rate": 9.21635599226782e-05,
-      "loss": 0.0132,
+      "epoch": 1.654949494949495,
+      "grad_norm": 0.007917758077383041,
+      "learning_rate": 6.623544631306599e-05,
+      "loss": 0.0036,
       "step": 512
     },
     {
-      "epoch": 1.3144412423951328,
-      "grad_norm": 0.025395115837454796,
-      "learning_rate": 9.211538500570925e-05,
-      "loss": 0.0093,
+      "epoch": 1.6581818181818182,
+      "grad_norm": 0.011790517717599869,
+      "learning_rate": 6.636481241914619e-05,
+      "loss": 0.005,
       "step": 513
     },
     {
-      "epoch": 1.3170028818443804,
-      "grad_norm": 0.03858427330851555,
-      "learning_rate": 9.206707513461551e-05,
-      "loss": 0.0159,
+      "epoch": 1.6614141414141415,
+      "grad_norm": 0.0130055733025074,
+      "learning_rate": 6.649417852522638e-05,
+      "loss": 0.0074,
       "step": 514
     },
     {
-      "epoch": 1.319564521293628,
-      "grad_norm": 0.018974190577864647,
-      "learning_rate": 9.201863046420065e-05,
-      "loss": 0.0076,
+      "epoch": 1.6646464646464647,
+      "grad_norm": 0.006904282607138157,
+      "learning_rate": 6.66235446313066e-05,
+      "loss": 0.0026,
       "step": 515
     },
     {
-      "epoch": 1.3221261607428754,
-      "grad_norm": 0.030375462025403976,
-      "learning_rate": 9.19700511497003e-05,
-      "loss": 0.0145,
+      "epoch": 1.667878787878788,
+      "grad_norm": 0.010409231297671795,
+      "learning_rate": 6.67529107373868e-05,
+      "loss": 0.0036,
       "step": 516
     },
     {
-      "epoch": 1.324687800192123,
-      "grad_norm": 0.02294369973242283,
-      "learning_rate": 9.192133734678155e-05,
-      "loss": 0.0104,
+      "epoch": 1.6711111111111112,
+      "grad_norm": 0.011454829946160316,
+      "learning_rate": 6.688227684346702e-05,
+      "loss": 0.004,
       "step": 517
     },
     {
-      "epoch": 1.3272494396413705,
-      "grad_norm": 0.033602748066186905,
-      "learning_rate": 9.187248921154239e-05,
-      "loss": 0.0098,
+      "epoch": 1.6743434343434345,
+      "grad_norm": 0.008268332108855247,
+      "learning_rate": 6.701164294954723e-05,
+      "loss": 0.0023,
       "step": 518
     },
     {
-      "epoch": 1.329811079090618,
-      "grad_norm": 0.019165707752108574,
-      "learning_rate": 9.182350690051133e-05,
-      "loss": 0.0111,
+      "epoch": 1.6775757575757577,
+      "grad_norm": 0.015307283960282803,
+      "learning_rate": 6.714100905562742e-05,
+      "loss": 0.0046,
       "step": 519
     },
     {
-      "epoch": 1.3323727185398655,
-      "grad_norm": 0.021407922729849815,
-      "learning_rate": 9.177439057064683e-05,
-      "loss": 0.0091,
+      "epoch": 1.680808080808081,
+      "grad_norm": 0.010210135951638222,
+      "learning_rate": 6.727037516170763e-05,
+      "loss": 0.003,
       "step": 520
     },
     {
-      "epoch": 1.334934357989113,
-      "grad_norm": 0.03057362511754036,
-      "learning_rate": 9.172514037933673e-05,
-      "loss": 0.0192,
+      "epoch": 1.684040404040404,
+      "grad_norm": 0.011168411001563072,
+      "learning_rate": 6.739974126778784e-05,
+      "loss": 0.003,
       "step": 521
     },
     {
-      "epoch": 1.3374959974383605,
-      "grad_norm": 0.016315093263983727,
-      "learning_rate": 9.167575648439788e-05,
-      "loss": 0.0072,
+      "epoch": 1.6872727272727273,
+      "grad_norm": 0.008964510634541512,
+      "learning_rate": 6.752910737386805e-05,
+      "loss": 0.0024,
       "step": 522
     },
     {
-      "epoch": 1.340057636887608,
-      "grad_norm": 0.025471271947026253,
-      "learning_rate": 9.162623904407553e-05,
-      "loss": 0.014,
+      "epoch": 1.6905050505050505,
+      "grad_norm": 0.01176072284579277,
+      "learning_rate": 6.765847347994826e-05,
+      "loss": 0.0024,
       "step": 523
     },
     {
-      "epoch": 1.3426192763368556,
-      "grad_norm": 0.022852519527077675,
-      "learning_rate": 9.15765882170429e-05,
-      "loss": 0.0102,
+      "epoch": 1.6937373737373738,
+      "grad_norm": 0.01260540820658207,
+      "learning_rate": 6.778783958602847e-05,
+      "loss": 0.0061,
       "step": 524
     },
     {
-      "epoch": 1.345180915786103,
-      "grad_norm": 0.015502731315791607,
-      "learning_rate": 9.152680416240059e-05,
-      "loss": 0.0073,
+      "epoch": 1.696969696969697,
+      "grad_norm": 0.010063033550977707,
+      "learning_rate": 6.791720569210867e-05,
+      "loss": 0.0046,
       "step": 525
     },
     {
-      "epoch": 1.345180915786103,
-      "eval_loss": 0.008211835287511349,
-      "eval_runtime": 15.7775,
-      "eval_samples_per_second": 6.338,
-      "eval_steps_per_second": 1.585,
+      "epoch": 1.696969696969697,
+      "eval_loss": 0.002695531351491809,
+      "eval_runtime": 15.4702,
+      "eval_samples_per_second": 6.464,
+      "eval_steps_per_second": 1.616,
       "step": 525
     },
     {
-      "epoch": 1.3477425552353506,
-      "grad_norm": 0.01804802566766739,
-      "learning_rate": 9.147688703967614e-05,
-      "loss": 0.0056,
+      "epoch": 1.70020202020202,
+      "grad_norm": 0.007502489257603884,
+      "learning_rate": 6.804657179818887e-05,
+      "loss": 0.0025,
       "step": 526
     },
     {
-      "epoch": 1.3503041946845982,
-      "grad_norm": 0.026565713807940483,
-      "learning_rate": 9.142683700882351e-05,
-      "loss": 0.0158,
+      "epoch": 1.7034343434343433,
+      "grad_norm": 0.006980999372899532,
+      "learning_rate": 6.817593790426908e-05,
+      "loss": 0.0027,
       "step": 527
     },
     {
-      "epoch": 1.3528658341338455,
-      "grad_norm": 0.023307789117097855,
-      "learning_rate": 9.13766542302225e-05,
-      "loss": 0.0132,
+      "epoch": 1.7066666666666666,
+      "grad_norm": 0.013577309437096119,
+      "learning_rate": 6.830530401034929e-05,
+      "loss": 0.0041,
       "step": 528
     },
     {
-      "epoch": 1.3554274735830931,
-      "grad_norm": 0.03050735965371132,
-      "learning_rate": 9.132633886467831e-05,
-      "loss": 0.0177,
+      "epoch": 1.7098989898989898,
+      "grad_norm": 0.008221564814448357,
+      "learning_rate": 6.84346701164295e-05,
+      "loss": 0.0029,
       "step": 529
     },
     {
-      "epoch": 1.3579891130323407,
-      "grad_norm": 0.014245600439608097,
-      "learning_rate": 9.1275891073421e-05,
-      "loss": 0.0072,
+      "epoch": 1.713131313131313,
+      "grad_norm": 0.0074736205860972404,
+      "learning_rate": 6.856403622250971e-05,
+      "loss": 0.0024,
       "step": 530
     },
     {
-      "epoch": 1.3605507524815883,
-      "grad_norm": 0.018063992261886597,
-      "learning_rate": 9.122531101810501e-05,
-      "loss": 0.0074,
+      "epoch": 1.7163636363636363,
+      "grad_norm": 0.007041532546281815,
+      "learning_rate": 6.869340232858992e-05,
+      "loss": 0.0021,
       "step": 531
     },
     {
-      "epoch": 1.3631123919308357,
-      "grad_norm": 0.03403697907924652,
-      "learning_rate": 9.117459886080854e-05,
-      "loss": 0.0233,
+      "epoch": 1.7195959595959596,
+      "grad_norm": 0.0065683722496032715,
+      "learning_rate": 6.882276843467012e-05,
+      "loss": 0.0024,
       "step": 532
     },
     {
-      "epoch": 1.3656740313800833,
-      "grad_norm": 0.02282620035111904,
-      "learning_rate": 9.112375476403312e-05,
-      "loss": 0.0109,
+      "epoch": 1.7228282828282828,
+      "grad_norm": 0.007500760722905397,
+      "learning_rate": 6.895213454075033e-05,
+      "loss": 0.0024,
       "step": 533
     },
     {
-      "epoch": 1.3682356708293308,
-      "grad_norm": 0.017634566873311996,
-      "learning_rate": 9.107277889070309e-05,
-      "loss": 0.0093,
+      "epoch": 1.726060606060606,
+      "grad_norm": 0.008781072683632374,
+      "learning_rate": 6.908150064683054e-05,
+      "loss": 0.0023,
       "step": 534
     },
     {
-      "epoch": 1.3707973102785782,
-      "grad_norm": 0.01788763515651226,
-      "learning_rate": 9.102167140416503e-05,
-      "loss": 0.0098,
+      "epoch": 1.7292929292929293,
+      "grad_norm": 0.013301175087690353,
+      "learning_rate": 6.921086675291075e-05,
+      "loss": 0.0046,
       "step": 535
     },
     {
-      "epoch": 1.3733589497278258,
-      "grad_norm": 0.02422240935266018,
-      "learning_rate": 9.097043246818727e-05,
-      "loss": 0.0112,
+      "epoch": 1.7325252525252526,
+      "grad_norm": 0.009006236679852009,
+      "learning_rate": 6.934023285899096e-05,
+      "loss": 0.004,
       "step": 536
     },
     {
-      "epoch": 1.3759205891770734,
-      "grad_norm": 0.02159043774008751,
-      "learning_rate": 9.091906224695935e-05,
-      "loss": 0.0137,
+      "epoch": 1.7357575757575758,
+      "grad_norm": 0.009585372172296047,
+      "learning_rate": 6.946959896507115e-05,
+      "loss": 0.0033,
       "step": 537
     },
     {
-      "epoch": 1.378482228626321,
-      "grad_norm": 0.031434234231710434,
-      "learning_rate": 9.08675609050915e-05,
-      "loss": 0.0159,
+      "epoch": 1.738989898989899,
+      "grad_norm": 0.018620749935507774,
+      "learning_rate": 6.959896507115135e-05,
+      "loss": 0.0093,
       "step": 538
     },
     {
-      "epoch": 1.3810438680755683,
-      "grad_norm": 0.029942264780402184,
-      "learning_rate": 9.081592860761412e-05,
-      "loss": 0.0156,
+      "epoch": 1.7422222222222223,
+      "grad_norm": 0.00692649045959115,
+      "learning_rate": 6.972833117723156e-05,
+      "loss": 0.0026,
       "step": 539
     },
     {
-      "epoch": 1.383605507524816,
-      "grad_norm": 0.02360542118549347,
-      "learning_rate": 9.076416551997721e-05,
-      "loss": 0.0064,
+      "epoch": 1.7454545454545456,
+      "grad_norm": 0.009817032143473625,
+      "learning_rate": 6.985769728331177e-05,
+      "loss": 0.0032,
       "step": 540
     },
     {
-      "epoch": 1.3861671469740635,
-      "grad_norm": 0.025887703523039818,
-      "learning_rate": 9.071227180804994e-05,
-      "loss": 0.0107,
+      "epoch": 1.7486868686868688,
+      "grad_norm": 0.007341981865465641,
+      "learning_rate": 6.998706338939198e-05,
+      "loss": 0.0027,
       "step": 541
     },
     {
-      "epoch": 1.3887287864233109,
-      "grad_norm": 0.018033858388662338,
-      "learning_rate": 9.066024763811997e-05,
-      "loss": 0.0092,
+      "epoch": 1.7519191919191919,
+      "grad_norm": 0.005250657442957163,
+      "learning_rate": 7.011642949547219e-05,
+      "loss": 0.0016,
       "step": 542
     },
     {
-      "epoch": 1.3912904258725585,
-      "grad_norm": 0.02424168959259987,
-      "learning_rate": 9.060809317689308e-05,
-      "loss": 0.0122,
+      "epoch": 1.7551515151515151,
+      "grad_norm": 0.012378359213471413,
+      "learning_rate": 7.02457956015524e-05,
+      "loss": 0.0065,
       "step": 543
     },
     {
-      "epoch": 1.393852065321806,
-      "grad_norm": 0.025377338752150536,
-      "learning_rate": 9.055580859149246e-05,
-      "loss": 0.0093,
+      "epoch": 1.7583838383838384,
+      "grad_norm": 0.006813595071434975,
+      "learning_rate": 7.037516170763261e-05,
+      "loss": 0.0038,
       "step": 544
     },
     {
-      "epoch": 1.3964137047710534,
-      "grad_norm": 0.02267155982553959,
-      "learning_rate": 9.050339404945833e-05,
-      "loss": 0.0177,
+      "epoch": 1.7616161616161616,
+      "grad_norm": 0.007977760396897793,
+      "learning_rate": 7.050452781371281e-05,
+      "loss": 0.003,
       "step": 545
     },
     {
-      "epoch": 1.398975344220301,
-      "grad_norm": 0.023138578981161118,
-      "learning_rate": 9.045084971874738e-05,
-      "loss": 0.0088,
+      "epoch": 1.7648484848484849,
+      "grad_norm": 0.009840245358645916,
+      "learning_rate": 7.063389391979302e-05,
+      "loss": 0.0035,
       "step": 546
     },
     {
-      "epoch": 1.4015369836695486,
-      "grad_norm": 0.0254532340914011,
-      "learning_rate": 9.03981757677321e-05,
-      "loss": 0.012,
+      "epoch": 1.768080808080808,
+      "grad_norm": 0.00743967154994607,
+      "learning_rate": 7.076326002587323e-05,
+      "loss": 0.0024,
       "step": 547
     },
     {
-      "epoch": 1.404098623118796,
-      "grad_norm": 0.043509308248758316,
-      "learning_rate": 9.034537236520039e-05,
-      "loss": 0.018,
+      "epoch": 1.7713131313131312,
+      "grad_norm": 0.006492244079709053,
+      "learning_rate": 7.089262613195343e-05,
+      "loss": 0.002,
       "step": 548
     },
     {
-      "epoch": 1.4066602625680436,
-      "grad_norm": 0.017051199451088905,
-      "learning_rate": 9.029243968035496e-05,
-      "loss": 0.0083,
+      "epoch": 1.7745454545454544,
+      "grad_norm": 0.009637167677283287,
+      "learning_rate": 7.102199223803364e-05,
+      "loss": 0.0038,
       "step": 549
     },
     {
-      "epoch": 1.4092219020172911,
-      "grad_norm": 0.023373804986476898,
-      "learning_rate": 9.023937788281278e-05,
-      "loss": 0.0114,
+      "epoch": 1.7777777777777777,
+      "grad_norm": 0.009506979957222939,
+      "learning_rate": 7.115135834411385e-05,
+      "loss": 0.004,
       "step": 550
     },
     {
-      "epoch": 1.4092219020172911,
-      "eval_loss": 0.009199035353958607,
-      "eval_runtime": 15.7746,
-      "eval_samples_per_second": 6.339,
-      "eval_steps_per_second": 1.585,
+      "epoch": 1.7777777777777777,
+      "eval_loss": 0.002756154164671898,
+      "eval_runtime": 15.5019,
+      "eval_samples_per_second": 6.451,
+      "eval_steps_per_second": 1.613,
       "step": 550
     },
     {
-      "epoch": 1.4117835414665385,
-      "grad_norm": 0.027553927153348923,
-      "learning_rate": 9.018618714260453e-05,
-      "loss": 0.0181,
+      "epoch": 1.781010101010101,
+      "grad_norm": 0.01711469143629074,
+      "learning_rate": 7.128072445019404e-05,
+      "loss": 0.0078,
       "step": 551
     },
     {
-      "epoch": 1.414345180915786,
-      "grad_norm": 0.013546819798648357,
-      "learning_rate": 9.013286763017415e-05,
-      "loss": 0.0057,
+      "epoch": 1.7842424242424242,
+      "grad_norm": 0.008916409686207771,
+      "learning_rate": 7.141009055627425e-05,
+      "loss": 0.0026,
       "step": 552
     },
     {
-      "epoch": 1.4169068203650337,
-      "grad_norm": 0.01957760564982891,
-      "learning_rate": 9.00794195163781e-05,
-      "loss": 0.007,
+      "epoch": 1.7874747474747474,
+      "grad_norm": 0.00987724307924509,
+      "learning_rate": 7.153945666235446e-05,
+      "loss": 0.003,
       "step": 553
     },
     {
-      "epoch": 1.419468459814281,
-      "grad_norm": 0.023621102795004845,
-      "learning_rate": 9.002584297248502e-05,
-      "loss": 0.0102,
+      "epoch": 1.7907070707070707,
+      "grad_norm": 0.012879880145192146,
+      "learning_rate": 7.166882276843467e-05,
+      "loss": 0.0057,
       "step": 554
     },
     {
-      "epoch": 1.4220300992635286,
-      "grad_norm": 0.027797885239124298,
-      "learning_rate": 8.997213817017507e-05,
-      "loss": 0.011,
+      "epoch": 1.793939393939394,
+      "grad_norm": 0.012945498339831829,
+      "learning_rate": 7.179818887451488e-05,
+      "loss": 0.0049,
       "step": 555
     },
     {
-      "epoch": 1.4245917387127762,
-      "grad_norm": 0.026531843468546867,
-      "learning_rate": 8.991830528153935e-05,
+      "epoch": 1.7971717171717172,
+      "grad_norm": 0.016738496720790863,
+      "learning_rate": 7.19275549805951e-05,
       "loss": 0.0073,
       "step": 556
     },
     {
-      "epoch": 1.4271533781620236,
-      "grad_norm": 0.02077258750796318,
-      "learning_rate": 8.986434447907948e-05,
-      "loss": 0.0085,
+      "epoch": 1.8004040404040405,
+      "grad_norm": 0.009467408992350101,
+      "learning_rate": 7.20569210866753e-05,
+      "loss": 0.0031,
       "step": 557
     },
     {
-      "epoch": 1.4297150176112712,
-      "grad_norm": 0.018663110211491585,
-      "learning_rate": 8.98102559357069e-05,
-      "loss": 0.0097,
+      "epoch": 1.8036363636363637,
+      "grad_norm": 0.009352928958833218,
+      "learning_rate": 7.21862871927555e-05,
+      "loss": 0.0035,
       "step": 558
     },
     {
-      "epoch": 1.4322766570605188,
-      "grad_norm": 0.02353561483323574,
-      "learning_rate": 8.97560398247424e-05,
-      "loss": 0.0092,
+      "epoch": 1.806868686868687,
+      "grad_norm": 0.008519312366843224,
+      "learning_rate": 7.231565329883571e-05,
+      "loss": 0.0024,
       "step": 559
     },
     {
-      "epoch": 1.4348382965097661,
-      "grad_norm": 0.02463211864233017,
-      "learning_rate": 8.970169631991556e-05,
-      "loss": 0.0096,
+      "epoch": 1.8101010101010102,
+      "grad_norm": 0.007112348452210426,
+      "learning_rate": 7.244501940491591e-05,
+      "loss": 0.0018,
       "step": 560
     },
     {
-      "epoch": 1.4373999359590137,
-      "grad_norm": 0.023898137733340263,
-      "learning_rate": 8.964722559536418e-05,
-      "loss": 0.0107,
+      "epoch": 1.8133333333333335,
+      "grad_norm": 0.013203850015997887,
+      "learning_rate": 7.257438551099612e-05,
+      "loss": 0.0044,
       "step": 561
     },
     {
-      "epoch": 1.4399615754082613,
-      "grad_norm": 0.018616532906889915,
-      "learning_rate": 8.959262782563372e-05,
-      "loss": 0.0124,
+      "epoch": 1.8165656565656567,
+      "grad_norm": 0.008245746605098248,
+      "learning_rate": 7.270375161707633e-05,
+      "loss": 0.0031,
       "step": 562
     },
     {
-      "epoch": 1.4425232148575087,
-      "grad_norm": 0.026767538860440254,
-      "learning_rate": 8.95379031856767e-05,
-      "loss": 0.016,
+      "epoch": 1.8197979797979797,
+      "grad_norm": 0.010106631554663181,
+      "learning_rate": 7.283311772315654e-05,
+      "loss": 0.0035,
       "step": 563
     },
     {
-      "epoch": 1.4450848543067563,
-      "grad_norm": 0.019774265587329865,
-      "learning_rate": 8.948305185085225e-05,
-      "loss": 0.0073,
+      "epoch": 1.823030303030303,
+      "grad_norm": 0.006693434435874224,
+      "learning_rate": 7.296248382923674e-05,
+      "loss": 0.0027,
       "step": 564
     },
     {
-      "epoch": 1.4476464937560038,
-      "grad_norm": 0.026325702667236328,
-      "learning_rate": 8.942807399692543e-05,
-      "loss": 0.0114,
+      "epoch": 1.8262626262626263,
+      "grad_norm": 0.011071219108998775,
+      "learning_rate": 7.309184993531695e-05,
+      "loss": 0.003,
       "step": 565
     },
     {
-      "epoch": 1.4502081332052514,
-      "grad_norm": 0.02122092805802822,
-      "learning_rate": 8.937296980006675e-05,
-      "loss": 0.0114,
+      "epoch": 1.8294949494949495,
+      "grad_norm": 0.01001444086432457,
+      "learning_rate": 7.322121604139716e-05,
+      "loss": 0.0042,
       "step": 566
     },
     {
-      "epoch": 1.4527697726544988,
-      "grad_norm": 0.017576023936271667,
-      "learning_rate": 8.931773943685155e-05,
-      "loss": 0.01,
+      "epoch": 1.8327272727272728,
+      "grad_norm": 0.004930165130645037,
+      "learning_rate": 7.335058214747737e-05,
+      "loss": 0.0014,
       "step": 567
     },
     {
-      "epoch": 1.4553314121037464,
-      "grad_norm": 0.01859590783715248,
-      "learning_rate": 8.926238308425945e-05,
-      "loss": 0.0142,
+      "epoch": 1.835959595959596,
+      "grad_norm": 0.013421786017715931,
+      "learning_rate": 7.347994825355758e-05,
+      "loss": 0.0037,
       "step": 568
     },
     {
-      "epoch": 1.457893051552994,
-      "grad_norm": 0.02905329316854477,
-      "learning_rate": 8.92069009196738e-05,
-      "loss": 0.0186,
+      "epoch": 1.839191919191919,
+      "grad_norm": 0.008933313190937042,
+      "learning_rate": 7.360931435963779e-05,
+      "loss": 0.0031,
       "step": 569
     },
     {
-      "epoch": 1.4604546910022416,
-      "grad_norm": 0.014791314490139484,
-      "learning_rate": 8.915129312088112e-05,
-      "loss": 0.0049,
+      "epoch": 1.8424242424242423,
+      "grad_norm": 0.00960726197808981,
+      "learning_rate": 7.3738680465718e-05,
+      "loss": 0.0034,
       "step": 570
     },
     {
-      "epoch": 1.463016330451489,
-      "grad_norm": 0.0218088086694479,
-      "learning_rate": 8.909555986607048e-05,
-      "loss": 0.0146,
+      "epoch": 1.8456565656565656,
+      "grad_norm": 0.008845716714859009,
+      "learning_rate": 7.38680465717982e-05,
+      "loss": 0.0033,
       "step": 571
     },
     {
-      "epoch": 1.4655779699007365,
-      "grad_norm": 0.014257200062274933,
-      "learning_rate": 8.903970133383297e-05,
-      "loss": 0.0059,
+      "epoch": 1.8488888888888888,
+      "grad_norm": 0.009193913079798222,
+      "learning_rate": 7.399741267787839e-05,
+      "loss": 0.0029,
       "step": 572
     },
     {
-      "epoch": 1.468139609349984,
-      "grad_norm": 0.02303312160074711,
-      "learning_rate": 8.898371770316111e-05,
-      "loss": 0.0114,
+      "epoch": 1.852121212121212,
+      "grad_norm": 0.008480757474899292,
+      "learning_rate": 7.41267787839586e-05,
+      "loss": 0.0027,
       "step": 573
     },
     {
-      "epoch": 1.4707012487992315,
-      "grad_norm": 0.014151007868349552,
-      "learning_rate": 8.892760915344834e-05,
-      "loss": 0.0067,
+      "epoch": 1.8553535353535353,
+      "grad_norm": 0.006112278904765844,
+      "learning_rate": 7.425614489003881e-05,
+      "loss": 0.0026,
       "step": 574
     },
     {
-      "epoch": 1.473262888248479,
-      "grad_norm": 0.019835256040096283,
-      "learning_rate": 8.88713758644883e-05,
-      "loss": 0.0085,
+      "epoch": 1.8585858585858586,
+      "grad_norm": 0.009303019382059574,
+      "learning_rate": 7.438551099611902e-05,
+      "loss": 0.0038,
       "step": 575
     },
     {
-      "epoch": 1.473262888248479,
-      "eval_loss": 0.008619890548288822,
-      "eval_runtime": 15.8078,
-      "eval_samples_per_second": 6.326,
-      "eval_steps_per_second": 1.582,
+      "epoch": 1.8585858585858586,
+      "eval_loss": 0.0026209026109427214,
+      "eval_runtime": 15.4958,
+      "eval_samples_per_second": 6.453,
+      "eval_steps_per_second": 1.613,
       "step": 575
     },
     {
-      "epoch": 1.4758245276977267,
-      "grad_norm": 0.026124896481633186,
-      "learning_rate": 8.88150180164744e-05,
-      "loss": 0.0131,
+      "epoch": 1.8618181818181818,
+      "grad_norm": 0.006630861200392246,
+      "learning_rate": 7.451487710219923e-05,
+      "loss": 0.0029,
       "step": 576
     },
     {
-      "epoch": 1.478386167146974,
-      "grad_norm": 0.033744972199201584,
-      "learning_rate": 8.875853578999918e-05,
-      "loss": 0.0187,
+      "epoch": 1.865050505050505,
+      "grad_norm": 0.009775327518582344,
+      "learning_rate": 7.464424320827943e-05,
+      "loss": 0.0036,
       "step": 577
     },
     {
-      "epoch": 1.4809478065962216,
-      "grad_norm": 0.02144186943769455,
-      "learning_rate": 8.870192936605372e-05,
-      "loss": 0.0114,
+      "epoch": 1.8682828282828283,
+      "grad_norm": 0.006040921900421381,
+      "learning_rate": 7.477360931435964e-05,
+      "loss": 0.0026,
       "step": 578
     },
     {
-      "epoch": 1.4835094460454692,
-      "grad_norm": 0.020475219935178757,
-      "learning_rate": 8.86451989260271e-05,
-      "loss": 0.0079,
+      "epoch": 1.8715151515151516,
+      "grad_norm": 0.010307783260941505,
+      "learning_rate": 7.490297542043985e-05,
+      "loss": 0.0036,
       "step": 579
     },
     {
-      "epoch": 1.4860710854947166,
-      "grad_norm": 0.01916237361729145,
-      "learning_rate": 8.858834465170576e-05,
-      "loss": 0.0075,
+      "epoch": 1.8747474747474748,
+      "grad_norm": 0.01584269292652607,
+      "learning_rate": 7.503234152652006e-05,
+      "loss": 0.0098,
       "step": 580
     },
     {
-      "epoch": 1.4886327249439641,
-      "grad_norm": 0.016654282808303833,
-      "learning_rate": 8.853136672527299e-05,
-      "loss": 0.0085,
+      "epoch": 1.877979797979798,
+      "grad_norm": 0.009038683027029037,
+      "learning_rate": 7.516170763260027e-05,
+      "loss": 0.0028,
       "step": 581
     },
     {
-      "epoch": 1.4911943643932117,
-      "grad_norm": 0.026485484093427658,
-      "learning_rate": 8.847426532930831e-05,
-      "loss": 0.0145,
+      "epoch": 1.8812121212121213,
+      "grad_norm": 0.010477108880877495,
+      "learning_rate": 7.529107373868047e-05,
+      "loss": 0.0049,
       "step": 582
     },
     {
-      "epoch": 1.493756003842459,
-      "grad_norm": 0.019145099446177483,
-      "learning_rate": 8.841704064678684e-05,
-      "loss": 0.0092,
+      "epoch": 1.8844444444444446,
+      "grad_norm": 0.00762271648272872,
+      "learning_rate": 7.542043984476068e-05,
+      "loss": 0.0024,
       "step": 583
     },
     {
-      "epoch": 1.4963176432917067,
-      "grad_norm": 0.024889731779694557,
-      "learning_rate": 8.835969286107881e-05,
-      "loss": 0.0085,
+      "epoch": 1.8876767676767678,
+      "grad_norm": 0.013287222012877464,
+      "learning_rate": 7.554980595084087e-05,
+      "loss": 0.0031,
       "step": 584
     },
     {
-      "epoch": 1.4988792827409543,
-      "grad_norm": 0.01563897170126438,
-      "learning_rate": 8.83022221559489e-05,
-      "loss": 0.0086,
+      "epoch": 1.8909090909090909,
+      "grad_norm": 0.00888756476342678,
+      "learning_rate": 7.567917205692108e-05,
+      "loss": 0.0028,
       "step": 585
     },
     {
-      "epoch": 1.5014409221902016,
-      "grad_norm": 0.02054821327328682,
-      "learning_rate": 8.824462871555568e-05,
-      "loss": 0.0075,
+      "epoch": 1.8941414141414141,
+      "grad_norm": 0.006059638224542141,
+      "learning_rate": 7.58085381630013e-05,
+      "loss": 0.0019,
       "step": 586
     },
     {
-      "epoch": 1.5040025616394492,
-      "grad_norm": 0.021489789709448814,
-      "learning_rate": 8.818691272445098e-05,
-      "loss": 0.008,
+      "epoch": 1.8973737373737374,
+      "grad_norm": 0.007643230725079775,
+      "learning_rate": 7.59379042690815e-05,
+      "loss": 0.0036,
       "step": 587
     },
     {
-      "epoch": 1.5065642010886968,
-      "grad_norm": 0.018730612471699715,
-      "learning_rate": 8.812907436757935e-05,
-      "loss": 0.009,
+      "epoch": 1.9006060606060606,
+      "grad_norm": 0.009327744133770466,
+      "learning_rate": 7.606727037516171e-05,
+      "loss": 0.0041,
       "step": 588
     },
     {
-      "epoch": 1.5091258405379442,
-      "grad_norm": 0.01786506175994873,
-      "learning_rate": 8.807111383027748e-05,
-      "loss": 0.0096,
+      "epoch": 1.9038383838383839,
+      "grad_norm": 0.010678495280444622,
+      "learning_rate": 7.619663648124192e-05,
+      "loss": 0.0046,
       "step": 589
     },
     {
-      "epoch": 1.5116874799871918,
-      "grad_norm": 0.026313120499253273,
-      "learning_rate": 8.801303129827352e-05,
-      "loss": 0.0169,
+      "epoch": 1.907070707070707,
+      "grad_norm": 0.005262179300189018,
+      "learning_rate": 7.632600258732212e-05,
+      "loss": 0.001,
       "step": 590
     },
     {
-      "epoch": 1.5142491194364394,
-      "grad_norm": 0.015450164675712585,
-      "learning_rate": 8.795482695768658e-05,
-      "loss": 0.0074,
+      "epoch": 1.9103030303030302,
+      "grad_norm": 0.012793856672942638,
+      "learning_rate": 7.645536869340233e-05,
+      "loss": 0.0038,
       "step": 591
     },
     {
-      "epoch": 1.5168107588856867,
-      "grad_norm": 0.02219979465007782,
-      "learning_rate": 8.789650099502606e-05,
-      "loss": 0.0091,
+      "epoch": 1.9135353535353534,
+      "grad_norm": 0.005518015474081039,
+      "learning_rate": 7.658473479948254e-05,
+      "loss": 0.0015,
       "step": 592
     },
     {
-      "epoch": 1.5193723983349343,
-      "grad_norm": 0.018070019781589508,
-      "learning_rate": 8.783805359719113e-05,
-      "loss": 0.0105,
+      "epoch": 1.9167676767676767,
+      "grad_norm": 0.011437000706791878,
+      "learning_rate": 7.671410090556275e-05,
+      "loss": 0.0046,
       "step": 593
     },
     {
-      "epoch": 1.521934037784182,
-      "grad_norm": 0.01956484653055668,
-      "learning_rate": 8.777948495147001e-05,
-      "loss": 0.0108,
+      "epoch": 1.92,
+      "grad_norm": 0.008323602378368378,
+      "learning_rate": 7.684346701164295e-05,
+      "loss": 0.0026,
       "step": 594
     },
     {
-      "epoch": 1.5244956772334293,
-      "grad_norm": 0.020401043817400932,
-      "learning_rate": 8.772079524553951e-05,
-      "loss": 0.009,
+      "epoch": 1.9232323232323232,
+      "grad_norm": 0.006219903938472271,
+      "learning_rate": 7.697283311772316e-05,
+      "loss": 0.0021,
       "step": 595
     },
     {
-      "epoch": 1.527057316682677,
-      "grad_norm": 0.022120250388979912,
-      "learning_rate": 8.766198466746437e-05,
-      "loss": 0.0129,
+      "epoch": 1.9264646464646464,
+      "grad_norm": 0.011064271442592144,
+      "learning_rate": 7.710219922380336e-05,
+      "loss": 0.004,
       "step": 596
     },
     {
-      "epoch": 1.5296189561319244,
-      "grad_norm": 0.023862425237894058,
-      "learning_rate": 8.760305340569661e-05,
-      "loss": 0.016,
+      "epoch": 1.9296969696969697,
+      "grad_norm": 0.008776338770985603,
+      "learning_rate": 7.723156532988357e-05,
+      "loss": 0.0031,
       "step": 597
     },
     {
-      "epoch": 1.5321805955811718,
-      "grad_norm": 0.01728702336549759,
-      "learning_rate": 8.754400164907497e-05,
-      "loss": 0.0085,
+      "epoch": 1.932929292929293,
+      "grad_norm": 0.007282125763595104,
+      "learning_rate": 7.736093143596378e-05,
+      "loss": 0.0028,
       "step": 598
     },
     {
-      "epoch": 1.5347422350304196,
-      "grad_norm": 0.018914831802248955,
-      "learning_rate": 8.748482958682433e-05,
-      "loss": 0.0082,
+      "epoch": 1.9361616161616162,
+      "grad_norm": 0.008244170807301998,
+      "learning_rate": 7.749029754204399e-05,
+      "loss": 0.002,
       "step": 599
     },
     {
-      "epoch": 1.537303874479667,
-      "grad_norm": 0.023846952244639397,
-      "learning_rate": 8.742553740855506e-05,
-      "loss": 0.0139,
+      "epoch": 1.9393939393939394,
+      "grad_norm": 0.011086437851190567,
+      "learning_rate": 7.76196636481242e-05,
+      "loss": 0.0064,
       "step": 600
     },
     {
-      "epoch": 1.537303874479667,
-      "eval_loss": 0.009778290055692196,
-      "eval_runtime": 15.8286,
-      "eval_samples_per_second": 6.318,
-      "eval_steps_per_second": 1.579,
+      "epoch": 1.9393939393939394,
+      "eval_loss": 0.0025780757423490286,
+      "eval_runtime": 15.4653,
+      "eval_samples_per_second": 6.466,
+      "eval_steps_per_second": 1.617,
       "step": 600
     },
     {
-      "epoch": 1.5398655139289144,
-      "grad_norm": 0.01700327731668949,
-      "learning_rate": 8.736612530426245e-05,
-      "loss": 0.0074,
+      "epoch": 1.9426262626262627,
+      "grad_norm": 0.006747119594365358,
+      "learning_rate": 7.774902975420441e-05,
+      "loss": 0.0019,
       "step": 601
     },
     {
-      "epoch": 1.5424271533781622,
-      "grad_norm": 0.016836758702993393,
-      "learning_rate": 8.730659346432602e-05,
-      "loss": 0.0068,
+      "epoch": 1.945858585858586,
+      "grad_norm": 0.005355107598006725,
+      "learning_rate": 7.787839586028462e-05,
+      "loss": 0.0022,
       "step": 602
     },
     {
-      "epoch": 1.5449887928274095,
-      "grad_norm": 0.020055629312992096,
-      "learning_rate": 8.724694207950906e-05,
-      "loss": 0.0073,
+      "epoch": 1.9490909090909092,
+      "grad_norm": 0.012666920199990273,
+      "learning_rate": 7.800776196636481e-05,
+      "loss": 0.005,
       "step": 603
     },
     {
-      "epoch": 1.547550432276657,
-      "grad_norm": 0.029256854206323624,
-      "learning_rate": 8.718717134095784e-05,
-      "loss": 0.0211,
+      "epoch": 1.9523232323232325,
+      "grad_norm": 0.005357131827622652,
+      "learning_rate": 7.813712807244502e-05,
+      "loss": 0.0017,
       "step": 604
     },
     {
-      "epoch": 1.5501120717259047,
-      "grad_norm": 0.03150349482893944,
-      "learning_rate": 8.712728144020118e-05,
-      "loss": 0.0211,
+      "epoch": 1.9555555555555557,
+      "grad_norm": 0.007205512840300798,
+      "learning_rate": 7.826649417852523e-05,
+      "loss": 0.0027,
       "step": 605
     },
     {
-      "epoch": 1.552673711175152,
-      "grad_norm": 0.02234642021358013,
-      "learning_rate": 8.706727256914961e-05,
-      "loss": 0.0106,
+      "epoch": 1.9587878787878787,
+      "grad_norm": 0.009989005513489246,
+      "learning_rate": 7.839586028460543e-05,
+      "loss": 0.0029,
       "step": 606
     },
     {
-      "epoch": 1.5552353506243997,
-      "grad_norm": 0.02711409516632557,
-      "learning_rate": 8.700714492009501e-05,
-      "loss": 0.0151,
+      "epoch": 1.962020202020202,
+      "grad_norm": 0.00870651938021183,
+      "learning_rate": 7.852522639068564e-05,
+      "loss": 0.0033,
       "step": 607
     },
     {
-      "epoch": 1.5577969900736472,
-      "grad_norm": 0.021641191095113754,
-      "learning_rate": 8.694689868570984e-05,
-      "loss": 0.0099,
+      "epoch": 1.9652525252525253,
+      "grad_norm": 0.009465783834457397,
+      "learning_rate": 7.865459249676585e-05,
+      "loss": 0.0047,
       "step": 608
     },
     {
-      "epoch": 1.5603586295228946,
-      "grad_norm": 0.02513721212744713,
-      "learning_rate": 8.688653405904652e-05,
-      "loss": 0.0174,
+      "epoch": 1.9684848484848485,
+      "grad_norm": 0.012961370870471,
+      "learning_rate": 7.878395860284605e-05,
+      "loss": 0.0027,
       "step": 609
     },
     {
-      "epoch": 1.5629202689721422,
-      "grad_norm": 0.017550567165017128,
-      "learning_rate": 8.682605123353685e-05,
-      "loss": 0.0075,
+      "epoch": 1.9717171717171718,
+      "grad_norm": 0.007300790399312973,
+      "learning_rate": 7.891332470892626e-05,
+      "loss": 0.0028,
       "step": 610
     },
     {
-      "epoch": 1.5654819084213898,
-      "grad_norm": 0.04532385244965553,
-      "learning_rate": 8.676545040299145e-05,
-      "loss": 0.018,
+      "epoch": 1.9749494949494948,
+      "grad_norm": 0.010898030363023281,
+      "learning_rate": 7.904269081500647e-05,
+      "loss": 0.0062,
       "step": 611
     },
     {
-      "epoch": 1.5680435478706372,
-      "grad_norm": 0.021223684772849083,
-      "learning_rate": 8.670473176159896e-05,
-      "loss": 0.011,
+      "epoch": 1.978181818181818,
+      "grad_norm": 0.010487280786037445,
+      "learning_rate": 7.917205692108668e-05,
+      "loss": 0.0052,
       "step": 612
     },
     {
-      "epoch": 1.5706051873198847,
-      "grad_norm": 0.02921074442565441,
-      "learning_rate": 8.664389550392564e-05,
-      "loss": 0.014,
+      "epoch": 1.9814141414141413,
+      "grad_norm": 0.011068948544561863,
+      "learning_rate": 7.930142302716689e-05,
+      "loss": 0.0038,
       "step": 613
     },
     {
-      "epoch": 1.5731668267691323,
-      "grad_norm": 0.01453994307667017,
-      "learning_rate": 8.658294182491457e-05,
-      "loss": 0.0067,
+      "epoch": 1.9846464646464645,
+      "grad_norm": 0.008286312222480774,
+      "learning_rate": 7.94307891332471e-05,
+      "loss": 0.0038,
       "step": 614
     },
     {
-      "epoch": 1.5757284662183797,
-      "grad_norm": 0.02055704966187477,
-      "learning_rate": 8.652187091988517e-05,
-      "loss": 0.0136,
+      "epoch": 1.9878787878787878,
+      "grad_norm": 0.011803009547293186,
+      "learning_rate": 7.956015523932731e-05,
+      "loss": 0.0058,
       "step": 615
     },
     {
-      "epoch": 1.5782901056676273,
-      "grad_norm": 0.020393170416355133,
-      "learning_rate": 8.646068298453238e-05,
-      "loss": 0.0085,
+      "epoch": 1.991111111111111,
+      "grad_norm": 0.009356028400361538,
+      "learning_rate": 7.968952134540751e-05,
+      "loss": 0.0029,
       "step": 616
     },
     {
-      "epoch": 1.5808517451168749,
-      "grad_norm": 0.018596651032567024,
-      "learning_rate": 8.639937821492624e-05,
-      "loss": 0.0093,
+      "epoch": 1.9943434343434343,
+      "grad_norm": 0.009298199787735939,
+      "learning_rate": 7.981888745148772e-05,
+      "loss": 0.0032,
       "step": 617
     },
     {
-      "epoch": 1.5834133845661222,
-      "grad_norm": 0.01737053133547306,
-      "learning_rate": 8.633795680751116e-05,
-      "loss": 0.0085,
+      "epoch": 1.9975757575757576,
+      "grad_norm": 0.006886545103043318,
+      "learning_rate": 7.994825355756791e-05,
+      "loss": 0.0025,
       "step": 618
     },
     {
-      "epoch": 1.5859750240153698,
-      "grad_norm": 0.035250473767519,
-      "learning_rate": 8.62764189591053e-05,
-      "loss": 0.0162,
+      "epoch": 2.000808080808081,
+      "grad_norm": 0.01841968297958374,
+      "learning_rate": 8.007761966364812e-05,
+      "loss": 0.0098,
       "step": 619
     },
     {
-      "epoch": 1.5885366634646174,
-      "grad_norm": 0.01704743504524231,
-      "learning_rate": 8.621476486689992e-05,
-      "loss": 0.0082,
+      "epoch": 2.004040404040404,
+      "grad_norm": 0.006127892527729273,
+      "learning_rate": 8.020698576972833e-05,
+      "loss": 0.0022,
       "step": 620
     },
     {
-      "epoch": 1.5910983029138648,
-      "grad_norm": 0.020262829959392548,
-      "learning_rate": 8.615299472845877e-05,
-      "loss": 0.0152,
+      "epoch": 2.0072727272727273,
+      "grad_norm": 0.007111063692718744,
+      "learning_rate": 8.033635187580855e-05,
+      "loss": 0.0017,
       "step": 621
     },
     {
-      "epoch": 1.5936599423631124,
-      "grad_norm": 0.02265034057199955,
-      "learning_rate": 8.609110874171751e-05,
-      "loss": 0.0114,
+      "epoch": 2.0105050505050506,
+      "grad_norm": 0.006170625798404217,
+      "learning_rate": 8.046571798188874e-05,
+      "loss": 0.0021,
       "step": 622
     },
     {
-      "epoch": 1.59622158181236,
-      "grad_norm": 0.02068282850086689,
-      "learning_rate": 8.602910710498295e-05,
-      "loss": 0.0112,
+      "epoch": 2.013737373737374,
+      "grad_norm": 0.008881688117980957,
+      "learning_rate": 8.059508408796895e-05,
+      "loss": 0.0038,
       "step": 623
     },
     {
-      "epoch": 1.5987832212616073,
-      "grad_norm": 0.022094765678048134,
-      "learning_rate": 8.596699001693255e-05,
-      "loss": 0.0096,
+      "epoch": 2.016969696969697,
+      "grad_norm": 0.009946111589670181,
+      "learning_rate": 8.072445019404916e-05,
+      "loss": 0.0028,
       "step": 624
     },
     {
-      "epoch": 1.601344860710855,
-      "grad_norm": 0.03619355335831642,
-      "learning_rate": 8.59047576766137e-05,
-      "loss": 0.0101,
+      "epoch": 2.0202020202020203,
+      "grad_norm": 0.009753699414432049,
+      "learning_rate": 8.085381630012937e-05,
+      "loss": 0.0044,
       "step": 625
     },
     {
-      "epoch": 1.601344860710855,
-      "eval_loss": 0.008696233853697777,
-      "eval_runtime": 15.7787,
-      "eval_samples_per_second": 6.338,
-      "eval_steps_per_second": 1.584,
+      "epoch": 2.0202020202020203,
+      "eval_loss": 0.0025918101891875267,
+      "eval_runtime": 15.4934,
+      "eval_samples_per_second": 6.454,
+      "eval_steps_per_second": 1.614,
       "step": 625
     },
     {
-      "epoch": 1.6039065001601025,
-      "grad_norm": 0.038272786885499954,
-      "learning_rate": 8.584241028344308e-05,
-      "loss": 0.0146,
+      "epoch": 2.0234343434343436,
+      "grad_norm": 0.00842776708304882,
+      "learning_rate": 8.098318240620958e-05,
+      "loss": 0.0028,
       "step": 626
     },
     {
-      "epoch": 1.6064681396093499,
-      "grad_norm": 0.02607797272503376,
-      "learning_rate": 8.577994803720606e-05,
-      "loss": 0.0108,
+      "epoch": 2.026666666666667,
+      "grad_norm": 0.010153642855584621,
+      "learning_rate": 8.111254851228979e-05,
+      "loss": 0.0034,
       "step": 627
     },
     {
-      "epoch": 1.6090297790585975,
-      "grad_norm": 0.02368808165192604,
-      "learning_rate": 8.571737113805606e-05,
-      "loss": 0.0157,
+      "epoch": 2.02989898989899,
+      "grad_norm": 0.011135900393128395,
+      "learning_rate": 8.124191461836999e-05,
+      "loss": 0.0032,
       "step": 628
     },
     {
-      "epoch": 1.611591418507845,
-      "grad_norm": 0.017954347655177116,
-      "learning_rate": 8.565467978651387e-05,
-      "loss": 0.0058,
+      "epoch": 2.0331313131313133,
+      "grad_norm": 0.012128735892474651,
+      "learning_rate": 8.13712807244502e-05,
+      "loss": 0.0042,
       "step": 629
     },
     {
-      "epoch": 1.6141530579570924,
-      "grad_norm": 0.03589998185634613,
-      "learning_rate": 8.559187418346703e-05,
-      "loss": 0.0234,
+      "epoch": 2.036363636363636,
+      "grad_norm": 0.007697262801229954,
+      "learning_rate": 8.15006468305304e-05,
+      "loss": 0.0023,
       "step": 630
     },
     {
-      "epoch": 1.6167146974063402,
-      "grad_norm": 0.02698587067425251,
-      "learning_rate": 8.552895453016922e-05,
-      "loss": 0.0136,
+      "epoch": 2.0395959595959594,
+      "grad_norm": 0.00962636899203062,
+      "learning_rate": 8.163001293661061e-05,
+      "loss": 0.0031,
       "step": 631
     },
     {
-      "epoch": 1.6192763368555876,
-      "grad_norm": 0.02276044897735119,
-      "learning_rate": 8.546592102823955e-05,
-      "loss": 0.0094,
+      "epoch": 2.0428282828282827,
+      "grad_norm": 0.009567685425281525,
+      "learning_rate": 8.175937904269082e-05,
+      "loss": 0.0044,
       "step": 632
     },
     {
-      "epoch": 1.621837976304835,
-      "grad_norm": 0.018764210864901543,
-      "learning_rate": 8.540277387966194e-05,
-      "loss": 0.0074,
+      "epoch": 2.046060606060606,
+      "grad_norm": 0.006336756516247988,
+      "learning_rate": 8.188874514877103e-05,
+      "loss": 0.0021,
       "step": 633
     },
     {
-      "epoch": 1.6243996157540828,
-      "grad_norm": 0.019631950184702873,
-      "learning_rate": 8.53395132867845e-05,
-      "loss": 0.0077,
+      "epoch": 2.049292929292929,
+      "grad_norm": 0.008533505722880363,
+      "learning_rate": 8.201811125485124e-05,
+      "loss": 0.0026,
       "step": 634
     },
     {
-      "epoch": 1.6269612552033301,
-      "grad_norm": 0.017163440585136414,
-      "learning_rate": 8.527613945231885e-05,
-      "loss": 0.0088,
+      "epoch": 2.0525252525252524,
+      "grad_norm": 0.013600981794297695,
+      "learning_rate": 8.214747736093143e-05,
+      "loss": 0.0063,
       "step": 635
     },
     {
-      "epoch": 1.6295228946525775,
-      "grad_norm": 0.01980568654835224,
-      "learning_rate": 8.521265257933948e-05,
-      "loss": 0.0105,
+      "epoch": 2.0557575757575757,
+      "grad_norm": 0.005397825501859188,
+      "learning_rate": 8.227684346701164e-05,
+      "loss": 0.0012,
       "step": 636
     },
     {
-      "epoch": 1.6320845341018253,
-      "grad_norm": 0.0192473866045475,
-      "learning_rate": 8.51490528712831e-05,
-      "loss": 0.0096,
+      "epoch": 2.058989898989899,
+      "grad_norm": 0.012294902466237545,
+      "learning_rate": 8.240620957309186e-05,
+      "loss": 0.0041,
       "step": 637
     },
     {
-      "epoch": 1.6346461735510727,
-      "grad_norm": 0.023643024265766144,
-      "learning_rate": 8.508534053194798e-05,
-      "loss": 0.0115,
+      "epoch": 2.062222222222222,
+      "grad_norm": 0.007750959601253271,
+      "learning_rate": 8.253557567917207e-05,
+      "loss": 0.0025,
       "step": 638
     },
     {
-      "epoch": 1.6372078130003203,
-      "grad_norm": 0.031089499592781067,
-      "learning_rate": 8.502151576549333e-05,
-      "loss": 0.0218,
+      "epoch": 2.0654545454545454,
+      "grad_norm": 0.006924534682184458,
+      "learning_rate": 8.266494178525228e-05,
+      "loss": 0.0029,
       "step": 639
     },
     {
-      "epoch": 1.6397694524495678,
-      "grad_norm": 0.021543102338910103,
-      "learning_rate": 8.495757877643858e-05,
-      "loss": 0.0111,
+      "epoch": 2.0686868686868687,
+      "grad_norm": 0.007069743238389492,
+      "learning_rate": 8.279430789133247e-05,
+      "loss": 0.0024,
       "step": 640
     },
     {
-      "epoch": 1.6423310918988152,
-      "grad_norm": 0.019607752561569214,
-      "learning_rate": 8.48935297696628e-05,
-      "loss": 0.0124,
+      "epoch": 2.071919191919192,
+      "grad_norm": 0.009360691532492638,
+      "learning_rate": 8.292367399741268e-05,
+      "loss": 0.0034,
       "step": 641
     },
     {
-      "epoch": 1.6448927313480628,
-      "grad_norm": 0.012617550790309906,
-      "learning_rate": 8.482936895040403e-05,
-      "loss": 0.005,
+      "epoch": 2.075151515151515,
+      "grad_norm": 0.008420188911259174,
+      "learning_rate": 8.305304010349288e-05,
+      "loss": 0.0033,
       "step": 642
     },
     {
-      "epoch": 1.6474543707973104,
-      "grad_norm": 0.019957462325692177,
-      "learning_rate": 8.476509652425851e-05,
-      "loss": 0.0069,
+      "epoch": 2.0783838383838384,
+      "grad_norm": 0.005898126866668463,
+      "learning_rate": 8.318240620957309e-05,
+      "loss": 0.0016,
       "step": 643
     },
     {
-      "epoch": 1.6500160102465578,
-      "grad_norm": 0.02268521673977375,
-      "learning_rate": 8.470071269718019e-05,
-      "loss": 0.0091,
+      "epoch": 2.0816161616161617,
+      "grad_norm": 0.007344219367951155,
+      "learning_rate": 8.33117723156533e-05,
+      "loss": 0.0023,
       "step": 644
     },
     {
-      "epoch": 1.6525776496958053,
-      "grad_norm": 0.0193330105394125,
-      "learning_rate": 8.463621767547998e-05,
-      "loss": 0.0095,
+      "epoch": 2.084848484848485,
+      "grad_norm": 0.004704777151346207,
+      "learning_rate": 8.344113842173351e-05,
+      "loss": 0.0016,
       "step": 645
     },
     {
-      "epoch": 1.655139289145053,
-      "grad_norm": 0.021894952282309532,
-      "learning_rate": 8.457161166582507e-05,
-      "loss": 0.008,
+      "epoch": 2.088080808080808,
+      "grad_norm": 0.0050371866673231125,
+      "learning_rate": 8.357050452781372e-05,
+      "loss": 0.0013,
       "step": 646
     },
     {
-      "epoch": 1.6577009285943003,
-      "grad_norm": 0.021809278056025505,
-      "learning_rate": 8.450689487523836e-05,
-      "loss": 0.0063,
+      "epoch": 2.0913131313131315,
+      "grad_norm": 0.01141134649515152,
+      "learning_rate": 8.369987063389393e-05,
+      "loss": 0.0023,
       "step": 647
     },
     {
-      "epoch": 1.6602625680435479,
-      "grad_norm": 0.02643033117055893,
-      "learning_rate": 8.444206751109767e-05,
-      "loss": 0.0117,
+      "epoch": 2.0945454545454547,
+      "grad_norm": 0.007947816513478756,
+      "learning_rate": 8.382923673997413e-05,
+      "loss": 0.003,
       "step": 648
     },
     {
-      "epoch": 1.6628242074927955,
-      "grad_norm": 0.033799465745687485,
-      "learning_rate": 8.437712978113516e-05,
-      "loss": 0.0214,
+      "epoch": 2.097777777777778,
+      "grad_norm": 0.009138518944382668,
+      "learning_rate": 8.395860284605434e-05,
+      "loss": 0.0027,
       "step": 649
     },
     {
-      "epoch": 1.6653858469420428,
-      "grad_norm": 0.024205051362514496,
-      "learning_rate": 8.43120818934367e-05,
-      "loss": 0.0111,
+      "epoch": 2.101010101010101,
+      "grad_norm": 0.0078082941472530365,
+      "learning_rate": 8.408796895213455e-05,
+      "loss": 0.0022,
       "step": 650
     },
     {
-      "epoch": 1.6653858469420428,
-      "eval_loss": 0.008477776311337948,
-      "eval_runtime": 15.7954,
-      "eval_samples_per_second": 6.331,
-      "eval_steps_per_second": 1.583,
+      "epoch": 2.101010101010101,
+      "eval_loss": 0.0024645410012453794,
+      "eval_runtime": 15.4794,
+      "eval_samples_per_second": 6.46,
+      "eval_steps_per_second": 1.615,
       "step": 650
     },
     {
-      "epoch": 1.6679474863912904,
-      "grad_norm": 0.021455738693475723,
-      "learning_rate": 8.424692405644102e-05,
-      "loss": 0.0076,
+      "epoch": 2.1042424242424245,
+      "grad_norm": 0.011119055561721325,
+      "learning_rate": 8.421733505821476e-05,
+      "loss": 0.0034,
       "step": 651
     },
     {
-      "epoch": 1.670509125840538,
-      "grad_norm": 0.012590169906616211,
-      "learning_rate": 8.418165647893931e-05,
-      "loss": 0.0048,
+      "epoch": 2.1074747474747473,
+      "grad_norm": 0.008150856010615826,
+      "learning_rate": 8.434670116429496e-05,
+      "loss": 0.0018,
       "step": 652
     },
     {
-      "epoch": 1.6730707652897854,
-      "grad_norm": 0.025595275685191154,
-      "learning_rate": 8.411627937007433e-05,
-      "loss": 0.0127,
+      "epoch": 2.1107070707070705,
+      "grad_norm": 0.007952040992677212,
+      "learning_rate": 8.447606727037517e-05,
+      "loss": 0.0029,
       "step": 653
     },
     {
-      "epoch": 1.675632404739033,
-      "grad_norm": 0.02153213880956173,
-      "learning_rate": 8.405079293933986e-05,
-      "loss": 0.0086,
+      "epoch": 2.113939393939394,
+      "grad_norm": 0.008000900037586689,
+      "learning_rate": 8.460543337645536e-05,
+      "loss": 0.0024,
       "step": 654
     },
     {
-      "epoch": 1.6781940441882806,
-      "grad_norm": 0.03660966455936432,
-      "learning_rate": 8.398519739657996e-05,
-      "loss": 0.0196,
+      "epoch": 2.117171717171717,
+      "grad_norm": 0.010997539386153221,
+      "learning_rate": 8.473479948253557e-05,
+      "loss": 0.0041,
       "step": 655
     },
     {
-      "epoch": 1.680755683637528,
-      "grad_norm": 0.019230369478464127,
-      "learning_rate": 8.391949295198836e-05,
-      "loss": 0.0086,
+      "epoch": 2.1204040404040403,
+      "grad_norm": 0.00813035573810339,
+      "learning_rate": 8.486416558861578e-05,
+      "loss": 0.0029,
       "step": 656
     },
     {
-      "epoch": 1.6833173230867755,
-      "grad_norm": 0.020435554906725883,
-      "learning_rate": 8.38536798161077e-05,
-      "loss": 0.0079,
+      "epoch": 2.1236363636363635,
+      "grad_norm": 0.009984957054257393,
+      "learning_rate": 8.499353169469599e-05,
+      "loss": 0.0043,
       "step": 657
     },
     {
-      "epoch": 1.685878962536023,
-      "grad_norm": 0.03471748158335686,
-      "learning_rate": 8.378775819982896e-05,
-      "loss": 0.0216,
+      "epoch": 2.126868686868687,
+      "grad_norm": 0.011672861874103546,
+      "learning_rate": 8.51228978007762e-05,
+      "loss": 0.0052,
       "step": 658
     },
     {
-      "epoch": 1.6884406019852705,
-      "grad_norm": 0.026042545214295387,
-      "learning_rate": 8.372172831439072e-05,
-      "loss": 0.01,
+      "epoch": 2.13010101010101,
+      "grad_norm": 0.012015840969979763,
+      "learning_rate": 8.525226390685641e-05,
+      "loss": 0.0033,
       "step": 659
     },
     {
-      "epoch": 1.691002241434518,
-      "grad_norm": 0.021988200023770332,
-      "learning_rate": 8.365559037137852e-05,
-      "loss": 0.0081,
+      "epoch": 2.1333333333333333,
+      "grad_norm": 0.009852144867181778,
+      "learning_rate": 8.538163001293662e-05,
+      "loss": 0.0035,
       "step": 660
     },
     {
-      "epoch": 1.6935638808837656,
-      "grad_norm": 0.020411528646945953,
-      "learning_rate": 8.358934458272409e-05,
-      "loss": 0.0089,
+      "epoch": 2.1365656565656566,
+      "grad_norm": 0.007703031413257122,
+      "learning_rate": 8.551099611901682e-05,
+      "loss": 0.003,
       "step": 661
     },
     {
-      "epoch": 1.696125520333013,
-      "grad_norm": 0.03231273591518402,
-      "learning_rate": 8.352299116070481e-05,
-      "loss": 0.0139,
+      "epoch": 2.13979797979798,
+      "grad_norm": 0.005590579938143492,
+      "learning_rate": 8.564036222509703e-05,
+      "loss": 0.0012,
       "step": 662
     },
     {
-      "epoch": 1.6986871597822606,
-      "grad_norm": 0.02595127373933792,
-      "learning_rate": 8.345653031794292e-05,
-      "loss": 0.0046,
+      "epoch": 2.143030303030303,
+      "grad_norm": 0.00858419667929411,
+      "learning_rate": 8.576972833117724e-05,
+      "loss": 0.0027,
       "step": 663
     },
     {
-      "epoch": 1.7012487992315082,
-      "grad_norm": 0.02997523732483387,
-      "learning_rate": 8.33899622674049e-05,
-      "loss": 0.0136,
+      "epoch": 2.1462626262626263,
+      "grad_norm": 0.010091731324791908,
+      "learning_rate": 8.589909443725744e-05,
+      "loss": 0.0019,
       "step": 664
     },
     {
-      "epoch": 1.7038104386807555,
-      "grad_norm": 0.020632093772292137,
-      "learning_rate": 8.332328722240073e-05,
-      "loss": 0.0101,
+      "epoch": 2.1494949494949496,
+      "grad_norm": 0.014338680543005466,
+      "learning_rate": 8.602846054333765e-05,
+      "loss": 0.0038,
       "step": 665
     },
     {
-      "epoch": 1.7063720781300034,
-      "grad_norm": 0.016184786334633827,
-      "learning_rate": 8.32565053965833e-05,
-      "loss": 0.0069,
+      "epoch": 2.152727272727273,
+      "grad_norm": 0.006007707212120295,
+      "learning_rate": 8.615782664941786e-05,
+      "loss": 0.002,
       "step": 666
     },
     {
-      "epoch": 1.7089337175792507,
-      "grad_norm": 0.02015138790011406,
-      "learning_rate": 8.31896170039476e-05,
-      "loss": 0.0093,
+      "epoch": 2.155959595959596,
+      "grad_norm": 0.006070902571082115,
+      "learning_rate": 8.628719275549805e-05,
+      "loss": 0.0021,
       "step": 667
     },
     {
-      "epoch": 1.711495357028498,
-      "grad_norm": 0.023183198645710945,
-      "learning_rate": 8.312262225883019e-05,
-      "loss": 0.0095,
+      "epoch": 2.1591919191919193,
+      "grad_norm": 0.008970859460532665,
+      "learning_rate": 8.641655886157827e-05,
+      "loss": 0.0029,
       "step": 668
     },
     {
-      "epoch": 1.714056996477746,
-      "grad_norm": 0.02725174091756344,
-      "learning_rate": 8.305552137590832e-05,
-      "loss": 0.0146,
+      "epoch": 2.1624242424242426,
+      "grad_norm": 0.009294882416725159,
+      "learning_rate": 8.654592496765848e-05,
+      "loss": 0.003,
       "step": 669
     },
     {
-      "epoch": 1.7166186359269933,
-      "grad_norm": 0.022738030180335045,
-      "learning_rate": 8.298831457019942e-05,
-      "loss": 0.0081,
+      "epoch": 2.165656565656566,
+      "grad_norm": 0.012971431016921997,
+      "learning_rate": 8.667529107373869e-05,
+      "loss": 0.0045,
       "step": 670
     },
     {
-      "epoch": 1.7191802753762406,
-      "grad_norm": 0.018372569233179092,
-      "learning_rate": 8.29210020570603e-05,
-      "loss": 0.0071,
+      "epoch": 2.168888888888889,
+      "grad_norm": 0.011960712261497974,
+      "learning_rate": 8.68046571798189e-05,
+      "loss": 0.0033,
       "step": 671
     },
     {
-      "epoch": 1.7217419148254884,
-      "grad_norm": 0.03261580318212509,
-      "learning_rate": 8.285358405218655e-05,
-      "loss": 0.0257,
+      "epoch": 2.172121212121212,
+      "grad_norm": 0.007202782668173313,
+      "learning_rate": 8.69340232858991e-05,
+      "loss": 0.0022,
       "step": 672
     },
     {
-      "epoch": 1.7243035542747358,
-      "grad_norm": 0.03144574537873268,
-      "learning_rate": 8.27860607716117e-05,
-      "loss": 0.0147,
+      "epoch": 2.175353535353535,
+      "grad_norm": 0.006524711847305298,
+      "learning_rate": 8.706338939197932e-05,
+      "loss": 0.0026,
       "step": 673
     },
     {
-      "epoch": 1.7268651937239834,
-      "grad_norm": 0.019830984994769096,
-      "learning_rate": 8.271843243170677e-05,
-      "loss": 0.008,
+      "epoch": 2.1785858585858584,
+      "grad_norm": 0.005754322744905949,
+      "learning_rate": 8.719275549805951e-05,
+      "loss": 0.0017,
       "step": 674
     },
     {
-      "epoch": 1.729426833173231,
-      "grad_norm": 0.01626870594918728,
-      "learning_rate": 8.265069924917925e-05,
-      "loss": 0.0075,
+      "epoch": 2.1818181818181817,
+      "grad_norm": 0.005853266920894384,
+      "learning_rate": 8.732212160413972e-05,
+      "loss": 0.0019,
       "step": 675
     },
     {
-      "epoch": 1.729426833173231,
-      "eval_loss": 0.007765541784465313,
-      "eval_runtime": 15.7645,
-      "eval_samples_per_second": 6.343,
-      "eval_steps_per_second": 1.586,
+      "epoch": 2.1818181818181817,
+      "eval_loss": 0.00235073734074831,
+      "eval_runtime": 15.4542,
+      "eval_samples_per_second": 6.471,
+      "eval_steps_per_second": 1.618,
       "step": 675
     },
     {
-      "epoch": 1.7319884726224783,
-      "grad_norm": 0.030844878405332565,
-      "learning_rate": 8.258286144107276e-05,
-      "loss": 0.0109,
+      "epoch": 2.185050505050505,
+      "grad_norm": 0.007134635932743549,
+      "learning_rate": 8.745148771021992e-05,
+      "loss": 0.0025,
       "step": 676
     },
     {
-      "epoch": 1.734550112071726,
-      "grad_norm": 0.019611435011029243,
-      "learning_rate": 8.251491922476609e-05,
-      "loss": 0.0069,
+      "epoch": 2.188282828282828,
+      "grad_norm": 0.006569647695869207,
+      "learning_rate": 8.758085381630013e-05,
+      "loss": 0.0021,
       "step": 677
     },
     {
-      "epoch": 1.7371117515209735,
-      "grad_norm": 0.02465008571743965,
-      "learning_rate": 8.244687281797255e-05,
-      "loss": 0.0112,
+      "epoch": 2.1915151515151514,
+      "grad_norm": 0.007616804912686348,
+      "learning_rate": 8.771021992238034e-05,
+      "loss": 0.0031,
       "step": 678
     },
     {
-      "epoch": 1.739673390970221,
-      "grad_norm": 0.015665439888834953,
-      "learning_rate": 8.237872243873945e-05,
-      "loss": 0.0059,
+      "epoch": 2.1947474747474747,
+      "grad_norm": 0.009457720443606377,
+      "learning_rate": 8.783958602846055e-05,
+      "loss": 0.0024,
       "step": 679
     },
     {
-      "epoch": 1.7422350304194685,
-      "grad_norm": 0.02098284289240837,
-      "learning_rate": 8.231046830544716e-05,
-      "loss": 0.0078,
+      "epoch": 2.197979797979798,
+      "grad_norm": 0.013973189517855644,
+      "learning_rate": 8.796895213454075e-05,
+      "loss": 0.0058,
       "step": 680
     },
     {
-      "epoch": 1.744796669868716,
-      "grad_norm": 0.02483079396188259,
-      "learning_rate": 8.224211063680853e-05,
-      "loss": 0.0139,
+      "epoch": 2.201212121212121,
+      "grad_norm": 0.007282907143235207,
+      "learning_rate": 8.809831824062096e-05,
+      "loss": 0.0018,
       "step": 681
     },
     {
-      "epoch": 1.7473583093179634,
-      "grad_norm": 0.014874100685119629,
-      "learning_rate": 8.217364965186824e-05,
-      "loss": 0.0083,
+      "epoch": 2.2044444444444444,
+      "grad_norm": 0.009663967415690422,
+      "learning_rate": 8.822768434670117e-05,
+      "loss": 0.0033,
       "step": 682
     },
     {
-      "epoch": 1.749919948767211,
-      "grad_norm": 0.017594069242477417,
-      "learning_rate": 8.2105085570002e-05,
-      "loss": 0.0092,
+      "epoch": 2.2076767676767677,
+      "grad_norm": 0.007305040489882231,
+      "learning_rate": 8.835705045278138e-05,
+      "loss": 0.0027,
       "step": 683
     },
     {
-      "epoch": 1.7524815882164586,
-      "grad_norm": 0.01697852835059166,
-      "learning_rate": 8.203641861091583e-05,
-      "loss": 0.0077,
+      "epoch": 2.210909090909091,
+      "grad_norm": 0.009345808997750282,
+      "learning_rate": 8.848641655886159e-05,
+      "loss": 0.0023,
       "step": 684
     },
     {
-      "epoch": 1.755043227665706,
-      "grad_norm": 0.019407397136092186,
-      "learning_rate": 8.196764899464551e-05,
-      "loss": 0.0064,
+      "epoch": 2.214141414141414,
+      "grad_norm": 0.006080677267163992,
+      "learning_rate": 8.86157826649418e-05,
+      "loss": 0.002,
       "step": 685
     },
     {
-      "epoch": 1.7576048671149536,
-      "grad_norm": 0.024107495322823524,
-      "learning_rate": 8.189877694155568e-05,
-      "loss": 0.0143,
+      "epoch": 2.2173737373737374,
+      "grad_norm": 0.0093191247433424,
+      "learning_rate": 8.8745148771022e-05,
+      "loss": 0.0026,
       "step": 686
     },
     {
-      "epoch": 1.7601665065642011,
-      "grad_norm": 0.033490199595689774,
-      "learning_rate": 8.182980267233929e-05,
-      "loss": 0.0077,
+      "epoch": 2.2206060606060607,
+      "grad_norm": 0.009632972069084644,
+      "learning_rate": 8.88745148771022e-05,
+      "loss": 0.003,
       "step": 687
     },
     {
-      "epoch": 1.7627281460134485,
-      "grad_norm": 0.02660991996526718,
-      "learning_rate": 8.17607264080168e-05,
-      "loss": 0.0087,
+      "epoch": 2.223838383838384,
+      "grad_norm": 0.011811980977654457,
+      "learning_rate": 8.90038809831824e-05,
+      "loss": 0.0028,
       "step": 688
     },
     {
-      "epoch": 1.765289785462696,
-      "grad_norm": 0.020847605541348457,
-      "learning_rate": 8.169154836993551e-05,
-      "loss": 0.0085,
+      "epoch": 2.227070707070707,
+      "grad_norm": 0.009458387270569801,
+      "learning_rate": 8.913324708926261e-05,
+      "loss": 0.0029,
       "step": 689
     },
     {
-      "epoch": 1.7678514249119437,
-      "grad_norm": 0.01973964460194111,
-      "learning_rate": 8.162226877976887e-05,
-      "loss": 0.0083,
+      "epoch": 2.2303030303030305,
+      "grad_norm": 0.008128968067467213,
+      "learning_rate": 8.926261319534282e-05,
+      "loss": 0.0028,
       "step": 690
     },
     {
-      "epoch": 1.770413064361191,
-      "grad_norm": 0.01618560589849949,
-      "learning_rate": 8.155288785951569e-05,
-      "loss": 0.009,
+      "epoch": 2.2335353535353537,
+      "grad_norm": 0.00910105463117361,
+      "learning_rate": 8.939197930142303e-05,
+      "loss": 0.0029,
       "step": 691
     },
     {
-      "epoch": 1.7729747038104386,
-      "grad_norm": 0.0218571238219738,
-      "learning_rate": 8.148340583149949e-05,
-      "loss": 0.0088,
+      "epoch": 2.236767676767677,
+      "grad_norm": 0.01368822529911995,
+      "learning_rate": 8.952134540750324e-05,
+      "loss": 0.0068,
       "step": 692
     },
     {
-      "epoch": 1.7755363432596862,
-      "grad_norm": 0.015629569068551064,
-      "learning_rate": 8.141382291836784e-05,
-      "loss": 0.0062,
+      "epoch": 2.24,
+      "grad_norm": 0.00982614140957594,
+      "learning_rate": 8.965071151358344e-05,
+      "loss": 0.0042,
       "step": 693
     },
     {
-      "epoch": 1.7780979827089336,
-      "grad_norm": 0.01955568976700306,
-      "learning_rate": 8.134413934309151e-05,
-      "loss": 0.0107,
+      "epoch": 2.2432323232323235,
+      "grad_norm": 0.008709289133548737,
+      "learning_rate": 8.978007761966365e-05,
+      "loss": 0.0023,
       "step": 694
     },
     {
-      "epoch": 1.7806596221581812,
-      "grad_norm": 0.028023671358823776,
-      "learning_rate": 8.127435532896388e-05,
-      "loss": 0.0139,
+      "epoch": 2.2464646464646463,
+      "grad_norm": 0.007170096971094608,
+      "learning_rate": 8.990944372574386e-05,
+      "loss": 0.0025,
       "step": 695
     },
     {
-      "epoch": 1.7832212616074288,
-      "grad_norm": 0.018414363265037537,
-      "learning_rate": 8.120447109960014e-05,
-      "loss": 0.0082,
+      "epoch": 2.2496969696969695,
+      "grad_norm": 0.008302228525280952,
+      "learning_rate": 9.003880983182407e-05,
+      "loss": 0.0021,
       "step": 696
     },
     {
-      "epoch": 1.7857829010566761,
-      "grad_norm": 0.024728137999773026,
-      "learning_rate": 8.113448687893664e-05,
-      "loss": 0.0111,
+      "epoch": 2.252929292929293,
+      "grad_norm": 0.011192730627954006,
+      "learning_rate": 9.016817593790428e-05,
+      "loss": 0.0024,
       "step": 697
     },
     {
-      "epoch": 1.788344540505924,
-      "grad_norm": 0.04410293325781822,
-      "learning_rate": 8.106440289123011e-05,
-      "loss": 0.0265,
+      "epoch": 2.256161616161616,
+      "grad_norm": 0.011657851748168468,
+      "learning_rate": 9.029754204398448e-05,
+      "loss": 0.0042,
       "step": 698
     },
     {
-      "epoch": 1.7909061799551713,
-      "grad_norm": 0.012759803794324398,
-      "learning_rate": 8.099421936105702e-05,
-      "loss": 0.0047,
+      "epoch": 2.2593939393939393,
+      "grad_norm": 0.009274295531213284,
+      "learning_rate": 9.042690815006469e-05,
+      "loss": 0.0031,
       "step": 699
     },
     {
-      "epoch": 1.7934678194044187,
-      "grad_norm": 0.017002368345856667,
-      "learning_rate": 8.092393651331275e-05,
-      "loss": 0.0077,
+      "epoch": 2.2626262626262625,
+      "grad_norm": 0.015825092792510986,
+      "learning_rate": 9.055627425614489e-05,
+      "loss": 0.0057,
       "step": 700
     },
     {
-      "epoch": 1.7934678194044187,
-      "eval_loss": 0.008254488930106163,
-      "eval_runtime": 15.7858,
-      "eval_samples_per_second": 6.335,
-      "eval_steps_per_second": 1.584,
+      "epoch": 2.2626262626262625,
+      "eval_loss": 0.002384421182796359,
+      "eval_runtime": 15.4875,
+      "eval_samples_per_second": 6.457,
+      "eval_steps_per_second": 1.614,
       "step": 700
     },
     {
-      "epoch": 1.7960294588536665,
-      "grad_norm": 0.017149921506643295,
-      "learning_rate": 8.085355457321098e-05,
-      "loss": 0.0085,
+      "epoch": 2.265858585858586,
+      "grad_norm": 0.012725941836833954,
+      "learning_rate": 9.06856403622251e-05,
+      "loss": 0.0038,
       "step": 701
     },
     {
-      "epoch": 1.7985910983029139,
-      "grad_norm": 0.013030176050961018,
-      "learning_rate": 8.07830737662829e-05,
-      "loss": 0.0058,
+      "epoch": 2.269090909090909,
+      "grad_norm": 0.006863633170723915,
+      "learning_rate": 9.08150064683053e-05,
+      "loss": 0.0021,
       "step": 702
     },
     {
-      "epoch": 1.8011527377521612,
-      "grad_norm": 0.02273356355726719,
-      "learning_rate": 8.071249431837655e-05,
-      "loss": 0.0071,
+      "epoch": 2.2723232323232323,
+      "grad_norm": 0.010940243490040302,
+      "learning_rate": 9.094437257438552e-05,
+      "loss": 0.0045,
       "step": 703
     },
     {
-      "epoch": 1.803714377201409,
-      "grad_norm": 0.017565427348017693,
-      "learning_rate": 8.064181645565596e-05,
-      "loss": 0.0093,
+      "epoch": 2.2755555555555556,
+      "grad_norm": 0.008155724965035915,
+      "learning_rate": 9.107373868046573e-05,
+      "loss": 0.0026,
       "step": 704
     },
     {
-      "epoch": 1.8062760166506564,
-      "grad_norm": 0.019855517894029617,
-      "learning_rate": 8.057104040460062e-05,
-      "loss": 0.0113,
+      "epoch": 2.278787878787879,
+      "grad_norm": 0.011424032971262932,
+      "learning_rate": 9.120310478654594e-05,
+      "loss": 0.0028,
       "step": 705
     },
     {
-      "epoch": 1.808837656099904,
-      "grad_norm": 0.014578789472579956,
-      "learning_rate": 8.050016639200458e-05,
-      "loss": 0.0068,
+      "epoch": 2.282020202020202,
+      "grad_norm": 0.00900313537567854,
+      "learning_rate": 9.133247089262613e-05,
+      "loss": 0.0032,
       "step": 706
     },
     {
-      "epoch": 1.8113992955491516,
-      "grad_norm": 0.02611469104886055,
-      "learning_rate": 8.042919464497584e-05,
-      "loss": 0.0173,
+      "epoch": 2.2852525252525253,
+      "grad_norm": 0.013425501994788647,
+      "learning_rate": 9.146183699870634e-05,
+      "loss": 0.0068,
       "step": 707
     },
     {
-      "epoch": 1.813960934998399,
-      "grad_norm": 0.017057769000530243,
-      "learning_rate": 8.035812539093557e-05,
-      "loss": 0.0088,
+      "epoch": 2.2884848484848486,
+      "grad_norm": 0.007079846691340208,
+      "learning_rate": 9.159120310478655e-05,
+      "loss": 0.0025,
       "step": 708
     },
     {
-      "epoch": 1.8165225744476465,
-      "grad_norm": 0.01934835873544216,
-      "learning_rate": 8.028695885761738e-05,
-      "loss": 0.0103,
+      "epoch": 2.291717171717172,
+      "grad_norm": 0.009416084736585617,
+      "learning_rate": 9.172056921086676e-05,
+      "loss": 0.0032,
       "step": 709
     },
     {
-      "epoch": 1.8190842138968941,
-      "grad_norm": 0.026960475370287895,
-      "learning_rate": 8.021569527306662e-05,
-      "loss": 0.0161,
+      "epoch": 2.294949494949495,
+      "grad_norm": 0.00754723371937871,
+      "learning_rate": 9.184993531694696e-05,
+      "loss": 0.0022,
       "step": 710
     },
     {
-      "epoch": 1.8216458533461415,
-      "grad_norm": 0.022418903186917305,
-      "learning_rate": 8.014433486563961e-05,
-      "loss": 0.0072,
+      "epoch": 2.2981818181818183,
+      "grad_norm": 0.0076440623961389065,
+      "learning_rate": 9.197930142302717e-05,
+      "loss": 0.0027,
       "step": 711
     },
     {
-      "epoch": 1.824207492795389,
-      "grad_norm": 0.017390521243214607,
-      "learning_rate": 8.007287786400295e-05,
-      "loss": 0.0063,
+      "epoch": 2.3014141414141416,
+      "grad_norm": 0.006314253434538841,
+      "learning_rate": 9.210866752910737e-05,
+      "loss": 0.002,
       "step": 712
     },
     {
-      "epoch": 1.8267691322446367,
-      "grad_norm": 0.018205352127552032,
-      "learning_rate": 8.000132449713272e-05,
-      "loss": 0.0089,
+      "epoch": 2.304646464646465,
+      "grad_norm": 0.009443306364119053,
+      "learning_rate": 9.223803363518758e-05,
+      "loss": 0.003,
       "step": 713
     },
     {
-      "epoch": 1.829330771693884,
-      "grad_norm": 0.025925058871507645,
-      "learning_rate": 7.992967499431386e-05,
-      "loss": 0.0182,
+      "epoch": 2.3078787878787876,
+      "grad_norm": 0.004945446737110615,
+      "learning_rate": 9.236739974126779e-05,
+      "loss": 0.0015,
       "step": 714
     },
     {
-      "epoch": 1.8318924111431316,
-      "grad_norm": 0.025157589465379715,
-      "learning_rate": 7.985792958513931e-05,
-      "loss": 0.0093,
+      "epoch": 2.311111111111111,
+      "grad_norm": 0.008110949769616127,
+      "learning_rate": 9.2496765847348e-05,
+      "loss": 0.0024,
       "step": 715
     },
     {
-      "epoch": 1.8344540505923792,
-      "grad_norm": 0.024524198845028877,
-      "learning_rate": 7.978608849950935e-05,
-      "loss": 0.0099,
+      "epoch": 2.314343434343434,
+      "grad_norm": 0.011624221689999104,
+      "learning_rate": 9.262613195342821e-05,
+      "loss": 0.0035,
       "step": 716
     },
     {
-      "epoch": 1.8370156900416266,
-      "grad_norm": 0.02669604681432247,
-      "learning_rate": 7.971415196763088e-05,
-      "loss": 0.0134,
+      "epoch": 2.3175757575757574,
+      "grad_norm": 0.006861218251287937,
+      "learning_rate": 9.275549805950842e-05,
+      "loss": 0.002,
       "step": 717
     },
     {
-      "epoch": 1.8395773294908742,
-      "grad_norm": 0.016144055873155594,
-      "learning_rate": 7.964212022001658e-05,
-      "loss": 0.0086,
+      "epoch": 2.3208080808080807,
+      "grad_norm": 0.00581625010818243,
+      "learning_rate": 9.288486416558863e-05,
+      "loss": 0.0017,
       "step": 718
     },
     {
-      "epoch": 1.8421389689401217,
-      "grad_norm": 0.019035059958696365,
-      "learning_rate": 7.956999348748427e-05,
-      "loss": 0.0089,
+      "epoch": 2.324040404040404,
+      "grad_norm": 0.009781553409993649,
+      "learning_rate": 9.301423027166883e-05,
+      "loss": 0.0027,
       "step": 719
     },
     {
-      "epoch": 1.8447006083893691,
-      "grad_norm": 0.026147911325097084,
-      "learning_rate": 7.949777200115616e-05,
-      "loss": 0.0095,
+      "epoch": 2.327272727272727,
+      "grad_norm": 0.007891007699072361,
+      "learning_rate": 9.314359637774904e-05,
+      "loss": 0.0026,
       "step": 720
     },
     {
-      "epoch": 1.8472622478386167,
-      "grad_norm": 0.01551744807511568,
-      "learning_rate": 7.942545599245805e-05,
-      "loss": 0.0071,
+      "epoch": 2.3305050505050504,
+      "grad_norm": 0.007546356413513422,
+      "learning_rate": 9.327296248382925e-05,
+      "loss": 0.0021,
       "step": 721
     },
     {
-      "epoch": 1.8498238872878643,
-      "grad_norm": 0.02032046765089035,
-      "learning_rate": 7.935304569311867e-05,
-      "loss": 0.0088,
+      "epoch": 2.3337373737373737,
+      "grad_norm": 0.00752625335007906,
+      "learning_rate": 9.340232858990944e-05,
+      "loss": 0.0033,
       "step": 722
     },
     {
-      "epoch": 1.8523855267371117,
-      "grad_norm": 0.014912649989128113,
-      "learning_rate": 7.928054133516884e-05,
-      "loss": 0.0062,
+      "epoch": 2.336969696969697,
+      "grad_norm": 0.012712947092950344,
+      "learning_rate": 9.353169469598965e-05,
+      "loss": 0.0075,
       "step": 723
     },
     {
-      "epoch": 1.8549471661863592,
-      "grad_norm": 0.021932363510131836,
-      "learning_rate": 7.920794315094084e-05,
-      "loss": 0.0129,
+      "epoch": 2.34020202020202,
+      "grad_norm": 0.006651640869677067,
+      "learning_rate": 9.366106080206986e-05,
+      "loss": 0.0022,
       "step": 724
     },
     {
-      "epoch": 1.8575088056356068,
-      "grad_norm": 0.023239847272634506,
-      "learning_rate": 7.913525137306756e-05,
-      "loss": 0.0105,
+      "epoch": 2.3434343434343434,
+      "grad_norm": 0.005778448656201363,
+      "learning_rate": 9.379042690815006e-05,
+      "loss": 0.0025,
       "step": 725
     },
     {
-      "epoch": 1.8575088056356068,
-      "eval_loss": 0.00722088199108839,
-      "eval_runtime": 15.8215,
-      "eval_samples_per_second": 6.321,
-      "eval_steps_per_second": 1.58,
+      "epoch": 2.3434343434343434,
+      "eval_loss": 0.002363896928727627,
+      "eval_runtime": 15.4937,
+      "eval_samples_per_second": 6.454,
+      "eval_steps_per_second": 1.614,
       "step": 725
     },
     {
-      "epoch": 1.8600704450848542,
-      "grad_norm": 0.014573454856872559,
-      "learning_rate": 7.906246623448183e-05,
-      "loss": 0.0066,
+      "epoch": 2.3466666666666667,
+      "grad_norm": 0.007105280179530382,
+      "learning_rate": 9.391979301423027e-05,
+      "loss": 0.0022,
       "step": 726
     },
     {
-      "epoch": 1.8626320845341018,
-      "grad_norm": 0.017306188121438026,
-      "learning_rate": 7.898958796841563e-05,
-      "loss": 0.007,
+      "epoch": 2.34989898989899,
+      "grad_norm": 0.006694236770272255,
+      "learning_rate": 9.404915912031048e-05,
+      "loss": 0.0027,
       "step": 727
     },
     {
-      "epoch": 1.8651937239833494,
-      "grad_norm": 0.02267426811158657,
-      "learning_rate": 7.891661680839932e-05,
-      "loss": 0.01,
+      "epoch": 2.353131313131313,
+      "grad_norm": 0.007940470241010189,
+      "learning_rate": 9.417852522639069e-05,
+      "loss": 0.0021,
       "step": 728
     },
     {
-      "epoch": 1.8677553634325967,
-      "grad_norm": 0.0263497456908226,
-      "learning_rate": 7.884355298826102e-05,
-      "loss": 0.0114,
+      "epoch": 2.3563636363636364,
+      "grad_norm": 0.011152379214763641,
+      "learning_rate": 9.43078913324709e-05,
+      "loss": 0.0033,
       "step": 729
     },
     {
-      "epoch": 1.8703170028818443,
-      "grad_norm": 0.020164979621767998,
-      "learning_rate": 7.877039674212569e-05,
-      "loss": 0.0081,
+      "epoch": 2.3595959595959597,
+      "grad_norm": 0.007020501885563135,
+      "learning_rate": 9.443725743855111e-05,
+      "loss": 0.0021,
       "step": 730
     },
     {
-      "epoch": 1.872878642331092,
-      "grad_norm": 0.013917827978730202,
-      "learning_rate": 7.869714830441448e-05,
-      "loss": 0.006,
+      "epoch": 2.362828282828283,
+      "grad_norm": 0.010429047048091888,
+      "learning_rate": 9.456662354463132e-05,
+      "loss": 0.002,
       "step": 731
     },
     {
-      "epoch": 1.8754402817803393,
-      "grad_norm": 0.027471866458654404,
-      "learning_rate": 7.862380790984396e-05,
-      "loss": 0.0115,
+      "epoch": 2.366060606060606,
+      "grad_norm": 0.006889528129249811,
+      "learning_rate": 9.469598965071152e-05,
+      "loss": 0.002,
       "step": 732
     },
     {
-      "epoch": 1.878001921229587,
-      "grad_norm": 0.022532496601343155,
-      "learning_rate": 7.855037579342539e-05,
-      "loss": 0.0094,
+      "epoch": 2.3692929292929295,
+      "grad_norm": 0.007243814878165722,
+      "learning_rate": 9.482535575679173e-05,
+      "loss": 0.0016,
       "step": 733
     },
     {
-      "epoch": 1.8805635606788345,
-      "grad_norm": 0.01607378013432026,
-      "learning_rate": 7.84768521904639e-05,
-      "loss": 0.0055,
+      "epoch": 2.3725252525252527,
+      "grad_norm": 0.010159321129322052,
+      "learning_rate": 9.495472186287193e-05,
+      "loss": 0.0032,
       "step": 734
     },
     {
-      "epoch": 1.8831252001280818,
-      "grad_norm": 0.016221843659877777,
-      "learning_rate": 7.840323733655778e-05,
-      "loss": 0.0075,
+      "epoch": 2.375757575757576,
+      "grad_norm": 0.008961687795817852,
+      "learning_rate": 9.508408796895214e-05,
+      "loss": 0.0028,
       "step": 735
     },
     {
-      "epoch": 1.8856868395773296,
-      "grad_norm": 0.03252291679382324,
-      "learning_rate": 7.832953146759781e-05,
-      "loss": 0.0171,
+      "epoch": 2.378989898989899,
+      "grad_norm": 0.007584400940686464,
+      "learning_rate": 9.521345407503235e-05,
+      "loss": 0.002,
       "step": 736
     },
     {
-      "epoch": 1.888248479026577,
-      "grad_norm": 0.021966032683849335,
-      "learning_rate": 7.825573481976629e-05,
-      "loss": 0.0082,
+      "epoch": 2.3822222222222225,
+      "grad_norm": 0.009180300869047642,
+      "learning_rate": 9.534282018111256e-05,
+      "loss": 0.0024,
       "step": 737
     },
     {
-      "epoch": 1.8908101184758244,
-      "grad_norm": 0.021348660811781883,
-      "learning_rate": 7.818184762953648e-05,
-      "loss": 0.0074,
+      "epoch": 2.3854545454545453,
+      "grad_norm": 0.0067786043509840965,
+      "learning_rate": 9.547218628719275e-05,
+      "loss": 0.0021,
       "step": 738
     },
     {
-      "epoch": 1.8933717579250722,
-      "grad_norm": 0.03429243341088295,
-      "learning_rate": 7.810787013367178e-05,
-      "loss": 0.0161,
+      "epoch": 2.3886868686868685,
+      "grad_norm": 0.014267507009208202,
+      "learning_rate": 9.560155239327296e-05,
+      "loss": 0.0053,
       "step": 739
     },
     {
-      "epoch": 1.8959333973743195,
-      "grad_norm": 0.019160069525241852,
-      "learning_rate": 7.803380256922495e-05,
-      "loss": 0.0096,
+      "epoch": 2.391919191919192,
+      "grad_norm": 0.010943619534373283,
+      "learning_rate": 9.573091849935317e-05,
+      "loss": 0.0031,
       "step": 740
     },
     {
-      "epoch": 1.8984950368235671,
-      "grad_norm": 0.019925393164157867,
-      "learning_rate": 7.795964517353735e-05,
-      "loss": 0.0073,
+      "epoch": 2.395151515151515,
+      "grad_norm": 0.009818903170526028,
+      "learning_rate": 9.586028460543338e-05,
+      "loss": 0.0021,
       "step": 741
     },
     {
-      "epoch": 1.9010566762728147,
-      "grad_norm": 0.02276492677628994,
-      "learning_rate": 7.788539818423822e-05,
-      "loss": 0.0084,
+      "epoch": 2.3983838383838383,
+      "grad_norm": 0.006286942400038242,
+      "learning_rate": 9.59896507115136e-05,
+      "loss": 0.0016,
       "step": 742
     },
     {
-      "epoch": 1.903618315722062,
-      "grad_norm": 0.025005770847201347,
-      "learning_rate": 7.781106183924385e-05,
-      "loss": 0.0095,
+      "epoch": 2.4016161616161615,
+      "grad_norm": 0.00991823710501194,
+      "learning_rate": 9.61190168175938e-05,
+      "loss": 0.003,
       "step": 743
     },
     {
-      "epoch": 1.9061799551713097,
-      "grad_norm": 0.0237042848020792,
-      "learning_rate": 7.773663637675694e-05,
-      "loss": 0.0115,
+      "epoch": 2.404848484848485,
+      "grad_norm": 0.008815119974315166,
+      "learning_rate": 9.6248382923674e-05,
+      "loss": 0.0027,
       "step": 744
     },
     {
-      "epoch": 1.9087415946205573,
-      "grad_norm": 0.02170964702963829,
-      "learning_rate": 7.76621220352657e-05,
-      "loss": 0.0112,
+      "epoch": 2.408080808080808,
+      "grad_norm": 0.008804360404610634,
+      "learning_rate": 9.63777490297542e-05,
+      "loss": 0.003,
       "step": 745
     },
     {
-      "epoch": 1.9113032340698046,
-      "grad_norm": 0.01486461516469717,
-      "learning_rate": 7.758751905354311e-05,
-      "loss": 0.0058,
+      "epoch": 2.4113131313131313,
+      "grad_norm": 0.006885518319904804,
+      "learning_rate": 9.650711513583441e-05,
+      "loss": 0.002,
       "step": 746
     },
     {
-      "epoch": 1.9138648735190522,
-      "grad_norm": 0.018541356548666954,
-      "learning_rate": 7.751282767064627e-05,
-      "loss": 0.0087,
+      "epoch": 2.4145454545454546,
+      "grad_norm": 0.008454122580587864,
+      "learning_rate": 9.663648124191462e-05,
+      "loss": 0.0029,
       "step": 747
     },
     {
-      "epoch": 1.9164265129682998,
-      "grad_norm": 0.016604872420430183,
-      "learning_rate": 7.74380481259155e-05,
-      "loss": 0.0066,
+      "epoch": 2.417777777777778,
+      "grad_norm": 0.009298015385866165,
+      "learning_rate": 9.676584734799483e-05,
+      "loss": 0.0022,
       "step": 748
     },
     {
-      "epoch": 1.9189881524175472,
-      "grad_norm": 0.022073280066251755,
-      "learning_rate": 7.736318065897362e-05,
-      "loss": 0.0114,
+      "epoch": 2.421010101010101,
+      "grad_norm": 0.0124392444267869,
+      "learning_rate": 9.689521345407504e-05,
+      "loss": 0.0033,
       "step": 749
     },
     {
-      "epoch": 1.9215497918667948,
-      "grad_norm": 0.021899666637182236,
-      "learning_rate": 7.728822550972523e-05,
-      "loss": 0.0089,
+      "epoch": 2.4242424242424243,
+      "grad_norm": 0.008510610088706017,
+      "learning_rate": 9.702457956015525e-05,
+      "loss": 0.0031,
       "step": 750
     },
     {
-      "epoch": 1.9215497918667948,
-      "eval_loss": 0.0077704195864498615,
-      "eval_runtime": 15.7892,
-      "eval_samples_per_second": 6.333,
-      "eval_steps_per_second": 1.583,
+      "epoch": 2.4242424242424243,
+      "eval_loss": 0.0023907506838440895,
+      "eval_runtime": 15.488,
+      "eval_samples_per_second": 6.457,
+      "eval_steps_per_second": 1.614,
       "step": 750
     },
     {
-      "epoch": 1.9241114313160423,
-      "grad_norm": 0.024171432480216026,
-      "learning_rate": 7.721318291835586e-05,
-      "loss": 0.0084,
+      "epoch": 2.4274747474747476,
+      "grad_norm": 0.005241287872195244,
+      "learning_rate": 9.715394566623545e-05,
+      "loss": 0.0018,
       "step": 751
     },
     {
-      "epoch": 1.9266730707652897,
-      "grad_norm": 0.027852538973093033,
-      "learning_rate": 7.713805312533126e-05,
-      "loss": 0.0116,
+      "epoch": 2.430707070707071,
+      "grad_norm": 0.00899229571223259,
+      "learning_rate": 9.728331177231566e-05,
+      "loss": 0.003,
       "step": 752
     },
     {
-      "epoch": 1.9292347102145373,
-      "grad_norm": 0.02583443932235241,
-      "learning_rate": 7.706283637139658e-05,
-      "loss": 0.01,
+      "epoch": 2.433939393939394,
+      "grad_norm": 0.007302114740014076,
+      "learning_rate": 9.741267787839587e-05,
+      "loss": 0.0019,
       "step": 753
     },
     {
-      "epoch": 1.9317963496637849,
-      "grad_norm": 0.022306324914097786,
-      "learning_rate": 7.698753289757565e-05,
-      "loss": 0.0111,
+      "epoch": 2.4371717171717173,
+      "grad_norm": 0.008491559885442257,
+      "learning_rate": 9.754204398447608e-05,
+      "loss": 0.0032,
       "step": 754
     },
     {
-      "epoch": 1.9343579891130323,
-      "grad_norm": 0.023564163595438004,
-      "learning_rate": 7.69121429451702e-05,
-      "loss": 0.0094,
+      "epoch": 2.4404040404040406,
+      "grad_norm": 0.008841630071401596,
+      "learning_rate": 9.767141009055629e-05,
+      "loss": 0.0028,
       "step": 755
     },
     {
-      "epoch": 1.9369196285622798,
-      "grad_norm": 0.02647082880139351,
-      "learning_rate": 7.683666675575901e-05,
-      "loss": 0.0131,
+      "epoch": 2.443636363636364,
+      "grad_norm": 0.00954421330243349,
+      "learning_rate": 9.780077619663648e-05,
+      "loss": 0.003,
       "step": 756
     },
     {
-      "epoch": 1.9394812680115274,
-      "grad_norm": 0.021390896290540695,
-      "learning_rate": 7.676110457119727e-05,
-      "loss": 0.0073,
+      "epoch": 2.4468686868686866,
+      "grad_norm": 0.006806389894336462,
+      "learning_rate": 9.793014230271668e-05,
+      "loss": 0.0021,
       "step": 757
     },
     {
-      "epoch": 1.9420429074607748,
-      "grad_norm": 0.022772209718823433,
-      "learning_rate": 7.66854566336157e-05,
-      "loss": 0.0107,
+      "epoch": 2.45010101010101,
+      "grad_norm": 0.007156576961278915,
+      "learning_rate": 9.805950840879689e-05,
+      "loss": 0.0023,
       "step": 758
     },
     {
-      "epoch": 1.9446045469100224,
-      "grad_norm": 0.021806808188557625,
-      "learning_rate": 7.660972318541982e-05,
-      "loss": 0.0104,
+      "epoch": 2.453333333333333,
+      "grad_norm": 0.0072568911127746105,
+      "learning_rate": 9.81888745148771e-05,
+      "loss": 0.0026,
       "step": 759
     },
     {
-      "epoch": 1.94716618635927,
-      "grad_norm": 0.03225119784474373,
-      "learning_rate": 7.653390446928909e-05,
-      "loss": 0.0206,
+      "epoch": 2.4565656565656564,
+      "grad_norm": 0.006635196506977081,
+      "learning_rate": 9.831824062095731e-05,
+      "loss": 0.0024,
       "step": 760
     },
     {
-      "epoch": 1.9497278258085173,
-      "grad_norm": 0.03250499814748764,
-      "learning_rate": 7.64580007281763e-05,
-      "loss": 0.0158,
+      "epoch": 2.4597979797979797,
+      "grad_norm": 0.005860018078237772,
+      "learning_rate": 9.844760672703752e-05,
+      "loss": 0.0024,
       "step": 761
     },
     {
-      "epoch": 1.952289465257765,
-      "grad_norm": 0.023754483088850975,
-      "learning_rate": 7.638201220530665e-05,
-      "loss": 0.0119,
+      "epoch": 2.463030303030303,
+      "grad_norm": 0.007157980464398861,
+      "learning_rate": 9.857697283311773e-05,
+      "loss": 0.0021,
       "step": 762
     },
     {
-      "epoch": 1.9548511047070125,
-      "grad_norm": 0.014481264166533947,
-      "learning_rate": 7.630593914417699e-05,
-      "loss": 0.0063,
+      "epoch": 2.466262626262626,
+      "grad_norm": 0.006852725055068731,
+      "learning_rate": 9.870633893919794e-05,
+      "loss": 0.0024,
       "step": 763
     },
     {
-      "epoch": 1.9574127441562599,
-      "grad_norm": 0.02314913645386696,
-      "learning_rate": 7.622978178855511e-05,
-      "loss": 0.0126,
+      "epoch": 2.4694949494949494,
+      "grad_norm": 0.007933683693408966,
+      "learning_rate": 9.883570504527814e-05,
+      "loss": 0.0025,
       "step": 764
     },
     {
-      "epoch": 1.9599743836055077,
-      "grad_norm": 0.025454895570874214,
-      "learning_rate": 7.615354038247888e-05,
-      "loss": 0.0097,
+      "epoch": 2.4727272727272727,
+      "grad_norm": 0.006870505400002003,
+      "learning_rate": 9.896507115135835e-05,
+      "loss": 0.0025,
       "step": 765
     },
     {
-      "epoch": 1.962536023054755,
-      "grad_norm": 0.02636459842324257,
-      "learning_rate": 7.607721517025552e-05,
-      "loss": 0.0077,
+      "epoch": 2.475959595959596,
+      "grad_norm": 0.01359974592924118,
+      "learning_rate": 9.909443725743856e-05,
+      "loss": 0.0047,
       "step": 766
     },
     {
-      "epoch": 1.9650976625040024,
-      "grad_norm": 0.024306008592247963,
-      "learning_rate": 7.600080639646077e-05,
-      "loss": 0.0096,
+      "epoch": 2.479191919191919,
+      "grad_norm": 0.006239316891878843,
+      "learning_rate": 9.922380336351877e-05,
+      "loss": 0.0019,
       "step": 767
     },
     {
-      "epoch": 1.9676593019532502,
-      "grad_norm": 0.017479566857218742,
-      "learning_rate": 7.592431430593818e-05,
-      "loss": 0.0062,
+      "epoch": 2.4824242424242424,
+      "grad_norm": 0.008975119329988956,
+      "learning_rate": 9.935316946959897e-05,
+      "loss": 0.0028,
       "step": 768
     },
     {
-      "epoch": 1.9702209414024976,
-      "grad_norm": 0.023251403123140335,
-      "learning_rate": 7.584773914379823e-05,
-      "loss": 0.0125,
+      "epoch": 2.4856565656565657,
+      "grad_norm": 0.005580460652709007,
+      "learning_rate": 9.948253557567918e-05,
+      "loss": 0.002,
       "step": 769
     },
     {
-      "epoch": 1.972782580851745,
-      "grad_norm": 0.019164303317666054,
-      "learning_rate": 7.577108115541761e-05,
-      "loss": 0.009,
+      "epoch": 2.488888888888889,
+      "grad_norm": 0.007628625258803368,
+      "learning_rate": 9.961190168175937e-05,
+      "loss": 0.0021,
       "step": 770
     },
     {
-      "epoch": 1.9753442203009928,
-      "grad_norm": 0.016869869083166122,
-      "learning_rate": 7.569434058643844e-05,
-      "loss": 0.0085,
+      "epoch": 2.492121212121212,
+      "grad_norm": 0.012194107286632061,
+      "learning_rate": 9.974126778783958e-05,
+      "loss": 0.0043,
       "step": 771
     },
     {
-      "epoch": 1.9779058597502401,
-      "grad_norm": 0.02194080501794815,
-      "learning_rate": 7.561751768276744e-05,
-      "loss": 0.0107,
+      "epoch": 2.4953535353535354,
+      "grad_norm": 0.008045129477977753,
+      "learning_rate": 9.98706338939198e-05,
+      "loss": 0.0023,
       "step": 772
     },
     {
-      "epoch": 1.9804674991994877,
-      "grad_norm": 0.019984381273388863,
-      "learning_rate": 7.554061269057515e-05,
-      "loss": 0.0098,
+      "epoch": 2.4985858585858587,
+      "grad_norm": 0.005020644515752792,
+      "learning_rate": 0.0001,
+      "loss": 0.0017,
       "step": 773
     },
     {
-      "epoch": 1.9830291386487353,
-      "grad_norm": 0.013338621705770493,
-      "learning_rate": 7.546362585629518e-05,
-      "loss": 0.0052,
+      "epoch": 2.501818181818182,
+      "grad_norm": 0.008611057884991169,
+      "learning_rate": 9.999999489471233e-05,
+      "loss": 0.0032,
       "step": 774
     },
     {
-      "epoch": 1.9855907780979827,
-      "grad_norm": 0.028268542140722275,
-      "learning_rate": 7.53865574266234e-05,
-      "loss": 0.0132,
+      "epoch": 2.505050505050505,
+      "grad_norm": 0.008335341699421406,
+      "learning_rate": 9.99999795788503e-05,
+      "loss": 0.0027,
       "step": 775
     },
     {
-      "epoch": 1.9855907780979827,
-      "eval_loss": 0.008421124890446663,
-      "eval_runtime": 15.7765,
-      "eval_samples_per_second": 6.339,
-      "eval_steps_per_second": 1.585,
+      "epoch": 2.505050505050505,
+      "eval_loss": 0.0022718999534845352,
+      "eval_runtime": 15.5876,
+      "eval_samples_per_second": 6.415,
+      "eval_steps_per_second": 1.604,
       "step": 775
     },
     {
-      "epoch": 1.9881524175472303,
-      "grad_norm": 0.015591204166412354,
-      "learning_rate": 7.530940764851708e-05,
-      "loss": 0.0083,
+      "epoch": 2.5082828282828284,
+      "grad_norm": 0.01257743127644062,
+      "learning_rate": 9.99999540524171e-05,
+      "loss": 0.0048,
       "step": 776
     },
     {
-      "epoch": 1.9907140569964779,
-      "grad_norm": 0.01851005293428898,
-      "learning_rate": 7.523217676919427e-05,
-      "loss": 0.0078,
+      "epoch": 2.5115151515151517,
+      "grad_norm": 0.009661512449383736,
+      "learning_rate": 9.999991831541789e-05,
+      "loss": 0.0035,
       "step": 777
     },
     {
-      "epoch": 1.9932756964457252,
-      "grad_norm": 0.014990904368460178,
-      "learning_rate": 7.515486503613278e-05,
-      "loss": 0.0061,
+      "epoch": 2.514747474747475,
+      "grad_norm": 0.0056405263021588326,
+      "learning_rate": 9.999987236786e-05,
+      "loss": 0.0017,
       "step": 778
     },
     {
-      "epoch": 1.9958373358949728,
-      "grad_norm": 0.018165841698646545,
-      "learning_rate": 7.507747269706962e-05,
-      "loss": 0.0073,
+      "epoch": 2.517979797979798,
+      "grad_norm": 0.01253304723650217,
+      "learning_rate": 9.999981620975281e-05,
+      "loss": 0.004,
       "step": 779
     },
     {
-      "epoch": 1.9983989753442204,
-      "grad_norm": 0.021395908668637276,
-      "learning_rate": 7.500000000000001e-05,
-      "loss": 0.0084,
+      "epoch": 2.5212121212121215,
+      "grad_norm": 0.00832959171384573,
+      "learning_rate": 9.999974984110779e-05,
+      "loss": 0.0029,
       "step": 780
     },
     {
-      "epoch": 2.0012808197246237,
-      "grad_norm": 0.12048190832138062,
-      "learning_rate": 7.492244719317668e-05,
-      "loss": 0.0199,
+      "epoch": 2.5244444444444447,
+      "grad_norm": 0.007301663514226675,
+      "learning_rate": 9.999967326193847e-05,
+      "loss": 0.0022,
       "step": 781
     },
     {
-      "epoch": 2.0038424591738715,
-      "grad_norm": 0.027717135846614838,
-      "learning_rate": 7.484481452510913e-05,
-      "loss": 0.0102,
+      "epoch": 2.5276767676767675,
+      "grad_norm": 0.00891867559403181,
+      "learning_rate": 9.999958647226049e-05,
+      "loss": 0.0026,
       "step": 782
     },
     {
-      "epoch": 2.006404098623119,
-      "grad_norm": 0.036011483520269394,
-      "learning_rate": 7.476710224456267e-05,
-      "loss": 0.0092,
+      "epoch": 2.5309090909090908,
+      "grad_norm": 0.00586441857740283,
+      "learning_rate": 9.999948947209162e-05,
+      "loss": 0.0015,
       "step": 783
     },
     {
-      "epoch": 2.0089657380723662,
-      "grad_norm": 0.030577300116419792,
-      "learning_rate": 7.468931060055777e-05,
-      "loss": 0.0082,
+      "epoch": 2.534141414141414,
+      "grad_norm": 0.008907037787139416,
+      "learning_rate": 9.999938226145161e-05,
+      "loss": 0.0032,
       "step": 784
     },
     {
-      "epoch": 2.011527377521614,
-      "grad_norm": 0.033715181052684784,
-      "learning_rate": 7.461143984236924e-05,
-      "loss": 0.0123,
+      "epoch": 2.5373737373737373,
+      "grad_norm": 0.005364672746509314,
+      "learning_rate": 9.999926484036237e-05,
+      "loss": 0.002,
       "step": 785
     },
     {
-      "epoch": 2.0140890169708614,
-      "grad_norm": 0.01578759402036667,
-      "learning_rate": 7.453349021952533e-05,
-      "loss": 0.0049,
+      "epoch": 2.5406060606060605,
+      "grad_norm": 0.00744849955663085,
+      "learning_rate": 9.999913720884791e-05,
+      "loss": 0.0022,
       "step": 786
     },
     {
-      "epoch": 2.0166506564201088,
-      "grad_norm": 0.017665740102529526,
-      "learning_rate": 7.44554619818071e-05,
-      "loss": 0.0078,
+      "epoch": 2.543838383838384,
+      "grad_norm": 0.011406635865569115,
+      "learning_rate": 9.999899936693426e-05,
+      "loss": 0.0044,
       "step": 787
     },
     {
-      "epoch": 2.0192122958693566,
-      "grad_norm": 0.0218870397657156,
-      "learning_rate": 7.437735537924742e-05,
-      "loss": 0.0073,
+      "epoch": 2.547070707070707,
+      "grad_norm": 0.01190891582518816,
+      "learning_rate": 9.99988513146496e-05,
+      "loss": 0.0047,
       "step": 788
     },
     {
-      "epoch": 2.021773935318604,
-      "grad_norm": 0.02893979661166668,
-      "learning_rate": 7.42991706621303e-05,
-      "loss": 0.01,
+      "epoch": 2.5503030303030303,
+      "grad_norm": 0.006621699780225754,
+      "learning_rate": 9.999869305202412e-05,
+      "loss": 0.0018,
       "step": 789
     },
     {
-      "epoch": 2.0243355747678513,
-      "grad_norm": 0.03510141000151634,
-      "learning_rate": 7.422090808099014e-05,
-      "loss": 0.0101,
+      "epoch": 2.5535353535353535,
+      "grad_norm": 0.006613705772906542,
+      "learning_rate": 9.999852457909018e-05,
+      "loss": 0.0016,
       "step": 790
     },
     {
-      "epoch": 2.026897214217099,
-      "grad_norm": 0.020581314340233803,
-      "learning_rate": 7.414256788661074e-05,
-      "loss": 0.0069,
+      "epoch": 2.556767676767677,
+      "grad_norm": 0.00954991951584816,
+      "learning_rate": 9.999834589588217e-05,
+      "loss": 0.0035,
       "step": 791
     },
     {
-      "epoch": 2.0294588536663465,
-      "grad_norm": 0.025523610413074493,
-      "learning_rate": 7.406415033002463e-05,
-      "loss": 0.0094,
+      "epoch": 2.56,
+      "grad_norm": 0.006031880155205727,
+      "learning_rate": 9.999815700243656e-05,
+      "loss": 0.0014,
       "step": 792
     },
     {
-      "epoch": 2.032020493115594,
-      "grad_norm": 0.017144009470939636,
-      "learning_rate": 7.398565566251232e-05,
-      "loss": 0.005,
+      "epoch": 2.5632323232323233,
+      "grad_norm": 0.008622254244983196,
+      "learning_rate": 9.999795789879196e-05,
+      "loss": 0.003,
       "step": 793
     },
     {
-      "epoch": 2.0345821325648417,
-      "grad_norm": 0.02107195369899273,
-      "learning_rate": 7.390708413560129e-05,
-      "loss": 0.0072,
+      "epoch": 2.5664646464646466,
+      "grad_norm": 0.00705391401425004,
+      "learning_rate": 9.9997748584989e-05,
+      "loss": 0.0021,
       "step": 794
     },
     {
-      "epoch": 2.037143772014089,
-      "grad_norm": 0.012858357280492783,
-      "learning_rate": 7.38284360010654e-05,
-      "loss": 0.0036,
+      "epoch": 2.56969696969697,
+      "grad_norm": 0.006877865642309189,
+      "learning_rate": 9.999752906107042e-05,
+      "loss": 0.0018,
       "step": 795
     },
     {
-      "epoch": 2.0397054114633364,
-      "grad_norm": 0.04638303071260452,
-      "learning_rate": 7.374971151092395e-05,
-      "loss": 0.0129,
+      "epoch": 2.572929292929293,
+      "grad_norm": 0.009295830503106117,
+      "learning_rate": 9.999729932708109e-05,
+      "loss": 0.003,
       "step": 796
     },
     {
-      "epoch": 2.042267050912584,
-      "grad_norm": 0.03213483840227127,
-      "learning_rate": 7.367091091744089e-05,
-      "loss": 0.0083,
+      "epoch": 2.5761616161616163,
+      "grad_norm": 0.007250424008816481,
+      "learning_rate": 9.999705938306789e-05,
+      "loss": 0.0022,
       "step": 797
     },
     {
-      "epoch": 2.0448286903618316,
-      "grad_norm": 0.021228497847914696,
-      "learning_rate": 7.35920344731241e-05,
+      "epoch": 2.579393939393939,
+      "grad_norm": 0.01407887414097786,
+      "learning_rate": 9.999680922907982e-05,
       "loss": 0.0069,
       "step": 798
     },
     {
-      "epoch": 2.047390329811079,
-      "grad_norm": 0.037191636860370636,
-      "learning_rate": 7.35130824307245e-05,
-      "loss": 0.0097,
+      "epoch": 2.5826262626262624,
+      "grad_norm": 0.010157939046621323,
+      "learning_rate": 9.999654886516798e-05,
+      "loss": 0.003,
       "step": 799
     },
     {
-      "epoch": 2.0499519692603267,
-      "grad_norm": 0.026018405333161354,
-      "learning_rate": 7.343405504323519e-05,
-      "loss": 0.0104,
+      "epoch": 2.5858585858585856,
+      "grad_norm": 0.006922993343323469,
+      "learning_rate": 9.999627829138554e-05,
+      "loss": 0.0024,
       "step": 800
     },
     {
-      "epoch": 2.0499519692603267,
-      "eval_loss": 0.009080244228243828,
-      "eval_runtime": 15.7393,
-      "eval_samples_per_second": 6.354,
-      "eval_steps_per_second": 1.588,
+      "epoch": 2.5858585858585856,
+      "eval_loss": 0.0022230970207601786,
+      "eval_runtime": 15.5011,
+      "eval_samples_per_second": 6.451,
+      "eval_steps_per_second": 1.613,
       "step": 800
     },
     {
-      "epoch": 2.052513608709574,
-      "grad_norm": 0.02520827390253544,
-      "learning_rate": 7.335495256389078e-05,
-      "loss": 0.0063,
+      "epoch": 2.589090909090909,
+      "grad_norm": 0.010659018531441689,
+      "learning_rate": 9.999599750778772e-05,
+      "loss": 0.0039,
       "step": 801
     },
     {
-      "epoch": 2.0550752481588215,
-      "grad_norm": 0.03224215283989906,
-      "learning_rate": 7.327577524616646e-05,
-      "loss": 0.0122,
+      "epoch": 2.592323232323232,
+      "grad_norm": 0.006377291399985552,
+      "learning_rate": 9.999570651443191e-05,
+      "loss": 0.002,
       "step": 802
     },
     {
-      "epoch": 2.0576368876080693,
-      "grad_norm": 0.02962084673345089,
-      "learning_rate": 7.319652334377725e-05,
-      "loss": 0.0105,
+      "epoch": 2.5955555555555554,
+      "grad_norm": 0.005982244852930307,
+      "learning_rate": 9.99954053113775e-05,
+      "loss": 0.0022,
       "step": 803
     },
     {
-      "epoch": 2.0601985270573167,
-      "grad_norm": 0.019580401480197906,
-      "learning_rate": 7.311719711067717e-05,
-      "loss": 0.0062,
+      "epoch": 2.5987878787878786,
+      "grad_norm": 0.007271333131939173,
+      "learning_rate": 9.9995093898686e-05,
+      "loss": 0.0025,
       "step": 804
     },
     {
-      "epoch": 2.062760166506564,
-      "grad_norm": 0.03741897642612457,
-      "learning_rate": 7.303779680105843e-05,
-      "loss": 0.0133,
+      "epoch": 2.602020202020202,
+      "grad_norm": 0.011588825844228268,
+      "learning_rate": 9.999477227642103e-05,
+      "loss": 0.0034,
       "step": 805
     },
     {
-      "epoch": 2.065321805955812,
-      "grad_norm": 0.02220587432384491,
-      "learning_rate": 7.295832266935059e-05,
-      "loss": 0.0096,
+      "epoch": 2.605252525252525,
+      "grad_norm": 0.006561601534485817,
+      "learning_rate": 9.999444044464823e-05,
+      "loss": 0.0019,
       "step": 806
     },
     {
-      "epoch": 2.067883445405059,
-      "grad_norm": 0.020196085795760155,
-      "learning_rate": 7.287877497021978e-05,
-      "loss": 0.01,
+      "epoch": 2.6084848484848484,
+      "grad_norm": 0.00749508012086153,
+      "learning_rate": 9.999409840343539e-05,
+      "loss": 0.0027,
       "step": 807
     },
     {
-      "epoch": 2.0704450848543066,
-      "grad_norm": 0.014655430801212788,
-      "learning_rate": 7.279915395856782e-05,
-      "loss": 0.0047,
+      "epoch": 2.6117171717171717,
+      "grad_norm": 0.010126384906470776,
+      "learning_rate": 9.999374615285236e-05,
+      "loss": 0.0033,
       "step": 808
     },
     {
-      "epoch": 2.0730067243035544,
-      "grad_norm": 0.026845693588256836,
-      "learning_rate": 7.271945988953157e-05,
-      "loss": 0.0089,
+      "epoch": 2.614949494949495,
+      "grad_norm": 0.012112539261579514,
+      "learning_rate": 9.999338369297106e-05,
+      "loss": 0.0036,
       "step": 809
     },
     {
-      "epoch": 2.0755683637528017,
-      "grad_norm": 0.024815455079078674,
-      "learning_rate": 7.263969301848188e-05,
-      "loss": 0.0069,
+      "epoch": 2.618181818181818,
+      "grad_norm": 0.009163744747638702,
+      "learning_rate": 9.999301102386553e-05,
+      "loss": 0.0035,
       "step": 810
     },
     {
-      "epoch": 2.078130003202049,
-      "grad_norm": 0.026393510401248932,
-      "learning_rate": 7.255985360102291e-05,
-      "loss": 0.0083,
+      "epoch": 2.6214141414141414,
+      "grad_norm": 0.007010017987340689,
+      "learning_rate": 9.999262814561185e-05,
+      "loss": 0.0023,
       "step": 811
     },
     {
-      "epoch": 2.080691642651297,
-      "grad_norm": 0.015269361436367035,
-      "learning_rate": 7.247994189299132e-05,
-      "loss": 0.0045,
+      "epoch": 2.6246464646464647,
+      "grad_norm": 0.01088892575353384,
+      "learning_rate": 9.999223505828821e-05,
+      "loss": 0.0033,
       "step": 812
     },
     {
-      "epoch": 2.0832532821005443,
-      "grad_norm": 0.020210297778248787,
-      "learning_rate": 7.239995815045542e-05,
-      "loss": 0.008,
+      "epoch": 2.627878787878788,
+      "grad_norm": 0.006594639737159014,
+      "learning_rate": 9.999183176197491e-05,
+      "loss": 0.0018,
       "step": 813
     },
     {
-      "epoch": 2.085814921549792,
-      "grad_norm": 0.027442224323749542,
-      "learning_rate": 7.231990262971428e-05,
-      "loss": 0.0121,
+      "epoch": 2.631111111111111,
+      "grad_norm": 0.010258231312036514,
+      "learning_rate": 9.999141825675426e-05,
+      "loss": 0.003,
       "step": 814
     },
     {
-      "epoch": 2.0883765609990395,
-      "grad_norm": 0.023403724655508995,
-      "learning_rate": 7.223977558729706e-05,
-      "loss": 0.0091,
+      "epoch": 2.6343434343434344,
+      "grad_norm": 0.005483047105371952,
+      "learning_rate": 9.999099454271074e-05,
+      "loss": 0.0019,
       "step": 815
     },
     {
-      "epoch": 2.090938200448287,
-      "grad_norm": 0.034363970160484314,
-      "learning_rate": 7.215957727996207e-05,
-      "loss": 0.0104,
+      "epoch": 2.6375757575757577,
+      "grad_norm": 0.010156993754208088,
+      "learning_rate": 9.999056061993089e-05,
+      "loss": 0.0056,
       "step": 816
     },
     {
-      "epoch": 2.0934998398975346,
-      "grad_norm": 0.0232424046844244,
-      "learning_rate": 7.207930796469595e-05,
-      "loss": 0.0088,
+      "epoch": 2.640808080808081,
+      "grad_norm": 0.006410804111510515,
+      "learning_rate": 9.999011648850329e-05,
+      "loss": 0.0016,
       "step": 817
     },
     {
-      "epoch": 2.096061479346782,
-      "grad_norm": 0.02351643331348896,
-      "learning_rate": 7.19989678987129e-05,
-      "loss": 0.0128,
+      "epoch": 2.644040404040404,
+      "grad_norm": 0.009594564326107502,
+      "learning_rate": 9.998966214851864e-05,
+      "loss": 0.004,
       "step": 818
     },
     {
-      "epoch": 2.0986231187960294,
-      "grad_norm": 0.02835538238286972,
-      "learning_rate": 7.191855733945387e-05,
-      "loss": 0.0108,
+      "epoch": 2.6472727272727274,
+      "grad_norm": 0.013452129438519478,
+      "learning_rate": 9.998919760006972e-05,
+      "loss": 0.0044,
       "step": 819
     },
     {
-      "epoch": 2.101184758245277,
-      "grad_norm": 0.019840311259031296,
-      "learning_rate": 7.183807654458564e-05,
-      "loss": 0.0064,
+      "epoch": 2.6505050505050507,
+      "grad_norm": 0.0037844020407646894,
+      "learning_rate": 9.998872284325142e-05,
+      "loss": 0.0011,
       "step": 820
     },
     {
-      "epoch": 2.1037463976945245,
-      "grad_norm": 0.022916074842214584,
-      "learning_rate": 7.175752577200007e-05,
-      "loss": 0.009,
+      "epoch": 2.653737373737374,
+      "grad_norm": 0.008337401784956455,
+      "learning_rate": 9.998823787816066e-05,
+      "loss": 0.0023,
       "step": 821
     },
     {
-      "epoch": 2.106308037143772,
-      "grad_norm": 0.019078776240348816,
-      "learning_rate": 7.167690527981328e-05,
-      "loss": 0.0058,
+      "epoch": 2.656969696969697,
+      "grad_norm": 0.009283684194087982,
+      "learning_rate": 9.99877427048965e-05,
+      "loss": 0.0028,
       "step": 822
     },
     {
-      "epoch": 2.1088696765930197,
-      "grad_norm": 0.018315337598323822,
-      "learning_rate": 7.159621532636477e-05,
-      "loss": 0.006,
+      "epoch": 2.6602020202020205,
+      "grad_norm": 0.006520859897136688,
+      "learning_rate": 9.998723732356006e-05,
+      "loss": 0.0017,
       "step": 823
     },
     {
-      "epoch": 2.111431316042267,
-      "grad_norm": 0.027668718248605728,
-      "learning_rate": 7.151545617021663e-05,
-      "loss": 0.0109,
+      "epoch": 2.6634343434343437,
+      "grad_norm": 0.006291633006185293,
+      "learning_rate": 9.998672173425452e-05,
+      "loss": 0.0016,
       "step": 824
     },
     {
-      "epoch": 2.1139929554915144,
-      "grad_norm": 0.025120360776782036,
-      "learning_rate": 7.143462807015271e-05,
-      "loss": 0.0084,
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.011474277824163437,
+      "learning_rate": 9.998619593708518e-05,
+      "loss": 0.0027,
       "step": 825
     },
     {
-      "epoch": 2.1139929554915144,
-      "eval_loss": 0.008172542788088322,
-      "eval_runtime": 15.7492,
-      "eval_samples_per_second": 6.35,
-      "eval_steps_per_second": 1.587,
+      "epoch": 2.6666666666666665,
+      "eval_loss": 0.0022249266039580107,
+      "eval_runtime": 15.4934,
+      "eval_samples_per_second": 6.454,
+      "eval_steps_per_second": 1.614,
       "step": 825
     },
     {
-      "epoch": 2.1165545949407623,
-      "grad_norm": 0.01747150532901287,
-      "learning_rate": 7.135373128517778e-05,
-      "loss": 0.0066,
+      "epoch": 2.6698989898989898,
+      "grad_norm": 0.007792735006660223,
+      "learning_rate": 9.998565993215943e-05,
+      "loss": 0.0025,
       "step": 826
     },
     {
-      "epoch": 2.1191162343900096,
-      "grad_norm": 0.024094821885228157,
-      "learning_rate": 7.12727660745167e-05,
-      "loss": 0.0124,
+      "epoch": 2.673131313131313,
+      "grad_norm": 0.006040231324732304,
+      "learning_rate": 9.998511371958672e-05,
+      "loss": 0.0021,
       "step": 827
     },
     {
-      "epoch": 2.121677873839257,
-      "grad_norm": 0.03347666934132576,
-      "learning_rate": 7.119173269761358e-05,
-      "loss": 0.0149,
+      "epoch": 2.6763636363636363,
+      "grad_norm": 0.009344964288175106,
+      "learning_rate": 9.998455729947858e-05,
+      "loss": 0.0024,
       "step": 828
     },
     {
-      "epoch": 2.124239513288505,
-      "grad_norm": 0.01727481372654438,
-      "learning_rate": 7.111063141413098e-05,
-      "loss": 0.0047,
+      "epoch": 2.6795959595959595,
+      "grad_norm": 0.009222546592354774,
+      "learning_rate": 9.998399067194864e-05,
+      "loss": 0.0033,
       "step": 829
     },
     {
-      "epoch": 2.126801152737752,
-      "grad_norm": 0.052980441600084305,
-      "learning_rate": 7.102946248394909e-05,
-      "loss": 0.0136,
+      "epoch": 2.682828282828283,
+      "grad_norm": 0.004181350115686655,
+      "learning_rate": 9.998341383711263e-05,
+      "loss": 0.0012,
       "step": 830
     },
     {
-      "epoch": 2.1293627921869995,
-      "grad_norm": 0.02497376874089241,
-      "learning_rate": 7.094822616716476e-05,
-      "loss": 0.0096,
+      "epoch": 2.686060606060606,
+      "grad_norm": 0.007057964336127043,
+      "learning_rate": 9.998282679508835e-05,
+      "loss": 0.0024,
       "step": 831
     },
     {
-      "epoch": 2.1319244316362473,
-      "grad_norm": 0.025692777708172798,
-      "learning_rate": 7.08669227240909e-05,
-      "loss": 0.0062,
+      "epoch": 2.6892929292929293,
+      "grad_norm": 0.006828635931015015,
+      "learning_rate": 9.998222954599563e-05,
+      "loss": 0.0019,
       "step": 832
     },
     {
-      "epoch": 2.1344860710854947,
-      "grad_norm": 0.019108694046735764,
-      "learning_rate": 7.078555241525545e-05,
-      "loss": 0.0075,
+      "epoch": 2.6925252525252525,
+      "grad_norm": 0.007364185526967049,
+      "learning_rate": 9.99816220899565e-05,
+      "loss": 0.0022,
       "step": 833
     },
     {
-      "epoch": 2.137047710534742,
-      "grad_norm": 0.020909873768687248,
-      "learning_rate": 7.07041155014006e-05,
-      "loss": 0.0082,
+      "epoch": 2.695757575757576,
+      "grad_norm": 0.006567430216819048,
+      "learning_rate": 9.998100442709497e-05,
+      "loss": 0.002,
       "step": 834
     },
     {
-      "epoch": 2.13960934998399,
-      "grad_norm": 0.02559494785964489,
-      "learning_rate": 7.062261224348203e-05,
-      "loss": 0.0103,
+      "epoch": 2.698989898989899,
+      "grad_norm": 0.0076549774967134,
+      "learning_rate": 9.998037655753717e-05,
+      "loss": 0.0021,
       "step": 835
     },
     {
-      "epoch": 2.1421709894332372,
-      "grad_norm": 0.02691330574452877,
-      "learning_rate": 7.054104290266796e-05,
-      "loss": 0.0079,
+      "epoch": 2.7022222222222223,
+      "grad_norm": 0.006928616669028997,
+      "learning_rate": 9.997973848141137e-05,
+      "loss": 0.0022,
       "step": 836
     },
     {
-      "epoch": 2.1447326288824846,
-      "grad_norm": 0.01753104105591774,
-      "learning_rate": 7.045940774033838e-05,
-      "loss": 0.0062,
+      "epoch": 2.7054545454545456,
+      "grad_norm": 0.008701048791408539,
+      "learning_rate": 9.997909019884781e-05,
+      "loss": 0.0029,
       "step": 837
     },
     {
-      "epoch": 2.1472942683317324,
-      "grad_norm": 0.011550121009349823,
-      "learning_rate": 7.037770701808421e-05,
-      "loss": 0.0036,
+      "epoch": 2.708686868686869,
+      "grad_norm": 0.010769462212920189,
+      "learning_rate": 9.99784317099789e-05,
+      "loss": 0.0034,
       "step": 838
     },
     {
-      "epoch": 2.14985590778098,
-      "grad_norm": 0.027242513373494148,
-      "learning_rate": 7.029594099770642e-05,
-      "loss": 0.0062,
+      "epoch": 2.711919191919192,
+      "grad_norm": 0.006520561873912811,
+      "learning_rate": 9.997776301493914e-05,
+      "loss": 0.0019,
       "step": 839
     },
     {
-      "epoch": 2.152417547230227,
-      "grad_norm": 0.019376887008547783,
-      "learning_rate": 7.021410994121525e-05,
-      "loss": 0.0072,
+      "epoch": 2.7151515151515153,
+      "grad_norm": 0.006793071515858173,
+      "learning_rate": 9.997708411386501e-05,
+      "loss": 0.0021,
       "step": 840
     },
     {
-      "epoch": 2.154979186679475,
-      "grad_norm": 0.012071039527654648,
-      "learning_rate": 7.013221411082935e-05,
-      "loss": 0.0038,
+      "epoch": 2.718383838383838,
+      "grad_norm": 0.008864209987223148,
+      "learning_rate": 9.997639500689523e-05,
+      "loss": 0.0026,
       "step": 841
     },
     {
-      "epoch": 2.1575408261287223,
-      "grad_norm": 0.018893806263804436,
-      "learning_rate": 7.005025376897489e-05,
-      "loss": 0.0054,
+      "epoch": 2.7216161616161614,
+      "grad_norm": 0.011382322758436203,
+      "learning_rate": 9.997569569417049e-05,
+      "loss": 0.0041,
       "step": 842
     },
     {
-      "epoch": 2.1601024655779697,
-      "grad_norm": 0.01803342066705227,
-      "learning_rate": 6.996822917828477e-05,
-      "loss": 0.0065,
+      "epoch": 2.7248484848484846,
+      "grad_norm": 0.006367840804159641,
+      "learning_rate": 9.997498617583358e-05,
+      "loss": 0.0021,
       "step": 843
     },
     {
-      "epoch": 2.1626641050272175,
-      "grad_norm": 0.017229989171028137,
-      "learning_rate": 6.988614060159783e-05,
-      "loss": 0.0039,
+      "epoch": 2.728080808080808,
+      "grad_norm": 0.00768255116418004,
+      "learning_rate": 9.997426645202943e-05,
+      "loss": 0.0028,
       "step": 844
     },
     {
-      "epoch": 2.165225744476465,
-      "grad_norm": 0.024557722732424736,
-      "learning_rate": 6.980398830195785e-05,
-      "loss": 0.006,
+      "epoch": 2.731313131313131,
+      "grad_norm": 0.010961061343550682,
+      "learning_rate": 9.9973536522905e-05,
+      "loss": 0.0039,
       "step": 845
     },
     {
-      "epoch": 2.1677873839257122,
-      "grad_norm": 0.025856252759695053,
-      "learning_rate": 6.972177254261287e-05,
-      "loss": 0.0072,
+      "epoch": 2.7345454545454544,
+      "grad_norm": 0.007170588709414005,
+      "learning_rate": 9.997279638860933e-05,
+      "loss": 0.0026,
       "step": 846
     },
     {
-      "epoch": 2.17034902337496,
-      "grad_norm": 0.014871695078909397,
-      "learning_rate": 6.963949358701424e-05,
-      "loss": 0.0057,
+      "epoch": 2.7377777777777776,
+      "grad_norm": 0.006400369573384523,
+      "learning_rate": 9.99720460492936e-05,
+      "loss": 0.0016,
       "step": 847
     },
     {
-      "epoch": 2.1729106628242074,
-      "grad_norm": 0.020170217379927635,
-      "learning_rate": 6.955715169881585e-05,
-      "loss": 0.0061,
+      "epoch": 2.741010101010101,
+      "grad_norm": 0.009051467292010784,
+      "learning_rate": 9.997128550511099e-05,
+      "loss": 0.0035,
       "step": 848
     },
     {
-      "epoch": 2.1754723022734552,
-      "grad_norm": 0.021420400589704514,
-      "learning_rate": 6.947474714187324e-05,
-      "loss": 0.0061,
+      "epoch": 2.744242424242424,
+      "grad_norm": 0.006981056183576584,
+      "learning_rate": 9.997051475621687e-05,
+      "loss": 0.0024,
       "step": 849
     },
     {
-      "epoch": 2.1780339417227026,
-      "grad_norm": 0.02425125241279602,
-      "learning_rate": 6.939228018024275e-05,
-      "loss": 0.0081,
+      "epoch": 2.7474747474747474,
+      "grad_norm": 0.009143094532191753,
+      "learning_rate": 9.996973380276857e-05,
+      "loss": 0.0023,
       "step": 850
     },
     {
-      "epoch": 2.1780339417227026,
-      "eval_loss": 0.008289722725749016,
-      "eval_runtime": 15.7617,
-      "eval_samples_per_second": 6.344,
-      "eval_steps_per_second": 1.586,
+      "epoch": 2.7474747474747474,
+      "eval_loss": 0.0022530267015099525,
+      "eval_runtime": 15.5117,
+      "eval_samples_per_second": 6.447,
+      "eval_steps_per_second": 1.612,
       "step": 850
     },
     {
-      "epoch": 2.18059558117195,
-      "grad_norm": 0.015109110623598099,
-      "learning_rate": 6.93097510781807e-05,
-      "loss": 0.0057,
+      "epoch": 2.7507070707070707,
+      "grad_norm": 0.00784252118319273,
+      "learning_rate": 9.996894264492563e-05,
+      "loss": 0.0029,
       "step": 851
     },
     {
-      "epoch": 2.1831572206211978,
-      "grad_norm": 0.020135734230279922,
-      "learning_rate": 6.922716010014255e-05,
-      "loss": 0.0079,
+      "epoch": 2.753939393939394,
+      "grad_norm": 0.01717456988990307,
+      "learning_rate": 9.99681412828496e-05,
+      "loss": 0.0018,
       "step": 852
     },
     {
-      "epoch": 2.185718860070445,
-      "grad_norm": 0.020137479528784752,
-      "learning_rate": 6.914450751078199e-05,
-      "loss": 0.0099,
+      "epoch": 2.757171717171717,
+      "grad_norm": 0.006478489376604557,
+      "learning_rate": 9.996732971670408e-05,
+      "loss": 0.002,
       "step": 853
     },
     {
-      "epoch": 2.1882804995196925,
-      "grad_norm": 0.015187174081802368,
-      "learning_rate": 6.906179357495016e-05,
-      "loss": 0.0066,
+      "epoch": 2.7604040404040404,
+      "grad_norm": 0.006418657023459673,
+      "learning_rate": 9.996650794665487e-05,
+      "loss": 0.0022,
       "step": 854
     },
     {
-      "epoch": 2.1908421389689403,
-      "grad_norm": 0.017928671091794968,
-      "learning_rate": 6.897901855769483e-05,
-      "loss": 0.0065,
+      "epoch": 2.7636363636363637,
+      "grad_norm": 0.007500687148422003,
+      "learning_rate": 9.996567597286974e-05,
+      "loss": 0.0025,
       "step": 855
     },
     {
-      "epoch": 2.1934037784181877,
-      "grad_norm": 0.022515594959259033,
-      "learning_rate": 6.88961827242594e-05,
-      "loss": 0.0093,
+      "epoch": 2.766868686868687,
+      "grad_norm": 0.007358520291745663,
+      "learning_rate": 9.996483379551861e-05,
+      "loss": 0.0022,
       "step": 856
     },
     {
-      "epoch": 2.195965417867435,
-      "grad_norm": 0.019031262025237083,
-      "learning_rate": 6.88132863400822e-05,
-      "loss": 0.0078,
+      "epoch": 2.77010101010101,
+      "grad_norm": 0.005068556405603886,
+      "learning_rate": 9.996398141477344e-05,
+      "loss": 0.0016,
       "step": 857
     },
     {
-      "epoch": 2.198527057316683,
-      "grad_norm": 0.018223704770207405,
-      "learning_rate": 6.873032967079561e-05,
-      "loss": 0.0081,
+      "epoch": 2.7733333333333334,
+      "grad_norm": 0.007919540628790855,
+      "learning_rate": 9.996311883080832e-05,
+      "loss": 0.002,
       "step": 858
     },
     {
-      "epoch": 2.20108869676593,
-      "grad_norm": 0.024755829945206642,
-      "learning_rate": 6.864731298222515e-05,
-      "loss": 0.0078,
+      "epoch": 2.7765656565656567,
+      "grad_norm": 0.006633463315665722,
+      "learning_rate": 9.996224604379938e-05,
+      "loss": 0.0021,
       "step": 859
     },
     {
-      "epoch": 2.2036503362151776,
-      "grad_norm": 0.01624191738665104,
-      "learning_rate": 6.856423654038868e-05,
-      "loss": 0.0049,
+      "epoch": 2.77979797979798,
+      "grad_norm": 0.007528850343078375,
+      "learning_rate": 9.996136305392487e-05,
+      "loss": 0.0022,
       "step": 860
     },
     {
-      "epoch": 2.2062119756644254,
-      "grad_norm": 0.01638570800423622,
-      "learning_rate": 6.848110061149556e-05,
-      "loss": 0.006,
+      "epoch": 2.783030303030303,
+      "grad_norm": 0.007386321201920509,
+      "learning_rate": 9.996046986136509e-05,
+      "loss": 0.0029,
       "step": 861
     },
     {
-      "epoch": 2.2087736151136728,
-      "grad_norm": 0.01753126084804535,
-      "learning_rate": 6.839790546194569e-05,
-      "loss": 0.0064,
+      "epoch": 2.7862626262626264,
+      "grad_norm": 0.007898709736764431,
+      "learning_rate": 9.995956646630246e-05,
+      "loss": 0.0041,
       "step": 862
     },
     {
-      "epoch": 2.21133525456292,
-      "grad_norm": 0.02570418082177639,
-      "learning_rate": 6.831465135832883e-05,
-      "loss": 0.0077,
+      "epoch": 2.7894949494949497,
+      "grad_norm": 0.010537680238485336,
+      "learning_rate": 9.995865286892145e-05,
+      "loss": 0.0027,
       "step": 863
     },
     {
-      "epoch": 2.213896894012168,
-      "grad_norm": 0.03022204339504242,
-      "learning_rate": 6.823133856742359e-05,
-      "loss": 0.0122,
+      "epoch": 2.792727272727273,
+      "grad_norm": 0.009241768158972263,
+      "learning_rate": 9.995772906940864e-05,
+      "loss": 0.0029,
       "step": 864
     },
     {
-      "epoch": 2.2164585334614153,
-      "grad_norm": 0.021783053874969482,
-      "learning_rate": 6.814796735619663e-05,
-      "loss": 0.0064,
+      "epoch": 2.795959595959596,
+      "grad_norm": 0.008436708711087704,
+      "learning_rate": 9.995679506795264e-05,
+      "loss": 0.0027,
       "step": 865
     },
     {
-      "epoch": 2.2190201729106627,
-      "grad_norm": 0.0161015372723341,
-      "learning_rate": 6.806453799180191e-05,
-      "loss": 0.0052,
+      "epoch": 2.7991919191919195,
+      "grad_norm": 0.006161965895444155,
+      "learning_rate": 9.995585086474424e-05,
+      "loss": 0.0015,
       "step": 866
     },
     {
-      "epoch": 2.2215818123599105,
-      "grad_norm": 0.01819256693124771,
-      "learning_rate": 6.798105074157958e-05,
-      "loss": 0.0078,
+      "epoch": 2.8024242424242423,
+      "grad_norm": 0.00874664168804884,
+      "learning_rate": 9.995489645997622e-05,
+      "loss": 0.0039,
       "step": 867
     },
     {
-      "epoch": 2.224143451809158,
-      "grad_norm": 0.015615181997418404,
-      "learning_rate": 6.789750587305544e-05,
-      "loss": 0.0053,
+      "epoch": 2.8056565656565655,
+      "grad_norm": 0.016317609697580338,
+      "learning_rate": 9.99539318538435e-05,
+      "loss": 0.0069,
       "step": 868
     },
     {
-      "epoch": 2.226705091258405,
-      "grad_norm": 0.026892973110079765,
-      "learning_rate": 6.78139036539398e-05,
-      "loss": 0.0075,
+      "epoch": 2.8088888888888888,
+      "grad_norm": 0.006787766702473164,
+      "learning_rate": 9.995295704654304e-05,
+      "loss": 0.0021,
       "step": 869
     },
     {
-      "epoch": 2.229266730707653,
-      "grad_norm": 0.02187221497297287,
-      "learning_rate": 6.773024435212678e-05,
-      "loss": 0.0064,
+      "epoch": 2.812121212121212,
+      "grad_norm": 0.010913880541920662,
+      "learning_rate": 9.995197203827393e-05,
+      "loss": 0.0063,
       "step": 870
     },
     {
-      "epoch": 2.2318283701569004,
-      "grad_norm": 0.0163852721452713,
-      "learning_rate": 6.764652823569344e-05,
-      "loss": 0.0074,
+      "epoch": 2.8153535353535353,
+      "grad_norm": 0.005392727442085743,
+      "learning_rate": 9.995097682923733e-05,
+      "loss": 0.0019,
       "step": 871
     },
     {
-      "epoch": 2.2343900096061478,
-      "grad_norm": 0.014609121717512608,
-      "learning_rate": 6.75627555728989e-05,
-      "loss": 0.005,
+      "epoch": 2.8185858585858585,
+      "grad_norm": 0.010516703128814697,
+      "learning_rate": 9.994997141963644e-05,
+      "loss": 0.0038,
       "step": 872
     },
     {
-      "epoch": 2.2369516490553956,
-      "grad_norm": 0.03244280070066452,
-      "learning_rate": 6.74789266321834e-05,
-      "loss": 0.0126,
+      "epoch": 2.821818181818182,
+      "grad_norm": 0.011369968764483929,
+      "learning_rate": 9.994895580967658e-05,
+      "loss": 0.0045,
       "step": 873
     },
     {
-      "epoch": 2.239513288504643,
-      "grad_norm": 0.017335107550024986,
-      "learning_rate": 6.739504168216762e-05,
-      "loss": 0.0083,
+      "epoch": 2.825050505050505,
+      "grad_norm": 0.007756918203085661,
+      "learning_rate": 9.994792999956518e-05,
+      "loss": 0.0025,
       "step": 874
     },
     {
-      "epoch": 2.2420749279538903,
-      "grad_norm": 0.01658845879137516,
-      "learning_rate": 6.731110099165164e-05,
-      "loss": 0.0065,
+      "epoch": 2.8282828282828283,
+      "grad_norm": 0.007660750765353441,
+      "learning_rate": 9.994689398951169e-05,
+      "loss": 0.0019,
       "step": 875
     },
     {
-      "epoch": 2.2420749279538903,
-      "eval_loss": 0.007803728803992271,
-      "eval_runtime": 15.7858,
-      "eval_samples_per_second": 6.335,
-      "eval_steps_per_second": 1.584,
+      "epoch": 2.8282828282828283,
+      "eval_loss": 0.002162019954994321,
+      "eval_runtime": 15.4977,
+      "eval_samples_per_second": 6.453,
+      "eval_steps_per_second": 1.613,
       "step": 875
     },
     {
-      "epoch": 2.244636567403138,
-      "grad_norm": 0.020299553871154785,
-      "learning_rate": 6.722710482961419e-05,
-      "loss": 0.0096,
+      "epoch": 2.8315151515151515,
+      "grad_norm": 0.007931852713227272,
+      "learning_rate": 9.994584777972769e-05,
+      "loss": 0.0025,
       "step": 876
     },
     {
-      "epoch": 2.2471982068523855,
-      "grad_norm": 0.017362618818879128,
-      "learning_rate": 6.714305346521174e-05,
-      "loss": 0.0059,
+      "epoch": 2.834747474747475,
+      "grad_norm": 0.008434126153588295,
+      "learning_rate": 9.994479137042683e-05,
+      "loss": 0.0023,
       "step": 877
     },
     {
-      "epoch": 2.249759846301633,
-      "grad_norm": 0.017628263682127,
-      "learning_rate": 6.705894716777768e-05,
-      "loss": 0.0069,
+      "epoch": 2.837979797979798,
+      "grad_norm": 0.006743580102920532,
+      "learning_rate": 9.994372476182484e-05,
+      "loss": 0.0022,
       "step": 878
     },
     {
-      "epoch": 2.2523214857508806,
-      "grad_norm": 0.019219299778342247,
-      "learning_rate": 6.697478620682137e-05,
-      "loss": 0.0084,
+      "epoch": 2.8412121212121213,
+      "grad_norm": 0.008293837308883667,
+      "learning_rate": 9.994264795413953e-05,
+      "loss": 0.0022,
       "step": 879
     },
     {
-      "epoch": 2.254883125200128,
-      "grad_norm": 0.016255104914307594,
-      "learning_rate": 6.689057085202737e-05,
-      "loss": 0.006,
+      "epoch": 2.8444444444444446,
+      "grad_norm": 0.011419296264648438,
+      "learning_rate": 9.99415609475908e-05,
+      "loss": 0.0029,
       "step": 880
     },
     {
-      "epoch": 2.2574447646493754,
-      "grad_norm": 0.016757622361183167,
-      "learning_rate": 6.680630137325453e-05,
-      "loss": 0.0053,
+      "epoch": 2.847676767676768,
+      "grad_norm": 0.019320234656333923,
+      "learning_rate": 9.994046374240062e-05,
+      "loss": 0.0032,
       "step": 881
     },
     {
-      "epoch": 2.260006404098623,
-      "grad_norm": 0.012115413323044777,
-      "learning_rate": 6.672197804053516e-05,
-      "loss": 0.0044,
+      "epoch": 2.850909090909091,
+      "grad_norm": 0.010916640050709248,
+      "learning_rate": 9.993935633879306e-05,
+      "loss": 0.0033,
       "step": 882
     },
     {
-      "epoch": 2.2625680435478706,
-      "grad_norm": 0.01831790618598461,
-      "learning_rate": 6.663760112407411e-05,
-      "loss": 0.0079,
+      "epoch": 2.854141414141414,
+      "grad_norm": 0.008513251319527626,
+      "learning_rate": 9.993823873699426e-05,
+      "loss": 0.0033,
       "step": 883
     },
     {
-      "epoch": 2.2651296829971184,
-      "grad_norm": 0.015823351219296455,
-      "learning_rate": 6.65531708942479e-05,
-      "loss": 0.0055,
+      "epoch": 2.857373737373737,
+      "grad_norm": 0.006323203910142183,
+      "learning_rate": 9.993711093723245e-05,
+      "loss": 0.0015,
       "step": 884
     },
     {
-      "epoch": 2.2676913224463657,
-      "grad_norm": 0.026958780363202095,
-      "learning_rate": 6.646868762160399e-05,
-      "loss": 0.012,
+      "epoch": 2.8606060606060604,
+      "grad_norm": 0.01390031073242426,
+      "learning_rate": 9.993597293973796e-05,
+      "loss": 0.0045,
       "step": 885
     },
     {
-      "epoch": 2.270252961895613,
-      "grad_norm": 0.022467253729701042,
-      "learning_rate": 6.63841515768597e-05,
-      "loss": 0.0082,
+      "epoch": 2.8638383838383836,
+      "grad_norm": 0.0072150928899645805,
+      "learning_rate": 9.993482474474314e-05,
+      "loss": 0.0021,
       "step": 886
     },
     {
-      "epoch": 2.2728146013448605,
-      "grad_norm": 0.024623310193419456,
-      "learning_rate": 6.629956303090151e-05,
-      "loss": 0.0088,
+      "epoch": 2.867070707070707,
+      "grad_norm": 0.015069859102368355,
+      "learning_rate": 9.99336663524825e-05,
+      "loss": 0.003,
       "step": 887
     },
     {
-      "epoch": 2.2753762407941083,
-      "grad_norm": 0.022642867639660835,
-      "learning_rate": 6.621492225478414e-05,
-      "loss": 0.0087,
+      "epoch": 2.87030303030303,
+      "grad_norm": 0.007027952000498772,
+      "learning_rate": 9.993249776319258e-05,
+      "loss": 0.0025,
       "step": 888
     },
     {
-      "epoch": 2.2779378802433556,
-      "grad_norm": 0.01657482422888279,
-      "learning_rate": 6.613022951972963e-05,
-      "loss": 0.0046,
+      "epoch": 2.8735353535353534,
+      "grad_norm": 0.010848205536603928,
+      "learning_rate": 9.993131897711202e-05,
+      "loss": 0.0049,
       "step": 889
     },
     {
-      "epoch": 2.2804995196926034,
-      "grad_norm": 0.017057588323950768,
-      "learning_rate": 6.604548509712658e-05,
-      "loss": 0.0071,
+      "epoch": 2.8767676767676766,
+      "grad_norm": 0.0068895635195076466,
+      "learning_rate": 9.993012999448154e-05,
+      "loss": 0.0026,
       "step": 890
     },
     {
-      "epoch": 2.283061159141851,
-      "grad_norm": 0.01688271202147007,
-      "learning_rate": 6.596068925852916e-05,
-      "loss": 0.0063,
+      "epoch": 2.88,
+      "grad_norm": 0.02316826395690441,
+      "learning_rate": 9.992893081554397e-05,
+      "loss": 0.0045,
       "step": 891
     },
     {
-      "epoch": 2.285622798591098,
-      "grad_norm": 0.024504490196704865,
-      "learning_rate": 6.587584227565631e-05,
-      "loss": 0.0059,
+      "epoch": 2.883232323232323,
+      "grad_norm": 0.009288731031119823,
+      "learning_rate": 9.992772144054415e-05,
+      "loss": 0.003,
       "step": 892
     },
     {
-      "epoch": 2.288184438040346,
-      "grad_norm": 0.016235264018177986,
-      "learning_rate": 6.579094442039091e-05,
-      "loss": 0.0068,
+      "epoch": 2.8864646464646464,
+      "grad_norm": 0.006021425593644381,
+      "learning_rate": 9.992650186972909e-05,
+      "loss": 0.0018,
       "step": 893
     },
     {
-      "epoch": 2.2907460774895934,
-      "grad_norm": 0.018847160041332245,
-      "learning_rate": 6.570599596477876e-05,
-      "loss": 0.0065,
+      "epoch": 2.8896969696969697,
+      "grad_norm": 0.006727719213813543,
+      "learning_rate": 9.99252721033478e-05,
+      "loss": 0.0026,
       "step": 894
     },
     {
-      "epoch": 2.2933077169388407,
-      "grad_norm": 0.029839105904102325,
-      "learning_rate": 6.562099718102788e-05,
-      "loss": 0.0119,
+      "epoch": 2.892929292929293,
+      "grad_norm": 0.008267631754279137,
+      "learning_rate": 9.992403214165147e-05,
+      "loss": 0.0022,
       "step": 895
     },
     {
-      "epoch": 2.2958693563880885,
-      "grad_norm": 0.015048625878989697,
-      "learning_rate": 6.553594834150754e-05,
-      "loss": 0.0062,
+      "epoch": 2.896161616161616,
+      "grad_norm": 0.008259359747171402,
+      "learning_rate": 9.992278198489327e-05,
+      "loss": 0.0037,
       "step": 896
     },
     {
-      "epoch": 2.298430995837336,
-      "grad_norm": 0.01649918593466282,
-      "learning_rate": 6.545084971874738e-05,
-      "loss": 0.0048,
+      "epoch": 2.8993939393939394,
+      "grad_norm": 0.0065229423344135284,
+      "learning_rate": 9.99215216333285e-05,
+      "loss": 0.0026,
       "step": 897
     },
     {
-      "epoch": 2.3009926352865833,
-      "grad_norm": 0.02212654985487461,
-      "learning_rate": 6.53657015854366e-05,
-      "loss": 0.0065,
+      "epoch": 2.9026262626262627,
+      "grad_norm": 0.005497989244759083,
+      "learning_rate": 9.992025108721454e-05,
+      "loss": 0.0017,
       "step": 898
     },
     {
-      "epoch": 2.303554274735831,
-      "grad_norm": 0.016795290634036064,
-      "learning_rate": 6.528050421442306e-05,
-      "loss": 0.0051,
+      "epoch": 2.905858585858586,
+      "grad_norm": 0.00781367626041174,
+      "learning_rate": 9.991897034681087e-05,
+      "loss": 0.0026,
       "step": 899
     },
     {
-      "epoch": 2.3061159141850784,
-      "grad_norm": 0.014818985015153885,
-      "learning_rate": 6.519525787871235e-05,
-      "loss": 0.0055,
+      "epoch": 2.909090909090909,
+      "grad_norm": 0.008026889525353909,
+      "learning_rate": 9.9917679412379e-05,
+      "loss": 0.002,
       "step": 900
     },
     {
-      "epoch": 2.3061159141850784,
-      "eval_loss": 0.007624690420925617,
-      "eval_runtime": 15.7609,
-      "eval_samples_per_second": 6.345,
-      "eval_steps_per_second": 1.586,
+      "epoch": 2.909090909090909,
+      "eval_loss": 0.0020985938608646393,
+      "eval_runtime": 15.5052,
+      "eval_samples_per_second": 6.449,
+      "eval_steps_per_second": 1.612,
       "step": 900
     },
     {
-      "epoch": 2.308677553634326,
-      "grad_norm": 0.020696334540843964,
-      "learning_rate": 6.5109962851467e-05,
-      "loss": 0.0103,
+      "epoch": 2.9123232323232324,
+      "grad_norm": 0.00950902234762907,
+      "learning_rate": 9.99163782841826e-05,
+      "loss": 0.0032,
       "step": 901
     },
     {
-      "epoch": 2.3112391930835736,
-      "grad_norm": 0.02347083017230034,
-      "learning_rate": 6.502461940600555e-05,
-      "loss": 0.0078,
+      "epoch": 2.9155555555555557,
+      "grad_norm": 0.00970569159835577,
+      "learning_rate": 9.991506696248731e-05,
+      "loss": 0.0029,
       "step": 902
     },
     {
-      "epoch": 2.313800832532821,
-      "grad_norm": 0.01726224087178707,
-      "learning_rate": 6.49392278158017e-05,
-      "loss": 0.0064,
+      "epoch": 2.918787878787879,
+      "grad_norm": 0.009722487069666386,
+      "learning_rate": 9.991374544756098e-05,
+      "loss": 0.003,
       "step": 903
     },
     {
-      "epoch": 2.3163624719820683,
-      "grad_norm": 0.024909283965826035,
-      "learning_rate": 6.485378835448342e-05,
-      "loss": 0.0124,
+      "epoch": 2.922020202020202,
+      "grad_norm": 0.007319387048482895,
+      "learning_rate": 9.991241373967344e-05,
+      "loss": 0.0017,
       "step": 904
     },
     {
-      "epoch": 2.318924111431316,
-      "grad_norm": 0.01956700161099434,
-      "learning_rate": 6.476830129583206e-05,
-      "loss": 0.006,
+      "epoch": 2.9252525252525254,
+      "grad_norm": 0.011507434770464897,
+      "learning_rate": 9.991107183909664e-05,
+      "loss": 0.0026,
       "step": 905
     },
     {
-      "epoch": 2.3214857508805635,
-      "grad_norm": 0.030146628618240356,
-      "learning_rate": 6.468276691378155e-05,
-      "loss": 0.0081,
+      "epoch": 2.9284848484848487,
+      "grad_norm": 0.006943380925804377,
+      "learning_rate": 9.990971974610466e-05,
+      "loss": 0.002,
       "step": 906
     },
     {
-      "epoch": 2.324047390329811,
-      "grad_norm": 0.01578514277935028,
-      "learning_rate": 6.459718548241738e-05,
-      "loss": 0.0064,
+      "epoch": 2.931717171717172,
+      "grad_norm": 0.010192270390689373,
+      "learning_rate": 9.990835746097356e-05,
+      "loss": 0.0033,
       "step": 907
     },
     {
-      "epoch": 2.3266090297790587,
-      "grad_norm": 0.020581640303134918,
-      "learning_rate": 6.451155727597586e-05,
-      "loss": 0.0064,
+      "epoch": 2.934949494949495,
+      "grad_norm": 0.007816629484295845,
+      "learning_rate": 9.990698498398155e-05,
+      "loss": 0.0023,
       "step": 908
     },
     {
-      "epoch": 2.329170669228306,
-      "grad_norm": 0.01611717976629734,
-      "learning_rate": 6.442588256884318e-05,
-      "loss": 0.0061,
+      "epoch": 2.9381818181818184,
+      "grad_norm": 0.010911786928772926,
+      "learning_rate": 9.990560231540889e-05,
+      "loss": 0.0037,
       "step": 909
     },
     {
-      "epoch": 2.3317323086775534,
-      "grad_norm": 0.025673942640423775,
-      "learning_rate": 6.434016163555452e-05,
-      "loss": 0.0099,
+      "epoch": 2.9414141414141413,
+      "grad_norm": 0.008337637409567833,
+      "learning_rate": 9.990420945553797e-05,
+      "loss": 0.002,
       "step": 910
     },
     {
-      "epoch": 2.3342939481268012,
-      "grad_norm": 0.025654910132288933,
-      "learning_rate": 6.42543947507932e-05,
-      "loss": 0.0087,
+      "epoch": 2.9446464646464645,
+      "grad_norm": 0.006498595234006643,
+      "learning_rate": 9.990280640465321e-05,
+      "loss": 0.0022,
       "step": 911
     },
     {
-      "epoch": 2.3368555875760486,
-      "grad_norm": 0.01794145628809929,
-      "learning_rate": 6.416858218938976e-05,
-      "loss": 0.0063,
+      "epoch": 2.9478787878787878,
+      "grad_norm": 0.008626777678728104,
+      "learning_rate": 9.990139316304112e-05,
+      "loss": 0.003,
       "step": 912
     },
     {
-      "epoch": 2.3394172270252964,
-      "grad_norm": 0.019059333950281143,
-      "learning_rate": 6.408272422632114e-05,
-      "loss": 0.0063,
+      "epoch": 2.951111111111111,
+      "grad_norm": 0.00594016769900918,
+      "learning_rate": 9.989996973099032e-05,
+      "loss": 0.0016,
       "step": 913
     },
     {
-      "epoch": 2.341978866474544,
-      "grad_norm": 0.018267640843987465,
-      "learning_rate": 6.399682113670978e-05,
-      "loss": 0.0055,
+      "epoch": 2.9543434343434343,
+      "grad_norm": 0.007042631506919861,
+      "learning_rate": 9.989853610879147e-05,
+      "loss": 0.0027,
       "step": 914
     },
     {
-      "epoch": 2.344540505923791,
-      "grad_norm": 0.028884587809443474,
-      "learning_rate": 6.391087319582264e-05,
-      "loss": 0.0091,
+      "epoch": 2.9575757575757575,
+      "grad_norm": 0.006258354522287846,
+      "learning_rate": 9.989709229673736e-05,
+      "loss": 0.0017,
       "step": 915
     },
     {
-      "epoch": 2.3471021453730385,
-      "grad_norm": 0.02319938875734806,
-      "learning_rate": 6.382488067907047e-05,
-      "loss": 0.009,
+      "epoch": 2.9608080808080808,
+      "grad_norm": 0.007070127874612808,
+      "learning_rate": 9.98956382951228e-05,
+      "loss": 0.0025,
       "step": 916
     },
     {
-      "epoch": 2.3496637848222863,
-      "grad_norm": 0.019833754748106003,
-      "learning_rate": 6.373884386200689e-05,
-      "loss": 0.0074,
+      "epoch": 2.964040404040404,
+      "grad_norm": 0.011631196364760399,
+      "learning_rate": 9.989417410424475e-05,
+      "loss": 0.0023,
       "step": 917
     },
     {
-      "epoch": 2.3522254242715337,
-      "grad_norm": 0.02763204462826252,
-      "learning_rate": 6.365276302032737e-05,
-      "loss": 0.009,
+      "epoch": 2.9672727272727273,
+      "grad_norm": 0.009569251909852028,
+      "learning_rate": 9.98926997244022e-05,
+      "loss": 0.0016,
       "step": 918
     },
     {
-      "epoch": 2.3547870637207815,
-      "grad_norm": 0.024364588782191277,
-      "learning_rate": 6.356663842986854e-05,
-      "loss": 0.0091,
+      "epoch": 2.9705050505050505,
+      "grad_norm": 0.006656884681433439,
+      "learning_rate": 9.989121515589622e-05,
+      "loss": 0.0013,
       "step": 919
     },
     {
-      "epoch": 2.357348703170029,
-      "grad_norm": 0.01990056037902832,
-      "learning_rate": 6.34804703666072e-05,
-      "loss": 0.0097,
+      "epoch": 2.973737373737374,
+      "grad_norm": 0.01403870154172182,
+      "learning_rate": 9.988972039902997e-05,
+      "loss": 0.0044,
       "step": 920
     },
     {
-      "epoch": 2.3599103426192762,
-      "grad_norm": 0.024780521169304848,
-      "learning_rate": 6.339425910665941e-05,
-      "loss": 0.0095,
+      "epoch": 2.976969696969697,
+      "grad_norm": 0.011693855747580528,
+      "learning_rate": 9.988821545410874e-05,
+      "loss": 0.0058,
       "step": 921
     },
     {
-      "epoch": 2.362471982068524,
-      "grad_norm": 0.02050694450736046,
-      "learning_rate": 6.33080049262797e-05,
-      "loss": 0.0068,
+      "epoch": 2.9802020202020203,
+      "grad_norm": 0.008179626427590847,
+      "learning_rate": 9.988670032143981e-05,
+      "loss": 0.0023,
       "step": 922
     },
     {
-      "epoch": 2.3650336215177714,
-      "grad_norm": 0.01734190620481968,
-      "learning_rate": 6.322170810186012e-05,
-      "loss": 0.0065,
+      "epoch": 2.9834343434343435,
+      "grad_norm": 0.008626478724181652,
+      "learning_rate": 9.988517500133262e-05,
+      "loss": 0.0025,
       "step": 923
     },
     {
-      "epoch": 2.3675952609670188,
-      "grad_norm": 0.02429112046957016,
-      "learning_rate": 6.313536890992935e-05,
-      "loss": 0.0071,
+      "epoch": 2.986666666666667,
+      "grad_norm": 0.007475933991372585,
+      "learning_rate": 9.988363949409865e-05,
+      "loss": 0.002,
       "step": 924
     },
     {
-      "epoch": 2.3701569004162666,
-      "grad_norm": 0.01471394207328558,
-      "learning_rate": 6.304898762715186e-05,
-      "loss": 0.0043,
+      "epoch": 2.98989898989899,
+      "grad_norm": 0.010853493586182594,
+      "learning_rate": 9.988209380005144e-05,
+      "loss": 0.0025,
       "step": 925
     },
     {
-      "epoch": 2.3701569004162666,
-      "eval_loss": 0.007436561398208141,
-      "eval_runtime": 15.7935,
-      "eval_samples_per_second": 6.332,
-      "eval_steps_per_second": 1.583,
+      "epoch": 2.98989898989899,
+      "eval_loss": 0.0020750754047185183,
+      "eval_runtime": 15.4834,
+      "eval_samples_per_second": 6.459,
+      "eval_steps_per_second": 1.615,
       "step": 925
     },
     {
-      "epoch": 2.372718539865514,
-      "grad_norm": 0.019127381965517998,
-      "learning_rate": 6.296256453032698e-05,
-      "loss": 0.006,
+      "epoch": 2.993131313131313,
+      "grad_norm": 0.00716310553252697,
+      "learning_rate": 9.98805379195067e-05,
+      "loss": 0.0024,
       "step": 926
     },
     {
-      "epoch": 2.3752801793147613,
-      "grad_norm": 0.025896169245243073,
-      "learning_rate": 6.287609989638801e-05,
-      "loss": 0.0091,
+      "epoch": 2.996363636363636,
+      "grad_norm": 0.009280243888497353,
+      "learning_rate": 9.987897185278208e-05,
+      "loss": 0.0028,
       "step": 927
     },
     {
-      "epoch": 2.377841818764009,
-      "grad_norm": 0.021031564101576805,
-      "learning_rate": 6.278959400240141e-05,
-      "loss": 0.0089,
+      "epoch": 2.9995959595959594,
+      "grad_norm": 0.005309139844030142,
+      "learning_rate": 9.987739560019746e-05,
+      "loss": 0.0019,
       "step": 928
     },
     {
-      "epoch": 2.3804034582132565,
-      "grad_norm": 0.022658109664916992,
-      "learning_rate": 6.270304712556578e-05,
-      "loss": 0.0072,
+      "epoch": 3.0028282828282826,
+      "grad_norm": 0.010185063816606998,
+      "learning_rate": 9.987580916207468e-05,
+      "loss": 0.004,
       "step": 929
     },
     {
-      "epoch": 2.382965097662504,
-      "grad_norm": 0.03344877436757088,
-      "learning_rate": 6.261645954321109e-05,
-      "loss": 0.0077,
+      "epoch": 3.006060606060606,
+      "grad_norm": 0.011335788294672966,
+      "learning_rate": 9.987421253873775e-05,
+      "loss": 0.0035,
       "step": 930
     },
     {
-      "epoch": 2.3855267371117517,
-      "grad_norm": 0.02661379799246788,
-      "learning_rate": 6.252983153279775e-05,
-      "loss": 0.0092,
+      "epoch": 3.009292929292929,
+      "grad_norm": 0.0048143682070076466,
+      "learning_rate": 9.987260573051269e-05,
+      "loss": 0.0014,
       "step": 931
     },
     {
-      "epoch": 2.388088376560999,
-      "grad_norm": 0.016638459637761116,
-      "learning_rate": 6.244316337191567e-05,
-      "loss": 0.0056,
+      "epoch": 3.0125252525252524,
+      "grad_norm": 0.007049249950796366,
+      "learning_rate": 9.987098873772763e-05,
+      "loss": 0.0017,
       "step": 932
     },
     {
-      "epoch": 2.3906500160102464,
-      "grad_norm": 0.02403213083744049,
-      "learning_rate": 6.235645533828349e-05,
-      "loss": 0.0081,
+      "epoch": 3.0157575757575756,
+      "grad_norm": 0.009972291067242622,
+      "learning_rate": 9.986936156071278e-05,
+      "loss": 0.0024,
       "step": 933
     },
     {
-      "epoch": 2.393211655459494,
-      "grad_norm": 0.021478185430169106,
-      "learning_rate": 6.226970770974754e-05,
-      "loss": 0.0066,
+      "epoch": 3.018989898989899,
+      "grad_norm": 0.007895351387560368,
+      "learning_rate": 9.986772419980044e-05,
+      "loss": 0.0018,
       "step": 934
     },
     {
-      "epoch": 2.3957732949087416,
-      "grad_norm": 0.021162012591958046,
-      "learning_rate": 6.21829207642811e-05,
-      "loss": 0.0068,
+      "epoch": 3.022222222222222,
+      "grad_norm": 0.009303879924118519,
+      "learning_rate": 9.986607665532497e-05,
+      "loss": 0.0027,
       "step": 935
     },
     {
-      "epoch": 2.398334934357989,
-      "grad_norm": 0.016740530729293823,
-      "learning_rate": 6.209609477998338e-05,
-      "loss": 0.0063,
+      "epoch": 3.0254545454545454,
+      "grad_norm": 0.009594123810529709,
+      "learning_rate": 9.986441892762281e-05,
+      "loss": 0.0019,
       "step": 936
     },
     {
-      "epoch": 2.4008965738072368,
-      "grad_norm": 0.02144131250679493,
-      "learning_rate": 6.200923003507875e-05,
-      "loss": 0.008,
+      "epoch": 3.0286868686868686,
+      "grad_norm": 0.007033223286271095,
+      "learning_rate": 9.98627510170325e-05,
+      "loss": 0.0015,
       "step": 937
     },
     {
-      "epoch": 2.403458213256484,
-      "grad_norm": 0.032675664871931076,
-      "learning_rate": 6.192232680791567e-05,
-      "loss": 0.0119,
+      "epoch": 3.031919191919192,
+      "grad_norm": 0.009397522546350956,
+      "learning_rate": 9.986107292389464e-05,
+      "loss": 0.0021,
       "step": 938
     },
     {
-      "epoch": 2.4060198527057315,
-      "grad_norm": 0.02699361927807331,
-      "learning_rate": 6.183538537696605e-05,
-      "loss": 0.0086,
+      "epoch": 3.035151515151515,
+      "grad_norm": 0.010171709582209587,
+      "learning_rate": 9.985938464855191e-05,
+      "loss": 0.0015,
       "step": 939
     },
     {
-      "epoch": 2.4085814921549793,
-      "grad_norm": 0.018751805648207664,
-      "learning_rate": 6.174840602082412e-05,
-      "loss": 0.007,
+      "epoch": 3.0383838383838384,
+      "grad_norm": 0.01049469318240881,
+      "learning_rate": 9.985768619134909e-05,
+      "loss": 0.0018,
       "step": 940
     },
     {
-      "epoch": 2.4111431316042267,
-      "grad_norm": 0.013862990774214268,
-      "learning_rate": 6.166138901820567e-05,
-      "loss": 0.0044,
+      "epoch": 3.0416161616161617,
+      "grad_norm": 0.008499372750520706,
+      "learning_rate": 9.985597755263302e-05,
+      "loss": 0.0015,
       "step": 941
     },
     {
-      "epoch": 2.413704771053474,
-      "grad_norm": 0.018660085275769234,
-      "learning_rate": 6.157433464794716e-05,
-      "loss": 0.007,
+      "epoch": 3.044848484848485,
+      "grad_norm": 0.009324614889919758,
+      "learning_rate": 9.985425873275263e-05,
+      "loss": 0.002,
       "step": 942
     },
     {
-      "epoch": 2.416266410502722,
-      "grad_norm": 0.016144121065735817,
-      "learning_rate": 6.148724318900472e-05,
-      "loss": 0.0052,
+      "epoch": 3.048080808080808,
+      "grad_norm": 0.006538646295666695,
+      "learning_rate": 9.98525297320589e-05,
+      "loss": 0.0014,
       "step": 943
     },
     {
-      "epoch": 2.418828049951969,
-      "grad_norm": 0.020236199721693993,
-      "learning_rate": 6.140011492045338e-05,
-      "loss": 0.0072,
+      "epoch": 3.0513131313131314,
+      "grad_norm": 0.004879586398601532,
+      "learning_rate": 9.985079055090493e-05,
+      "loss": 0.0015,
       "step": 944
     },
     {
-      "epoch": 2.4213896894012166,
-      "grad_norm": 0.02062610536813736,
-      "learning_rate": 6.131295012148612e-05,
-      "loss": 0.0076,
+      "epoch": 3.0545454545454547,
+      "grad_norm": 0.005798091180622578,
+      "learning_rate": 9.984904118964588e-05,
+      "loss": 0.0013,
       "step": 945
     },
     {
-      "epoch": 2.4239513288504644,
-      "grad_norm": 0.0255962535738945,
-      "learning_rate": 6.122574907141291e-05,
-      "loss": 0.0108,
+      "epoch": 3.057777777777778,
+      "grad_norm": 0.008816920220851898,
+      "learning_rate": 9.984728164863898e-05,
+      "loss": 0.0021,
       "step": 946
     },
     {
-      "epoch": 2.4265129682997117,
-      "grad_norm": 0.01486723218113184,
-      "learning_rate": 6.113851204966001e-05,
-      "loss": 0.0068,
+      "epoch": 3.061010101010101,
+      "grad_norm": 0.009533471427857876,
+      "learning_rate": 9.984551192824355e-05,
+      "loss": 0.0024,
       "step": 947
     },
     {
-      "epoch": 2.4290746077489596,
-      "grad_norm": 0.01589619740843773,
-      "learning_rate": 6.105123933576882e-05,
-      "loss": 0.0051,
+      "epoch": 3.0642424242424244,
+      "grad_norm": 0.012373212724924088,
+      "learning_rate": 9.9843732028821e-05,
+      "loss": 0.0026,
       "step": 948
     },
     {
-      "epoch": 2.431636247198207,
-      "grad_norm": 0.018491653725504875,
-      "learning_rate": 6.096393120939516e-05,
-      "loss": 0.0075,
+      "epoch": 3.0674747474747477,
+      "grad_norm": 0.006702150218188763,
+      "learning_rate": 9.98419419507348e-05,
+      "loss": 0.0014,
       "step": 949
     },
     {
-      "epoch": 2.4341978866474543,
-      "grad_norm": 0.015982965007424355,
-      "learning_rate": 6.087658795030837e-05,
-      "loss": 0.0063,
+      "epoch": 3.0707070707070705,
+      "grad_norm": 0.00798215065151453,
+      "learning_rate": 9.98401416943505e-05,
+      "loss": 0.0014,
       "step": 950
     },
     {
-      "epoch": 2.4341978866474543,
-      "eval_loss": 0.0073355091735720634,
-      "eval_runtime": 15.7799,
-      "eval_samples_per_second": 6.337,
-      "eval_steps_per_second": 1.584,
+      "epoch": 3.0707070707070705,
+      "eval_loss": 0.002163056982681155,
+      "eval_runtime": 15.5739,
+      "eval_samples_per_second": 6.421,
+      "eval_steps_per_second": 1.605,
       "step": 950
     },
     {
-      "epoch": 2.4367595260967017,
-      "grad_norm": 0.016064204275608063,
-      "learning_rate": 6.078920983839031e-05,
-      "loss": 0.0054,
+      "epoch": 3.0739393939393937,
+      "grad_norm": 0.01455872692167759,
+      "learning_rate": 9.983833126003572e-05,
+      "loss": 0.0043,
       "step": 951
     },
     {
-      "epoch": 2.4393211655459495,
-      "grad_norm": 0.03509011119604111,
-      "learning_rate": 6.070179715363453e-05,
-      "loss": 0.0154,
+      "epoch": 3.077171717171717,
+      "grad_norm": 0.009544805623590946,
+      "learning_rate": 9.98365106481602e-05,
+      "loss": 0.0027,
       "step": 952
     },
     {
-      "epoch": 2.441882804995197,
-      "grad_norm": 0.018619479611516,
-      "learning_rate": 6.06143501761454e-05,
-      "loss": 0.0069,
+      "epoch": 3.0804040404040403,
+      "grad_norm": 0.008772431872785091,
+      "learning_rate": 9.983467985909573e-05,
+      "loss": 0.0017,
       "step": 953
     },
     {
-      "epoch": 2.4444444444444446,
-      "grad_norm": 0.019991964101791382,
-      "learning_rate": 6.052686918613712e-05,
-      "loss": 0.0083,
+      "epoch": 3.0836363636363635,
+      "grad_norm": 0.00674818130210042,
+      "learning_rate": 9.983283889321615e-05,
+      "loss": 0.0019,
       "step": 954
     },
     {
-      "epoch": 2.447006083893692,
-      "grad_norm": 0.01930798403918743,
-      "learning_rate": 6.043935446393294e-05,
-      "loss": 0.0079,
+      "epoch": 3.0868686868686868,
+      "grad_norm": 0.006830274593085051,
+      "learning_rate": 9.983098775089742e-05,
+      "loss": 0.0018,
       "step": 955
     },
     {
-      "epoch": 2.4495677233429394,
-      "grad_norm": 0.01885543018579483,
-      "learning_rate": 6.035180628996416e-05,
-      "loss": 0.0071,
+      "epoch": 3.09010101010101,
+      "grad_norm": 0.007949273101985455,
+      "learning_rate": 9.982912643251757e-05,
+      "loss": 0.002,
       "step": 956
     },
     {
-      "epoch": 2.452129362792187,
-      "grad_norm": 0.021524138748645782,
-      "learning_rate": 6.02642249447693e-05,
-      "loss": 0.0081,
+      "epoch": 3.0933333333333333,
+      "grad_norm": 0.007161784451454878,
+      "learning_rate": 9.98272549384567e-05,
+      "loss": 0.0019,
       "step": 957
     },
     {
-      "epoch": 2.4546910022414345,
-      "grad_norm": 0.020283209159970284,
-      "learning_rate": 6.017661070899314e-05,
-      "loss": 0.0057,
+      "epoch": 3.0965656565656565,
+      "grad_norm": 0.007934720255434513,
+      "learning_rate": 9.982537326909697e-05,
+      "loss": 0.0019,
       "step": 958
     },
     {
-      "epoch": 2.457252641690682,
-      "grad_norm": 0.02511567436158657,
-      "learning_rate": 6.0088963863385886e-05,
-      "loss": 0.0073,
+      "epoch": 3.0997979797979798,
+      "grad_norm": 0.003556368174031377,
+      "learning_rate": 9.982348142482269e-05,
+      "loss": 0.001,
       "step": 959
     },
     {
-      "epoch": 2.4598142811399297,
-      "grad_norm": 0.013706948608160019,
-      "learning_rate": 6.0001284688802226e-05,
-      "loss": 0.0039,
+      "epoch": 3.103030303030303,
+      "grad_norm": 0.007654329761862755,
+      "learning_rate": 9.982157940602014e-05,
+      "loss": 0.0016,
       "step": 960
     },
     {
-      "epoch": 2.462375920589177,
-      "grad_norm": 0.025410829111933708,
-      "learning_rate": 5.9913573466200444e-05,
-      "loss": 0.0113,
+      "epoch": 3.1062626262626263,
+      "grad_norm": 0.012930963188409805,
+      "learning_rate": 9.981966721307778e-05,
+      "loss": 0.0027,
       "step": 961
     },
     {
-      "epoch": 2.4649375600384245,
-      "grad_norm": 0.017684413120150566,
-      "learning_rate": 5.982583047664151e-05,
-      "loss": 0.0076,
+      "epoch": 3.1094949494949495,
+      "grad_norm": 0.012075499631464481,
+      "learning_rate": 9.981774484638606e-05,
+      "loss": 0.0043,
       "step": 962
     },
     {
-      "epoch": 2.4674991994876723,
-      "grad_norm": 0.017570337280631065,
-      "learning_rate": 5.97380560012882e-05,
-      "loss": 0.0075,
+      "epoch": 3.112727272727273,
+      "grad_norm": 0.0126312505453825,
+      "learning_rate": 9.981581230633758e-05,
+      "loss": 0.0021,
       "step": 963
     },
     {
-      "epoch": 2.4700608389369196,
-      "grad_norm": 0.020053574815392494,
-      "learning_rate": 5.965025032140419e-05,
-      "loss": 0.0109,
+      "epoch": 3.115959595959596,
+      "grad_norm": 0.009510237723588943,
+      "learning_rate": 9.981386959332697e-05,
+      "loss": 0.002,
       "step": 964
     },
     {
-      "epoch": 2.472622478386167,
-      "grad_norm": 0.01965426467359066,
-      "learning_rate": 5.956241371835312e-05,
-      "loss": 0.0056,
+      "epoch": 3.1191919191919193,
+      "grad_norm": 0.008141322061419487,
+      "learning_rate": 9.981191670775097e-05,
+      "loss": 0.0026,
       "step": 965
     },
     {
-      "epoch": 2.475184117835415,
-      "grad_norm": 0.020925339311361313,
-      "learning_rate": 5.947454647359774e-05,
-      "loss": 0.0082,
+      "epoch": 3.1224242424242425,
+      "grad_norm": 0.006920178886502981,
+      "learning_rate": 9.980995365000836e-05,
+      "loss": 0.0024,
       "step": 966
     },
     {
-      "epoch": 2.477745757284662,
-      "grad_norm": 0.017741505056619644,
-      "learning_rate": 5.938664886869897e-05,
-      "loss": 0.0056,
+      "epoch": 3.125656565656566,
+      "grad_norm": 0.007769599091261625,
+      "learning_rate": 9.980798042050004e-05,
+      "loss": 0.0026,
       "step": 967
     },
     {
-      "epoch": 2.4803073967339095,
-      "grad_norm": 0.028152042999863625,
-      "learning_rate": 5.9298721185315053e-05,
-      "loss": 0.0135,
+      "epoch": 3.128888888888889,
+      "grad_norm": 0.015179040841758251,
+      "learning_rate": 9.980599701962896e-05,
+      "loss": 0.0041,
       "step": 968
     },
     {
-      "epoch": 2.4828690361831574,
-      "grad_norm": 0.028088664636015892,
-      "learning_rate": 5.921076370520058e-05,
-      "loss": 0.0054,
+      "epoch": 3.1321212121212123,
+      "grad_norm": 0.007052835542708635,
+      "learning_rate": 9.980400344780015e-05,
+      "loss": 0.0019,
       "step": 969
     },
     {
-      "epoch": 2.4854306756324047,
-      "grad_norm": 0.011729970574378967,
-      "learning_rate": 5.912277671020564e-05,
-      "loss": 0.0038,
+      "epoch": 3.1353535353535356,
+      "grad_norm": 0.008246042765676975,
+      "learning_rate": 9.98019997054207e-05,
+      "loss": 0.0023,
       "step": 970
     },
     {
-      "epoch": 2.487992315081652,
-      "grad_norm": 0.031858205795288086,
-      "learning_rate": 5.9034760482274875e-05,
-      "loss": 0.0096,
+      "epoch": 3.1385858585858584,
+      "grad_norm": 0.006023083347827196,
+      "learning_rate": 9.979998579289984e-05,
+      "loss": 0.0021,
       "step": 971
     },
     {
-      "epoch": 2.4905539545309,
-      "grad_norm": 0.02154734544456005,
-      "learning_rate": 5.894671530344664e-05,
-      "loss": 0.0125,
+      "epoch": 3.1418181818181816,
+      "grad_norm": 0.00788182020187378,
+      "learning_rate": 9.979796171064881e-05,
+      "loss": 0.0023,
       "step": 972
     },
     {
-      "epoch": 2.4931155939801473,
-      "grad_norm": 0.021228833124041557,
-      "learning_rate": 5.8858641455852036e-05,
-      "loss": 0.0057,
+      "epoch": 3.145050505050505,
+      "grad_norm": 0.004013084806501865,
+      "learning_rate": 9.979592745908095e-05,
+      "loss": 0.0009,
       "step": 973
     },
     {
-      "epoch": 2.4956772334293946,
-      "grad_norm": 0.018846942111849785,
-      "learning_rate": 5.8770539221714014e-05,
-      "loss": 0.0064,
+      "epoch": 3.148282828282828,
+      "grad_norm": 0.005609092302620411,
+      "learning_rate": 9.979388303861169e-05,
+      "loss": 0.0017,
       "step": 974
     },
     {
-      "epoch": 2.4982388728786424,
-      "grad_norm": 0.02012551948428154,
-      "learning_rate": 5.868240888334653e-05,
-      "loss": 0.0056,
+      "epoch": 3.1515151515151514,
+      "grad_norm": 0.004820605274289846,
+      "learning_rate": 9.97918284496585e-05,
+      "loss": 0.0012,
       "step": 975
     },
     {
-      "epoch": 2.4982388728786424,
-      "eval_loss": 0.007884026505053043,
-      "eval_runtime": 15.7805,
-      "eval_samples_per_second": 6.337,
-      "eval_steps_per_second": 1.584,
+      "epoch": 3.1515151515151514,
+      "eval_loss": 0.0021045615430921316,
+      "eval_runtime": 15.6003,
+      "eval_samples_per_second": 6.41,
+      "eval_steps_per_second": 1.603,
       "step": 975
     },
     {
-      "epoch": 2.50080051232789,
-      "grad_norm": 0.023805588483810425,
-      "learning_rate": 5.8594250723153544e-05,
-      "loss": 0.0076,
+      "epoch": 3.1547474747474746,
+      "grad_norm": 0.007316002622246742,
+      "learning_rate": 9.978976369264098e-05,
+      "loss": 0.0016,
       "step": 976
     },
     {
-      "epoch": 2.5033621517771376,
-      "grad_norm": 0.01971352845430374,
-      "learning_rate": 5.850606502362822e-05,
-      "loss": 0.006,
+      "epoch": 3.157979797979798,
+      "grad_norm": 0.008189670741558075,
+      "learning_rate": 9.978768876798075e-05,
+      "loss": 0.0022,
       "step": 977
     },
     {
-      "epoch": 2.505923791226385,
-      "grad_norm": 0.015985120087862015,
-      "learning_rate": 5.841785206735192e-05,
-      "loss": 0.0056,
+      "epoch": 3.161212121212121,
+      "grad_norm": 0.005883866921067238,
+      "learning_rate": 9.978560367610156e-05,
+      "loss": 0.0019,
       "step": 978
     },
     {
-      "epoch": 2.5084854306756323,
-      "grad_norm": 0.022680046036839485,
-      "learning_rate": 5.8329612136993414e-05,
-      "loss": 0.0074,
+      "epoch": 3.1644444444444444,
+      "grad_norm": 0.011413908563554287,
+      "learning_rate": 9.978350841742919e-05,
+      "loss": 0.0042,
       "step": 979
     },
     {
-      "epoch": 2.5110470701248797,
-      "grad_norm": 0.019912639632821083,
-      "learning_rate": 5.824134551530783e-05,
-      "loss": 0.007,
+      "epoch": 3.1676767676767676,
+      "grad_norm": 0.005321170203387737,
+      "learning_rate": 9.978140299239152e-05,
+      "loss": 0.0012,
       "step": 980
     },
     {
-      "epoch": 2.5136087095741275,
-      "grad_norm": 0.022760851308703423,
-      "learning_rate": 5.815305248513587e-05,
-      "loss": 0.0069,
+      "epoch": 3.170909090909091,
+      "grad_norm": 0.007383297197520733,
+      "learning_rate": 9.977928740141851e-05,
+      "loss": 0.0024,
       "step": 981
     },
     {
-      "epoch": 2.516170349023375,
-      "grad_norm": 0.02517043985426426,
-      "learning_rate": 5.8064733329402856e-05,
-      "loss": 0.0087,
+      "epoch": 3.174141414141414,
+      "grad_norm": 0.007583739701658487,
+      "learning_rate": 9.977716164494217e-05,
+      "loss": 0.0019,
       "step": 982
     },
     {
-      "epoch": 2.5187319884726227,
-      "grad_norm": 0.027291610836982727,
-      "learning_rate": 5.7976388331117807e-05,
-      "loss": 0.0113,
+      "epoch": 3.1773737373737374,
+      "grad_norm": 0.011693987995386124,
+      "learning_rate": 9.977502572339664e-05,
+      "loss": 0.0023,
       "step": 983
     },
     {
-      "epoch": 2.52129362792187,
-      "grad_norm": 0.02210170030593872,
-      "learning_rate": 5.788801777337257e-05,
-      "loss": 0.0078,
+      "epoch": 3.1806060606060607,
+      "grad_norm": 0.005918224807828665,
+      "learning_rate": 9.977287963721804e-05,
+      "loss": 0.0015,
       "step": 984
     },
     {
-      "epoch": 2.5238552673711174,
-      "grad_norm": 0.01957295462489128,
-      "learning_rate": 5.7799621939340896e-05,
-      "loss": 0.0066,
+      "epoch": 3.183838383838384,
+      "grad_norm": 0.01007204968482256,
+      "learning_rate": 9.977072338684469e-05,
+      "loss": 0.0019,
       "step": 985
     },
     {
-      "epoch": 2.526416906820365,
-      "grad_norm": 0.019617103040218353,
-      "learning_rate": 5.771120111227751e-05,
-      "loss": 0.0074,
+      "epoch": 3.187070707070707,
+      "grad_norm": 0.006889281794428825,
+      "learning_rate": 9.976855697271689e-05,
+      "loss": 0.0016,
       "step": 986
     },
     {
-      "epoch": 2.5289785462696126,
-      "grad_norm": 0.028092211112380028,
-      "learning_rate": 5.762275557551727e-05,
-      "loss": 0.0071,
+      "epoch": 3.1903030303030304,
+      "grad_norm": 0.009370551444590092,
+      "learning_rate": 9.976638039527704e-05,
+      "loss": 0.0017,
       "step": 987
     },
     {
-      "epoch": 2.53154018571886,
-      "grad_norm": 0.024590661749243736,
-      "learning_rate": 5.753428561247416e-05,
-      "loss": 0.0084,
+      "epoch": 3.1935353535353537,
+      "grad_norm": 0.00849044881761074,
+      "learning_rate": 9.976419365496963e-05,
+      "loss": 0.0024,
       "step": 988
     },
     {
-      "epoch": 2.534101825168108,
-      "grad_norm": 0.024077923968434334,
-      "learning_rate": 5.744579150664046e-05,
-      "loss": 0.0086,
+      "epoch": 3.196767676767677,
+      "grad_norm": 0.010178672149777412,
+      "learning_rate": 9.976199675224123e-05,
+      "loss": 0.003,
       "step": 989
     },
     {
-      "epoch": 2.536663464617355,
-      "grad_norm": 0.017687125131487846,
-      "learning_rate": 5.735727354158581e-05,
-      "loss": 0.0066,
+      "epoch": 3.2,
+      "grad_norm": 0.00836561806499958,
+      "learning_rate": 9.975978968754045e-05,
+      "loss": 0.002,
       "step": 990
     },
     {
-      "epoch": 2.5392251040666025,
-      "grad_norm": 0.022700879722833633,
-      "learning_rate": 5.7268732000956304e-05,
-      "loss": 0.0096,
+      "epoch": 3.2032323232323234,
+      "grad_norm": 0.0063944170251488686,
+      "learning_rate": 9.975757246131803e-05,
+      "loss": 0.0017,
       "step": 991
     },
     {
-      "epoch": 2.54178674351585,
-      "grad_norm": 0.022096680477261543,
-      "learning_rate": 5.718016716847359e-05,
-      "loss": 0.0114,
+      "epoch": 3.2064646464646467,
+      "grad_norm": 0.005984049290418625,
+      "learning_rate": 9.975534507402671e-05,
+      "loss": 0.0018,
       "step": 992
     },
     {
-      "epoch": 2.5443483829650977,
-      "grad_norm": 0.01778617687523365,
-      "learning_rate": 5.709157932793394e-05,
-      "loss": 0.0061,
+      "epoch": 3.2096969696969695,
+      "grad_norm": 0.0061658755876123905,
+      "learning_rate": 9.975310752612137e-05,
+      "loss": 0.0012,
       "step": 993
     },
     {
-      "epoch": 2.546910022414345,
-      "grad_norm": 0.01681138202548027,
-      "learning_rate": 5.7002968763207364e-05,
-      "loss": 0.005,
+      "epoch": 3.2129292929292927,
+      "grad_norm": 0.007746810093522072,
+      "learning_rate": 9.975085981805897e-05,
+      "loss": 0.0021,
       "step": 994
     },
     {
-      "epoch": 2.549471661863593,
-      "grad_norm": 0.028049584478139877,
-      "learning_rate": 5.691433575823666e-05,
-      "loss": 0.0102,
+      "epoch": 3.216161616161616,
+      "grad_norm": 0.008047693409025669,
+      "learning_rate": 9.974860195029847e-05,
+      "loss": 0.0021,
       "step": 995
     },
     {
-      "epoch": 2.5520333013128402,
-      "grad_norm": 0.016301583498716354,
-      "learning_rate": 5.682568059703659e-05,
-      "loss": 0.0058,
+      "epoch": 3.2193939393939393,
+      "grad_norm": 0.009668510407209396,
+      "learning_rate": 9.974633392330097e-05,
+      "loss": 0.0022,
       "step": 996
     },
     {
-      "epoch": 2.5545949407620876,
-      "grad_norm": 0.016592754051089287,
-      "learning_rate": 5.6737003563692834e-05,
-      "loss": 0.0061,
+      "epoch": 3.2226262626262625,
+      "grad_norm": 0.007023985497653484,
+      "learning_rate": 9.974405573752965e-05,
+      "loss": 0.0018,
       "step": 997
     },
     {
-      "epoch": 2.5571565802113354,
-      "grad_norm": 0.019314013421535492,
-      "learning_rate": 5.664830494236126e-05,
-      "loss": 0.0083,
+      "epoch": 3.2258585858585858,
+      "grad_norm": 0.006368582136929035,
+      "learning_rate": 9.974176739344971e-05,
+      "loss": 0.0016,
       "step": 998
     },
     {
-      "epoch": 2.5597182196605828,
-      "grad_norm": 0.019145967438817024,
-      "learning_rate": 5.6559585017266826e-05,
-      "loss": 0.0069,
+      "epoch": 3.229090909090909,
+      "grad_norm": 0.00720590353012085,
+      "learning_rate": 9.973946889152847e-05,
+      "loss": 0.0015,
       "step": 999
     },
     {
-      "epoch": 2.56227985910983,
-      "grad_norm": 0.02137383259832859,
-      "learning_rate": 5.6470844072702764e-05,
-      "loss": 0.0076,
+      "epoch": 3.2323232323232323,
+      "grad_norm": 0.0065858145244419575,
+      "learning_rate": 9.973716023223531e-05,
+      "loss": 0.0018,
       "step": 1000
     },
     {
-      "epoch": 2.56227985910983,
-      "eval_loss": 0.0072718774899840355,
-      "eval_runtime": 15.766,
-      "eval_samples_per_second": 6.343,
-      "eval_steps_per_second": 1.586,
+      "epoch": 3.2323232323232323,
+      "eval_loss": 0.002038483275100589,
+      "eval_runtime": 15.6093,
+      "eval_samples_per_second": 6.406,
+      "eval_steps_per_second": 1.602,
       "step": 1000
     },
     {
-      "epoch": 2.564841498559078,
-      "grad_norm": 0.017583675682544708,
-      "learning_rate": 5.6382082393029746e-05,
-      "loss": 0.0059,
+      "epoch": 3.2355555555555555,
+      "grad_norm": 0.015130908228456974,
+      "learning_rate": 9.97348414160417e-05,
+      "loss": 0.0045,
       "step": 1001
     },
     {
-      "epoch": 2.5674031380083253,
-      "grad_norm": 0.018566062673926353,
-      "learning_rate": 5.6293300262674755e-05,
-      "loss": 0.0077,
+      "epoch": 3.2387878787878788,
+      "grad_norm": 0.009384035132825375,
+      "learning_rate": 9.973251244342114e-05,
+      "loss": 0.003,
       "step": 1002
     },
     {
-      "epoch": 2.5699647774575727,
-      "grad_norm": 0.02141863852739334,
-      "learning_rate": 5.62044979661304e-05,
-      "loss": 0.0081,
+      "epoch": 3.242020202020202,
+      "grad_norm": 0.0059922849759459496,
+      "learning_rate": 9.973017331484926e-05,
+      "loss": 0.0013,
       "step": 1003
     },
     {
-      "epoch": 2.5725264169068205,
-      "grad_norm": 0.01930375210940838,
-      "learning_rate": 5.611567578795392e-05,
-      "loss": 0.0067,
+      "epoch": 3.2452525252525253,
+      "grad_norm": 0.009899535216391087,
+      "learning_rate": 9.972782403080372e-05,
+      "loss": 0.002,
       "step": 1004
     },
     {
-      "epoch": 2.575088056356068,
-      "grad_norm": 0.025137510150671005,
-      "learning_rate": 5.602683401276615e-05,
-      "loss": 0.0087,
+      "epoch": 3.2484848484848485,
+      "grad_norm": 0.004808941390365362,
+      "learning_rate": 9.972546459176425e-05,
+      "loss": 0.0012,
       "step": 1005
     },
     {
-      "epoch": 2.5776496958053157,
-      "grad_norm": 0.021621698513627052,
-      "learning_rate": 5.593797292525087e-05,
-      "loss": 0.007,
+      "epoch": 3.251717171717172,
+      "grad_norm": 0.011834941804409027,
+      "learning_rate": 9.972309499821273e-05,
+      "loss": 0.0027,
       "step": 1006
     },
     {
-      "epoch": 2.580211335254563,
-      "grad_norm": 0.02559741400182247,
-      "learning_rate": 5.5849092810153666e-05,
-      "loss": 0.0116,
+      "epoch": 3.254949494949495,
+      "grad_norm": 0.00586290005594492,
+      "learning_rate": 9.972071525063303e-05,
+      "loss": 0.0014,
       "step": 1007
     },
     {
-      "epoch": 2.5827729747038104,
-      "grad_norm": 0.022138088941574097,
-      "learning_rate": 5.5760193952281056e-05,
-      "loss": 0.0083,
+      "epoch": 3.2581818181818183,
+      "grad_norm": 0.007359727751463652,
+      "learning_rate": 9.971832534951108e-05,
+      "loss": 0.0017,
       "step": 1008
     },
     {
-      "epoch": 2.5853346141530578,
-      "grad_norm": 0.01950952783226967,
-      "learning_rate": 5.5671276636499714e-05,
-      "loss": 0.0056,
+      "epoch": 3.2614141414141415,
+      "grad_norm": 0.014025827869772911,
+      "learning_rate": 9.9715925295335e-05,
+      "loss": 0.0065,
       "step": 1009
     },
     {
-      "epoch": 2.5878962536023056,
-      "grad_norm": 0.02397429198026657,
-      "learning_rate": 5.55823411477354e-05,
-      "loss": 0.0105,
+      "epoch": 3.264646464646465,
+      "grad_norm": 0.009781260043382645,
+      "learning_rate": 9.971351508859488e-05,
+      "loss": 0.0018,
       "step": 1010
     },
     {
-      "epoch": 2.590457893051553,
-      "grad_norm": 0.014748901128768921,
-      "learning_rate": 5.549338777097208e-05,
-      "loss": 0.0053,
+      "epoch": 3.267878787878788,
+      "grad_norm": 0.003897070651873946,
+      "learning_rate": 9.971109472978288e-05,
+      "loss": 0.0008,
       "step": 1011
     },
     {
-      "epoch": 2.5930195325008008,
-      "grad_norm": 0.02159014530479908,
-      "learning_rate": 5.540441679125112e-05,
-      "loss": 0.0071,
+      "epoch": 3.2711111111111113,
+      "grad_norm": 0.006243659649044275,
+      "learning_rate": 9.97086642193933e-05,
+      "loss": 0.0018,
       "step": 1012
     },
     {
-      "epoch": 2.595581171950048,
-      "grad_norm": 0.023283282294869423,
-      "learning_rate": 5.531542849367026e-05,
-      "loss": 0.0056,
+      "epoch": 3.274343434343434,
+      "grad_norm": 0.004899726249277592,
+      "learning_rate": 9.970622355792247e-05,
+      "loss": 0.0013,
       "step": 1013
     },
     {
-      "epoch": 2.5981428113992955,
-      "grad_norm": 0.019547423347830772,
-      "learning_rate": 5.522642316338268e-05,
-      "loss": 0.0066,
+      "epoch": 3.2775757575757574,
+      "grad_norm": 0.0059542288072407246,
+      "learning_rate": 9.970377274586879e-05,
+      "loss": 0.0011,
       "step": 1014
     },
     {
-      "epoch": 2.600704450848543,
-      "grad_norm": 0.02325627952814102,
-      "learning_rate": 5.513740108559622e-05,
-      "loss": 0.0056,
+      "epoch": 3.2808080808080806,
+      "grad_norm": 0.008120191283524036,
+      "learning_rate": 9.970131178373277e-05,
+      "loss": 0.0024,
       "step": 1015
     },
     {
-      "epoch": 2.6032660902977907,
-      "grad_norm": 0.015685738995671272,
-      "learning_rate": 5.5048362545572304e-05,
-      "loss": 0.0046,
+      "epoch": 3.284040404040404,
+      "grad_norm": 0.006169213447719812,
+      "learning_rate": 9.969884067201695e-05,
+      "loss": 0.0018,
       "step": 1016
     },
     {
-      "epoch": 2.605827729747038,
-      "grad_norm": 0.027276068925857544,
-      "learning_rate": 5.495930782862522e-05,
-      "loss": 0.0093,
+      "epoch": 3.287272727272727,
+      "grad_norm": 0.007144790142774582,
+      "learning_rate": 9.969635941122595e-05,
+      "loss": 0.0014,
       "step": 1017
     },
     {
-      "epoch": 2.608389369196286,
-      "grad_norm": 0.019772149622440338,
-      "learning_rate": 5.4870237220120976e-05,
-      "loss": 0.0056,
+      "epoch": 3.2905050505050504,
+      "grad_norm": 0.006319798994809389,
+      "learning_rate": 9.969386800186649e-05,
+      "loss": 0.0017,
       "step": 1018
     },
     {
-      "epoch": 2.610951008645533,
-      "grad_norm": 0.016364077106118202,
-      "learning_rate": 5.478115100547656e-05,
-      "loss": 0.0056,
+      "epoch": 3.2937373737373736,
+      "grad_norm": 0.006894100923091173,
+      "learning_rate": 9.969136644444731e-05,
+      "loss": 0.0016,
       "step": 1019
     },
     {
-      "epoch": 2.6135126480947806,
-      "grad_norm": 0.026881124824285507,
-      "learning_rate": 5.469204947015897e-05,
-      "loss": 0.0101,
+      "epoch": 3.296969696969697,
+      "grad_norm": 0.006226851139217615,
+      "learning_rate": 9.968885473947932e-05,
+      "loss": 0.0019,
       "step": 1020
     },
     {
-      "epoch": 2.616074287544028,
-      "grad_norm": 0.03398275375366211,
-      "learning_rate": 5.460293289968429e-05,
-      "loss": 0.0081,
+      "epoch": 3.30020202020202,
+      "grad_norm": 0.01570131815969944,
+      "learning_rate": 9.968633288747539e-05,
+      "loss": 0.0022,
       "step": 1021
     },
     {
-      "epoch": 2.6186359269932757,
-      "grad_norm": 0.028755880892276764,
-      "learning_rate": 5.451380157961676e-05,
-      "loss": 0.0085,
+      "epoch": 3.3034343434343434,
+      "grad_norm": 0.00846484862267971,
+      "learning_rate": 9.968380088895052e-05,
+      "loss": 0.0026,
       "step": 1022
     },
     {
-      "epoch": 2.621197566442523,
-      "grad_norm": 0.01581442728638649,
-      "learning_rate": 5.442465579556793e-05,
-      "loss": 0.0055,
+      "epoch": 3.3066666666666666,
+      "grad_norm": 0.007999323308467865,
+      "learning_rate": 9.968125874442179e-05,
+      "loss": 0.0021,
       "step": 1023
     },
     {
-      "epoch": 2.623759205891771,
-      "grad_norm": 0.01677510514855385,
-      "learning_rate": 5.433549583319565e-05,
-      "loss": 0.0051,
+      "epoch": 3.30989898989899,
+      "grad_norm": 0.007302125450223684,
+      "learning_rate": 9.96787064544083e-05,
+      "loss": 0.0023,
       "step": 1024
     },
     {
-      "epoch": 2.6263208453410183,
-      "grad_norm": 0.01671496219933033,
-      "learning_rate": 5.424632197820324e-05,
-      "loss": 0.0054,
+      "epoch": 3.313131313131313,
+      "grad_norm": 0.008102331310510635,
+      "learning_rate": 9.96761440194313e-05,
+      "loss": 0.0022,
       "step": 1025
     },
     {
-      "epoch": 2.6263208453410183,
-      "eval_loss": 0.0073104496113955975,
-      "eval_runtime": 15.8028,
-      "eval_samples_per_second": 6.328,
-      "eval_steps_per_second": 1.582,
+      "epoch": 3.313131313131313,
+      "eval_loss": 0.0019420969765633345,
+      "eval_runtime": 15.6195,
+      "eval_samples_per_second": 6.402,
+      "eval_steps_per_second": 1.601,
       "step": 1025
     },
     {
-      "epoch": 2.6288824847902657,
-      "grad_norm": 0.013602949678897858,
-      "learning_rate": 5.415713451633853e-05,
-      "loss": 0.0042,
+      "epoch": 3.3163636363636364,
+      "grad_norm": 0.00843795482069254,
+      "learning_rate": 9.967357144001403e-05,
+      "loss": 0.0027,
       "step": 1026
     },
     {
-      "epoch": 2.631444124239513,
-      "grad_norm": 0.02331901527941227,
-      "learning_rate": 5.4067933733392915e-05,
-      "loss": 0.0092,
+      "epoch": 3.3195959595959597,
+      "grad_norm": 0.007327123545110226,
+      "learning_rate": 9.967098871668186e-05,
+      "loss": 0.0026,
       "step": 1027
     },
     {
-      "epoch": 2.634005763688761,
-      "grad_norm": 0.027525432407855988,
-      "learning_rate": 5.397871991520054e-05,
-      "loss": 0.0103,
+      "epoch": 3.322828282828283,
+      "grad_norm": 0.005961123388260603,
+      "learning_rate": 9.966839584996222e-05,
+      "loss": 0.0016,
       "step": 1028
     },
     {
-      "epoch": 2.636567403138008,
-      "grad_norm": 0.025305457413196564,
-      "learning_rate": 5.388949334763724e-05,
-      "loss": 0.0096,
+      "epoch": 3.326060606060606,
+      "grad_norm": 0.007074032910168171,
+      "learning_rate": 9.96657928403846e-05,
+      "loss": 0.0022,
       "step": 1029
     },
     {
-      "epoch": 2.639129042587256,
-      "grad_norm": 0.017113184556365013,
-      "learning_rate": 5.3800254316619806e-05,
-      "loss": 0.0047,
+      "epoch": 3.3292929292929294,
+      "grad_norm": 0.007795639336109161,
+      "learning_rate": 9.966317968848054e-05,
+      "loss": 0.0024,
       "step": 1030
     },
     {
-      "epoch": 2.6416906820365034,
-      "grad_norm": 0.02039264515042305,
-      "learning_rate": 5.371100310810486e-05,
-      "loss": 0.0064,
+      "epoch": 3.3325252525252527,
+      "grad_norm": 0.007937250658869743,
+      "learning_rate": 9.966055639478369e-05,
+      "loss": 0.0026,
       "step": 1031
     },
     {
-      "epoch": 2.6442523214857507,
-      "grad_norm": 0.01747237518429756,
-      "learning_rate": 5.3621740008088126e-05,
-      "loss": 0.0064,
+      "epoch": 3.335757575757576,
+      "grad_norm": 0.006392105482518673,
+      "learning_rate": 9.965792295982978e-05,
+      "loss": 0.0015,
       "step": 1032
     },
     {
-      "epoch": 2.6468139609349985,
-      "grad_norm": 0.0253047663718462,
-      "learning_rate": 5.353246530260339e-05,
-      "loss": 0.0084,
+      "epoch": 3.338989898989899,
+      "grad_norm": 0.006848600227385759,
+      "learning_rate": 9.965527938415655e-05,
+      "loss": 0.0016,
       "step": 1033
     },
     {
-      "epoch": 2.649375600384246,
-      "grad_norm": 0.01626812294125557,
-      "learning_rate": 5.344317927772163e-05,
-      "loss": 0.0053,
+      "epoch": 3.3422222222222224,
+      "grad_norm": 0.00761762447655201,
+      "learning_rate": 9.965262566830388e-05,
+      "loss": 0.0021,
       "step": 1034
     },
     {
-      "epoch": 2.6519372398334933,
-      "grad_norm": 0.042853761464357376,
-      "learning_rate": 5.335388221955012e-05,
-      "loss": 0.0064,
+      "epoch": 3.3454545454545457,
+      "grad_norm": 0.005903237033635378,
+      "learning_rate": 9.964996181281367e-05,
+      "loss": 0.0012,
       "step": 1035
     },
     {
-      "epoch": 2.654498879282741,
-      "grad_norm": 0.017256347462534904,
-      "learning_rate": 5.3264574414231475e-05,
-      "loss": 0.0055,
+      "epoch": 3.348686868686869,
+      "grad_norm": 0.008324799127876759,
+      "learning_rate": 9.964728781822992e-05,
+      "loss": 0.0018,
       "step": 1036
     },
     {
-      "epoch": 2.6570605187319885,
-      "grad_norm": 0.020623289048671722,
-      "learning_rate": 5.317525614794273e-05,
-      "loss": 0.0083,
+      "epoch": 3.3519191919191917,
+      "grad_norm": 0.00987184513360262,
+      "learning_rate": 9.964460368509867e-05,
+      "loss": 0.0023,
       "step": 1037
     },
     {
-      "epoch": 2.659622158181236,
-      "grad_norm": 0.018989235162734985,
-      "learning_rate": 5.308592770689448e-05,
-      "loss": 0.0067,
+      "epoch": 3.355151515151515,
+      "grad_norm": 0.009272204712033272,
+      "learning_rate": 9.964190941396808e-05,
+      "loss": 0.0023,
       "step": 1038
     },
     {
-      "epoch": 2.6621837976304836,
-      "grad_norm": 0.014944621361792088,
-      "learning_rate": 5.299658937732988e-05,
-      "loss": 0.0055,
+      "epoch": 3.3583838383838382,
+      "grad_norm": 0.01208008173853159,
+      "learning_rate": 9.963920500538834e-05,
+      "loss": 0.0036,
       "step": 1039
     },
     {
-      "epoch": 2.664745437079731,
-      "grad_norm": 0.01713753305375576,
-      "learning_rate": 5.290724144552379e-05,
-      "loss": 0.0053,
+      "epoch": 3.3616161616161615,
+      "grad_norm": 0.007658164016902447,
+      "learning_rate": 9.963649045991173e-05,
+      "loss": 0.0019,
       "step": 1040
     },
     {
-      "epoch": 2.667307076528979,
-      "grad_norm": 0.027423840016126633,
-      "learning_rate": 5.281788419778187e-05,
-      "loss": 0.0091,
+      "epoch": 3.3648484848484848,
+      "grad_norm": 0.010208998806774616,
+      "learning_rate": 9.963376577809256e-05,
+      "loss": 0.0025,
       "step": 1041
     },
     {
-      "epoch": 2.669868715978226,
-      "grad_norm": 0.02789049781858921,
-      "learning_rate": 5.272851792043959e-05,
-      "loss": 0.0086,
+      "epoch": 3.368080808080808,
+      "grad_norm": 0.007051629014313221,
+      "learning_rate": 9.963103096048728e-05,
+      "loss": 0.0026,
       "step": 1042
     },
     {
-      "epoch": 2.6724303554274735,
-      "grad_norm": 0.01847296766936779,
-      "learning_rate": 5.263914289986136e-05,
-      "loss": 0.0067,
+      "epoch": 3.3713131313131313,
+      "grad_norm": 0.011365561746060848,
+      "learning_rate": 9.962828600765433e-05,
+      "loss": 0.0023,
       "step": 1043
     },
     {
-      "epoch": 2.674991994876721,
-      "grad_norm": 0.016148822382092476,
-      "learning_rate": 5.254975942243963e-05,
-      "loss": 0.0043,
+      "epoch": 3.3745454545454545,
+      "grad_norm": 0.005724266171455383,
+      "learning_rate": 9.96255309201543e-05,
+      "loss": 0.0015,
       "step": 1044
     },
     {
-      "epoch": 2.6775536343259687,
-      "grad_norm": 0.02496311441063881,
-      "learning_rate": 5.246036777459391e-05,
-      "loss": 0.0076,
+      "epoch": 3.3777777777777778,
+      "grad_norm": 0.008675566874444485,
+      "learning_rate": 9.962276569854977e-05,
+      "loss": 0.0022,
       "step": 1045
     },
     {
-      "epoch": 2.680115273775216,
-      "grad_norm": 0.014748088084161282,
-      "learning_rate": 5.2370968242769946e-05,
-      "loss": 0.0044,
+      "epoch": 3.381010101010101,
+      "grad_norm": 0.006316622719168663,
+      "learning_rate": 9.96199903434055e-05,
+      "loss": 0.0019,
       "step": 1046
     },
     {
-      "epoch": 2.682676913224464,
-      "grad_norm": 0.01946377381682396,
-      "learning_rate": 5.22815611134387e-05,
-      "loss": 0.0053,
+      "epoch": 3.3842424242424243,
+      "grad_norm": 0.006571315228939056,
+      "learning_rate": 9.961720485528819e-05,
+      "loss": 0.0021,
       "step": 1047
     },
     {
-      "epoch": 2.6852385526737113,
-      "grad_norm": 0.026910366490483284,
-      "learning_rate": 5.219214667309551e-05,
-      "loss": 0.0087,
+      "epoch": 3.3874747474747475,
+      "grad_norm": 0.006140459794551134,
+      "learning_rate": 9.961440923476666e-05,
+      "loss": 0.0018,
       "step": 1048
     },
     {
-      "epoch": 2.6878001921229586,
-      "grad_norm": 0.01077069342136383,
-      "learning_rate": 5.210272520825911e-05,
-      "loss": 0.0035,
+      "epoch": 3.390707070707071,
+      "grad_norm": 0.004743719473481178,
+      "learning_rate": 9.961160348241185e-05,
+      "loss": 0.0013,
       "step": 1049
     },
     {
-      "epoch": 2.690361831572206,
-      "grad_norm": 0.02103227563202381,
-      "learning_rate": 5.201329700547076e-05,
-      "loss": 0.01,
+      "epoch": 3.393939393939394,
+      "grad_norm": 0.006233834661543369,
+      "learning_rate": 9.96087875987967e-05,
+      "loss": 0.0018,
       "step": 1050
     },
     {
-      "epoch": 2.690361831572206,
-      "eval_loss": 0.006800453644245863,
-      "eval_runtime": 15.7754,
-      "eval_samples_per_second": 6.339,
-      "eval_steps_per_second": 1.585,
+      "epoch": 3.393939393939394,
+      "eval_loss": 0.0018019942799583077,
+      "eval_runtime": 15.6183,
+      "eval_samples_per_second": 6.403,
+      "eval_steps_per_second": 1.601,
       "step": 1050
     },
     {
-      "epoch": 2.692923471021454,
-      "grad_norm": 0.01600644364953041,
-      "learning_rate": 5.192386235129332e-05,
-      "loss": 0.0058,
+      "epoch": 3.3971717171717173,
+      "grad_norm": 0.006021643057465553,
+      "learning_rate": 9.960596158449627e-05,
+      "loss": 0.0014,
       "step": 1051
     },
     {
-      "epoch": 2.695485110470701,
-      "grad_norm": 0.01561508234590292,
-      "learning_rate": 5.183442153231034e-05,
-      "loss": 0.0055,
+      "epoch": 3.4004040404040405,
+      "grad_norm": 0.00728834280744195,
+      "learning_rate": 9.960312544008763e-05,
+      "loss": 0.0018,
       "step": 1052
     },
     {
-      "epoch": 2.698046749919949,
-      "grad_norm": 0.014669963158667088,
-      "learning_rate": 5.174497483512506e-05,
-      "loss": 0.0046,
+      "epoch": 3.403636363636364,
+      "grad_norm": 0.009629890322685242,
+      "learning_rate": 9.960027916614998e-05,
+      "loss": 0.0027,
       "step": 1053
     },
     {
-      "epoch": 2.7006083893691963,
-      "grad_norm": 0.02971458062529564,
-      "learning_rate": 5.165552254635959e-05,
-      "loss": 0.0115,
+      "epoch": 3.406868686868687,
+      "grad_norm": 0.007169786375015974,
+      "learning_rate": 9.959742276326456e-05,
+      "loss": 0.0021,
       "step": 1054
     },
     {
-      "epoch": 2.7031700288184437,
-      "grad_norm": 0.022039420902729034,
-      "learning_rate": 5.1566064952654014e-05,
-      "loss": 0.008,
+      "epoch": 3.41010101010101,
+      "grad_norm": 0.012614664621651173,
+      "learning_rate": 9.959455623201465e-05,
+      "loss": 0.0028,
       "step": 1055
     },
     {
-      "epoch": 2.705731668267691,
-      "grad_norm": 0.02216564118862152,
-      "learning_rate": 5.147660234066533e-05,
-      "loss": 0.0097,
+      "epoch": 3.413333333333333,
+      "grad_norm": 0.007051707245409489,
+      "learning_rate": 9.959167957298568e-05,
+      "loss": 0.0017,
       "step": 1056
     },
     {
-      "epoch": 2.708293307716939,
-      "grad_norm": 0.02106768824160099,
-      "learning_rate": 5.1387134997066634e-05,
-      "loss": 0.008,
+      "epoch": 3.4165656565656564,
+      "grad_norm": 0.006615843158215284,
+      "learning_rate": 9.958879278676506e-05,
+      "loss": 0.0016,
       "step": 1057
     },
     {
-      "epoch": 2.7108549471661862,
-      "grad_norm": 0.01805659383535385,
-      "learning_rate": 5.1297663208546234e-05,
-      "loss": 0.0061,
+      "epoch": 3.4197979797979796,
+      "grad_norm": 0.004922749008983374,
+      "learning_rate": 9.958589587394231e-05,
+      "loss": 0.0013,
       "step": 1058
     },
     {
-      "epoch": 2.713416586615434,
-      "grad_norm": 0.024404456838965416,
-      "learning_rate": 5.1208187261806615e-05,
-      "loss": 0.0077,
+      "epoch": 3.423030303030303,
+      "grad_norm": 0.005267519038170576,
+      "learning_rate": 9.958298883510903e-05,
+      "loss": 0.0014,
       "step": 1059
     },
     {
-      "epoch": 2.7159782260646814,
-      "grad_norm": 0.023735761642456055,
-      "learning_rate": 5.1118707443563665e-05,
-      "loss": 0.0105,
+      "epoch": 3.426262626262626,
+      "grad_norm": 0.00552243972197175,
+      "learning_rate": 9.958007167085886e-05,
+      "loss": 0.0017,
       "step": 1060
     },
     {
-      "epoch": 2.718539865513929,
-      "grad_norm": 0.022548731416463852,
-      "learning_rate": 5.1029224040545585e-05,
-      "loss": 0.0058,
+      "epoch": 3.4294949494949494,
+      "grad_norm": 0.005605652462691069,
+      "learning_rate": 9.95771443817875e-05,
+      "loss": 0.0018,
       "step": 1061
     },
     {
-      "epoch": 2.7211015049631766,
-      "grad_norm": 0.017163462936878204,
-      "learning_rate": 5.0939737339492124e-05,
-      "loss": 0.0049,
+      "epoch": 3.4327272727272726,
+      "grad_norm": 0.005629305727779865,
+      "learning_rate": 9.957420696849275e-05,
+      "loss": 0.0015,
       "step": 1062
     },
     {
-      "epoch": 2.723663144412424,
-      "grad_norm": 0.021907225251197815,
-      "learning_rate": 5.08502476271536e-05,
-      "loss": 0.0085,
+      "epoch": 3.435959595959596,
+      "grad_norm": 0.008712087757885456,
+      "learning_rate": 9.957125943157448e-05,
+      "loss": 0.0016,
       "step": 1063
     },
     {
-      "epoch": 2.7262247838616713,
-      "grad_norm": 0.01215510256588459,
-      "learning_rate": 5.0760755190289964e-05,
-      "loss": 0.0036,
+      "epoch": 3.439191919191919,
+      "grad_norm": 0.008017695508897305,
+      "learning_rate": 9.956830177163461e-05,
+      "loss": 0.0024,
       "step": 1064
     },
     {
-      "epoch": 2.728786423310919,
-      "grad_norm": 0.015955861657857895,
-      "learning_rate": 5.0671260315669875e-05,
-      "loss": 0.0048,
+      "epoch": 3.4424242424242424,
+      "grad_norm": 0.006030660588294268,
+      "learning_rate": 9.95653339892771e-05,
+      "loss": 0.0016,
       "step": 1065
     },
     {
-      "epoch": 2.7313480627601665,
-      "grad_norm": 0.01937129907310009,
-      "learning_rate": 5.0581763290069865e-05,
-      "loss": 0.0068,
+      "epoch": 3.4456565656565656,
+      "grad_norm": 0.00810298416763544,
+      "learning_rate": 9.956235608510802e-05,
+      "loss": 0.0023,
       "step": 1066
     },
     {
-      "epoch": 2.733909702209414,
-      "grad_norm": 0.02090214006602764,
-      "learning_rate": 5.049226440027328e-05,
-      "loss": 0.0073,
+      "epoch": 3.448888888888889,
+      "grad_norm": 0.00831377599388361,
+      "learning_rate": 9.95593680597355e-05,
+      "loss": 0.002,
       "step": 1067
     },
     {
-      "epoch": 2.7364713416586617,
-      "grad_norm": 0.021622899919748306,
-      "learning_rate": 5.0402763933069496e-05,
-      "loss": 0.007,
+      "epoch": 3.452121212121212,
+      "grad_norm": 0.007975469343364239,
+      "learning_rate": 9.955636991376971e-05,
+      "loss": 0.0017,
       "step": 1068
     },
     {
-      "epoch": 2.739032981107909,
-      "grad_norm": 0.01952539198100567,
-      "learning_rate": 5.031326217525296e-05,
-      "loss": 0.0051,
+      "epoch": 3.4553535353535354,
+      "grad_norm": 0.0072644660249352455,
+      "learning_rate": 9.955336164782292e-05,
+      "loss": 0.0018,
       "step": 1069
     },
     {
-      "epoch": 2.7415946205571564,
-      "grad_norm": 0.01868268847465515,
-      "learning_rate": 5.022375941362217e-05,
-      "loss": 0.0058,
+      "epoch": 3.4585858585858587,
+      "grad_norm": 0.009866764768958092,
+      "learning_rate": 9.955034326250946e-05,
+      "loss": 0.003,
       "step": 1070
     },
     {
-      "epoch": 2.7441562600064042,
-      "grad_norm": 0.022691085934638977,
-      "learning_rate": 5.0134255934978924e-05,
-      "loss": 0.0069,
+      "epoch": 3.461818181818182,
+      "grad_norm": 0.009485723450779915,
+      "learning_rate": 9.954731475844571e-05,
+      "loss": 0.0019,
       "step": 1071
     },
     {
-      "epoch": 2.7467178994556516,
-      "grad_norm": 0.025523167103528976,
-      "learning_rate": 5.004475202612731e-05,
-      "loss": 0.0076,
+      "epoch": 3.465050505050505,
+      "grad_norm": 0.01677337847650051,
+      "learning_rate": 9.954427613625013e-05,
+      "loss": 0.0016,
       "step": 1072
     },
     {
-      "epoch": 2.749279538904899,
-      "grad_norm": 0.018498631194233894,
-      "learning_rate": 4.995524797387272e-05,
-      "loss": 0.0059,
+      "epoch": 3.4682828282828284,
+      "grad_norm": 0.008975732140243053,
+      "learning_rate": 9.95412273965432e-05,
+      "loss": 0.0021,
       "step": 1073
     },
     {
-      "epoch": 2.7518411783541468,
-      "grad_norm": 0.015959061682224274,
-      "learning_rate": 4.986574406502108e-05,
-      "loss": 0.0053,
+      "epoch": 3.4715151515151517,
+      "grad_norm": 0.008125899359583855,
+      "learning_rate": 9.953816853994759e-05,
+      "loss": 0.0021,
       "step": 1074
     },
     {
-      "epoch": 2.754402817803394,
-      "grad_norm": 0.036607638001441956,
-      "learning_rate": 4.977624058637783e-05,
-      "loss": 0.0064,
+      "epoch": 3.474747474747475,
+      "grad_norm": 0.007210091687738895,
+      "learning_rate": 9.953509956708789e-05,
+      "loss": 0.002,
       "step": 1075
     },
     {
-      "epoch": 2.754402817803394,
-      "eval_loss": 0.007549135480076075,
-      "eval_runtime": 15.7053,
-      "eval_samples_per_second": 6.367,
-      "eval_steps_per_second": 1.592,
+      "epoch": 3.474747474747475,
+      "eval_loss": 0.001874809619039297,
+      "eval_runtime": 15.588,
+      "eval_samples_per_second": 6.415,
+      "eval_steps_per_second": 1.604,
       "step": 1075
     },
     {
-      "epoch": 2.756964457252642,
-      "grad_norm": 0.01559431292116642,
-      "learning_rate": 4.9686737824747064e-05,
-      "loss": 0.0052,
+      "epoch": 3.477979797979798,
+      "grad_norm": 0.010867740027606487,
+      "learning_rate": 9.953202047859085e-05,
+      "loss": 0.0027,
       "step": 1076
     },
     {
-      "epoch": 2.7595260967018893,
-      "grad_norm": 0.017336146906018257,
-      "learning_rate": 4.9597236066930516e-05,
-      "loss": 0.0053,
+      "epoch": 3.4812121212121214,
+      "grad_norm": 0.010386588983237743,
+      "learning_rate": 9.952893127508522e-05,
+      "loss": 0.0025,
       "step": 1077
     },
     {
-      "epoch": 2.7620877361511367,
-      "grad_norm": 0.0237768292427063,
-      "learning_rate": 4.9507735599726726e-05,
-      "loss": 0.009,
+      "epoch": 3.4844444444444447,
+      "grad_norm": 0.008947139605879784,
+      "learning_rate": 9.95258319572019e-05,
+      "loss": 0.0017,
       "step": 1078
     },
     {
-      "epoch": 2.764649375600384,
-      "grad_norm": 0.016704855486750603,
-      "learning_rate": 4.941823670993016e-05,
-      "loss": 0.0056,
+      "epoch": 3.4876767676767675,
+      "grad_norm": 0.01362928282469511,
+      "learning_rate": 9.952272252557378e-05,
+      "loss": 0.0012,
       "step": 1079
     },
     {
-      "epoch": 2.767211015049632,
-      "grad_norm": 0.02107900008559227,
-      "learning_rate": 4.9328739684330137e-05,
-      "loss": 0.0056,
+      "epoch": 3.4909090909090907,
+      "grad_norm": 0.007846825756132603,
+      "learning_rate": 9.951960298083583e-05,
+      "loss": 0.0012,
       "step": 1080
     },
     {
-      "epoch": 2.769772654498879,
-      "grad_norm": 0.01982649229466915,
-      "learning_rate": 4.923924480971004e-05,
-      "loss": 0.0051,
+      "epoch": 3.494141414141414,
+      "grad_norm": 0.00865482073277235,
+      "learning_rate": 9.95164733236251e-05,
+      "loss": 0.0029,
       "step": 1081
     },
     {
-      "epoch": 2.772334293948127,
-      "grad_norm": 0.017537608742713928,
-      "learning_rate": 4.9149752372846404e-05,
-      "loss": 0.0055,
+      "epoch": 3.4973737373737372,
+      "grad_norm": 0.007692219689488411,
+      "learning_rate": 9.951333355458072e-05,
+      "loss": 0.0023,
       "step": 1082
     },
     {
-      "epoch": 2.7748959333973744,
-      "grad_norm": 0.02003137394785881,
-      "learning_rate": 4.906026266050789e-05,
-      "loss": 0.006,
+      "epoch": 3.5006060606060605,
+      "grad_norm": 0.007061603479087353,
+      "learning_rate": 9.951018367434386e-05,
+      "loss": 0.0022,
       "step": 1083
     },
     {
-      "epoch": 2.7774575728466218,
-      "grad_norm": 0.02151995524764061,
-      "learning_rate": 4.8970775959454427e-05,
-      "loss": 0.0076,
+      "epoch": 3.5038383838383838,
+      "grad_norm": 0.004897949285805225,
+      "learning_rate": 9.950702368355775e-05,
+      "loss": 0.001,
       "step": 1084
     },
     {
-      "epoch": 2.780019212295869,
-      "grad_norm": 0.01217377558350563,
-      "learning_rate": 4.8881292556436354e-05,
-      "loss": 0.0037,
+      "epoch": 3.507070707070707,
+      "grad_norm": 0.011754898354411125,
+      "learning_rate": 9.950385358286772e-05,
+      "loss": 0.0047,
       "step": 1085
     },
     {
-      "epoch": 2.782580851745117,
-      "grad_norm": 0.018807468935847282,
-      "learning_rate": 4.87918127381934e-05,
-      "loss": 0.0058,
+      "epoch": 3.5103030303030303,
+      "grad_norm": 0.009107725694775581,
+      "learning_rate": 9.950067337292112e-05,
+      "loss": 0.0027,
       "step": 1086
     },
     {
-      "epoch": 2.7851424911943643,
-      "grad_norm": 0.02241814136505127,
-      "learning_rate": 4.870233679145378e-05,
-      "loss": 0.0073,
+      "epoch": 3.5135353535353535,
+      "grad_norm": 0.006025637499988079,
+      "learning_rate": 9.949748305436741e-05,
+      "loss": 0.0023,
       "step": 1087
     },
     {
-      "epoch": 2.787704130643612,
-      "grad_norm": 0.014323737472295761,
-      "learning_rate": 4.861286500293338e-05,
-      "loss": 0.0047,
+      "epoch": 3.5167676767676768,
+      "grad_norm": 0.0053115105256438255,
+      "learning_rate": 9.949428262785805e-05,
+      "loss": 0.0011,
       "step": 1088
     },
     {
-      "epoch": 2.7902657700928595,
-      "grad_norm": 0.01725595071911812,
-      "learning_rate": 4.8523397659334704e-05,
-      "loss": 0.005,
+      "epoch": 3.52,
+      "grad_norm": 0.005652850028127432,
+      "learning_rate": 9.949107209404665e-05,
+      "loss": 0.0015,
       "step": 1089
     },
     {
-      "epoch": 2.792827409542107,
-      "grad_norm": 0.032416146248579025,
-      "learning_rate": 4.8433935047346e-05,
-      "loss": 0.009,
+      "epoch": 3.5232323232323233,
+      "grad_norm": 0.005916878115385771,
+      "learning_rate": 9.948785145358879e-05,
+      "loss": 0.0019,
       "step": 1090
     },
     {
-      "epoch": 2.795389048991354,
-      "grad_norm": 0.02138088457286358,
-      "learning_rate": 4.834447745364041e-05,
-      "loss": 0.0064,
+      "epoch": 3.5264646464646465,
+      "grad_norm": 0.007432945538312197,
+      "learning_rate": 9.948462070714219e-05,
+      "loss": 0.002,
       "step": 1091
     },
     {
-      "epoch": 2.797950688440602,
-      "grad_norm": 0.012949190102517605,
-      "learning_rate": 4.825502516487497e-05,
-      "loss": 0.0036,
+      "epoch": 3.5296969696969698,
+      "grad_norm": 0.006576692219823599,
+      "learning_rate": 9.948137985536662e-05,
+      "loss": 0.0016,
       "step": 1092
     },
     {
-      "epoch": 2.8005123278898494,
-      "grad_norm": 0.021004170179367065,
-      "learning_rate": 4.8165578467689674e-05,
-      "loss": 0.0063,
+      "epoch": 3.532929292929293,
+      "grad_norm": 0.005020272918045521,
+      "learning_rate": 9.947812889892387e-05,
+      "loss": 0.0017,
       "step": 1093
     },
     {
-      "epoch": 2.803073967339097,
-      "grad_norm": 0.02376505918800831,
-      "learning_rate": 4.8076137648706674e-05,
-      "loss": 0.0067,
+      "epoch": 3.5361616161616163,
+      "grad_norm": 0.004623802378773689,
+      "learning_rate": 9.947486783847784e-05,
+      "loss": 0.0011,
       "step": 1094
     },
     {
-      "epoch": 2.8056356067883446,
-      "grad_norm": 0.01954122632741928,
-      "learning_rate": 4.798670299452926e-05,
-      "loss": 0.0054,
+      "epoch": 3.5393939393939395,
+      "grad_norm": 0.01020809356123209,
+      "learning_rate": 9.947159667469446e-05,
+      "loss": 0.003,
       "step": 1095
     },
     {
-      "epoch": 2.808197246237592,
-      "grad_norm": 0.019754689186811447,
-      "learning_rate": 4.78972747917409e-05,
-      "loss": 0.0073,
+      "epoch": 3.542626262626263,
+      "grad_norm": 0.008039414882659912,
+      "learning_rate": 9.946831540824175e-05,
+      "loss": 0.0017,
       "step": 1096
     },
     {
-      "epoch": 2.8107588856868397,
-      "grad_norm": 0.02440563589334488,
-      "learning_rate": 4.7807853326904504e-05,
-      "loss": 0.0072,
+      "epoch": 3.5458585858585856,
+      "grad_norm": 0.0057967244647443295,
+      "learning_rate": 9.94650240397898e-05,
+      "loss": 0.0019,
       "step": 1097
     },
     {
-      "epoch": 2.813320525136087,
-      "grad_norm": 0.027994103729724884,
-      "learning_rate": 4.7718438886561316e-05,
-      "loss": 0.0098,
+      "epoch": 3.549090909090909,
+      "grad_norm": 0.01112373173236847,
+      "learning_rate": 9.946172257001069e-05,
+      "loss": 0.0037,
       "step": 1098
     },
     {
-      "epoch": 2.8158821645853345,
-      "grad_norm": 0.021733155474066734,
-      "learning_rate": 4.7629031757230066e-05,
-      "loss": 0.0073,
+      "epoch": 3.552323232323232,
+      "grad_norm": 0.009685222990810871,
+      "learning_rate": 9.945841099957869e-05,
+      "loss": 0.0035,
       "step": 1099
     },
     {
-      "epoch": 2.8184438040345823,
-      "grad_norm": 0.01901175267994404,
-      "learning_rate": 4.7539632225406095e-05,
-      "loss": 0.0066,
+      "epoch": 3.5555555555555554,
+      "grad_norm": 0.008019673638045788,
+      "learning_rate": 9.945508932917001e-05,
+      "loss": 0.0019,
       "step": 1100
     },
     {
-      "epoch": 2.8184438040345823,
-      "eval_loss": 0.007056466303765774,
-      "eval_runtime": 15.7048,
-      "eval_samples_per_second": 6.367,
-      "eval_steps_per_second": 1.592,
+      "epoch": 3.5555555555555554,
+      "eval_loss": 0.0019529261626303196,
+      "eval_runtime": 15.632,
+      "eval_samples_per_second": 6.397,
+      "eval_steps_per_second": 1.599,
       "step": 1100
     },
     {
-      "epoch": 2.8210054434838296,
-      "grad_norm": 0.019699299708008766,
-      "learning_rate": 4.745024057756037e-05,
-      "loss": 0.0054,
+      "epoch": 3.5587878787878786,
+      "grad_norm": 0.012118584476411343,
+      "learning_rate": 9.9451757559463e-05,
+      "loss": 0.0039,
       "step": 1101
     },
     {
-      "epoch": 2.823567082933077,
-      "grad_norm": 0.02164098061621189,
-      "learning_rate": 4.736085710013865e-05,
-      "loss": 0.0078,
+      "epoch": 3.562020202020202,
+      "grad_norm": 0.006976007483899593,
+      "learning_rate": 9.944841569113803e-05,
+      "loss": 0.0018,
       "step": 1102
     },
     {
-      "epoch": 2.826128722382325,
-      "grad_norm": 0.019722308963537216,
-      "learning_rate": 4.727148207956042e-05,
-      "loss": 0.0061,
+      "epoch": 3.565252525252525,
+      "grad_norm": 0.005068677943199873,
+      "learning_rate": 9.944506372487754e-05,
+      "loss": 0.0014,
       "step": 1103
     },
     {
-      "epoch": 2.828690361831572,
-      "grad_norm": 0.019483914598822594,
-      "learning_rate": 4.7182115802218126e-05,
-      "loss": 0.0064,
+      "epoch": 3.5684848484848484,
+      "grad_norm": 0.007260710932314396,
+      "learning_rate": 9.944170166136607e-05,
+      "loss": 0.0017,
       "step": 1104
     },
     {
-      "epoch": 2.83125200128082,
-      "grad_norm": 0.01710042729973793,
-      "learning_rate": 4.709275855447621e-05,
-      "loss": 0.0051,
+      "epoch": 3.5717171717171716,
+      "grad_norm": 0.0048537589609622955,
+      "learning_rate": 9.943832950129018e-05,
+      "loss": 0.0014,
       "step": 1105
     },
     {
-      "epoch": 2.8338136407300674,
-      "grad_norm": 0.022047800943255424,
-      "learning_rate": 4.7003410622670134e-05,
-      "loss": 0.0081,
+      "epoch": 3.574949494949495,
+      "grad_norm": 0.009134343825280666,
+      "learning_rate": 9.943494724533848e-05,
+      "loss": 0.0026,
       "step": 1106
     },
     {
-      "epoch": 2.8363752801793147,
-      "grad_norm": 0.016609322279691696,
-      "learning_rate": 4.691407229310553e-05,
-      "loss": 0.006,
+      "epoch": 3.578181818181818,
+      "grad_norm": 0.009165911935269833,
+      "learning_rate": 9.943155489420169e-05,
+      "loss": 0.002,
       "step": 1107
     },
     {
-      "epoch": 2.838936919628562,
-      "grad_norm": 0.023373253643512726,
-      "learning_rate": 4.682474385205728e-05,
-      "loss": 0.0066,
+      "epoch": 3.5814141414141414,
+      "grad_norm": 0.005547838285565376,
+      "learning_rate": 9.942815244857256e-05,
+      "loss": 0.0017,
       "step": 1108
     },
     {
-      "epoch": 2.84149855907781,
-      "grad_norm": 0.021889397874474525,
-      "learning_rate": 4.6735425585768536e-05,
-      "loss": 0.0066,
+      "epoch": 3.5846464646464646,
+      "grad_norm": 0.008086358197033405,
+      "learning_rate": 9.942473990914593e-05,
+      "loss": 0.0022,
       "step": 1109
     },
     {
-      "epoch": 2.8440601985270573,
-      "grad_norm": 0.028596634045243263,
-      "learning_rate": 4.6646117780449876e-05,
-      "loss": 0.0087,
+      "epoch": 3.587878787878788,
+      "grad_norm": 0.008301950991153717,
+      "learning_rate": 9.942131727661863e-05,
+      "loss": 0.0025,
       "step": 1110
     },
     {
-      "epoch": 2.846621837976305,
-      "grad_norm": 0.0335087887942791,
-      "learning_rate": 4.6556820722278393e-05,
-      "loss": 0.01,
+      "epoch": 3.591111111111111,
+      "grad_norm": 0.00843408890068531,
+      "learning_rate": 9.941788455168965e-05,
+      "loss": 0.0026,
       "step": 1111
     },
     {
-      "epoch": 2.8491834774255524,
-      "grad_norm": 0.02548697032034397,
-      "learning_rate": 4.646753469739662e-05,
-      "loss": 0.0066,
+      "epoch": 3.5943434343434344,
+      "grad_norm": 0.008435909636318684,
+      "learning_rate": 9.941444173505997e-05,
+      "loss": 0.0021,
       "step": 1112
     },
     {
-      "epoch": 2.8517451168748,
-      "grad_norm": 0.026222119107842445,
-      "learning_rate": 4.6378259991911886e-05,
-      "loss": 0.008,
+      "epoch": 3.5975757575757576,
+      "grad_norm": 0.0071822963654994965,
+      "learning_rate": 9.941098882743267e-05,
+      "loss": 0.0023,
       "step": 1113
     },
     {
-      "epoch": 2.854306756324047,
-      "grad_norm": 0.018268585205078125,
-      "learning_rate": 4.628899689189516e-05,
-      "loss": 0.0058,
+      "epoch": 3.600808080808081,
+      "grad_norm": 0.010135437361896038,
+      "learning_rate": 9.940752582951283e-05,
+      "loss": 0.0031,
       "step": 1114
     },
     {
-      "epoch": 2.856868395773295,
-      "grad_norm": 0.012070003896951675,
-      "learning_rate": 4.6199745683380206e-05,
-      "loss": 0.0039,
+      "epoch": 3.604040404040404,
+      "grad_norm": 0.0061368634924292564,
+      "learning_rate": 9.940405274200769e-05,
+      "loss": 0.0019,
       "step": 1115
     },
     {
-      "epoch": 2.8594300352225424,
-      "grad_norm": 0.01760229282081127,
-      "learning_rate": 4.611050665236276e-05,
-      "loss": 0.0071,
+      "epoch": 3.6072727272727274,
+      "grad_norm": 0.009635472670197487,
+      "learning_rate": 9.940056956562645e-05,
+      "loss": 0.0027,
       "step": 1116
     },
     {
-      "epoch": 2.86199167467179,
-      "grad_norm": 0.021311108022928238,
-      "learning_rate": 4.602128008479949e-05,
-      "loss": 0.0077,
+      "epoch": 3.6105050505050507,
+      "grad_norm": 0.01018766313791275,
+      "learning_rate": 9.939707630108044e-05,
+      "loss": 0.0029,
       "step": 1117
     },
     {
-      "epoch": 2.8645533141210375,
-      "grad_norm": 0.020954597741365433,
-      "learning_rate": 4.593206626660709e-05,
-      "loss": 0.0091,
+      "epoch": 3.613737373737374,
+      "grad_norm": 0.012025324627757072,
+      "learning_rate": 9.939357294908301e-05,
+      "loss": 0.0039,
       "step": 1118
     },
     {
-      "epoch": 2.867114953570285,
-      "grad_norm": 0.024132655933499336,
-      "learning_rate": 4.584286548366148e-05,
-      "loss": 0.0066,
+      "epoch": 3.616969696969697,
+      "grad_norm": 0.008770022541284561,
+      "learning_rate": 9.939005951034959e-05,
+      "loss": 0.0028,
       "step": 1119
     },
     {
-      "epoch": 2.8696765930195323,
-      "grad_norm": 0.023584892973303795,
-      "learning_rate": 4.5753678021796755e-05,
-      "loss": 0.0091,
+      "epoch": 3.6202020202020204,
+      "grad_norm": 0.006850701291114092,
+      "learning_rate": 9.938653598559769e-05,
+      "loss": 0.0017,
       "step": 1120
     },
     {
-      "epoch": 2.87223823246878,
-      "grad_norm": 0.02193642593920231,
-      "learning_rate": 4.5664504166804356e-05,
-      "loss": 0.0081,
+      "epoch": 3.6234343434343437,
+      "grad_norm": 0.006951460614800453,
+      "learning_rate": 9.93830023755468e-05,
+      "loss": 0.0021,
       "step": 1121
     },
     {
-      "epoch": 2.8747998719180274,
-      "grad_norm": 0.015388750471174717,
-      "learning_rate": 4.5575344204432084e-05,
-      "loss": 0.0047,
+      "epoch": 3.626666666666667,
+      "grad_norm": 0.007340079639106989,
+      "learning_rate": 9.937945868091856e-05,
+      "loss": 0.0013,
       "step": 1122
     },
     {
-      "epoch": 2.8773615113672752,
-      "grad_norm": 0.018353408202528954,
-      "learning_rate": 4.5486198420383244e-05,
-      "loss": 0.0057,
+      "epoch": 3.6298989898989897,
+      "grad_norm": 0.012878290377557278,
+      "learning_rate": 9.937590490243665e-05,
+      "loss": 0.0028,
       "step": 1123
     },
     {
-      "epoch": 2.8799231508165226,
-      "grad_norm": 0.01442483440041542,
-      "learning_rate": 4.5397067100315734e-05,
-      "loss": 0.0053,
+      "epoch": 3.633131313131313,
+      "grad_norm": 0.008417353965342045,
+      "learning_rate": 9.937234104082676e-05,
+      "loss": 0.0018,
       "step": 1124
     },
     {
-      "epoch": 2.88248479026577,
-      "grad_norm": 0.039199162274599075,
-      "learning_rate": 4.530795052984104e-05,
-      "loss": 0.0168,
+      "epoch": 3.6363636363636362,
+      "grad_norm": 0.007263545412570238,
+      "learning_rate": 9.936876709681668e-05,
+      "loss": 0.0014,
       "step": 1125
     },
     {
-      "epoch": 2.88248479026577,
-      "eval_loss": 0.008591169491410255,
-      "eval_runtime": 15.6964,
-      "eval_samples_per_second": 6.371,
-      "eval_steps_per_second": 1.593,
+      "epoch": 3.6363636363636362,
+      "eval_loss": 0.001890599844045937,
+      "eval_runtime": 15.6433,
+      "eval_samples_per_second": 6.393,
+      "eval_steps_per_second": 1.598,
       "step": 1125
     },
     {
-      "epoch": 2.8850464297150173,
-      "grad_norm": 0.018906261771917343,
-      "learning_rate": 4.5218848994523446e-05,
-      "loss": 0.0057,
+      "epoch": 3.6395959595959595,
+      "grad_norm": 0.012460816651582718,
+      "learning_rate": 9.936518307113625e-05,
+      "loss": 0.004,
       "step": 1126
     },
     {
-      "epoch": 2.887608069164265,
-      "grad_norm": 0.024255679920315742,
-      "learning_rate": 4.512976277987904e-05,
-      "loss": 0.01,
+      "epoch": 3.6428282828282827,
+      "grad_norm": 0.007229766808450222,
+      "learning_rate": 9.936158896451737e-05,
+      "loss": 0.0019,
       "step": 1127
     },
     {
-      "epoch": 2.8901697086135125,
-      "grad_norm": 0.025274179875850677,
-      "learning_rate": 4.50406921713748e-05,
-      "loss": 0.0108,
+      "epoch": 3.646060606060606,
+      "grad_norm": 0.008500331081449986,
+      "learning_rate": 9.9357984777694e-05,
+      "loss": 0.0013,
       "step": 1128
     },
     {
-      "epoch": 2.8927313480627603,
-      "grad_norm": 0.020932083949446678,
-      "learning_rate": 4.495163745442769e-05,
-      "loss": 0.0063,
+      "epoch": 3.6492929292929293,
+      "grad_norm": 0.009302293881773949,
+      "learning_rate": 9.935437051140216e-05,
+      "loss": 0.0025,
       "step": 1129
     },
     {
-      "epoch": 2.8952929875120077,
-      "grad_norm": 0.013744296506047249,
-      "learning_rate": 4.48625989144038e-05,
-      "loss": 0.0049,
+      "epoch": 3.6525252525252525,
+      "grad_norm": 0.007230978459119797,
+      "learning_rate": 9.935074616637992e-05,
+      "loss": 0.0022,
       "step": 1130
     },
     {
-      "epoch": 2.897854626961255,
-      "grad_norm": 0.013666169717907906,
-      "learning_rate": 4.477357683661734e-05,
-      "loss": 0.0043,
+      "epoch": 3.6557575757575758,
+      "grad_norm": 0.005689865443855524,
+      "learning_rate": 9.934711174336742e-05,
+      "loss": 0.0015,
       "step": 1131
     },
     {
-      "epoch": 2.900416266410503,
-      "grad_norm": 0.015672704204916954,
-      "learning_rate": 4.468457150632975e-05,
-      "loss": 0.0064,
+      "epoch": 3.658989898989899,
+      "grad_norm": 0.005996790248900652,
+      "learning_rate": 9.934346724310684e-05,
+      "loss": 0.0016,
       "step": 1132
     },
     {
-      "epoch": 2.9029779058597502,
-      "grad_norm": 0.01978994347155094,
-      "learning_rate": 4.459558320874888e-05,
-      "loss": 0.0067,
+      "epoch": 3.6622222222222223,
+      "grad_norm": 0.010158374905586243,
+      "learning_rate": 9.933981266634243e-05,
+      "loss": 0.003,
       "step": 1133
     },
     {
-      "epoch": 2.9055395453089976,
-      "grad_norm": 0.03054462932050228,
-      "learning_rate": 4.450661222902793e-05,
-      "loss": 0.0163,
+      "epoch": 3.6654545454545455,
+      "grad_norm": 0.008783318102359772,
+      "learning_rate": 9.93361480138205e-05,
+      "loss": 0.0019,
       "step": 1134
     },
     {
-      "epoch": 2.9081011847582454,
-      "grad_norm": 0.0375843346118927,
-      "learning_rate": 4.4417658852264614e-05,
-      "loss": 0.0098,
+      "epoch": 3.6686868686868688,
+      "grad_norm": 0.006477830931544304,
+      "learning_rate": 9.933247328628944e-05,
+      "loss": 0.0017,
       "step": 1135
     },
     {
-      "epoch": 2.910662824207493,
-      "grad_norm": 0.019927896559238434,
-      "learning_rate": 4.43287233635003e-05,
-      "loss": 0.0084,
+      "epoch": 3.671919191919192,
+      "grad_norm": 0.007675418630242348,
+      "learning_rate": 9.93287884844996e-05,
+      "loss": 0.002,
       "step": 1136
     },
     {
-      "epoch": 2.91322446365674,
-      "grad_norm": 0.015810800716280937,
-      "learning_rate": 4.423980604771895e-05,
-      "loss": 0.0044,
+      "epoch": 3.6751515151515153,
+      "grad_norm": 0.008330308832228184,
+      "learning_rate": 9.932509360920353e-05,
+      "loss": 0.0018,
       "step": 1137
     },
     {
-      "epoch": 2.915786103105988,
-      "grad_norm": 0.0203675776720047,
-      "learning_rate": 4.4150907189846345e-05,
-      "loss": 0.0064,
+      "epoch": 3.6783838383838385,
+      "grad_norm": 0.00984620675444603,
+      "learning_rate": 9.932138866115574e-05,
+      "loss": 0.0023,
       "step": 1138
     },
     {
-      "epoch": 2.9183477425552353,
-      "grad_norm": 0.015811270102858543,
-      "learning_rate": 4.4062027074749126e-05,
-      "loss": 0.0055,
+      "epoch": 3.6816161616161613,
+      "grad_norm": 0.006650052033364773,
+      "learning_rate": 9.931767364111283e-05,
+      "loss": 0.002,
       "step": 1139
     },
     {
-      "epoch": 2.920909382004483,
-      "grad_norm": 0.02347410097718239,
-      "learning_rate": 4.397316598723385e-05,
-      "loss": 0.0088,
+      "epoch": 3.6848484848484846,
+      "grad_norm": 0.005721759051084518,
+      "learning_rate": 9.931394854983345e-05,
+      "loss": 0.0018,
       "step": 1140
     },
     {
-      "epoch": 2.9234710214537305,
-      "grad_norm": 0.017205510288476944,
-      "learning_rate": 4.38843242120461e-05,
-      "loss": 0.0058,
+      "epoch": 3.688080808080808,
+      "grad_norm": 0.008330756798386574,
+      "learning_rate": 9.931021338807828e-05,
+      "loss": 0.0023,
       "step": 1141
     },
     {
-      "epoch": 2.926032660902978,
-      "grad_norm": 0.015807706862688065,
-      "learning_rate": 4.37955020338696e-05,
-      "loss": 0.0058,
+      "epoch": 3.691313131313131,
+      "grad_norm": 0.006348080467432737,
+      "learning_rate": 9.93064681566101e-05,
+      "loss": 0.0016,
       "step": 1142
     },
     {
-      "epoch": 2.9285943003522252,
-      "grad_norm": 0.02101590670645237,
-      "learning_rate": 4.370669973732527e-05,
-      "loss": 0.0045,
+      "epoch": 3.6945454545454544,
+      "grad_norm": 0.0061849444173276424,
+      "learning_rate": 9.930271285619376e-05,
+      "loss": 0.002,
       "step": 1143
     },
     {
-      "epoch": 2.931155939801473,
-      "grad_norm": 0.015992263332009315,
-      "learning_rate": 4.361791760697027e-05,
-      "loss": 0.0056,
+      "epoch": 3.6977777777777776,
+      "grad_norm": 0.00802608672529459,
+      "learning_rate": 9.92989474875961e-05,
+      "loss": 0.0026,
       "step": 1144
     },
     {
-      "epoch": 2.9337175792507204,
-      "grad_norm": 0.0318165086209774,
-      "learning_rate": 4.352915592729723e-05,
-      "loss": 0.0125,
+      "epoch": 3.701010101010101,
+      "grad_norm": 0.010230571031570435,
+      "learning_rate": 9.929517205158605e-05,
+      "loss": 0.0031,
       "step": 1145
     },
     {
-      "epoch": 2.936279218699968,
-      "grad_norm": 0.016865696758031845,
-      "learning_rate": 4.34404149827332e-05,
-      "loss": 0.0053,
+      "epoch": 3.704242424242424,
+      "grad_norm": 0.005332183558493853,
+      "learning_rate": 9.929138654893462e-05,
+      "loss": 0.0015,
       "step": 1146
     },
     {
-      "epoch": 2.9388408581492156,
-      "grad_norm": 0.039752960205078125,
-      "learning_rate": 4.335169505763874e-05,
-      "loss": 0.0088,
+      "epoch": 3.7074747474747474,
+      "grad_norm": 0.006876831874251366,
+      "learning_rate": 9.928759098041482e-05,
+      "loss": 0.0019,
       "step": 1147
     },
     {
-      "epoch": 2.941402497598463,
-      "grad_norm": 0.022740483283996582,
-      "learning_rate": 4.326299643630716e-05,
-      "loss": 0.0065,
+      "epoch": 3.7107070707070706,
+      "grad_norm": 0.006538907065987587,
+      "learning_rate": 9.928378534680178e-05,
+      "loss": 0.0012,
       "step": 1148
     },
     {
-      "epoch": 2.9439641370477103,
-      "grad_norm": 0.023348230868577957,
-      "learning_rate": 4.317431940296343e-05,
-      "loss": 0.0074,
+      "epoch": 3.713939393939394,
+      "grad_norm": 0.005673635751008987,
+      "learning_rate": 9.927996964887265e-05,
+      "loss": 0.0017,
       "step": 1149
     },
     {
-      "epoch": 2.946525776496958,
-      "grad_norm": 0.04514496400952339,
-      "learning_rate": 4.308566424176336e-05,
-      "loss": 0.013,
+      "epoch": 3.717171717171717,
+      "grad_norm": 0.007973823696374893,
+      "learning_rate": 9.927614388740663e-05,
+      "loss": 0.0026,
       "step": 1150
     },
     {
-      "epoch": 2.946525776496958,
-      "eval_loss": 0.006918873637914658,
-      "eval_runtime": 15.7249,
-      "eval_samples_per_second": 6.359,
-      "eval_steps_per_second": 1.59,
+      "epoch": 3.717171717171717,
+      "eval_loss": 0.0018489748472347856,
+      "eval_runtime": 15.6022,
+      "eval_samples_per_second": 6.409,
+      "eval_steps_per_second": 1.602,
       "step": 1150
     },
     {
-      "epoch": 2.9490874159462055,
-      "grad_norm": 0.01825115457177162,
-      "learning_rate": 4.299703123679265e-05,
-      "loss": 0.0069,
+      "epoch": 3.7204040404040404,
+      "grad_norm": 0.007316716015338898,
+      "learning_rate": 9.9272308063185e-05,
+      "loss": 0.002,
       "step": 1151
     },
     {
-      "epoch": 2.9516490553954533,
-      "grad_norm": 0.022036191076040268,
-      "learning_rate": 4.290842067206608e-05,
-      "loss": 0.0075,
+      "epoch": 3.7236363636363636,
+      "grad_norm": 0.010193591937422752,
+      "learning_rate": 9.926846217699104e-05,
+      "loss": 0.0028,
       "step": 1152
     },
     {
-      "epoch": 2.9542106948447007,
-      "grad_norm": 0.016818994656205177,
-      "learning_rate": 4.281983283152643e-05,
-      "loss": 0.0049,
+      "epoch": 3.726868686868687,
+      "grad_norm": 0.005911578889936209,
+      "learning_rate": 9.926460622961016e-05,
+      "loss": 0.0013,
       "step": 1153
     },
     {
-      "epoch": 2.956772334293948,
-      "grad_norm": 0.01692839525640011,
-      "learning_rate": 4.27312679990437e-05,
-      "loss": 0.0039,
+      "epoch": 3.73010101010101,
+      "grad_norm": 0.005372088868170977,
+      "learning_rate": 9.926074022182979e-05,
+      "loss": 0.0015,
       "step": 1154
     },
     {
-      "epoch": 2.9593339737431954,
-      "grad_norm": 0.027776619419455528,
-      "learning_rate": 4.264272645841419e-05,
-      "loss": 0.0079,
+      "epoch": 3.7333333333333334,
+      "grad_norm": 0.006013571284711361,
+      "learning_rate": 9.92568641544394e-05,
+      "loss": 0.0014,
       "step": 1155
     },
     {
-      "epoch": 2.961895613192443,
-      "grad_norm": 0.021229857578873634,
-      "learning_rate": 4.255420849335956e-05,
-      "loss": 0.0056,
+      "epoch": 3.7365656565656566,
+      "grad_norm": 0.008860640227794647,
+      "learning_rate": 9.925297802823054e-05,
+      "loss": 0.003,
       "step": 1156
     },
     {
-      "epoch": 2.9644572526416906,
-      "grad_norm": 0.0324573740363121,
-      "learning_rate": 4.246571438752585e-05,
-      "loss": 0.0088,
+      "epoch": 3.73979797979798,
+      "grad_norm": 0.005647321231663227,
+      "learning_rate": 9.924908184399677e-05,
+      "loss": 0.0016,
       "step": 1157
     },
     {
-      "epoch": 2.9670188920909384,
-      "grad_norm": 0.021437501534819603,
-      "learning_rate": 4.237724442448273e-05,
-      "loss": 0.0093,
+      "epoch": 3.743030303030303,
+      "grad_norm": 0.008007604628801346,
+      "learning_rate": 9.924517560253378e-05,
+      "loss": 0.0031,
       "step": 1158
     },
     {
-      "epoch": 2.9695805315401858,
-      "grad_norm": 0.020703041926026344,
-      "learning_rate": 4.2288798887722494e-05,
-      "loss": 0.0077,
+      "epoch": 3.7462626262626264,
+      "grad_norm": 0.007285719737410545,
+      "learning_rate": 9.924125930463924e-05,
+      "loss": 0.0019,
       "step": 1159
     },
     {
-      "epoch": 2.972142170989433,
-      "grad_norm": 0.01659969985485077,
-      "learning_rate": 4.2200378060659116e-05,
-      "loss": 0.0038,
+      "epoch": 3.7494949494949497,
+      "grad_norm": 0.005834940355271101,
+      "learning_rate": 9.92373329511129e-05,
+      "loss": 0.0016,
       "step": 1160
     },
     {
-      "epoch": 2.9747038104386805,
-      "grad_norm": 0.018442777916789055,
-      "learning_rate": 4.2111982226627436e-05,
-      "loss": 0.005,
+      "epoch": 3.752727272727273,
+      "grad_norm": 0.008350113406777382,
+      "learning_rate": 9.92333965427566e-05,
+      "loss": 0.0036,
       "step": 1161
     },
     {
-      "epoch": 2.9772654498879283,
-      "grad_norm": 0.02515353262424469,
-      "learning_rate": 4.202361166888222e-05,
-      "loss": 0.0059,
+      "epoch": 3.755959595959596,
+      "grad_norm": 0.008788091130554676,
+      "learning_rate": 9.922945008037417e-05,
+      "loss": 0.0025,
       "step": 1162
     },
     {
-      "epoch": 2.9798270893371757,
-      "grad_norm": 0.029337922111153603,
-      "learning_rate": 4.193526667059716e-05,
-      "loss": 0.0087,
+      "epoch": 3.7591919191919194,
+      "grad_norm": 0.008776461705565453,
+      "learning_rate": 9.922549356477152e-05,
+      "loss": 0.0018,
       "step": 1163
     },
     {
-      "epoch": 2.9823887287864235,
-      "grad_norm": 0.030802778899669647,
-      "learning_rate": 4.184694751486414e-05,
-      "loss": 0.0112,
+      "epoch": 3.7624242424242427,
+      "grad_norm": 0.006701668258756399,
+      "learning_rate": 9.922152699675664e-05,
+      "loss": 0.0024,
       "step": 1164
     },
     {
-      "epoch": 2.984950368235671,
-      "grad_norm": 0.03156080096960068,
-      "learning_rate": 4.1758654484692186e-05,
-      "loss": 0.0117,
+      "epoch": 3.765656565656566,
+      "grad_norm": 0.007463917136192322,
+      "learning_rate": 9.921755037713952e-05,
+      "loss": 0.002,
       "step": 1165
     },
     {
-      "epoch": 2.987512007684918,
-      "grad_norm": 0.025577133521437645,
-      "learning_rate": 4.167038786300659e-05,
-      "loss": 0.0076,
+      "epoch": 3.7688888888888887,
+      "grad_norm": 0.005220974329859018,
+      "learning_rate": 9.921356370673225e-05,
+      "loss": 0.0013,
       "step": 1166
     },
     {
-      "epoch": 2.990073647134166,
-      "grad_norm": 0.030604520812630653,
-      "learning_rate": 4.1582147932648074e-05,
-      "loss": 0.012,
+      "epoch": 3.772121212121212,
+      "grad_norm": 0.009098538197577,
+      "learning_rate": 9.920956698634896e-05,
+      "loss": 0.0035,
       "step": 1167
     },
     {
-      "epoch": 2.9926352865834134,
-      "grad_norm": 0.023930441588163376,
-      "learning_rate": 4.14939349763718e-05,
-      "loss": 0.0056,
+      "epoch": 3.7753535353535352,
+      "grad_norm": 0.008859221823513508,
+      "learning_rate": 9.92055602168058e-05,
+      "loss": 0.002,
       "step": 1168
     },
     {
-      "epoch": 2.9951969260326607,
-      "grad_norm": 0.019746793434023857,
-      "learning_rate": 4.140574927684647e-05,
-      "loss": 0.0081,
+      "epoch": 3.7785858585858585,
+      "grad_norm": 0.00567187974229455,
+      "learning_rate": 9.920154339892104e-05,
+      "loss": 0.0016,
       "step": 1169
     },
     {
-      "epoch": 2.9977585654819086,
-      "grad_norm": 0.0151827996596694,
-      "learning_rate": 4.131759111665349e-05,
-      "loss": 0.0042,
+      "epoch": 3.7818181818181817,
+      "grad_norm": 0.011804512701928616,
+      "learning_rate": 9.919751653351493e-05,
+      "loss": 0.0029,
       "step": 1170
     },
     {
-      "epoch": 3.000640409862312,
-      "grad_norm": 0.061972830444574356,
-      "learning_rate": 4.1229460778286004e-05,
-      "loss": 0.02,
+      "epoch": 3.785050505050505,
+      "grad_norm": 0.009031563997268677,
+      "learning_rate": 9.919347962140979e-05,
+      "loss": 0.0027,
       "step": 1171
     },
     {
-      "epoch": 3.003202049311559,
-      "grad_norm": 0.010719524696469307,
-      "learning_rate": 4.1141358544147976e-05,
-      "loss": 0.0026,
+      "epoch": 3.7882828282828283,
+      "grad_norm": 0.009653420187532902,
+      "learning_rate": 9.918943266343004e-05,
+      "loss": 0.0028,
       "step": 1172
     },
     {
-      "epoch": 3.005763688760807,
-      "grad_norm": 0.012511296197772026,
-      "learning_rate": 4.105328469655336e-05,
-      "loss": 0.0048,
+      "epoch": 3.7915151515151515,
+      "grad_norm": 0.006550057325512171,
+      "learning_rate": 9.91853756604021e-05,
+      "loss": 0.0014,
       "step": 1173
     },
     {
-      "epoch": 3.0083253282100544,
-      "grad_norm": 0.02187875472009182,
-      "learning_rate": 4.0965239517725117e-05,
-      "loss": 0.0077,
+      "epoch": 3.7947474747474748,
+      "grad_norm": 0.007456491235643625,
+      "learning_rate": 9.918130861315444e-05,
+      "loss": 0.0016,
       "step": 1174
     },
     {
-      "epoch": 3.0108869676593017,
-      "grad_norm": 0.016843268647789955,
-      "learning_rate": 4.087722328979438e-05,
-      "loss": 0.0058,
+      "epoch": 3.797979797979798,
+      "grad_norm": 0.007812919095158577,
+      "learning_rate": 9.91772315225176e-05,
+      "loss": 0.0021,
       "step": 1175
     },
     {
-      "epoch": 3.0108869676593017,
-      "eval_loss": 0.007106189616024494,
-      "eval_runtime": 15.6924,
-      "eval_samples_per_second": 6.373,
-      "eval_steps_per_second": 1.593,
+      "epoch": 3.797979797979798,
+      "eval_loss": 0.0018577298615127802,
+      "eval_runtime": 15.6323,
+      "eval_samples_per_second": 6.397,
+      "eval_steps_per_second": 1.599,
       "step": 1175
     },
     {
-      "epoch": 3.0134486071085496,
-      "grad_norm": 0.014118392020463943,
-      "learning_rate": 4.078923629479943e-05,
-      "loss": 0.0039,
+      "epoch": 3.8012121212121213,
+      "grad_norm": 0.0034585888497531414,
+      "learning_rate": 9.917314438932421e-05,
+      "loss": 0.0009,
       "step": 1176
     },
     {
-      "epoch": 3.016010246557797,
-      "grad_norm": 0.017133532091975212,
-      "learning_rate": 4.0701278814684945e-05,
-      "loss": 0.007,
+      "epoch": 3.8044444444444445,
+      "grad_norm": 0.005305670201778412,
+      "learning_rate": 9.916904721440887e-05,
+      "loss": 0.001,
       "step": 1177
     },
     {
-      "epoch": 3.0185718860070443,
-      "grad_norm": 0.015284501016139984,
-      "learning_rate": 4.0613351131301036e-05,
-      "loss": 0.0054,
+      "epoch": 3.8076767676767678,
+      "grad_norm": 0.005096958950161934,
+      "learning_rate": 9.916493999860828e-05,
+      "loss": 0.0014,
       "step": 1178
     },
     {
-      "epoch": 3.021133525456292,
-      "grad_norm": 0.025085574015975,
-      "learning_rate": 4.052545352640228e-05,
-      "loss": 0.004,
+      "epoch": 3.810909090909091,
+      "grad_norm": 0.004448979161679745,
+      "learning_rate": 9.916082274276117e-05,
+      "loss": 0.0015,
       "step": 1179
     },
     {
-      "epoch": 3.0236951649055395,
-      "grad_norm": 0.018907958641648293,
-      "learning_rate": 4.043758628164688e-05,
-      "loss": 0.0044,
+      "epoch": 3.8141414141414143,
+      "grad_norm": 0.010591958649456501,
+      "learning_rate": 9.915669544770836e-05,
+      "loss": 0.0015,
       "step": 1180
     },
     {
-      "epoch": 3.0262568043547873,
-      "grad_norm": 0.017235899344086647,
-      "learning_rate": 4.034974967859582e-05,
-      "loss": 0.0042,
+      "epoch": 3.8173737373737375,
+      "grad_norm": 0.009333600290119648,
+      "learning_rate": 9.915255811429267e-05,
+      "loss": 0.0024,
       "step": 1181
     },
     {
-      "epoch": 3.0288184438040346,
-      "grad_norm": 0.01964586041867733,
-      "learning_rate": 4.026194399871181e-05,
-      "loss": 0.0038,
+      "epoch": 3.8206060606060603,
+      "grad_norm": 0.013536066748201847,
+      "learning_rate": 9.914841074335898e-05,
+      "loss": 0.005,
       "step": 1182
     },
     {
-      "epoch": 3.031380083253282,
-      "grad_norm": 0.019763899967074394,
-      "learning_rate": 4.017416952335849e-05,
-      "loss": 0.0037,
+      "epoch": 3.8238383838383836,
+      "grad_norm": 0.008231270126998425,
+      "learning_rate": 9.914425333575426e-05,
+      "loss": 0.002,
       "step": 1183
     },
     {
-      "epoch": 3.03394172270253,
-      "grad_norm": 0.020342646166682243,
-      "learning_rate": 4.008642653379957e-05,
-      "loss": 0.0042,
+      "epoch": 3.827070707070707,
+      "grad_norm": 0.006396579556167126,
+      "learning_rate": 9.914008589232749e-05,
+      "loss": 0.0022,
       "step": 1184
     },
     {
-      "epoch": 3.036503362151777,
-      "grad_norm": 0.021486425772309303,
-      "learning_rate": 3.9998715311197785e-05,
+      "epoch": 3.83030303030303,
+      "grad_norm": 0.012204843573272228,
+      "learning_rate": 9.91359084139297e-05,
       "loss": 0.005,
       "step": 1185
     },
     {
-      "epoch": 3.0390650016010246,
-      "grad_norm": 0.012984909117221832,
-      "learning_rate": 3.991103613661412e-05,
-      "loss": 0.0026,
+      "epoch": 3.8335353535353534,
+      "grad_norm": 0.006131699774414301,
+      "learning_rate": 9.913172090141399e-05,
+      "loss": 0.0016,
       "step": 1186
     },
     {
-      "epoch": 3.0416266410502724,
-      "grad_norm": 0.022616276517510414,
-      "learning_rate": 3.982338929100688e-05,
-      "loss": 0.0041,
+      "epoch": 3.8367676767676766,
+      "grad_norm": 0.004087884910404682,
+      "learning_rate": 9.912752335563548e-05,
+      "loss": 0.0011,
       "step": 1187
     },
     {
-      "epoch": 3.0441882804995197,
-      "grad_norm": 0.01583174802362919,
-      "learning_rate": 3.973577505523072e-05,
-      "loss": 0.0034,
+      "epoch": 3.84,
+      "grad_norm": 0.007036447525024414,
+      "learning_rate": 9.912331577745138e-05,
+      "loss": 0.0016,
       "step": 1188
     },
     {
-      "epoch": 3.046749919948767,
-      "grad_norm": 0.01628080941736698,
-      "learning_rate": 3.964819371003585e-05,
-      "loss": 0.0034,
+      "epoch": 3.843232323232323,
+      "grad_norm": 0.007033781614154577,
+      "learning_rate": 9.911909816772091e-05,
+      "loss": 0.0024,
       "step": 1189
     },
     {
-      "epoch": 3.049311559398015,
-      "grad_norm": 0.02016085386276245,
-      "learning_rate": 3.956064553606708e-05,
-      "loss": 0.0041,
+      "epoch": 3.8464646464646464,
+      "grad_norm": 0.008417130447924137,
+      "learning_rate": 9.911487052730537e-05,
+      "loss": 0.0037,
       "step": 1190
     },
     {
-      "epoch": 3.0518731988472623,
-      "grad_norm": 0.02444877102971077,
-      "learning_rate": 3.947313081386289e-05,
-      "loss": 0.0059,
+      "epoch": 3.8496969696969696,
+      "grad_norm": 0.004637300968170166,
+      "learning_rate": 9.911063285706808e-05,
+      "loss": 0.0014,
       "step": 1191
     },
     {
-      "epoch": 3.0544348382965096,
-      "grad_norm": 0.028850266709923744,
-      "learning_rate": 3.938564982385462e-05,
-      "loss": 0.0069,
+      "epoch": 3.852929292929293,
+      "grad_norm": 0.00659260805696249,
+      "learning_rate": 9.910638515787442e-05,
+      "loss": 0.0018,
       "step": 1192
     },
     {
-      "epoch": 3.0569964777457574,
-      "grad_norm": 0.024146832525730133,
-      "learning_rate": 3.9298202846365466e-05,
-      "loss": 0.004,
+      "epoch": 3.856161616161616,
+      "grad_norm": 0.00474232342094183,
+      "learning_rate": 9.910212743059182e-05,
+      "loss": 0.0014,
       "step": 1193
     },
     {
-      "epoch": 3.059558117195005,
-      "grad_norm": 0.03326459228992462,
-      "learning_rate": 3.92107901616097e-05,
-      "loss": 0.0064,
+      "epoch": 3.8593939393939394,
+      "grad_norm": 0.005664649419486523,
+      "learning_rate": 9.909785967608977e-05,
+      "loss": 0.0012,
       "step": 1194
     },
     {
-      "epoch": 3.062119756644252,
-      "grad_norm": 0.02041238360106945,
-      "learning_rate": 3.912341204969164e-05,
-      "loss": 0.0042,
+      "epoch": 3.8626262626262626,
+      "grad_norm": 0.00791127234697342,
+      "learning_rate": 9.909358189523978e-05,
+      "loss": 0.0019,
       "step": 1195
     },
     {
-      "epoch": 3.0646813960935,
-      "grad_norm": 0.02406291291117668,
-      "learning_rate": 3.903606879060483e-05,
-      "loss": 0.0044,
+      "epoch": 3.865858585858586,
+      "grad_norm": 0.01060640998184681,
+      "learning_rate": 9.908929408891542e-05,
+      "loss": 0.0035,
       "step": 1196
     },
     {
-      "epoch": 3.0672430355427474,
-      "grad_norm": 0.017291957512497902,
-      "learning_rate": 3.8948760664231195e-05,
-      "loss": 0.003,
+      "epoch": 3.869090909090909,
+      "grad_norm": 0.004038059618324041,
+      "learning_rate": 9.908499625799235e-05,
+      "loss": 0.0011,
       "step": 1197
     },
     {
-      "epoch": 3.0698046749919947,
-      "grad_norm": 0.013232097029685974,
-      "learning_rate": 3.886148795034001e-05,
-      "loss": 0.0034,
+      "epoch": 3.8723232323232324,
+      "grad_norm": 0.005425203591585159,
+      "learning_rate": 9.908068840334818e-05,
+      "loss": 0.0012,
       "step": 1198
     },
     {
-      "epoch": 3.0723663144412425,
-      "grad_norm": 0.026490524411201477,
-      "learning_rate": 3.877425092858708e-05,
-      "loss": 0.0058,
+      "epoch": 3.8755555555555556,
+      "grad_norm": 0.009089686907827854,
+      "learning_rate": 9.907637052586265e-05,
+      "loss": 0.0028,
       "step": 1199
     },
     {
-      "epoch": 3.07492795389049,
-      "grad_norm": 0.018783340230584145,
-      "learning_rate": 3.86870498785139e-05,
-      "loss": 0.0057,
+      "epoch": 3.878787878787879,
+      "grad_norm": 0.008087129332125187,
+      "learning_rate": 9.907204262641751e-05,
+      "loss": 0.0022,
       "step": 1200
     },
     {
-      "epoch": 3.07492795389049,
-      "eval_loss": 0.0069786058738827705,
-      "eval_runtime": 15.681,
-      "eval_samples_per_second": 6.377,
-      "eval_steps_per_second": 1.594,
+      "epoch": 3.878787878787879,
+      "eval_loss": 0.0019142035162076354,
+      "eval_runtime": 15.6022,
+      "eval_samples_per_second": 6.409,
+      "eval_steps_per_second": 1.602,
       "step": 1200
     },
     {
-      "epoch": 3.0774895933397373,
-      "grad_norm": 0.016674773767590523,
-      "learning_rate": 3.859988507954663e-05,
-      "loss": 0.0034,
+      "epoch": 3.882020202020202,
+      "grad_norm": 0.006272248458117247,
+      "learning_rate": 9.906770470589657e-05,
+      "loss": 0.0013,
       "step": 1201
     },
     {
-      "epoch": 3.080051232788985,
-      "grad_norm": 0.01833149790763855,
-      "learning_rate": 3.851275681099528e-05,
-      "loss": 0.0047,
+      "epoch": 3.8852525252525254,
+      "grad_norm": 0.0093921460211277,
+      "learning_rate": 9.90633567651857e-05,
+      "loss": 0.0032,
       "step": 1202
     },
     {
-      "epoch": 3.0826128722382324,
-      "grad_norm": 0.012066060677170753,
-      "learning_rate": 3.842566535205286e-05,
-      "loss": 0.0025,
+      "epoch": 3.8884848484848487,
+      "grad_norm": 0.006976392585784197,
+      "learning_rate": 9.905899880517278e-05,
+      "loss": 0.002,
       "step": 1203
     },
     {
-      "epoch": 3.08517451168748,
-      "grad_norm": 0.026396669447422028,
-      "learning_rate": 3.8338610981794344e-05,
-      "loss": 0.0052,
+      "epoch": 3.891717171717172,
+      "grad_norm": 0.009651237167418003,
+      "learning_rate": 9.905463082674778e-05,
+      "loss": 0.0031,
       "step": 1204
     },
     {
-      "epoch": 3.0877361511367276,
-      "grad_norm": 0.018845973536372185,
-      "learning_rate": 3.825159397917589e-05,
-      "loss": 0.0054,
+      "epoch": 3.894949494949495,
+      "grad_norm": 0.010465215891599655,
+      "learning_rate": 9.905025283080265e-05,
+      "loss": 0.0025,
       "step": 1205
     },
     {
-      "epoch": 3.090297790585975,
-      "grad_norm": 0.020172514021396637,
-      "learning_rate": 3.8164614623033975e-05,
-      "loss": 0.0032,
+      "epoch": 3.8981818181818184,
+      "grad_norm": 0.007280126214027405,
+      "learning_rate": 9.904586481823146e-05,
+      "loss": 0.0017,
       "step": 1206
     },
     {
-      "epoch": 3.0928594300352223,
-      "grad_norm": 0.025464298203587532,
-      "learning_rate": 3.807767319208435e-05,
-      "loss": 0.0055,
+      "epoch": 3.9014141414141417,
+      "grad_norm": 0.004892250057309866,
+      "learning_rate": 9.904146678993027e-05,
+      "loss": 0.0014,
       "step": 1207
     },
     {
-      "epoch": 3.09542106948447,
-      "grad_norm": 0.020005084574222565,
-      "learning_rate": 3.799076996492127e-05,
-      "loss": 0.004,
+      "epoch": 3.904646464646465,
+      "grad_norm": 0.007419066969305277,
+      "learning_rate": 9.903705874679724e-05,
+      "loss": 0.0021,
       "step": 1208
     },
     {
-      "epoch": 3.0979827089337175,
-      "grad_norm": 0.017038699239492416,
-      "learning_rate": 3.790390522001662e-05,
-      "loss": 0.0045,
+      "epoch": 3.9078787878787877,
+      "grad_norm": 0.0050157601945102215,
+      "learning_rate": 9.903264068973252e-05,
+      "loss": 0.001,
       "step": 1209
     },
     {
-      "epoch": 3.100544348382965,
-      "grad_norm": 0.025118080899119377,
-      "learning_rate": 3.781707923571891e-05,
-      "loss": 0.0046,
+      "epoch": 3.911111111111111,
+      "grad_norm": 0.004466901067644358,
+      "learning_rate": 9.902821261963833e-05,
+      "loss": 0.0014,
       "step": 1210
     },
     {
-      "epoch": 3.1031059878322127,
-      "grad_norm": 0.014508400112390518,
-      "learning_rate": 3.773029229025245e-05,
-      "loss": 0.0027,
+      "epoch": 3.9143434343434342,
+      "grad_norm": 0.011662506498396397,
+      "learning_rate": 9.902377453741893e-05,
+      "loss": 0.0031,
       "step": 1211
     },
     {
-      "epoch": 3.10566762728146,
-      "grad_norm": 0.01848450116813183,
-      "learning_rate": 3.764354466171652e-05,
-      "loss": 0.0047,
+      "epoch": 3.9175757575757575,
+      "grad_norm": 0.00633740471675992,
+      "learning_rate": 9.901932644398065e-05,
+      "loss": 0.0017,
       "step": 1212
     },
     {
-      "epoch": 3.1082292667307074,
-      "grad_norm": 0.019989987835288048,
-      "learning_rate": 3.755683662808434e-05,
-      "loss": 0.004,
+      "epoch": 3.9208080808080807,
+      "grad_norm": 0.007614588364958763,
+      "learning_rate": 9.901486834023182e-05,
+      "loss": 0.0028,
       "step": 1213
     },
     {
-      "epoch": 3.1107909061799552,
-      "grad_norm": 0.023956049233675003,
-      "learning_rate": 3.747016846720226e-05,
-      "loss": 0.0057,
+      "epoch": 3.924040404040404,
+      "grad_norm": 0.009498460218310356,
+      "learning_rate": 9.901040022708283e-05,
+      "loss": 0.0027,
       "step": 1214
     },
     {
-      "epoch": 3.1133525456292026,
-      "grad_norm": 0.015794293954968452,
-      "learning_rate": 3.738354045678891e-05,
-      "loss": 0.0031,
+      "epoch": 3.9272727272727272,
+      "grad_norm": 0.008266978897154331,
+      "learning_rate": 9.900592210544614e-05,
+      "loss": 0.0021,
       "step": 1215
     },
     {
-      "epoch": 3.1159141850784504,
-      "grad_norm": 0.018386952579021454,
-      "learning_rate": 3.7296952874434234e-05,
-      "loss": 0.0043,
+      "epoch": 3.9305050505050505,
+      "grad_norm": 0.007966354489326477,
+      "learning_rate": 9.900143397623622e-05,
+      "loss": 0.0014,
       "step": 1216
     },
     {
-      "epoch": 3.118475824527698,
-      "grad_norm": 0.03184031322598457,
-      "learning_rate": 3.7210405997598596e-05,
-      "loss": 0.0065,
+      "epoch": 3.9337373737373738,
+      "grad_norm": 0.005629670340567827,
+      "learning_rate": 9.899693584036959e-05,
+      "loss": 0.0021,
       "step": 1217
     },
     {
-      "epoch": 3.121037463976945,
-      "grad_norm": 0.02117329277098179,
-      "learning_rate": 3.712390010361198e-05,
-      "loss": 0.0053,
+      "epoch": 3.936969696969697,
+      "grad_norm": 0.006163495127111673,
+      "learning_rate": 9.899242769876486e-05,
+      "loss": 0.0021,
       "step": 1218
     },
     {
-      "epoch": 3.123599103426193,
-      "grad_norm": 0.02299274317920208,
-      "learning_rate": 3.703743546967304e-05,
-      "loss": 0.005,
+      "epoch": 3.9402020202020203,
+      "grad_norm": 0.007493765093386173,
+      "learning_rate": 9.89879095523426e-05,
+      "loss": 0.0017,
       "step": 1219
     },
     {
-      "epoch": 3.1261607428754403,
-      "grad_norm": 0.019891448318958282,
-      "learning_rate": 3.695101237284815e-05,
-      "loss": 0.0047,
+      "epoch": 3.9434343434343435,
+      "grad_norm": 0.010350759141147137,
+      "learning_rate": 9.898338140202549e-05,
+      "loss": 0.0021,
       "step": 1220
     },
     {
-      "epoch": 3.1287223823246877,
-      "grad_norm": 0.07732272148132324,
-      "learning_rate": 3.6864631090070655e-05,
-      "loss": 0.0061,
+      "epoch": 3.9466666666666668,
+      "grad_norm": 0.010328019969165325,
+      "learning_rate": 9.897884324873824e-05,
+      "loss": 0.0022,
       "step": 1221
     },
     {
-      "epoch": 3.1312840217739355,
-      "grad_norm": 0.019109439104795456,
-      "learning_rate": 3.67782918981399e-05,
-      "loss": 0.0042,
+      "epoch": 3.94989898989899,
+      "grad_norm": 0.005281366873532534,
+      "learning_rate": 9.897429509340756e-05,
+      "loss": 0.0013,
       "step": 1222
     },
     {
-      "epoch": 3.133845661223183,
-      "grad_norm": 0.020790351554751396,
-      "learning_rate": 3.6691995073720306e-05,
-      "loss": 0.0048,
+      "epoch": 3.9531313131313133,
+      "grad_norm": 0.006888162344694138,
+      "learning_rate": 9.896973693696228e-05,
+      "loss": 0.0019,
       "step": 1223
     },
     {
-      "epoch": 3.1364073006724302,
-      "grad_norm": 0.017758572474122047,
-      "learning_rate": 3.66057408933406e-05,
-      "loss": 0.0038,
+      "epoch": 3.9563636363636365,
+      "grad_norm": 0.007842673920094967,
+      "learning_rate": 9.896516878033317e-05,
+      "loss": 0.0019,
       "step": 1224
     },
     {
-      "epoch": 3.138968940121678,
-      "grad_norm": 0.034487027674913406,
-      "learning_rate": 3.651952963339282e-05,
-      "loss": 0.0075,
+      "epoch": 3.9595959595959593,
+      "grad_norm": 0.010139373131096363,
+      "learning_rate": 9.896059062445315e-05,
+      "loss": 0.0028,
       "step": 1225
     },
     {
-      "epoch": 3.138968940121678,
-      "eval_loss": 0.0067987870424985886,
-      "eval_runtime": 15.7069,
-      "eval_samples_per_second": 6.367,
-      "eval_steps_per_second": 1.592,
+      "epoch": 3.9595959595959593,
+      "eval_loss": 0.001911315368488431,
+      "eval_runtime": 15.6003,
+      "eval_samples_per_second": 6.41,
+      "eval_steps_per_second": 1.603,
       "step": 1225
     },
     {
-      "epoch": 3.1415305795709254,
-      "grad_norm": 0.022508880123496056,
-      "learning_rate": 3.643336157013147e-05,
-      "loss": 0.0056,
+      "epoch": 3.9628282828282826,
+      "grad_norm": 0.011555963195860386,
+      "learning_rate": 9.895600247025712e-05,
+      "loss": 0.0033,
       "step": 1226
     },
     {
-      "epoch": 3.1440922190201728,
-      "grad_norm": 0.01572093553841114,
-      "learning_rate": 3.634723697967263e-05,
-      "loss": 0.003,
+      "epoch": 3.966060606060606,
+      "grad_norm": 0.005696468986570835,
+      "learning_rate": 9.895140431868203e-05,
+      "loss": 0.0019,
       "step": 1227
     },
     {
-      "epoch": 3.1466538584694206,
-      "grad_norm": 0.02083379589021206,
-      "learning_rate": 3.6261156137993116e-05,
-      "loss": 0.004,
+      "epoch": 3.969292929292929,
+      "grad_norm": 0.008271520957350731,
+      "learning_rate": 9.894679617066687e-05,
+      "loss": 0.0027,
       "step": 1228
     },
     {
-      "epoch": 3.149215497918668,
-      "grad_norm": 0.0228340495377779,
-      "learning_rate": 3.617511932092953e-05,
-      "loss": 0.0055,
+      "epoch": 3.9725252525252523,
+      "grad_norm": 0.005810366943478584,
+      "learning_rate": 9.894217802715266e-05,
+      "loss": 0.0017,
       "step": 1229
     },
     {
-      "epoch": 3.1517771373679153,
-      "grad_norm": 0.020101532340049744,
-      "learning_rate": 3.608912680417737e-05,
-      "loss": 0.0037,
+      "epoch": 3.9757575757575756,
+      "grad_norm": 0.006310660857707262,
+      "learning_rate": 9.893754988908253e-05,
+      "loss": 0.0013,
       "step": 1230
     },
     {
-      "epoch": 3.154338776817163,
-      "grad_norm": 0.017341822385787964,
-      "learning_rate": 3.6003178863290235e-05,
-      "loss": 0.0037,
+      "epoch": 3.978989898989899,
+      "grad_norm": 0.005208928138017654,
+      "learning_rate": 9.893291175740156e-05,
+      "loss": 0.0018,
       "step": 1231
     },
     {
-      "epoch": 3.1569004162664105,
-      "grad_norm": 0.016037769615650177,
-      "learning_rate": 3.591727577367887e-05,
-      "loss": 0.0036,
+      "epoch": 3.982222222222222,
+      "grad_norm": 0.007240624632686377,
+      "learning_rate": 9.892826363305691e-05,
+      "loss": 0.0017,
       "step": 1232
     },
     {
-      "epoch": 3.159462055715658,
-      "grad_norm": 0.014761366881430149,
-      "learning_rate": 3.583141781061024e-05,
-      "loss": 0.0027,
+      "epoch": 3.9854545454545454,
+      "grad_norm": 0.008776954375207424,
+      "learning_rate": 9.892360551699779e-05,
+      "loss": 0.0032,
       "step": 1233
     },
     {
-      "epoch": 3.1620236951649057,
-      "grad_norm": 0.017608877271413803,
-      "learning_rate": 3.5745605249206816e-05,
-      "loss": 0.0045,
+      "epoch": 3.9886868686868686,
+      "grad_norm": 0.005627659149467945,
+      "learning_rate": 9.891893741017546e-05,
+      "loss": 0.0018,
       "step": 1234
     },
     {
-      "epoch": 3.164585334614153,
-      "grad_norm": 0.019648857414722443,
-      "learning_rate": 3.5659838364445505e-05,
-      "loss": 0.0042,
+      "epoch": 3.991919191919192,
+      "grad_norm": 0.00928240641951561,
+      "learning_rate": 9.891425931354317e-05,
+      "loss": 0.0032,
       "step": 1235
     },
     {
-      "epoch": 3.1671469740634004,
-      "grad_norm": 0.01928468607366085,
-      "learning_rate": 3.557411743115683e-05,
-      "loss": 0.0041,
+      "epoch": 3.995151515151515,
+      "grad_norm": 0.00606192834675312,
+      "learning_rate": 9.890957122805624e-05,
+      "loss": 0.0017,
       "step": 1236
     },
     {
-      "epoch": 3.169708613512648,
-      "grad_norm": 0.01620553433895111,
-      "learning_rate": 3.548844272402415e-05,
-      "loss": 0.0039,
+      "epoch": 3.9983838383838384,
+      "grad_norm": 0.008535273373126984,
+      "learning_rate": 9.890487315467205e-05,
+      "loss": 0.0025,
       "step": 1237
     },
     {
-      "epoch": 3.1722702529618956,
-      "grad_norm": 0.019136866554617882,
-      "learning_rate": 3.5402814517582644e-05,
-      "loss": 0.0047,
+      "epoch": 4.001616161616162,
+      "grad_norm": 0.007840057834982872,
+      "learning_rate": 9.890016509434998e-05,
+      "loss": 0.0022,
       "step": 1238
     },
     {
-      "epoch": 3.174831892411143,
-      "grad_norm": 0.022867608815431595,
-      "learning_rate": 3.531723308621847e-05,
-      "loss": 0.0046,
+      "epoch": 4.004848484848485,
+      "grad_norm": 0.007088125683367252,
+      "learning_rate": 9.889544704805151e-05,
+      "loss": 0.0025,
       "step": 1239
     },
     {
-      "epoch": 3.1773935318603908,
-      "grad_norm": 0.027643466368317604,
-      "learning_rate": 3.523169870416795e-05,
-      "loss": 0.0055,
+      "epoch": 4.008080808080808,
+      "grad_norm": 0.004251295235008001,
+      "learning_rate": 9.889071901674008e-05,
+      "loss": 0.0011,
       "step": 1240
     },
     {
-      "epoch": 3.179955171309638,
-      "grad_norm": 0.029165679588913918,
-      "learning_rate": 3.5146211645516615e-05,
-      "loss": 0.0069,
+      "epoch": 4.011313131313131,
+      "grad_norm": 0.00552331143990159,
+      "learning_rate": 9.888598100138123e-05,
+      "loss": 0.0018,
       "step": 1241
     },
     {
-      "epoch": 3.1825168107588855,
-      "grad_norm": 0.021566109731793404,
-      "learning_rate": 3.506077218419832e-05,
-      "loss": 0.005,
+      "epoch": 4.014545454545455,
+      "grad_norm": 0.006929084192961454,
+      "learning_rate": 9.888123300294249e-05,
+      "loss": 0.0019,
       "step": 1242
     },
     {
-      "epoch": 3.1850784502081333,
-      "grad_norm": 0.023212000727653503,
-      "learning_rate": 3.497538059399447e-05,
-      "loss": 0.0045,
+      "epoch": 4.017777777777778,
+      "grad_norm": 0.008191565982997417,
+      "learning_rate": 9.887647502239348e-05,
+      "loss": 0.0016,
       "step": 1243
     },
     {
-      "epoch": 3.1876400896573807,
-      "grad_norm": 0.022735675796866417,
-      "learning_rate": 3.489003714853303e-05,
-      "loss": 0.005,
+      "epoch": 4.021010101010101,
+      "grad_norm": 0.005747498944401741,
+      "learning_rate": 9.887170706070584e-05,
+      "loss": 0.001,
       "step": 1244
     },
     {
-      "epoch": 3.1902017291066285,
-      "grad_norm": 0.018199926242232323,
-      "learning_rate": 3.480474212128766e-05,
-      "loss": 0.0038,
+      "epoch": 4.024242424242424,
+      "grad_norm": 0.00564873032271862,
+      "learning_rate": 9.886692911885322e-05,
+      "loss": 0.0014,
       "step": 1245
     },
     {
-      "epoch": 3.192763368555876,
-      "grad_norm": 0.027111992239952087,
-      "learning_rate": 3.471949578557695e-05,
-      "loss": 0.0051,
+      "epoch": 4.027474747474748,
+      "grad_norm": 0.008652864024043083,
+      "learning_rate": 9.886214119781137e-05,
+      "loss": 0.0025,
       "step": 1246
     },
     {
-      "epoch": 3.195325008005123,
-      "grad_norm": 0.0206306129693985,
-      "learning_rate": 3.463429841456339e-05,
-      "loss": 0.005,
+      "epoch": 4.030707070707071,
+      "grad_norm": 0.004413561895489693,
+      "learning_rate": 9.885734329855798e-05,
+      "loss": 0.0009,
       "step": 1247
     },
     {
-      "epoch": 3.1978866474543706,
-      "grad_norm": 0.015546630136668682,
-      "learning_rate": 3.4549150281252636e-05,
-      "loss": 0.0029,
+      "epoch": 4.033939393939394,
+      "grad_norm": 0.010519410483539104,
+      "learning_rate": 9.885253542207288e-05,
+      "loss": 0.0022,
       "step": 1248
     },
     {
-      "epoch": 3.2004482869036184,
-      "grad_norm": 0.017362765967845917,
-      "learning_rate": 3.446405165849248e-05,
-      "loss": 0.0041,
+      "epoch": 4.037171717171717,
+      "grad_norm": 0.007141730282455683,
+      "learning_rate": 9.884771756933788e-05,
+      "loss": 0.0018,
       "step": 1249
     },
     {
-      "epoch": 3.2030099263528657,
-      "grad_norm": 0.016959575936198235,
-      "learning_rate": 3.4379002818972124e-05,
-      "loss": 0.0035,
+      "epoch": 4.040404040404041,
+      "grad_norm": 0.004019517917186022,
+      "learning_rate": 9.884288974133684e-05,
+      "loss": 0.0009,
       "step": 1250
     },
     {
-      "epoch": 3.2030099263528657,
-      "eval_loss": 0.007326140534132719,
-      "eval_runtime": 15.7231,
-      "eval_samples_per_second": 6.36,
-      "eval_steps_per_second": 1.59,
+      "epoch": 4.040404040404041,
+      "eval_loss": 0.001998462714254856,
+      "eval_runtime": 15.6607,
+      "eval_samples_per_second": 6.385,
+      "eval_steps_per_second": 1.596,
       "step": 1250
     },
     {
-      "epoch": 3.2055715658021136,
-      "grad_norm": 0.019312283024191856,
-      "learning_rate": 3.429400403522125e-05,
-      "loss": 0.0047,
+      "epoch": 4.043636363636364,
+      "grad_norm": 0.009633654728531837,
+      "learning_rate": 9.883805193905567e-05,
+      "loss": 0.0015,
       "step": 1251
     },
     {
-      "epoch": 3.208133205251361,
-      "grad_norm": 0.018146058544516563,
-      "learning_rate": 3.420905557960911e-05,
-      "loss": 0.0039,
+      "epoch": 4.046868686868687,
+      "grad_norm": 0.01062705833464861,
+      "learning_rate": 9.883320416348227e-05,
+      "loss": 0.0018,
       "step": 1252
     },
     {
-      "epoch": 3.2106948447006083,
-      "grad_norm": 0.016642430797219276,
-      "learning_rate": 3.412415772434368e-05,
-      "loss": 0.0037,
+      "epoch": 4.05010101010101,
+      "grad_norm": 0.008980722166597843,
+      "learning_rate": 9.882834641560666e-05,
+      "loss": 0.0015,
       "step": 1253
     },
     {
-      "epoch": 3.213256484149856,
-      "grad_norm": 0.018599150702357292,
-      "learning_rate": 3.4039310741470855e-05,
-      "loss": 0.0046,
+      "epoch": 4.053333333333334,
+      "grad_norm": 0.004700644873082638,
+      "learning_rate": 9.882347869642081e-05,
+      "loss": 0.0012,
       "step": 1254
     },
     {
-      "epoch": 3.2158181235991035,
-      "grad_norm": 0.01898500695824623,
-      "learning_rate": 3.3954514902873425e-05,
-      "loss": 0.0053,
+      "epoch": 4.056565656565657,
+      "grad_norm": 0.00533246947452426,
+      "learning_rate": 9.881860100691878e-05,
+      "loss": 0.0012,
       "step": 1255
     },
     {
-      "epoch": 3.218379763048351,
-      "grad_norm": 0.035850610584020615,
-      "learning_rate": 3.3869770480270354e-05,
-      "loss": 0.0046,
+      "epoch": 4.05979797979798,
+      "grad_norm": 0.00716199679300189,
+      "learning_rate": 9.881371334809666e-05,
+      "loss": 0.0012,
       "step": 1256
     },
     {
-      "epoch": 3.2209414024975986,
-      "grad_norm": 0.017346277832984924,
-      "learning_rate": 3.3785077745215873e-05,
-      "loss": 0.0043,
+      "epoch": 4.063030303030303,
+      "grad_norm": 0.006241815630346537,
+      "learning_rate": 9.880881572095256e-05,
+      "loss": 0.0013,
       "step": 1257
     },
     {
-      "epoch": 3.223503041946846,
-      "grad_norm": 0.02053334005177021,
-      "learning_rate": 3.370043696909849e-05,
-      "loss": 0.005,
+      "epoch": 4.066262626262627,
+      "grad_norm": 0.010392213240265846,
+      "learning_rate": 9.880390812648662e-05,
+      "loss": 0.0022,
       "step": 1258
     },
     {
-      "epoch": 3.2260646813960934,
-      "grad_norm": 0.012268176302313805,
-      "learning_rate": 3.36158484231403e-05,
-      "loss": 0.0022,
+      "epoch": 4.069494949494949,
+      "grad_norm": 0.007912958040833473,
+      "learning_rate": 9.879899056570101e-05,
+      "loss": 0.0019,
       "step": 1259
     },
     {
-      "epoch": 3.228626320845341,
-      "grad_norm": 0.021979577839374542,
-      "learning_rate": 3.3531312378396026e-05,
-      "loss": 0.006,
+      "epoch": 4.072727272727272,
+      "grad_norm": 0.005262615624815226,
+      "learning_rate": 9.879406303959999e-05,
+      "loss": 0.0013,
       "step": 1260
     },
     {
-      "epoch": 3.2311879602945885,
-      "grad_norm": 0.01328719686716795,
-      "learning_rate": 3.34468291057521e-05,
-      "loss": 0.003,
+      "epoch": 4.075959595959596,
+      "grad_norm": 0.006423169281333685,
+      "learning_rate": 9.878912554918982e-05,
+      "loss": 0.0014,
       "step": 1261
     },
     {
-      "epoch": 3.233749599743836,
-      "grad_norm": 0.019954100251197815,
-      "learning_rate": 3.3362398875925904e-05,
-      "loss": 0.0057,
+      "epoch": 4.079191919191919,
+      "grad_norm": 0.0050147781148552895,
+      "learning_rate": 9.878417809547878e-05,
+      "loss": 0.001,
       "step": 1262
     },
     {
-      "epoch": 3.2363112391930837,
-      "grad_norm": 0.015923641622066498,
-      "learning_rate": 3.327802195946486e-05,
-      "loss": 0.0042,
+      "epoch": 4.082424242424242,
+      "grad_norm": 0.007317288313060999,
+      "learning_rate": 9.877922067947717e-05,
+      "loss": 0.0013,
       "step": 1263
     },
     {
-      "epoch": 3.238872878642331,
-      "grad_norm": 0.018225105479359627,
-      "learning_rate": 3.319369862674547e-05,
-      "loss": 0.0047,
+      "epoch": 4.085656565656565,
+      "grad_norm": 0.0065062460489571095,
+      "learning_rate": 9.877425330219739e-05,
+      "loss": 0.0014,
       "step": 1264
     },
     {
-      "epoch": 3.2414345180915785,
-      "grad_norm": 0.02089548483490944,
-      "learning_rate": 3.310942914797265e-05,
-      "loss": 0.0057,
+      "epoch": 4.088888888888889,
+      "grad_norm": 0.011141772381961346,
+      "learning_rate": 9.876927596465381e-05,
+      "loss": 0.0019,
       "step": 1265
     },
     {
-      "epoch": 3.2439961575408263,
-      "grad_norm": 0.015877477824687958,
-      "learning_rate": 3.3025213793178646e-05,
-      "loss": 0.004,
+      "epoch": 4.092121212121212,
+      "grad_norm": 0.007965835742652416,
+      "learning_rate": 9.876428866786288e-05,
+      "loss": 0.0018,
       "step": 1266
     },
     {
-      "epoch": 3.2465577969900736,
-      "grad_norm": 0.020146911963820457,
-      "learning_rate": 3.294105283222234e-05,
-      "loss": 0.0044,
+      "epoch": 4.095353535353535,
+      "grad_norm": 0.006193566136062145,
+      "learning_rate": 9.875929141284306e-05,
+      "loss": 0.0013,
       "step": 1267
     },
     {
-      "epoch": 3.249119436439321,
-      "grad_norm": 0.017449114471673965,
-      "learning_rate": 3.285694653478827e-05,
-      "loss": 0.0036,
+      "epoch": 4.098585858585858,
+      "grad_norm": 0.00881227944046259,
+      "learning_rate": 9.875428420061483e-05,
+      "loss": 0.0023,
       "step": 1268
     },
     {
-      "epoch": 3.251681075888569,
-      "grad_norm": 0.019350802525877953,
-      "learning_rate": 3.277289517038582e-05,
-      "loss": 0.0046,
+      "epoch": 4.101818181818182,
+      "grad_norm": 0.00721417460590601,
+      "learning_rate": 9.874926703220073e-05,
+      "loss": 0.0019,
       "step": 1269
     },
     {
-      "epoch": 3.254242715337816,
-      "grad_norm": 0.019596491008996964,
-      "learning_rate": 3.2688899008348386e-05,
-      "loss": 0.005,
+      "epoch": 4.105050505050505,
+      "grad_norm": 0.005325607024133205,
+      "learning_rate": 9.874423990862533e-05,
+      "loss": 0.0015,
       "step": 1270
     },
     {
-      "epoch": 3.2568043547870635,
-      "grad_norm": 0.037712205201387405,
-      "learning_rate": 3.26049583178324e-05,
-      "loss": 0.006,
+      "epoch": 4.108282828282828,
+      "grad_norm": 0.007099227048456669,
+      "learning_rate": 9.873920283091521e-05,
+      "loss": 0.0016,
       "step": 1271
     },
     {
-      "epoch": 3.2593659942363113,
-      "grad_norm": 0.016859978437423706,
-      "learning_rate": 3.25210733678166e-05,
-      "loss": 0.0036,
+      "epoch": 4.111515151515151,
+      "grad_norm": 0.006921656895428896,
+      "learning_rate": 9.873415580009901e-05,
+      "loss": 0.0017,
       "step": 1272
     },
     {
-      "epoch": 3.2619276336855587,
-      "grad_norm": 0.01952943578362465,
-      "learning_rate": 3.243724442710112e-05,
-      "loss": 0.0043,
+      "epoch": 4.114747474747475,
+      "grad_norm": 0.011020069941878319,
+      "learning_rate": 9.872909881720741e-05,
+      "loss": 0.0019,
       "step": 1273
     },
     {
-      "epoch": 3.2644892731348065,
-      "grad_norm": 0.021825823932886124,
-      "learning_rate": 3.235347176430656e-05,
-      "loss": 0.0054,
+      "epoch": 4.117979797979798,
+      "grad_norm": 0.0061460561119019985,
+      "learning_rate": 9.872403188327308e-05,
+      "loss": 0.0014,
       "step": 1274
     },
     {
-      "epoch": 3.267050912584054,
-      "grad_norm": 0.027096325531601906,
-      "learning_rate": 3.226975564787322e-05,
-      "loss": 0.0065,
+      "epoch": 4.121212121212121,
+      "grad_norm": 0.012745308689773083,
+      "learning_rate": 9.871895499933075e-05,
+      "loss": 0.0015,
       "step": 1275
     },
     {
-      "epoch": 3.267050912584054,
-      "eval_loss": 0.0074441577307879925,
-      "eval_runtime": 15.7576,
-      "eval_samples_per_second": 6.346,
-      "eval_steps_per_second": 1.587,
+      "epoch": 4.121212121212121,
+      "eval_loss": 0.0018330957973375916,
+      "eval_runtime": 15.6082,
+      "eval_samples_per_second": 6.407,
+      "eval_steps_per_second": 1.602,
       "step": 1275
     },
     {
-      "epoch": 3.2696125520333013,
-      "grad_norm": 0.017573341727256775,
-      "learning_rate": 3.218609634606021e-05,
-      "loss": 0.004,
+      "epoch": 4.124444444444444,
+      "grad_norm": 0.008913307450711727,
+      "learning_rate": 9.871386816641718e-05,
+      "loss": 0.0017,
       "step": 1276
     },
     {
-      "epoch": 3.2721741914825486,
-      "grad_norm": 0.017359409481287003,
-      "learning_rate": 3.2102494126944574e-05,
-      "loss": 0.0039,
+      "epoch": 4.127676767676768,
+      "grad_norm": 0.004997677635401487,
+      "learning_rate": 9.870877138557116e-05,
+      "loss": 0.001,
       "step": 1277
     },
     {
-      "epoch": 3.2747358309317964,
-      "grad_norm": 0.025902550667524338,
-      "learning_rate": 3.201894925842041e-05,
-      "loss": 0.0046,
+      "epoch": 4.130909090909091,
+      "grad_norm": 0.003410317236557603,
+      "learning_rate": 9.870366465783351e-05,
+      "loss": 0.001,
       "step": 1278
     },
     {
-      "epoch": 3.277297470381044,
-      "grad_norm": 0.02130158431828022,
-      "learning_rate": 3.1935462008198116e-05,
-      "loss": 0.0034,
+      "epoch": 4.134141414141414,
+      "grad_norm": 0.008104338310658932,
+      "learning_rate": 9.869854798424709e-05,
+      "loss": 0.0023,
       "step": 1279
     },
     {
-      "epoch": 3.2798591098302916,
-      "grad_norm": 0.021285217255353928,
-      "learning_rate": 3.185203264380338e-05,
-      "loss": 0.0045,
+      "epoch": 4.137373737373737,
+      "grad_norm": 0.007407724391669035,
+      "learning_rate": 9.869342136585677e-05,
+      "loss": 0.0013,
       "step": 1280
     },
     {
-      "epoch": 3.282420749279539,
-      "grad_norm": 0.017796866595745087,
-      "learning_rate": 3.176866143257643e-05,
-      "loss": 0.0032,
+      "epoch": 4.140606060606061,
+      "grad_norm": 0.011552196927368641,
+      "learning_rate": 9.868828480370948e-05,
+      "loss": 0.0019,
       "step": 1281
     },
     {
-      "epoch": 3.2849823887287863,
-      "grad_norm": 0.022058211266994476,
-      "learning_rate": 3.1685348641671196e-05,
-      "loss": 0.004,
+      "epoch": 4.143838383838384,
+      "grad_norm": 0.005420271307229996,
+      "learning_rate": 9.868313829885414e-05,
+      "loss": 0.0011,
       "step": 1282
     },
     {
-      "epoch": 3.2875440281780337,
-      "grad_norm": 0.018691932782530785,
-      "learning_rate": 3.1602094538054326e-05,
-      "loss": 0.0037,
+      "epoch": 4.147070707070707,
+      "grad_norm": 0.00629067700356245,
+      "learning_rate": 9.867798185234176e-05,
+      "loss": 0.001,
       "step": 1283
     },
     {
-      "epoch": 3.2901056676272815,
-      "grad_norm": 0.013120202347636223,
-      "learning_rate": 3.151889938850445e-05,
-      "loss": 0.0019,
+      "epoch": 4.15030303030303,
+      "grad_norm": 0.007845763117074966,
+      "learning_rate": 9.867281546522533e-05,
+      "loss": 0.0017,
       "step": 1284
     },
     {
-      "epoch": 3.292667307076529,
-      "grad_norm": 0.02618294209241867,
-      "learning_rate": 3.143576345961132e-05,
-      "loss": 0.0046,
+      "epoch": 4.153535353535354,
+      "grad_norm": 0.00859200581908226,
+      "learning_rate": 9.866763913855988e-05,
+      "loss": 0.0018,
       "step": 1285
     },
     {
-      "epoch": 3.2952289465257767,
-      "grad_norm": 0.018655234947800636,
-      "learning_rate": 3.1352687017774864e-05,
-      "loss": 0.0042,
+      "epoch": 4.156767676767677,
+      "grad_norm": 0.0041308156214654446,
+      "learning_rate": 9.866245287340247e-05,
+      "loss": 0.0007,
       "step": 1286
     },
     {
-      "epoch": 3.297790585975024,
-      "grad_norm": 0.023432090878486633,
-      "learning_rate": 3.12696703292044e-05,
-      "loss": 0.0044,
+      "epoch": 4.16,
+      "grad_norm": 0.007610360160470009,
+      "learning_rate": 9.865725667081221e-05,
+      "loss": 0.0012,
       "step": 1287
     },
     {
-      "epoch": 3.3003522254242714,
-      "grad_norm": 0.026196971535682678,
-      "learning_rate": 3.118671365991781e-05,
-      "loss": 0.0049,
+      "epoch": 4.163232323232323,
+      "grad_norm": 0.009420975111424923,
+      "learning_rate": 9.865205053185023e-05,
+      "loss": 0.002,
       "step": 1288
     },
     {
-      "epoch": 3.3029138648735192,
-      "grad_norm": 0.0182013139128685,
-      "learning_rate": 3.110381727574063e-05,
-      "loss": 0.0034,
+      "epoch": 4.166464646464647,
+      "grad_norm": 0.004176929127424955,
+      "learning_rate": 9.864683445757966e-05,
+      "loss": 0.0007,
       "step": 1289
     },
     {
-      "epoch": 3.3054755043227666,
-      "grad_norm": 0.021666232496500015,
-      "learning_rate": 3.1020981442305184e-05,
-      "loss": 0.0039,
+      "epoch": 4.16969696969697,
+      "grad_norm": 0.011829263530671597,
+      "learning_rate": 9.86416084490657e-05,
+      "loss": 0.0049,
       "step": 1290
     },
     {
-      "epoch": 3.308037143772014,
-      "grad_norm": 0.021782975643873215,
-      "learning_rate": 3.093820642504983e-05,
-      "loss": 0.0047,
+      "epoch": 4.172929292929293,
+      "grad_norm": 0.005556262563914061,
+      "learning_rate": 9.863637250737557e-05,
+      "loss": 0.0012,
       "step": 1291
     },
     {
-      "epoch": 3.3105987832212618,
-      "grad_norm": 0.022409774363040924,
-      "learning_rate": 3.085549248921803e-05,
-      "loss": 0.0037,
+      "epoch": 4.176161616161616,
+      "grad_norm": 0.006624159868806601,
+      "learning_rate": 9.863112663357847e-05,
+      "loss": 0.0015,
       "step": 1292
     },
     {
-      "epoch": 3.313160422670509,
-      "grad_norm": 0.017817934975028038,
-      "learning_rate": 3.0772839899857464e-05,
-      "loss": 0.0038,
+      "epoch": 4.17939393939394,
+      "grad_norm": 0.007349417544901371,
+      "learning_rate": 9.86258708287457e-05,
+      "loss": 0.0018,
       "step": 1293
     },
     {
-      "epoch": 3.3157220621197565,
-      "grad_norm": 0.024781012907624245,
-      "learning_rate": 3.06902489218193e-05,
-      "loss": 0.005,
+      "epoch": 4.182626262626263,
+      "grad_norm": 0.005895329639315605,
+      "learning_rate": 9.862060509395056e-05,
+      "loss": 0.0013,
       "step": 1294
     },
     {
-      "epoch": 3.3182837015690043,
-      "grad_norm": 0.023359034210443497,
-      "learning_rate": 3.060771981975726e-05,
-      "loss": 0.0056,
+      "epoch": 4.185858585858586,
+      "grad_norm": 0.004604201763868332,
+      "learning_rate": 9.861532943026836e-05,
+      "loss": 0.0009,
       "step": 1295
     },
     {
-      "epoch": 3.3208453410182517,
-      "grad_norm": 0.01976236328482628,
-      "learning_rate": 3.052525285812677e-05,
-      "loss": 0.0054,
+      "epoch": 4.189090909090909,
+      "grad_norm": 0.008079406805336475,
+      "learning_rate": 9.861004383877645e-05,
+      "loss": 0.0015,
       "step": 1296
     },
     {
-      "epoch": 3.323406980467499,
-      "grad_norm": 0.030070582404732704,
-      "learning_rate": 3.0442848301184158e-05,
-      "loss": 0.0062,
+      "epoch": 4.192323232323233,
+      "grad_norm": 0.01320496667176485,
+      "learning_rate": 9.860474832055421e-05,
+      "loss": 0.0018,
       "step": 1297
     },
     {
-      "epoch": 3.325968619916747,
-      "grad_norm": 0.028238259255886078,
-      "learning_rate": 3.036050641298578e-05,
-      "loss": 0.006,
+      "epoch": 4.195555555555556,
+      "grad_norm": 0.009637363255023956,
+      "learning_rate": 9.859944287668306e-05,
+      "loss": 0.0026,
       "step": 1298
     },
     {
-      "epoch": 3.3285302593659942,
-      "grad_norm": 0.015943922102451324,
-      "learning_rate": 3.0278227457387147e-05,
-      "loss": 0.0036,
+      "epoch": 4.198787878787879,
+      "grad_norm": 0.010264774784445763,
+      "learning_rate": 9.85941275082464e-05,
+      "loss": 0.0017,
       "step": 1299
     },
     {
-      "epoch": 3.3310918988152416,
-      "grad_norm": 0.018986733630299568,
-      "learning_rate": 3.019601169804216e-05,
-      "loss": 0.0039,
+      "epoch": 4.202020202020202,
+      "grad_norm": 0.005758476909250021,
+      "learning_rate": 9.858880221632973e-05,
+      "loss": 0.0011,
       "step": 1300
     },
     {
-      "epoch": 3.3310918988152416,
-      "eval_loss": 0.007093509659171104,
-      "eval_runtime": 15.7479,
-      "eval_samples_per_second": 6.35,
-      "eval_steps_per_second": 1.588,
+      "epoch": 4.202020202020202,
+      "eval_loss": 0.0018192220013588667,
+      "eval_runtime": 15.6044,
+      "eval_samples_per_second": 6.408,
+      "eval_steps_per_second": 1.602,
       "step": 1300
     },
     {
-      "epoch": 3.3336535382644894,
-      "grad_norm": 0.018647341057658195,
-      "learning_rate": 3.0113859398402188e-05,
-      "loss": 0.0035,
+      "epoch": 4.205252525252526,
+      "grad_norm": 0.005597962997853756,
+      "learning_rate": 9.85834670020205e-05,
+      "loss": 0.0011,
       "step": 1301
     },
     {
-      "epoch": 3.3362151777137368,
-      "grad_norm": 0.04162399098277092,
-      "learning_rate": 3.003177082171523e-05,
-      "loss": 0.0153,
+      "epoch": 4.208484848484849,
+      "grad_norm": 0.005339125171303749,
+      "learning_rate": 9.857812186640824e-05,
+      "loss": 0.0013,
       "step": 1302
     },
     {
-      "epoch": 3.338776817162984,
-      "grad_norm": 0.038681481033563614,
-      "learning_rate": 2.994974623102512e-05,
-      "loss": 0.0063,
+      "epoch": 4.211717171717171,
+      "grad_norm": 0.006229112856090069,
+      "learning_rate": 9.857276681058452e-05,
+      "loss": 0.0015,
       "step": 1303
     },
     {
-      "epoch": 3.341338456612232,
-      "grad_norm": 0.02795899473130703,
-      "learning_rate": 2.9867785889170662e-05,
-      "loss": 0.0055,
+      "epoch": 4.214949494949495,
+      "grad_norm": 0.004702513571828604,
+      "learning_rate": 9.856740183564284e-05,
+      "loss": 0.0011,
       "step": 1304
     },
     {
-      "epoch": 3.3439000960614793,
-      "grad_norm": 0.021580150350928307,
-      "learning_rate": 2.978589005878476e-05,
-      "loss": 0.0048,
+      "epoch": 4.218181818181818,
+      "grad_norm": 0.013774113729596138,
+      "learning_rate": 9.856202694267882e-05,
+      "loss": 0.0024,
       "step": 1305
     },
     {
-      "epoch": 3.3464617355107267,
-      "grad_norm": 0.02223021350800991,
-      "learning_rate": 2.9704059002293593e-05,
-      "loss": 0.0067,
+      "epoch": 4.221414141414141,
+      "grad_norm": 0.010066622868180275,
+      "learning_rate": 9.855664213279008e-05,
+      "loss": 0.0014,
       "step": 1306
     },
     {
-      "epoch": 3.3490233749599745,
-      "grad_norm": 0.016967158764600754,
-      "learning_rate": 2.9622292981915806e-05,
-      "loss": 0.0037,
+      "epoch": 4.224646464646464,
+      "grad_norm": 0.007124657742679119,
+      "learning_rate": 9.855124740707627e-05,
+      "loss": 0.002,
       "step": 1307
     },
     {
-      "epoch": 3.351585014409222,
-      "grad_norm": 0.022687457501888275,
-      "learning_rate": 2.9540592259661637e-05,
-      "loss": 0.0058,
+      "epoch": 4.227878787878788,
+      "grad_norm": 0.006717347074300051,
+      "learning_rate": 9.854584276663903e-05,
+      "loss": 0.0013,
       "step": 1308
     },
     {
-      "epoch": 3.3541466538584697,
-      "grad_norm": 0.022553442046046257,
-      "learning_rate": 2.945895709733205e-05,
-      "loss": 0.0035,
+      "epoch": 4.231111111111111,
+      "grad_norm": 0.0060617681592702866,
+      "learning_rate": 9.854042821258205e-05,
+      "loss": 0.0011,
       "step": 1309
     },
     {
-      "epoch": 3.356708293307717,
-      "grad_norm": 0.022646233439445496,
-      "learning_rate": 2.9377387756517982e-05,
-      "loss": 0.0056,
+      "epoch": 4.234343434343434,
+      "grad_norm": 0.006181581411510706,
+      "learning_rate": 9.853500374601106e-05,
+      "loss": 0.0013,
       "step": 1310
     },
     {
-      "epoch": 3.3592699327569644,
-      "grad_norm": 0.035119421780109406,
-      "learning_rate": 2.9295884498599414e-05,
-      "loss": 0.009,
+      "epoch": 4.237575757575757,
+      "grad_norm": 0.005808467045426369,
+      "learning_rate": 9.85295693680338e-05,
+      "loss": 0.0016,
       "step": 1311
     },
     {
-      "epoch": 3.3618315722062118,
-      "grad_norm": 0.015311050228774548,
-      "learning_rate": 2.9214447584744565e-05,
-      "loss": 0.0041,
+      "epoch": 4.240808080808081,
+      "grad_norm": 0.004074588883668184,
+      "learning_rate": 9.852412507976002e-05,
+      "loss": 0.0009,
       "step": 1312
     },
     {
-      "epoch": 3.3643932116554596,
-      "grad_norm": 0.021092480048537254,
-      "learning_rate": 2.9133077275909108e-05,
-      "loss": 0.0047,
+      "epoch": 4.244040404040404,
+      "grad_norm": 0.0065651689656078815,
+      "learning_rate": 9.851867088230152e-05,
+      "loss": 0.0018,
       "step": 1313
     },
     {
-      "epoch": 3.366954851104707,
-      "grad_norm": 0.017000602558255196,
-      "learning_rate": 2.905177383283526e-05,
-      "loss": 0.0041,
+      "epoch": 4.247272727272727,
+      "grad_norm": 0.007975611835718155,
+      "learning_rate": 9.85132067767721e-05,
+      "loss": 0.0015,
       "step": 1314
     },
     {
-      "epoch": 3.3695164905539547,
-      "grad_norm": 0.020628029480576515,
-      "learning_rate": 2.897053751605093e-05,
-      "loss": 0.0049,
+      "epoch": 4.25050505050505,
+      "grad_norm": 0.010979430750012398,
+      "learning_rate": 9.850773276428759e-05,
+      "loss": 0.0016,
       "step": 1315
     },
     {
-      "epoch": 3.372078130003202,
-      "grad_norm": 0.014183783903717995,
-      "learning_rate": 2.8889368585869014e-05,
-      "loss": 0.0029,
+      "epoch": 4.253737373737374,
+      "grad_norm": 0.006963154301047325,
+      "learning_rate": 9.850224884596585e-05,
+      "loss": 0.0013,
       "step": 1316
     },
     {
-      "epoch": 3.3746397694524495,
-      "grad_norm": 0.020699163898825645,
-      "learning_rate": 2.880826730238643e-05,
-      "loss": 0.0052,
+      "epoch": 4.256969696969697,
+      "grad_norm": 0.00852319598197937,
+      "learning_rate": 9.849675502292676e-05,
+      "loss": 0.0017,
       "step": 1317
     },
     {
-      "epoch": 3.377201408901697,
-      "grad_norm": 0.017421424388885498,
-      "learning_rate": 2.872723392548332e-05,
-      "loss": 0.0051,
+      "epoch": 4.26020202020202,
+      "grad_norm": 0.004744053818285465,
+      "learning_rate": 9.849125129629224e-05,
+      "loss": 0.0011,
       "step": 1318
     },
     {
-      "epoch": 3.3797630483509447,
-      "grad_norm": 0.016957523301243782,
-      "learning_rate": 2.8646268714822227e-05,
-      "loss": 0.0038,
+      "epoch": 4.263434343434343,
+      "grad_norm": 0.005866665858775377,
+      "learning_rate": 9.848573766718617e-05,
+      "loss": 0.0013,
       "step": 1319
     },
     {
-      "epoch": 3.382324687800192,
-      "grad_norm": 0.019371114671230316,
-      "learning_rate": 2.8565371929847284e-05,
-      "loss": 0.0036,
+      "epoch": 4.266666666666667,
+      "grad_norm": 0.009565476328134537,
+      "learning_rate": 9.848021413673454e-05,
+      "loss": 0.0016,
       "step": 1320
     },
     {
-      "epoch": 3.38488632724944,
-      "grad_norm": 0.01660952717065811,
-      "learning_rate": 2.8484543829783378e-05,
-      "loss": 0.0045,
+      "epoch": 4.26989898989899,
+      "grad_norm": 0.01027013547718525,
+      "learning_rate": 9.847468070606529e-05,
+      "loss": 0.0023,
       "step": 1321
     },
     {
-      "epoch": 3.387447966698687,
-      "grad_norm": 0.014318636618554592,
-      "learning_rate": 2.840378467363523e-05,
-      "loss": 0.0034,
+      "epoch": 4.273131313131313,
+      "grad_norm": 0.009343735873699188,
+      "learning_rate": 9.846913737630843e-05,
+      "loss": 0.0022,
       "step": 1322
     },
     {
-      "epoch": 3.3900096061479346,
-      "grad_norm": 0.02716119773685932,
-      "learning_rate": 2.8323094720186728e-05,
-      "loss": 0.0077,
+      "epoch": 4.276363636363636,
+      "grad_norm": 0.004626363515853882,
+      "learning_rate": 9.846358414859597e-05,
+      "loss": 0.001,
       "step": 1323
     },
     {
-      "epoch": 3.3925712455971824,
-      "grad_norm": 0.01730690523982048,
-      "learning_rate": 2.824247422799995e-05,
-      "loss": 0.0039,
+      "epoch": 4.27959595959596,
+      "grad_norm": 0.007942027412354946,
+      "learning_rate": 9.845802102406192e-05,
+      "loss": 0.0016,
       "step": 1324
     },
     {
-      "epoch": 3.3951328850464297,
-      "grad_norm": 0.018140211701393127,
-      "learning_rate": 2.8161923455414367e-05,
-      "loss": 0.0047,
+      "epoch": 4.282828282828283,
+      "grad_norm": 0.008563620038330555,
+      "learning_rate": 9.845244800384237e-05,
+      "loss": 0.002,
       "step": 1325
     },
     {
-      "epoch": 3.3951328850464297,
-      "eval_loss": 0.006955994758754969,
-      "eval_runtime": 15.7837,
-      "eval_samples_per_second": 6.336,
-      "eval_steps_per_second": 1.584,
+      "epoch": 4.282828282828283,
+      "eval_loss": 0.0018768340814858675,
+      "eval_runtime": 15.5889,
+      "eval_samples_per_second": 6.415,
+      "eval_steps_per_second": 1.604,
       "step": 1325
     },
     {
-      "epoch": 3.397694524495677,
-      "grad_norm": 0.017170537263154984,
-      "learning_rate": 2.8081442660546125e-05,
-      "loss": 0.0033,
+      "epoch": 4.286060606060606,
+      "grad_norm": 0.007845155894756317,
+      "learning_rate": 9.844686508907537e-05,
+      "loss": 0.0018,
       "step": 1326
     },
     {
-      "epoch": 3.400256163944925,
-      "grad_norm": 0.01794111169874668,
-      "learning_rate": 2.8001032101287095e-05,
-      "loss": 0.0036,
+      "epoch": 4.289292929292929,
+      "grad_norm": 0.010610824450850487,
+      "learning_rate": 9.844127228090102e-05,
+      "loss": 0.0021,
       "step": 1327
     },
     {
-      "epoch": 3.4028178033941723,
-      "grad_norm": 0.035149358212947845,
-      "learning_rate": 2.7920692035304052e-05,
-      "loss": 0.0076,
+      "epoch": 4.292525252525253,
+      "grad_norm": 0.007378740701824427,
+      "learning_rate": 9.843566958046145e-05,
+      "loss": 0.0017,
       "step": 1328
     },
     {
-      "epoch": 3.4053794428434196,
-      "grad_norm": 0.015975119546055794,
-      "learning_rate": 2.784042272003794e-05,
-      "loss": 0.0032,
+      "epoch": 4.295757575757576,
+      "grad_norm": 0.013690670020878315,
+      "learning_rate": 9.843005698890076e-05,
+      "loss": 0.0031,
       "step": 1329
     },
     {
-      "epoch": 3.4079410822926675,
-      "grad_norm": 0.018400490283966064,
-      "learning_rate": 2.776022441270295e-05,
-      "loss": 0.0046,
+      "epoch": 4.298989898989899,
+      "grad_norm": 0.005793304182589054,
+      "learning_rate": 9.842443450736514e-05,
+      "loss": 0.0013,
       "step": 1330
     },
     {
-      "epoch": 3.410502721741915,
-      "grad_norm": 0.01995062455534935,
-      "learning_rate": 2.7680097370285723e-05,
-      "loss": 0.0051,
+      "epoch": 4.302222222222222,
+      "grad_norm": 0.005671793594956398,
+      "learning_rate": 9.841880213700274e-05,
+      "loss": 0.0014,
       "step": 1331
     },
     {
-      "epoch": 3.413064361191162,
-      "grad_norm": 0.022594042122364044,
-      "learning_rate": 2.760004184954459e-05,
-      "loss": 0.0056,
+      "epoch": 4.305454545454546,
+      "grad_norm": 0.006172803230583668,
+      "learning_rate": 9.841315987896378e-05,
+      "loss": 0.0014,
       "step": 1332
     },
     {
-      "epoch": 3.41562600064041,
-      "grad_norm": 0.024578507989645004,
-      "learning_rate": 2.752005810700869e-05,
-      "loss": 0.0047,
+      "epoch": 4.308686868686869,
+      "grad_norm": 0.0046432907693088055,
+      "learning_rate": 9.840750773440046e-05,
+      "loss": 0.0009,
       "step": 1333
     },
     {
-      "epoch": 3.4181876400896574,
-      "grad_norm": 0.01541770901530981,
-      "learning_rate": 2.7440146398977097e-05,
-      "loss": 0.0036,
+      "epoch": 4.311919191919192,
+      "grad_norm": 0.010229202918708324,
+      "learning_rate": 9.840184570446702e-05,
+      "loss": 0.0027,
       "step": 1334
     },
     {
-      "epoch": 3.4207492795389047,
-      "grad_norm": 0.015968957915902138,
-      "learning_rate": 2.7360306981518146e-05,
-      "loss": 0.0038,
+      "epoch": 4.315151515151515,
+      "grad_norm": 0.00660282326862216,
+      "learning_rate": 9.839617379031971e-05,
+      "loss": 0.0012,
       "step": 1335
     },
     {
-      "epoch": 3.4233109189881525,
-      "grad_norm": 0.023852817714214325,
-      "learning_rate": 2.7280540110468455e-05,
-      "loss": 0.005,
+      "epoch": 4.318383838383839,
+      "grad_norm": 0.008186656050384045,
+      "learning_rate": 9.839049199311679e-05,
+      "loss": 0.0023,
       "step": 1336
     },
     {
-      "epoch": 3.4258725584374,
-      "grad_norm": 0.018949022516608238,
-      "learning_rate": 2.720084604143218e-05,
-      "loss": 0.0047,
+      "epoch": 4.321616161616162,
+      "grad_norm": 0.004670700989663601,
+      "learning_rate": 9.838480031401856e-05,
+      "loss": 0.001,
       "step": 1337
     },
     {
-      "epoch": 3.4284341978866473,
-      "grad_norm": 0.018564656376838684,
-      "learning_rate": 2.712122502978024e-05,
-      "loss": 0.0049,
+      "epoch": 4.324848484848485,
+      "grad_norm": 0.008229521103203297,
+      "learning_rate": 9.837909875418732e-05,
+      "loss": 0.002,
       "step": 1338
     },
     {
-      "epoch": 3.430995837335895,
-      "grad_norm": 0.01999684050679207,
-      "learning_rate": 2.7041677330649407e-05,
-      "loss": 0.0041,
+      "epoch": 4.328080808080808,
+      "grad_norm": 0.004257467109709978,
+      "learning_rate": 9.837338731478741e-05,
+      "loss": 0.0011,
       "step": 1339
     },
     {
-      "epoch": 3.4335574767851424,
-      "grad_norm": 0.014954477548599243,
-      "learning_rate": 2.6962203198941587e-05,
-      "loss": 0.0026,
+      "epoch": 4.331313131313132,
+      "grad_norm": 0.006889670621603727,
+      "learning_rate": 9.836766599698514e-05,
+      "loss": 0.0014,
       "step": 1340
     },
     {
-      "epoch": 3.43611911623439,
-      "grad_norm": 0.023138830438256264,
-      "learning_rate": 2.6882802889322834e-05,
-      "loss": 0.0051,
+      "epoch": 4.334545454545455,
+      "grad_norm": 0.0060333143919706345,
+      "learning_rate": 9.83619348019489e-05,
+      "loss": 0.0011,
       "step": 1341
     },
     {
-      "epoch": 3.4386807556836376,
-      "grad_norm": 0.01983695477247238,
-      "learning_rate": 2.680347665622275e-05,
-      "loss": 0.0032,
+      "epoch": 4.337777777777778,
+      "grad_norm": 0.00593153852969408,
+      "learning_rate": 9.835619373084905e-05,
+      "loss": 0.0012,
       "step": 1342
     },
     {
-      "epoch": 3.441242395132885,
-      "grad_norm": 0.02612839825451374,
-      "learning_rate": 2.672422475383356e-05,
-      "loss": 0.0053,
+      "epoch": 4.3410101010101005,
+      "grad_norm": 0.006328411400318146,
+      "learning_rate": 9.835044278485799e-05,
+      "loss": 0.0017,
       "step": 1343
     },
     {
-      "epoch": 3.443804034582133,
-      "grad_norm": 0.015107940882444382,
-      "learning_rate": 2.6645047436109226e-05,
-      "loss": 0.0037,
+      "epoch": 4.344242424242424,
+      "grad_norm": 0.006277166772633791,
+      "learning_rate": 9.834468196515011e-05,
+      "loss": 0.0014,
       "step": 1344
     },
     {
-      "epoch": 3.44636567403138,
-      "grad_norm": 0.013161804527044296,
-      "learning_rate": 2.656594495676482e-05,
-      "loss": 0.003,
+      "epoch": 4.347474747474747,
+      "grad_norm": 0.007049013860523701,
+      "learning_rate": 9.833891127290187e-05,
+      "loss": 0.0014,
       "step": 1345
     },
     {
-      "epoch": 3.4489273134806275,
-      "grad_norm": 0.015442034229636192,
-      "learning_rate": 2.6486917569275528e-05,
-      "loss": 0.0031,
+      "epoch": 4.35070707070707,
+      "grad_norm": 0.009100987575948238,
+      "learning_rate": 9.833313070929167e-05,
+      "loss": 0.0015,
       "step": 1346
     },
     {
-      "epoch": 3.451488952929875,
-      "grad_norm": 0.019541800022125244,
-      "learning_rate": 2.64079655268759e-05,
-      "loss": 0.0045,
+      "epoch": 4.3539393939393936,
+      "grad_norm": 0.008106924593448639,
+      "learning_rate": 9.83273402755e-05,
+      "loss": 0.0018,
       "step": 1347
     },
     {
-      "epoch": 3.4540505923791227,
-      "grad_norm": 0.01665390469133854,
-      "learning_rate": 2.632908908255911e-05,
-      "loss": 0.0031,
+      "epoch": 4.357171717171717,
+      "grad_norm": 0.008905386552214622,
+      "learning_rate": 9.832153997270934e-05,
+      "loss": 0.0012,
       "step": 1348
     },
     {
-      "epoch": 3.45661223182837,
-      "grad_norm": 0.018763909116387367,
-      "learning_rate": 2.6250288489076074e-05,
-      "loss": 0.0038,
+      "epoch": 4.36040404040404,
+      "grad_norm": 0.00858787726610899,
+      "learning_rate": 9.831572980210413e-05,
+      "loss": 0.0013,
       "step": 1349
     },
     {
-      "epoch": 3.459173871277618,
-      "grad_norm": 0.031971413642168045,
-      "learning_rate": 2.6171563998934605e-05,
-      "loss": 0.0046,
+      "epoch": 4.363636363636363,
+      "grad_norm": 0.010237335227429867,
+      "learning_rate": 9.830990976487094e-05,
+      "loss": 0.0014,
       "step": 1350
     },
     {
-      "epoch": 3.459173871277618,
-      "eval_loss": 0.006668676156550646,
-      "eval_runtime": 15.761,
-      "eval_samples_per_second": 6.345,
-      "eval_steps_per_second": 1.586,
+      "epoch": 4.363636363636363,
+      "eval_loss": 0.0019568479619920254,
+      "eval_runtime": 15.6057,
+      "eval_samples_per_second": 6.408,
+      "eval_steps_per_second": 1.602,
       "step": 1350
     },
     {
-      "epoch": 3.4617355107268653,
-      "grad_norm": 0.020659996196627617,
-      "learning_rate": 2.6092915864398722e-05,
-      "loss": 0.0036,
+      "epoch": 4.366868686868687,
+      "grad_norm": 0.006910400930792093,
+      "learning_rate": 9.830407986219823e-05,
+      "loss": 0.0013,
       "step": 1351
     },
     {
-      "epoch": 3.4642971501761126,
-      "grad_norm": 0.02322494424879551,
-      "learning_rate": 2.6014344337487707e-05,
-      "loss": 0.0041,
+      "epoch": 4.37010101010101,
+      "grad_norm": 0.008381280116736889,
+      "learning_rate": 9.829824009527657e-05,
+      "loss": 0.0013,
       "step": 1352
     },
     {
-      "epoch": 3.46685878962536,
-      "grad_norm": 0.018928896635770798,
-      "learning_rate": 2.5935849669975378e-05,
-      "loss": 0.0043,
+      "epoch": 4.373333333333333,
+      "grad_norm": 0.009742970578372478,
+      "learning_rate": 9.82923904652985e-05,
+      "loss": 0.0017,
       "step": 1353
     },
     {
-      "epoch": 3.469420429074608,
-      "grad_norm": 0.02125394716858864,
-      "learning_rate": 2.5857432113389273e-05,
-      "loss": 0.0047,
+      "epoch": 4.376565656565656,
+      "grad_norm": 0.006199553608894348,
+      "learning_rate": 9.828653097345857e-05,
+      "loss": 0.001,
       "step": 1354
     },
     {
-      "epoch": 3.471982068523855,
-      "grad_norm": 0.02026151679456234,
-      "learning_rate": 2.5779091919009877e-05,
-      "loss": 0.0031,
+      "epoch": 4.37979797979798,
+      "grad_norm": 0.016486119478940964,
+      "learning_rate": 9.828066162095335e-05,
+      "loss": 0.0013,
       "step": 1355
     },
     {
-      "epoch": 3.474543707973103,
-      "grad_norm": 0.01922105997800827,
-      "learning_rate": 2.57008293378697e-05,
-      "loss": 0.004,
+      "epoch": 4.383030303030303,
+      "grad_norm": 0.007773060817271471,
+      "learning_rate": 9.827478240898145e-05,
+      "loss": 0.0011,
       "step": 1356
     },
     {
-      "epoch": 3.4771053474223503,
-      "grad_norm": 0.024818353354930878,
-      "learning_rate": 2.5622644620752606e-05,
-      "loss": 0.0043,
+      "epoch": 4.386262626262626,
+      "grad_norm": 0.012599975802004337,
+      "learning_rate": 9.826889333874348e-05,
+      "loss": 0.002,
       "step": 1357
     },
     {
-      "epoch": 3.4796669868715977,
-      "grad_norm": 0.03687607869505882,
-      "learning_rate": 2.554453801819291e-05,
-      "loss": 0.007,
+      "epoch": 4.389494949494949,
+      "grad_norm": 0.009155334904789925,
+      "learning_rate": 9.826299441144203e-05,
+      "loss": 0.0016,
       "step": 1358
     },
     {
-      "epoch": 3.4822286263208455,
-      "grad_norm": 0.02132255584001541,
-      "learning_rate": 2.546650978047467e-05,
-      "loss": 0.0046,
+      "epoch": 4.392727272727273,
+      "grad_norm": 0.008478849194943905,
+      "learning_rate": 9.825708562828173e-05,
+      "loss": 0.0013,
       "step": 1359
     },
     {
-      "epoch": 3.484790265770093,
-      "grad_norm": 0.021544774994254112,
-      "learning_rate": 2.5388560157630765e-05,
-      "loss": 0.0045,
+      "epoch": 4.395959595959596,
+      "grad_norm": 0.008158011361956596,
+      "learning_rate": 9.825116699046923e-05,
+      "loss": 0.0015,
       "step": 1360
     },
     {
-      "epoch": 3.4873519052193402,
-      "grad_norm": 0.018406540155410767,
-      "learning_rate": 2.5310689399442223e-05,
-      "loss": 0.0043,
+      "epoch": 4.399191919191919,
+      "grad_norm": 0.006658365484327078,
+      "learning_rate": 9.82452384992132e-05,
+      "loss": 0.0017,
       "step": 1361
     },
     {
-      "epoch": 3.489913544668588,
-      "grad_norm": 0.02386508323252201,
-      "learning_rate": 2.5232897755437344e-05,
-      "loss": 0.0035,
+      "epoch": 4.402424242424242,
+      "grad_norm": 0.006244409829378128,
+      "learning_rate": 9.823930015572428e-05,
+      "loss": 0.0014,
       "step": 1362
     },
     {
-      "epoch": 3.4924751841178354,
-      "grad_norm": 0.020438361912965775,
-      "learning_rate": 2.5155185474890897e-05,
-      "loss": 0.0042,
+      "epoch": 4.405656565656566,
+      "grad_norm": 0.01533744391053915,
+      "learning_rate": 9.823335196121517e-05,
+      "loss": 0.0021,
       "step": 1363
     },
     {
-      "epoch": 3.495036823567083,
-      "grad_norm": 0.0224450696259737,
-      "learning_rate": 2.5077552806823318e-05,
-      "loss": 0.0053,
+      "epoch": 4.408888888888889,
+      "grad_norm": 0.007020717486739159,
+      "learning_rate": 9.822739391690054e-05,
+      "loss": 0.0016,
       "step": 1364
     },
     {
-      "epoch": 3.4975984630163306,
-      "grad_norm": 0.024398161098361015,
-      "learning_rate": 2.500000000000001e-05,
-      "loss": 0.0051,
+      "epoch": 4.412121212121212,
+      "grad_norm": 0.007510787807404995,
+      "learning_rate": 9.82214260239971e-05,
+      "loss": 0.0012,
       "step": 1365
     },
     {
-      "epoch": 3.500160102465578,
-      "grad_norm": 0.020685985684394836,
-      "learning_rate": 2.4922527302930388e-05,
-      "loss": 0.0033,
+      "epoch": 4.415353535353535,
+      "grad_norm": 0.00850648619234562,
+      "learning_rate": 9.821544828372357e-05,
+      "loss": 0.002,
       "step": 1366
     },
     {
-      "epoch": 3.5027217419148253,
-      "grad_norm": 0.027152569964528084,
-      "learning_rate": 2.4845134963867207e-05,
-      "loss": 0.0061,
+      "epoch": 4.418585858585859,
+      "grad_norm": 0.006544474977999926,
+      "learning_rate": 9.820946069730066e-05,
+      "loss": 0.0014,
       "step": 1367
     },
     {
-      "epoch": 3.505283381364073,
-      "grad_norm": 0.022943297401070595,
-      "learning_rate": 2.476782323080574e-05,
-      "loss": 0.0064,
+      "epoch": 4.421818181818182,
+      "grad_norm": 0.009068132378160954,
+      "learning_rate": 9.820346326595111e-05,
+      "loss": 0.0016,
       "step": 1368
     },
     {
-      "epoch": 3.5078450208133205,
-      "grad_norm": 0.020930426195263863,
-      "learning_rate": 2.4690592351482934e-05,
-      "loss": 0.0039,
+      "epoch": 4.425050505050505,
+      "grad_norm": 0.005028106737881899,
+      "learning_rate": 9.819745599089967e-05,
+      "loss": 0.0011,
       "step": 1369
     },
     {
-      "epoch": 3.510406660262568,
-      "grad_norm": 0.02447371743619442,
-      "learning_rate": 2.461344257337662e-05,
-      "loss": 0.0049,
+      "epoch": 4.428282828282828,
+      "grad_norm": 0.006823462434113026,
+      "learning_rate": 9.819143887337308e-05,
+      "loss": 0.0017,
       "step": 1370
     },
     {
-      "epoch": 3.5129682997118157,
-      "grad_norm": 0.016021478921175003,
-      "learning_rate": 2.4536374143704838e-05,
-      "loss": 0.0033,
+      "epoch": 4.431515151515152,
+      "grad_norm": 0.008006379939615726,
+      "learning_rate": 9.818541191460011e-05,
+      "loss": 0.0017,
       "step": 1371
     },
     {
-      "epoch": 3.515529939161063,
-      "grad_norm": 0.0182783342897892,
-      "learning_rate": 2.4459387309424865e-05,
-      "loss": 0.0039,
+      "epoch": 4.434747474747475,
+      "grad_norm": 0.008464152924716473,
+      "learning_rate": 9.817937511581155e-05,
+      "loss": 0.0014,
       "step": 1372
     },
     {
-      "epoch": 3.518091578610311,
-      "grad_norm": 0.020285986363887787,
-      "learning_rate": 2.4382482317232568e-05,
-      "loss": 0.0034,
+      "epoch": 4.437979797979798,
+      "grad_norm": 0.007001511752605438,
+      "learning_rate": 9.817332847824017e-05,
+      "loss": 0.0015,
       "step": 1373
     },
     {
-      "epoch": 3.520653218059558,
-      "grad_norm": 0.023470548912882805,
-      "learning_rate": 2.430565941356157e-05,
-      "loss": 0.0047,
+      "epoch": 4.441212121212121,
+      "grad_norm": 0.004775685258209705,
+      "learning_rate": 9.816727200312076e-05,
+      "loss": 0.001,
       "step": 1374
     },
     {
-      "epoch": 3.5232148575088056,
-      "grad_norm": 0.02091079019010067,
-      "learning_rate": 2.422891884458241e-05,
-      "loss": 0.0047,
+      "epoch": 4.444444444444445,
+      "grad_norm": 0.004762740805745125,
+      "learning_rate": 9.816120569169013e-05,
+      "loss": 0.0011,
       "step": 1375
     },
     {
-      "epoch": 3.5232148575088056,
-      "eval_loss": 0.0071318005211651325,
-      "eval_runtime": 15.7705,
-      "eval_samples_per_second": 6.341,
-      "eval_steps_per_second": 1.585,
+      "epoch": 4.444444444444445,
+      "eval_loss": 0.0018742467509582639,
+      "eval_runtime": 15.6028,
+      "eval_samples_per_second": 6.409,
+      "eval_steps_per_second": 1.602,
       "step": 1375
     },
     {
-      "epoch": 3.525776496958053,
-      "grad_norm": 0.028989002108573914,
-      "learning_rate": 2.4152260856201787e-05,
-      "loss": 0.0064,
+      "epoch": 4.447676767676768,
+      "grad_norm": 0.007123338058590889,
+      "learning_rate": 9.815512954518709e-05,
+      "loss": 0.0018,
       "step": 1376
     },
     {
-      "epoch": 3.5283381364073008,
-      "grad_norm": 0.023004917427897453,
-      "learning_rate": 2.4075685694061822e-05,
-      "loss": 0.0034,
+      "epoch": 4.450909090909091,
+      "grad_norm": 0.006350385956466198,
+      "learning_rate": 9.814904356485245e-05,
+      "loss": 0.0015,
       "step": 1377
     },
     {
-      "epoch": 3.530899775856548,
-      "grad_norm": 0.02438783086836338,
-      "learning_rate": 2.399919360353923e-05,
-      "loss": 0.005,
+      "epoch": 4.454141414141414,
+      "grad_norm": 0.004663221072405577,
+      "learning_rate": 9.814294775192904e-05,
+      "loss": 0.0012,
       "step": 1378
     },
     {
-      "epoch": 3.533461415305796,
-      "grad_norm": 0.016736894845962524,
-      "learning_rate": 2.39227848297445e-05,
-      "loss": 0.004,
+      "epoch": 4.457373737373738,
+      "grad_norm": 0.0077123455703258514,
+      "learning_rate": 9.81368421076617e-05,
+      "loss": 0.0015,
       "step": 1379
     },
     {
-      "epoch": 3.5360230547550433,
-      "grad_norm": 0.019711444154381752,
-      "learning_rate": 2.3846459617521128e-05,
-      "loss": 0.0048,
+      "epoch": 4.460606060606061,
+      "grad_norm": 0.00417733658105135,
+      "learning_rate": 9.813072663329728e-05,
+      "loss": 0.0012,
       "step": 1380
     },
     {
-      "epoch": 3.5385846942042907,
-      "grad_norm": 0.020827678963541985,
-      "learning_rate": 2.3770218211444907e-05,
-      "loss": 0.0044,
+      "epoch": 4.463838383838384,
+      "grad_norm": 0.005968569312244654,
+      "learning_rate": 9.812460133008463e-05,
+      "loss": 0.0017,
       "step": 1381
     },
     {
-      "epoch": 3.541146333653538,
-      "grad_norm": 0.02978433296084404,
-      "learning_rate": 2.369406085582302e-05,
-      "loss": 0.0061,
+      "epoch": 4.467070707070707,
+      "grad_norm": 0.01048695482313633,
+      "learning_rate": 9.811846619927459e-05,
+      "loss": 0.0023,
       "step": 1382
     },
     {
-      "epoch": 3.543707973102786,
-      "grad_norm": 0.022353358566761017,
-      "learning_rate": 2.361798779469336e-05,
-      "loss": 0.0037,
+      "epoch": 4.470303030303031,
+      "grad_norm": 0.00757136195898056,
+      "learning_rate": 9.811232124212002e-05,
+      "loss": 0.0018,
       "step": 1383
     },
     {
-      "epoch": 3.546269612552033,
-      "grad_norm": 0.017815707251429558,
-      "learning_rate": 2.354199927182371e-05,
-      "loss": 0.0035,
+      "epoch": 4.473535353535354,
+      "grad_norm": 0.010669861920177937,
+      "learning_rate": 9.810616645987582e-05,
+      "loss": 0.0019,
       "step": 1384
     },
     {
-      "epoch": 3.548831252001281,
-      "grad_norm": 0.03284938260912895,
-      "learning_rate": 2.346609553071093e-05,
-      "loss": 0.0053,
+      "epoch": 4.476767676767677,
+      "grad_norm": 0.006885967683047056,
+      "learning_rate": 9.810000185379884e-05,
+      "loss": 0.0018,
       "step": 1385
     },
     {
-      "epoch": 3.5513928914505284,
-      "grad_norm": 0.016769099980592728,
-      "learning_rate": 2.3390276814580204e-05,
-      "loss": 0.0039,
+      "epoch": 4.48,
+      "grad_norm": 0.004515382461249828,
+      "learning_rate": 9.8093827425148e-05,
+      "loss": 0.0009,
       "step": 1386
     },
     {
-      "epoch": 3.5539545308997758,
-      "grad_norm": 0.01981167122721672,
-      "learning_rate": 2.331454336638431e-05,
-      "loss": 0.0046,
+      "epoch": 4.483232323232324,
+      "grad_norm": 0.006351994350552559,
+      "learning_rate": 9.808764317518413e-05,
+      "loss": 0.0012,
       "step": 1387
     },
     {
-      "epoch": 3.556516170349023,
-      "grad_norm": 0.016274696215987206,
-      "learning_rate": 2.323889542880273e-05,
-      "loss": 0.0032,
+      "epoch": 4.486464646464647,
+      "grad_norm": 0.009356615133583546,
+      "learning_rate": 9.808144910517019e-05,
+      "loss": 0.0018,
       "step": 1388
     },
     {
-      "epoch": 3.559077809798271,
-      "grad_norm": 0.02920779027044773,
-      "learning_rate": 2.3163333244240987e-05,
-      "loss": 0.0057,
+      "epoch": 4.489696969696969,
+      "grad_norm": 0.02052316442131996,
+      "learning_rate": 9.807524521637102e-05,
+      "loss": 0.0014,
       "step": 1389
     },
     {
-      "epoch": 3.5616394492475183,
-      "grad_norm": 0.016290957108139992,
-      "learning_rate": 2.308785705482982e-05,
-      "loss": 0.003,
+      "epoch": 4.4929292929292926,
+      "grad_norm": 0.009456136263906956,
+      "learning_rate": 9.806903151005357e-05,
+      "loss": 0.0019,
       "step": 1390
     },
     {
-      "epoch": 3.564201088696766,
-      "grad_norm": 0.014574081636965275,
-      "learning_rate": 2.3012467102424373e-05,
-      "loss": 0.0033,
+      "epoch": 4.496161616161616,
+      "grad_norm": 0.012844165787100792,
+      "learning_rate": 9.806280798748674e-05,
+      "loss": 0.0025,
       "step": 1391
     },
     {
-      "epoch": 3.5667627281460135,
-      "grad_norm": 0.020299769937992096,
-      "learning_rate": 2.2937163628603435e-05,
-      "loss": 0.0035,
+      "epoch": 4.499393939393939,
+      "grad_norm": 0.007946658879518509,
+      "learning_rate": 9.805657464994144e-05,
+      "loss": 0.0013,
       "step": 1392
     },
     {
-      "epoch": 3.569324367595261,
-      "grad_norm": 0.029202919453382492,
-      "learning_rate": 2.286194687466875e-05,
-      "loss": 0.0033,
+      "epoch": 4.502626262626262,
+      "grad_norm": 0.012202220037579536,
+      "learning_rate": 9.80503314986906e-05,
+      "loss": 0.0024,
       "step": 1393
     },
     {
-      "epoch": 3.5718860070445086,
-      "grad_norm": 0.024974342435598373,
-      "learning_rate": 2.2786817081644147e-05,
-      "loss": 0.0042,
+      "epoch": 4.505858585858586,
+      "grad_norm": 0.006191349122673273,
+      "learning_rate": 9.804407853500912e-05,
+      "loss": 0.0014,
       "step": 1394
     },
     {
-      "epoch": 3.574447646493756,
-      "grad_norm": 0.028322022408246994,
-      "learning_rate": 2.2711774490274766e-05,
-      "loss": 0.0048,
+      "epoch": 4.509090909090909,
+      "grad_norm": 0.005740044638514519,
+      "learning_rate": 9.803781576017394e-05,
+      "loss": 0.0011,
       "step": 1395
     },
     {
-      "epoch": 3.5770092859430034,
-      "grad_norm": 0.01611924171447754,
-      "learning_rate": 2.263681934102639e-05,
-      "loss": 0.0031,
+      "epoch": 4.512323232323232,
+      "grad_norm": 0.011826829053461552,
+      "learning_rate": 9.8031543175464e-05,
+      "loss": 0.002,
       "step": 1396
     },
     {
-      "epoch": 3.579570925392251,
-      "grad_norm": 0.026567207649350166,
-      "learning_rate": 2.2561951874084524e-05,
-      "loss": 0.0046,
+      "epoch": 4.515555555555555,
+      "grad_norm": 0.0054328362457454205,
+      "learning_rate": 9.802526078216021e-05,
+      "loss": 0.0012,
       "step": 1397
     },
     {
-      "epoch": 3.5821325648414986,
-      "grad_norm": 0.017477042973041534,
-      "learning_rate": 2.2487172329353746e-05,
-      "loss": 0.0029,
+      "epoch": 4.518787878787879,
+      "grad_norm": 0.015383337624371052,
+      "learning_rate": 9.801896858154553e-05,
+      "loss": 0.0014,
       "step": 1398
     },
     {
-      "epoch": 3.584694204290746,
-      "grad_norm": 0.03424292802810669,
-      "learning_rate": 2.2412480946456893e-05,
-      "loss": 0.0057,
+      "epoch": 4.522020202020202,
+      "grad_norm": 0.0068153864704072475,
+      "learning_rate": 9.80126665749049e-05,
+      "loss": 0.0014,
       "step": 1399
     },
     {
-      "epoch": 3.5872558437399937,
-      "grad_norm": 0.02512899786233902,
-      "learning_rate": 2.233787796473432e-05,
-      "loss": 0.0039,
+      "epoch": 4.525252525252525,
+      "grad_norm": 0.010167761705815792,
+      "learning_rate": 9.800635476352525e-05,
+      "loss": 0.0021,
       "step": 1400
     },
     {
-      "epoch": 3.5872558437399937,
-      "eval_loss": 0.00698334164917469,
-      "eval_runtime": 15.7713,
-      "eval_samples_per_second": 6.341,
-      "eval_steps_per_second": 1.585,
+      "epoch": 4.525252525252525,
+      "eval_loss": 0.001835559611208737,
+      "eval_runtime": 15.611,
+      "eval_samples_per_second": 6.406,
+      "eval_steps_per_second": 1.601,
       "step": 1400
     },
     {
-      "epoch": 3.589817483189241,
-      "grad_norm": 0.02017313428223133,
-      "learning_rate": 2.2263363623243054e-05,
-      "loss": 0.0043,
+      "epoch": 4.528484848484848,
+      "grad_norm": 0.009218861348927021,
+      "learning_rate": 9.800003314869552e-05,
+      "loss": 0.0018,
       "step": 1401
     },
     {
-      "epoch": 3.592379122638489,
-      "grad_norm": 0.037246350198984146,
-      "learning_rate": 2.218893816075615e-05,
-      "loss": 0.0069,
+      "epoch": 4.531717171717172,
+      "grad_norm": 0.007689634338021278,
+      "learning_rate": 9.799370173170667e-05,
+      "loss": 0.0017,
       "step": 1402
     },
     {
-      "epoch": 3.5949407620877363,
-      "grad_norm": 0.015876617282629013,
-      "learning_rate": 2.2114601815761808e-05,
-      "loss": 0.0035,
+      "epoch": 4.534949494949495,
+      "grad_norm": 0.006554129999130964,
+      "learning_rate": 9.798736051385165e-05,
+      "loss": 0.0014,
       "step": 1403
     },
     {
-      "epoch": 3.5975024015369836,
-      "grad_norm": 0.01643594354391098,
-      "learning_rate": 2.2040354826462668e-05,
-      "loss": 0.0039,
+      "epoch": 4.538181818181818,
+      "grad_norm": 0.0066703082993626595,
+      "learning_rate": 9.79810094964254e-05,
+      "loss": 0.0011,
       "step": 1404
     },
     {
-      "epoch": 3.600064040986231,
-      "grad_norm": 0.023170500993728638,
-      "learning_rate": 2.1966197430775053e-05,
-      "loss": 0.0045,
+      "epoch": 4.541414141414141,
+      "grad_norm": 0.010076484642922878,
+      "learning_rate": 9.797464868072488e-05,
+      "loss": 0.0023,
       "step": 1405
     },
     {
-      "epoch": 3.602625680435479,
-      "grad_norm": 0.03175827115774155,
-      "learning_rate": 2.1892129866328232e-05,
-      "loss": 0.0063,
+      "epoch": 4.544646464646465,
+      "grad_norm": 0.005947070196270943,
+      "learning_rate": 9.796827806804902e-05,
+      "loss": 0.0015,
       "step": 1406
     },
     {
-      "epoch": 3.605187319884726,
-      "grad_norm": 0.011946959421038628,
-      "learning_rate": 2.1818152370463523e-05,
-      "loss": 0.0025,
+      "epoch": 4.547878787878788,
+      "grad_norm": 0.007635572925209999,
+      "learning_rate": 9.79618976596988e-05,
+      "loss": 0.0017,
       "step": 1407
     },
     {
-      "epoch": 3.607748959333974,
-      "grad_norm": 0.02114485576748848,
-      "learning_rate": 2.174426518023373e-05,
-      "loss": 0.0045,
+      "epoch": 4.551111111111111,
+      "grad_norm": 0.007745967246592045,
+      "learning_rate": 9.795550745697716e-05,
+      "loss": 0.0022,
       "step": 1408
     },
     {
-      "epoch": 3.6103105987832214,
-      "grad_norm": 0.017472948879003525,
-      "learning_rate": 2.1670468532402215e-05,
-      "loss": 0.0035,
+      "epoch": 4.554343434343434,
+      "grad_norm": 0.005735979415476322,
+      "learning_rate": 9.794910746118904e-05,
+      "loss": 0.0012,
       "step": 1409
     },
     {
-      "epoch": 3.6128722382324687,
-      "grad_norm": 0.02146715112030506,
-      "learning_rate": 2.1596762663442218e-05,
-      "loss": 0.004,
+      "epoch": 4.557575757575758,
+      "grad_norm": 0.006444281432777643,
+      "learning_rate": 9.794269767364143e-05,
+      "loss": 0.0016,
       "step": 1410
     },
     {
-      "epoch": 3.615433877681716,
-      "grad_norm": 0.037003446370363235,
-      "learning_rate": 2.152314780953611e-05,
-      "loss": 0.0025,
+      "epoch": 4.560808080808081,
+      "grad_norm": 0.00770741468295455,
+      "learning_rate": 9.793627809564324e-05,
+      "loss": 0.0017,
       "step": 1411
     },
     {
-      "epoch": 3.617995517130964,
-      "grad_norm": 0.02539096027612686,
-      "learning_rate": 2.144962420657461e-05,
-      "loss": 0.0043,
+      "epoch": 4.564040404040404,
+      "grad_norm": 0.010197889991104603,
+      "learning_rate": 9.792984872850546e-05,
+      "loss": 0.002,
       "step": 1412
     },
     {
-      "epoch": 3.6205571565802113,
-      "grad_norm": 0.020174555480480194,
-      "learning_rate": 2.1376192090156044e-05,
-      "loss": 0.0042,
+      "epoch": 4.567272727272727,
+      "grad_norm": 0.0056621721014380455,
+      "learning_rate": 9.792340957354101e-05,
+      "loss": 0.0013,
       "step": 1413
     },
     {
-      "epoch": 3.623118796029459,
-      "grad_norm": 0.028370022773742676,
-      "learning_rate": 2.1302851695585535e-05,
-      "loss": 0.0049,
+      "epoch": 4.570505050505051,
+      "grad_norm": 0.007511416915804148,
+      "learning_rate": 9.791696063206484e-05,
+      "loss": 0.0015,
       "step": 1414
     },
     {
-      "epoch": 3.6256804354787064,
-      "grad_norm": 0.03755459561944008,
-      "learning_rate": 2.122960325787432e-05,
-      "loss": 0.0044,
+      "epoch": 4.573737373737374,
+      "grad_norm": 0.007925822399556637,
+      "learning_rate": 9.791050190539393e-05,
+      "loss": 0.0015,
       "step": 1415
     },
     {
-      "epoch": 3.628242074927954,
-      "grad_norm": 0.021341577172279358,
-      "learning_rate": 2.115644701173899e-05,
-      "loss": 0.004,
+      "epoch": 4.576969696969697,
+      "grad_norm": 0.007809912320226431,
+      "learning_rate": 9.790403339484718e-05,
+      "loss": 0.0012,
       "step": 1416
     },
     {
-      "epoch": 3.630803714377201,
-      "grad_norm": 0.01897352933883667,
-      "learning_rate": 2.1083383191600674e-05,
-      "loss": 0.0041,
+      "epoch": 4.58020202020202,
+      "grad_norm": 0.004711519461125135,
+      "learning_rate": 9.789755510174559e-05,
+      "loss": 0.0008,
       "step": 1417
     },
     {
-      "epoch": 3.633365353826449,
-      "grad_norm": 0.013568758964538574,
-      "learning_rate": 2.1010412031584386e-05,
-      "loss": 0.0028,
+      "epoch": 4.583434343434344,
+      "grad_norm": 0.00534845981746912,
+      "learning_rate": 9.789106702741204e-05,
+      "loss": 0.0015,
       "step": 1418
     },
     {
-      "epoch": 3.6359269932756964,
-      "grad_norm": 0.01961873658001423,
-      "learning_rate": 2.0937533765518187e-05,
-      "loss": 0.0037,
+      "epoch": 4.586666666666667,
+      "grad_norm": 0.008164715953171253,
+      "learning_rate": 9.788456917317153e-05,
+      "loss": 0.002,
       "step": 1419
     },
     {
-      "epoch": 3.638488632724944,
-      "grad_norm": 0.04281854256987572,
-      "learning_rate": 2.086474862693244e-05,
-      "loss": 0.0062,
+      "epoch": 4.58989898989899,
+      "grad_norm": 0.00691712461411953,
+      "learning_rate": 9.787806154035095e-05,
+      "loss": 0.0017,
       "step": 1420
     },
     {
-      "epoch": 3.6410502721741915,
-      "grad_norm": 0.0375184640288353,
-      "learning_rate": 2.0792056849059155e-05,
-      "loss": 0.0059,
+      "epoch": 4.593131313131313,
+      "grad_norm": 0.006845057941973209,
+      "learning_rate": 9.787154413027926e-05,
+      "loss": 0.0018,
       "step": 1421
     },
     {
-      "epoch": 3.643611911623439,
-      "grad_norm": 0.027061646804213524,
-      "learning_rate": 2.0719458664831165e-05,
-      "loss": 0.0062,
+      "epoch": 4.596363636363637,
+      "grad_norm": 0.00543162040412426,
+      "learning_rate": 9.786501694428738e-05,
+      "loss": 0.0013,
       "step": 1422
     },
     {
-      "epoch": 3.6461735510726863,
-      "grad_norm": 0.01808282546699047,
-      "learning_rate": 2.064695430688134e-05,
-      "loss": 0.0033,
+      "epoch": 4.59959595959596,
+      "grad_norm": 0.0072128651663661,
+      "learning_rate": 9.785847998370823e-05,
+      "loss": 0.0011,
       "step": 1423
     },
     {
-      "epoch": 3.648735190521934,
-      "grad_norm": 0.023773295804858208,
-      "learning_rate": 2.0574544007541962e-05,
-      "loss": 0.0041,
+      "epoch": 4.602828282828283,
+      "grad_norm": 0.005033902823925018,
+      "learning_rate": 9.785193324987673e-05,
+      "loss": 0.0015,
       "step": 1424
     },
     {
-      "epoch": 3.6512968299711814,
-      "grad_norm": 0.022816844284534454,
-      "learning_rate": 2.050222799884387e-05,
-      "loss": 0.0049,
+      "epoch": 4.606060606060606,
+      "grad_norm": 0.010515904985368252,
+      "learning_rate": 9.784537674412984e-05,
+      "loss": 0.0025,
       "step": 1425
     },
     {
-      "epoch": 3.6512968299711814,
-      "eval_loss": 0.007150666788220406,
-      "eval_runtime": 15.8192,
-      "eval_samples_per_second": 6.321,
-      "eval_steps_per_second": 1.58,
+      "epoch": 4.606060606060606,
+      "eval_loss": 0.0018574610585346818,
+      "eval_runtime": 15.6005,
+      "eval_samples_per_second": 6.41,
+      "eval_steps_per_second": 1.603,
       "step": 1425
     },
     {
-      "epoch": 3.6538584694204292,
-      "grad_norm": 0.0231269933283329,
-      "learning_rate": 2.043000651251575e-05,
-      "loss": 0.0059,
-      "step": 1426
-    },
-    {
-      "epoch": 3.6564201088696766,
-      "grad_norm": 0.018816029652953148,
-      "learning_rate": 2.0357879779983435e-05,
-      "loss": 0.0039,
-      "step": 1427
-    },
-    {
-      "epoch": 3.658981748318924,
-      "grad_norm": 0.023829815909266472,
-      "learning_rate": 2.0285848032369137e-05,
-      "loss": 0.0074,
-      "step": 1428
-    },
-    {
-      "epoch": 3.661543387768172,
-      "grad_norm": 0.03465423360466957,
-      "learning_rate": 2.0213911500490647e-05,
-      "loss": 0.0054,
-      "step": 1429
-    },
-    {
-      "epoch": 3.664105027217419,
-      "grad_norm": 0.024734264239668846,
-      "learning_rate": 2.0142070414860704e-05,
-      "loss": 0.0045,
-      "step": 1430
+      "epoch": 4.606060606060606,
+      "step": 1425,
+      "total_flos": 2.884558791231996e+18,
+      "train_loss": 0.003744818231442192,
+      "train_runtime": 22029.3077,
+      "train_samples_per_second": 11.235,
+      "train_steps_per_second": 0.351
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 7725,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 25,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 5,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 4
+      }
     },
-    {
-      "epoch": 3.6666666666666665,
-      "grad_norm": 0.015112842433154583,
-      "learning_rate": 2.0070325005686147e-05,
-      "loss": 0.0024,
-      "step": 1431
-    },
-    {
-      "epoch": 3.6692283061159143,
-      "grad_norm": 0.017952393740415573,
-      "learning_rate": 1.9998675502867286e-05,
-      "loss": 0.0043,
-      "step": 1432
-    },
-    {
-      "epoch": 3.6717899455651617,
-      "grad_norm": 0.020078269764780998,
-      "learning_rate": 1.9927122135997063e-05,
-      "loss": 0.0041,
-      "step": 1433
-    },
-    {
-      "epoch": 3.674351585014409,
-      "grad_norm": 0.01992672309279442,
-      "learning_rate": 1.9855665134360386e-05,
-      "loss": 0.0043,
-      "step": 1434
-    },
-    {
-      "epoch": 3.676913224463657,
-      "grad_norm": 0.016725733876228333,
-      "learning_rate": 1.9784304726933383e-05,
-      "loss": 0.0035,
-      "step": 1435
-    },
-    {
-      "epoch": 3.6794748639129042,
-      "grad_norm": 0.021456018090248108,
-      "learning_rate": 1.971304114238263e-05,
-      "loss": 0.0038,
-      "step": 1436
-    },
-    {
-      "epoch": 3.682036503362152,
-      "grad_norm": 0.017350321635603905,
-      "learning_rate": 1.9641874609064443e-05,
-      "loss": 0.0043,
-      "step": 1437
-    },
-    {
-      "epoch": 3.6845981428113994,
-      "grad_norm": 0.02247788943350315,
-      "learning_rate": 1.957080535502418e-05,
-      "loss": 0.0065,
-      "step": 1438
-    },
-    {
-      "epoch": 3.687159782260647,
-      "grad_norm": 0.021594872698187828,
-      "learning_rate": 1.949983360799543e-05,
-      "loss": 0.0037,
-      "step": 1439
-    },
-    {
-      "epoch": 3.689721421709894,
-      "grad_norm": 0.020270483568310738,
-      "learning_rate": 1.942895959539939e-05,
-      "loss": 0.0034,
-      "step": 1440
-    },
-    {
-      "epoch": 3.692283061159142,
-      "grad_norm": 0.020229430869221687,
-      "learning_rate": 1.9358183544344044e-05,
-      "loss": 0.0046,
-      "step": 1441
-    },
-    {
-      "epoch": 3.6948447006083893,
-      "grad_norm": 0.017405830323696136,
-      "learning_rate": 1.928750568162347e-05,
-      "loss": 0.0033,
-      "step": 1442
-    },
-    {
-      "epoch": 3.697406340057637,
-      "grad_norm": 0.02286132425069809,
-      "learning_rate": 1.9216926233717085e-05,
-      "loss": 0.0053,
-      "step": 1443
-    },
-    {
-      "epoch": 3.6999679795068845,
-      "grad_norm": 0.02939784526824951,
-      "learning_rate": 1.914644542678903e-05,
-      "loss": 0.0074,
-      "step": 1444
-    },
-    {
-      "epoch": 3.702529618956132,
-      "grad_norm": 0.014939618296921253,
-      "learning_rate": 1.9076063486687256e-05,
-      "loss": 0.0027,
-      "step": 1445
-    },
-    {
-      "epoch": 3.7050912584053792,
-      "grad_norm": 0.024590235203504562,
-      "learning_rate": 1.9005780638942982e-05,
-      "loss": 0.0084,
-      "step": 1446
-    },
-    {
-      "epoch": 3.707652897854627,
-      "grad_norm": 0.017727019265294075,
-      "learning_rate": 1.8935597108769893e-05,
-      "loss": 0.0045,
-      "step": 1447
-    },
-    {
-      "epoch": 3.7102145373038744,
-      "grad_norm": 0.020679691806435585,
-      "learning_rate": 1.8865513121063383e-05,
-      "loss": 0.0042,
-      "step": 1448
-    },
-    {
-      "epoch": 3.712776176753122,
-      "grad_norm": 0.019846631214022636,
-      "learning_rate": 1.879552890039987e-05,
-      "loss": 0.0035,
-      "step": 1449
-    },
-    {
-      "epoch": 3.7153378162023696,
-      "grad_norm": 0.01740499958395958,
-      "learning_rate": 1.8725644671036126e-05,
-      "loss": 0.0045,
-      "step": 1450
-    },
-    {
-      "epoch": 3.7153378162023696,
-      "eval_loss": 0.0069673252291977406,
-      "eval_runtime": 15.759,
-      "eval_samples_per_second": 6.346,
-      "eval_steps_per_second": 1.586,
-      "step": 1450
-    },
-    {
-      "epoch": 3.717899455651617,
-      "grad_norm": 0.022572670131921768,
-      "learning_rate": 1.8655860656908497e-05,
-      "loss": 0.0071,
-      "step": 1451
-    },
-    {
-      "epoch": 3.7204610951008643,
-      "grad_norm": 0.01428269874304533,
-      "learning_rate": 1.8586177081632162e-05,
-      "loss": 0.0028,
-      "step": 1452
-    },
-    {
-      "epoch": 3.723022734550112,
-      "grad_norm": 0.027408719062805176,
-      "learning_rate": 1.8516594168500513e-05,
-      "loss": 0.0023,
-      "step": 1453
-    },
-    {
-      "epoch": 3.7255843739993595,
-      "grad_norm": 0.016997095197439194,
-      "learning_rate": 1.844711214048434e-05,
-      "loss": 0.0031,
-      "step": 1454
-    },
-    {
-      "epoch": 3.7281460134486073,
-      "grad_norm": 0.027528347447514534,
-      "learning_rate": 1.837773122023114e-05,
-      "loss": 0.0044,
-      "step": 1455
-    },
-    {
-      "epoch": 3.7307076528978547,
-      "grad_norm": 0.015830373391509056,
-      "learning_rate": 1.8308451630064484e-05,
-      "loss": 0.0035,
-      "step": 1456
-    },
-    {
-      "epoch": 3.733269292347102,
-      "grad_norm": 0.022276075556874275,
-      "learning_rate": 1.823927359198321e-05,
-      "loss": 0.0046,
-      "step": 1457
-    },
-    {
-      "epoch": 3.7358309317963494,
-      "grad_norm": 0.021877508610486984,
-      "learning_rate": 1.8170197327660733e-05,
-      "loss": 0.004,
-      "step": 1458
-    },
-    {
-      "epoch": 3.738392571245597,
-      "grad_norm": 0.017875684425234795,
-      "learning_rate": 1.8101223058444333e-05,
-      "loss": 0.0035,
-      "step": 1459
-    },
-    {
-      "epoch": 3.7409542106948446,
-      "grad_norm": 0.022446325048804283,
-      "learning_rate": 1.803235100535452e-05,
-      "loss": 0.0059,
-      "step": 1460
-    },
-    {
-      "epoch": 3.7435158501440924,
-      "grad_norm": 0.02225191332399845,
-      "learning_rate": 1.7963581389084178e-05,
-      "loss": 0.003,
-      "step": 1461
-    },
-    {
-      "epoch": 3.7460774895933397,
-      "grad_norm": 0.017582381144165993,
-      "learning_rate": 1.7894914429998012e-05,
-      "loss": 0.0029,
-      "step": 1462
-    },
-    {
-      "epoch": 3.748639129042587,
-      "grad_norm": 0.018602004274725914,
-      "learning_rate": 1.782635034813176e-05,
-      "loss": 0.0035,
-      "step": 1463
-    },
-    {
-      "epoch": 3.751200768491835,
-      "grad_norm": 0.017477497458457947,
-      "learning_rate": 1.7757889363191483e-05,
-      "loss": 0.0032,
-      "step": 1464
-    },
-    {
-      "epoch": 3.7537624079410823,
-      "grad_norm": 0.0241811852902174,
-      "learning_rate": 1.7689531694552863e-05,
-      "loss": 0.0049,
-      "step": 1465
-    },
-    {
-      "epoch": 3.7563240473903297,
-      "grad_norm": 0.012605165131390095,
-      "learning_rate": 1.762127756126056e-05,
-      "loss": 0.0024,
-      "step": 1466
-    },
-    {
-      "epoch": 3.7588856868395775,
-      "grad_norm": 0.028144102543592453,
-      "learning_rate": 1.7553127182027455e-05,
-      "loss": 0.0064,
-      "step": 1467
-    },
-    {
-      "epoch": 3.761447326288825,
-      "grad_norm": 0.01943473145365715,
-      "learning_rate": 1.748508077523393e-05,
-      "loss": 0.0042,
-      "step": 1468
-    },
-    {
-      "epoch": 3.764008965738072,
-      "grad_norm": 0.020696593448519707,
-      "learning_rate": 1.7417138558927244e-05,
-      "loss": 0.0042,
-      "step": 1469
-    },
-    {
-      "epoch": 3.76657060518732,
-      "grad_norm": 0.021954044699668884,
-      "learning_rate": 1.734930075082076e-05,
-      "loss": 0.0039,
-      "step": 1470
-    },
-    {
-      "epoch": 3.7691322446365674,
-      "grad_norm": 0.022788338363170624,
-      "learning_rate": 1.728156756829325e-05,
-      "loss": 0.0043,
-      "step": 1471
-    },
-    {
-      "epoch": 3.771693884085815,
-      "grad_norm": 0.019136719405651093,
-      "learning_rate": 1.7213939228388287e-05,
-      "loss": 0.0029,
-      "step": 1472
-    },
-    {
-      "epoch": 3.7742555235350626,
-      "grad_norm": 0.03452589362859726,
-      "learning_rate": 1.714641594781347e-05,
-      "loss": 0.0046,
-      "step": 1473
-    },
-    {
-      "epoch": 3.77681716298431,
-      "grad_norm": 0.02517046593129635,
-      "learning_rate": 1.7078997942939695e-05,
-      "loss": 0.0057,
-      "step": 1474
-    },
-    {
-      "epoch": 3.7793788024335573,
-      "grad_norm": 0.02047286182641983,
-      "learning_rate": 1.7011685429800595e-05,
-      "loss": 0.0043,
-      "step": 1475
-    },
-    {
-      "epoch": 3.7793788024335573,
-      "eval_loss": 0.006774227134883404,
-      "eval_runtime": 15.8345,
-      "eval_samples_per_second": 6.315,
-      "eval_steps_per_second": 1.579,
-      "step": 1475
-    },
-    {
-      "epoch": 3.781940441882805,
-      "grad_norm": 0.018452035263180733,
-      "learning_rate": 1.6944478624091704e-05,
-      "loss": 0.0036,
-      "step": 1476
-    },
-    {
-      "epoch": 3.7845020813320525,
-      "grad_norm": 0.020415108650922775,
-      "learning_rate": 1.6877377741169827e-05,
-      "loss": 0.0025,
-      "step": 1477
-    },
-    {
-      "epoch": 3.7870637207813003,
-      "grad_norm": 0.030411340296268463,
-      "learning_rate": 1.6810382996052394e-05,
-      "loss": 0.0052,
-      "step": 1478
-    },
-    {
-      "epoch": 3.7896253602305476,
-      "grad_norm": 0.016582481563091278,
-      "learning_rate": 1.6743494603416714e-05,
-      "loss": 0.004,
-      "step": 1479
-    },
-    {
-      "epoch": 3.792186999679795,
-      "grad_norm": 0.022396402433514595,
-      "learning_rate": 1.6676712777599273e-05,
-      "loss": 0.0051,
-      "step": 1480
-    },
-    {
-      "epoch": 3.7947486391290424,
-      "grad_norm": 0.03476893901824951,
-      "learning_rate": 1.6610037732595123e-05,
-      "loss": 0.0093,
-      "step": 1481
-    },
-    {
-      "epoch": 3.79731027857829,
-      "grad_norm": 0.019850311800837517,
-      "learning_rate": 1.6543469682057106e-05,
-      "loss": 0.0044,
-      "step": 1482
-    },
-    {
-      "epoch": 3.7998719180275375,
-      "grad_norm": 0.025984439998865128,
-      "learning_rate": 1.6477008839295203e-05,
-      "loss": 0.0043,
-      "step": 1483
-    },
-    {
-      "epoch": 3.8024335574767854,
-      "grad_norm": 0.01581578701734543,
-      "learning_rate": 1.6410655417275917e-05,
-      "loss": 0.0033,
-      "step": 1484
-    },
-    {
-      "epoch": 3.8049951969260327,
-      "grad_norm": 0.013997646979987621,
-      "learning_rate": 1.6344409628621484e-05,
-      "loss": 0.0037,
-      "step": 1485
-    },
-    {
-      "epoch": 3.80755683637528,
-      "grad_norm": 0.020644215866923332,
-      "learning_rate": 1.6278271685609274e-05,
-      "loss": 0.0052,
-      "step": 1486
-    },
-    {
-      "epoch": 3.8101184758245275,
-      "grad_norm": 0.023151680827140808,
-      "learning_rate": 1.6212241800171052e-05,
-      "loss": 0.004,
-      "step": 1487
-    },
-    {
-      "epoch": 3.8126801152737753,
-      "grad_norm": 0.020557168871164322,
-      "learning_rate": 1.614632018389231e-05,
-      "loss": 0.0039,
-      "step": 1488
-    },
-    {
-      "epoch": 3.8152417547230226,
-      "grad_norm": 0.018763143569231033,
-      "learning_rate": 1.6080507048011666e-05,
-      "loss": 0.0037,
-      "step": 1489
-    },
-    {
-      "epoch": 3.8178033941722704,
-      "grad_norm": 0.019956249743700027,
-      "learning_rate": 1.6014802603420044e-05,
-      "loss": 0.0036,
-      "step": 1490
-    },
-    {
-      "epoch": 3.820365033621518,
-      "grad_norm": 0.01779358461499214,
-      "learning_rate": 1.5949207060660138e-05,
-      "loss": 0.0032,
-      "step": 1491
-    },
-    {
-      "epoch": 3.822926673070765,
-      "grad_norm": 0.01736687310039997,
-      "learning_rate": 1.5883720629925687e-05,
-      "loss": 0.004,
-      "step": 1492
-    },
-    {
-      "epoch": 3.8254883125200125,
-      "grad_norm": 0.036861103028059006,
-      "learning_rate": 1.58183435210607e-05,
-      "loss": 0.0054,
-      "step": 1493
-    },
-    {
-      "epoch": 3.8280499519692603,
-      "grad_norm": 0.023185743018984795,
-      "learning_rate": 1.5753075943558987e-05,
-      "loss": 0.0057,
-      "step": 1494
-    },
-    {
-      "epoch": 3.8306115914185077,
-      "grad_norm": 0.02160869538784027,
-      "learning_rate": 1.5687918106563326e-05,
-      "loss": 0.0057,
-      "step": 1495
-    },
-    {
-      "epoch": 3.8331732308677555,
-      "grad_norm": 0.021889541298151016,
-      "learning_rate": 1.5622870218864833e-05,
-      "loss": 0.0039,
-      "step": 1496
-    },
-    {
-      "epoch": 3.835734870317003,
-      "grad_norm": 0.020682981237769127,
-      "learning_rate": 1.5557932488902338e-05,
-      "loss": 0.0065,
-      "step": 1497
-    },
-    {
-      "epoch": 3.8382965097662503,
-      "grad_norm": 0.018665673211216927,
-      "learning_rate": 1.5493105124761657e-05,
-      "loss": 0.004,
-      "step": 1498
-    },
-    {
-      "epoch": 3.840858149215498,
-      "grad_norm": 0.015613999217748642,
-      "learning_rate": 1.542838833417493e-05,
-      "loss": 0.0036,
-      "step": 1499
-    },
-    {
-      "epoch": 3.8434197886647454,
-      "grad_norm": 0.018136585131287575,
-      "learning_rate": 1.536378232452003e-05,
-      "loss": 0.0036,
-      "step": 1500
-    },
-    {
-      "epoch": 3.8434197886647454,
-      "eval_loss": 0.006774452514946461,
-      "eval_runtime": 15.7383,
-      "eval_samples_per_second": 6.354,
-      "eval_steps_per_second": 1.588,
-      "step": 1500
-    },
-    {
-      "epoch": 3.845981428113993,
-      "grad_norm": 0.026844680309295654,
-      "learning_rate": 1.529928730281983e-05,
-      "loss": 0.0031,
-      "step": 1501
-    },
-    {
-      "epoch": 3.8485430675632406,
-      "grad_norm": 0.023822171613574028,
-      "learning_rate": 1.5234903475741508e-05,
-      "loss": 0.005,
-      "step": 1502
-    },
-    {
-      "epoch": 3.851104707012488,
-      "grad_norm": 0.0172001700848341,
-      "learning_rate": 1.5170631049595996e-05,
-      "loss": 0.0027,
-      "step": 1503
-    },
-    {
-      "epoch": 3.8536663464617353,
-      "grad_norm": 0.017186421900987625,
-      "learning_rate": 1.5106470230337194e-05,
-      "loss": 0.0036,
-      "step": 1504
-    },
-    {
-      "epoch": 3.856227985910983,
-      "grad_norm": 0.02007218264043331,
-      "learning_rate": 1.5042421223561432e-05,
-      "loss": 0.0052,
-      "step": 1505
-    },
-    {
-      "epoch": 3.8587896253602305,
-      "grad_norm": 0.021068302914500237,
-      "learning_rate": 1.4978484234506684e-05,
-      "loss": 0.0037,
-      "step": 1506
-    },
-    {
-      "epoch": 3.8613512648094783,
-      "grad_norm": 0.026851925998926163,
-      "learning_rate": 1.4914659468052022e-05,
-      "loss": 0.0055,
-      "step": 1507
-    },
-    {
-      "epoch": 3.8639129042587257,
-      "grad_norm": 0.015978677198290825,
-      "learning_rate": 1.4850947128716913e-05,
-      "loss": 0.0032,
-      "step": 1508
-    },
-    {
-      "epoch": 3.866474543707973,
-      "grad_norm": 0.018305087462067604,
-      "learning_rate": 1.4787347420660541e-05,
-      "loss": 0.0036,
-      "step": 1509
-    },
-    {
-      "epoch": 3.8690361831572204,
-      "grad_norm": 0.018359962850809097,
-      "learning_rate": 1.4723860547681162e-05,
-      "loss": 0.0038,
-      "step": 1510
-    },
-    {
-      "epoch": 3.8715978226064682,
-      "grad_norm": 0.01923317462205887,
-      "learning_rate": 1.4660486713215516e-05,
-      "loss": 0.0035,
-      "step": 1511
-    },
-    {
-      "epoch": 3.8741594620557156,
-      "grad_norm": 0.0211176760494709,
-      "learning_rate": 1.459722612033807e-05,
-      "loss": 0.0038,
-      "step": 1512
-    },
-    {
-      "epoch": 3.8767211015049634,
-      "grad_norm": 0.036694832146167755,
-      "learning_rate": 1.4534078971760451e-05,
-      "loss": 0.0068,
-      "step": 1513
-    },
-    {
-      "epoch": 3.8792827409542108,
-      "grad_norm": 0.019058063626289368,
-      "learning_rate": 1.4471045469830786e-05,
-      "loss": 0.0038,
-      "step": 1514
-    },
-    {
-      "epoch": 3.881844380403458,
-      "grad_norm": 0.017758546397089958,
-      "learning_rate": 1.440812581653298e-05,
-      "loss": 0.0033,
-      "step": 1515
-    },
-    {
-      "epoch": 3.8844060198527055,
-      "grad_norm": 0.025525173172354698,
-      "learning_rate": 1.4345320213486146e-05,
-      "loss": 0.0043,
-      "step": 1516
-    },
-    {
-      "epoch": 3.8869676593019533,
-      "grad_norm": 0.016888508573174477,
-      "learning_rate": 1.4282628861943959e-05,
-      "loss": 0.0036,
-      "step": 1517
-    },
-    {
-      "epoch": 3.8895292987512007,
-      "grad_norm": 0.02030453458428383,
-      "learning_rate": 1.422005196279395e-05,
-      "loss": 0.0041,
-      "step": 1518
-    },
-    {
-      "epoch": 3.8920909382004485,
-      "grad_norm": 0.018481934443116188,
-      "learning_rate": 1.4157589716556923e-05,
-      "loss": 0.0032,
-      "step": 1519
-    },
-    {
-      "epoch": 3.894652577649696,
-      "grad_norm": 0.01700839400291443,
-      "learning_rate": 1.4095242323386303e-05,
-      "loss": 0.0031,
-      "step": 1520
-    },
-    {
-      "epoch": 3.8972142170989432,
-      "grad_norm": 0.017972050234675407,
-      "learning_rate": 1.4033009983067452e-05,
-      "loss": 0.0049,
-      "step": 1521
-    },
-    {
-      "epoch": 3.8997758565481906,
-      "grad_norm": 0.01810053363442421,
-      "learning_rate": 1.3970892895017052e-05,
-      "loss": 0.0033,
-      "step": 1522
-    },
-    {
-      "epoch": 3.9023374959974384,
-      "grad_norm": 0.020266929641366005,
-      "learning_rate": 1.390889125828249e-05,
-      "loss": 0.0036,
-      "step": 1523
-    },
-    {
-      "epoch": 3.9048991354466858,
-      "grad_norm": 0.02565910294651985,
-      "learning_rate": 1.3847005271541236e-05,
-      "loss": 0.0053,
-      "step": 1524
-    },
-    {
-      "epoch": 3.9074607748959336,
-      "grad_norm": 0.02397267334163189,
-      "learning_rate": 1.3785235133100088e-05,
-      "loss": 0.0056,
-      "step": 1525
-    },
-    {
-      "epoch": 3.9074607748959336,
-      "eval_loss": 0.006717777345329523,
-      "eval_runtime": 15.7612,
-      "eval_samples_per_second": 6.345,
-      "eval_steps_per_second": 1.586,
-      "step": 1525
-    },
-    {
-      "epoch": 3.910022414345181,
-      "grad_norm": 0.018318889662623405,
-      "learning_rate": 1.3723581040894712e-05,
-      "loss": 0.0032,
-      "step": 1526
-    },
-    {
-      "epoch": 3.9125840537944283,
-      "grad_norm": 0.04019472375512123,
-      "learning_rate": 1.3662043192488849e-05,
-      "loss": 0.0055,
-      "step": 1527
-    },
-    {
-      "epoch": 3.915145693243676,
-      "grad_norm": 0.01894393004477024,
-      "learning_rate": 1.360062178507377e-05,
-      "loss": 0.0033,
-      "step": 1528
-    },
-    {
-      "epoch": 3.9177073326929235,
-      "grad_norm": 0.02132059633731842,
-      "learning_rate": 1.3539317015467629e-05,
-      "loss": 0.0036,
-      "step": 1529
-    },
-    {
-      "epoch": 3.920268972142171,
-      "grad_norm": 0.016668565571308136,
-      "learning_rate": 1.3478129080114848e-05,
-      "loss": 0.0042,
-      "step": 1530
-    },
-    {
-      "epoch": 3.9228306115914187,
-      "grad_norm": 0.022684797644615173,
-      "learning_rate": 1.3417058175085435e-05,
-      "loss": 0.0051,
-      "step": 1531
-    },
-    {
-      "epoch": 3.925392251040666,
-      "grad_norm": 0.020761920139193535,
-      "learning_rate": 1.3356104496074368e-05,
-      "loss": 0.0033,
-      "step": 1532
-    },
-    {
-      "epoch": 3.9279538904899134,
-      "grad_norm": 0.01829049177467823,
-      "learning_rate": 1.3295268238401059e-05,
-      "loss": 0.0034,
-      "step": 1533
-    },
-    {
-      "epoch": 3.930515529939161,
-      "grad_norm": 0.028276601806282997,
-      "learning_rate": 1.3234549597008571e-05,
-      "loss": 0.0071,
-      "step": 1534
-    },
-    {
-      "epoch": 3.9330771693884086,
-      "grad_norm": 0.018241243436932564,
-      "learning_rate": 1.3173948766463145e-05,
-      "loss": 0.0041,
-      "step": 1535
-    },
-    {
-      "epoch": 3.9356388088376564,
-      "grad_norm": 0.022500475868582726,
-      "learning_rate": 1.3113465940953495e-05,
-      "loss": 0.0058,
-      "step": 1536
-    },
-    {
-      "epoch": 3.9382004482869037,
-      "grad_norm": 0.022990485653281212,
-      "learning_rate": 1.3053101314290177e-05,
-      "loss": 0.0034,
-      "step": 1537
-    },
-    {
-      "epoch": 3.940762087736151,
-      "grad_norm": 0.029217321425676346,
-      "learning_rate": 1.2992855079904998e-05,
-      "loss": 0.0074,
-      "step": 1538
-    },
-    {
-      "epoch": 3.9433237271853985,
-      "grad_norm": 0.016778865829110146,
-      "learning_rate": 1.29327274308504e-05,
-      "loss": 0.0028,
-      "step": 1539
-    },
-    {
-      "epoch": 3.9458853666346463,
-      "grad_norm": 0.02236914075911045,
-      "learning_rate": 1.2872718559798853e-05,
-      "loss": 0.0044,
-      "step": 1540
-    },
-    {
-      "epoch": 3.9484470060838937,
-      "grad_norm": 0.017543382942676544,
-      "learning_rate": 1.2812828659042159e-05,
-      "loss": 0.0035,
-      "step": 1541
-    },
-    {
-      "epoch": 3.9510086455331415,
-      "grad_norm": 0.025278670713305473,
-      "learning_rate": 1.2753057920490952e-05,
-      "loss": 0.0055,
-      "step": 1542
-    },
-    {
-      "epoch": 3.953570284982389,
-      "grad_norm": 0.015434582717716694,
-      "learning_rate": 1.2693406535673996e-05,
-      "loss": 0.0032,
-      "step": 1543
-    },
-    {
-      "epoch": 3.956131924431636,
-      "grad_norm": 0.03193724900484085,
-      "learning_rate": 1.2633874695737574e-05,
-      "loss": 0.0084,
-      "step": 1544
-    },
-    {
-      "epoch": 3.9586935638808836,
-      "grad_norm": 0.01635214500129223,
-      "learning_rate": 1.257446259144494e-05,
-      "loss": 0.0032,
-      "step": 1545
-    },
-    {
-      "epoch": 3.9612552033301314,
-      "grad_norm": 0.024026844650506973,
-      "learning_rate": 1.2515170413175687e-05,
-      "loss": 0.0045,
-      "step": 1546
-    },
-    {
-      "epoch": 3.9638168427793787,
-      "grad_norm": 0.015127202495932579,
-      "learning_rate": 1.245599835092504e-05,
-      "loss": 0.003,
-      "step": 1547
-    },
-    {
-      "epoch": 3.9663784822286265,
-      "grad_norm": 0.023889271542429924,
-      "learning_rate": 1.239694659430341e-05,
-      "loss": 0.0041,
-      "step": 1548
-    },
-    {
-      "epoch": 3.968940121677874,
-      "grad_norm": 0.017018994316458702,
-      "learning_rate": 1.2338015332535645e-05,
-      "loss": 0.0034,
-      "step": 1549
-    },
-    {
-      "epoch": 3.9715017611271213,
-      "grad_norm": 0.039207857102155685,
-      "learning_rate": 1.2279204754460493e-05,
-      "loss": 0.0049,
-      "step": 1550
-    },
-    {
-      "epoch": 3.9715017611271213,
-      "eval_loss": 0.006608300376683474,
-      "eval_runtime": 15.7851,
-      "eval_samples_per_second": 6.335,
-      "eval_steps_per_second": 1.584,
-      "step": 1550
-    },
-    {
-      "epoch": 3.9740634005763686,
-      "grad_norm": 0.014196470379829407,
-      "learning_rate": 1.2220515048529996e-05,
-      "loss": 0.0018,
-      "step": 1551
-    },
-    {
-      "epoch": 3.9766250400256165,
-      "grad_norm": 0.02193739451467991,
-      "learning_rate": 1.2161946402808888e-05,
-      "loss": 0.0046,
-      "step": 1552
-    },
-    {
-      "epoch": 3.979186679474864,
-      "grad_norm": 0.01906512677669525,
-      "learning_rate": 1.2103499004973934e-05,
-      "loss": 0.0036,
-      "step": 1553
-    },
-    {
-      "epoch": 3.9817483189241116,
-      "grad_norm": 0.01951034553349018,
-      "learning_rate": 1.204517304231343e-05,
-      "loss": 0.0049,
-      "step": 1554
-    },
-    {
-      "epoch": 3.984309958373359,
-      "grad_norm": 0.01712062582373619,
-      "learning_rate": 1.1986968701726491e-05,
-      "loss": 0.0031,
-      "step": 1555
-    },
-    {
-      "epoch": 3.9868715978226064,
-      "grad_norm": 0.02322932332754135,
-      "learning_rate": 1.1928886169722531e-05,
-      "loss": 0.0056,
-      "step": 1556
-    },
-    {
-      "epoch": 3.9894332372718537,
-      "grad_norm": 0.025899920612573624,
-      "learning_rate": 1.1870925632420648e-05,
-      "loss": 0.005,
-      "step": 1557
-    },
-    {
-      "epoch": 3.9919948767211015,
-      "grad_norm": 0.020725306123495102,
-      "learning_rate": 1.1813087275549023e-05,
-      "loss": 0.0039,
-      "step": 1558
-    },
-    {
-      "epoch": 3.994556516170349,
-      "grad_norm": 0.01383247971534729,
-      "learning_rate": 1.1755371284444328e-05,
-      "loss": 0.0025,
-      "step": 1559
-    },
-    {
-      "epoch": 3.9971181556195967,
-      "grad_norm": 0.020084239542484283,
-      "learning_rate": 1.1697777844051105e-05,
-      "loss": 0.0041,
-      "step": 1560
-    },
-    {
-      "epoch": 3.999679795068844,
-      "grad_norm": 0.041470300406217575,
-      "learning_rate": 1.1640307138921191e-05,
-      "loss": 0.0067,
-      "step": 1561
-    },
-    {
-      "epoch": 4.002561639449247,
-      "grad_norm": 0.021700385957956314,
-      "learning_rate": 1.1582959353213175e-05,
-      "loss": 0.0029,
-      "step": 1562
-    },
-    {
-      "epoch": 4.005123278898495,
-      "grad_norm": 0.009706800803542137,
-      "learning_rate": 1.1525734670691701e-05,
-      "loss": 0.0018,
-      "step": 1563
-    },
-    {
-      "epoch": 4.007684918347743,
-      "grad_norm": 0.01918914169073105,
-      "learning_rate": 1.146863327472701e-05,
-      "loss": 0.0022,
-      "step": 1564
-    },
-    {
-      "epoch": 4.01024655779699,
-      "grad_norm": 0.011788219213485718,
-      "learning_rate": 1.1411655348294247e-05,
-      "loss": 0.0021,
-      "step": 1565
-    },
-    {
-      "epoch": 4.012808197246238,
-      "grad_norm": 0.012416953220963478,
-      "learning_rate": 1.135480107397292e-05,
-      "loss": 0.0019,
-      "step": 1566
-    },
-    {
-      "epoch": 4.015369836695485,
-      "grad_norm": 0.0157302413135767,
-      "learning_rate": 1.1298070633946284e-05,
-      "loss": 0.0027,
-      "step": 1567
-    },
-    {
-      "epoch": 4.0179314761447324,
-      "grad_norm": 0.01494533196091652,
-      "learning_rate": 1.1241464210000829e-05,
-      "loss": 0.002,
-      "step": 1568
-    },
-    {
-      "epoch": 4.02049311559398,
-      "grad_norm": 0.020875081419944763,
-      "learning_rate": 1.11849819835256e-05,
-      "loss": 0.0033,
-      "step": 1569
-    },
-    {
-      "epoch": 4.023054755043228,
-      "grad_norm": 0.01567363180220127,
-      "learning_rate": 1.1128624135511712e-05,
-      "loss": 0.0026,
-      "step": 1570
-    },
-    {
-      "epoch": 4.025616394492475,
-      "grad_norm": 0.015228018164634705,
-      "learning_rate": 1.1072390846551679e-05,
-      "loss": 0.0029,
-      "step": 1571
-    },
-    {
-      "epoch": 4.028178033941723,
-      "grad_norm": 0.012901673093438148,
-      "learning_rate": 1.1016282296838887e-05,
-      "loss": 0.0024,
-      "step": 1572
-    },
-    {
-      "epoch": 4.03073967339097,
-      "grad_norm": 0.021614767611026764,
-      "learning_rate": 1.096029866616704e-05,
-      "loss": 0.0037,
-      "step": 1573
-    },
-    {
-      "epoch": 4.0333013128402175,
-      "grad_norm": 0.016034532338380814,
-      "learning_rate": 1.0904440133929539e-05,
-      "loss": 0.0026,
-      "step": 1574
-    },
-    {
-      "epoch": 4.035862952289465,
-      "grad_norm": 0.012547748163342476,
-      "learning_rate": 1.0848706879118892e-05,
-      "loss": 0.002,
-      "step": 1575
-    },
-    {
-      "epoch": 4.035862952289465,
-      "eval_loss": 0.006813474930822849,
-      "eval_runtime": 15.8112,
-      "eval_samples_per_second": 6.325,
-      "eval_steps_per_second": 1.581,
-      "step": 1575
-    },
-    {
-      "epoch": 4.038424591738713,
-      "grad_norm": 0.019763262942433357,
-      "learning_rate": 1.0793099080326214e-05,
-      "loss": 0.0028,
-      "step": 1576
-    },
-    {
-      "epoch": 4.0409862311879605,
-      "grad_norm": 0.009770766831934452,
-      "learning_rate": 1.0737616915740561e-05,
-      "loss": 0.0015,
-      "step": 1577
-    },
-    {
-      "epoch": 4.043547870637208,
-      "grad_norm": 0.028357144445180893,
-      "learning_rate": 1.0682260563148466e-05,
-      "loss": 0.0036,
-      "step": 1578
-    },
-    {
-      "epoch": 4.046109510086455,
-      "grad_norm": 0.018938632681965828,
-      "learning_rate": 1.0627030199933252e-05,
-      "loss": 0.0042,
-      "step": 1579
-    },
-    {
-      "epoch": 4.048671149535703,
-      "grad_norm": 0.016851453110575676,
-      "learning_rate": 1.0571926003074561e-05,
-      "loss": 0.0023,
-      "step": 1580
-    },
-    {
-      "epoch": 4.05123278898495,
-      "grad_norm": 0.022341042757034302,
-      "learning_rate": 1.0516948149147754e-05,
-      "loss": 0.0031,
-      "step": 1581
-    },
-    {
-      "epoch": 4.053794428434198,
-      "grad_norm": 0.01736963354051113,
-      "learning_rate": 1.0462096814323313e-05,
-      "loss": 0.0032,
-      "step": 1582
-    },
-    {
-      "epoch": 4.056356067883446,
-      "grad_norm": 0.01756434515118599,
-      "learning_rate": 1.0407372174366297e-05,
-      "loss": 0.0026,
-      "step": 1583
-    },
-    {
-      "epoch": 4.058917707332693,
-      "grad_norm": 0.014218315482139587,
-      "learning_rate": 1.0352774404635829e-05,
-      "loss": 0.0022,
-      "step": 1584
-    },
-    {
-      "epoch": 4.06147934678194,
-      "grad_norm": 0.012512991204857826,
-      "learning_rate": 1.0298303680084448e-05,
-      "loss": 0.0016,
-      "step": 1585
-    },
-    {
-      "epoch": 4.064040986231188,
-      "grad_norm": 0.02227507159113884,
-      "learning_rate": 1.0243960175257606e-05,
-      "loss": 0.0034,
-      "step": 1586
-    },
-    {
-      "epoch": 4.066602625680435,
-      "grad_norm": 0.013525206595659256,
-      "learning_rate": 1.0189744064293121e-05,
-      "loss": 0.0022,
-      "step": 1587
-    },
-    {
-      "epoch": 4.069164265129683,
-      "grad_norm": 0.01833437569439411,
-      "learning_rate": 1.013565552092054e-05,
-      "loss": 0.0028,
-      "step": 1588
-    },
-    {
-      "epoch": 4.071725904578931,
-      "grad_norm": 0.015551027841866016,
-      "learning_rate": 1.0081694718460655e-05,
-      "loss": 0.0021,
-      "step": 1589
-    },
-    {
-      "epoch": 4.074287544028178,
-      "grad_norm": 0.027464207261800766,
-      "learning_rate": 1.0027861829824952e-05,
-      "loss": 0.0039,
-      "step": 1590
-    },
-    {
-      "epoch": 4.076849183477425,
-      "grad_norm": 0.01209068764001131,
-      "learning_rate": 9.97415702751498e-06,
-      "loss": 0.0022,
-      "step": 1591
-    },
-    {
-      "epoch": 4.079410822926673,
-      "grad_norm": 0.02232500910758972,
-      "learning_rate": 9.9205804836219e-06,
-      "loss": 0.0036,
-      "step": 1592
-    },
-    {
-      "epoch": 4.08197246237592,
-      "grad_norm": 0.02386978268623352,
-      "learning_rate": 9.86713236982586e-06,
-      "loss": 0.0023,
-      "step": 1593
-    },
-    {
-      "epoch": 4.084534101825168,
-      "grad_norm": 0.02061832696199417,
-      "learning_rate": 9.81381285739547e-06,
-      "loss": 0.0026,
-      "step": 1594
-    },
-    {
-      "epoch": 4.087095741274416,
-      "grad_norm": 0.024451183155179024,
-      "learning_rate": 9.760622117187235e-06,
-      "loss": 0.003,
-      "step": 1595
-    },
-    {
-      "epoch": 4.089657380723663,
-      "grad_norm": 0.019360024482011795,
-      "learning_rate": 9.707560319645042e-06,
-      "loss": 0.0023,
-      "step": 1596
-    },
-    {
-      "epoch": 4.0922190201729105,
-      "grad_norm": 0.01975105330348015,
-      "learning_rate": 9.654627634799618e-06,
-      "loss": 0.0023,
-      "step": 1597
-    },
-    {
-      "epoch": 4.094780659622158,
-      "grad_norm": 0.01753416284918785,
-      "learning_rate": 9.601824232267909e-06,
-      "loss": 0.0025,
-      "step": 1598
-    },
-    {
-      "epoch": 4.097342299071406,
-      "grad_norm": 0.017939181998372078,
-      "learning_rate": 9.549150281252633e-06,
-      "loss": 0.0023,
-      "step": 1599
-    },
-    {
-      "epoch": 4.0999039385206535,
-      "grad_norm": 0.016956206411123276,
-      "learning_rate": 9.496605950541676e-06,
-      "loss": 0.0038,
-      "step": 1600
-    },
-    {
-      "epoch": 4.0999039385206535,
-      "eval_loss": 0.007028018124401569,
-      "eval_runtime": 15.8036,
-      "eval_samples_per_second": 6.328,
-      "eval_steps_per_second": 1.582,
-      "step": 1600
-    },
-    {
-      "epoch": 4.102465577969901,
-      "grad_norm": 0.02734898217022419,
-      "learning_rate": 9.444191408507559e-06,
-      "loss": 0.0029,
-      "step": 1601
-    },
-    {
-      "epoch": 4.105027217419148,
-      "grad_norm": 0.0182413998991251,
-      "learning_rate": 9.391906823106933e-06,
-      "loss": 0.0028,
-      "step": 1602
-    },
-    {
-      "epoch": 4.107588856868396,
-      "grad_norm": 0.020263666287064552,
-      "learning_rate": 9.339752361880028e-06,
-      "loss": 0.0029,
-      "step": 1603
-    },
-    {
-      "epoch": 4.110150496317643,
-      "grad_norm": 0.018666667863726616,
-      "learning_rate": 9.287728191950057e-06,
-      "loss": 0.0027,
-      "step": 1604
-    },
-    {
-      "epoch": 4.112712135766891,
-      "grad_norm": 0.01849486492574215,
-      "learning_rate": 9.235834480022787e-06,
-      "loss": 0.0023,
-      "step": 1605
-    },
-    {
-      "epoch": 4.115273775216139,
-      "grad_norm": 0.014099624007940292,
-      "learning_rate": 9.184071392385895e-06,
-      "loss": 0.0022,
-      "step": 1606
-    },
-    {
-      "epoch": 4.117835414665386,
-      "grad_norm": 0.019898248836398125,
-      "learning_rate": 9.132439094908501e-06,
-      "loss": 0.0017,
-      "step": 1607
-    },
-    {
-      "epoch": 4.120397054114633,
-      "grad_norm": 0.02428223192691803,
-      "learning_rate": 9.080937753040646e-06,
-      "loss": 0.003,
-      "step": 1608
-    },
-    {
-      "epoch": 4.122958693563881,
-      "grad_norm": 0.021522926166653633,
-      "learning_rate": 9.029567531812738e-06,
-      "loss": 0.0029,
-      "step": 1609
-    },
-    {
-      "epoch": 4.125520333013128,
-      "grad_norm": 0.019226526841521263,
-      "learning_rate": 8.978328595834984e-06,
-      "loss": 0.0023,
-      "step": 1610
-    },
-    {
-      "epoch": 4.128081972462376,
-      "grad_norm": 0.038743916898965836,
-      "learning_rate": 8.927221109296919e-06,
-      "loss": 0.0022,
-      "step": 1611
-    },
-    {
-      "epoch": 4.130643611911624,
-      "grad_norm": 0.018456993624567986,
-      "learning_rate": 8.876245235966885e-06,
-      "loss": 0.0018,
-      "step": 1612
-    },
-    {
-      "epoch": 4.133205251360871,
-      "grad_norm": 0.01764741726219654,
-      "learning_rate": 8.825401139191475e-06,
-      "loss": 0.0024,
-      "step": 1613
-    },
-    {
-      "epoch": 4.135766890810118,
-      "grad_norm": 0.015948278829455376,
-      "learning_rate": 8.774688981894997e-06,
-      "loss": 0.0028,
-      "step": 1614
-    },
-    {
-      "epoch": 4.138328530259366,
-      "grad_norm": 0.023764314129948616,
-      "learning_rate": 8.724108926578999e-06,
-      "loss": 0.0029,
-      "step": 1615
-    },
-    {
-      "epoch": 4.140890169708613,
-      "grad_norm": 0.01793825626373291,
-      "learning_rate": 8.673661135321709e-06,
-      "loss": 0.0026,
-      "step": 1616
-    },
-    {
-      "epoch": 4.143451809157861,
-      "grad_norm": 0.019253414124250412,
-      "learning_rate": 8.623345769777514e-06,
-      "loss": 0.0021,
-      "step": 1617
-    },
-    {
-      "epoch": 4.146013448607109,
-      "grad_norm": 0.02996036969125271,
-      "learning_rate": 8.573162991176498e-06,
-      "loss": 0.0041,
-      "step": 1618
-    },
-    {
-      "epoch": 4.148575088056356,
-      "grad_norm": 0.01609407179057598,
-      "learning_rate": 8.523112960323864e-06,
-      "loss": 0.0021,
-      "step": 1619
-    },
-    {
-      "epoch": 4.1511367275056035,
-      "grad_norm": 0.018650375306606293,
-      "learning_rate": 8.473195837599418e-06,
-      "loss": 0.0021,
-      "step": 1620
-    },
-    {
-      "epoch": 4.153698366954851,
-      "grad_norm": 0.020306214690208435,
-      "learning_rate": 8.423411782957119e-06,
-      "loss": 0.0031,
-      "step": 1621
-    },
-    {
-      "epoch": 4.156260006404098,
-      "grad_norm": 0.014326635748147964,
-      "learning_rate": 8.373760955924487e-06,
-      "loss": 0.0015,
-      "step": 1622
-    },
-    {
-      "epoch": 4.1588216458533465,
-      "grad_norm": 0.029303543269634247,
-      "learning_rate": 8.324243515602137e-06,
-      "loss": 0.0033,
-      "step": 1623
-    },
-    {
-      "epoch": 4.161383285302594,
-      "grad_norm": 0.029846735298633575,
-      "learning_rate": 8.274859620663277e-06,
-      "loss": 0.0033,
-      "step": 1624
-    },
-    {
-      "epoch": 4.163944924751841,
-      "grad_norm": 0.011236918158829212,
-      "learning_rate": 8.225609429353187e-06,
-      "loss": 0.0017,
-      "step": 1625
-    },
-    {
-      "epoch": 4.163944924751841,
-      "eval_loss": 0.007391571532934904,
-      "eval_runtime": 15.7924,
-      "eval_samples_per_second": 6.332,
-      "eval_steps_per_second": 1.583,
-      "step": 1625
-    },
-    {
-      "epoch": 4.166506564201089,
-      "grad_norm": 0.030453361570835114,
-      "learning_rate": 8.176493099488663e-06,
-      "loss": 0.0023,
-      "step": 1626
-    },
-    {
-      "epoch": 4.169068203650336,
-      "grad_norm": 0.020724887028336525,
-      "learning_rate": 8.12751078845762e-06,
-      "loss": 0.0026,
-      "step": 1627
-    },
-    {
-      "epoch": 4.171629843099584,
-      "grad_norm": 0.028344491496682167,
-      "learning_rate": 8.078662653218477e-06,
-      "loss": 0.0032,
-      "step": 1628
-    },
-    {
-      "epoch": 4.1741914825488315,
-      "grad_norm": 0.03142205998301506,
-      "learning_rate": 8.029948850299706e-06,
-      "loss": 0.0035,
-      "step": 1629
-    },
-    {
-      "epoch": 4.176753121998079,
-      "grad_norm": 0.020122842863202095,
-      "learning_rate": 7.981369535799354e-06,
-      "loss": 0.0028,
-      "step": 1630
-    },
-    {
-      "epoch": 4.179314761447326,
-      "grad_norm": 0.021725695580244064,
-      "learning_rate": 7.93292486538449e-06,
-      "loss": 0.0033,
-      "step": 1631
-    },
-    {
-      "epoch": 4.181876400896574,
-      "grad_norm": 0.01623694784939289,
-      "learning_rate": 7.884614994290756e-06,
-      "loss": 0.0022,
-      "step": 1632
-    },
-    {
-      "epoch": 4.184438040345821,
-      "grad_norm": 0.03401968255639076,
-      "learning_rate": 7.836440077321815e-06,
-      "loss": 0.0034,
-      "step": 1633
-    },
-    {
-      "epoch": 4.186999679795069,
-      "grad_norm": 0.017616091296076775,
-      "learning_rate": 7.788400268848884e-06,
-      "loss": 0.0021,
-      "step": 1634
-    },
-    {
-      "epoch": 4.189561319244317,
-      "grad_norm": 0.031969357281923294,
-      "learning_rate": 7.740495722810271e-06,
-      "loss": 0.0041,
-      "step": 1635
-    },
-    {
-      "epoch": 4.192122958693564,
-      "grad_norm": 0.020527997985482216,
-      "learning_rate": 7.692726592710808e-06,
-      "loss": 0.0025,
-      "step": 1636
-    },
-    {
-      "epoch": 4.194684598142811,
-      "grad_norm": 0.014912785962224007,
-      "learning_rate": 7.645093031621447e-06,
-      "loss": 0.0017,
-      "step": 1637
-    },
-    {
-      "epoch": 4.197246237592059,
-      "grad_norm": 0.023326560854911804,
-      "learning_rate": 7.597595192178702e-06,
-      "loss": 0.0034,
-      "step": 1638
-    },
-    {
-      "epoch": 4.199807877041306,
-      "grad_norm": 0.019423892721533775,
-      "learning_rate": 7.550233226584174e-06,
-      "loss": 0.0024,
-      "step": 1639
-    },
-    {
-      "epoch": 4.202369516490554,
-      "grad_norm": 0.01979582943022251,
-      "learning_rate": 7.503007286604069e-06,
-      "loss": 0.0027,
-      "step": 1640
-    },
-    {
-      "epoch": 4.204931155939802,
-      "grad_norm": 0.021676816046237946,
-      "learning_rate": 7.455917523568751e-06,
-      "loss": 0.0022,
-      "step": 1641
-    },
-    {
-      "epoch": 4.207492795389049,
-      "grad_norm": 0.025906957685947418,
-      "learning_rate": 7.408964088372161e-06,
-      "loss": 0.0028,
-      "step": 1642
-    },
-    {
-      "epoch": 4.2100544348382964,
-      "grad_norm": 0.02123456820845604,
-      "learning_rate": 7.362147131471437e-06,
-      "loss": 0.0038,
-      "step": 1643
-    },
-    {
-      "epoch": 4.212616074287544,
-      "grad_norm": 0.019668417051434517,
-      "learning_rate": 7.315466802886401e-06,
-      "loss": 0.0025,
-      "step": 1644
-    },
-    {
-      "epoch": 4.215177713736791,
-      "grad_norm": 0.022003300487995148,
-      "learning_rate": 7.268923252198989e-06,
-      "loss": 0.003,
-      "step": 1645
-    },
-    {
-      "epoch": 4.217739353186039,
-      "grad_norm": 0.019475610926747322,
-      "learning_rate": 7.222516628552922e-06,
-      "loss": 0.0023,
-      "step": 1646
-    },
-    {
-      "epoch": 4.220300992635287,
-      "grad_norm": 0.020722266286611557,
-      "learning_rate": 7.1762470806531426e-06,
-      "loss": 0.0027,
-      "step": 1647
-    },
-    {
-      "epoch": 4.222862632084534,
-      "grad_norm": 0.022313257679343224,
-      "learning_rate": 7.1301147567653e-06,
-      "loss": 0.0026,
-      "step": 1648
-    },
-    {
-      "epoch": 4.2254242715337815,
-      "grad_norm": 0.0169731006026268,
-      "learning_rate": 7.084119804715389e-06,
-      "loss": 0.003,
-      "step": 1649
-    },
-    {
-      "epoch": 4.227985910983029,
-      "grad_norm": 0.02195361629128456,
-      "learning_rate": 7.038262371889159e-06,
-      "loss": 0.0029,
-      "step": 1650
-    },
-    {
-      "epoch": 4.227985910983029,
-      "eval_loss": 0.007039368152618408,
-      "eval_runtime": 15.7807,
-      "eval_samples_per_second": 6.337,
-      "eval_steps_per_second": 1.584,
-      "step": 1650
-    },
-    {
-      "epoch": 4.230547550432276,
-      "grad_norm": 0.014441386796534061,
-      "learning_rate": 6.992542605231739e-06,
-      "loss": 0.0019,
-      "step": 1651
-    },
-    {
-      "epoch": 4.2331091898815245,
-      "grad_norm": 0.018818169832229614,
-      "learning_rate": 6.94696065124707e-06,
-      "loss": 0.0022,
-      "step": 1652
-    },
-    {
-      "epoch": 4.235670829330772,
-      "grad_norm": 0.021508438512682915,
-      "learning_rate": 6.901516655997536e-06,
-      "loss": 0.0027,
-      "step": 1653
-    },
-    {
-      "epoch": 4.238232468780019,
-      "grad_norm": 0.02906384877860546,
-      "learning_rate": 6.8562107651034325e-06,
-      "loss": 0.0025,
-      "step": 1654
-    },
-    {
-      "epoch": 4.240794108229267,
-      "grad_norm": 0.01837456226348877,
-      "learning_rate": 6.811043123742494e-06,
-      "loss": 0.0026,
-      "step": 1655
-    },
-    {
-      "epoch": 4.243355747678514,
-      "grad_norm": 0.04199704900383949,
-      "learning_rate": 6.7660138766494685e-06,
-      "loss": 0.0038,
-      "step": 1656
-    },
-    {
-      "epoch": 4.245917387127761,
-      "grad_norm": 0.020847132429480553,
-      "learning_rate": 6.721123168115645e-06,
-      "loss": 0.0021,
-      "step": 1657
-    },
-    {
-      "epoch": 4.24847902657701,
-      "grad_norm": 0.019165808334946632,
-      "learning_rate": 6.676371141988346e-06,
-      "loss": 0.0024,
-      "step": 1658
-    },
-    {
-      "epoch": 4.251040666026257,
-      "grad_norm": 0.02482159435749054,
-      "learning_rate": 6.631757941670535e-06,
-      "loss": 0.0039,
-      "step": 1659
-    },
-    {
-      "epoch": 4.253602305475504,
-      "grad_norm": 0.01521696150302887,
-      "learning_rate": 6.587283710120324e-06,
-      "loss": 0.002,
-      "step": 1660
-    },
-    {
-      "epoch": 4.256163944924752,
-      "grad_norm": 0.016666557639837265,
-      "learning_rate": 6.542948589850489e-06,
-      "loss": 0.0022,
-      "step": 1661
-    },
-    {
-      "epoch": 4.258725584373999,
-      "grad_norm": 0.017305411398410797,
-      "learning_rate": 6.498752722928042e-06,
-      "loss": 0.0023,
-      "step": 1662
-    },
-    {
-      "epoch": 4.261287223823247,
-      "grad_norm": 0.02088834159076214,
-      "learning_rate": 6.4546962509738075e-06,
-      "loss": 0.0031,
-      "step": 1663
-    },
-    {
-      "epoch": 4.263848863272495,
-      "grad_norm": 0.01768411323428154,
-      "learning_rate": 6.410779315161886e-06,
-      "loss": 0.0022,
-      "step": 1664
-    },
-    {
-      "epoch": 4.266410502721742,
-      "grad_norm": 0.020017629489302635,
-      "learning_rate": 6.367002056219284e-06,
-      "loss": 0.0033,
-      "step": 1665
-    },
-    {
-      "epoch": 4.268972142170989,
-      "grad_norm": 0.024552103132009506,
-      "learning_rate": 6.323364614425431e-06,
-      "loss": 0.0041,
-      "step": 1666
-    },
-    {
-      "epoch": 4.271533781620237,
-      "grad_norm": 0.01518949493765831,
-      "learning_rate": 6.2798671296117106e-06,
-      "loss": 0.0019,
-      "step": 1667
-    },
-    {
-      "epoch": 4.274095421069484,
-      "grad_norm": 0.01741175912320614,
-      "learning_rate": 6.236509741161012e-06,
-      "loss": 0.0025,
-      "step": 1668
-    },
-    {
-      "epoch": 4.276657060518732,
-      "grad_norm": 0.016028808429837227,
-      "learning_rate": 6.193292588007349e-06,
-      "loss": 0.0028,
-      "step": 1669
-    },
-    {
-      "epoch": 4.27921869996798,
-      "grad_norm": 0.02235426940023899,
-      "learning_rate": 6.150215808635335e-06,
-      "loss": 0.0026,
-      "step": 1670
-    },
-    {
-      "epoch": 4.281780339417227,
-      "grad_norm": 0.015044658444821835,
-      "learning_rate": 6.107279541079769e-06,
-      "loss": 0.0018,
-      "step": 1671
-    },
-    {
-      "epoch": 4.2843419788664745,
-      "grad_norm": 0.01653240993618965,
-      "learning_rate": 6.0644839229252234e-06,
-      "loss": 0.0015,
-      "step": 1672
-    },
-    {
-      "epoch": 4.286903618315722,
-      "grad_norm": 0.02017267420887947,
-      "learning_rate": 6.021829091305542e-06,
-      "loss": 0.0025,
-      "step": 1673
-    },
-    {
-      "epoch": 4.289465257764969,
-      "grad_norm": 0.020671404898166656,
-      "learning_rate": 5.979315182903433e-06,
-      "loss": 0.0027,
-      "step": 1674
-    },
-    {
-      "epoch": 4.2920268972142175,
-      "grad_norm": 0.026510581374168396,
-      "learning_rate": 5.936942333950063e-06,
-      "loss": 0.0031,
-      "step": 1675
-    },
-    {
-      "epoch": 4.2920268972142175,
-      "eval_loss": 0.007185637950897217,
-      "eval_runtime": 15.7785,
-      "eval_samples_per_second": 6.338,
-      "eval_steps_per_second": 1.584,
-      "step": 1675
-    },
-    {
-      "epoch": 4.294588536663465,
-      "grad_norm": 0.029353618621826172,
-      "learning_rate": 5.894710680224574e-06,
-      "loss": 0.0022,
-      "step": 1676
-    },
-    {
-      "epoch": 4.297150176112712,
-      "grad_norm": 0.01747061312198639,
-      "learning_rate": 5.852620357053651e-06,
-      "loss": 0.0027,
-      "step": 1677
-    },
-    {
-      "epoch": 4.29971181556196,
-      "grad_norm": 0.018502507358789444,
-      "learning_rate": 5.810671499311126e-06,
-      "loss": 0.0028,
-      "step": 1678
-    },
-    {
-      "epoch": 4.302273455011207,
-      "grad_norm": 0.019391067326068878,
-      "learning_rate": 5.768864241417493e-06,
-      "loss": 0.003,
-      "step": 1679
-    },
-    {
-      "epoch": 4.304835094460454,
-      "grad_norm": 0.020265663042664528,
-      "learning_rate": 5.727198717339511e-06,
-      "loss": 0.0037,
-      "step": 1680
-    },
-    {
-      "epoch": 4.307396733909703,
-      "grad_norm": 0.022747740149497986,
-      "learning_rate": 5.685675060589779e-06,
-      "loss": 0.0028,
-      "step": 1681
-    },
-    {
-      "epoch": 4.30995837335895,
-      "grad_norm": 0.0249753687530756,
-      "learning_rate": 5.644293404226314e-06,
-      "loss": 0.0026,
-      "step": 1682
-    },
-    {
-      "epoch": 4.312520012808197,
-      "grad_norm": 0.0229217316955328,
-      "learning_rate": 5.603053880852066e-06,
-      "loss": 0.0031,
-      "step": 1683
-    },
-    {
-      "epoch": 4.315081652257445,
-      "grad_norm": 0.026928609237074852,
-      "learning_rate": 5.561956622614556e-06,
-      "loss": 0.0023,
-      "step": 1684
-    },
-    {
-      "epoch": 4.317643291706692,
-      "grad_norm": 0.021758759394288063,
-      "learning_rate": 5.521001761205441e-06,
-      "loss": 0.0029,
-      "step": 1685
-    },
-    {
-      "epoch": 4.320204931155939,
-      "grad_norm": 0.018240762874484062,
-      "learning_rate": 5.480189427860089e-06,
-      "loss": 0.0024,
-      "step": 1686
-    },
-    {
-      "epoch": 4.322766570605188,
-      "grad_norm": 0.016551947221159935,
-      "learning_rate": 5.439519753357108e-06,
-      "loss": 0.0023,
-      "step": 1687
-    },
-    {
-      "epoch": 4.325328210054435,
-      "grad_norm": 0.019051391631364822,
-      "learning_rate": 5.398992868018033e-06,
-      "loss": 0.0027,
-      "step": 1688
-    },
-    {
-      "epoch": 4.327889849503682,
-      "grad_norm": 0.021947545930743217,
-      "learning_rate": 5.358608901706802e-06,
-      "loss": 0.0036,
-      "step": 1689
-    },
-    {
-      "epoch": 4.33045148895293,
-      "grad_norm": 0.019278889521956444,
-      "learning_rate": 5.318367983829392e-06,
-      "loss": 0.0023,
-      "step": 1690
-    },
-    {
-      "epoch": 4.333013128402177,
-      "grad_norm": 0.025414612144231796,
-      "learning_rate": 5.2782702433334105e-06,
-      "loss": 0.003,
-      "step": 1691
-    },
-    {
-      "epoch": 4.3355747678514245,
-      "grad_norm": 0.023709530010819435,
-      "learning_rate": 5.2383158087076664e-06,
-      "loss": 0.0026,
-      "step": 1692
-    },
-    {
-      "epoch": 4.338136407300673,
-      "grad_norm": 0.02490607649087906,
-      "learning_rate": 5.198504807981746e-06,
-      "loss": 0.0029,
-      "step": 1693
-    },
-    {
-      "epoch": 4.34069804674992,
-      "grad_norm": 0.015029078349471092,
-      "learning_rate": 5.158837368725649e-06,
-      "loss": 0.0019,
-      "step": 1694
-    },
-    {
-      "epoch": 4.3432596861991675,
-      "grad_norm": 0.03754875063896179,
-      "learning_rate": 5.1193136180493095e-06,
-      "loss": 0.0035,
-      "step": 1695
-    },
-    {
-      "epoch": 4.345821325648415,
-      "grad_norm": 0.014525432139635086,
-      "learning_rate": 5.079933682602234e-06,
-      "loss": 0.002,
-      "step": 1696
-    },
-    {
-      "epoch": 4.348382965097662,
-      "grad_norm": 0.01714390702545643,
-      "learning_rate": 5.040697688573115e-06,
-      "loss": 0.0026,
-      "step": 1697
-    },
-    {
-      "epoch": 4.3509446045469105,
-      "grad_norm": 0.014638411812484264,
-      "learning_rate": 5.001605761689398e-06,
-      "loss": 0.0018,
-      "step": 1698
-    },
-    {
-      "epoch": 4.353506243996158,
-      "grad_norm": 0.025071274489164352,
-      "learning_rate": 4.962658027216837e-06,
-      "loss": 0.0025,
-      "step": 1699
-    },
-    {
-      "epoch": 4.356067883445405,
-      "grad_norm": 0.01735590770840645,
-      "learning_rate": 4.9238546099592e-06,
-      "loss": 0.0022,
-      "step": 1700
-    },
-    {
-      "epoch": 4.356067883445405,
-      "eval_loss": 0.007100006565451622,
-      "eval_runtime": 15.7627,
-      "eval_samples_per_second": 6.344,
-      "eval_steps_per_second": 1.586,
-      "step": 1700
-    },
-    {
-      "epoch": 4.3586295228946526,
-      "grad_norm": 0.01827225089073181,
-      "learning_rate": 4.885195634257761e-06,
-      "loss": 0.0023,
-      "step": 1701
-    },
-    {
-      "epoch": 4.3611911623439,
-      "grad_norm": 0.023278500884771347,
-      "learning_rate": 4.846681223990951e-06,
-      "loss": 0.0021,
-      "step": 1702
-    },
-    {
-      "epoch": 4.363752801793147,
-      "grad_norm": 0.022773802280426025,
-      "learning_rate": 4.8083115025739756e-06,
-      "loss": 0.0025,
-      "step": 1703
-    },
-    {
-      "epoch": 4.3663144412423955,
-      "grad_norm": 0.025192998349666595,
-      "learning_rate": 4.770086592958389e-06,
-      "loss": 0.0031,
-      "step": 1704
-    },
-    {
-      "epoch": 4.368876080691643,
-      "grad_norm": 0.016672352328896523,
-      "learning_rate": 4.732006617631729e-06,
-      "loss": 0.0023,
-      "step": 1705
-    },
-    {
-      "epoch": 4.37143772014089,
-      "grad_norm": 0.017816077917814255,
-      "learning_rate": 4.6940716986170865e-06,
-      "loss": 0.0022,
-      "step": 1706
-    },
-    {
-      "epoch": 4.373999359590138,
-      "grad_norm": 0.01937803253531456,
-      "learning_rate": 4.65628195747273e-06,
-      "loss": 0.0027,
-      "step": 1707
-    },
-    {
-      "epoch": 4.376560999039385,
-      "grad_norm": 0.02456626296043396,
-      "learning_rate": 4.618637515291746e-06,
-      "loss": 0.0031,
-      "step": 1708
-    },
-    {
-      "epoch": 4.379122638488632,
-      "grad_norm": 0.01984013430774212,
-      "learning_rate": 4.581138492701609e-06,
-      "loss": 0.003,
-      "step": 1709
-    },
-    {
-      "epoch": 4.381684277937881,
-      "grad_norm": 0.03520813584327698,
-      "learning_rate": 4.54378500986381e-06,
-      "loss": 0.0025,
-      "step": 1710
-    },
-    {
-      "epoch": 4.384245917387128,
-      "grad_norm": 0.02249833196401596,
-      "learning_rate": 4.5065771864735094e-06,
-      "loss": 0.0031,
-      "step": 1711
-    },
-    {
-      "epoch": 4.386807556836375,
-      "grad_norm": 0.02302868850529194,
-      "learning_rate": 4.469515141759068e-06,
-      "loss": 0.0025,
-      "step": 1712
-    },
-    {
-      "epoch": 4.389369196285623,
-      "grad_norm": 0.02689419314265251,
-      "learning_rate": 4.432598994481729e-06,
-      "loss": 0.0031,
-      "step": 1713
-    },
-    {
-      "epoch": 4.39193083573487,
-      "grad_norm": 0.015539591200649738,
-      "learning_rate": 4.395828862935247e-06,
-      "loss": 0.0018,
-      "step": 1714
-    },
-    {
-      "epoch": 4.3944924751841175,
-      "grad_norm": 0.01954408548772335,
-      "learning_rate": 4.3592048649454594e-06,
-      "loss": 0.0024,
-      "step": 1715
-    },
-    {
-      "epoch": 4.397054114633366,
-      "grad_norm": 0.020380103960633278,
-      "learning_rate": 4.322727117869951e-06,
-      "loss": 0.0034,
-      "step": 1716
-    },
-    {
-      "epoch": 4.399615754082613,
-      "grad_norm": 0.015975268557667732,
-      "learning_rate": 4.286395738597659e-06,
-      "loss": 0.0019,
-      "step": 1717
-    },
-    {
-      "epoch": 4.40217739353186,
-      "grad_norm": 0.016705699265003204,
-      "learning_rate": 4.250210843548496e-06,
-      "loss": 0.0025,
-      "step": 1718
-    },
-    {
-      "epoch": 4.404739032981108,
-      "grad_norm": 0.02079274319112301,
-      "learning_rate": 4.214172548672968e-06,
-      "loss": 0.0024,
-      "step": 1719
-    },
-    {
-      "epoch": 4.407300672430355,
-      "grad_norm": 0.02368057705461979,
-      "learning_rate": 4.178280969451853e-06,
-      "loss": 0.0026,
-      "step": 1720
-    },
-    {
-      "epoch": 4.4098623118796025,
-      "grad_norm": 0.0249889325350523,
-      "learning_rate": 4.142536220895749e-06,
-      "loss": 0.0024,
-      "step": 1721
-    },
-    {
-      "epoch": 4.412423951328851,
-      "grad_norm": 0.024965455755591393,
-      "learning_rate": 4.1069384175448035e-06,
-      "loss": 0.0031,
-      "step": 1722
-    },
-    {
-      "epoch": 4.414985590778098,
-      "grad_norm": 0.01613733358681202,
-      "learning_rate": 4.0714876734682495e-06,
-      "loss": 0.0021,
-      "step": 1723
-    },
-    {
-      "epoch": 4.4175472302273455,
-      "grad_norm": 0.03153941407799721,
-      "learning_rate": 4.036184102264112e-06,
-      "loss": 0.0026,
-      "step": 1724
-    },
-    {
-      "epoch": 4.420108869676593,
-      "grad_norm": 0.01852312684059143,
-      "learning_rate": 4.001027817058789e-06,
-      "loss": 0.0024,
-      "step": 1725
-    },
-    {
-      "epoch": 4.420108869676593,
-      "eval_loss": 0.007159404922276735,
-      "eval_runtime": 15.833,
-      "eval_samples_per_second": 6.316,
-      "eval_steps_per_second": 1.579,
-      "step": 1725
-    },
-    {
-      "epoch": 4.42267050912584,
-      "grad_norm": 0.026458974927663803,
-      "learning_rate": 3.966018930506754e-06,
-      "loss": 0.0035,
-      "step": 1726
-    },
-    {
-      "epoch": 4.425232148575088,
-      "grad_norm": 0.01900397054851055,
-      "learning_rate": 3.931157554790138e-06,
-      "loss": 0.0027,
-      "step": 1727
-    },
-    {
-      "epoch": 4.427793788024336,
-      "grad_norm": 0.01978159509599209,
-      "learning_rate": 3.896443801618377e-06,
-      "loss": 0.0023,
-      "step": 1728
-    },
-    {
-      "epoch": 4.430355427473583,
-      "grad_norm": 0.017388010397553444,
-      "learning_rate": 3.861877782227885e-06,
-      "loss": 0.0015,
-      "step": 1729
-    },
-    {
-      "epoch": 4.432917066922831,
-      "grad_norm": 0.013536867685616016,
-      "learning_rate": 3.827459607381678e-06,
-      "loss": 0.0019,
-      "step": 1730
-    },
-    {
-      "epoch": 4.435478706372078,
-      "grad_norm": 0.024279840290546417,
-      "learning_rate": 3.793189387369006e-06,
-      "loss": 0.003,
-      "step": 1731
-    },
-    {
-      "epoch": 4.438040345821325,
-      "grad_norm": 0.022974256426095963,
-      "learning_rate": 3.7590672320050312e-06,
-      "loss": 0.0033,
-      "step": 1732
-    },
-    {
-      "epoch": 4.440601985270574,
-      "grad_norm": 0.02040516957640648,
-      "learning_rate": 3.7250932506304627e-06,
-      "loss": 0.0032,
-      "step": 1733
-    },
-    {
-      "epoch": 4.443163624719821,
-      "grad_norm": 0.0270087830722332,
-      "learning_rate": 3.691267552111183e-06,
-      "loss": 0.0029,
-      "step": 1734
-    },
-    {
-      "epoch": 4.445725264169068,
-      "grad_norm": 0.01492816861718893,
-      "learning_rate": 3.6575902448379105e-06,
-      "loss": 0.0019,
-      "step": 1735
-    },
-    {
-      "epoch": 4.448286903618316,
-      "grad_norm": 0.020600121468305588,
-      "learning_rate": 3.6240614367259064e-06,
-      "loss": 0.0025,
-      "step": 1736
-    },
-    {
-      "epoch": 4.450848543067563,
-      "grad_norm": 0.01913149654865265,
-      "learning_rate": 3.5906812352145268e-06,
-      "loss": 0.0025,
-      "step": 1737
-    },
-    {
-      "epoch": 4.45341018251681,
-      "grad_norm": 0.021030202507972717,
-      "learning_rate": 3.5574497472669722e-06,
-      "loss": 0.0025,
-      "step": 1738
-    },
-    {
-      "epoch": 4.455971821966059,
-      "grad_norm": 0.016762861981987953,
-      "learning_rate": 3.524367079369906e-06,
-      "loss": 0.0021,
-      "step": 1739
-    },
-    {
-      "epoch": 4.458533461415306,
-      "grad_norm": 0.02296973019838333,
-      "learning_rate": 3.4914333375330898e-06,
-      "loss": 0.0032,
-      "step": 1740
-    },
-    {
-      "epoch": 4.461095100864553,
-      "grad_norm": 0.02116263285279274,
-      "learning_rate": 3.4586486272890793e-06,
-      "loss": 0.0023,
-      "step": 1741
-    },
-    {
-      "epoch": 4.463656740313801,
-      "grad_norm": 0.01815374568104744,
-      "learning_rate": 3.426013053692878e-06,
-      "loss": 0.0026,
-      "step": 1742
-    },
-    {
-      "epoch": 4.466218379763048,
-      "grad_norm": 0.018989766016602516,
-      "learning_rate": 3.393526721321616e-06,
-      "loss": 0.0024,
-      "step": 1743
-    },
-    {
-      "epoch": 4.4687800192122955,
-      "grad_norm": 0.023553112521767616,
-      "learning_rate": 3.361189734274145e-06,
-      "loss": 0.0028,
-      "step": 1744
-    },
-    {
-      "epoch": 4.471341658661544,
-      "grad_norm": 0.017866475507616997,
-      "learning_rate": 3.329002196170816e-06,
-      "loss": 0.0026,
-      "step": 1745
-    },
-    {
-      "epoch": 4.473903298110791,
-      "grad_norm": 0.016270101070404053,
-      "learning_rate": 3.296964210153042e-06,
-      "loss": 0.0021,
-      "step": 1746
-    },
-    {
-      "epoch": 4.4764649375600385,
-      "grad_norm": 0.02642947994172573,
-      "learning_rate": 3.265075878883034e-06,
-      "loss": 0.0029,
-      "step": 1747
-    },
-    {
-      "epoch": 4.479026577009286,
-      "grad_norm": 0.015611767768859863,
-      "learning_rate": 3.2333373045434435e-06,
-      "loss": 0.0017,
-      "step": 1748
-    },
-    {
-      "epoch": 4.481588216458533,
-      "grad_norm": 0.017911631613969803,
-      "learning_rate": 3.2017485888370556e-06,
-      "loss": 0.0018,
-      "step": 1749
-    },
-    {
-      "epoch": 4.484149855907781,
-      "grad_norm": 0.021472591906785965,
-      "learning_rate": 3.1703098329864233e-06,
-      "loss": 0.0025,
-      "step": 1750
-    },
-    {
-      "epoch": 4.484149855907781,
-      "eval_loss": 0.0070798140950500965,
-      "eval_runtime": 15.7859,
-      "eval_samples_per_second": 6.335,
-      "eval_steps_per_second": 1.584,
-      "step": 1750
-    },
-    {
-      "epoch": 4.486711495357029,
-      "grad_norm": 0.01979091763496399,
-      "learning_rate": 3.1390211377336053e-06,
-      "loss": 0.0023,
-      "step": 1751
-    },
-    {
-      "epoch": 4.489273134806276,
-      "grad_norm": 0.024668654426932335,
-      "learning_rate": 3.1078826033397843e-06,
-      "loss": 0.0027,
-      "step": 1752
-    },
-    {
-      "epoch": 4.491834774255524,
-      "grad_norm": 0.029574980959296227,
-      "learning_rate": 3.076894329584956e-06,
-      "loss": 0.0024,
-      "step": 1753
-    },
-    {
-      "epoch": 4.494396413704771,
-      "grad_norm": 0.022667400538921356,
-      "learning_rate": 3.0460564157676506e-06,
-      "loss": 0.0029,
-      "step": 1754
-    },
-    {
-      "epoch": 4.496958053154018,
-      "grad_norm": 0.02057446911931038,
-      "learning_rate": 3.0153689607045845e-06,
-      "loss": 0.0018,
-      "step": 1755
-    },
-    {
-      "epoch": 4.499519692603266,
-      "grad_norm": 0.019535532221198082,
-      "learning_rate": 2.9848320627303193e-06,
-      "loss": 0.0028,
-      "step": 1756
-    },
-    {
-      "epoch": 4.502081332052514,
-      "grad_norm": 0.025299862027168274,
-      "learning_rate": 2.95444581969701e-06,
-      "loss": 0.0026,
-      "step": 1757
-    },
-    {
-      "epoch": 4.504642971501761,
-      "grad_norm": 0.025970950722694397,
-      "learning_rate": 2.9242103289740198e-06,
-      "loss": 0.0026,
-      "step": 1758
-    },
-    {
-      "epoch": 4.507204610951009,
-      "grad_norm": 0.015390222892165184,
-      "learning_rate": 2.894125687447652e-06,
-      "loss": 0.002,
-      "step": 1759
-    },
-    {
-      "epoch": 4.509766250400256,
-      "grad_norm": 0.03658923879265785,
-      "learning_rate": 2.864191991520848e-06,
-      "loss": 0.0022,
-      "step": 1760
-    },
-    {
-      "epoch": 4.512327889849503,
-      "grad_norm": 0.017071478068828583,
-      "learning_rate": 2.8344093371128424e-06,
-      "loss": 0.0028,
-      "step": 1761
-    },
-    {
-      "epoch": 4.514889529298751,
-      "grad_norm": 0.018991444259881973,
-      "learning_rate": 2.8047778196588937e-06,
-      "loss": 0.0024,
-      "step": 1762
-    },
-    {
-      "epoch": 4.517451168747999,
-      "grad_norm": 0.016508381813764572,
-      "learning_rate": 2.775297534109922e-06,
-      "loss": 0.0017,
-      "step": 1763
-    },
-    {
-      "epoch": 4.520012808197246,
-      "grad_norm": 0.015772221609950066,
-      "learning_rate": 2.7459685749322726e-06,
-      "loss": 0.0021,
-      "step": 1764
-    },
-    {
-      "epoch": 4.522574447646494,
-      "grad_norm": 0.019358433783054352,
-      "learning_rate": 2.71679103610738e-06,
-      "loss": 0.0034,
-      "step": 1765
-    },
-    {
-      "epoch": 4.525136087095741,
-      "grad_norm": 0.01851905882358551,
-      "learning_rate": 2.687765011131449e-06,
-      "loss": 0.0021,
-      "step": 1766
-    },
-    {
-      "epoch": 4.5276977265449885,
-      "grad_norm": 0.01947377622127533,
-      "learning_rate": 2.6588905930152065e-06,
-      "loss": 0.0025,
-      "step": 1767
-    },
-    {
-      "epoch": 4.530259365994237,
-      "grad_norm": 0.023187410086393356,
-      "learning_rate": 2.63016787428354e-06,
-      "loss": 0.0026,
-      "step": 1768
-    },
-    {
-      "epoch": 4.532821005443484,
-      "grad_norm": 0.018168721348047256,
-      "learning_rate": 2.601596946975243e-06,
-      "loss": 0.0026,
-      "step": 1769
-    },
-    {
-      "epoch": 4.5353826448927315,
-      "grad_norm": 0.022810759022831917,
-      "learning_rate": 2.573177902642726e-06,
-      "loss": 0.0029,
-      "step": 1770
-    },
-    {
-      "epoch": 4.537944284341979,
-      "grad_norm": 0.02491903305053711,
-      "learning_rate": 2.544910832351688e-06,
-      "loss": 0.0033,
-      "step": 1771
-    },
-    {
-      "epoch": 4.540505923791226,
-      "grad_norm": 0.022031614556908607,
-      "learning_rate": 2.5167958266808456e-06,
-      "loss": 0.0017,
-      "step": 1772
-    },
-    {
-      "epoch": 4.543067563240474,
-      "grad_norm": 0.015833253040909767,
-      "learning_rate": 2.488832975721661e-06,
-      "loss": 0.0018,
-      "step": 1773
-    },
-    {
-      "epoch": 4.545629202689721,
-      "grad_norm": 0.01989520899951458,
-      "learning_rate": 2.4610223690780087e-06,
-      "loss": 0.0023,
-      "step": 1774
-    },
-    {
-      "epoch": 4.548190842138969,
-      "grad_norm": 0.016461314633488655,
-      "learning_rate": 2.4333640958659143e-06,
-      "loss": 0.0026,
-      "step": 1775
-    },
-    {
-      "epoch": 4.548190842138969,
-      "eval_loss": 0.007215212564915419,
-      "eval_runtime": 15.7922,
-      "eval_samples_per_second": 6.332,
-      "eval_steps_per_second": 1.583,
-      "step": 1775
-    },
-    {
-      "epoch": 4.5507524815882165,
-      "grad_norm": 0.015237953513860703,
-      "learning_rate": 2.405858244713283e-06,
-      "loss": 0.0018,
-      "step": 1776
-    },
-    {
-      "epoch": 4.553314121037464,
-      "grad_norm": 0.023646904155611992,
-      "learning_rate": 2.3785049037595987e-06,
-      "loss": 0.0023,
-      "step": 1777
-    },
-    {
-      "epoch": 4.555875760486711,
-      "grad_norm": 0.027399618178606033,
-      "learning_rate": 2.3513041606556375e-06,
-      "loss": 0.0029,
-      "step": 1778
-    },
-    {
-      "epoch": 4.558437399935959,
-      "grad_norm": 0.019109679386019707,
-      "learning_rate": 2.324256102563188e-06,
-      "loss": 0.0021,
-      "step": 1779
-    },
-    {
-      "epoch": 4.560999039385207,
-      "grad_norm": 0.023158298805356026,
-      "learning_rate": 2.2973608161547753e-06,
-      "loss": 0.0026,
-      "step": 1780
-    },
-    {
-      "epoch": 4.563560678834454,
-      "grad_norm": 0.017553139477968216,
-      "learning_rate": 2.2706183876134045e-06,
-      "loss": 0.0019,
-      "step": 1781
-    },
-    {
-      "epoch": 4.566122318283702,
-      "grad_norm": 0.02320447377860546,
-      "learning_rate": 2.244028902632239e-06,
-      "loss": 0.0034,
-      "step": 1782
-    },
-    {
-      "epoch": 4.568683957732949,
-      "grad_norm": 0.022288013249635696,
-      "learning_rate": 2.2175924464143627e-06,
-      "loss": 0.0032,
-      "step": 1783
-    },
-    {
-      "epoch": 4.571245597182196,
-      "grad_norm": 0.02146083116531372,
-      "learning_rate": 2.1913091036725062e-06,
-      "loss": 0.0018,
-      "step": 1784
-    },
-    {
-      "epoch": 4.573807236631444,
-      "grad_norm": 0.019133323803544044,
-      "learning_rate": 2.1651789586287442e-06,
-      "loss": 0.002,
-      "step": 1785
-    },
-    {
-      "epoch": 4.576368876080692,
-      "grad_norm": 0.02038615196943283,
-      "learning_rate": 2.1392020950142533e-06,
-      "loss": 0.0023,
-      "step": 1786
-    },
-    {
-      "epoch": 4.578930515529939,
-      "grad_norm": 0.02124660089612007,
-      "learning_rate": 2.1133785960690487e-06,
-      "loss": 0.0024,
-      "step": 1787
-    },
-    {
-      "epoch": 4.581492154979187,
-      "grad_norm": 0.024046916514635086,
-      "learning_rate": 2.087708544541689e-06,
-      "loss": 0.0026,
-      "step": 1788
-    },
-    {
-      "epoch": 4.584053794428434,
-      "grad_norm": 0.01588166505098343,
-      "learning_rate": 2.0621920226890355e-06,
-      "loss": 0.0022,
-      "step": 1789
-    },
-    {
-      "epoch": 4.5866154338776814,
-      "grad_norm": 0.024788126349449158,
-      "learning_rate": 2.03682911227599e-06,
-      "loss": 0.0024,
-      "step": 1790
-    },
-    {
-      "epoch": 4.58917707332693,
-      "grad_norm": 0.025243714451789856,
-      "learning_rate": 2.011619894575201e-06,
-      "loss": 0.0026,
-      "step": 1791
-    },
-    {
-      "epoch": 4.591738712776177,
-      "grad_norm": 0.023342831060290337,
-      "learning_rate": 1.9865644503668345e-06,
-      "loss": 0.0028,
-      "step": 1792
-    },
-    {
-      "epoch": 4.594300352225424,
-      "grad_norm": 0.020319147035479546,
-      "learning_rate": 1.961662859938318e-06,
-      "loss": 0.0028,
-      "step": 1793
-    },
-    {
-      "epoch": 4.596861991674672,
-      "grad_norm": 0.016949201002717018,
-      "learning_rate": 1.9369152030840556e-06,
-      "loss": 0.0024,
-      "step": 1794
-    },
-    {
-      "epoch": 4.599423631123919,
-      "grad_norm": 0.025084851309657097,
-      "learning_rate": 1.9123215591052013e-06,
-      "loss": 0.0036,
-      "step": 1795
-    },
-    {
-      "epoch": 4.6019852705731665,
-      "grad_norm": 0.018058422952890396,
-      "learning_rate": 1.8878820068093884e-06,
-      "loss": 0.002,
-      "step": 1796
-    },
-    {
-      "epoch": 4.604546910022414,
-      "grad_norm": 0.022538725286722183,
-      "learning_rate": 1.8635966245104664e-06,
-      "loss": 0.0029,
-      "step": 1797
-    },
-    {
-      "epoch": 4.607108549471662,
-      "grad_norm": 0.01928574964404106,
-      "learning_rate": 1.8394654900282804e-06,
-      "loss": 0.0025,
-      "step": 1798
-    },
-    {
-      "epoch": 4.6096701889209095,
-      "grad_norm": 0.018158117309212685,
-      "learning_rate": 1.8154886806883985e-06,
-      "loss": 0.0027,
-      "step": 1799
-    },
-    {
-      "epoch": 4.612231828370157,
-      "grad_norm": 0.024838857352733612,
-      "learning_rate": 1.7916662733218847e-06,
-      "loss": 0.0023,
-      "step": 1800
-    },
-    {
-      "epoch": 4.612231828370157,
-      "eval_loss": 0.007183325942605734,
-      "eval_runtime": 15.7766,
-      "eval_samples_per_second": 6.339,
-      "eval_steps_per_second": 1.585,
-      "step": 1800
-    },
-    {
-      "epoch": 4.614793467819404,
-      "grad_norm": 0.018951160833239555,
-      "learning_rate": 1.7679983442650095e-06,
-      "loss": 0.0021,
-      "step": 1801
-    },
-    {
-      "epoch": 4.617355107268652,
-      "grad_norm": 0.02853049337863922,
-      "learning_rate": 1.744484969359056e-06,
-      "loss": 0.0027,
-      "step": 1802
-    },
-    {
-      "epoch": 4.6199167467179,
-      "grad_norm": 0.018323928117752075,
-      "learning_rate": 1.7211262239500591e-06,
-      "loss": 0.0025,
-      "step": 1803
-    },
-    {
-      "epoch": 4.622478386167147,
-      "grad_norm": 0.021388335153460503,
-      "learning_rate": 1.6979221828885393e-06,
-      "loss": 0.0021,
-      "step": 1804
-    },
-    {
-      "epoch": 4.625040025616395,
-      "grad_norm": 0.020138802006840706,
-      "learning_rate": 1.6748729205293023e-06,
-      "loss": 0.0021,
-      "step": 1805
-    },
-    {
-      "epoch": 4.627601665065642,
-      "grad_norm": 0.026121465489268303,
-      "learning_rate": 1.6519785107311891e-06,
-      "loss": 0.0024,
-      "step": 1806
-    },
-    {
-      "epoch": 4.630163304514889,
-      "grad_norm": 0.017890939489006996,
-      "learning_rate": 1.6292390268568104e-06,
-      "loss": 0.0017,
-      "step": 1807
-    },
-    {
-      "epoch": 4.632724943964137,
-      "grad_norm": 0.02806764468550682,
-      "learning_rate": 1.6066545417723343e-06,
-      "loss": 0.004,
-      "step": 1808
-    },
-    {
-      "epoch": 4.635286583413385,
-      "grad_norm": 0.024905608966946602,
-      "learning_rate": 1.584225127847283e-06,
-      "loss": 0.0019,
-      "step": 1809
-    },
-    {
-      "epoch": 4.637848222862632,
-      "grad_norm": 0.021152952685952187,
-      "learning_rate": 1.5619508569542363e-06,
-      "loss": 0.0021,
-      "step": 1810
-    },
-    {
-      "epoch": 4.64040986231188,
-      "grad_norm": 0.018304862082004547,
-      "learning_rate": 1.5398318004686496e-06,
-      "loss": 0.0018,
-      "step": 1811
-    },
-    {
-      "epoch": 4.642971501761127,
-      "grad_norm": 0.018207866698503494,
-      "learning_rate": 1.5178680292686154e-06,
-      "loss": 0.0027,
-      "step": 1812
-    },
-    {
-      "epoch": 4.645533141210374,
-      "grad_norm": 0.026254169642925262,
-      "learning_rate": 1.496059613734624e-06,
-      "loss": 0.0042,
-      "step": 1813
-    },
-    {
-      "epoch": 4.648094780659622,
-      "grad_norm": 0.019822141155600548,
-      "learning_rate": 1.4744066237493415e-06,
-      "loss": 0.0024,
-      "step": 1814
-    },
-    {
-      "epoch": 4.65065642010887,
-      "grad_norm": 0.021628927439451218,
-      "learning_rate": 1.4529091286973995e-06,
-      "loss": 0.0022,
-      "step": 1815
-    },
-    {
-      "epoch": 4.653218059558117,
-      "grad_norm": 0.022384101524949074,
-      "learning_rate": 1.4315671974651612e-06,
-      "loss": 0.003,
-      "step": 1816
-    },
-    {
-      "epoch": 4.655779699007365,
-      "grad_norm": 0.020145338028669357,
-      "learning_rate": 1.4103808984404942e-06,
-      "loss": 0.0021,
-      "step": 1817
-    },
-    {
-      "epoch": 4.658341338456612,
-      "grad_norm": 0.016928505152463913,
-      "learning_rate": 1.3893502995125818e-06,
-      "loss": 0.002,
-      "step": 1818
-    },
-    {
-      "epoch": 4.6609029779058595,
-      "grad_norm": 0.01927059143781662,
-      "learning_rate": 1.3684754680716506e-06,
-      "loss": 0.003,
-      "step": 1819
-    },
-    {
-      "epoch": 4.663464617355107,
-      "grad_norm": 0.017414642497897148,
-      "learning_rate": 1.3477564710088098e-06,
-      "loss": 0.002,
-      "step": 1820
-    },
-    {
-      "epoch": 4.666026256804355,
-      "grad_norm": 0.0180311668664217,
-      "learning_rate": 1.3271933747158128e-06,
-      "loss": 0.002,
-      "step": 1821
-    },
-    {
-      "epoch": 4.6685878962536025,
-      "grad_norm": 0.01807776838541031,
-      "learning_rate": 1.30678624508484e-06,
-      "loss": 0.0021,
-      "step": 1822
-    },
-    {
-      "epoch": 4.67114953570285,
-      "grad_norm": 0.014348401688039303,
-      "learning_rate": 1.2865351475082998e-06,
-      "loss": 0.0019,
-      "step": 1823
-    },
-    {
-      "epoch": 4.673711175152097,
-      "grad_norm": 0.013491121120750904,
-      "learning_rate": 1.2664401468786114e-06,
-      "loss": 0.0016,
-      "step": 1824
-    },
-    {
-      "epoch": 4.676272814601345,
-      "grad_norm": 0.0232559721916914,
-      "learning_rate": 1.2465013075879883e-06,
-      "loss": 0.0023,
-      "step": 1825
-    },
-    {
-      "epoch": 4.676272814601345,
-      "eval_loss": 0.007197246421128511,
-      "eval_runtime": 15.823,
-      "eval_samples_per_second": 6.32,
-      "eval_steps_per_second": 1.58,
-      "step": 1825
-    },
-    {
-      "epoch": 4.678834454050593,
-      "grad_norm": 0.02286207117140293,
-      "learning_rate": 1.226718693528256e-06,
-      "loss": 0.0026,
-      "step": 1826
-    },
-    {
-      "epoch": 4.68139609349984,
-      "grad_norm": 0.020442461594939232,
-      "learning_rate": 1.2070923680906231e-06,
-      "loss": 0.002,
-      "step": 1827
-    },
-    {
-      "epoch": 4.683957732949088,
-      "grad_norm": 0.017635615542531013,
-      "learning_rate": 1.1876223941655051e-06,
-      "loss": 0.0022,
-      "step": 1828
-    },
-    {
-      "epoch": 4.686519372398335,
-      "grad_norm": 0.019227273762226105,
-      "learning_rate": 1.1683088341422843e-06,
-      "loss": 0.0026,
-      "step": 1829
-    },
-    {
-      "epoch": 4.689081011847582,
-      "grad_norm": 0.031491003930568695,
-      "learning_rate": 1.1491517499091498e-06,
-      "loss": 0.0036,
-      "step": 1830
-    },
-    {
-      "epoch": 4.69164265129683,
-      "grad_norm": 0.02115483582019806,
-      "learning_rate": 1.1301512028528694e-06,
-      "loss": 0.0028,
-      "step": 1831
-    },
-    {
-      "epoch": 4.694204290746077,
-      "grad_norm": 0.013307468965649605,
-      "learning_rate": 1.111307253858601e-06,
-      "loss": 0.0012,
-      "step": 1832
-    },
-    {
-      "epoch": 4.696765930195325,
-      "grad_norm": 0.017989305779337883,
-      "learning_rate": 1.0926199633097157e-06,
-      "loss": 0.0024,
-      "step": 1833
-    },
-    {
-      "epoch": 4.699327569644573,
-      "grad_norm": 0.02980799227952957,
-      "learning_rate": 1.0740893910875904e-06,
-      "loss": 0.0037,
-      "step": 1834
-    },
-    {
-      "epoch": 4.70188920909382,
-      "grad_norm": 0.016261843964457512,
-      "learning_rate": 1.055715596571405e-06,
-      "loss": 0.0019,
-      "step": 1835
-    },
-    {
-      "epoch": 4.704450848543067,
-      "grad_norm": 0.02803937904536724,
-      "learning_rate": 1.037498638637957e-06,
-      "loss": 0.003,
-      "step": 1836
-    },
-    {
-      "epoch": 4.707012487992315,
-      "grad_norm": 0.01880720630288124,
-      "learning_rate": 1.019438575661491e-06,
-      "loss": 0.0025,
-      "step": 1837
-    },
-    {
-      "epoch": 4.709574127441563,
-      "grad_norm": 0.019532645121216774,
-      "learning_rate": 1.0015354655135034e-06,
-      "loss": 0.0019,
-      "step": 1838
-    },
-    {
-      "epoch": 4.71213576689081,
-      "grad_norm": 0.020650824531912804,
-      "learning_rate": 9.837893655625318e-07,
-      "loss": 0.0025,
-      "step": 1839
-    },
-    {
-      "epoch": 4.714697406340058,
-      "grad_norm": 0.03180498257279396,
-      "learning_rate": 9.662003326740166e-07,
-      "loss": 0.0037,
-      "step": 1840
-    },
-    {
-      "epoch": 4.717259045789305,
-      "grad_norm": 0.014423294924199581,
-      "learning_rate": 9.487684232100669e-07,
-      "loss": 0.0013,
-      "step": 1841
-    },
-    {
-      "epoch": 4.7198206852385525,
-      "grad_norm": 0.03218596428632736,
-      "learning_rate": 9.314936930293283e-07,
-      "loss": 0.0026,
-      "step": 1842
-    },
-    {
-      "epoch": 4.7223823246878,
-      "grad_norm": 0.01891743578016758,
-      "learning_rate": 9.143761974867771e-07,
-      "loss": 0.0021,
-      "step": 1843
-    },
-    {
-      "epoch": 4.724943964137048,
-      "grad_norm": 0.016995957121253014,
-      "learning_rate": 8.974159914335534e-07,
-      "loss": 0.0017,
-      "step": 1844
-    },
-    {
-      "epoch": 4.7275056035862955,
-      "grad_norm": 0.023755529895424843,
-      "learning_rate": 8.806131292167618e-07,
-      "loss": 0.0027,
-      "step": 1845
-    },
-    {
-      "epoch": 4.730067243035543,
-      "grad_norm": 0.01824359968304634,
-      "learning_rate": 8.639676646793382e-07,
-      "loss": 0.002,
-      "step": 1846
-    },
-    {
-      "epoch": 4.73262888248479,
-      "grad_norm": 0.018043162301182747,
-      "learning_rate": 8.474796511598382e-07,
-      "loss": 0.0027,
-      "step": 1847
-    },
-    {
-      "epoch": 4.7351905219340376,
-      "grad_norm": 0.020727921277284622,
-      "learning_rate": 8.311491414922934e-07,
-      "loss": 0.0025,
-      "step": 1848
-    },
-    {
-      "epoch": 4.737752161383285,
-      "grad_norm": 0.020384375005960464,
-      "learning_rate": 8.149761880060281e-07,
-      "loss": 0.003,
-      "step": 1849
-    },
-    {
-      "epoch": 4.740313800832533,
-      "grad_norm": 0.018516646698117256,
-      "learning_rate": 7.989608425254924e-07,
-      "loss": 0.0019,
-      "step": 1850
-    },
-    {
-      "epoch": 4.740313800832533,
-      "eval_loss": 0.007202571257948875,
-      "eval_runtime": 15.7826,
-      "eval_samples_per_second": 6.336,
-      "eval_steps_per_second": 1.584,
-      "step": 1850
-    },
-    {
-      "epoch": 4.7428754402817805,
-      "grad_norm": 0.015421532094478607,
-      "learning_rate": 7.83103156370113e-07,
-      "loss": 0.0023,
-      "step": 1851
-    },
-    {
-      "epoch": 4.745437079731028,
-      "grad_norm": 0.013238969258964062,
-      "learning_rate": 7.674031803540926e-07,
-      "loss": 0.0019,
-      "step": 1852
-    },
-    {
-      "epoch": 4.747998719180275,
-      "grad_norm": 0.02619444951415062,
-      "learning_rate": 7.518609647862718e-07,
-      "loss": 0.0034,
-      "step": 1853
-    },
-    {
-      "epoch": 4.750560358629523,
-      "grad_norm": 0.025059690698981285,
-      "learning_rate": 7.364765594699951e-07,
-      "loss": 0.003,
-      "step": 1854
-    },
-    {
-      "epoch": 4.75312199807877,
-      "grad_norm": 0.019235456362366676,
-      "learning_rate": 7.212500137028788e-07,
-      "loss": 0.002,
-      "step": 1855
-    },
-    {
-      "epoch": 4.755683637528018,
-      "grad_norm": 0.021623438224196434,
-      "learning_rate": 7.061813762767211e-07,
-      "loss": 0.0024,
-      "step": 1856
-    },
-    {
-      "epoch": 4.758245276977266,
-      "grad_norm": 0.017953049391508102,
-      "learning_rate": 6.912706954773251e-07,
-      "loss": 0.0023,
-      "step": 1857
-    },
-    {
-      "epoch": 4.760806916426513,
-      "grad_norm": 0.03642131760716438,
-      "learning_rate": 6.765180190843212e-07,
-      "loss": 0.0031,
-      "step": 1858
-    },
-    {
-      "epoch": 4.76336855587576,
-      "grad_norm": 0.03756588324904442,
-      "learning_rate": 6.61923394371039e-07,
-      "loss": 0.0036,
-      "step": 1859
-    },
-    {
-      "epoch": 4.765930195325008,
-      "grad_norm": 0.018120767548680305,
-      "learning_rate": 6.474868681043578e-07,
-      "loss": 0.0018,
-      "step": 1860
-    },
-    {
-      "epoch": 4.768491834774256,
-      "grad_norm": 0.021253090351819992,
-      "learning_rate": 6.332084865445342e-07,
-      "loss": 0.0026,
-      "step": 1861
-    },
-    {
-      "epoch": 4.771053474223503,
-      "grad_norm": 0.02116861194372177,
-      "learning_rate": 6.190882954450694e-07,
-      "loss": 0.0027,
-      "step": 1862
-    },
-    {
-      "epoch": 4.773615113672751,
-      "grad_norm": 0.016063787043094635,
-      "learning_rate": 6.051263400525753e-07,
-      "loss": 0.0019,
-      "step": 1863
-    },
-    {
-      "epoch": 4.776176753121998,
-      "grad_norm": 0.014829093590378761,
-      "learning_rate": 5.913226651065973e-07,
-      "loss": 0.0018,
-      "step": 1864
-    },
-    {
-      "epoch": 4.778738392571245,
-      "grad_norm": 0.020974157378077507,
-      "learning_rate": 5.776773148394976e-07,
-      "loss": 0.0029,
-      "step": 1865
-    },
-    {
-      "epoch": 4.781300032020493,
-      "grad_norm": 0.02501363679766655,
-      "learning_rate": 5.641903329762998e-07,
-      "loss": 0.003,
-      "step": 1866
-    },
-    {
-      "epoch": 4.78386167146974,
-      "grad_norm": 0.020856289193034172,
-      "learning_rate": 5.508617627345558e-07,
-      "loss": 0.0023,
-      "step": 1867
-    },
-    {
-      "epoch": 4.786423310918988,
-      "grad_norm": 0.018088124692440033,
-      "learning_rate": 5.376916468242066e-07,
-      "loss": 0.0018,
-      "step": 1868
-    },
-    {
-      "epoch": 4.788984950368236,
-      "grad_norm": 0.0162031389772892,
-      "learning_rate": 5.246800274474439e-07,
-      "loss": 0.0026,
-      "step": 1869
-    },
-    {
-      "epoch": 4.791546589817483,
-      "grad_norm": 0.0184140894562006,
-      "learning_rate": 5.118269462985714e-07,
-      "loss": 0.0017,
-      "step": 1870
-    },
-    {
-      "epoch": 4.7941082292667305,
-      "grad_norm": 0.01935194432735443,
-      "learning_rate": 4.991324445638712e-07,
-      "loss": 0.0022,
-      "step": 1871
-    },
-    {
-      "epoch": 4.796669868715978,
-      "grad_norm": 0.021204661577939987,
-      "learning_rate": 4.865965629214819e-07,
-      "loss": 0.0027,
-      "step": 1872
-    },
-    {
-      "epoch": 4.799231508165226,
-      "grad_norm": 0.035143256187438965,
-      "learning_rate": 4.7421934154126545e-07,
-      "loss": 0.0031,
-      "step": 1873
-    },
-    {
-      "epoch": 4.8017931476144735,
-      "grad_norm": 0.02215544879436493,
-      "learning_rate": 4.6200082008466263e-07,
-      "loss": 0.0025,
-      "step": 1874
-    },
-    {
-      "epoch": 4.804354787063721,
-      "grad_norm": 0.052750926464796066,
-      "learning_rate": 4.4994103770457653e-07,
-      "loss": 0.0025,
-      "step": 1875
-    },
-    {
-      "epoch": 4.804354787063721,
-      "eval_loss": 0.007240465842187405,
-      "eval_runtime": 15.7811,
-      "eval_samples_per_second": 6.337,
-      "eval_steps_per_second": 1.584,
-      "step": 1875
-    },
-    {
-      "epoch": 4.806916426512968,
-      "grad_norm": 0.015583074651658535,
-      "learning_rate": 4.3804003304526717e-07,
-      "loss": 0.0017,
-      "step": 1876
-    },
-    {
-      "epoch": 4.809478065962216,
-      "grad_norm": 0.01865261234343052,
-      "learning_rate": 4.2629784424219056e-07,
-      "loss": 0.0022,
-      "step": 1877
-    },
-    {
-      "epoch": 4.812039705411463,
-      "grad_norm": 0.018412189558148384,
-      "learning_rate": 4.1471450892189846e-07,
-      "loss": 0.0028,
-      "step": 1878
-    },
-    {
-      "epoch": 4.814601344860711,
-      "grad_norm": 0.025617893785238266,
-      "learning_rate": 4.0329006420192215e-07,
-      "loss": 0.0033,
-      "step": 1879
-    },
-    {
-      "epoch": 4.817162984309959,
-      "grad_norm": 0.02348766289651394,
-      "learning_rate": 3.920245466906391e-07,
-      "loss": 0.0029,
-      "step": 1880
-    },
-    {
-      "epoch": 4.819724623759206,
-      "grad_norm": 0.018552901223301888,
-      "learning_rate": 3.809179924871564e-07,
-      "loss": 0.0025,
-      "step": 1881
-    },
-    {
-      "epoch": 4.822286263208453,
-      "grad_norm": 0.021071987226605415,
-      "learning_rate": 3.6997043718121626e-07,
-      "loss": 0.0024,
-      "step": 1882
-    },
-    {
-      "epoch": 4.824847902657701,
-      "grad_norm": 0.020377142354846,
-      "learning_rate": 3.5918191585305183e-07,
-      "loss": 0.0023,
-      "step": 1883
-    },
-    {
-      "epoch": 4.827409542106948,
-      "grad_norm": 0.03159242868423462,
-      "learning_rate": 3.4855246307328727e-07,
-      "loss": 0.0024,
-      "step": 1884
-    },
-    {
-      "epoch": 4.829971181556196,
-      "grad_norm": 0.023257169872522354,
-      "learning_rate": 3.380821129028489e-07,
-      "loss": 0.0023,
-      "step": 1885
-    },
-    {
-      "epoch": 4.832532821005444,
-      "grad_norm": 0.015475527383387089,
-      "learning_rate": 3.2777089889282075e-07,
-      "loss": 0.0018,
-      "step": 1886
-    },
-    {
-      "epoch": 4.835094460454691,
-      "grad_norm": 0.026007868349552155,
-      "learning_rate": 3.1761885408435054e-07,
-      "loss": 0.0022,
-      "step": 1887
-    },
-    {
-      "epoch": 4.837656099903938,
-      "grad_norm": 0.02517171949148178,
-      "learning_rate": 3.076260110085549e-07,
-      "loss": 0.0032,
-      "step": 1888
-    },
-    {
-      "epoch": 4.840217739353186,
-      "grad_norm": 0.021184207871556282,
-      "learning_rate": 2.9779240168639754e-07,
-      "loss": 0.0023,
-      "step": 1889
-    },
-    {
-      "epoch": 4.842779378802433,
-      "grad_norm": 0.016074050217866898,
-      "learning_rate": 2.8811805762860576e-07,
-      "loss": 0.0024,
-      "step": 1890
-    },
-    {
-      "epoch": 4.845341018251681,
-      "grad_norm": 0.024288572371006012,
-      "learning_rate": 2.78603009835543e-07,
-      "loss": 0.0039,
-      "step": 1891
-    },
-    {
-      "epoch": 4.847902657700929,
-      "grad_norm": 0.014958859421312809,
-      "learning_rate": 2.6924728879714754e-07,
-      "loss": 0.0016,
-      "step": 1892
-    },
-    {
-      "epoch": 4.850464297150176,
-      "grad_norm": 0.014642615802586079,
-      "learning_rate": 2.600509244927829e-07,
-      "loss": 0.0017,
-      "step": 1893
-    },
-    {
-      "epoch": 4.8530259365994235,
-      "grad_norm": 0.019684061408042908,
-      "learning_rate": 2.510139463912042e-07,
-      "loss": 0.0025,
-      "step": 1894
-    },
-    {
-      "epoch": 4.855587576048671,
-      "grad_norm": 0.028054427355527878,
-      "learning_rate": 2.421363834504087e-07,
-      "loss": 0.0023,
-      "step": 1895
-    },
-    {
-      "epoch": 4.858149215497919,
-      "grad_norm": 0.02247677557170391,
-      "learning_rate": 2.334182641175686e-07,
-      "loss": 0.0031,
-      "step": 1896
-    },
-    {
-      "epoch": 4.8607108549471665,
-      "grad_norm": 0.020722435787320137,
-      "learning_rate": 2.2485961632894292e-07,
-      "loss": 0.0026,
-      "step": 1897
-    },
-    {
-      "epoch": 4.863272494396414,
-      "grad_norm": 0.028742695227265358,
-      "learning_rate": 2.1646046750978254e-07,
-      "loss": 0.0046,
-      "step": 1898
-    },
-    {
-      "epoch": 4.865834133845661,
-      "grad_norm": 0.020024484023451805,
-      "learning_rate": 2.082208445742251e-07,
-      "loss": 0.0025,
-      "step": 1899
-    },
-    {
-      "epoch": 4.868395773294909,
-      "grad_norm": 0.018740255385637283,
-      "learning_rate": 2.0014077392525031e-07,
-      "loss": 0.0023,
-      "step": 1900
-    },
-    {
-      "epoch": 4.868395773294909,
-      "eval_loss": 0.007208134513348341,
-      "eval_runtime": 15.7883,
-      "eval_samples_per_second": 6.334,
-      "eval_steps_per_second": 1.583,
-      "step": 1900
-    },
-    {
-      "epoch": 4.870957412744156,
-      "grad_norm": 0.017158091068267822,
-      "learning_rate": 1.9222028145454707e-07,
-      "loss": 0.0021,
-      "step": 1901
-    },
-    {
-      "epoch": 4.873519052193403,
-      "grad_norm": 0.03217052295804024,
-      "learning_rate": 1.844593925424687e-07,
-      "loss": 0.0031,
-      "step": 1902
-    },
-    {
-      "epoch": 4.876080691642652,
-      "grad_norm": 0.02281808853149414,
-      "learning_rate": 1.7685813205793323e-07,
-      "loss": 0.0028,
-      "step": 1903
-    },
-    {
-      "epoch": 4.878642331091899,
-      "grad_norm": 0.020626462996006012,
-      "learning_rate": 1.694165243583401e-07,
-      "loss": 0.0022,
-      "step": 1904
-    },
-    {
-      "epoch": 4.881203970541146,
-      "grad_norm": 0.013555928133428097,
-      "learning_rate": 1.6213459328950352e-07,
-      "loss": 0.0019,
-      "step": 1905
-    },
-    {
-      "epoch": 4.883765609990394,
-      "grad_norm": 0.026738880202174187,
-      "learning_rate": 1.5501236218556925e-07,
-      "loss": 0.0038,
-      "step": 1906
-    },
-    {
-      "epoch": 4.886327249439641,
-      "grad_norm": 0.01642564684152603,
-      "learning_rate": 1.4804985386895342e-07,
-      "loss": 0.0018,
-      "step": 1907
-    },
-    {
-      "epoch": 4.888888888888889,
-      "grad_norm": 0.030033448711037636,
-      "learning_rate": 1.4124709065024832e-07,
-      "loss": 0.0029,
-      "step": 1908
-    },
-    {
-      "epoch": 4.891450528338137,
-      "grad_norm": 0.01564773917198181,
-      "learning_rate": 1.3460409432815012e-07,
-      "loss": 0.0022,
-      "step": 1909
-    },
-    {
-      "epoch": 4.894012167787384,
-      "grad_norm": 0.019766777753829956,
-      "learning_rate": 1.281208861894201e-07,
-      "loss": 0.0025,
-      "step": 1910
-    },
-    {
-      "epoch": 4.896573807236631,
-      "grad_norm": 0.018657788634300232,
-      "learning_rate": 1.2179748700879012e-07,
-      "loss": 0.0022,
-      "step": 1911
-    },
-    {
-      "epoch": 4.899135446685879,
-      "grad_norm": 0.02429956942796707,
-      "learning_rate": 1.1563391704888515e-07,
-      "loss": 0.0022,
-      "step": 1912
-    },
-    {
-      "epoch": 4.901697086135126,
-      "grad_norm": 0.03048797696828842,
-      "learning_rate": 1.0963019606019532e-07,
-      "loss": 0.0031,
-      "step": 1913
-    },
-    {
-      "epoch": 4.904258725584374,
-      "grad_norm": 0.0223206989467144,
-      "learning_rate": 1.0378634328099269e-07,
-      "loss": 0.003,
-      "step": 1914
-    },
-    {
-      "epoch": 4.906820365033622,
-      "grad_norm": 0.01883537694811821,
-      "learning_rate": 9.810237743724803e-08,
-      "loss": 0.0025,
-      "step": 1915
-    },
-    {
-      "epoch": 4.909382004482869,
-      "grad_norm": 0.014887621626257896,
-      "learning_rate": 9.257831674261418e-08,
-      "loss": 0.0018,
-      "step": 1916
-    },
-    {
-      "epoch": 4.9119436439321165,
-      "grad_norm": 0.018005670979619026,
-      "learning_rate": 8.72141788983316e-08,
-      "loss": 0.0023,
-      "step": 1917
-    },
-    {
-      "epoch": 4.914505283381364,
-      "grad_norm": 0.02114519104361534,
-      "learning_rate": 8.200998109319513e-08,
-      "loss": 0.0023,
-      "step": 1918
-    },
-    {
-      "epoch": 4.917066922830612,
-      "grad_norm": 0.019435500726103783,
-      "learning_rate": 7.696574000348178e-08,
-      "loss": 0.0028,
-      "step": 1919
-    },
-    {
-      "epoch": 4.9196285622798595,
-      "grad_norm": 0.020121954381465912,
-      "learning_rate": 7.208147179291192e-08,
-      "loss": 0.0027,
-      "step": 1920
-    },
-    {
-      "epoch": 4.922190201729107,
-      "grad_norm": 0.023929473012685776,
-      "learning_rate": 6.735719211257708e-08,
-      "loss": 0.0026,
-      "step": 1921
-    },
-    {
-      "epoch": 4.924751841178354,
-      "grad_norm": 0.02012651227414608,
-      "learning_rate": 6.27929161009233e-08,
-      "loss": 0.0028,
-      "step": 1922
-    },
-    {
-      "epoch": 4.9273134806276015,
-      "grad_norm": 0.019554229453206062,
-      "learning_rate": 5.838865838366792e-08,
-      "loss": 0.0022,
-      "step": 1923
-    },
-    {
-      "epoch": 4.929875120076849,
-      "grad_norm": 0.01740524172782898,
-      "learning_rate": 5.4144433073771707e-08,
-      "loss": 0.0021,
-      "step": 1924
-    },
-    {
-      "epoch": 4.932436759526096,
-      "grad_norm": 0.024000542238354683,
-      "learning_rate": 5.006025377138901e-08,
-      "loss": 0.0021,
-      "step": 1925
-    },
-    {
-      "epoch": 4.932436759526096,
-      "eval_loss": 0.007247826550155878,
-      "eval_runtime": 15.7615,
-      "eval_samples_per_second": 6.345,
-      "eval_steps_per_second": 1.586,
-      "step": 1925
-    },
-    {
-      "epoch": 4.9349983989753445,
-      "grad_norm": 0.017517974600195885,
-      "learning_rate": 4.6136133563817743e-08,
-      "loss": 0.0021,
-      "step": 1926
-    },
-    {
-      "epoch": 4.937560038424592,
-      "grad_norm": 0.02366037294268608,
-      "learning_rate": 4.2372085025477184e-08,
-      "loss": 0.0022,
-      "step": 1927
-    },
-    {
-      "epoch": 4.940121677873839,
-      "grad_norm": 0.02913198620080948,
-      "learning_rate": 3.8768120217846927e-08,
-      "loss": 0.0028,
-      "step": 1928
-    },
-    {
-      "epoch": 4.942683317323087,
-      "grad_norm": 0.02785470150411129,
-      "learning_rate": 3.532425068942802e-08,
-      "loss": 0.0021,
-      "step": 1929
-    },
-    {
-      "epoch": 4.945244956772334,
-      "grad_norm": 0.020093509927392006,
-      "learning_rate": 3.2040487475731854e-08,
-      "loss": 0.0022,
-      "step": 1930
-    },
-    {
-      "epoch": 4.947806596221582,
-      "grad_norm": 0.033045586198568344,
-      "learning_rate": 2.8916841099213554e-08,
-      "loss": 0.0032,
-      "step": 1931
-    },
-    {
-      "epoch": 4.95036823567083,
-      "grad_norm": 0.06443755328655243,
-      "learning_rate": 2.595332156925534e-08,
-      "loss": 0.0022,
-      "step": 1932
-    },
-    {
-      "epoch": 4.952929875120077,
-      "grad_norm": 0.022899948060512543,
-      "learning_rate": 2.3149938382122095e-08,
-      "loss": 0.0029,
-      "step": 1933
-    },
-    {
-      "epoch": 4.955491514569324,
-      "grad_norm": 0.023586386814713478,
-      "learning_rate": 2.0506700520961374e-08,
-      "loss": 0.0022,
-      "step": 1934
-    },
-    {
-      "epoch": 4.958053154018572,
-      "grad_norm": 0.020165063440799713,
-      "learning_rate": 1.802361645573125e-08,
-      "loss": 0.0019,
-      "step": 1935
-    },
-    {
-      "epoch": 4.960614793467819,
-      "grad_norm": 0.015795165672898293,
-      "learning_rate": 1.570069414319475e-08,
-      "loss": 0.0018,
-      "step": 1936
-    },
-    {
-      "epoch": 4.9631764329170664,
-      "grad_norm": 0.013385736383497715,
-      "learning_rate": 1.3537941026914303e-08,
-      "loss": 0.0019,
-      "step": 1937
-    },
-    {
-      "epoch": 4.965738072366315,
-      "grad_norm": 0.02121737413108349,
-      "learning_rate": 1.1535364037190688e-08,
-      "loss": 0.0022,
-      "step": 1938
-    },
-    {
-      "epoch": 4.968299711815562,
-      "grad_norm": 0.01854584738612175,
-      "learning_rate": 9.692969591057478e-09,
-      "loss": 0.0028,
-      "step": 1939
-    },
-    {
-      "epoch": 4.970861351264809,
-      "grad_norm": 0.034055206924676895,
-      "learning_rate": 8.010763592264381e-09,
-      "loss": 0.0034,
-      "step": 1940
-    },
-    {
-      "epoch": 4.973422990714057,
-      "grad_norm": 0.019612444564700127,
-      "learning_rate": 6.488751431266149e-09,
-      "loss": 0.003,
-      "step": 1941
-    },
-    {
-      "epoch": 4.975984630163304,
-      "grad_norm": 0.016554439440369606,
-      "learning_rate": 5.126937985172608e-09,
-      "loss": 0.0018,
-      "step": 1942
-    },
-    {
-      "epoch": 4.978546269612552,
-      "grad_norm": 0.02089136466383934,
-      "learning_rate": 3.9253276177708685e-09,
-      "loss": 0.0047,
-      "step": 1943
-    },
-    {
-      "epoch": 4.9811079090618,
-      "grad_norm": 0.0305600855499506,
-      "learning_rate": 2.8839241794864636e-09,
-      "loss": 0.0028,
-      "step": 1944
-    },
-    {
-      "epoch": 4.983669548511047,
-      "grad_norm": 0.015379647724330425,
-      "learning_rate": 2.0027310073833518e-09,
-      "loss": 0.002,
-      "step": 1945
-    },
-    {
-      "epoch": 4.9862311879602945,
-      "grad_norm": 0.023715125396847725,
-      "learning_rate": 1.2817509251472627e-09,
-      "loss": 0.0027,
-      "step": 1946
-    },
-    {
-      "epoch": 4.988792827409542,
-      "grad_norm": 0.022410869598388672,
-      "learning_rate": 7.209862430801462e-10,
-      "loss": 0.0022,
-      "step": 1947
-    },
-    {
-      "epoch": 4.991354466858789,
-      "grad_norm": 0.018450425937771797,
-      "learning_rate": 3.204387580946211e-10,
-      "loss": 0.002,
-      "step": 1948
-    },
-    {
-      "epoch": 4.9939161063080375,
-      "grad_norm": 0.022843310609459877,
-      "learning_rate": 8.010975369732166e-11,
-      "loss": 0.0026,
-      "step": 1949
-    },
-    {
-      "epoch": 4.996477745757285,
-      "grad_norm": 0.017766928300261497,
-      "learning_rate": 0.0,
-      "loss": 0.0028,
-      "step": 1950
-    },
-    {
-      "epoch": 4.996477745757285,
-      "eval_loss": 0.00722610903903842,
-      "eval_runtime": 15.8268,
-      "eval_samples_per_second": 6.318,
-      "eval_steps_per_second": 1.58,
-      "step": 1950
-    },
-    {
-      "epoch": 4.996477745757285,
-      "step": 1950,
-      "total_flos": 3.832695062431138e+18,
-      "train_loss": 0.009293928827003887,
-      "train_runtime": 29383.8827,
-      "train_samples_per_second": 2.125,
-      "train_steps_per_second": 0.066
-    }
-  ],
-  "logging_steps": 1,
-  "max_steps": 1950,
-  "num_input_tokens_seen": 0,
-  "num_train_epochs": 5,
-  "save_steps": 100,
-  "stateful_callbacks": {
     "TrainerControl": {
       "args": {
         "should_epoch_stop": false,
         "should_evaluate": false,
         "should_log": false,
         "should_save": true,
-        "should_training_stop": true
+        "should_training_stop": false
       },
       "attributes": {}
     }
   },
-  "total_flos": 3.832695062431138e+18,
+  "total_flos": 2.884558791231996e+18,
   "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null