diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -3,3412 +3,8690 @@
   "best_model_checkpoint": null,
   "epoch": 1.0,
   "eval_steps": 500,
-  "global_step": 2414,
+  "global_step": 6181,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.00041425020712510354,
-      "grad_norm": 24.199668764830136,
-      "learning_rate": 4.132231404958678e-08,
-      "loss": 1.3991,
+      "epoch": 0.00016178611875101117,
+      "grad_norm": 2.989531072795082,
+      "learning_rate": 1.6155088852988694e-08,
+      "loss": 1.6714,
       "step": 1
     },
     {
-      "epoch": 0.002071251035625518,
-      "grad_norm": 23.414571187298456,
-      "learning_rate": 2.066115702479339e-07,
-      "loss": 1.4112,
+      "epoch": 0.0008089305937550558,
+      "grad_norm": 3.0740133028448042,
+      "learning_rate": 8.077544426494346e-08,
+      "loss": 1.6455,
       "step": 5
     },
     {
-      "epoch": 0.004142502071251036,
-      "grad_norm": 14.988102359744419,
-      "learning_rate": 4.132231404958678e-07,
-      "loss": 1.3662,
+      "epoch": 0.0016178611875101116,
+      "grad_norm": 3.05751861374171,
+      "learning_rate": 1.6155088852988693e-07,
+      "loss": 1.6578,
       "step": 10
     },
     {
-      "epoch": 0.006213753106876553,
-      "grad_norm": 8.653198188938534,
-      "learning_rate": 6.198347107438018e-07,
-      "loss": 1.2641,
+      "epoch": 0.0024267917812651673,
+      "grad_norm": 3.05989958650721,
+      "learning_rate": 2.4232633279483037e-07,
+      "loss": 1.6633,
       "step": 15
     },
     {
-      "epoch": 0.008285004142502071,
-      "grad_norm": 10.616310133905728,
-      "learning_rate": 8.264462809917356e-07,
-      "loss": 1.1368,
+      "epoch": 0.003235722375020223,
+      "grad_norm": 3.024001026912918,
+      "learning_rate": 3.2310177705977386e-07,
+      "loss": 1.6386,
       "step": 20
     },
     {
-      "epoch": 0.010356255178127589,
-      "grad_norm": 4.668322782816374,
-      "learning_rate": 1.0330578512396695e-06,
-      "loss": 1.0379,
+      "epoch": 0.004044652968775279,
+      "grad_norm": 3.1122902778444006,
+      "learning_rate": 4.038772213247173e-07,
+      "loss": 1.6687,
       "step": 25
     },
     {
-      "epoch": 0.012427506213753107,
-      "grad_norm": 3.5182741608551535,
-      "learning_rate": 1.2396694214876035e-06,
-      "loss": 0.9937,
+      "epoch": 0.004853583562530335,
+      "grad_norm": 2.8890112766655296,
+      "learning_rate": 4.846526655896607e-07,
+      "loss": 1.6414,
       "step": 30
     },
     {
-      "epoch": 0.014498757249378625,
-      "grad_norm": 3.3411660569699504,
-      "learning_rate": 1.4462809917355372e-06,
-      "loss": 0.9572,
+      "epoch": 0.0056625141562853904,
+      "grad_norm": 2.6523925300683246,
+      "learning_rate": 5.654281098546043e-07,
+      "loss": 1.59,
       "step": 35
     },
     {
-      "epoch": 0.016570008285004142,
-      "grad_norm": 3.037596563067607,
-      "learning_rate": 1.6528925619834712e-06,
-      "loss": 0.9289,
+      "epoch": 0.006471444750040446,
+      "grad_norm": 2.603827366657395,
+      "learning_rate": 6.462035541195477e-07,
+      "loss": 1.6336,
       "step": 40
     },
     {
-      "epoch": 0.018641259320629662,
-      "grad_norm": 3.2528069002854374,
-      "learning_rate": 1.859504132231405e-06,
-      "loss": 0.9214,
+      "epoch": 0.007280375343795502,
+      "grad_norm": 2.577297564219146,
+      "learning_rate": 7.269789983844912e-07,
+      "loss": 1.6402,
       "step": 45
     },
     {
-      "epoch": 0.020712510356255178,
-      "grad_norm": 3.2107014690984803,
-      "learning_rate": 2.066115702479339e-06,
-      "loss": 0.9086,
+      "epoch": 0.008089305937550558,
+      "grad_norm": 2.439515573088333,
+      "learning_rate": 8.077544426494346e-07,
+      "loss": 1.6193,
       "step": 50
     },
     {
-      "epoch": 0.022783761391880698,
-      "grad_norm": 2.9859094433454554,
-      "learning_rate": 2.2727272727272728e-06,
-      "loss": 0.8991,
+      "epoch": 0.008898236531305614,
+      "grad_norm": 2.4333132929087764,
+      "learning_rate": 8.885298869143781e-07,
+      "loss": 1.6122,
       "step": 55
     },
     {
-      "epoch": 0.024855012427506214,
-      "grad_norm": 3.0590584148615627,
-      "learning_rate": 2.479338842975207e-06,
-      "loss": 0.8916,
+      "epoch": 0.00970716712506067,
+      "grad_norm": 2.1113677966611464,
+      "learning_rate": 9.693053311793215e-07,
+      "loss": 1.6185,
       "step": 60
     },
     {
-      "epoch": 0.026926263463131733,
-      "grad_norm": 2.9914820848818446,
-      "learning_rate": 2.6859504132231405e-06,
-      "loss": 0.8785,
+      "epoch": 0.010516097718815726,
+      "grad_norm": 1.8886265008348362,
+      "learning_rate": 1.0500807754442651e-06,
+      "loss": 1.5725,
       "step": 65
     },
     {
-      "epoch": 0.02899751449875725,
-      "grad_norm": 2.9775144313765445,
-      "learning_rate": 2.8925619834710743e-06,
-      "loss": 0.8897,
+      "epoch": 0.011325028312570781,
+      "grad_norm": 1.8864578193982546,
+      "learning_rate": 1.1308562197092086e-06,
+      "loss": 1.5734,
       "step": 70
     },
     {
-      "epoch": 0.03106876553438277,
-      "grad_norm": 3.128569537433572,
-      "learning_rate": 3.0991735537190086e-06,
-      "loss": 0.8834,
+      "epoch": 0.012133958906325838,
+      "grad_norm": 1.6120301708706668,
+      "learning_rate": 1.211631663974152e-06,
+      "loss": 1.544,
       "step": 75
     },
     {
-      "epoch": 0.033140016570008285,
-      "grad_norm": 3.1281090024503757,
-      "learning_rate": 3.3057851239669424e-06,
-      "loss": 0.8651,
+      "epoch": 0.012942889500080892,
+      "grad_norm": 1.5440235686664168,
+      "learning_rate": 1.2924071082390954e-06,
+      "loss": 1.5381,
       "step": 80
     },
     {
-      "epoch": 0.035211267605633804,
-      "grad_norm": 3.0548507580726625,
-      "learning_rate": 3.5123966942148763e-06,
-      "loss": 0.8737,
+      "epoch": 0.013751820093835949,
+      "grad_norm": 1.3513181383313533,
+      "learning_rate": 1.3731825525040387e-06,
+      "loss": 1.5246,
       "step": 85
     },
     {
-      "epoch": 0.037282518641259324,
-      "grad_norm": 3.088551853774246,
-      "learning_rate": 3.71900826446281e-06,
-      "loss": 0.8623,
+      "epoch": 0.014560750687591004,
+      "grad_norm": 1.4327471253077833,
+      "learning_rate": 1.4539579967689823e-06,
+      "loss": 1.4991,
       "step": 90
     },
     {
-      "epoch": 0.03935376967688484,
-      "grad_norm": 3.117570527250618,
-      "learning_rate": 3.925619834710744e-06,
-      "loss": 0.838,
+      "epoch": 0.01536968128134606,
+      "grad_norm": 1.2301698886041794,
+      "learning_rate": 1.5347334410339258e-06,
+      "loss": 1.4618,
       "step": 95
     },
     {
-      "epoch": 0.041425020712510356,
-      "grad_norm": 3.248544738727649,
-      "learning_rate": 4.132231404958678e-06,
-      "loss": 0.8675,
+      "epoch": 0.016178611875101116,
+      "grad_norm": 1.162576921517532,
+      "learning_rate": 1.6155088852988692e-06,
+      "loss": 1.4728,
       "step": 100
     },
     {
-      "epoch": 0.043496271748135876,
-      "grad_norm": 3.096171447698008,
-      "learning_rate": 4.338842975206612e-06,
-      "loss": 0.8394,
+      "epoch": 0.01698754246885617,
+      "grad_norm": 1.0168699965353747,
+      "learning_rate": 1.6962843295638126e-06,
+      "loss": 1.46,
       "step": 105
     },
     {
-      "epoch": 0.045567522783761395,
-      "grad_norm": 3.233248587377124,
-      "learning_rate": 4.5454545454545455e-06,
-      "loss": 0.857,
+      "epoch": 0.01779647306261123,
+      "grad_norm": 1.0395803267996115,
+      "learning_rate": 1.7770597738287563e-06,
+      "loss": 1.4594,
       "step": 110
     },
     {
-      "epoch": 0.04763877381938691,
-      "grad_norm": 3.6001636024321293,
-      "learning_rate": 4.75206611570248e-06,
-      "loss": 0.8321,
+      "epoch": 0.018605403656366284,
+      "grad_norm": 0.9917353590665066,
+      "learning_rate": 1.8578352180936995e-06,
+      "loss": 1.4397,
       "step": 115
     },
     {
-      "epoch": 0.04971002485501243,
-      "grad_norm": 2.9956895468112084,
-      "learning_rate": 4.958677685950414e-06,
-      "loss": 0.833,
+      "epoch": 0.01941433425012134,
+      "grad_norm": 0.9523144102295756,
+      "learning_rate": 1.938610662358643e-06,
+      "loss": 1.4234,
       "step": 120
     },
     {
-      "epoch": 0.05178127589063795,
-      "grad_norm": 3.2710428266169953,
-      "learning_rate": 5.165289256198347e-06,
-      "loss": 0.847,
+      "epoch": 0.020223264843876397,
+      "grad_norm": 0.9258023739630329,
+      "learning_rate": 2.0193861066235864e-06,
+      "loss": 1.4226,
       "step": 125
     },
     {
-      "epoch": 0.053852526926263466,
-      "grad_norm": 3.249727950554207,
-      "learning_rate": 5.371900826446281e-06,
-      "loss": 0.8491,
+      "epoch": 0.021032195437631452,
+      "grad_norm": 0.9483758496421177,
+      "learning_rate": 2.1001615508885302e-06,
+      "loss": 1.3762,
       "step": 130
     },
     {
-      "epoch": 0.05592377796188898,
-      "grad_norm": 3.638097729325915,
-      "learning_rate": 5.578512396694216e-06,
-      "loss": 0.8218,
+      "epoch": 0.021841126031386507,
+      "grad_norm": 0.9530895080844857,
+      "learning_rate": 2.1809369951534733e-06,
+      "loss": 1.354,
       "step": 135
     },
     {
-      "epoch": 0.0579950289975145,
-      "grad_norm": 3.136515122305162,
-      "learning_rate": 5.785123966942149e-06,
-      "loss": 0.8367,
+      "epoch": 0.022650056625141562,
+      "grad_norm": 0.9891655010686481,
+      "learning_rate": 2.261712439418417e-06,
+      "loss": 1.3418,
       "step": 140
     },
     {
-      "epoch": 0.06006628003314002,
-      "grad_norm": 2.992975297054502,
-      "learning_rate": 5.991735537190083e-06,
-      "loss": 0.8331,
+      "epoch": 0.02345898721889662,
+      "grad_norm": 1.0770341521239468,
+      "learning_rate": 2.34248788368336e-06,
+      "loss": 1.3239,
       "step": 145
     },
     {
-      "epoch": 0.06213753106876554,
-      "grad_norm": 3.1773736798479653,
-      "learning_rate": 6.198347107438017e-06,
-      "loss": 0.8159,
+      "epoch": 0.024267917812651675,
+      "grad_norm": 1.1321453887674335,
+      "learning_rate": 2.423263327948304e-06,
+      "loss": 1.2818,
       "step": 150
     },
     {
-      "epoch": 0.06420878210439106,
-      "grad_norm": 3.2512013569633,
-      "learning_rate": 6.404958677685951e-06,
-      "loss": 0.8261,
+      "epoch": 0.02507684840640673,
+      "grad_norm": 1.116259450785514,
+      "learning_rate": 2.5040387722132474e-06,
+      "loss": 1.2499,
       "step": 155
     },
     {
-      "epoch": 0.06628003314001657,
-      "grad_norm": 3.0823463887606852,
-      "learning_rate": 6.611570247933885e-06,
-      "loss": 0.8376,
+      "epoch": 0.025885779000161785,
+      "grad_norm": 1.1299351548572238,
+      "learning_rate": 2.584814216478191e-06,
+      "loss": 1.2403,
       "step": 160
     },
     {
-      "epoch": 0.06835128417564208,
-      "grad_norm": 3.1208323742016897,
-      "learning_rate": 6.818181818181818e-06,
-      "loss": 0.8275,
+      "epoch": 0.026694709593916843,
+      "grad_norm": 1.1335363316132652,
+      "learning_rate": 2.6655896607431343e-06,
+      "loss": 1.214,
       "step": 165
     },
     {
-      "epoch": 0.07042253521126761,
-      "grad_norm": 3.0745780302802332,
-      "learning_rate": 7.0247933884297525e-06,
-      "loss": 0.8249,
+      "epoch": 0.027503640187671898,
+      "grad_norm": 1.1757124749289842,
+      "learning_rate": 2.7463651050080773e-06,
+      "loss": 1.163,
       "step": 170
     },
     {
-      "epoch": 0.07249378624689312,
-      "grad_norm": 3.119480709134969,
-      "learning_rate": 7.231404958677687e-06,
-      "loss": 0.8042,
+      "epoch": 0.028312570781426953,
+      "grad_norm": 1.2651810945082007,
+      "learning_rate": 2.827140549273021e-06,
+      "loss": 1.1479,
       "step": 175
     },
     {
-      "epoch": 0.07456503728251865,
-      "grad_norm": 3.3248978246402663,
-      "learning_rate": 7.43801652892562e-06,
-      "loss": 0.8284,
+      "epoch": 0.029121501375182008,
+      "grad_norm": 1.2624388452227249,
+      "learning_rate": 2.9079159935379646e-06,
+      "loss": 1.0948,
       "step": 180
     },
     {
-      "epoch": 0.07663628831814416,
-      "grad_norm": 3.101088437755532,
-      "learning_rate": 7.644628099173555e-06,
-      "loss": 0.8158,
+      "epoch": 0.029930431968937066,
+      "grad_norm": 1.2111655416186609,
+      "learning_rate": 2.988691437802908e-06,
+      "loss": 1.0487,
       "step": 185
     },
     {
-      "epoch": 0.07870753935376967,
-      "grad_norm": 3.3761252638628707,
-      "learning_rate": 7.851239669421489e-06,
-      "loss": 0.8082,
+      "epoch": 0.03073936256269212,
+      "grad_norm": 0.8837452655216318,
+      "learning_rate": 3.0694668820678515e-06,
+      "loss": 1.0465,
       "step": 190
     },
     {
-      "epoch": 0.0807787903893952,
-      "grad_norm": 3.211138010347251,
-      "learning_rate": 8.057851239669421e-06,
-      "loss": 0.8066,
+      "epoch": 0.03154829315644718,
+      "grad_norm": 0.8221498455856253,
+      "learning_rate": 3.1502423263327954e-06,
+      "loss": 1.0297,
       "step": 195
     },
     {
-      "epoch": 0.08285004142502071,
-      "grad_norm": 3.390864723290171,
-      "learning_rate": 8.264462809917356e-06,
-      "loss": 0.8071,
+      "epoch": 0.03235722375020223,
+      "grad_norm": 0.8439527908802734,
+      "learning_rate": 3.2310177705977384e-06,
+      "loss": 1.0015,
       "step": 200
     },
     {
-      "epoch": 0.08492129246064622,
-      "grad_norm": 2.9884296341537775,
-      "learning_rate": 8.47107438016529e-06,
-      "loss": 0.8173,
+      "epoch": 0.03316615434395729,
+      "grad_norm": 0.7410910924751422,
+      "learning_rate": 3.311793214862682e-06,
+      "loss": 1.0173,
       "step": 205
     },
     {
-      "epoch": 0.08699254349627175,
-      "grad_norm": 2.9929265463289774,
-      "learning_rate": 8.677685950413224e-06,
-      "loss": 0.8139,
+      "epoch": 0.03397508493771234,
+      "grad_norm": 0.6701077230967789,
+      "learning_rate": 3.3925686591276253e-06,
+      "loss": 0.999,
       "step": 210
     },
     {
-      "epoch": 0.08906379453189726,
-      "grad_norm": 3.178365772310986,
-      "learning_rate": 8.884297520661158e-06,
-      "loss": 0.8077,
+      "epoch": 0.0347840155314674,
+      "grad_norm": 0.6355481671837402,
+      "learning_rate": 3.473344103392569e-06,
+      "loss": 0.9725,
       "step": 215
     },
     {
-      "epoch": 0.09113504556752279,
-      "grad_norm": 3.0990381736081556,
-      "learning_rate": 9.090909090909091e-06,
-      "loss": 0.7962,
+      "epoch": 0.03559294612522246,
+      "grad_norm": 0.5875107208972431,
+      "learning_rate": 3.5541195476575126e-06,
+      "loss": 0.9867,
       "step": 220
     },
     {
-      "epoch": 0.0932062966031483,
-      "grad_norm": 3.3389255178168322,
-      "learning_rate": 9.297520661157025e-06,
-      "loss": 0.7985,
+      "epoch": 0.03640187671897751,
+      "grad_norm": 0.5994183367628626,
+      "learning_rate": 3.6348949919224556e-06,
+      "loss": 0.9943,
       "step": 225
     },
     {
-      "epoch": 0.09527754763877382,
-      "grad_norm": 3.3752236468584456,
-      "learning_rate": 9.50413223140496e-06,
-      "loss": 0.796,
+      "epoch": 0.03721080731273257,
+      "grad_norm": 0.6182303752035966,
+      "learning_rate": 3.715670436187399e-06,
+      "loss": 0.9985,
       "step": 230
     },
     {
-      "epoch": 0.09734879867439934,
-      "grad_norm": 3.0456999521377877,
-      "learning_rate": 9.710743801652894e-06,
-      "loss": 0.7983,
+      "epoch": 0.038019737906487626,
+      "grad_norm": 0.5701097398554537,
+      "learning_rate": 3.796445880452343e-06,
+      "loss": 0.9736,
       "step": 235
     },
     {
-      "epoch": 0.09942004971002485,
-      "grad_norm": 3.2120044737399187,
-      "learning_rate": 9.917355371900828e-06,
-      "loss": 0.8055,
+      "epoch": 0.03882866850024268,
+      "grad_norm": 0.5584494985009502,
+      "learning_rate": 3.877221324717286e-06,
+      "loss": 0.9875,
       "step": 240
     },
     {
-      "epoch": 0.10149130074565037,
-      "grad_norm": 3.132342089203479,
-      "learning_rate": 9.999952928077044e-06,
-      "loss": 0.8021,
+      "epoch": 0.039637599093997736,
+      "grad_norm": 0.5782425116149067,
+      "learning_rate": 3.95799676898223e-06,
+      "loss": 0.9701,
       "step": 245
     },
     {
-      "epoch": 0.1035625517812759,
-      "grad_norm": 2.9894402985924766,
-      "learning_rate": 9.999665269535307e-06,
-      "loss": 0.8013,
+      "epoch": 0.040446529687752794,
+      "grad_norm": 0.6299578118181779,
+      "learning_rate": 4.038772213247173e-06,
+      "loss": 0.9671,
       "step": 250
     },
     {
-      "epoch": 0.1056338028169014,
-      "grad_norm": 3.046903074957559,
-      "learning_rate": 9.99911611854702e-06,
-      "loss": 0.7979,
+      "epoch": 0.041255460281507846,
+      "grad_norm": 0.5812417163133361,
+      "learning_rate": 4.119547657512117e-06,
+      "loss": 0.9579,
       "step": 255
     },
     {
-      "epoch": 0.10770505385252693,
-      "grad_norm": 2.8665185067493626,
-      "learning_rate": 9.998305503833872e-06,
-      "loss": 0.8073,
+      "epoch": 0.042064390875262904,
+      "grad_norm": 0.5781954014931824,
+      "learning_rate": 4.2003231017770605e-06,
+      "loss": 0.9685,
       "step": 260
     },
     {
-      "epoch": 0.10977630488815245,
-      "grad_norm": 3.2409413642928278,
-      "learning_rate": 9.997233467792626e-06,
-      "loss": 0.7976,
+      "epoch": 0.042873321469017955,
+      "grad_norm": 0.608052455969015,
+      "learning_rate": 4.2810985460420035e-06,
+      "loss": 0.9701,
       "step": 265
     },
     {
-      "epoch": 0.11184755592377796,
-      "grad_norm": 2.944077687828054,
-      "learning_rate": 9.995900066492902e-06,
-      "loss": 0.8016,
+      "epoch": 0.043682252062773014,
+      "grad_norm": 0.5966466875482974,
+      "learning_rate": 4.3618739903069465e-06,
+      "loss": 0.9702,
       "step": 270
     },
     {
-      "epoch": 0.11391880695940348,
-      "grad_norm": 2.9912116980156145,
-      "learning_rate": 9.994305369674242e-06,
-      "loss": 0.7853,
+      "epoch": 0.04449118265652807,
+      "grad_norm": 0.5596371503375434,
+      "learning_rate": 4.44264943457189e-06,
+      "loss": 0.9563,
       "step": 275
     },
     {
-      "epoch": 0.115990057995029,
-      "grad_norm": 2.892932100424063,
-      "learning_rate": 9.992449460742464e-06,
-      "loss": 0.8046,
+      "epoch": 0.045300113250283124,
+      "grad_norm": 0.5994461579702871,
+      "learning_rate": 4.523424878836834e-06,
+      "loss": 0.9578,
       "step": 280
     },
     {
-      "epoch": 0.11806130903065451,
-      "grad_norm": 3.055501054823764,
-      "learning_rate": 9.9903324367653e-06,
-      "loss": 0.7794,
+      "epoch": 0.04610904384403818,
+      "grad_norm": 0.5595808732955236,
+      "learning_rate": 4.604200323101777e-06,
+      "loss": 0.9607,
       "step": 285
     },
     {
-      "epoch": 0.12013256006628004,
-      "grad_norm": 2.7659846238569052,
-      "learning_rate": 9.98795440846732e-06,
-      "loss": 0.7814,
+      "epoch": 0.04691797443779324,
+      "grad_norm": 0.669537499667742,
+      "learning_rate": 4.68497576736672e-06,
+      "loss": 0.9476,
       "step": 290
     },
     {
-      "epoch": 0.12220381110190555,
-      "grad_norm": 3.0121040906813126,
-      "learning_rate": 9.985315500224135e-06,
-      "loss": 0.7809,
+      "epoch": 0.04772690503154829,
+      "grad_norm": 0.5511679026608909,
+      "learning_rate": 4.765751211631664e-06,
+      "loss": 0.9464,
       "step": 295
     },
     {
-      "epoch": 0.12427506213753108,
-      "grad_norm": 3.01982583757692,
-      "learning_rate": 9.982415850055902e-06,
-      "loss": 0.781,
+      "epoch": 0.04853583562530335,
+      "grad_norm": 0.5939912843219857,
+      "learning_rate": 4.846526655896608e-06,
+      "loss": 0.9429,
       "step": 300
     },
     {
-      "epoch": 0.1263463131731566,
-      "grad_norm": 2.716394804747616,
-      "learning_rate": 9.979255609620095e-06,
-      "loss": 0.7734,
+      "epoch": 0.0493447662190584,
+      "grad_norm": 0.5798565127301802,
+      "learning_rate": 4.927302100161551e-06,
+      "loss": 0.9485,
       "step": 305
     },
     {
-      "epoch": 0.12841756420878211,
-      "grad_norm": 2.690802383393494,
-      "learning_rate": 9.975834944203581e-06,
-      "loss": 0.7667,
+      "epoch": 0.05015369681281346,
+      "grad_norm": 0.5756338877596261,
+      "learning_rate": 5.008077544426495e-06,
+      "loss": 0.9572,
       "step": 310
     },
     {
-      "epoch": 0.1304888152444076,
-      "grad_norm": 2.9488089043416124,
-      "learning_rate": 9.972154032713973e-06,
-      "loss": 0.7805,
+      "epoch": 0.05096262740656852,
+      "grad_norm": 0.5842007801259245,
+      "learning_rate": 5.088852988691439e-06,
+      "loss": 0.9219,
       "step": 315
     },
     {
-      "epoch": 0.13256006628003314,
-      "grad_norm": 2.719266979190075,
-      "learning_rate": 9.968213067670265e-06,
-      "loss": 0.7763,
+      "epoch": 0.05177155800032357,
+      "grad_norm": 0.5909234064830663,
+      "learning_rate": 5.169628432956382e-06,
+      "loss": 0.9463,
       "step": 320
     },
     {
-      "epoch": 0.13463131731565867,
-      "grad_norm": 2.812329679333236,
-      "learning_rate": 9.964012255192776e-06,
-      "loss": 0.7533,
+      "epoch": 0.05258048859407863,
+      "grad_norm": 0.5964209125322439,
+      "learning_rate": 5.250403877221325e-06,
+      "loss": 0.9137,
       "step": 325
     },
     {
-      "epoch": 0.13670256835128416,
-      "grad_norm": 2.939179450726379,
-      "learning_rate": 9.959551814992364e-06,
-      "loss": 0.7538,
+      "epoch": 0.05338941918783369,
+      "grad_norm": 0.6196287292433933,
+      "learning_rate": 5.331179321486269e-06,
+      "loss": 0.9275,
       "step": 330
     },
     {
-      "epoch": 0.1387738193869097,
-      "grad_norm": 3.0343927565408566,
-      "learning_rate": 9.954831980358928e-06,
-      "loss": 0.7761,
+      "epoch": 0.05419834978158874,
+      "grad_norm": 0.6885711787979655,
+      "learning_rate": 5.411954765751212e-06,
+      "loss": 0.9308,
       "step": 335
     },
     {
-      "epoch": 0.14084507042253522,
-      "grad_norm": 3.051369274408743,
-      "learning_rate": 9.949852998149217e-06,
-      "loss": 0.7623,
+      "epoch": 0.055007280375343796,
+      "grad_norm": 0.5882034901608182,
+      "learning_rate": 5.492730210016155e-06,
+      "loss": 0.9226,
       "step": 340
     },
     {
-      "epoch": 0.14291632145816072,
-      "grad_norm": 2.8926023239726666,
-      "learning_rate": 9.944615128773911e-06,
-      "loss": 0.7449,
+      "epoch": 0.055816210969098855,
+      "grad_norm": 0.6167809439808443,
+      "learning_rate": 5.573505654281099e-06,
+      "loss": 0.9329,
       "step": 345
     },
     {
-      "epoch": 0.14498757249378624,
-      "grad_norm": 3.0274876780182725,
-      "learning_rate": 9.939118646184007e-06,
-      "loss": 0.7564,
+      "epoch": 0.056625141562853906,
+      "grad_norm": 0.6473624549151995,
+      "learning_rate": 5.654281098546042e-06,
+      "loss": 0.9342,
       "step": 350
     },
     {
-      "epoch": 0.14705882352941177,
-      "grad_norm": 3.030086237395326,
-      "learning_rate": 9.933363837856485e-06,
-      "loss": 0.7636,
+      "epoch": 0.057434072156608965,
+      "grad_norm": 0.6268324107561812,
+      "learning_rate": 5.735056542810986e-06,
+      "loss": 0.9444,
       "step": 355
     },
     {
-      "epoch": 0.1491300745650373,
-      "grad_norm": 2.7216718925144443,
-      "learning_rate": 9.927351004779275e-06,
-      "loss": 0.7523,
+      "epoch": 0.058243002750364016,
+      "grad_norm": 0.6081881922413113,
+      "learning_rate": 5.815831987075929e-06,
+      "loss": 0.9266,
       "step": 360
     },
     {
-      "epoch": 0.1512013256006628,
-      "grad_norm": 2.8963011471510423,
-      "learning_rate": 9.921080461435522e-06,
-      "loss": 0.747,
+      "epoch": 0.059051933344119074,
+      "grad_norm": 0.623515938924003,
+      "learning_rate": 5.896607431340873e-06,
+      "loss": 0.927,
       "step": 365
     },
     {
-      "epoch": 0.15327257663628832,
-      "grad_norm": 2.908273407418608,
-      "learning_rate": 9.914552535787122e-06,
-      "loss": 0.7325,
+      "epoch": 0.05986086393787413,
+      "grad_norm": 0.595326688964164,
+      "learning_rate": 5.977382875605816e-06,
+      "loss": 0.941,
       "step": 370
     },
     {
-      "epoch": 0.15534382767191385,
-      "grad_norm": 2.8676437077981722,
-      "learning_rate": 9.90776756925758e-06,
-      "loss": 0.7349,
+      "epoch": 0.060669794531629184,
+      "grad_norm": 0.7035346939039838,
+      "learning_rate": 6.058158319870759e-06,
+      "loss": 0.9243,
       "step": 375
     },
     {
-      "epoch": 0.15741507870753935,
-      "grad_norm": 2.8556939586506687,
-      "learning_rate": 9.900725916714157e-06,
-      "loss": 0.7524,
+      "epoch": 0.06147872512538424,
+      "grad_norm": 0.6486917875962197,
+      "learning_rate": 6.138933764135703e-06,
+      "loss": 0.945,
       "step": 380
     },
     {
-      "epoch": 0.15948632974316487,
-      "grad_norm": 2.87186116366176,
-      "learning_rate": 9.893427946449297e-06,
-      "loss": 0.7214,
+      "epoch": 0.0622876557191393,
+      "grad_norm": 0.6272725751694639,
+      "learning_rate": 6.219709208400647e-06,
+      "loss": 0.9333,
       "step": 385
     },
     {
-      "epoch": 0.1615575807787904,
-      "grad_norm": 2.8757200745024485,
-      "learning_rate": 9.885874040161373e-06,
-      "loss": 0.7326,
+      "epoch": 0.06309658631289436,
+      "grad_norm": 0.6847897267332435,
+      "learning_rate": 6.300484652665591e-06,
+      "loss": 0.9201,
       "step": 390
     },
     {
-      "epoch": 0.1636288318144159,
-      "grad_norm": 3.215961320189587,
-      "learning_rate": 9.878064592934723e-06,
-      "loss": 0.7227,
+      "epoch": 0.0639055169066494,
+      "grad_norm": 0.6063331771589294,
+      "learning_rate": 6.381260096930534e-06,
+      "loss": 0.9301,
       "step": 395
     },
     {
-      "epoch": 0.16570008285004142,
-      "grad_norm": 2.9059437980483693,
-      "learning_rate": 9.87000001321898e-06,
-      "loss": 0.7509,
+      "epoch": 0.06471444750040446,
+      "grad_norm": 0.6535104175430165,
+      "learning_rate": 6.462035541195477e-06,
+      "loss": 0.9268,
       "step": 400
     },
     {
-      "epoch": 0.16777133388566695,
-      "grad_norm": 2.685626883478994,
-      "learning_rate": 9.86168072280772e-06,
-      "loss": 0.7223,
+      "epoch": 0.06552337809415952,
+      "grad_norm": 0.6298942805679355,
+      "learning_rate": 6.542810985460421e-06,
+      "loss": 0.9126,
       "step": 405
     },
     {
-      "epoch": 0.16984258492129245,
-      "grad_norm": 3.033273475512452,
-      "learning_rate": 9.853107156816393e-06,
-      "loss": 0.7385,
+      "epoch": 0.06633230868791458,
+      "grad_norm": 0.687088041257841,
+      "learning_rate": 6.623586429725364e-06,
+      "loss": 0.8997,
       "step": 410
     },
     {
-      "epoch": 0.17191383595691798,
-      "grad_norm": 2.7784947677053955,
-      "learning_rate": 9.844279763659566e-06,
-      "loss": 0.7237,
+      "epoch": 0.06714123928166964,
+      "grad_norm": 0.6809022568244187,
+      "learning_rate": 6.7043618739903075e-06,
+      "loss": 0.9221,
       "step": 415
     },
     {
-      "epoch": 0.1739850869925435,
-      "grad_norm": 2.9117670580445543,
-      "learning_rate": 9.835199005027477e-06,
-      "loss": 0.7161,
+      "epoch": 0.06795016987542468,
+      "grad_norm": 0.6595953972527364,
+      "learning_rate": 6.7851373182552505e-06,
+      "loss": 0.9351,
       "step": 420
     },
     {
-      "epoch": 0.176056338028169,
-      "grad_norm": 2.8843325957862764,
-      "learning_rate": 9.825865355861878e-06,
-      "loss": 0.7256,
+      "epoch": 0.06875910046917974,
+      "grad_norm": 0.6526975586915988,
+      "learning_rate": 6.865912762520195e-06,
+      "loss": 0.9046,
       "step": 425
     },
     {
-      "epoch": 0.17812758906379453,
-      "grad_norm": 2.856399843872642,
-      "learning_rate": 9.816279304331202e-06,
-      "loss": 0.7142,
+      "epoch": 0.0695680310629348,
+      "grad_norm": 0.6747286601047916,
+      "learning_rate": 6.946688206785138e-06,
+      "loss": 0.9384,
       "step": 430
     },
     {
-      "epoch": 0.18019884009942005,
-      "grad_norm": 2.9556721265181163,
-      "learning_rate": 9.806441351805025e-06,
-      "loss": 0.7306,
+      "epoch": 0.07037696165668986,
+      "grad_norm": 0.705527948534626,
+      "learning_rate": 7.027463651050081e-06,
+      "loss": 0.9171,
       "step": 435
     },
     {
-      "epoch": 0.18227009113504558,
-      "grad_norm": 2.9810648069102172,
-      "learning_rate": 9.79635201282785e-06,
-      "loss": 0.705,
+      "epoch": 0.07118589225044492,
+      "grad_norm": 0.6636402242054834,
+      "learning_rate": 7.108239095315025e-06,
+      "loss": 0.9311,
       "step": 440
     },
     {
-      "epoch": 0.18434134217067108,
-      "grad_norm": 2.650650945302262,
-      "learning_rate": 9.786011815092193e-06,
-      "loss": 0.7026,
+      "epoch": 0.07199482284419997,
+      "grad_norm": 0.6340851719954015,
+      "learning_rate": 7.189014539579968e-06,
+      "loss": 0.9194,
       "step": 445
     },
     {
-      "epoch": 0.1864125932062966,
-      "grad_norm": 2.7614731795144727,
-      "learning_rate": 9.775421299410977e-06,
-      "loss": 0.6993,
+      "epoch": 0.07280375343795502,
+      "grad_norm": 0.7041525654790047,
+      "learning_rate": 7.269789983844911e-06,
+      "loss": 0.906,
       "step": 450
     },
     {
-      "epoch": 0.18848384424192213,
-      "grad_norm": 2.8508808311574487,
-      "learning_rate": 9.764581019689255e-06,
-      "loss": 0.707,
+      "epoch": 0.07361268403171008,
+      "grad_norm": 1.0272938347497549,
+      "learning_rate": 7.350565428109855e-06,
+      "loss": 0.9156,
       "step": 455
     },
     {
-      "epoch": 0.19055509527754763,
-      "grad_norm": 3.1150961713118646,
-      "learning_rate": 9.753491542895237e-06,
-      "loss": 0.7186,
+      "epoch": 0.07442161462546514,
+      "grad_norm": 0.7023049681523049,
+      "learning_rate": 7.431340872374798e-06,
+      "loss": 0.8947,
       "step": 460
     },
     {
-      "epoch": 0.19262634631317316,
-      "grad_norm": 3.0323106493041325,
-      "learning_rate": 9.742153449030639e-06,
-      "loss": 0.6893,
+      "epoch": 0.0752305452192202,
+      "grad_norm": 0.6957484415499231,
+      "learning_rate": 7.512116316639743e-06,
+      "loss": 0.9088,
       "step": 465
     },
     {
-      "epoch": 0.19469759734879868,
-      "grad_norm": 3.0499193466413654,
-      "learning_rate": 9.730567331100333e-06,
-      "loss": 0.7106,
+      "epoch": 0.07603947581297525,
+      "grad_norm": 0.6699201382717078,
+      "learning_rate": 7.592891760904686e-06,
+      "loss": 0.9086,
       "step": 470
     },
     {
-      "epoch": 0.19676884838442418,
-      "grad_norm": 2.8549221082922,
-      "learning_rate": 9.71873379508136e-06,
-      "loss": 0.6801,
+      "epoch": 0.0768484064067303,
+      "grad_norm": 0.7101796332013423,
+      "learning_rate": 7.673667205169629e-06,
+      "loss": 0.9037,
       "step": 475
     },
     {
-      "epoch": 0.1988400994200497,
-      "grad_norm": 2.8978651489251703,
-      "learning_rate": 9.706653459891207e-06,
-      "loss": 0.6906,
+      "epoch": 0.07765733700048535,
+      "grad_norm": 0.7497394327378185,
+      "learning_rate": 7.754442649434572e-06,
+      "loss": 0.929,
       "step": 480
     },
     {
-      "epoch": 0.20091135045567524,
-      "grad_norm": 2.7131289500779237,
-      "learning_rate": 9.694326957355452e-06,
-      "loss": 0.6793,
+      "epoch": 0.07846626759424041,
+      "grad_norm": 0.7422612177167345,
+      "learning_rate": 7.835218093699516e-06,
+      "loss": 0.8977,
       "step": 485
     },
     {
-      "epoch": 0.20298260149130073,
-      "grad_norm": 2.7450048929619033,
-      "learning_rate": 9.681754932174719e-06,
-      "loss": 0.6987,
+      "epoch": 0.07927519818799547,
+      "grad_norm": 0.7103824932867557,
+      "learning_rate": 7.91599353796446e-06,
+      "loss": 0.9151,
       "step": 490
     },
     {
-      "epoch": 0.20505385252692626,
-      "grad_norm": 2.752895531078258,
-      "learning_rate": 9.668938041890952e-06,
-      "loss": 0.693,
+      "epoch": 0.08008412878175053,
+      "grad_norm": 0.7605669414288259,
+      "learning_rate": 7.996768982229403e-06,
+      "loss": 0.9234,
       "step": 495
     },
     {
-      "epoch": 0.2071251035625518,
-      "grad_norm": 2.808137119286013,
-      "learning_rate": 9.655876956853025e-06,
-      "loss": 0.681,
+      "epoch": 0.08089305937550559,
+      "grad_norm": 0.7765227746326537,
+      "learning_rate": 8.077544426494346e-06,
+      "loss": 0.9317,
       "step": 500
     },
     {
-      "epoch": 0.2091963545981773,
-      "grad_norm": 2.724476444787799,
-      "learning_rate": 9.64257236018169e-06,
-      "loss": 0.6716,
+      "epoch": 0.08170198996926063,
+      "grad_norm": 0.7114224103403521,
+      "learning_rate": 8.15831987075929e-06,
+      "loss": 0.8977,
       "step": 505
     },
     {
-      "epoch": 0.2112676056338028,
-      "grad_norm": 2.748867424602178,
-      "learning_rate": 9.629024947733836e-06,
-      "loss": 0.6944,
+      "epoch": 0.08251092056301569,
+      "grad_norm": 0.709653797696482,
+      "learning_rate": 8.239095315024233e-06,
+      "loss": 0.9048,
       "step": 510
     },
     {
-      "epoch": 0.21333885666942834,
-      "grad_norm": 2.9490226550436676,
-      "learning_rate": 9.615235428066106e-06,
-      "loss": 0.6844,
+      "epoch": 0.08331985115677075,
+      "grad_norm": 0.6694669312530299,
+      "learning_rate": 8.319870759289176e-06,
+      "loss": 0.9079,
       "step": 515
     },
     {
-      "epoch": 0.21541010770505387,
-      "grad_norm": 2.6625828983237634,
-      "learning_rate": 9.601204522397826e-06,
-      "loss": 0.6876,
+      "epoch": 0.08412878175052581,
+      "grad_norm": 0.6918229992578886,
+      "learning_rate": 8.400646203554121e-06,
+      "loss": 0.9161,
       "step": 520
     },
     {
-      "epoch": 0.21748135874067936,
-      "grad_norm": 2.8580713323817672,
-      "learning_rate": 9.586932964573298e-06,
-      "loss": 0.6675,
+      "epoch": 0.08493771234428087,
+      "grad_norm": 0.826690733304416,
+      "learning_rate": 8.481421647819064e-06,
+      "loss": 0.9179,
       "step": 525
     },
     {
-      "epoch": 0.2195526097763049,
-      "grad_norm": 2.65085890829761,
-      "learning_rate": 9.572421501023403e-06,
-      "loss": 0.6871,
+      "epoch": 0.08574664293803591,
+      "grad_norm": 0.8502072042848492,
+      "learning_rate": 8.562197092084007e-06,
+      "loss": 0.8922,
       "step": 530
     },
     {
-      "epoch": 0.22162386081193042,
-      "grad_norm": 2.8073591225727066,
-      "learning_rate": 9.557670890726576e-06,
-      "loss": 0.6806,
+      "epoch": 0.08655557353179097,
+      "grad_norm": 0.7888391565118393,
+      "learning_rate": 8.64297253634895e-06,
+      "loss": 0.9087,
       "step": 535
     },
     {
-      "epoch": 0.22369511184755592,
-      "grad_norm": 2.890518621456693,
-      "learning_rate": 9.5426819051691e-06,
-      "loss": 0.6842,
+      "epoch": 0.08736450412554603,
+      "grad_norm": 0.7612155866818295,
+      "learning_rate": 8.723747980613893e-06,
+      "loss": 0.9067,
       "step": 540
     },
     {
-      "epoch": 0.22576636288318144,
-      "grad_norm": 3.037365740542274,
-      "learning_rate": 9.527455328304756e-06,
-      "loss": 0.6706,
+      "epoch": 0.08817343471930109,
+      "grad_norm": 0.7340496045333647,
+      "learning_rate": 8.804523424878838e-06,
+      "loss": 0.9122,
       "step": 545
     },
     {
-      "epoch": 0.22783761391880697,
-      "grad_norm": 3.007466890807781,
-      "learning_rate": 9.511991956513828e-06,
-      "loss": 0.6768,
+      "epoch": 0.08898236531305614,
+      "grad_norm": 0.6807951345118415,
+      "learning_rate": 8.88529886914378e-06,
+      "loss": 0.9209,
       "step": 550
     },
     {
-      "epoch": 0.22990886495443247,
-      "grad_norm": 2.991724717996849,
-      "learning_rate": 9.496292598561445e-06,
-      "loss": 0.6399,
+      "epoch": 0.08979129590681119,
+      "grad_norm": 0.7182636447883889,
+      "learning_rate": 8.966074313408725e-06,
+      "loss": 0.9036,
       "step": 555
     },
     {
-      "epoch": 0.231980115990058,
-      "grad_norm": 2.6643291004806144,
-      "learning_rate": 9.480358075555278e-06,
-      "loss": 0.6745,
+      "epoch": 0.09060022650056625,
+      "grad_norm": 0.6906380745424776,
+      "learning_rate": 9.046849757673668e-06,
+      "loss": 0.8775,
       "step": 560
     },
     {
-      "epoch": 0.23405136702568352,
-      "grad_norm": 2.7494300595409076,
-      "learning_rate": 9.464189220902603e-06,
-      "loss": 0.6659,
+      "epoch": 0.0914091570943213,
+      "grad_norm": 0.7205121865769839,
+      "learning_rate": 9.127625201938612e-06,
+      "loss": 0.8799,
       "step": 565
     },
     {
-      "epoch": 0.23612261806130902,
-      "grad_norm": 2.778050795482066,
-      "learning_rate": 9.447786880266706e-06,
-      "loss": 0.6682,
+      "epoch": 0.09221808768807636,
+      "grad_norm": 0.699683431919587,
+      "learning_rate": 9.208400646203555e-06,
+      "loss": 0.8776,
       "step": 570
     },
     {
-      "epoch": 0.23819386909693455,
-      "grad_norm": 2.8808954872434107,
-      "learning_rate": 9.431151911522656e-06,
-      "loss": 0.6603,
+      "epoch": 0.09302701828183142,
+      "grad_norm": 0.7092973559328101,
+      "learning_rate": 9.289176090468498e-06,
+      "loss": 0.8976,
       "step": 575
     },
     {
-      "epoch": 0.24026512013256007,
-      "grad_norm": 2.7631310193416008,
-      "learning_rate": 9.414285184712432e-06,
-      "loss": 0.6479,
+      "epoch": 0.09383594887558648,
+      "grad_norm": 0.7013577925584726,
+      "learning_rate": 9.36995153473344e-06,
+      "loss": 0.8915,
       "step": 580
     },
     {
-      "epoch": 0.24233637116818557,
-      "grad_norm": 2.7504091304269505,
-      "learning_rate": 9.397187581999424e-06,
-      "loss": 0.656,
+      "epoch": 0.09464487946934153,
+      "grad_norm": 0.7360966395662988,
+      "learning_rate": 9.450726978998385e-06,
+      "loss": 0.9049,
       "step": 585
     },
     {
-      "epoch": 0.2444076222038111,
-      "grad_norm": 2.7704909990609345,
-      "learning_rate": 9.37985999762229e-06,
-      "loss": 0.6494,
+      "epoch": 0.09545381006309658,
+      "grad_norm": 0.7793159637124253,
+      "learning_rate": 9.531502423263328e-06,
+      "loss": 0.8924,
       "step": 590
     },
     {
-      "epoch": 0.24647887323943662,
-      "grad_norm": 2.7880341771333086,
-      "learning_rate": 9.362303337848188e-06,
-      "loss": 0.6498,
+      "epoch": 0.09626274065685164,
+      "grad_norm": 0.7373798826049844,
+      "learning_rate": 9.612277867528273e-06,
+      "loss": 0.9125,
       "step": 595
     },
     {
-      "epoch": 0.24855012427506215,
-      "grad_norm": 2.7372277502845064,
-      "learning_rate": 9.344518520925377e-06,
-      "loss": 0.635,
+      "epoch": 0.0970716712506067,
+      "grad_norm": 0.7155304623776628,
+      "learning_rate": 9.693053311793216e-06,
+      "loss": 0.8922,
       "step": 600
     },
     {
-      "epoch": 0.2506213753106877,
-      "grad_norm": 2.700220180234652,
-      "learning_rate": 9.326506477035179e-06,
-      "loss": 0.6205,
+      "epoch": 0.09788060184436176,
+      "grad_norm": 0.722457498504299,
+      "learning_rate": 9.773828756058159e-06,
+      "loss": 0.9133,
       "step": 605
     },
     {
-      "epoch": 0.2526926263463132,
-      "grad_norm": 2.9446333047323714,
-      "learning_rate": 9.308268148243355e-06,
-      "loss": 0.6377,
+      "epoch": 0.0986895324381168,
+      "grad_norm": 0.7006683969526154,
+      "learning_rate": 9.854604200323102e-06,
+      "loss": 0.8794,
       "step": 610
     },
     {
-      "epoch": 0.2547638773819387,
-      "grad_norm": 2.851475449493436,
-      "learning_rate": 9.289804488450805e-06,
-      "loss": 0.6395,
+      "epoch": 0.09949846303187186,
+      "grad_norm": 0.7854584772109673,
+      "learning_rate": 9.935379644588045e-06,
+      "loss": 0.9029,
       "step": 615
     },
     {
-      "epoch": 0.25683512841756423,
-      "grad_norm": 2.868033711564291,
-      "learning_rate": 9.271116463343692e-06,
-      "loss": 0.6421,
+      "epoch": 0.10030739362562692,
+      "grad_norm": 0.7970045944993353,
+      "learning_rate": 9.999999202413539e-06,
+      "loss": 0.9009,
       "step": 620
     },
     {
-      "epoch": 0.25890637945318973,
-      "grad_norm": 2.6514436144580307,
-      "learning_rate": 9.25220505034293e-06,
-      "loss": 0.6372,
+      "epoch": 0.10111632421938198,
+      "grad_norm": 0.8001952905335415,
+      "learning_rate": 9.999971286914108e-06,
+      "loss": 0.8689,
       "step": 625
     },
     {
-      "epoch": 0.2609776304888152,
-      "grad_norm": 2.9338066764643598,
-      "learning_rate": 9.23307123855307e-06,
-      "loss": 0.636,
+      "epoch": 0.10192525481313704,
+      "grad_norm": 0.7947330109524953,
+      "learning_rate": 9.999903492346063e-06,
+      "loss": 0.8836,
       "step": 630
     },
     {
-      "epoch": 0.2630488815244408,
-      "grad_norm": 3.023298105516956,
-      "learning_rate": 9.213716028710558e-06,
-      "loss": 0.6103,
+      "epoch": 0.1027341854068921,
+      "grad_norm": 0.6517743774036119,
+      "learning_rate": 9.999795819250126e-06,
+      "loss": 0.8937,
       "step": 635
     },
     {
-      "epoch": 0.2651201325600663,
-      "grad_norm": 2.871190920425405,
-      "learning_rate": 9.194140433131397e-06,
-      "loss": 0.6159,
+      "epoch": 0.10354311600064714,
+      "grad_norm": 0.7153316955972873,
+      "learning_rate": 9.99964826848508e-06,
+      "loss": 0.8952,
       "step": 640
     },
     {
-      "epoch": 0.2671913835956918,
-      "grad_norm": 2.7631786024911187,
-      "learning_rate": 9.174345475658208e-06,
-      "loss": 0.6228,
+      "epoch": 0.1043520465944022,
+      "grad_norm": 0.7470689793235693,
+      "learning_rate": 9.99946084122777e-06,
+      "loss": 0.905,
       "step": 645
     },
     {
-      "epoch": 0.26926263463131733,
-      "grad_norm": 2.998454912972155,
-      "learning_rate": 9.154332191606671e-06,
-      "loss": 0.6335,
+      "epoch": 0.10516097718815726,
+      "grad_norm": 0.8088466120200852,
+      "learning_rate": 9.99923353897309e-06,
+      "loss": 0.8982,
       "step": 650
     },
     {
-      "epoch": 0.27133388566694283,
-      "grad_norm": 2.803386942098811,
-      "learning_rate": 9.134101627711384e-06,
-      "loss": 0.6119,
+      "epoch": 0.10596990778191231,
+      "grad_norm": 0.6912247019886972,
+      "learning_rate": 9.998966363533972e-06,
+      "loss": 0.8895,
       "step": 655
     },
     {
-      "epoch": 0.27340513670256833,
-      "grad_norm": 2.6219329149002277,
-      "learning_rate": 9.113654842071114e-06,
-      "loss": 0.6171,
+      "epoch": 0.10677883837566737,
+      "grad_norm": 0.7942586835594388,
+      "learning_rate": 9.998659317041367e-06,
+      "loss": 0.8913,
       "step": 660
     },
     {
-      "epoch": 0.2754763877381939,
-      "grad_norm": 2.61515633927928,
-      "learning_rate": 9.092992904093451e-06,
-      "loss": 0.6165,
+      "epoch": 0.10758776896942242,
+      "grad_norm": 0.7841096349195548,
+      "learning_rate": 9.998312401944237e-06,
+      "loss": 0.8774,
       "step": 665
     },
     {
-      "epoch": 0.2775476387738194,
-      "grad_norm": 2.8142275849395473,
-      "learning_rate": 9.072116894438885e-06,
-      "loss": 0.6091,
+      "epoch": 0.10839669956317748,
+      "grad_norm": 0.7485526011092998,
+      "learning_rate": 9.997925621009527e-06,
+      "loss": 0.8955,
       "step": 670
     },
     {
-      "epoch": 0.2796188898094449,
-      "grad_norm": 2.8036261300008856,
-      "learning_rate": 9.051027904964279e-06,
-      "loss": 0.6266,
+      "epoch": 0.10920563015693253,
+      "grad_norm": 0.7271239921076256,
+      "learning_rate": 9.997498977322146e-06,
+      "loss": 0.8871,
       "step": 675
     },
     {
-      "epoch": 0.28169014084507044,
-      "grad_norm": 2.5728545238650122,
-      "learning_rate": 9.029727038665765e-06,
-      "loss": 0.625,
+      "epoch": 0.11001456075068759,
+      "grad_norm": 0.7148944860864371,
+      "learning_rate": 9.997032474284949e-06,
+      "loss": 0.9008,
       "step": 680
     },
     {
-      "epoch": 0.28376139188069593,
-      "grad_norm": 2.911382813259161,
-      "learning_rate": 9.008215409621053e-06,
-      "loss": 0.6023,
+      "epoch": 0.11082349134444265,
+      "grad_norm": 0.7486275146950349,
+      "learning_rate": 9.996526115618694e-06,
+      "loss": 0.8786,
       "step": 685
     },
     {
-      "epoch": 0.28583264291632143,
-      "grad_norm": 2.8794319732821956,
-      "learning_rate": 8.986494142931168e-06,
-      "loss": 0.6113,
+      "epoch": 0.11163242193819771,
+      "grad_norm": 0.7878464787195636,
+      "learning_rate": 9.995979905362028e-06,
+      "loss": 0.8805,
       "step": 690
     },
     {
-      "epoch": 0.287903893951947,
-      "grad_norm": 2.6994578026382956,
-      "learning_rate": 8.964564374661597e-06,
-      "loss": 0.6083,
+      "epoch": 0.11244135253195275,
+      "grad_norm": 0.7844738669203999,
+      "learning_rate": 9.995393847871446e-06,
+      "loss": 0.8768,
       "step": 695
     },
     {
-      "epoch": 0.2899751449875725,
-      "grad_norm": 2.901823826406794,
-      "learning_rate": 8.94242725178288e-06,
-      "loss": 0.616,
+      "epoch": 0.11325028312570781,
+      "grad_norm": 0.7008587939833357,
+      "learning_rate": 9.994767947821261e-06,
+      "loss": 0.8848,
       "step": 700
     },
     {
-      "epoch": 0.292046396023198,
-      "grad_norm": 2.653196617958314,
-      "learning_rate": 8.920083932110608e-06,
-      "loss": 0.6196,
+      "epoch": 0.11405921371946287,
+      "grad_norm": 0.7176705233353148,
+      "learning_rate": 9.994102210203567e-06,
+      "loss": 0.9161,
       "step": 705
     },
     {
-      "epoch": 0.29411764705882354,
-      "grad_norm": 2.6749585402475264,
-      "learning_rate": 8.89753558424488e-06,
-      "loss": 0.597,
+      "epoch": 0.11486814431321793,
+      "grad_norm": 0.7984767869907693,
+      "learning_rate": 9.993396640328191e-06,
+      "loss": 0.8984,
       "step": 710
     },
     {
-      "epoch": 0.29618889809444904,
-      "grad_norm": 2.7369834176016723,
-      "learning_rate": 8.874783387509181e-06,
-      "loss": 0.6012,
+      "epoch": 0.11567707490697299,
+      "grad_norm": 0.7660329081563221,
+      "learning_rate": 9.992651243822658e-06,
+      "loss": 0.8918,
       "step": 715
     },
     {
-      "epoch": 0.2982601491300746,
-      "grad_norm": 2.845681780523314,
-      "learning_rate": 8.851828531888692e-06,
-      "loss": 0.6223,
+      "epoch": 0.11648600550072803,
+      "grad_norm": 0.7692169317757526,
+      "learning_rate": 9.991866026632146e-06,
+      "loss": 0.8845,
       "step": 720
     },
     {
-      "epoch": 0.3003314001657001,
-      "grad_norm": 2.7849231768312883,
-      "learning_rate": 8.828672217968055e-06,
-      "loss": 0.6192,
+      "epoch": 0.11729493609448309,
+      "grad_norm": 0.7728932225344195,
+      "learning_rate": 9.991040995019441e-06,
+      "loss": 0.8944,
       "step": 725
     },
     {
-      "epoch": 0.3024026512013256,
-      "grad_norm": 2.7476355382279385,
-      "learning_rate": 8.805315656868587e-06,
-      "loss": 0.6053,
+      "epoch": 0.11810386668823815,
+      "grad_norm": 0.7481133899054797,
+      "learning_rate": 9.990176155564874e-06,
+      "loss": 0.8782,
       "step": 730
     },
     {
-      "epoch": 0.30447390223695114,
-      "grad_norm": 2.686604793998387,
-      "learning_rate": 8.781760070184933e-06,
-      "loss": 0.5934,
+      "epoch": 0.11891279728199321,
+      "grad_norm": 0.7221002980302099,
+      "learning_rate": 9.989271515166287e-06,
+      "loss": 0.8762,
       "step": 735
     },
     {
-      "epoch": 0.30654515327257664,
-      "grad_norm": 2.997703713470155,
-      "learning_rate": 8.75800668992117e-06,
-      "loss": 0.5956,
+      "epoch": 0.11972172787574827,
+      "grad_norm": 0.7251326764742602,
+      "learning_rate": 9.988327081038962e-06,
+      "loss": 0.865,
       "step": 740
     },
     {
-      "epoch": 0.30861640430820214,
-      "grad_norm": 2.621004060213254,
-      "learning_rate": 8.734056758426367e-06,
-      "loss": 0.5974,
+      "epoch": 0.12053065846950331,
+      "grad_norm": 0.70263798156615,
+      "learning_rate": 9.987342860715575e-06,
+      "loss": 0.8885,
       "step": 745
     },
     {
-      "epoch": 0.3106876553438277,
-      "grad_norm": 2.6549364932267303,
-      "learning_rate": 8.709911528329623e-06,
-      "loss": 0.5948,
+      "epoch": 0.12133958906325837,
+      "grad_norm": 0.7527476269587685,
+      "learning_rate": 9.986318862046129e-06,
+      "loss": 0.9032,
       "step": 750
     },
     {
-      "epoch": 0.3127589063794532,
-      "grad_norm": 2.7965881278503435,
-      "learning_rate": 8.685572262474538e-06,
-      "loss": 0.5967,
+      "epoch": 0.12214851965701343,
+      "grad_norm": 0.7398355773627006,
+      "learning_rate": 9.98525509319789e-06,
+      "loss": 0.8788,
       "step": 755
     },
     {
-      "epoch": 0.3148301574150787,
-      "grad_norm": 2.5906530552932185,
-      "learning_rate": 8.661040233853166e-06,
-      "loss": 0.5824,
+      "epoch": 0.12295745025076849,
+      "grad_norm": 0.7459365223897065,
+      "learning_rate": 9.984151562655333e-06,
+      "loss": 0.8754,
       "step": 760
     },
     {
-      "epoch": 0.31690140845070425,
-      "grad_norm": 2.96203173026875,
-      "learning_rate": 8.636316725539445e-06,
-      "loss": 0.5833,
+      "epoch": 0.12376638084452354,
+      "grad_norm": 0.7806602396452277,
+      "learning_rate": 9.983008279220062e-06,
+      "loss": 0.8716,
       "step": 765
     },
     {
-      "epoch": 0.31897265948632975,
-      "grad_norm": 2.6456839069829967,
-      "learning_rate": 8.611403030622074e-06,
-      "loss": 0.5913,
+      "epoch": 0.1245753114382786,
+      "grad_norm": 0.7479182177455358,
+      "learning_rate": 9.981825252010743e-06,
+      "loss": 0.8849,
       "step": 770
     },
     {
-      "epoch": 0.32104391052195524,
-      "grad_norm": 2.552077771616609,
-      "learning_rate": 8.586300452136895e-06,
-      "loss": 0.5604,
+      "epoch": 0.12538424203203366,
+      "grad_norm": 0.7644325690144114,
+      "learning_rate": 9.980602490463037e-06,
+      "loss": 0.8742,
       "step": 775
     },
     {
-      "epoch": 0.3231151615575808,
-      "grad_norm": 2.681514794026766,
-      "learning_rate": 8.561010302998734e-06,
-      "loss": 0.5878,
+      "epoch": 0.12619317262578872,
+      "grad_norm": 0.8269622902407507,
+      "learning_rate": 9.979340004329516e-06,
+      "loss": 0.8747,
       "step": 780
     },
     {
-      "epoch": 0.3251864125932063,
-      "grad_norm": 2.810277058352099,
-      "learning_rate": 8.535533905932739e-06,
-      "loss": 0.5904,
+      "epoch": 0.12700210321954375,
+      "grad_norm": 0.709710236101346,
+      "learning_rate": 9.978037803679595e-06,
+      "loss": 0.8929,
       "step": 785
     },
     {
-      "epoch": 0.3272576636288318,
-      "grad_norm": 2.823366500468815,
-      "learning_rate": 8.509872593405189e-06,
-      "loss": 0.5785,
+      "epoch": 0.1278110338132988,
+      "grad_norm": 0.7487057568451865,
+      "learning_rate": 9.97669589889944e-06,
+      "loss": 0.8944,
       "step": 790
     },
     {
-      "epoch": 0.32932891466445735,
-      "grad_norm": 2.537302794266039,
-      "learning_rate": 8.484027707553818e-06,
-      "loss": 0.5749,
+      "epoch": 0.12861996440705387,
+      "grad_norm": 0.8332024529186928,
+      "learning_rate": 9.975314300691898e-06,
+      "loss": 0.8803,
       "step": 795
     },
     {
-      "epoch": 0.33140016570008285,
-      "grad_norm": 2.679914960811432,
-      "learning_rate": 8.458000600117604e-06,
-      "loss": 0.5758,
+      "epoch": 0.12942889500080892,
+      "grad_norm": 0.7617229913207306,
+      "learning_rate": 9.973893020076402e-06,
+      "loss": 0.8764,
       "step": 800
     },
     {
-      "epoch": 0.33347141673570835,
-      "grad_norm": 2.714722412588362,
-      "learning_rate": 8.43179263236608e-06,
-      "loss": 0.5703,
+      "epoch": 0.13023782559456398,
+      "grad_norm": 0.7990666084729844,
+      "learning_rate": 9.972432068388885e-06,
+      "loss": 0.8841,
       "step": 805
     },
     {
-      "epoch": 0.3355426677713339,
-      "grad_norm": 2.8825027456443593,
-      "learning_rate": 8.40540517502813e-06,
-      "loss": 0.5741,
+      "epoch": 0.13104675618831904,
+      "grad_norm": 0.8144767375507436,
+      "learning_rate": 9.970931457281693e-06,
+      "loss": 0.8751,
       "step": 810
     },
     {
-      "epoch": 0.3376139188069594,
-      "grad_norm": 2.69779454090861,
-      "learning_rate": 8.378839608220304e-06,
-      "loss": 0.5758,
+      "epoch": 0.1318556867820741,
+      "grad_norm": 0.8471484705899229,
+      "learning_rate": 9.96939119872349e-06,
+      "loss": 0.878,
       "step": 815
     },
     {
-      "epoch": 0.3396851698425849,
-      "grad_norm": 2.591195744363521,
-      "learning_rate": 8.35209732137463e-06,
-      "loss": 0.5703,
+      "epoch": 0.13266461737582916,
+      "grad_norm": 0.7419857259389397,
+      "learning_rate": 9.96781130499916e-06,
+      "loss": 0.8593,
       "step": 820
     },
     {
-      "epoch": 0.34175642087821045,
-      "grad_norm": 2.5928484463176056,
-      "learning_rate": 8.32517971316595e-06,
-      "loss": 0.5789,
+      "epoch": 0.13347354796958422,
+      "grad_norm": 0.6966615337982099,
+      "learning_rate": 9.966191788709716e-06,
+      "loss": 0.8754,
       "step": 825
     },
     {
-      "epoch": 0.34382767191383595,
-      "grad_norm": 2.8392376730929962,
-      "learning_rate": 8.298088191438753e-06,
-      "loss": 0.5573,
+      "epoch": 0.13428247856333927,
+      "grad_norm": 0.7552286075445729,
+      "learning_rate": 9.96453266277219e-06,
+      "loss": 0.8729,
       "step": 830
     },
     {
-      "epoch": 0.34589892294946145,
-      "grad_norm": 2.717204690264741,
-      "learning_rate": 8.270824173133563e-06,
-      "loss": 0.5553,
+      "epoch": 0.13509140915709433,
+      "grad_norm": 0.8299494943538954,
+      "learning_rate": 9.96283394041954e-06,
+      "loss": 0.881,
       "step": 835
     },
     {
-      "epoch": 0.347970173985087,
-      "grad_norm": 2.6426626328945124,
-      "learning_rate": 8.243389084212808e-06,
-      "loss": 0.5503,
+      "epoch": 0.13590033975084936,
+      "grad_norm": 0.734889508057368,
+      "learning_rate": 9.961095635200536e-06,
+      "loss": 0.8642,
       "step": 840
     },
     {
-      "epoch": 0.3500414250207125,
-      "grad_norm": 2.6361988946032806,
-      "learning_rate": 8.215784359586257e-06,
-      "loss": 0.5498,
+      "epoch": 0.13670927034460442,
+      "grad_norm": 0.7083874584173164,
+      "learning_rate": 9.959317760979654e-06,
+      "loss": 0.87,
       "step": 845
     },
     {
-      "epoch": 0.352112676056338,
-      "grad_norm": 2.452756091881905,
-      "learning_rate": 8.188011443035962e-06,
-      "loss": 0.538,
+      "epoch": 0.13751820093835948,
+      "grad_norm": 0.6795303692054026,
+      "learning_rate": 9.957500331936971e-06,
+      "loss": 0.8916,
       "step": 850
     },
     {
-      "epoch": 0.35418392709196356,
-      "grad_norm": 2.740977671213533,
-      "learning_rate": 8.160071787140742e-06,
-      "loss": 0.5452,
+      "epoch": 0.13832713153211454,
+      "grad_norm": 0.7711107263091647,
+      "learning_rate": 9.955643362568048e-06,
+      "loss": 0.8867,
       "step": 855
     },
     {
-      "epoch": 0.35625517812758906,
-      "grad_norm": 2.650236162364244,
-      "learning_rate": 8.131966853200226e-06,
-      "loss": 0.5485,
+      "epoch": 0.1391360621258696,
+      "grad_norm": 0.7822464459121171,
+      "learning_rate": 9.953746867683807e-06,
+      "loss": 0.875,
       "step": 860
     },
     {
-      "epoch": 0.35832642916321455,
-      "grad_norm": 2.701846208365303,
-      "learning_rate": 8.103698111158405e-06,
-      "loss": 0.5495,
+      "epoch": 0.13994499271962466,
+      "grad_norm": 0.8269771768105255,
+      "learning_rate": 9.951810862410426e-06,
+      "loss": 0.8762,
       "step": 865
     },
     {
-      "epoch": 0.3603976801988401,
-      "grad_norm": 2.7179474541169104,
-      "learning_rate": 8.075267039526764e-06,
-      "loss": 0.548,
+      "epoch": 0.14075392331337971,
+      "grad_norm": 0.766471476016186,
+      "learning_rate": 9.949835362189215e-06,
+      "loss": 0.8635,
       "step": 870
     },
     {
-      "epoch": 0.3624689312344656,
-      "grad_norm": 2.833260463644198,
-      "learning_rate": 8.046675125306948e-06,
-      "loss": 0.5632,
+      "epoch": 0.14156285390713477,
+      "grad_norm": 0.8196823476460824,
+      "learning_rate": 9.947820382776482e-06,
+      "loss": 0.8933,
       "step": 875
     },
     {
-      "epoch": 0.36454018227009116,
-      "grad_norm": 2.7066133568619724,
-      "learning_rate": 8.017923863912989e-06,
-      "loss": 0.5475,
+      "epoch": 0.14237178450088983,
+      "grad_norm": 0.7502232673514019,
+      "learning_rate": 9.945765940243422e-06,
+      "loss": 0.9162,
       "step": 880
     },
     {
-      "epoch": 0.36661143330571666,
-      "grad_norm": 2.554105810407388,
-      "learning_rate": 7.989014759093095e-06,
-      "loss": 0.5455,
+      "epoch": 0.1431807150946449,
+      "grad_norm": 0.8243799834314466,
+      "learning_rate": 9.943672050975979e-06,
+      "loss": 0.8973,
       "step": 885
     },
     {
-      "epoch": 0.36868268434134216,
-      "grad_norm": 2.530083794168148,
-      "learning_rate": 7.959949322850994e-06,
-      "loss": 0.5397,
+      "epoch": 0.14398964568839995,
+      "grad_norm": 0.7327053926662375,
+      "learning_rate": 9.94153873167472e-06,
+      "loss": 0.8754,
       "step": 890
     },
     {
-      "epoch": 0.3707539353769677,
-      "grad_norm": 2.6016133268460937,
-      "learning_rate": 7.930729075366867e-06,
-      "loss": 0.542,
+      "epoch": 0.14479857628215498,
+      "grad_norm": 0.7710933741203212,
+      "learning_rate": 9.9393659993547e-06,
+      "loss": 0.8989,
       "step": 895
     },
     {
-      "epoch": 0.3728251864125932,
-      "grad_norm": 2.5754320630229754,
-      "learning_rate": 7.901355544917827e-06,
-      "loss": 0.5362,
+      "epoch": 0.14560750687591004,
+      "grad_norm": 0.6969476376928585,
+      "learning_rate": 9.937153871345326e-06,
+      "loss": 0.8851,
       "step": 900
     },
     {
-      "epoch": 0.3748964374482187,
-      "grad_norm": 2.576938951435503,
-      "learning_rate": 7.87183026779799e-06,
-      "loss": 0.5414,
+      "epoch": 0.1464164374696651,
+      "grad_norm": 0.7537226761385668,
+      "learning_rate": 9.934902365290222e-06,
+      "loss": 0.8678,
       "step": 905
     },
     {
-      "epoch": 0.37696768848384427,
-      "grad_norm": 2.6785725784997796,
-      "learning_rate": 7.842154788238124e-06,
-      "loss": 0.5512,
+      "epoch": 0.14722536806342015,
+      "grad_norm": 0.7880392328433623,
+      "learning_rate": 9.932611499147082e-06,
+      "loss": 0.8785,
       "step": 910
     },
     {
-      "epoch": 0.37903893951946976,
-      "grad_norm": 2.6795394480092414,
-      "learning_rate": 7.812330658324884e-06,
-      "loss": 0.5375,
+      "epoch": 0.1480342986571752,
+      "grad_norm": 0.7979672909632645,
+      "learning_rate": 9.930281291187534e-06,
+      "loss": 0.9031,
       "step": 915
     },
     {
-      "epoch": 0.38111019055509526,
-      "grad_norm": 2.672382602041246,
-      "learning_rate": 7.782359437919644e-06,
-      "loss": 0.5457,
+      "epoch": 0.14884322925093027,
+      "grad_norm": 0.769227717875029,
+      "learning_rate": 9.927911759996989e-06,
+      "loss": 0.8752,
       "step": 920
     },
     {
-      "epoch": 0.3831814415907208,
-      "grad_norm": 2.884729781500387,
-      "learning_rate": 7.75224269457689e-06,
-      "loss": 0.5239,
+      "epoch": 0.14965215984468533,
+      "grad_norm": 0.7501410474884525,
+      "learning_rate": 9.925502924474495e-06,
+      "loss": 0.8818,
       "step": 925
     },
     {
-      "epoch": 0.3852526926263463,
-      "grad_norm": 2.56507336027056,
-      "learning_rate": 7.721982003462255e-06,
-      "loss": 0.5405,
+      "epoch": 0.1504610904384404,
+      "grad_norm": 0.7900709039002747,
+      "learning_rate": 9.923054803832585e-06,
+      "loss": 0.8549,
       "step": 930
     },
     {
-      "epoch": 0.3873239436619718,
-      "grad_norm": 2.5473419393915,
-      "learning_rate": 7.691578947270122e-06,
-      "loss": 0.5438,
+      "epoch": 0.15127002103219545,
+      "grad_norm": 0.727528278106019,
+      "learning_rate": 9.920567417597127e-06,
+      "loss": 0.869,
       "step": 935
     },
     {
-      "epoch": 0.38939519469759737,
-      "grad_norm": 2.783657224970263,
-      "learning_rate": 7.661035116140856e-06,
-      "loss": 0.5198,
+      "epoch": 0.1520789516259505,
+      "grad_norm": 0.7478261446374327,
+      "learning_rate": 9.918040785607163e-06,
+      "loss": 0.8803,
       "step": 940
     },
     {
-      "epoch": 0.39146644573322287,
-      "grad_norm": 2.6274430429540927,
-      "learning_rate": 7.63035210757763e-06,
-      "loss": 0.5535,
+      "epoch": 0.15288788221970556,
+      "grad_norm": 0.7798141297679039,
+      "learning_rate": 9.915474928014754e-06,
+      "loss": 0.8601,
       "step": 945
     },
     {
-      "epoch": 0.39353769676884837,
-      "grad_norm": 2.7442552412406958,
-      "learning_rate": 7.599531526362873e-06,
-      "loss": 0.5304,
+      "epoch": 0.1536968128134606,
+      "grad_norm": 0.743300400063168,
+      "learning_rate": 9.912869865284821e-06,
+      "loss": 0.873,
       "step": 950
     },
     {
-      "epoch": 0.3956089478044739,
-      "grad_norm": 2.7785345881187786,
-      "learning_rate": 7.568574984474335e-06,
-      "loss": 0.5225,
+      "epoch": 0.15450574340721565,
+      "grad_norm": 0.7855652729779465,
+      "learning_rate": 9.91022561819498e-06,
+      "loss": 0.8556,
       "step": 955
     },
     {
-      "epoch": 0.3976801988400994,
-      "grad_norm": 2.720728749217881,
-      "learning_rate": 7.537484101000787e-06,
-      "loss": 0.5217,
+      "epoch": 0.1553146740009707,
+      "grad_norm": 0.7512935005312259,
+      "learning_rate": 9.90754220783537e-06,
+      "loss": 0.8761,
       "step": 960
     },
     {
-      "epoch": 0.3997514498757249,
-      "grad_norm": 2.637160358823368,
-      "learning_rate": 7.506260502057325e-06,
-      "loss": 0.5297,
+      "epoch": 0.15612360459472577,
+      "grad_norm": 0.7480633701848225,
+      "learning_rate": 9.9048196556085e-06,
+      "loss": 0.884,
       "step": 965
     },
     {
-      "epoch": 0.40182270091135047,
-      "grad_norm": 2.6335954311355216,
-      "learning_rate": 7.474905820700334e-06,
-      "loss": 0.5291,
+      "epoch": 0.15693253518848083,
+      "grad_norm": 0.7649617436679113,
+      "learning_rate": 9.902057983229059e-06,
+      "loss": 0.8558,
       "step": 970
     },
     {
-      "epoch": 0.40389395194697597,
-      "grad_norm": 2.656261545637842,
-      "learning_rate": 7.443421696842066e-06,
-      "loss": 0.5076,
+      "epoch": 0.15774146578223588,
+      "grad_norm": 0.7423506733665634,
+      "learning_rate": 9.89925721272376e-06,
+      "loss": 0.8773,
       "step": 975
     },
     {
-      "epoch": 0.40596520298260147,
-      "grad_norm": 2.6508585415601362,
-      "learning_rate": 7.411809777164873e-06,
-      "loss": 0.5183,
+      "epoch": 0.15855039637599094,
+      "grad_norm": 0.7931165102529991,
+      "learning_rate": 9.89641736643116e-06,
+      "loss": 0.8846,
       "step": 980
     },
     {
-      "epoch": 0.408036454018227,
-      "grad_norm": 2.665237164150542,
-      "learning_rate": 7.380071715035089e-06,
-      "loss": 0.5241,
+      "epoch": 0.159359326969746,
+      "grad_norm": 0.7550133131247714,
+      "learning_rate": 9.893538467001466e-06,
+      "loss": 0.8727,
       "step": 985
     },
     {
-      "epoch": 0.4101077050538525,
-      "grad_norm": 2.5770691509075565,
-      "learning_rate": 7.3482091704165405e-06,
-      "loss": 0.5164,
+      "epoch": 0.16016825756350106,
+      "grad_norm": 0.773118150891364,
+      "learning_rate": 9.89062053739638e-06,
+      "loss": 0.8808,
       "step": 990
     },
     {
-      "epoch": 0.412178956089478,
-      "grad_norm": 2.5586140708653584,
-      "learning_rate": 7.316223809783745e-06,
-      "loss": 0.4982,
+      "epoch": 0.16097718815725612,
+      "grad_norm": 0.75632598887103,
+      "learning_rate": 9.887663600888897e-06,
+      "loss": 0.8713,
       "step": 995
     },
     {
-      "epoch": 0.4142502071251036,
-      "grad_norm": 2.6304400469333697,
-      "learning_rate": 7.284117306034733e-06,
-      "loss": 0.514,
+      "epoch": 0.16178611875101118,
+      "grad_norm": 0.8316783703999336,
+      "learning_rate": 9.88466768106313e-06,
+      "loss": 0.8571,
       "step": 1000
     },
     {
-      "epoch": 0.4163214581607291,
-      "grad_norm": 2.5369318681193667,
-      "learning_rate": 7.2518913384035685e-06,
-      "loss": 0.5119,
+      "epoch": 0.1625950493447662,
+      "grad_norm": 0.7454307910719469,
+      "learning_rate": 9.881632801814112e-06,
+      "loss": 0.8602,
       "step": 1005
     },
     {
-      "epoch": 0.4183927091963546,
-      "grad_norm": 2.7061992404592683,
-      "learning_rate": 7.219547592372512e-06,
-      "loss": 0.5175,
+      "epoch": 0.16340397993852127,
+      "grad_norm": 0.7564730753355293,
+      "learning_rate": 9.878558987347613e-06,
+      "loss": 0.8694,
       "step": 1010
     },
     {
-      "epoch": 0.4204639602319801,
-      "grad_norm": 2.420325437701678,
-      "learning_rate": 7.187087759583869e-06,
-      "loss": 0.5063,
+      "epoch": 0.16421291053227632,
+      "grad_norm": 0.7924034403437432,
+      "learning_rate": 9.875446262179948e-06,
+      "loss": 0.8695,
       "step": 1015
     },
     {
-      "epoch": 0.4225352112676056,
-      "grad_norm": 2.7148650041029514,
-      "learning_rate": 7.15451353775151e-06,
-      "loss": 0.5237,
+      "epoch": 0.16502184112603138,
+      "grad_norm": 0.809312814940399,
+      "learning_rate": 9.872294651137773e-06,
+      "loss": 0.8639,
       "step": 1020
     },
     {
-      "epoch": 0.4246064623032311,
-      "grad_norm": 2.654478692235913,
-      "learning_rate": 7.121826630572084e-06,
-      "loss": 0.4966,
+      "epoch": 0.16583077171978644,
+      "grad_norm": 0.8496318363889708,
+      "learning_rate": 9.869104179357898e-06,
+      "loss": 0.8656,
       "step": 1025
     },
     {
-      "epoch": 0.4266777133388567,
-      "grad_norm": 2.6379021813183354,
-      "learning_rate": 7.089028747635908e-06,
-      "loss": 0.4938,
+      "epoch": 0.1666397023135415,
+      "grad_norm": 0.8463627183621499,
+      "learning_rate": 9.865874872287076e-06,
+      "loss": 0.8721,
       "step": 1030
     },
     {
-      "epoch": 0.4287489643744822,
-      "grad_norm": 2.713654107850754,
-      "learning_rate": 7.056121604337554e-06,
-      "loss": 0.5074,
+      "epoch": 0.16744863290729656,
+      "grad_norm": 0.7538421574954219,
+      "learning_rate": 9.862606755681805e-06,
+      "loss": 0.887,
       "step": 1035
     },
     {
-      "epoch": 0.43082021541010773,
-      "grad_norm": 2.540085709590772,
-      "learning_rate": 7.023106921786118e-06,
-      "loss": 0.4995,
+      "epoch": 0.16825756350105162,
+      "grad_norm": 0.8015977734669262,
+      "learning_rate": 9.859299855608127e-06,
+      "loss": 0.8652,
       "step": 1040
     },
     {
-      "epoch": 0.43289146644573323,
-      "grad_norm": 2.5985604745407374,
-      "learning_rate": 6.9899864267152275e-06,
-      "loss": 0.5318,
+      "epoch": 0.16906649409480667,
+      "grad_norm": 0.7608756372365878,
+      "learning_rate": 9.855954198441411e-06,
+      "loss": 0.8882,
       "step": 1045
     },
     {
-      "epoch": 0.43496271748135873,
-      "grad_norm": 2.4411457867844417,
-      "learning_rate": 6.956761851392706e-06,
-      "loss": 0.4858,
+      "epoch": 0.16987542468856173,
+      "grad_norm": 0.8188556394828449,
+      "learning_rate": 9.852569810866148e-06,
+      "loss": 0.8694,
       "step": 1050
     },
     {
-      "epoch": 0.4370339685169843,
-      "grad_norm": 2.4714643955598072,
-      "learning_rate": 6.9234349335299835e-06,
-      "loss": 0.5171,
+      "epoch": 0.1706843552823168,
+      "grad_norm": 0.805253431195849,
+      "learning_rate": 9.849146719875737e-06,
+      "loss": 0.8754,
       "step": 1055
     },
     {
-      "epoch": 0.4391052195526098,
-      "grad_norm": 2.4510153141350868,
-      "learning_rate": 6.890007416191209e-06,
-      "loss": 0.4875,
+      "epoch": 0.17149328587607182,
+      "grad_norm": 0.7590085927711259,
+      "learning_rate": 9.845684952772274e-06,
+      "loss": 0.8741,
       "step": 1060
     },
     {
-      "epoch": 0.4411764705882353,
-      "grad_norm": 2.502867427304289,
-      "learning_rate": 6.8564810477020835e-06,
-      "loss": 0.4982,
+      "epoch": 0.17230221646982688,
+      "grad_norm": 0.7493004472784411,
+      "learning_rate": 9.842184537166326e-06,
+      "loss": 0.8784,
       "step": 1065
     },
     {
-      "epoch": 0.44324772162386084,
-      "grad_norm": 2.5224619340497183,
-      "learning_rate": 6.822857581558423e-06,
-      "loss": 0.4971,
+      "epoch": 0.17311114706358194,
+      "grad_norm": 0.8005839214503973,
+      "learning_rate": 9.838645500976716e-06,
+      "loss": 0.8668,
       "step": 1070
     },
     {
-      "epoch": 0.44531897265948633,
-      "grad_norm": 2.622904137029192,
-      "learning_rate": 6.789138776334441e-06,
-      "loss": 0.5063,
+      "epoch": 0.173920077657337,
+      "grad_norm": 0.7331632983706363,
+      "learning_rate": 9.835067872430297e-06,
+      "loss": 0.8577,
       "step": 1075
     },
     {
-      "epoch": 0.44739022369511183,
-      "grad_norm": 2.4481597241493174,
-      "learning_rate": 6.7553263955907755e-06,
-      "loss": 0.4841,
+      "epoch": 0.17472900825109206,
+      "grad_norm": 0.8212489518002695,
+      "learning_rate": 9.831451680061735e-06,
+      "loss": 0.8503,
       "step": 1080
     },
     {
-      "epoch": 0.4494614747307374,
-      "grad_norm": 2.505241922300399,
-      "learning_rate": 6.721422207782249e-06,
-      "loss": 0.486,
+      "epoch": 0.1755379388448471,
+      "grad_norm": 0.7423733165945756,
+      "learning_rate": 9.82779695271327e-06,
+      "loss": 0.849,
       "step": 1085
     },
     {
-      "epoch": 0.4515327257663629,
-      "grad_norm": 2.51038092005137,
-      "learning_rate": 6.687427986165379e-06,
-      "loss": 0.4866,
+      "epoch": 0.17634686943860217,
+      "grad_norm": 0.8436334149956825,
+      "learning_rate": 9.824103719534497e-06,
+      "loss": 0.8739,
       "step": 1090
     },
     {
-      "epoch": 0.4536039768019884,
-      "grad_norm": 2.618228992641671,
-      "learning_rate": 6.653345508705629e-06,
-      "loss": 0.4889,
+      "epoch": 0.17715580003235723,
+      "grad_norm": 0.8371560523056507,
+      "learning_rate": 9.820372009982122e-06,
+      "loss": 0.8885,
       "step": 1095
     },
     {
-      "epoch": 0.45567522783761394,
-      "grad_norm": 2.6041268512577873,
-      "learning_rate": 6.6191765579844205e-06,
-      "loss": 0.4957,
+      "epoch": 0.1779647306261123,
+      "grad_norm": 0.8549576764605578,
+      "learning_rate": 9.816601853819739e-06,
+      "loss": 0.8905,
       "step": 1100
     },
     {
-      "epoch": 0.45774647887323944,
-      "grad_norm": 2.5703382132992414,
-      "learning_rate": 6.584922921105894e-06,
-      "loss": 0.4805,
+      "epoch": 0.17877366121986735,
+      "grad_norm": 0.8302136713634986,
+      "learning_rate": 9.81279328111758e-06,
+      "loss": 0.863,
       "step": 1105
     },
     {
-      "epoch": 0.45981772990886494,
-      "grad_norm": 2.4810077168179894,
-      "learning_rate": 6.550586389603451e-06,
-      "loss": 0.4843,
+      "epoch": 0.17958259181362238,
+      "grad_norm": 0.7692548326236497,
+      "learning_rate": 9.80894632225229e-06,
+      "loss": 0.8942,
       "step": 1110
     },
     {
-      "epoch": 0.4618889809444905,
-      "grad_norm": 2.7146694842288492,
-      "learning_rate": 6.5161687593460395e-06,
-      "loss": 0.4805,
+      "epoch": 0.18039152240737744,
+      "grad_norm": 0.8190077972884791,
+      "learning_rate": 9.805061007906668e-06,
+      "loss": 0.8721,
       "step": 1115
     },
     {
-      "epoch": 0.463960231980116,
-      "grad_norm": 2.661840488355652,
-      "learning_rate": 6.481671830444243e-06,
-      "loss": 0.4766,
+      "epoch": 0.1812004530011325,
+      "grad_norm": 0.8021866541591391,
+      "learning_rate": 9.801137369069441e-06,
+      "loss": 0.8587,
       "step": 1120
     },
     {
-      "epoch": 0.4660314830157415,
-      "grad_norm": 2.542982024246636,
-      "learning_rate": 6.447097407156114e-06,
-      "loss": 0.4714,
+      "epoch": 0.18200938359488755,
+      "grad_norm": 0.7295302337882422,
+      "learning_rate": 9.797175437034997e-06,
+      "loss": 0.8809,
       "step": 1125
     },
     {
-      "epoch": 0.46810273405136704,
-      "grad_norm": 2.545352453639695,
-      "learning_rate": 6.412447297792818e-06,
-      "loss": 0.4627,
+      "epoch": 0.1828183141886426,
+      "grad_norm": 0.7303087771888767,
+      "learning_rate": 9.79317524340315e-06,
+      "loss": 0.864,
       "step": 1130
     },
     {
-      "epoch": 0.47017398508699254,
-      "grad_norm": 2.601762512884624,
-      "learning_rate": 6.377723314624057e-06,
-      "loss": 0.4617,
+      "epoch": 0.18362724478239767,
+      "grad_norm": 0.7668374049535479,
+      "learning_rate": 9.789136820078884e-06,
+      "loss": 0.8768,
       "step": 1135
     },
     {
-      "epoch": 0.47224523612261804,
-      "grad_norm": 2.5271935902650746,
-      "learning_rate": 6.3429272737832726e-06,
-      "loss": 0.482,
+      "epoch": 0.18443617537615273,
+      "grad_norm": 0.7779749351860815,
+      "learning_rate": 9.785060199272096e-06,
+      "loss": 0.8871,
       "step": 1140
     },
     {
-      "epoch": 0.4743164871582436,
-      "grad_norm": 2.8257956277427265,
-      "learning_rate": 6.308060995172673e-06,
-      "loss": 0.4875,
+      "epoch": 0.1852451059699078,
+      "grad_norm": 0.7605751732056839,
+      "learning_rate": 9.780945413497339e-06,
+      "loss": 0.8756,
       "step": 1145
     },
     {
-      "epoch": 0.4763877381938691,
-      "grad_norm": 2.6203637102255843,
-      "learning_rate": 6.273126302368037e-06,
-      "loss": 0.4725,
+      "epoch": 0.18605403656366284,
+      "grad_norm": 0.7608509607177488,
+      "learning_rate": 9.776792495573567e-06,
+      "loss": 0.8651,
       "step": 1150
     },
     {
-      "epoch": 0.4784589892294946,
-      "grad_norm": 2.4615255657443855,
-      "learning_rate": 6.238125022523343e-06,
-      "loss": 0.462,
+      "epoch": 0.1868629671574179,
+      "grad_norm": 0.7649020527995475,
+      "learning_rate": 9.772601478623871e-06,
+      "loss": 0.8522,
       "step": 1155
     },
     {
-      "epoch": 0.48053024026512015,
-      "grad_norm": 2.399337323008823,
-      "learning_rate": 6.203058986275207e-06,
-      "loss": 0.4706,
+      "epoch": 0.18767189775117296,
+      "grad_norm": 0.8284338499023652,
+      "learning_rate": 9.768372396075213e-06,
+      "loss": 0.8986,
       "step": 1160
     },
     {
-      "epoch": 0.48260149130074564,
-      "grad_norm": 2.568546498877998,
-      "learning_rate": 6.1679300276471285e-06,
-      "loss": 0.46,
+      "epoch": 0.188480828344928,
+      "grad_norm": 0.7588241076036676,
+      "learning_rate": 9.764105281658161e-06,
+      "loss": 0.8727,
       "step": 1165
     },
     {
-      "epoch": 0.48467274233637114,
-      "grad_norm": 2.5774193338530886,
-      "learning_rate": 6.132739983953579e-06,
-      "loss": 0.4639,
+      "epoch": 0.18928975893868305,
+      "grad_norm": 0.7815552643338736,
+      "learning_rate": 9.759800169406621e-06,
+      "loss": 0.8639,
       "step": 1170
     },
     {
-      "epoch": 0.4867439933719967,
-      "grad_norm": 2.433169814085788,
-      "learning_rate": 6.097490695703896e-06,
-      "loss": 0.4706,
+      "epoch": 0.1900986895324381,
+      "grad_norm": 0.7935036774949533,
+      "learning_rate": 9.755457093657562e-06,
+      "loss": 0.8651,
       "step": 1175
     },
     {
-      "epoch": 0.4888152444076222,
-      "grad_norm": 2.5672509798739975,
-      "learning_rate": 6.062184006506027e-06,
-      "loss": 0.469,
+      "epoch": 0.19090762012619317,
+      "grad_norm": 0.7411492264667076,
+      "learning_rate": 9.751076089050747e-06,
+      "loss": 0.8675,
       "step": 1180
     },
     {
-      "epoch": 0.4908864954432477,
-      "grad_norm": 2.566412335613826,
-      "learning_rate": 6.026821762970102e-06,
-      "loss": 0.4667,
+      "epoch": 0.19171655071994823,
+      "grad_norm": 0.7597654203339239,
+      "learning_rate": 9.746657190528454e-06,
+      "loss": 0.8421,
       "step": 1185
     },
     {
-      "epoch": 0.49295774647887325,
-      "grad_norm": 2.4468040851684885,
-      "learning_rate": 5.991405814611855e-06,
-      "loss": 0.4547,
+      "epoch": 0.19252548131370328,
+      "grad_norm": 0.7871189163253436,
+      "learning_rate": 9.742200433335196e-06,
+      "loss": 0.877,
       "step": 1190
     },
     {
-      "epoch": 0.49502899751449875,
-      "grad_norm": 2.490658198568155,
-      "learning_rate": 5.955938013755888e-06,
-      "loss": 0.4628,
+      "epoch": 0.19333441190745834,
+      "grad_norm": 0.7773214820602937,
+      "learning_rate": 9.737705853017442e-06,
+      "loss": 0.8721,
       "step": 1195
     },
     {
-      "epoch": 0.4971002485501243,
-      "grad_norm": 2.463293970038353,
-      "learning_rate": 5.920420215438794e-06,
-      "loss": 0.4804,
+      "epoch": 0.1941433425012134,
+      "grad_norm": 0.8230817780724599,
+      "learning_rate": 9.733173485423333e-06,
+      "loss": 0.8714,
       "step": 1200
     },
     {
-      "epoch": 0.4991714995857498,
-      "grad_norm": 2.6047443491478393,
-      "learning_rate": 5.8848542773121285e-06,
-      "loss": 0.468,
+      "epoch": 0.19495227309496846,
+      "grad_norm": 0.7681885249706483,
+      "learning_rate": 9.7286033667024e-06,
+      "loss": 0.8594,
       "step": 1205
     },
     {
-      "epoch": 0.5012427506213754,
-      "grad_norm": 2.4826419630150407,
-      "learning_rate": 5.849242059545259e-06,
-      "loss": 0.4565,
+      "epoch": 0.19576120368872352,
+      "grad_norm": 0.7640895548306019,
+      "learning_rate": 9.723995533305262e-06,
+      "loss": 0.8621,
       "step": 1210
     },
     {
-      "epoch": 0.5033140016570008,
-      "grad_norm": 2.757538020740051,
-      "learning_rate": 5.81358542472807e-06,
-      "loss": 0.4598,
+      "epoch": 0.19657013428247858,
+      "grad_norm": 0.876302015000407,
+      "learning_rate": 9.719350021983356e-06,
+      "loss": 0.873,
       "step": 1215
     },
     {
-      "epoch": 0.5053852526926264,
-      "grad_norm": 2.522678699637508,
-      "learning_rate": 5.777886237773542e-06,
-      "loss": 0.4451,
+      "epoch": 0.1973790648762336,
+      "grad_norm": 0.7687883497172476,
+      "learning_rate": 9.714666869788622e-06,
+      "loss": 0.8837,
       "step": 1220
     },
     {
-      "epoch": 0.5074565037282519,
-      "grad_norm": 2.697025952581437,
-      "learning_rate": 5.742146365820223e-06,
-      "loss": 0.4606,
+      "epoch": 0.19818799546998866,
+      "grad_norm": 0.7918479859070253,
+      "learning_rate": 9.709946114073231e-06,
+      "loss": 0.8742,
       "step": 1225
     },
     {
-      "epoch": 0.5095277547638773,
-      "grad_norm": 2.5255536162735095,
-      "learning_rate": 5.706367678134562e-06,
-      "loss": 0.4587,
+      "epoch": 0.19899692606374372,
+      "grad_norm": 0.7704108440790387,
+      "learning_rate": 9.705187792489263e-06,
+      "loss": 0.8831,
       "step": 1230
     },
     {
-      "epoch": 0.5115990057995029,
-      "grad_norm": 2.5706493150820866,
-      "learning_rate": 5.670552046013151e-06,
-      "loss": 0.4626,
+      "epoch": 0.19980585665749878,
+      "grad_norm": 0.7958664141923848,
+      "learning_rate": 9.700391942988422e-06,
+      "loss": 0.8815,
       "step": 1235
     },
     {
-      "epoch": 0.5136702568351285,
-      "grad_norm": 2.478635300584232,
-      "learning_rate": 5.634701342684852e-06,
-      "loss": 0.4495,
+      "epoch": 0.20061478725125384,
+      "grad_norm": 0.7299908461383041,
+      "learning_rate": 9.695558603821735e-06,
+      "loss": 0.8705,
       "step": 1240
     },
     {
-      "epoch": 0.5157415078707539,
-      "grad_norm": 2.413081651555011,
-      "learning_rate": 5.598817443212813e-06,
-      "loss": 0.4545,
+      "epoch": 0.2014237178450089,
+      "grad_norm": 0.7921888129305676,
+      "learning_rate": 9.69068781353923e-06,
+      "loss": 0.8651,
       "step": 1245
     },
     {
-      "epoch": 0.5178127589063795,
-      "grad_norm": 2.4514043208682614,
-      "learning_rate": 5.562902224396416e-06,
-      "loss": 0.4487,
+      "epoch": 0.20223264843876396,
+      "grad_norm": 0.783261666055412,
+      "learning_rate": 9.68577961098965e-06,
+      "loss": 0.8657,
       "step": 1250
     },
     {
-      "epoch": 0.519884009942005,
-      "grad_norm": 2.8553325471084894,
-      "learning_rate": 5.526957564673098e-06,
-      "loss": 0.4491,
+      "epoch": 0.20304157903251902,
+      "grad_norm": 0.7640970071314145,
+      "learning_rate": 9.680834035320127e-06,
+      "loss": 0.8613,
       "step": 1255
     },
     {
-      "epoch": 0.5219552609776305,
-      "grad_norm": 2.3499605624954043,
-      "learning_rate": 5.49098534402012e-06,
-      "loss": 0.455,
+      "epoch": 0.20385050962627407,
+      "grad_norm": 0.7480591135260592,
+      "learning_rate": 9.675851125975879e-06,
+      "loss": 0.8522,
       "step": 1260
     },
     {
-      "epoch": 0.524026512013256,
-      "grad_norm": 2.511645835881994,
-      "learning_rate": 5.454987443856235e-06,
-      "loss": 0.4405,
+      "epoch": 0.20465944022002913,
+      "grad_norm": 0.764427552884641,
+      "learning_rate": 9.67083092269989e-06,
+      "loss": 0.8846,
       "step": 1265
     },
     {
-      "epoch": 0.5260977630488816,
-      "grad_norm": 2.5573246473567663,
-      "learning_rate": 5.418965746943281e-06,
-      "loss": 0.4573,
+      "epoch": 0.2054683708137842,
+      "grad_norm": 0.7562918652667725,
+      "learning_rate": 9.665773465532597e-06,
+      "loss": 0.8538,
       "step": 1270
     },
     {
-      "epoch": 0.528169014084507,
-      "grad_norm": 2.6069496932706446,
-      "learning_rate": 5.3829221372877175e-06,
-      "loss": 0.4385,
+      "epoch": 0.20627730140753922,
+      "grad_norm": 0.7308192790835369,
+      "learning_rate": 9.660678794811569e-06,
+      "loss": 0.8758,
       "step": 1275
     },
     {
-      "epoch": 0.5302402651201326,
-      "grad_norm": 2.3669759764626286,
-      "learning_rate": 5.34685850004208e-06,
-      "loss": 0.4474,
+      "epoch": 0.20708623200129428,
+      "grad_norm": 0.7327237540533151,
+      "learning_rate": 9.65554695117118e-06,
+      "loss": 0.8556,
       "step": 1280
     },
     {
-      "epoch": 0.5323115161557581,
-      "grad_norm": 2.5262162943745428,
-      "learning_rate": 5.310776721406392e-06,
-      "loss": 0.4465,
+      "epoch": 0.20789516259504934,
+      "grad_norm": 0.8111511896192851,
+      "learning_rate": 9.650377975542298e-06,
+      "loss": 0.8845,
       "step": 1285
     },
     {
-      "epoch": 0.5343827671913836,
-      "grad_norm": 2.367388706229585,
-      "learning_rate": 5.2746786885295034e-06,
-      "loss": 0.4477,
+      "epoch": 0.2087040931888044,
+      "grad_norm": 0.7725083000676402,
+      "learning_rate": 9.645171909151944e-06,
+      "loss": 0.8713,
       "step": 1290
     },
     {
-      "epoch": 0.5364540182270091,
-      "grad_norm": 2.4424670349956834,
-      "learning_rate": 5.238566289410396e-06,
-      "loss": 0.4347,
+      "epoch": 0.20951302378255945,
+      "grad_norm": 0.8087330967911632,
+      "learning_rate": 9.639928793522976e-06,
+      "loss": 0.8807,
       "step": 1295
     },
     {
-      "epoch": 0.5385252692626347,
-      "grad_norm": 2.374597209278358,
-      "learning_rate": 5.2024414127994325e-06,
-      "loss": 0.4414,
+      "epoch": 0.2103219543763145,
+      "grad_norm": 0.7987317654845778,
+      "learning_rate": 9.634648670473743e-06,
+      "loss": 0.8887,
       "step": 1300
     },
     {
-      "epoch": 0.5405965202982601,
-      "grad_norm": 2.363317895955179,
-      "learning_rate": 5.166305948099574e-06,
-      "loss": 0.4211,
+      "epoch": 0.21113088497006957,
+      "grad_norm": 0.7222991798791983,
+      "learning_rate": 9.629331582117766e-06,
+      "loss": 0.8617,
       "step": 1305
     },
     {
-      "epoch": 0.5426677713338857,
-      "grad_norm": 2.445116140490052,
-      "learning_rate": 5.13016178526756e-06,
-      "loss": 0.4341,
+      "epoch": 0.21193981556382463,
+      "grad_norm": 0.8205396256241525,
+      "learning_rate": 9.623977570863398e-06,
+      "loss": 0.8512,
       "step": 1310
     },
     {
-      "epoch": 0.5447390223695112,
-      "grad_norm": 2.451842751824045,
-      "learning_rate": 5.094010814715062e-06,
-      "loss": 0.4323,
+      "epoch": 0.2127487461575797,
+      "grad_norm": 0.7704920928191922,
+      "learning_rate": 9.618586679413477e-06,
+      "loss": 0.8545,
       "step": 1315
     },
     {
-      "epoch": 0.5468102734051367,
-      "grad_norm": 2.4108281284288116,
-      "learning_rate": 5.057854927209804e-06,
-      "loss": 0.4363,
+      "epoch": 0.21355767675133475,
+      "grad_norm": 0.8121022691153029,
+      "learning_rate": 9.613158950764996e-06,
+      "loss": 0.8647,
       "step": 1320
     },
     {
-      "epoch": 0.5488815244407622,
-      "grad_norm": 2.47478290810414,
-      "learning_rate": 5.0216960137766805e-06,
-      "loss": 0.429,
+      "epoch": 0.2143666073450898,
+      "grad_norm": 0.8068844595908223,
+      "learning_rate": 9.60769442820876e-06,
+      "loss": 0.8511,
       "step": 1325
     },
     {
-      "epoch": 0.5509527754763878,
-      "grad_norm": 2.462991177532584,
-      "learning_rate": 4.985535965598843e-06,
-      "loss": 0.4504,
+      "epoch": 0.21517553793884484,
+      "grad_norm": 0.779446090355673,
+      "learning_rate": 9.602193155329029e-06,
+      "loss": 0.8623,
       "step": 1330
     },
     {
-      "epoch": 0.5530240265120132,
-      "grad_norm": 2.437965270822768,
-      "learning_rate": 4.949376673918802e-06,
-      "loss": 0.432,
+      "epoch": 0.2159844685325999,
+      "grad_norm": 0.7763229071537453,
+      "learning_rate": 9.596655176003185e-06,
+      "loss": 0.8662,
       "step": 1335
     },
     {
-      "epoch": 0.5550952775476388,
-      "grad_norm": 2.588594999778493,
-      "learning_rate": 4.913220029939491e-06,
-      "loss": 0.4335,
+      "epoch": 0.21679339912635495,
+      "grad_norm": 0.7877601933457381,
+      "learning_rate": 9.591080534401371e-06,
+      "loss": 0.87,
       "step": 1340
     },
     {
-      "epoch": 0.5571665285832643,
-      "grad_norm": 2.624887614176281,
-      "learning_rate": 4.877067924725368e-06,
-      "loss": 0.4268,
+      "epoch": 0.21760232972011,
+      "grad_norm": 0.7986846021359395,
+      "learning_rate": 9.585469274986148e-06,
+      "loss": 0.8716,
       "step": 1345
     },
     {
-      "epoch": 0.5592377796188898,
-      "grad_norm": 2.59768554910079,
-      "learning_rate": 4.840922249103506e-06,
-      "loss": 0.4304,
+      "epoch": 0.21841126031386507,
+      "grad_norm": 0.7790084651864484,
+      "learning_rate": 9.579821442512131e-06,
+      "loss": 0.8595,
       "step": 1350
     },
     {
-      "epoch": 0.5613090306545153,
-      "grad_norm": 2.3988418599389063,
-      "learning_rate": 4.804784893564697e-06,
-      "loss": 0.4411,
+      "epoch": 0.21922019090762013,
+      "grad_norm": 0.7565290946235815,
+      "learning_rate": 9.574137082025639e-06,
+      "loss": 0.878,
       "step": 1355
     },
     {
-      "epoch": 0.5633802816901409,
-      "grad_norm": 2.580765462139435,
-      "learning_rate": 4.7686577481645745e-06,
-      "loss": 0.4082,
+      "epoch": 0.22002912150137519,
+      "grad_norm": 0.7534079459249721,
+      "learning_rate": 9.568416238864335e-06,
+      "loss": 0.8542,
       "step": 1360
     },
     {
-      "epoch": 0.5654515327257663,
-      "grad_norm": 2.4467281357535855,
-      "learning_rate": 4.732542702424759e-06,
-      "loss": 0.4382,
+      "epoch": 0.22083805209513024,
+      "grad_norm": 0.7934001624755948,
+      "learning_rate": 9.562658958656856e-06,
+      "loss": 0.8662,
       "step": 1365
     },
     {
-      "epoch": 0.5675227837613919,
-      "grad_norm": 2.6370177774043073,
-      "learning_rate": 4.696441645234042e-06,
-      "loss": 0.4271,
+      "epoch": 0.2216469826888853,
+      "grad_norm": 0.7851224446308791,
+      "learning_rate": 9.556865287322464e-06,
+      "loss": 0.8708,
       "step": 1370
     },
     {
-      "epoch": 0.5695940347970174,
-      "grad_norm": 2.475111070321159,
-      "learning_rate": 4.660356464749578e-06,
-      "loss": 0.4185,
+      "epoch": 0.22245591328264036,
+      "grad_norm": 0.7292260657124646,
+      "learning_rate": 9.551035271070665e-06,
+      "loss": 0.8496,
       "step": 1375
     },
     {
-      "epoch": 0.5716652858326429,
-      "grad_norm": 2.6994305643163776,
-      "learning_rate": 4.624289048298147e-06,
-      "loss": 0.4259,
+      "epoch": 0.22326484387639542,
+      "grad_norm": 0.8178256842698678,
+      "learning_rate": 9.54516895640085e-06,
+      "loss": 0.8615,
       "step": 1380
     },
     {
-      "epoch": 0.5737365368682684,
-      "grad_norm": 2.4520066328265977,
-      "learning_rate": 4.588241282277428e-06,
-      "loss": 0.4179,
+      "epoch": 0.22407377447015045,
+      "grad_norm": 0.7761082648853441,
+      "learning_rate": 9.539266390101922e-06,
+      "loss": 0.8596,
       "step": 1385
     },
     {
-      "epoch": 0.575807787903894,
-      "grad_norm": 2.3795955584858386,
-      "learning_rate": 4.55221505205734e-06,
-      "loss": 0.4287,
+      "epoch": 0.2248827050639055,
+      "grad_norm": 0.7510699850597663,
+      "learning_rate": 9.533327619251921e-06,
+      "loss": 0.8591,
       "step": 1390
     },
     {
-      "epoch": 0.5778790389395194,
-      "grad_norm": 2.381541096732435,
-      "learning_rate": 4.516212241881448e-06,
-      "loss": 0.4157,
+      "epoch": 0.22569163565766057,
+      "grad_norm": 0.7901677982076935,
+      "learning_rate": 9.527352691217649e-06,
+      "loss": 0.856,
       "step": 1395
     },
     {
-      "epoch": 0.579950289975145,
-      "grad_norm": 2.566306679289895,
-      "learning_rate": 4.480234734768393e-06,
-      "loss": 0.4213,
+      "epoch": 0.22650056625141562,
+      "grad_norm": 0.8121283027645855,
+      "learning_rate": 9.52134165365429e-06,
+      "loss": 0.8686,
       "step": 1400
     },
     {
-      "epoch": 0.5820215410107705,
-      "grad_norm": 2.396097079278926,
-      "learning_rate": 4.444284412413418e-06,
-      "loss": 0.4172,
+      "epoch": 0.22730949684517068,
+      "grad_norm": 0.8027876769163085,
+      "learning_rate": 9.515294554505039e-06,
+      "loss": 0.8609,
       "step": 1405
     },
     {
-      "epoch": 0.584092792046396,
-      "grad_norm": 2.3514162631646274,
-      "learning_rate": 4.408363155089952e-06,
-      "loss": 0.4211,
+      "epoch": 0.22811842743892574,
+      "grad_norm": 0.7677214802906873,
+      "learning_rate": 9.509211442000705e-06,
+      "loss": 0.8842,
       "step": 1410
     },
     {
-      "epoch": 0.5861640430820215,
-      "grad_norm": 2.3172982824966906,
-      "learning_rate": 4.3724728415512585e-06,
-      "loss": 0.4193,
+      "epoch": 0.2289273580326808,
+      "grad_norm": 0.7935787903248027,
+      "learning_rate": 9.503092364659343e-06,
+      "loss": 0.8571,
       "step": 1415
     },
     {
-      "epoch": 0.5882352941176471,
-      "grad_norm": 2.472751185964712,
-      "learning_rate": 4.3366153489321855e-06,
-      "loss": 0.4264,
+      "epoch": 0.22973628862643586,
+      "grad_norm": 0.7885818327639111,
+      "learning_rate": 9.496937371285852e-06,
+      "loss": 0.8368,
       "step": 1420
     },
     {
-      "epoch": 0.5903065451532725,
-      "grad_norm": 2.4154833816633365,
-      "learning_rate": 4.30079255265098e-06,
-      "loss": 0.4048,
+      "epoch": 0.23054521922019092,
+      "grad_norm": 0.7664542560047006,
+      "learning_rate": 9.490746510971595e-06,
+      "loss": 0.8586,
       "step": 1425
     },
     {
-      "epoch": 0.5923777961888981,
-      "grad_norm": 2.5211717793789803,
-      "learning_rate": 4.265006326311199e-06,
-      "loss": 0.4089,
+      "epoch": 0.23135414981394598,
+      "grad_norm": 0.7915598638517407,
+      "learning_rate": 9.484519833094006e-06,
+      "loss": 0.8594,
       "step": 1430
     },
     {
-      "epoch": 0.5944490472245236,
-      "grad_norm": 2.3701199367093504,
-      "learning_rate": 4.229258541603723e-06,
-      "loss": 0.4156,
+      "epoch": 0.232163080407701,
+      "grad_norm": 0.8224568006509917,
+      "learning_rate": 9.478257387316189e-06,
+      "loss": 0.8604,
       "step": 1435
     },
     {
-      "epoch": 0.5965202982601492,
-      "grad_norm": 2.3994950566503994,
-      "learning_rate": 4.1935510682088545e-06,
-      "loss": 0.4065,
+      "epoch": 0.23297201100145606,
+      "grad_norm": 0.7646329363431662,
+      "learning_rate": 9.471959223586535e-06,
+      "loss": 0.8667,
       "step": 1440
     },
     {
-      "epoch": 0.5985915492957746,
-      "grad_norm": 2.4968713927832273,
-      "learning_rate": 4.157885773698535e-06,
-      "loss": 0.4057,
+      "epoch": 0.23378094159521112,
+      "grad_norm": 0.8054133554377607,
+      "learning_rate": 9.465625392138314e-06,
+      "loss": 0.8491,
       "step": 1445
     },
     {
-      "epoch": 0.6006628003314002,
-      "grad_norm": 2.416064472998044,
-      "learning_rate": 4.122264523438668e-06,
-      "loss": 0.3931,
+      "epoch": 0.23458987218896618,
+      "grad_norm": 0.7999914607438723,
+      "learning_rate": 9.459255943489271e-06,
+      "loss": 0.8787,
       "step": 1450
     },
     {
-      "epoch": 0.6027340513670257,
-      "grad_norm": 2.2816034379976355,
-      "learning_rate": 4.086689180491554e-06,
-      "loss": 0.4099,
+      "epoch": 0.23539880278272124,
+      "grad_norm": 0.7770791430956507,
+      "learning_rate": 9.45285092844124e-06,
+      "loss": 0.8526,
       "step": 1455
     },
     {
-      "epoch": 0.6048053024026512,
-      "grad_norm": 2.40826781123055,
-      "learning_rate": 4.051161605518453e-06,
-      "loss": 0.4149,
+      "epoch": 0.2362077333764763,
+      "grad_norm": 0.817105925992962,
+      "learning_rate": 9.446410398079716e-06,
+      "loss": 0.8605,
       "step": 1460
     },
     {
-      "epoch": 0.6068765534382767,
-      "grad_norm": 2.538687087204467,
-      "learning_rate": 4.015683656682255e-06,
-      "loss": 0.4175,
+      "epoch": 0.23701666397023136,
+      "grad_norm": 0.7238122855218315,
+      "learning_rate": 9.439934403773468e-06,
+      "loss": 0.8727,
       "step": 1465
     },
     {
-      "epoch": 0.6089478044739023,
-      "grad_norm": 2.430104423449275,
-      "learning_rate": 3.980257189550316e-06,
-      "loss": 0.4053,
+      "epoch": 0.23782559456398641,
+      "grad_norm": 0.7489559356823628,
+      "learning_rate": 9.433422997174113e-06,
+      "loss": 0.8672,
       "step": 1470
     },
     {
-      "epoch": 0.6110190555095277,
-      "grad_norm": 2.3721045844407143,
-      "learning_rate": 3.94488405699739e-06,
-      "loss": 0.4031,
+      "epoch": 0.23863452515774147,
+      "grad_norm": 0.8073337490099621,
+      "learning_rate": 9.42687623021572e-06,
+      "loss": 0.8631,
       "step": 1475
     },
     {
-      "epoch": 0.6130903065451533,
-      "grad_norm": 2.338789854463612,
-      "learning_rate": 3.909566109108727e-06,
-      "loss": 0.3972,
+      "epoch": 0.23944345575149653,
+      "grad_norm": 0.7504637131946496,
+      "learning_rate": 9.42029415511438e-06,
+      "loss": 0.85,
       "step": 1480
     },
     {
-      "epoch": 0.6151615575807788,
-      "grad_norm": 2.2787527608868943,
-      "learning_rate": 3.874305193083313e-06,
-      "loss": 0.411,
+      "epoch": 0.2402523863452516,
+      "grad_norm": 0.7888833671784519,
+      "learning_rate": 9.4136768243678e-06,
+      "loss": 0.8602,
       "step": 1485
     },
     {
-      "epoch": 0.6172328086164043,
-      "grad_norm": 2.454450620709443,
-      "learning_rate": 3.839103153137247e-06,
-      "loss": 0.4008,
+      "epoch": 0.24106131693900662,
+      "grad_norm": 0.8101599644677917,
+      "learning_rate": 9.40702429075488e-06,
+      "loss": 0.8535,
       "step": 1490
     },
     {
-      "epoch": 0.6193040596520298,
-      "grad_norm": 2.4225686330793823,
-      "learning_rate": 3.803961830407297e-06,
-      "loss": 0.3974,
+      "epoch": 0.24187024753276168,
+      "grad_norm": 0.7871350703464342,
+      "learning_rate": 9.400336607335294e-06,
+      "loss": 0.8486,
       "step": 1495
     },
     {
-      "epoch": 0.6213753106876554,
-      "grad_norm": 2.564739658366535,
-      "learning_rate": 3.768883062854598e-06,
-      "loss": 0.4152,
+      "epoch": 0.24267917812651674,
+      "grad_norm": 0.7878053882806549,
+      "learning_rate": 9.393613827449064e-06,
+      "loss": 0.8537,
       "step": 1500
     },
     {
-      "epoch": 0.6234465617232808,
-      "grad_norm": 2.4055933426253837,
-      "learning_rate": 3.7338686851685267e-06,
-      "loss": 0.4021,
+      "epoch": 0.2434881087202718,
+      "grad_norm": 0.7716899022882726,
+      "learning_rate": 9.38685600471614e-06,
+      "loss": 0.8751,
       "step": 1505
     },
     {
-      "epoch": 0.6255178127589064,
-      "grad_norm": 2.3297777756780986,
-      "learning_rate": 3.6989205286707398e-06,
-      "loss": 0.4087,
+      "epoch": 0.24429703931402685,
+      "grad_norm": 0.7804644122014117,
+      "learning_rate": 9.380063193035968e-06,
+      "loss": 0.8496,
       "step": 1510
     },
     {
-      "epoch": 0.6275890637945319,
-      "grad_norm": 2.276174828544462,
-      "learning_rate": 3.664040421219393e-06,
-      "loss": 0.4,
+      "epoch": 0.2451059699077819,
+      "grad_norm": 0.8146517496395704,
+      "learning_rate": 9.373235446587055e-06,
+      "loss": 0.8678,
       "step": 1515
     },
     {
-      "epoch": 0.6296603148301574,
-      "grad_norm": 2.558523087497906,
-      "learning_rate": 3.6292301871135425e-06,
-      "loss": 0.3919,
+      "epoch": 0.24591490050153697,
+      "grad_norm": 0.7819896556165521,
+      "learning_rate": 9.366372819826553e-06,
+      "loss": 0.8572,
       "step": 1520
     },
     {
-      "epoch": 0.6317315658657829,
-      "grad_norm": 2.599732843239929,
-      "learning_rate": 3.59449164699773e-06,
-      "loss": 0.4016,
+      "epoch": 0.24672383109529203,
+      "grad_norm": 0.7667264574062115,
+      "learning_rate": 9.359475367489805e-06,
+      "loss": 0.8627,
       "step": 1525
     },
     {
-      "epoch": 0.6338028169014085,
-      "grad_norm": 2.4726991401898952,
-      "learning_rate": 3.55982661776676e-06,
-      "loss": 0.3986,
+      "epoch": 0.2475327616890471,
+      "grad_norm": 0.8236534571118577,
+      "learning_rate": 9.352543144589923e-06,
+      "loss": 0.8668,
       "step": 1530
     },
     {
-      "epoch": 0.6358740679370339,
-      "grad_norm": 2.321690978258901,
-      "learning_rate": 3.5252369124706697e-06,
-      "loss": 0.3978,
+      "epoch": 0.24834169228280215,
+      "grad_norm": 0.7670049125342286,
+      "learning_rate": 9.345576206417345e-06,
+      "loss": 0.8418,
       "step": 1535
     },
     {
-      "epoch": 0.6379453189726595,
-      "grad_norm": 2.6754380236168664,
-      "learning_rate": 3.4907243402199013e-06,
-      "loss": 0.4113,
+      "epoch": 0.2491506228765572,
+      "grad_norm": 0.7827102030650855,
+      "learning_rate": 9.338574608539389e-06,
+      "loss": 0.8439,
       "step": 1540
     },
     {
-      "epoch": 0.640016570008285,
-      "grad_norm": 2.4052468560081475,
-      "learning_rate": 3.4562907060906908e-06,
-      "loss": 0.3813,
+      "epoch": 0.24995955347031223,
+      "grad_norm": 0.7839079136756119,
+      "learning_rate": 9.331538406799815e-06,
+      "loss": 0.8549,
       "step": 1545
     },
     {
-      "epoch": 0.6420878210439105,
-      "grad_norm": 2.4570458861707696,
-      "learning_rate": 3.4219378110306523e-06,
-      "loss": 0.392,
+      "epoch": 0.2507684840640673,
+      "grad_norm": 0.8178409940371506,
+      "learning_rate": 9.324467657318384e-06,
+      "loss": 0.8534,
       "step": 1550
     },
     {
-      "epoch": 0.644159072079536,
-      "grad_norm": 2.3101859891504866,
-      "learning_rate": 3.3876674517645815e-06,
-      "loss": 0.3996,
+      "epoch": 0.25157741465782235,
+      "grad_norm": 0.8099902787097547,
+      "learning_rate": 9.317362416490396e-06,
+      "loss": 0.8589,
       "step": 1555
     },
     {
-      "epoch": 0.6462303231151616,
-      "grad_norm": 2.4539409882344483,
-      "learning_rate": 3.353481420700495e-06,
-      "loss": 0.3979,
+      "epoch": 0.25238634525157744,
+      "grad_norm": 0.8397966939088385,
+      "learning_rate": 9.310222740986258e-06,
+      "loss": 0.8695,
       "step": 1560
     },
     {
-      "epoch": 0.648301574150787,
-      "grad_norm": 2.2847760045122087,
-      "learning_rate": 3.319381505835868e-06,
-      "loss": 0.3749,
+      "epoch": 0.25319527584533247,
+      "grad_norm": 0.8012721851788485,
+      "learning_rate": 9.303048687751016e-06,
+      "loss": 0.8647,
       "step": 1565
     },
     {
-      "epoch": 0.6503728251864126,
-      "grad_norm": 2.2918085600506717,
-      "learning_rate": 3.285369490664133e-06,
-      "loss": 0.3895,
+      "epoch": 0.2540042064390875,
+      "grad_norm": 0.7848654081297167,
+      "learning_rate": 9.29584031400391e-06,
+      "loss": 0.848,
       "step": 1570
     },
     {
-      "epoch": 0.6524440762220381,
-      "grad_norm": 2.4126038265192538,
-      "learning_rate": 3.251447154081394e-06,
-      "loss": 0.3885,
+      "epoch": 0.2548131370328426,
+      "grad_norm": 0.7276468748203139,
+      "learning_rate": 9.288597677237918e-06,
+      "loss": 0.8451,
       "step": 1575
     },
     {
-      "epoch": 0.6545153272576636,
-      "grad_norm": 2.4831233203755603,
-      "learning_rate": 3.2176162702933816e-06,
-      "loss": 0.413,
+      "epoch": 0.2556220676265976,
+      "grad_norm": 0.8006865536918176,
+      "learning_rate": 9.281320835219294e-06,
+      "loss": 0.847,
       "step": 1580
     },
     {
-      "epoch": 0.6565865782932891,
-      "grad_norm": 2.385955156565656,
-      "learning_rate": 3.183878608722669e-06,
-      "loss": 0.3838,
+      "epoch": 0.2564309982203527,
+      "grad_norm": 0.7628111941438427,
+      "learning_rate": 9.274009845987106e-06,
+      "loss": 0.8712,
       "step": 1585
     },
     {
-      "epoch": 0.6586578293289147,
-      "grad_norm": 2.5426984744026124,
-      "learning_rate": 3.150235933916115e-06,
-      "loss": 0.3755,
+      "epoch": 0.25723992881410773,
+      "grad_norm": 0.7881647592364686,
+      "learning_rate": 9.26666476785278e-06,
+      "loss": 0.8678,
       "step": 1590
     },
     {
-      "epoch": 0.6607290803645401,
-      "grad_norm": 2.388904551962294,
-      "learning_rate": 3.1166900054525873e-06,
-      "loss": 0.3868,
+      "epoch": 0.2580488594078628,
+      "grad_norm": 0.8116538712282235,
+      "learning_rate": 9.259285659399624e-06,
+      "loss": 0.8535,
       "step": 1595
     },
     {
-      "epoch": 0.6628003314001657,
-      "grad_norm": 2.5869530818499147,
-      "learning_rate": 3.0832425778509235e-06,
-      "loss": 0.3821,
+      "epoch": 0.25885779000161785,
+      "grad_norm": 0.7813279263856877,
+      "learning_rate": 9.251872579482373e-06,
+      "loss": 0.845,
       "step": 1600
     },
     {
-      "epoch": 0.6648715824357913,
-      "grad_norm": 2.47920960360778,
-      "learning_rate": 3.049895400478174e-06,
-      "loss": 0.3935,
+      "epoch": 0.25966672059537294,
+      "grad_norm": 0.7940356068886795,
+      "learning_rate": 9.244425587226708e-06,
+      "loss": 0.8492,
       "step": 1605
     },
     {
-      "epoch": 0.6669428334714167,
-      "grad_norm": 2.4320164826856234,
-      "learning_rate": 3.0166502174581012e-06,
-      "loss": 0.3732,
+      "epoch": 0.26047565118912797,
+      "grad_norm": 0.8041883146042144,
+      "learning_rate": 9.236944742028797e-06,
+      "loss": 0.869,
       "step": 1610
     },
     {
-      "epoch": 0.6690140845070423,
-      "grad_norm": 2.3447158283526783,
-      "learning_rate": 2.983508767579956e-06,
-      "loss": 0.3805,
+      "epoch": 0.26128458178288305,
+      "grad_norm": 0.8411011003169626,
+      "learning_rate": 9.229430103554808e-06,
+      "loss": 0.8702,
       "step": 1615
     },
     {
-      "epoch": 0.6710853355426678,
-      "grad_norm": 2.336702612441604,
-      "learning_rate": 2.950472784207544e-06,
-      "loss": 0.3814,
+      "epoch": 0.2620935123766381,
+      "grad_norm": 0.7994752148577935,
+      "learning_rate": 9.221881731740442e-06,
+      "loss": 0.846,
       "step": 1620
     },
     {
-      "epoch": 0.6731565865782932,
-      "grad_norm": 2.403683822019676,
-      "learning_rate": 2.917543995188562e-06,
-      "loss": 0.3799,
+      "epoch": 0.2629024429703931,
+      "grad_norm": 0.7904404440251053,
+      "learning_rate": 9.214299686790453e-06,
+      "loss": 0.8479,
       "step": 1625
     },
     {
-      "epoch": 0.6752278376139188,
-      "grad_norm": 2.32245444127231,
-      "learning_rate": 2.8847241227642255e-06,
-      "loss": 0.3878,
+      "epoch": 0.2637113735641482,
+      "grad_norm": 0.8285353436413048,
+      "learning_rate": 9.206684029178166e-06,
+      "loss": 0.867,
       "step": 1630
     },
     {
-      "epoch": 0.6772990886495444,
-      "grad_norm": 2.2336601101273894,
-      "learning_rate": 2.852014883479198e-06,
-      "loss": 0.3668,
+      "epoch": 0.26452030415790323,
+      "grad_norm": 0.8167896861662431,
+      "learning_rate": 9.199034819644997e-06,
+      "loss": 0.845,
       "step": 1635
     },
     {
-      "epoch": 0.6793703396851698,
-      "grad_norm": 2.4572013362431666,
-      "learning_rate": 2.819417988091814e-06,
-      "loss": 0.3789,
+      "epoch": 0.2653292347516583,
+      "grad_norm": 0.7625190041745681,
+      "learning_rate": 9.191352119199965e-06,
+      "loss": 0.8715,
       "step": 1640
     },
     {
-      "epoch": 0.6814415907207954,
-      "grad_norm": 2.374979599510779,
-      "learning_rate": 2.786935141484586e-06,
-      "loss": 0.3776,
+      "epoch": 0.26613816534541335,
+      "grad_norm": 0.766576459090851,
+      "learning_rate": 9.183635989119211e-06,
+      "loss": 0.8554,
       "step": 1645
     },
     {
-      "epoch": 0.6835128417564209,
-      "grad_norm": 2.387695271521185,
-      "learning_rate": 2.754568042575061e-06,
-      "loss": 0.3785,
+      "epoch": 0.26694709593916843,
+      "grad_norm": 0.7598503790298559,
+      "learning_rate": 9.175886490945505e-06,
+      "loss": 0.8657,
       "step": 1650
     },
     {
-      "epoch": 0.6855840927920464,
-      "grad_norm": 2.4380470816357573,
-      "learning_rate": 2.7223183842269442e-06,
-      "loss": 0.3762,
+      "epoch": 0.26775602653292346,
+      "grad_norm": 0.7436415672250105,
+      "learning_rate": 9.168103686487755e-06,
+      "loss": 0.8618,
       "step": 1655
     },
     {
-      "epoch": 0.6876553438276719,
-      "grad_norm": 2.2802764216409996,
-      "learning_rate": 2.6901878531615677e-06,
-      "loss": 0.377,
+      "epoch": 0.26856495712667855,
+      "grad_norm": 0.8147678677174188,
+      "learning_rate": 9.160287637820514e-06,
+      "loss": 0.8591,
       "step": 1660
     },
     {
-      "epoch": 0.6897265948632975,
-      "grad_norm": 2.3338582614060748,
-      "learning_rate": 2.658178129869672e-06,
-      "loss": 0.382,
+      "epoch": 0.2693738877204336,
+      "grad_norm": 0.7963714833005615,
+      "learning_rate": 9.152438407283493e-06,
+      "loss": 0.8651,
       "step": 1665
     },
     {
-      "epoch": 0.6917978458989229,
-      "grad_norm": 2.4939738702851932,
-      "learning_rate": 2.6262908885235046e-06,
-      "loss": 0.3789,
+      "epoch": 0.27018281831418867,
+      "grad_norm": 0.7937419589693009,
+      "learning_rate": 9.144556057481048e-06,
+      "loss": 0.8373,
       "step": 1670
     },
     {
-      "epoch": 0.6938690969345485,
-      "grad_norm": 2.409154689853176,
-      "learning_rate": 2.594527796889265e-06,
-      "loss": 0.3859,
+      "epoch": 0.2709917489079437,
+      "grad_norm": 0.8239141053085791,
+      "learning_rate": 9.136640651281694e-06,
+      "loss": 0.8712,
       "step": 1675
     },
     {
-      "epoch": 0.695940347970174,
-      "grad_norm": 2.4232651472750724,
-      "learning_rate": 2.5628905162398797e-06,
-      "loss": 0.3759,
+      "epoch": 0.2718006795016987,
+      "grad_norm": 0.7797626292094625,
+      "learning_rate": 9.128692251817602e-06,
+      "loss": 0.8629,
       "step": 1680
     },
     {
-      "epoch": 0.6980115990057995,
-      "grad_norm": 2.438606094251203,
-      "learning_rate": 2.531380701268108e-06,
-      "loss": 0.3707,
+      "epoch": 0.2726096100954538,
+      "grad_norm": 0.7697095287067577,
+      "learning_rate": 9.120710922484089e-06,
+      "loss": 0.8307,
       "step": 1685
     },
     {
-      "epoch": 0.700082850041425,
-      "grad_norm": 2.2068267980004164,
-      "learning_rate": 2.5000000000000015e-06,
-      "loss": 0.3635,
+      "epoch": 0.27341854068920884,
+      "grad_norm": 0.7710196952612974,
+      "learning_rate": 9.112696726939112e-06,
+      "loss": 0.8513,
       "step": 1690
     },
     {
-      "epoch": 0.7021541010770506,
-      "grad_norm": 2.4849851469693487,
-      "learning_rate": 2.4687500537087027e-06,
-      "loss": 0.3765,
+      "epoch": 0.27422747128296393,
+      "grad_norm": 0.7800639021416464,
+      "learning_rate": 9.104649729102774e-06,
+      "loss": 0.8582,
       "step": 1695
     },
     {
-      "epoch": 0.704225352112676,
-      "grad_norm": 2.3676710626755604,
-      "learning_rate": 2.4376324968286154e-06,
-      "loss": 0.3706,
+      "epoch": 0.27503640187671896,
+      "grad_norm": 0.8044706731480418,
+      "learning_rate": 9.096569993156797e-06,
+      "loss": 0.8406,
       "step": 1700
     },
     {
-      "epoch": 0.7062966031483016,
-      "grad_norm": 2.4665247552413194,
-      "learning_rate": 2.40664895686991e-06,
-      "loss": 0.359,
+      "epoch": 0.27584533247047405,
+      "grad_norm": 0.7699086496292621,
+      "learning_rate": 9.088457583544022e-06,
+      "loss": 0.8562,
       "step": 1705
     },
     {
-      "epoch": 0.7083678541839271,
-      "grad_norm": 2.4228989801005882,
-      "learning_rate": 2.375801054333409e-06,
-      "loss": 0.3574,
+      "epoch": 0.2766542630642291,
+      "grad_norm": 0.8340399404633732,
+      "learning_rate": 9.080312564967884e-06,
+      "loss": 0.8694,
       "step": 1710
     },
     {
-      "epoch": 0.7104391052195526,
-      "grad_norm": 2.447240340812193,
-      "learning_rate": 2.345090402625822e-06,
-      "loss": 0.3629,
+      "epoch": 0.27746319365798416,
+      "grad_norm": 0.7919025521073914,
+      "learning_rate": 9.072135002391912e-06,
+      "loss": 0.8769,
       "step": 1715
     },
     {
-      "epoch": 0.7125103562551781,
-      "grad_norm": 2.494665611756937,
-      "learning_rate": 2.3145186079753685e-06,
-      "loss": 0.3607,
+      "epoch": 0.2782721242517392,
+      "grad_norm": 0.7754055380467696,
+      "learning_rate": 9.063924961039195e-06,
+      "loss": 0.8852,
       "step": 1720
     },
     {
-      "epoch": 0.7145816072908037,
-      "grad_norm": 2.3466738040218433,
-      "learning_rate": 2.2840872693477694e-06,
-      "loss": 0.3582,
+      "epoch": 0.2790810548454943,
+      "grad_norm": 0.76639361552101,
+      "learning_rate": 9.055682506391866e-06,
+      "loss": 0.8655,
       "step": 1725
     },
     {
-      "epoch": 0.7166528583264291,
-      "grad_norm": 2.4579034785541274,
-      "learning_rate": 2.253797978362617e-06,
-      "loss": 0.37,
+      "epoch": 0.2798899854392493,
+      "grad_norm": 0.7860193334912077,
+      "learning_rate": 9.04740770419059e-06,
+      "loss": 0.8572,
       "step": 1730
     },
     {
-      "epoch": 0.7187241093620547,
-      "grad_norm": 2.294277539263346,
-      "learning_rate": 2.2236523192101264e-06,
-      "loss": 0.371,
+      "epoch": 0.28069891603300434,
+      "grad_norm": 0.7656637491035695,
+      "learning_rate": 9.039100620434025e-06,
+      "loss": 0.8634,
       "step": 1735
     },
     {
-      "epoch": 0.7207953603976802,
-      "grad_norm": 2.3165105524366676,
-      "learning_rate": 2.193651868568285e-06,
-      "loss": 0.3533,
+      "epoch": 0.28150784662675943,
+      "grad_norm": 0.8287569040947624,
+      "learning_rate": 9.030761321378303e-06,
+      "loss": 0.8539,
       "step": 1740
     },
     {
-      "epoch": 0.7228666114333057,
-      "grad_norm": 2.3412049707939766,
-      "learning_rate": 2.16379819552038e-06,
-      "loss": 0.3598,
+      "epoch": 0.28231677722051446,
+      "grad_norm": 0.778379469664662,
+      "learning_rate": 9.022389873536505e-06,
+      "loss": 0.8543,
       "step": 1745
     },
     {
-      "epoch": 0.7249378624689312,
-      "grad_norm": 2.37970184238566,
-      "learning_rate": 2.1340928614729445e-06,
-      "loss": 0.3553,
+      "epoch": 0.28312570781426954,
+      "grad_norm": 0.8341873388515572,
+      "learning_rate": 9.01398634367812e-06,
+      "loss": 0.8378,
       "step": 1750
     },
     {
-      "epoch": 0.7270091135045568,
-      "grad_norm": 2.5101331147915733,
-      "learning_rate": 2.1045374200740863e-06,
-      "loss": 0.3589,
+      "epoch": 0.2839346384080246,
+      "grad_norm": 0.807630286757475,
+      "learning_rate": 9.005550798828521e-06,
+      "loss": 0.8661,
       "step": 1755
     },
     {
-      "epoch": 0.7290803645401823,
-      "grad_norm": 2.381929011616353,
-      "learning_rate": 2.075133417132223e-06,
-      "loss": 0.3544,
+      "epoch": 0.28474356900177966,
+      "grad_norm": 0.7492013925485693,
+      "learning_rate": 8.997083306268434e-06,
+      "loss": 0.8646,
       "step": 1760
     },
     {
-      "epoch": 0.7311516155758078,
-      "grad_norm": 2.64277485599053,
-      "learning_rate": 2.045882390535248e-06,
-      "loss": 0.354,
+      "epoch": 0.2855524995955347,
+      "grad_norm": 0.820329567734956,
+      "learning_rate": 8.988583933533384e-06,
+      "loss": 0.866,
       "step": 1765
     },
     {
-      "epoch": 0.7332228666114333,
-      "grad_norm": 2.2983110375410756,
-      "learning_rate": 2.016785870170079e-06,
-      "loss": 0.359,
+      "epoch": 0.2863614301892898,
+      "grad_norm": 0.8427359735718537,
+      "learning_rate": 8.980052748413177e-06,
+      "loss": 0.8393,
       "step": 1770
     },
     {
-      "epoch": 0.7352941176470589,
-      "grad_norm": 2.338818197474944,
-      "learning_rate": 1.987845377842656e-06,
-      "loss": 0.3515,
+      "epoch": 0.2871703607830448,
+      "grad_norm": 0.7315290649122431,
+      "learning_rate": 8.971489818951347e-06,
+      "loss": 0.8561,
       "step": 1775
     },
     {
-      "epoch": 0.7373653686826843,
-      "grad_norm": 2.412156205295859,
-      "learning_rate": 1.9590624271983406e-06,
-      "loss": 0.3707,
+      "epoch": 0.2879792913767999,
+      "grad_norm": 0.7968151029522487,
+      "learning_rate": 8.962895213444618e-06,
+      "loss": 0.8449,
       "step": 1780
     },
     {
-      "epoch": 0.7394366197183099,
-      "grad_norm": 2.3221544897617394,
-      "learning_rate": 1.9304385236427505e-06,
-      "loss": 0.3522,
+      "epoch": 0.2887882219705549,
+      "grad_norm": 0.8497262991663913,
+      "learning_rate": 8.954269000442353e-06,
+      "loss": 0.8614,
       "step": 1785
     },
     {
-      "epoch": 0.7415078707539354,
-      "grad_norm": 2.4316222609963427,
-      "learning_rate": 1.9019751642630252e-06,
-      "loss": 0.3574,
+      "epoch": 0.28959715256430996,
+      "grad_norm": 0.8884396786756781,
+      "learning_rate": 8.945611248746015e-06,
+      "loss": 0.873,
       "step": 1790
     },
     {
-      "epoch": 0.7435791217895609,
-      "grad_norm": 2.2641479623423635,
-      "learning_rate": 1.8736738377495196e-06,
-      "loss": 0.3592,
+      "epoch": 0.29040608315806504,
+      "grad_norm": 0.7822894663944296,
+      "learning_rate": 8.936922027408618e-06,
+      "loss": 0.8504,
       "step": 1795
     },
     {
-      "epoch": 0.7456503728251864,
-      "grad_norm": 2.3420647549229265,
-      "learning_rate": 1.8455360243179537e-06,
-      "loss": 0.3542,
+      "epoch": 0.2912150137518201,
+      "grad_norm": 0.7512013193436944,
+      "learning_rate": 8.928201405734172e-06,
+      "loss": 0.8638,
       "step": 1800
     },
     {
-      "epoch": 0.747721623860812,
-      "grad_norm": 2.456638840774933,
-      "learning_rate": 1.8175631956319823e-06,
-      "loss": 0.3542,
+      "epoch": 0.29202394434557516,
+      "grad_norm": 0.808524098300231,
+      "learning_rate": 8.919449453277124e-06,
+      "loss": 0.8555,
       "step": 1805
     },
     {
-      "epoch": 0.7497928748964374,
-      "grad_norm": 2.315911960148082,
-      "learning_rate": 1.7897568147262323e-06,
-      "loss": 0.3651,
+      "epoch": 0.2928328749393302,
+      "grad_norm": 0.8205569397216151,
+      "learning_rate": 8.910666239841824e-06,
+      "loss": 0.8393,
       "step": 1810
     },
     {
-      "epoch": 0.751864125932063,
-      "grad_norm": 2.430456263244821,
-      "learning_rate": 1.7621183359297817e-06,
-      "loss": 0.3513,
+      "epoch": 0.2936418055330853,
+      "grad_norm": 0.7685972280574218,
+      "learning_rate": 8.901851835481947e-06,
+      "loss": 0.8616,
       "step": 1815
     },
     {
-      "epoch": 0.7539353769676885,
-      "grad_norm": 2.2820850020652137,
-      "learning_rate": 1.7346492047900897e-06,
-      "loss": 0.344,
+      "epoch": 0.2944507361268403,
+      "grad_norm": 0.7888992044706635,
+      "learning_rate": 8.893006310499941e-06,
+      "loss": 0.827,
       "step": 1820
     },
     {
-      "epoch": 0.756006628003314,
-      "grad_norm": 2.4006568265757755,
-      "learning_rate": 1.7073508579973996e-06,
-      "loss": 0.3526,
+      "epoch": 0.2952596667205954,
+      "grad_norm": 0.8377015460290415,
+      "learning_rate": 8.884129735446471e-06,
+      "loss": 0.8699,
       "step": 1825
     },
     {
-      "epoch": 0.7580778790389395,
-      "grad_norm": 2.2901041487912774,
-      "learning_rate": 1.6802247233095914e-06,
-      "loss": 0.349,
+      "epoch": 0.2960685973143504,
+      "grad_norm": 0.8237212938659825,
+      "learning_rate": 8.875222181119859e-06,
+      "loss": 0.8503,
       "step": 1830
     },
     {
-      "epoch": 0.7601491300745651,
-      "grad_norm": 2.4009693108682204,
-      "learning_rate": 1.6532722194775108e-06,
-      "loss": 0.3537,
+      "epoch": 0.2968775279081055,
+      "grad_norm": 0.7665032383816565,
+      "learning_rate": 8.866283718565498e-06,
+      "loss": 0.8589,
       "step": 1835
     },
     {
-      "epoch": 0.7622203811101905,
-      "grad_norm": 2.343710211021654,
-      "learning_rate": 1.626494756170765e-06,
-      "loss": 0.3536,
+      "epoch": 0.29768645850186054,
+      "grad_norm": 0.8141188029046369,
+      "learning_rate": 8.857314419075316e-06,
+      "loss": 0.8571,
       "step": 1840
     },
     {
-      "epoch": 0.7642916321458161,
-      "grad_norm": 2.3742949969171363,
-      "learning_rate": 1.5998937339039889e-06,
-      "loss": 0.3542,
+      "epoch": 0.29849538909561557,
+      "grad_norm": 0.7761079464199027,
+      "learning_rate": 8.848314354187184e-06,
+      "loss": 0.8463,
       "step": 1845
     },
     {
-      "epoch": 0.7663628831814416,
-      "grad_norm": 2.3672793937996093,
-      "learning_rate": 1.5734705439636017e-06,
-      "loss": 0.3428,
+      "epoch": 0.29930431968937066,
+      "grad_norm": 0.8097442274081034,
+      "learning_rate": 8.839283595684355e-06,
+      "loss": 0.8581,
       "step": 1850
     },
     {
-      "epoch": 0.7684341342170671,
-      "grad_norm": 2.2889634008759803,
-      "learning_rate": 1.5472265683350397e-06,
-      "loss": 0.3535,
+      "epoch": 0.3001132502831257,
+      "grad_norm": 0.8095252549308872,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.8557,
       "step": 1855
     },
     {
-      "epoch": 0.7705053852526926,
-      "grad_norm": 2.2816526862864164,
-      "learning_rate": 1.5211631796304721e-06,
-      "loss": 0.3452,
+      "epoch": 0.3009221808768808,
+      "grad_norm": 0.7980409130367505,
+      "learning_rate": 8.821130286191086e-06,
+      "loss": 0.8671,
       "step": 1860
     },
     {
-      "epoch": 0.7725766362883182,
-      "grad_norm": 2.256494769623704,
-      "learning_rate": 1.495281741017016e-06,
-      "loss": 0.3569,
+      "epoch": 0.3017311114706358,
+      "grad_norm": 0.8173815138889032,
+      "learning_rate": 8.81200787998889e-06,
+      "loss": 0.8594,
       "step": 1865
     },
     {
-      "epoch": 0.7746478873239436,
-      "grad_norm": 2.2248161459647333,
-      "learning_rate": 1.46958360614543e-06,
-      "loss": 0.3439,
+      "epoch": 0.3025400420643909,
+      "grad_norm": 0.8047823405503607,
+      "learning_rate": 8.802855069747338e-06,
+      "loss": 0.8596,
       "step": 1870
     },
     {
-      "epoch": 0.7767191383595692,
-      "grad_norm": 2.3902740970389065,
-      "learning_rate": 1.4440701190793278e-06,
-      "loss": 0.3301,
+      "epoch": 0.3033489726581459,
+      "grad_norm": 0.804911943505487,
+      "learning_rate": 8.793671928467953e-06,
+      "loss": 0.8397,
       "step": 1875
     },
     {
-      "epoch": 0.7787903893951947,
-      "grad_norm": 2.439067358960243,
-      "learning_rate": 1.4187426142248723e-06,
-      "loss": 0.3462,
+      "epoch": 0.304157903251901,
+      "grad_norm": 0.7613233673132125,
+      "learning_rate": 8.784458529394185e-06,
+      "loss": 0.8407,
       "step": 1880
     },
     {
-      "epoch": 0.7808616404308202,
-      "grad_norm": 2.3392746632984585,
-      "learning_rate": 1.3936024162609897e-06,
-      "loss": 0.3408,
+      "epoch": 0.30496683384565604,
+      "grad_norm": 0.7397028565696792,
+      "learning_rate": 8.775214946010806e-06,
+      "loss": 0.8476,
       "step": 1885
     },
     {
-      "epoch": 0.7829328914664457,
-      "grad_norm": 2.377498700024158,
-      "learning_rate": 1.3686508400700787e-06,
-      "loss": 0.3549,
+      "epoch": 0.3057757644394111,
+      "grad_norm": 0.7728205441604445,
+      "learning_rate": 8.765941252043341e-06,
+      "loss": 0.8384,
       "step": 1890
     },
     {
-      "epoch": 0.7850041425020713,
-      "grad_norm": 2.3496620916951767,
-      "learning_rate": 1.3438891906692447e-06,
-      "loss": 0.3472,
+      "epoch": 0.30658469503316615,
+      "grad_norm": 0.8229918467819616,
+      "learning_rate": 8.756637521457473e-06,
+      "loss": 0.8488,
       "step": 1895
     },
     {
-      "epoch": 0.7870753935376967,
-      "grad_norm": 2.289494721325453,
-      "learning_rate": 1.3193187631420462e-06,
-      "loss": 0.3355,
+      "epoch": 0.3073936256269212,
+      "grad_norm": 0.7848703050564307,
+      "learning_rate": 8.747303828458446e-06,
+      "loss": 0.8488,
       "step": 1900
     },
     {
-      "epoch": 0.7891466445733223,
-      "grad_norm": 2.2947285370456396,
-      "learning_rate": 1.2949408425707566e-06,
-      "loss": 0.3394,
+      "epoch": 0.30820255622067627,
+      "grad_norm": 0.8185735473058204,
+      "learning_rate": 8.737940247490487e-06,
+      "loss": 0.8447,
       "step": 1905
     },
     {
-      "epoch": 0.7912178956089478,
-      "grad_norm": 2.3617970794893126,
-      "learning_rate": 1.2707567039691505e-06,
-      "loss": 0.3497,
+      "epoch": 0.3090114868144313,
+      "grad_norm": 0.8175009758279596,
+      "learning_rate": 8.728546853236202e-06,
+      "loss": 0.8468,
       "step": 1910
     },
     {
-      "epoch": 0.7932891466445733,
-      "grad_norm": 2.321811184391228,
-      "learning_rate": 1.2467676122158224e-06,
-      "loss": 0.3412,
+      "epoch": 0.3098204174081864,
+      "grad_norm": 0.8234001884738343,
+      "learning_rate": 8.71912372061598e-06,
+      "loss": 0.8579,
       "step": 1915
     },
     {
-      "epoch": 0.7953603976801988,
-      "grad_norm": 2.3211063248548967,
-      "learning_rate": 1.222974821988024e-06,
-      "loss": 0.3436,
+      "epoch": 0.3106293480019414,
+      "grad_norm": 0.8135018815005348,
+      "learning_rate": 8.70967092478741e-06,
+      "loss": 0.8333,
       "step": 1920
     },
     {
-      "epoch": 0.7974316487158244,
-      "grad_norm": 2.367048239376449,
-      "learning_rate": 1.1993795776960498e-06,
-      "loss": 0.3387,
+      "epoch": 0.3114382785956965,
+      "grad_norm": 0.8316517603780106,
+      "learning_rate": 8.700188541144658e-06,
+      "loss": 0.8152,
       "step": 1925
     },
     {
-      "epoch": 0.7995028997514498,
-      "grad_norm": 2.415017409117694,
-      "learning_rate": 1.1759831134181504e-06,
-      "loss": 0.3477,
+      "epoch": 0.31224720918945154,
+      "grad_norm": 0.8542272626940559,
+      "learning_rate": 8.690676645317886e-06,
+      "loss": 0.8302,
       "step": 1930
     },
     {
-      "epoch": 0.8015741507870754,
-      "grad_norm": 2.2878955632108737,
-      "learning_rate": 1.1527866528359805e-06,
-      "loss": 0.3424,
+      "epoch": 0.3130561397832066,
+      "grad_norm": 0.7850582360070165,
+      "learning_rate": 8.68113531317264e-06,
+      "loss": 0.8494,
       "step": 1935
     },
     {
-      "epoch": 0.8036454018227009,
-      "grad_norm": 2.4168712983822407,
-      "learning_rate": 1.1297914091706086e-06,
-      "loss": 0.3395,
+      "epoch": 0.31386507037696165,
+      "grad_norm": 0.8046763932380032,
+      "learning_rate": 8.671564620809243e-06,
+      "loss": 0.8512,
       "step": 1940
     },
     {
-      "epoch": 0.8057166528583264,
-      "grad_norm": 2.332403241899359,
-      "learning_rate": 1.1069985851190524e-06,
-      "loss": 0.3413,
+      "epoch": 0.31467400097071674,
+      "grad_norm": 0.8472122957190853,
+      "learning_rate": 8.661964644562194e-06,
+      "loss": 0.8481,
       "step": 1945
     },
     {
-      "epoch": 0.8077879038939519,
-      "grad_norm": 2.245314869032349,
-      "learning_rate": 1.0844093727913868e-06,
-      "loss": 0.3302,
+      "epoch": 0.31548293156447177,
+      "grad_norm": 0.8098454336630103,
+      "learning_rate": 8.652335460999554e-06,
+      "loss": 0.8386,
       "step": 1950
     },
     {
-      "epoch": 0.8098591549295775,
-      "grad_norm": 2.3766452170038814,
-      "learning_rate": 1.062024953648384e-06,
-      "loss": 0.3363,
+      "epoch": 0.3162918621582268,
+      "grad_norm": 0.81377246582292,
+      "learning_rate": 8.64267714692234e-06,
+      "loss": 0.8109,
       "step": 1955
     },
     {
-      "epoch": 0.8119304059652029,
-      "grad_norm": 2.3725445423461076,
-      "learning_rate": 1.039846498439727e-06,
-      "loss": 0.3312,
+      "epoch": 0.3171007927519819,
+      "grad_norm": 0.7941773089013205,
+      "learning_rate": 8.632989779363907e-06,
+      "loss": 0.8223,
       "step": 1960
     },
     {
-      "epoch": 0.8140016570008285,
-      "grad_norm": 2.400378472961016,
-      "learning_rate": 1.0178751671427755e-06,
-      "loss": 0.3406,
+      "epoch": 0.3179097233457369,
+      "grad_norm": 0.7988879204951932,
+      "learning_rate": 8.623273435589338e-06,
+      "loss": 0.8536,
       "step": 1965
     },
     {
-      "epoch": 0.816072908036454,
-      "grad_norm": 2.600161757489673,
-      "learning_rate": 9.961121089018933e-07,
-      "loss": 0.3286,
+      "epoch": 0.318718653939492,
+      "grad_norm": 0.7645549102866793,
+      "learning_rate": 8.613528193094826e-06,
+      "loss": 0.841,
       "step": 1970
     },
     {
-      "epoch": 0.8181441590720795,
-      "grad_norm": 2.4280753174812597,
-      "learning_rate": 9.745584619683524e-07,
-      "loss": 0.3439,
+      "epoch": 0.31952758453324703,
+      "grad_norm": 0.7971671512555748,
+      "learning_rate": 8.603754129607055e-06,
+      "loss": 0.8418,
       "step": 1975
     },
     {
-      "epoch": 0.820215410107705,
-      "grad_norm": 2.260545698321657,
-      "learning_rate": 9.532153536407923e-07,
-      "loss": 0.3317,
+      "epoch": 0.3203365151270021,
+      "grad_norm": 0.7570319344787492,
+      "learning_rate": 8.593951323082586e-06,
+      "loss": 0.8583,
       "step": 1980
     },
     {
-      "epoch": 0.8222866611433306,
-      "grad_norm": 2.4227916089483204,
-      "learning_rate": 9.320839002062682e-07,
-      "loss": 0.3312,
+      "epoch": 0.32114544572075715,
+      "grad_norm": 0.8210274942358864,
+      "learning_rate": 8.584119851707224e-06,
+      "loss": 0.8493,
       "step": 1985
     },
     {
-      "epoch": 0.824357912178956,
-      "grad_norm": 2.8477868846172067,
-      "learning_rate": 9.111652068818621e-07,
-      "loss": 0.3334,
+      "epoch": 0.32195437631451224,
+      "grad_norm": 0.7921398145158576,
+      "learning_rate": 8.574259793895404e-06,
+      "loss": 0.8373,
       "step": 1990
     },
     {
-      "epoch": 0.8264291632145816,
-      "grad_norm": 2.370728174821394,
-      "learning_rate": 8.904603677568785e-07,
-      "loss": 0.3278,
+      "epoch": 0.32276330690826727,
+      "grad_norm": 0.8093679320396422,
+      "learning_rate": 8.564371228289563e-06,
+      "loss": 0.8353,
       "step": 1995
     },
     {
-      "epoch": 0.8285004142502072,
-      "grad_norm": 2.3115016796994183,
-      "learning_rate": 8.699704657356195e-07,
-      "loss": 0.3358,
+      "epoch": 0.32357223750202235,
+      "grad_norm": 0.7347855423064408,
+      "learning_rate": 8.554454233759508e-06,
+      "loss": 0.856,
       "step": 2000
     },
     {
-      "epoch": 0.8305716652858326,
-      "grad_norm": 2.253238724272844,
-      "learning_rate": 8.496965724807516e-07,
-      "loss": 0.3348,
+      "epoch": 0.3243811680957774,
+      "grad_norm": 0.7460223907838642,
+      "learning_rate": 8.544508889401799e-06,
+      "loss": 0.8437,
       "step": 2005
     },
     {
-      "epoch": 0.8326429163214581,
-      "grad_norm": 2.3696595779036573,
-      "learning_rate": 8.296397483572515e-07,
-      "loss": 0.3322,
+      "epoch": 0.3251900986895324,
+      "grad_norm": 0.7539924040446289,
+      "learning_rate": 8.534535274539103e-06,
+      "loss": 0.8448,
       "step": 2010
     },
     {
-      "epoch": 0.8347141673570837,
-      "grad_norm": 2.3994710782454143,
-      "learning_rate": 8.098010423769503e-07,
-      "loss": 0.3159,
+      "epoch": 0.3259990292832875,
+      "grad_norm": 0.8431518796240063,
+      "learning_rate": 8.524533468719569e-06,
+      "loss": 0.8545,
       "step": 2015
     },
     {
-      "epoch": 0.8367854183927091,
-      "grad_norm": 2.1347268704516638,
-      "learning_rate": 7.901814921436624e-07,
-      "loss": 0.3318,
+      "epoch": 0.32680795987704253,
+      "grad_norm": 0.8116074595962663,
+      "learning_rate": 8.5145035517162e-06,
+      "loss": 0.8262,
       "step": 2020
     },
     {
-      "epoch": 0.8388566694283347,
-      "grad_norm": 2.2628490853732397,
-      "learning_rate": 7.70782123798921e-07,
-      "loss": 0.3318,
+      "epoch": 0.3276168904707976,
+      "grad_norm": 0.7709832113170016,
+      "learning_rate": 8.504445603526202e-06,
+      "loss": 0.8718,
       "step": 2025
     },
     {
-      "epoch": 0.8409279204639603,
-      "grad_norm": 2.2582184248706683,
-      "learning_rate": 7.516039519683105e-07,
-      "loss": 0.3234,
+      "epoch": 0.32842582106455265,
+      "grad_norm": 0.8030729878881923,
+      "learning_rate": 8.494359704370357e-06,
+      "loss": 0.8588,
       "step": 2030
     },
     {
-      "epoch": 0.8429991714995857,
-      "grad_norm": 2.476131958227405,
-      "learning_rate": 7.326479797083963e-07,
-      "loss": 0.3278,
+      "epoch": 0.32923475165830773,
+      "grad_norm": 0.8146354616260442,
+      "learning_rate": 8.484245934692379e-06,
+      "loss": 0.8448,
       "step": 2035
     },
     {
-      "epoch": 0.8450704225352113,
-      "grad_norm": 2.3728222901700677,
-      "learning_rate": 7.139151984542636e-07,
-      "loss": 0.3306,
+      "epoch": 0.33004368225206276,
+      "grad_norm": 0.7977438419631868,
+      "learning_rate": 8.474104375158277e-06,
+      "loss": 0.8472,
       "step": 2040
     },
     {
-      "epoch": 0.8471416735708368,
-      "grad_norm": 2.3665157862844994,
-      "learning_rate": 6.954065879676653e-07,
-      "loss": 0.3309,
+      "epoch": 0.33085261284581785,
+      "grad_norm": 0.7577130929037066,
+      "learning_rate": 8.463935106655705e-06,
+      "loss": 0.849,
       "step": 2045
     },
     {
-      "epoch": 0.8492129246064622,
-      "grad_norm": 2.5008625602985304,
-      "learning_rate": 6.771231162857722e-07,
-      "loss": 0.3362,
+      "epoch": 0.3316615434395729,
+      "grad_norm": 0.7702289198050427,
+      "learning_rate": 8.453738210293316e-06,
+      "loss": 0.8646,
       "step": 2050
     },
     {
-      "epoch": 0.8512841756420878,
-      "grad_norm": 2.2712774856509688,
-      "learning_rate": 6.590657396705525e-07,
-      "loss": 0.3215,
+      "epoch": 0.33247047403332797,
+      "grad_norm": 0.8007959864809794,
+      "learning_rate": 8.443513767400126e-06,
+      "loss": 0.8488,
       "step": 2055
     },
     {
-      "epoch": 0.8533554266777134,
-      "grad_norm": 2.535681331577295,
-      "learning_rate": 6.412354025587509e-07,
-      "loss": 0.3229,
+      "epoch": 0.333279404627083,
+      "grad_norm": 0.8288190327751878,
+      "learning_rate": 8.433261859524856e-06,
+      "loss": 0.8473,
       "step": 2060
     },
     {
-      "epoch": 0.8554266777133389,
-      "grad_norm": 2.547194182489716,
-      "learning_rate": 6.236330375124921e-07,
-      "loss": 0.3461,
+      "epoch": 0.33408833522083803,
+      "grad_norm": 0.7899197293508395,
+      "learning_rate": 8.422982568435283e-06,
+      "loss": 0.8503,
       "step": 2065
     },
     {
-      "epoch": 0.8574979287489644,
-      "grad_norm": 2.371373910066343,
-      "learning_rate": 6.062595651705111e-07,
-      "loss": 0.3389,
+      "epoch": 0.3348972658145931,
+      "grad_norm": 0.7977808601108305,
+      "learning_rate": 8.412675976117585e-06,
+      "loss": 0.8711,
       "step": 2070
     },
     {
-      "epoch": 0.8595691797845899,
-      "grad_norm": 2.439136346587766,
-      "learning_rate": 5.891158941999959e-07,
-      "loss": 0.3299,
+      "epoch": 0.33570619640834815,
+      "grad_norm": 0.8550202055185513,
+      "learning_rate": 8.4023421647757e-06,
+      "loss": 0.8296,
       "step": 2075
     },
     {
-      "epoch": 0.8616404308202155,
-      "grad_norm": 2.466656635870232,
-      "learning_rate": 5.722029212490666e-07,
-      "loss": 0.3298,
+      "epoch": 0.33651512700210323,
+      "grad_norm": 0.8370007865096064,
+      "learning_rate": 8.391981216830651e-06,
+      "loss": 0.8789,
       "step": 2080
     },
     {
-      "epoch": 0.8637116818558409,
-      "grad_norm": 2.288892577992347,
-      "learning_rate": 5.555215308998779e-07,
-      "loss": 0.3223,
+      "epoch": 0.33732405759585826,
+      "grad_norm": 0.7648678747284082,
+      "learning_rate": 8.381593214919905e-06,
+      "loss": 0.8615,
       "step": 2085
     },
     {
-      "epoch": 0.8657829328914665,
-      "grad_norm": 2.401858716049078,
-      "learning_rate": 5.390725956223531e-07,
-      "loss": 0.3218,
+      "epoch": 0.33813298818961335,
+      "grad_norm": 0.8338949429924307,
+      "learning_rate": 8.371178241896708e-06,
+      "loss": 0.8546,
       "step": 2090
     },
     {
-      "epoch": 0.867854183927092,
-      "grad_norm": 2.33005782306982,
-      "learning_rate": 5.22856975728554e-07,
-      "loss": 0.3216,
+      "epoch": 0.3389419187833684,
+      "grad_norm": 0.7923771910011903,
+      "learning_rate": 8.36073638082942e-06,
+      "loss": 0.8421,
       "step": 2095
     },
     {
-      "epoch": 0.8699254349627175,
-      "grad_norm": 2.4143761257261303,
-      "learning_rate": 5.068755193276798e-07,
-      "loss": 0.3307,
+      "epoch": 0.33975084937712347,
+      "grad_norm": 0.7717061524027085,
+      "learning_rate": 8.350267715000857e-06,
+      "loss": 0.8459,
       "step": 2100
     },
     {
-      "epoch": 0.871996685998343,
-      "grad_norm": 2.352096096095919,
-      "learning_rate": 4.911290622817161e-07,
-      "loss": 0.3204,
+      "epoch": 0.3405597799708785,
+      "grad_norm": 0.7706870416392109,
+      "learning_rate": 8.339772327907627e-06,
+      "loss": 0.839,
       "step": 2105
     },
     {
-      "epoch": 0.8740679370339686,
-      "grad_norm": 2.3702490110719303,
-      "learning_rate": 4.756184281617121e-07,
-      "loss": 0.3245,
+      "epoch": 0.3413687105646336,
+      "grad_norm": 0.8629826885755151,
+      "learning_rate": 8.329250303259466e-06,
+      "loss": 0.847,
       "step": 2110
     },
     {
-      "epoch": 0.876139188069594,
-      "grad_norm": 2.2456862499766306,
-      "learning_rate": 4.6034442820471037e-07,
-      "loss": 0.3182,
+      "epoch": 0.3421776411583886,
+      "grad_norm": 0.7813813713851825,
+      "learning_rate": 8.318701724978564e-06,
+      "loss": 0.8627,
       "step": 2115
     },
     {
-      "epoch": 0.8782104391052196,
-      "grad_norm": 2.2534896280263816,
-      "learning_rate": 4.4530786127131575e-07,
-      "loss": 0.3312,
+      "epoch": 0.34298657175214364,
+      "grad_norm": 0.7760292657058412,
+      "learning_rate": 8.308126677198896e-06,
+      "loss": 0.843,
       "step": 2120
     },
     {
-      "epoch": 0.8802816901408451,
-      "grad_norm": 2.309108572623016,
-      "learning_rate": 4.305095138039106e-07,
-      "loss": 0.326,
+      "epoch": 0.34379550234589873,
+      "grad_norm": 0.8222906943010696,
+      "learning_rate": 8.29752524426556e-06,
+      "loss": 0.8362,
       "step": 2125
     },
     {
-      "epoch": 0.8823529411764706,
-      "grad_norm": 2.5387736296808563,
-      "learning_rate": 4.159501597855287e-07,
-      "loss": 0.3327,
+      "epoch": 0.34460443293965376,
+      "grad_norm": 0.8292261329180289,
+      "learning_rate": 8.286897510734098e-06,
+      "loss": 0.8548,
       "step": 2130
     },
     {
-      "epoch": 0.8844241922120961,
-      "grad_norm": 2.4068093993598962,
-      "learning_rate": 4.0163056069936757e-07,
-      "loss": 0.3265,
+      "epoch": 0.34541336353340885,
+      "grad_norm": 0.7772581002720523,
+      "learning_rate": 8.276243561369815e-06,
+      "loss": 0.8475,
       "step": 2135
     },
     {
-      "epoch": 0.8864954432477217,
-      "grad_norm": 2.410912519062069,
-      "learning_rate": 3.8755146548896784e-07,
-      "loss": 0.3286,
+      "epoch": 0.3462222941271639,
+      "grad_norm": 0.8663068319361349,
+      "learning_rate": 8.265563481147118e-06,
+      "loss": 0.8656,
       "step": 2140
     },
     {
-      "epoch": 0.8885666942833471,
-      "grad_norm": 2.3441979713103476,
-      "learning_rate": 3.737136105190337e-07,
-      "loss": 0.3215,
+      "epoch": 0.34703122472091896,
+      "grad_norm": 0.759426780159734,
+      "learning_rate": 8.254857355248823e-06,
+      "loss": 0.8272,
       "step": 2145
     },
     {
-      "epoch": 0.8906379453189727,
-      "grad_norm": 2.4124574145028843,
-      "learning_rate": 3.6011771953693044e-07,
-      "loss": 0.3222,
+      "epoch": 0.347840155314674,
+      "grad_norm": 0.8045042800571909,
+      "learning_rate": 8.244125269065492e-06,
+      "loss": 0.8284,
       "step": 2150
     },
     {
-      "epoch": 0.8927091963545982,
-      "grad_norm": 2.3579671682789467,
-      "learning_rate": 3.4676450363481937e-07,
-      "loss": 0.3227,
+      "epoch": 0.3486490859084291,
+      "grad_norm": 0.7696869214740462,
+      "learning_rate": 8.233367308194735e-06,
+      "loss": 0.8441,
       "step": 2155
     },
     {
-      "epoch": 0.8947804473902237,
-      "grad_norm": 2.3464323541089214,
-      "learning_rate": 3.336546612124758e-07,
-      "loss": 0.3197,
+      "epoch": 0.3494580165021841,
+      "grad_norm": 0.8076486642638429,
+      "learning_rate": 8.222583558440531e-06,
+      "loss": 0.845,
       "step": 2160
     },
     {
-      "epoch": 0.8968516984258492,
-      "grad_norm": 2.4445949626198242,
-      "learning_rate": 3.20788877940757e-07,
-      "loss": 0.3164,
+      "epoch": 0.35026694709593914,
+      "grad_norm": 0.7966361567902642,
+      "learning_rate": 8.21177410581256e-06,
+      "loss": 0.8512,
       "step": 2165
     },
     {
-      "epoch": 0.8989229494614748,
-      "grad_norm": 2.668457673876882,
-      "learning_rate": 3.081678267257404e-07,
-      "loss": 0.3233,
+      "epoch": 0.3510758776896942,
+      "grad_norm": 0.8192704344463675,
+      "learning_rate": 8.200939036525495e-06,
+      "loss": 0.8475,
       "step": 2170
     },
     {
-      "epoch": 0.9009942004971002,
-      "grad_norm": 2.2794632035182865,
-      "learning_rate": 2.9579216767352815e-07,
-      "loss": 0.314,
+      "epoch": 0.35188480828344926,
+      "grad_norm": 0.7866644986853055,
+      "learning_rate": 8.190078436998326e-06,
+      "loss": 0.8481,
       "step": 2175
     },
     {
-      "epoch": 0.9030654515327258,
-      "grad_norm": 2.511600762449477,
-      "learning_rate": 2.836625480557265e-07,
-      "loss": 0.3195,
+      "epoch": 0.35269373887720434,
+      "grad_norm": 0.7784262615486034,
+      "learning_rate": 8.179192393853667e-06,
+      "loss": 0.8482,
       "step": 2180
     },
     {
-      "epoch": 0.9051367025683513,
-      "grad_norm": 2.290095274854777,
-      "learning_rate": 2.7177960227558863e-07,
-      "loss": 0.323,
+      "epoch": 0.3535026694709594,
+      "grad_norm": 0.7987375076221339,
+      "learning_rate": 8.168280993917078e-06,
+      "loss": 0.8315,
       "step": 2185
     },
     {
-      "epoch": 0.9072079536039768,
-      "grad_norm": 2.490832241300732,
-      "learning_rate": 2.601439518348331e-07,
-      "loss": 0.3229,
+      "epoch": 0.35431160006471446,
+      "grad_norm": 0.7513918277520256,
+      "learning_rate": 8.15734432421634e-06,
+      "loss": 0.8547,
       "step": 2190
     },
     {
-      "epoch": 0.9092792046396023,
-      "grad_norm": 2.3369810868473833,
-      "learning_rate": 2.487562053011422e-07,
-      "loss": 0.3298,
+      "epoch": 0.3551205306584695,
+      "grad_norm": 0.8667321468753193,
+      "learning_rate": 8.146382471980803e-06,
+      "loss": 0.8583,
       "step": 2195
     },
     {
-      "epoch": 0.9113504556752279,
-      "grad_norm": 2.511875787240971,
-      "learning_rate": 2.376169582763288e-07,
-      "loss": 0.3273,
+      "epoch": 0.3559294612522246,
+      "grad_norm": 0.7568485546003565,
+      "learning_rate": 8.135395524640659e-06,
+      "loss": 0.8342,
       "step": 2200
     },
     {
-      "epoch": 0.9134217067108533,
-      "grad_norm": 2.4194421741312078,
-      "learning_rate": 2.2672679336518789e-07,
-      "loss": 0.3187,
+      "epoch": 0.3567383918459796,
+      "grad_norm": 0.8356531732827182,
+      "learning_rate": 8.124383569826253e-06,
+      "loss": 0.8642,
       "step": 2205
     },
     {
-      "epoch": 0.9154929577464789,
-      "grad_norm": 2.4729733542961516,
-      "learning_rate": 2.1608628014502364e-07,
-      "loss": 0.3209,
+      "epoch": 0.3575473224397347,
+      "grad_norm": 0.8080998046421258,
+      "learning_rate": 8.113346695367393e-06,
+      "loss": 0.8619,
       "step": 2210
     },
     {
-      "epoch": 0.9175642087821044,
-      "grad_norm": 2.379186052198249,
-      "learning_rate": 2.0569597513586004e-07,
-      "loss": 0.3229,
+      "epoch": 0.3583562530334897,
+      "grad_norm": 0.7192475502436593,
+      "learning_rate": 8.102284989292639e-06,
+      "loss": 0.858,
       "step": 2215
     },
     {
-      "epoch": 0.9196354598177299,
-      "grad_norm": 2.3319074686966674,
-      "learning_rate": 1.955564217713335e-07,
-      "loss": 0.3181,
+      "epoch": 0.35916518362724476,
+      "grad_norm": 0.8273826603550964,
+      "learning_rate": 8.091198539828601e-06,
+      "loss": 0.8433,
       "step": 2220
     },
     {
-      "epoch": 0.9217067108533554,
-      "grad_norm": 2.6517343047945334,
-      "learning_rate": 1.8566815037026897e-07,
-      "loss": 0.323,
+      "epoch": 0.35997411422099984,
+      "grad_norm": 0.8607381682601364,
+      "learning_rate": 8.080087435399248e-06,
+      "loss": 0.8382,
       "step": 2225
     },
     {
-      "epoch": 0.923777961888981,
-      "grad_norm": 2.232882122217696,
-      "learning_rate": 1.7603167810894662e-07,
-      "loss": 0.3196,
+      "epoch": 0.36078304481475487,
+      "grad_norm": 0.8001621243743261,
+      "learning_rate": 8.068951764625186e-06,
+      "loss": 0.8406,
       "step": 2230
     },
     {
-      "epoch": 0.9258492129246064,
-      "grad_norm": 2.3988532707568186,
-      "learning_rate": 1.6664750899404892e-07,
-      "loss": 0.3183,
+      "epoch": 0.36159197540850996,
+      "grad_norm": 0.7615698239071352,
+      "learning_rate": 8.057791616322958e-06,
+      "loss": 0.8316,
       "step": 2235
     },
     {
-      "epoch": 0.927920463960232,
-      "grad_norm": 2.3744099648372403,
-      "learning_rate": 1.5751613383630128e-07,
-      "loss": 0.3181,
+      "epoch": 0.362400906002265,
+      "grad_norm": 0.8191316557232389,
+      "learning_rate": 8.046607079504345e-06,
+      "loss": 0.8585,
       "step": 2240
     },
     {
-      "epoch": 0.9299917149958575,
-      "grad_norm": 2.419787212184274,
-      "learning_rate": 1.4863803022480362e-07,
-      "loss": 0.3244,
+      "epoch": 0.3632098365960201,
+      "grad_norm": 0.8329244880945085,
+      "learning_rate": 8.035398243375636e-06,
+      "loss": 0.8237,
       "step": 2245
     },
     {
-      "epoch": 0.932062966031483,
-      "grad_norm": 2.3333109862483856,
-      "learning_rate": 1.4001366250204762e-07,
-      "loss": 0.325,
+      "epoch": 0.3640187671897751,
+      "grad_norm": 0.8151507773643624,
+      "learning_rate": 8.024165197336934e-06,
+      "loss": 0.875,
       "step": 2250
     },
     {
-      "epoch": 0.9341342170671085,
-      "grad_norm": 2.5299591194957505,
-      "learning_rate": 1.3164348173963392e-07,
-      "loss": 0.3251,
+      "epoch": 0.3648276977835302,
+      "grad_norm": 0.8450488670993342,
+      "learning_rate": 8.012908030981442e-06,
+      "loss": 0.8336,
       "step": 2255
     },
     {
-      "epoch": 0.9362054681027341,
-      "grad_norm": 2.415328780928961,
-      "learning_rate": 1.235279257146804e-07,
-      "loss": 0.3317,
+      "epoch": 0.3656366283772852,
+      "grad_norm": 0.7623122756894671,
+      "learning_rate": 8.00162683409473e-06,
+      "loss": 0.8535,
       "step": 2260
     },
     {
-      "epoch": 0.9382767191383595,
-      "grad_norm": 2.39855625868432,
-      "learning_rate": 1.1566741888692168e-07,
-      "loss": 0.3106,
+      "epoch": 0.3664455589710403,
+      "grad_norm": 0.7904848999057988,
+      "learning_rate": 7.990321696654044e-06,
+      "loss": 0.8566,
       "step": 2265
     },
     {
-      "epoch": 0.9403479701739851,
-      "grad_norm": 2.51301146564224,
-      "learning_rate": 1.080623723765134e-07,
-      "loss": 0.3207,
+      "epoch": 0.36725448956479534,
+      "grad_norm": 0.7803510833606502,
+      "learning_rate": 7.978992708827571e-06,
+      "loss": 0.8544,
       "step": 2270
     },
     {
-      "epoch": 0.9424192212096106,
-      "grad_norm": 2.5451397411628425,
-      "learning_rate": 1.0071318394252849e-07,
-      "loss": 0.321,
+      "epoch": 0.36806342015855037,
+      "grad_norm": 0.8005678354543746,
+      "learning_rate": 7.967639960973727e-06,
+      "loss": 0.8461,
       "step": 2275
     },
     {
-      "epoch": 0.9444904722452361,
-      "grad_norm": 2.3238922238777753,
-      "learning_rate": 9.362023796215036e-08,
-      "loss": 0.3115,
+      "epoch": 0.36887235075230546,
+      "grad_norm": 0.8007519176589866,
+      "learning_rate": 7.956263543640432e-06,
+      "loss": 0.8409,
       "step": 2280
     },
     {
-      "epoch": 0.9465617232808616,
-      "grad_norm": 2.3960161103125595,
-      "learning_rate": 8.678390541057512e-08,
-      "loss": 0.3334,
+      "epoch": 0.3696812813460605,
+      "grad_norm": 0.8474393359723089,
+      "learning_rate": 7.944863547564396e-06,
+      "loss": 0.8585,
       "step": 2285
     },
     {
-      "epoch": 0.9486329743164872,
-      "grad_norm": 2.455291465574181,
-      "learning_rate": 8.020454384160437e-08,
-      "loss": 0.321,
+      "epoch": 0.3704902119398156,
+      "grad_norm": 0.8486471733965024,
+      "learning_rate": 7.933440063670383e-06,
+      "loss": 0.8399,
       "step": 2290
     },
     {
-      "epoch": 0.9507042253521126,
-      "grad_norm": 2.313389085031741,
-      "learning_rate": 7.388249736894615e-08,
-      "loss": 0.3155,
+      "epoch": 0.3712991425335706,
+      "grad_norm": 0.7861689750278411,
+      "learning_rate": 7.921993183070497e-06,
+      "loss": 0.8496,
       "step": 2295
     },
     {
-      "epoch": 0.9527754763877382,
-      "grad_norm": 2.4859967083644507,
-      "learning_rate": 6.78180966482156e-08,
-      "loss": 0.3189,
+      "epoch": 0.3721080731273257,
+      "grad_norm": 0.8805931639748436,
+      "learning_rate": 7.910522997063451e-06,
+      "loss": 0.8351,
       "step": 2300
     },
     {
-      "epoch": 0.9548467274233637,
-      "grad_norm": 2.5150728286365878,
-      "learning_rate": 6.201165885964311e-08,
-      "loss": 0.3196,
+      "epoch": 0.3729170037210807,
+      "grad_norm": 0.7946263432456658,
+      "learning_rate": 7.899029597133836e-06,
+      "loss": 0.8444,
       "step": 2305
     },
     {
-      "epoch": 0.9569179784589892,
-      "grad_norm": 2.5646449395309134,
-      "learning_rate": 5.646348769148491e-08,
-      "loss": 0.32,
+      "epoch": 0.3737259343148358,
+      "grad_norm": 0.8337819610743522,
+      "learning_rate": 7.887513074951397e-06,
+      "loss": 0.8491,
       "step": 2310
     },
     {
-      "epoch": 0.9589892294946147,
-      "grad_norm": 2.4648895794572145,
-      "learning_rate": 5.117387332413737e-08,
-      "loss": 0.3142,
+      "epoch": 0.37453486490859084,
+      "grad_norm": 0.8157784141608686,
+      "learning_rate": 7.875973522370294e-06,
+      "loss": 0.835,
       "step": 2315
     },
     {
-      "epoch": 0.9610604805302403,
-      "grad_norm": 2.4075943066752292,
-      "learning_rate": 4.6143092414961396e-08,
-      "loss": 0.3275,
+      "epoch": 0.3753437955023459,
+      "grad_norm": 0.8127539829554938,
+      "learning_rate": 7.864411031428379e-06,
+      "loss": 0.8425,
       "step": 2320
     },
     {
-      "epoch": 0.9631317315658657,
-      "grad_norm": 2.5717189064452604,
-      "learning_rate": 4.1371408083815126e-08,
-      "loss": 0.3326,
+      "epoch": 0.37615272609610095,
+      "grad_norm": 0.7999538507762206,
+      "learning_rate": 7.852825694346455e-06,
+      "loss": 0.8364,
       "step": 2325
     },
     {
-      "epoch": 0.9652029826014913,
-      "grad_norm": 2.2405081152048982,
-      "learning_rate": 3.685906989928656e-08,
-      "loss": 0.3198,
+      "epoch": 0.376961656689856,
+      "grad_norm": 0.8590002826350661,
+      "learning_rate": 7.841217603527545e-06,
+      "loss": 0.8415,
       "step": 2330
     },
     {
-      "epoch": 0.9672742336371168,
-      "grad_norm": 2.493323605860437,
-      "learning_rate": 3.2606313865646276e-08,
-      "loss": 0.3158,
+      "epoch": 0.37777058728361107,
+      "grad_norm": 0.8424318909298105,
+      "learning_rate": 7.82958685155615e-06,
+      "loss": 0.8559,
       "step": 2335
     },
     {
-      "epoch": 0.9693454846727423,
-      "grad_norm": 2.3221489971336307,
-      "learning_rate": 2.861336241050061e-08,
-      "loss": 0.3192,
+      "epoch": 0.3785795178773661,
+      "grad_norm": 0.7727695207961051,
+      "learning_rate": 7.817933531197513e-06,
+      "loss": 0.8479,
       "step": 2340
     },
     {
-      "epoch": 0.9714167357083678,
-      "grad_norm": 2.224686624765676,
-      "learning_rate": 2.488042437315985e-08,
-      "loss": 0.3144,
+      "epoch": 0.3793884484711212,
+      "grad_norm": 0.8080185911172371,
+      "learning_rate": 7.806257735396879e-06,
+      "loss": 0.832,
       "step": 2345
     },
     {
-      "epoch": 0.9734879867439934,
-      "grad_norm": 2.517150826241084,
-      "learning_rate": 2.1407694993714755e-08,
-      "loss": 0.3201,
+      "epoch": 0.3801973790648762,
+      "grad_norm": 0.8144308118806884,
+      "learning_rate": 7.794559557278757e-06,
+      "loss": 0.8449,
       "step": 2350
     },
     {
-      "epoch": 0.9755592377796188,
-      "grad_norm": 2.4950963422344126,
-      "learning_rate": 1.8195355902824174e-08,
-      "loss": 0.3094,
+      "epoch": 0.3810063096586313,
+      "grad_norm": 0.8257947013812411,
+      "learning_rate": 7.782839090146172e-06,
+      "loss": 0.8579,
       "step": 2355
     },
     {
-      "epoch": 0.9776304888152444,
-      "grad_norm": 2.3719000652949376,
-      "learning_rate": 1.5243575112218744e-08,
-      "loss": 0.3069,
+      "epoch": 0.38181524025238633,
+      "grad_norm": 0.7945614016073755,
+      "learning_rate": 7.771096427479925e-06,
+      "loss": 0.8298,
       "step": 2360
     },
     {
-      "epoch": 0.97970173985087,
-      "grad_norm": 2.532095428796904,
-      "learning_rate": 1.2552507005909042e-08,
-      "loss": 0.317,
+      "epoch": 0.3826241708461414,
+      "grad_norm": 0.8410802459537005,
+      "learning_rate": 7.759331662937841e-06,
+      "loss": 0.8491,
       "step": 2365
     },
     {
-      "epoch": 0.9817729908864954,
-      "grad_norm": 2.508218831541988,
-      "learning_rate": 1.0122292332114814e-08,
-      "loss": 0.3214,
+      "epoch": 0.38343310143989645,
+      "grad_norm": 0.8308755908758182,
+      "learning_rate": 7.747544890354031e-06,
+      "loss": 0.8573,
       "step": 2370
     },
     {
-      "epoch": 0.9838442419221209,
-      "grad_norm": 2.3230365500226156,
-      "learning_rate": 7.953058195900864e-09,
-      "loss": 0.3214,
+      "epoch": 0.38424203203365154,
+      "grad_norm": 0.8040651764511801,
+      "learning_rate": 7.735736203738137e-06,
+      "loss": 0.8344,
       "step": 2375
     },
     {
-      "epoch": 0.9859154929577465,
-      "grad_norm": 2.3842329300121565,
-      "learning_rate": 6.044918052531268e-09,
-      "loss": 0.3267,
+      "epoch": 0.38505096262740657,
+      "grad_norm": 0.8125819462471874,
+      "learning_rate": 7.723905697274586e-06,
+      "loss": 0.8424,
       "step": 2380
     },
     {
-      "epoch": 0.987986743993372,
-      "grad_norm": 2.450379118891722,
-      "learning_rate": 4.397971701533554e-09,
-      "loss": 0.3231,
+      "epoch": 0.3858598932211616,
+      "grad_norm": 0.7539076215491992,
+      "learning_rate": 7.71205346532183e-06,
+      "loss": 0.8495,
       "step": 2385
     },
     {
-      "epoch": 0.9900579950289975,
-      "grad_norm": 2.3274963425198423,
-      "learning_rate": 3.0123052814812203e-09,
-      "loss": 0.3284,
+      "epoch": 0.3866688238149167,
+      "grad_norm": 0.8155202307231507,
+      "learning_rate": 7.700179602411615e-06,
+      "loss": 0.8342,
       "step": 2390
     },
     {
-      "epoch": 0.992129246064623,
-      "grad_norm": 2.4027725490076075,
-      "learning_rate": 1.887991265486222e-09,
-      "loss": 0.309,
+      "epoch": 0.3874777544086717,
+      "grad_norm": 0.8220478382780729,
+      "learning_rate": 7.688284203248197e-06,
+      "loss": 0.846,
       "step": 2395
     },
     {
-      "epoch": 0.9942004971002486,
-      "grad_norm": 2.5888360240441015,
-      "learning_rate": 1.025088457409229e-09,
-      "loss": 0.3177,
+      "epoch": 0.3882866850024268,
+      "grad_norm": 0.8673932844304086,
+      "learning_rate": 7.67636736270761e-06,
+      "loss": 0.875,
       "step": 2400
     },
     {
-      "epoch": 0.996271748135874,
-      "grad_norm": 2.5018345143420917,
-      "learning_rate": 4.2364198878597216e-10,
-      "loss": 0.3297,
+      "epoch": 0.38909561559618183,
+      "grad_norm": 0.803133188698084,
+      "learning_rate": 7.664429175836903e-06,
+      "loss": 0.8672,
       "step": 2405
     },
     {
-      "epoch": 0.9983429991714996,
-      "grad_norm": 2.22608195027372,
-      "learning_rate": 8.368331646302353e-11,
-      "loss": 0.3161,
+      "epoch": 0.3899045461899369,
+      "grad_norm": 0.7796141487305905,
+      "learning_rate": 7.652469737853372e-06,
+      "loss": 0.8542,
       "step": 2410
     },
+    {
+      "epoch": 0.39071347678369195,
+      "grad_norm": 0.7721547412514925,
+      "learning_rate": 7.64048914414382e-06,
+      "loss": 0.8274,
+      "step": 2415
+    },
+    {
+      "epoch": 0.39152240737744703,
+      "grad_norm": 0.8076354344928516,
+      "learning_rate": 7.628487490263779e-06,
+      "loss": 0.8315,
+      "step": 2420
+    },
+    {
+      "epoch": 0.39233133797120207,
+      "grad_norm": 0.8310201097209512,
+      "learning_rate": 7.616464871936748e-06,
+      "loss": 0.8478,
+      "step": 2425
+    },
+    {
+      "epoch": 0.39314026856495715,
+      "grad_norm": 0.8088889309070267,
+      "learning_rate": 7.60442138505345e-06,
+      "loss": 0.8434,
+      "step": 2430
+    },
+    {
+      "epoch": 0.3939491991587122,
+      "grad_norm": 0.8845883583499743,
+      "learning_rate": 7.5923571256710396e-06,
+      "loss": 0.827,
+      "step": 2435
+    },
+    {
+      "epoch": 0.3947581297524672,
+      "grad_norm": 0.8075412429174051,
+      "learning_rate": 7.580272190012357e-06,
+      "loss": 0.8395,
+      "step": 2440
+    },
+    {
+      "epoch": 0.3955670603462223,
+      "grad_norm": 0.781477252019071,
+      "learning_rate": 7.5681666744651505e-06,
+      "loss": 0.8157,
+      "step": 2445
+    },
+    {
+      "epoch": 0.39637599093997733,
+      "grad_norm": 0.7999629302259256,
+      "learning_rate": 7.556040675581311e-06,
+      "loss": 0.8288,
+      "step": 2450
+    },
+    {
+      "epoch": 0.3971849215337324,
+      "grad_norm": 0.8292218777187262,
+      "learning_rate": 7.5438942900761035e-06,
+      "loss": 0.8417,
+      "step": 2455
+    },
+    {
+      "epoch": 0.39799385212748745,
+      "grad_norm": 0.7796119585780503,
+      "learning_rate": 7.531727614827392e-06,
+      "loss": 0.8512,
+      "step": 2460
+    },
+    {
+      "epoch": 0.39880278272124253,
+      "grad_norm": 1.0387110538612172,
+      "learning_rate": 7.519540746874869e-06,
+      "loss": 0.8261,
+      "step": 2465
+    },
+    {
+      "epoch": 0.39961171331499756,
+      "grad_norm": 0.8332505684625972,
+      "learning_rate": 7.507333783419282e-06,
+      "loss": 0.8643,
+      "step": 2470
+    },
+    {
+      "epoch": 0.40042064390875265,
+      "grad_norm": 0.80731246004178,
+      "learning_rate": 7.495106821821656e-06,
+      "loss": 0.8542,
+      "step": 2475
+    },
+    {
+      "epoch": 0.4012295745025077,
+      "grad_norm": 0.8894304174472868,
+      "learning_rate": 7.482859959602518e-06,
+      "loss": 0.8403,
+      "step": 2480
+    },
+    {
+      "epoch": 0.40203850509626277,
+      "grad_norm": 0.8754692672700708,
+      "learning_rate": 7.470593294441124e-06,
+      "loss": 0.8516,
+      "step": 2485
+    },
+    {
+      "epoch": 0.4028474356900178,
+      "grad_norm": 0.8789987496906461,
+      "learning_rate": 7.4583069241746696e-06,
+      "loss": 0.834,
+      "step": 2490
+    },
+    {
+      "epoch": 0.4036563662837728,
+      "grad_norm": 0.841200352411981,
+      "learning_rate": 7.446000946797519e-06,
+      "loss": 0.8241,
+      "step": 2495
+    },
+    {
+      "epoch": 0.4044652968775279,
+      "grad_norm": 0.8282889658691387,
+      "learning_rate": 7.433675460460419e-06,
+      "loss": 0.845,
+      "step": 2500
+    },
+    {
+      "epoch": 0.40527422747128294,
+      "grad_norm": 0.8171176352536702,
+      "learning_rate": 7.421330563469717e-06,
+      "loss": 0.8361,
+      "step": 2505
+    },
+    {
+      "epoch": 0.40608315806503803,
+      "grad_norm": 0.7424149733511415,
+      "learning_rate": 7.408966354286575e-06,
+      "loss": 0.8394,
+      "step": 2510
+    },
+    {
+      "epoch": 0.40689208865879306,
+      "grad_norm": 0.772435215520305,
+      "learning_rate": 7.396582931526194e-06,
+      "loss": 0.8364,
+      "step": 2515
+    },
+    {
+      "epoch": 0.40770101925254815,
+      "grad_norm": 0.8577362984373131,
+      "learning_rate": 7.38418039395701e-06,
+      "loss": 0.8522,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4085099498463032,
+      "grad_norm": 0.8324922368152244,
+      "learning_rate": 7.371758840499921e-06,
+      "loss": 0.837,
+      "step": 2525
+    },
+    {
+      "epoch": 0.40931888044005826,
+      "grad_norm": 0.8167673718582185,
+      "learning_rate": 7.359318370227494e-06,
+      "loss": 0.8551,
+      "step": 2530
+    },
+    {
+      "epoch": 0.4101278110338133,
+      "grad_norm": 0.8183481719062728,
+      "learning_rate": 7.346859082363172e-06,
+      "loss": 0.8653,
+      "step": 2535
+    },
+    {
+      "epoch": 0.4109367416275684,
+      "grad_norm": 0.8526948175700703,
+      "learning_rate": 7.334381076280483e-06,
+      "loss": 0.8321,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4117456722213234,
+      "grad_norm": 0.8683127108850153,
+      "learning_rate": 7.321884451502252e-06,
+      "loss": 0.8391,
+      "step": 2545
+    },
+    {
+      "epoch": 0.41255460281507844,
+      "grad_norm": 0.8156307062993907,
+      "learning_rate": 7.309369307699802e-06,
+      "loss": 0.8281,
+      "step": 2550
+    },
+    {
+      "epoch": 0.41336353340883353,
+      "grad_norm": 0.831536181417552,
+      "learning_rate": 7.296835744692163e-06,
+      "loss": 0.8417,
+      "step": 2555
+    },
+    {
+      "epoch": 0.41417246400258856,
+      "grad_norm": 0.840572967544839,
+      "learning_rate": 7.2842838624452725e-06,
+      "loss": 0.8172,
+      "step": 2560
+    },
+    {
+      "epoch": 0.41498139459634364,
+      "grad_norm": 0.8063608871323279,
+      "learning_rate": 7.271713761071181e-06,
+      "loss": 0.8533,
+      "step": 2565
+    },
+    {
+      "epoch": 0.4157903251900987,
+      "grad_norm": 0.9598879655118182,
+      "learning_rate": 7.259125540827248e-06,
+      "loss": 0.8323,
+      "step": 2570
+    },
+    {
+      "epoch": 0.41659925578385376,
+      "grad_norm": 0.77131131850428,
+      "learning_rate": 7.246519302115355e-06,
+      "loss": 0.8202,
+      "step": 2575
+    },
+    {
+      "epoch": 0.4174081863776088,
+      "grad_norm": 0.7740177253879231,
+      "learning_rate": 7.233895145481086e-06,
+      "loss": 0.8169,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4182171169713639,
+      "grad_norm": 0.7993458055375153,
+      "learning_rate": 7.221253171612944e-06,
+      "loss": 0.8281,
+      "step": 2585
+    },
+    {
+      "epoch": 0.4190260475651189,
+      "grad_norm": 0.8017047616887952,
+      "learning_rate": 7.208593481341536e-06,
+      "loss": 0.8551,
+      "step": 2590
+    },
+    {
+      "epoch": 0.419834978158874,
+      "grad_norm": 0.7947124171074315,
+      "learning_rate": 7.195916175638773e-06,
+      "loss": 0.8347,
+      "step": 2595
+    },
+    {
+      "epoch": 0.420643908752629,
+      "grad_norm": 0.8878564841823726,
+      "learning_rate": 7.183221355617065e-06,
+      "loss": 0.8377,
+      "step": 2600
+    },
+    {
+      "epoch": 0.42145283934638406,
+      "grad_norm": 0.8125554403726852,
+      "learning_rate": 7.170509122528511e-06,
+      "loss": 0.8322,
+      "step": 2605
+    },
+    {
+      "epoch": 0.42226176994013914,
+      "grad_norm": 0.8652065540252103,
+      "learning_rate": 7.157779577764099e-06,
+      "loss": 0.8367,
+      "step": 2610
+    },
+    {
+      "epoch": 0.4230707005338942,
+      "grad_norm": 0.8610622568512487,
+      "learning_rate": 7.145032822852889e-06,
+      "loss": 0.8378,
+      "step": 2615
+    },
+    {
+      "epoch": 0.42387963112764926,
+      "grad_norm": 0.7876508067233663,
+      "learning_rate": 7.132268959461209e-06,
+      "loss": 0.8375,
+      "step": 2620
+    },
+    {
+      "epoch": 0.4246885617214043,
+      "grad_norm": 0.8282333937739974,
+      "learning_rate": 7.119488089391836e-06,
+      "loss": 0.8406,
+      "step": 2625
+    },
+    {
+      "epoch": 0.4254974923151594,
+      "grad_norm": 0.8305925610445289,
+      "learning_rate": 7.106690314583199e-06,
+      "loss": 0.8414,
+      "step": 2630
+    },
+    {
+      "epoch": 0.4263064229089144,
+      "grad_norm": 0.789770771549953,
+      "learning_rate": 7.0938757371085485e-06,
+      "loss": 0.853,
+      "step": 2635
+    },
+    {
+      "epoch": 0.4271153535026695,
+      "grad_norm": 0.8456657346625689,
+      "learning_rate": 7.081044459175159e-06,
+      "loss": 0.8308,
+      "step": 2640
+    },
+    {
+      "epoch": 0.4279242840964245,
+      "grad_norm": 0.8352288349586666,
+      "learning_rate": 7.068196583123495e-06,
+      "loss": 0.8269,
+      "step": 2645
+    },
+    {
+      "epoch": 0.4287332146901796,
+      "grad_norm": 0.8448828847445977,
+      "learning_rate": 7.055332211426417e-06,
+      "loss": 0.8388,
+      "step": 2650
+    },
+    {
+      "epoch": 0.42954214528393464,
+      "grad_norm": 0.858853034822238,
+      "learning_rate": 7.042451446688342e-06,
+      "loss": 0.8298,
+      "step": 2655
+    },
+    {
+      "epoch": 0.43035107587768967,
+      "grad_norm": 0.8391759096265385,
+      "learning_rate": 7.029554391644441e-06,
+      "loss": 0.8371,
+      "step": 2660
+    },
+    {
+      "epoch": 0.43116000647144476,
+      "grad_norm": 0.7973420420214934,
+      "learning_rate": 7.016641149159816e-06,
+      "loss": 0.8377,
+      "step": 2665
+    },
+    {
+      "epoch": 0.4319689370651998,
+      "grad_norm": 0.8010191599865539,
+      "learning_rate": 7.00371182222867e-06,
+      "loss": 0.8483,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4327778676589549,
+      "grad_norm": 0.8013088566432139,
+      "learning_rate": 6.9907665139735035e-06,
+      "loss": 0.8366,
+      "step": 2675
+    },
+    {
+      "epoch": 0.4335867982527099,
+      "grad_norm": 0.8481738940133525,
+      "learning_rate": 6.977805327644275e-06,
+      "loss": 0.8311,
+      "step": 2680
+    },
+    {
+      "epoch": 0.434395728846465,
+      "grad_norm": 0.7925446967652235,
+      "learning_rate": 6.964828366617583e-06,
+      "loss": 0.8466,
+      "step": 2685
+    },
+    {
+      "epoch": 0.43520465944022,
+      "grad_norm": 0.810554276504509,
+      "learning_rate": 6.95183573439585e-06,
+      "loss": 0.8468,
+      "step": 2690
+    },
+    {
+      "epoch": 0.4360135900339751,
+      "grad_norm": 0.8018391904385581,
+      "learning_rate": 6.938827534606484e-06,
+      "loss": 0.8356,
+      "step": 2695
+    },
+    {
+      "epoch": 0.43682252062773014,
+      "grad_norm": 0.8273266309236444,
+      "learning_rate": 6.925803871001058e-06,
+      "loss": 0.8358,
+      "step": 2700
+    },
+    {
+      "epoch": 0.4376314512214852,
+      "grad_norm": 0.867800658645478,
+      "learning_rate": 6.912764847454485e-06,
+      "loss": 0.8332,
+      "step": 2705
+    },
+    {
+      "epoch": 0.43844038181524025,
+      "grad_norm": 0.8366321913701985,
+      "learning_rate": 6.899710567964184e-06,
+      "loss": 0.8289,
+      "step": 2710
+    },
+    {
+      "epoch": 0.4392493124089953,
+      "grad_norm": 0.851318956301739,
+      "learning_rate": 6.8866411366492546e-06,
+      "loss": 0.8252,
+      "step": 2715
+    },
+    {
+      "epoch": 0.44005824300275037,
+      "grad_norm": 0.8661629567364111,
+      "learning_rate": 6.873556657749646e-06,
+      "loss": 0.8433,
+      "step": 2720
+    },
+    {
+      "epoch": 0.4408671735965054,
+      "grad_norm": 0.8890212561668989,
+      "learning_rate": 6.860457235625322e-06,
+      "loss": 0.8327,
+      "step": 2725
+    },
+    {
+      "epoch": 0.4416761041902605,
+      "grad_norm": 0.8479920834756562,
+      "learning_rate": 6.847342974755435e-06,
+      "loss": 0.8408,
+      "step": 2730
+    },
+    {
+      "epoch": 0.4424850347840155,
+      "grad_norm": 0.8040960156855093,
+      "learning_rate": 6.834213979737488e-06,
+      "loss": 0.8667,
+      "step": 2735
+    },
+    {
+      "epoch": 0.4432939653777706,
+      "grad_norm": 0.8449566845357666,
+      "learning_rate": 6.821070355286498e-06,
+      "loss": 0.8318,
+      "step": 2740
+    },
+    {
+      "epoch": 0.44410289597152564,
+      "grad_norm": 0.8008355458582818,
+      "learning_rate": 6.807912206234168e-06,
+      "loss": 0.8233,
+      "step": 2745
+    },
+    {
+      "epoch": 0.4449118265652807,
+      "grad_norm": 0.8051699329730306,
+      "learning_rate": 6.794739637528051e-06,
+      "loss": 0.8386,
+      "step": 2750
+    },
+    {
+      "epoch": 0.44572075715903575,
+      "grad_norm": 0.8602857948468763,
+      "learning_rate": 6.7815527542307e-06,
+      "loss": 0.8399,
+      "step": 2755
+    },
+    {
+      "epoch": 0.44652968775279084,
+      "grad_norm": 0.8187853844017233,
+      "learning_rate": 6.768351661518845e-06,
+      "loss": 0.8176,
+      "step": 2760
+    },
+    {
+      "epoch": 0.44733861834654587,
+      "grad_norm": 0.8053208482891812,
+      "learning_rate": 6.755136464682546e-06,
+      "loss": 0.8475,
+      "step": 2765
+    },
+    {
+      "epoch": 0.4481475489403009,
+      "grad_norm": 0.8750589437015153,
+      "learning_rate": 6.741907269124358e-06,
+      "loss": 0.8279,
+      "step": 2770
+    },
+    {
+      "epoch": 0.448956479534056,
+      "grad_norm": 0.8815876030660124,
+      "learning_rate": 6.728664180358487e-06,
+      "loss": 0.8438,
+      "step": 2775
+    },
+    {
+      "epoch": 0.449765410127811,
+      "grad_norm": 0.8318847722342249,
+      "learning_rate": 6.715407304009948e-06,
+      "loss": 0.852,
+      "step": 2780
+    },
+    {
+      "epoch": 0.4505743407215661,
+      "grad_norm": 0.8913600813483957,
+      "learning_rate": 6.702136745813721e-06,
+      "loss": 0.8435,
+      "step": 2785
+    },
+    {
+      "epoch": 0.45138327131532113,
+      "grad_norm": 0.8207757002268261,
+      "learning_rate": 6.688852611613921e-06,
+      "loss": 0.8432,
+      "step": 2790
+    },
+    {
+      "epoch": 0.4521922019090762,
+      "grad_norm": 0.8964348508265096,
+      "learning_rate": 6.675555007362931e-06,
+      "loss": 0.8492,
+      "step": 2795
+    },
+    {
+      "epoch": 0.45300113250283125,
+      "grad_norm": 0.88701161837332,
+      "learning_rate": 6.662244039120575e-06,
+      "loss": 0.8488,
+      "step": 2800
+    },
+    {
+      "epoch": 0.45381006309658634,
+      "grad_norm": 0.7897608660219696,
+      "learning_rate": 6.648919813053266e-06,
+      "loss": 0.8436,
+      "step": 2805
+    },
+    {
+      "epoch": 0.45461899369034137,
+      "grad_norm": 0.8277360620764583,
+      "learning_rate": 6.635582435433161e-06,
+      "loss": 0.8473,
+      "step": 2810
+    },
+    {
+      "epoch": 0.4554279242840964,
+      "grad_norm": 0.8476898435569198,
+      "learning_rate": 6.6222320126373105e-06,
+      "loss": 0.8248,
+      "step": 2815
+    },
+    {
+      "epoch": 0.4562368548778515,
+      "grad_norm": 0.8213309435302791,
+      "learning_rate": 6.6088686511468106e-06,
+      "loss": 0.8208,
+      "step": 2820
+    },
+    {
+      "epoch": 0.4570457854716065,
+      "grad_norm": 0.8314089865069589,
+      "learning_rate": 6.595492457545953e-06,
+      "loss": 0.8654,
+      "step": 2825
+    },
+    {
+      "epoch": 0.4578547160653616,
+      "grad_norm": 0.8098173131527258,
+      "learning_rate": 6.582103538521383e-06,
+      "loss": 0.817,
+      "step": 2830
+    },
+    {
+      "epoch": 0.45866364665911663,
+      "grad_norm": 0.8530608603749926,
+      "learning_rate": 6.568702000861234e-06,
+      "loss": 0.858,
+      "step": 2835
+    },
+    {
+      "epoch": 0.4594725772528717,
+      "grad_norm": 0.884209021072577,
+      "learning_rate": 6.5552879514542915e-06,
+      "loss": 0.8314,
+      "step": 2840
+    },
+    {
+      "epoch": 0.46028150784662675,
+      "grad_norm": 0.8215391384460337,
+      "learning_rate": 6.541861497289126e-06,
+      "loss": 0.8284,
+      "step": 2845
+    },
+    {
+      "epoch": 0.46109043844038183,
+      "grad_norm": 0.7972506723621342,
+      "learning_rate": 6.528422745453251e-06,
+      "loss": 0.8439,
+      "step": 2850
+    },
+    {
+      "epoch": 0.46189936903413686,
+      "grad_norm": 0.8602949746737556,
+      "learning_rate": 6.514971803132264e-06,
+      "loss": 0.8366,
+      "step": 2855
+    },
+    {
+      "epoch": 0.46270829962789195,
+      "grad_norm": 0.8506988739569268,
+      "learning_rate": 6.50150877760899e-06,
+      "loss": 0.86,
+      "step": 2860
+    },
+    {
+      "epoch": 0.463517230221647,
+      "grad_norm": 0.7981906008336666,
+      "learning_rate": 6.488033776262632e-06,
+      "loss": 0.8531,
+      "step": 2865
+    },
+    {
+      "epoch": 0.464326160815402,
+      "grad_norm": 0.7798400777450988,
+      "learning_rate": 6.474546906567905e-06,
+      "loss": 0.8457,
+      "step": 2870
+    },
+    {
+      "epoch": 0.4651350914091571,
+      "grad_norm": 0.8391735146467554,
+      "learning_rate": 6.46104827609419e-06,
+      "loss": 0.8408,
+      "step": 2875
+    },
+    {
+      "epoch": 0.46594402200291213,
+      "grad_norm": 0.8052808401629455,
+      "learning_rate": 6.447537992504663e-06,
+      "loss": 0.8486,
+      "step": 2880
+    },
+    {
+      "epoch": 0.4667529525966672,
+      "grad_norm": 0.7988941689518605,
+      "learning_rate": 6.434016163555452e-06,
+      "loss": 0.846,
+      "step": 2885
+    },
+    {
+      "epoch": 0.46756188319042224,
+      "grad_norm": 0.8165894688556871,
+      "learning_rate": 6.420482897094764e-06,
+      "loss": 0.8319,
+      "step": 2890
+    },
+    {
+      "epoch": 0.46837081378417733,
+      "grad_norm": 0.8121894000178076,
+      "learning_rate": 6.406938301062031e-06,
+      "loss": 0.8388,
+      "step": 2895
+    },
+    {
+      "epoch": 0.46917974437793236,
+      "grad_norm": 0.835869232540653,
+      "learning_rate": 6.393382483487049e-06,
+      "loss": 0.8228,
+      "step": 2900
+    },
+    {
+      "epoch": 0.46998867497168745,
+      "grad_norm": 0.912311843916525,
+      "learning_rate": 6.379815552489112e-06,
+      "loss": 0.819,
+      "step": 2905
+    },
+    {
+      "epoch": 0.4707976055654425,
+      "grad_norm": 0.8202079981924546,
+      "learning_rate": 6.366237616276161e-06,
+      "loss": 0.825,
+      "step": 2910
+    },
+    {
+      "epoch": 0.47160653615919756,
+      "grad_norm": 0.82471875114036,
+      "learning_rate": 6.3526487831439045e-06,
+      "loss": 0.8238,
+      "step": 2915
+    },
+    {
+      "epoch": 0.4724154667529526,
+      "grad_norm": 0.9037615912491918,
+      "learning_rate": 6.339049161474965e-06,
+      "loss": 0.8202,
+      "step": 2920
+    },
+    {
+      "epoch": 0.4732243973467076,
+      "grad_norm": 0.8152831104583218,
+      "learning_rate": 6.3254388597380165e-06,
+      "loss": 0.8448,
+      "step": 2925
+    },
+    {
+      "epoch": 0.4740333279404627,
+      "grad_norm": 0.8118468395706068,
+      "learning_rate": 6.311817986486917e-06,
+      "loss": 0.8342,
+      "step": 2930
+    },
+    {
+      "epoch": 0.47484225853421774,
+      "grad_norm": 0.8075212743777721,
+      "learning_rate": 6.298186650359832e-06,
+      "loss": 0.8274,
+      "step": 2935
+    },
+    {
+      "epoch": 0.47565118912797283,
+      "grad_norm": 0.859392711702304,
+      "learning_rate": 6.284544960078387e-06,
+      "loss": 0.8401,
+      "step": 2940
+    },
+    {
+      "epoch": 0.47646011972172786,
+      "grad_norm": 0.8858782454028935,
+      "learning_rate": 6.270893024446788e-06,
+      "loss": 0.8624,
+      "step": 2945
+    },
+    {
+      "epoch": 0.47726905031548295,
+      "grad_norm": 0.8427992758295506,
+      "learning_rate": 6.257230952350954e-06,
+      "loss": 0.839,
+      "step": 2950
+    },
+    {
+      "epoch": 0.478077980909238,
+      "grad_norm": 0.8680923960328171,
+      "learning_rate": 6.243558852757654e-06,
+      "loss": 0.8577,
+      "step": 2955
+    },
+    {
+      "epoch": 0.47888691150299306,
+      "grad_norm": 0.8682921557831497,
+      "learning_rate": 6.229876834713633e-06,
+      "loss": 0.839,
+      "step": 2960
+    },
+    {
+      "epoch": 0.4796958420967481,
+      "grad_norm": 0.8040852911882872,
+      "learning_rate": 6.216185007344745e-06,
+      "loss": 0.8501,
+      "step": 2965
+    },
+    {
+      "epoch": 0.4805047726905032,
+      "grad_norm": 0.7631057139150006,
+      "learning_rate": 6.202483479855083e-06,
+      "loss": 0.8268,
+      "step": 2970
+    },
+    {
+      "epoch": 0.4813137032842582,
+      "grad_norm": 0.813256135509401,
+      "learning_rate": 6.188772361526104e-06,
+      "loss": 0.8196,
+      "step": 2975
+    },
+    {
+      "epoch": 0.48212263387801324,
+      "grad_norm": 0.8297884897498811,
+      "learning_rate": 6.175051761715762e-06,
+      "loss": 0.848,
+      "step": 2980
+    },
+    {
+      "epoch": 0.4829315644717683,
+      "grad_norm": 0.8334678385544815,
+      "learning_rate": 6.161321789857635e-06,
+      "loss": 0.8367,
+      "step": 2985
+    },
+    {
+      "epoch": 0.48374049506552336,
+      "grad_norm": 0.8985811450220941,
+      "learning_rate": 6.147582555460048e-06,
+      "loss": 0.843,
+      "step": 2990
+    },
+    {
+      "epoch": 0.48454942565927844,
+      "grad_norm": 0.8362819535768706,
+      "learning_rate": 6.133834168105206e-06,
+      "loss": 0.8385,
+      "step": 2995
+    },
+    {
+      "epoch": 0.4853583562530335,
+      "grad_norm": 0.8317479151299757,
+      "learning_rate": 6.120076737448314e-06,
+      "loss": 0.8331,
+      "step": 3000
+    },
+    {
+      "epoch": 0.48616728684678856,
+      "grad_norm": 0.816034350431961,
+      "learning_rate": 6.106310373216706e-06,
+      "loss": 0.8389,
+      "step": 3005
+    },
+    {
+      "epoch": 0.4869762174405436,
+      "grad_norm": 0.8435438416835686,
+      "learning_rate": 6.092535185208973e-06,
+      "loss": 0.8144,
+      "step": 3010
+    },
+    {
+      "epoch": 0.4877851480342987,
+      "grad_norm": 0.8650473536522931,
+      "learning_rate": 6.078751283294075e-06,
+      "loss": 0.8415,
+      "step": 3015
+    },
+    {
+      "epoch": 0.4885940786280537,
+      "grad_norm": 0.8631976632792485,
+      "learning_rate": 6.0649587774104775e-06,
+      "loss": 0.8384,
+      "step": 3020
+    },
+    {
+      "epoch": 0.4894030092218088,
+      "grad_norm": 0.8051930942211001,
+      "learning_rate": 6.0511577775652744e-06,
+      "loss": 0.8307,
+      "step": 3025
+    },
+    {
+      "epoch": 0.4902119398155638,
+      "grad_norm": 0.9000033790636423,
+      "learning_rate": 6.037348393833298e-06,
+      "loss": 0.8288,
+      "step": 3030
+    },
+    {
+      "epoch": 0.49102087040931885,
+      "grad_norm": 0.8370590038208652,
+      "learning_rate": 6.0235307363562524e-06,
+      "loss": 0.8331,
+      "step": 3035
+    },
+    {
+      "epoch": 0.49182980100307394,
+      "grad_norm": 0.8387976240360758,
+      "learning_rate": 6.009704915341835e-06,
+      "loss": 0.8429,
+      "step": 3040
+    },
+    {
+      "epoch": 0.49263873159682897,
+      "grad_norm": 0.849367045833896,
+      "learning_rate": 5.9958710410628515e-06,
+      "loss": 0.8245,
+      "step": 3045
+    },
+    {
+      "epoch": 0.49344766219058406,
+      "grad_norm": 0.817264343403589,
+      "learning_rate": 5.9820292238563404e-06,
+      "loss": 0.8279,
+      "step": 3050
+    },
+    {
+      "epoch": 0.4942565927843391,
+      "grad_norm": 0.8397242273647842,
+      "learning_rate": 5.96817957412269e-06,
+      "loss": 0.8211,
+      "step": 3055
+    },
+    {
+      "epoch": 0.4950655233780942,
+      "grad_norm": 0.8292950717170571,
+      "learning_rate": 5.954322202324759e-06,
+      "loss": 0.8447,
+      "step": 3060
+    },
+    {
+      "epoch": 0.4958744539718492,
+      "grad_norm": 0.8153225989185483,
+      "learning_rate": 5.940457218987003e-06,
+      "loss": 0.849,
+      "step": 3065
+    },
+    {
+      "epoch": 0.4966833845656043,
+      "grad_norm": 0.8817065793496368,
+      "learning_rate": 5.926584734694579e-06,
+      "loss": 0.831,
+      "step": 3070
+    },
+    {
+      "epoch": 0.4974923151593593,
+      "grad_norm": 0.8223135029609022,
+      "learning_rate": 5.912704860092473e-06,
+      "loss": 0.8442,
+      "step": 3075
+    },
+    {
+      "epoch": 0.4983012457531144,
+      "grad_norm": 0.8254057255135826,
+      "learning_rate": 5.898817705884615e-06,
+      "loss": 0.8231,
+      "step": 3080
+    },
+    {
+      "epoch": 0.49911017634686944,
+      "grad_norm": 0.8516771176045018,
+      "learning_rate": 5.8849233828329964e-06,
+      "loss": 0.8325,
+      "step": 3085
+    },
+    {
+      "epoch": 0.49991910694062447,
+      "grad_norm": 0.8770116756575128,
+      "learning_rate": 5.871022001756786e-06,
+      "loss": 0.8353,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5007280375343796,
+      "grad_norm": 0.8416392905846662,
+      "learning_rate": 5.857113673531446e-06,
+      "loss": 0.8411,
+      "step": 3095
+    },
+    {
+      "epoch": 0.5015369681281346,
+      "grad_norm": 0.8727681516631783,
+      "learning_rate": 5.843198509087848e-06,
+      "loss": 0.8327,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5023458987218896,
+      "grad_norm": 0.8122836804739928,
+      "learning_rate": 5.829276619411392e-06,
+      "loss": 0.8156,
+      "step": 3105
+    },
+    {
+      "epoch": 0.5031548293156447,
+      "grad_norm": 0.8419116014056326,
+      "learning_rate": 5.815348115541112e-06,
+      "loss": 0.8406,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5039637599093998,
+      "grad_norm": 0.8496791968992736,
+      "learning_rate": 5.801413108568798e-06,
+      "loss": 0.8428,
+      "step": 3115
+    },
+    {
+      "epoch": 0.5047726905031549,
+      "grad_norm": 0.8370866951694025,
+      "learning_rate": 5.787471709638109e-06,
+      "loss": 0.8397,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5055816210969099,
+      "grad_norm": 0.8807740954559549,
+      "learning_rate": 5.7735240299436825e-06,
+      "loss": 0.8281,
+      "step": 3125
+    },
+    {
+      "epoch": 0.5063905516906649,
+      "grad_norm": 0.880842454890424,
+      "learning_rate": 5.759570180730255e-06,
+      "loss": 0.8261,
+      "step": 3130
+    },
+    {
+      "epoch": 0.50719948228442,
+      "grad_norm": 0.8501925674960217,
+      "learning_rate": 5.745610273291766e-06,
+      "loss": 0.8182,
+      "step": 3135
+    },
+    {
+      "epoch": 0.508008412878175,
+      "grad_norm": 0.7885937426308233,
+      "learning_rate": 5.731644418970478e-06,
+      "loss": 0.8006,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5088173434719301,
+      "grad_norm": 0.8453255098699243,
+      "learning_rate": 5.717672729156082e-06,
+      "loss": 0.8408,
+      "step": 3145
+    },
+    {
+      "epoch": 0.5096262740656852,
+      "grad_norm": 0.9027259053341657,
+      "learning_rate": 5.703695315284814e-06,
+      "loss": 0.8415,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5104352046594403,
+      "grad_norm": 0.8760720225211055,
+      "learning_rate": 5.689712288838561e-06,
+      "loss": 0.8459,
+      "step": 3155
+    },
+    {
+      "epoch": 0.5112441352531952,
+      "grad_norm": 0.8371540278282537,
+      "learning_rate": 5.675723761343981e-06,
+      "loss": 0.8217,
+      "step": 3160
+    },
+    {
+      "epoch": 0.5120530658469503,
+      "grad_norm": 0.8832530229547365,
+      "learning_rate": 5.661729844371602e-06,
+      "loss": 0.8321,
+      "step": 3165
+    },
+    {
+      "epoch": 0.5128619964407054,
+      "grad_norm": 0.8466920163343052,
+      "learning_rate": 5.64773064953494e-06,
+      "loss": 0.8443,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5136709270344605,
+      "grad_norm": 0.8386270391117758,
+      "learning_rate": 5.633726288489609e-06,
+      "loss": 0.8215,
+      "step": 3175
+    },
+    {
+      "epoch": 0.5144798576282155,
+      "grad_norm": 0.8968915867410968,
+      "learning_rate": 5.619716872932422e-06,
+      "loss": 0.8369,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5152887882219706,
+      "grad_norm": 0.8572970506555265,
+      "learning_rate": 5.6057025146005125e-06,
+      "loss": 0.8312,
+      "step": 3185
+    },
+    {
+      "epoch": 0.5160977188157256,
+      "grad_norm": 0.856451059153177,
+      "learning_rate": 5.591683325270434e-06,
+      "loss": 0.828,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5169066494094806,
+      "grad_norm": 0.8777484476404416,
+      "learning_rate": 5.577659416757267e-06,
+      "loss": 0.8408,
+      "step": 3195
+    },
+    {
+      "epoch": 0.5177155800032357,
+      "grad_norm": 0.8571851342757004,
+      "learning_rate": 5.5636309009137404e-06,
+      "loss": 0.8337,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5185245105969908,
+      "grad_norm": 0.829598576934011,
+      "learning_rate": 5.549597889629325e-06,
+      "loss": 0.8507,
+      "step": 3205
+    },
+    {
+      "epoch": 0.5193334411907459,
+      "grad_norm": 0.8763914543424554,
+      "learning_rate": 5.535560494829345e-06,
+      "loss": 0.8363,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5201423717845008,
+      "grad_norm": 0.8614187326018883,
+      "learning_rate": 5.521518828474092e-06,
+      "loss": 0.8254,
+      "step": 3215
+    },
+    {
+      "epoch": 0.5209513023782559,
+      "grad_norm": 0.8803171537853229,
+      "learning_rate": 5.507473002557922e-06,
+      "loss": 0.822,
+      "step": 3220
+    },
+    {
+      "epoch": 0.521760232972011,
+      "grad_norm": 0.8826978372002413,
+      "learning_rate": 5.493423129108373e-06,
+      "loss": 0.8249,
+      "step": 3225
+    },
+    {
+      "epoch": 0.5225691635657661,
+      "grad_norm": 0.8712751190362638,
+      "learning_rate": 5.47936932018526e-06,
+      "loss": 0.8423,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5233780941595211,
+      "grad_norm": 0.8238326567914839,
+      "learning_rate": 5.465311687879785e-06,
+      "loss": 0.824,
+      "step": 3235
+    },
+    {
+      "epoch": 0.5241870247532762,
+      "grad_norm": 0.8899415075708137,
+      "learning_rate": 5.4512503443136555e-06,
+      "loss": 0.8663,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5249959553470313,
+      "grad_norm": 0.9621216340368667,
+      "learning_rate": 5.437185401638168e-06,
+      "loss": 0.8414,
+      "step": 3245
+    },
+    {
+      "epoch": 0.5258048859407862,
+      "grad_norm": 0.8156708167135952,
+      "learning_rate": 5.423116972033332e-06,
+      "loss": 0.8249,
+      "step": 3250
+    },
+    {
+      "epoch": 0.5266138165345413,
+      "grad_norm": 0.8179857114204797,
+      "learning_rate": 5.409045167706962e-06,
+      "loss": 0.8421,
+      "step": 3255
+    },
+    {
+      "epoch": 0.5274227471282964,
+      "grad_norm": 0.9127466183923784,
+      "learning_rate": 5.394970100893797e-06,
+      "loss": 0.8359,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5282316777220515,
+      "grad_norm": 0.8430058361090182,
+      "learning_rate": 5.380891883854591e-06,
+      "loss": 0.8152,
+      "step": 3265
+    },
+    {
+      "epoch": 0.5290406083158065,
+      "grad_norm": 0.9184787098267886,
+      "learning_rate": 5.366810628875226e-06,
+      "loss": 0.8253,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5298495389095615,
+      "grad_norm": 0.8924101947820497,
+      "learning_rate": 5.352726448265808e-06,
+      "loss": 0.8386,
+      "step": 3275
+    },
+    {
+      "epoch": 0.5306584695033166,
+      "grad_norm": 0.7686705583327001,
+      "learning_rate": 5.338639454359792e-06,
+      "loss": 0.8305,
+      "step": 3280
+    },
+    {
+      "epoch": 0.5314674000970717,
+      "grad_norm": 0.8264073173393284,
+      "learning_rate": 5.324549759513058e-06,
+      "loss": 0.8279,
+      "step": 3285
+    },
+    {
+      "epoch": 0.5322763306908267,
+      "grad_norm": 0.7898714553275273,
+      "learning_rate": 5.310457476103033e-06,
+      "loss": 0.832,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5330852612845818,
+      "grad_norm": 0.915049917158178,
+      "learning_rate": 5.2963627165277884e-06,
+      "loss": 0.8423,
+      "step": 3295
+    },
+    {
+      "epoch": 0.5338941918783369,
+      "grad_norm": 0.8390857749633213,
+      "learning_rate": 5.28226559320515e-06,
+      "loss": 0.7937,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5347031224720918,
+      "grad_norm": 0.8766888011605986,
+      "learning_rate": 5.268166218571792e-06,
+      "loss": 0.8409,
+      "step": 3305
+    },
+    {
+      "epoch": 0.5355120530658469,
+      "grad_norm": 0.9055264757342382,
+      "learning_rate": 5.254064705082345e-06,
+      "loss": 0.8374,
+      "step": 3310
+    },
+    {
+      "epoch": 0.536320983659602,
+      "grad_norm": 0.9082065836332707,
+      "learning_rate": 5.239961165208499e-06,
+      "loss": 0.8424,
+      "step": 3315
+    },
+    {
+      "epoch": 0.5371299142533571,
+      "grad_norm": 0.8630555734437255,
+      "learning_rate": 5.2258557114381085e-06,
+      "loss": 0.8492,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5379388448471121,
+      "grad_norm": 0.8387927972292225,
+      "learning_rate": 5.211748456274291e-06,
+      "loss": 0.8162,
+      "step": 3325
+    },
+    {
+      "epoch": 0.5387477754408672,
+      "grad_norm": 0.8838319893113061,
+      "learning_rate": 5.197639512234532e-06,
+      "loss": 0.8302,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5395567060346222,
+      "grad_norm": 0.8810506540467933,
+      "learning_rate": 5.183528991849784e-06,
+      "loss": 0.8246,
+      "step": 3335
+    },
+    {
+      "epoch": 0.5403656366283773,
+      "grad_norm": 0.9032568956242902,
+      "learning_rate": 5.16941700766358e-06,
+      "loss": 0.8302,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5411745672221323,
+      "grad_norm": 0.9878346022756306,
+      "learning_rate": 5.155303672231123e-06,
+      "loss": 0.8429,
+      "step": 3345
+    },
+    {
+      "epoch": 0.5419834978158874,
+      "grad_norm": 0.7990363456989189,
+      "learning_rate": 5.141189098118392e-06,
+      "loss": 0.8295,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5427924284096425,
+      "grad_norm": 0.849668666846281,
+      "learning_rate": 5.127073397901248e-06,
+      "loss": 0.8447,
+      "step": 3355
+    },
+    {
+      "epoch": 0.5436013590033975,
+      "grad_norm": 0.8870097311869496,
+      "learning_rate": 5.112956684164532e-06,
+      "loss": 0.8291,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5444102895971525,
+      "grad_norm": 0.8786319797950302,
+      "learning_rate": 5.09883906950117e-06,
+      "loss": 0.8325,
+      "step": 3365
+    },
+    {
+      "epoch": 0.5452192201909076,
+      "grad_norm": 0.8559879223353866,
+      "learning_rate": 5.084720666511276e-06,
+      "loss": 0.8343,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5460281507846627,
+      "grad_norm": 0.8393771435379772,
+      "learning_rate": 5.070601587801246e-06,
+      "loss": 0.8264,
+      "step": 3375
+    },
+    {
+      "epoch": 0.5468370813784177,
+      "grad_norm": 0.8756680886902389,
+      "learning_rate": 5.056481945982871e-06,
+      "loss": 0.8363,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5476460119721728,
+      "grad_norm": 0.8643664631148287,
+      "learning_rate": 5.042361853672429e-06,
+      "loss": 0.8508,
+      "step": 3385
+    },
+    {
+      "epoch": 0.5484549425659279,
+      "grad_norm": 0.8396247416018382,
+      "learning_rate": 5.0282414234897905e-06,
+      "loss": 0.8313,
+      "step": 3390
+    },
+    {
+      "epoch": 0.549263873159683,
+      "grad_norm": 0.8799962735756717,
+      "learning_rate": 5.014120768057526e-06,
+      "loss": 0.8394,
+      "step": 3395
+    },
+    {
+      "epoch": 0.5500728037534379,
+      "grad_norm": 0.8585807475699035,
+      "learning_rate": 5e-06,
+      "loss": 0.8259,
+      "step": 3400
+    },
+    {
+      "epoch": 0.550881734347193,
+      "grad_norm": 0.8967809635243733,
+      "learning_rate": 4.985879231942475e-06,
+      "loss": 0.8232,
+      "step": 3405
+    },
+    {
+      "epoch": 0.5516906649409481,
+      "grad_norm": 0.8853595174049823,
+      "learning_rate": 4.97175857651021e-06,
+      "loss": 0.8196,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5524995955347031,
+      "grad_norm": 0.839604935690722,
+      "learning_rate": 4.957638146327575e-06,
+      "loss": 0.8361,
+      "step": 3415
+    },
+    {
+      "epoch": 0.5533085261284582,
+      "grad_norm": 0.8391467792592279,
+      "learning_rate": 4.943518054017131e-06,
+      "loss": 0.8272,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5541174567222132,
+      "grad_norm": 0.8844616497057926,
+      "learning_rate": 4.929398412198756e-06,
+      "loss": 0.8233,
+      "step": 3425
+    },
+    {
+      "epoch": 0.5549263873159683,
+      "grad_norm": 0.8052750334117075,
+      "learning_rate": 4.915279333488726e-06,
+      "loss": 0.8479,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5557353179097233,
+      "grad_norm": 0.8647851317448605,
+      "learning_rate": 4.90116093049883e-06,
+      "loss": 0.819,
+      "step": 3435
+    },
+    {
+      "epoch": 0.5565442485034784,
+      "grad_norm": 0.8453777869125878,
+      "learning_rate": 4.887043315835469e-06,
+      "loss": 0.8536,
+      "step": 3440
+    },
+    {
+      "epoch": 0.5573531790972335,
+      "grad_norm": 0.8705052285115737,
+      "learning_rate": 4.872926602098756e-06,
+      "loss": 0.8277,
+      "step": 3445
+    },
+    {
+      "epoch": 0.5581621096909886,
+      "grad_norm": 0.861071586334625,
+      "learning_rate": 4.85881090188161e-06,
+      "loss": 0.8081,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5589710402847435,
+      "grad_norm": 0.8558895543087007,
+      "learning_rate": 4.844696327768878e-06,
+      "loss": 0.8086,
+      "step": 3455
+    },
+    {
+      "epoch": 0.5597799708784986,
+      "grad_norm": 0.9566303499060257,
+      "learning_rate": 4.83058299233642e-06,
+      "loss": 0.8388,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5605889014722537,
+      "grad_norm": 0.8480580281776552,
+      "learning_rate": 4.8164710081502165e-06,
+      "loss": 0.8464,
+      "step": 3465
+    },
+    {
+      "epoch": 0.5613978320660087,
+      "grad_norm": 0.8577968798548672,
+      "learning_rate": 4.802360487765471e-06,
+      "loss": 0.8331,
+      "step": 3470
+    },
+    {
+      "epoch": 0.5622067626597638,
+      "grad_norm": 0.8683642459249713,
+      "learning_rate": 4.788251543725711e-06,
+      "loss": 0.8379,
+      "step": 3475
+    },
+    {
+      "epoch": 0.5630156932535189,
+      "grad_norm": 0.8893404896273497,
+      "learning_rate": 4.774144288561893e-06,
+      "loss": 0.8138,
+      "step": 3480
+    },
+    {
+      "epoch": 0.5638246238472739,
+      "grad_norm": 0.8886059419047985,
+      "learning_rate": 4.7600388347915025e-06,
+      "loss": 0.8306,
+      "step": 3485
+    },
+    {
+      "epoch": 0.5646335544410289,
+      "grad_norm": 0.8630514971290992,
+      "learning_rate": 4.745935294917658e-06,
+      "loss": 0.8348,
+      "step": 3490
+    },
+    {
+      "epoch": 0.565442485034784,
+      "grad_norm": 0.8662086181368763,
+      "learning_rate": 4.731833781428208e-06,
+      "loss": 0.8272,
+      "step": 3495
+    },
+    {
+      "epoch": 0.5662514156285391,
+      "grad_norm": 0.8982952976447754,
+      "learning_rate": 4.7177344067948526e-06,
+      "loss": 0.8525,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5670603462222942,
+      "grad_norm": 0.8422016831549223,
+      "learning_rate": 4.703637283472213e-06,
+      "loss": 0.8501,
+      "step": 3505
+    },
+    {
+      "epoch": 0.5678692768160492,
+      "grad_norm": 0.9548881231159024,
+      "learning_rate": 4.689542523896969e-06,
+      "loss": 0.8328,
+      "step": 3510
+    },
+    {
+      "epoch": 0.5686782074098042,
+      "grad_norm": 0.9166786357525355,
+      "learning_rate": 4.6754502404869434e-06,
+      "loss": 0.8205,
+      "step": 3515
+    },
+    {
+      "epoch": 0.5694871380035593,
+      "grad_norm": 0.9061529794904933,
+      "learning_rate": 4.661360545640209e-06,
+      "loss": 0.8352,
+      "step": 3520
+    },
+    {
+      "epoch": 0.5702960685973143,
+      "grad_norm": 0.8412734635963403,
+      "learning_rate": 4.647273551734193e-06,
+      "loss": 0.8348,
+      "step": 3525
+    },
+    {
+      "epoch": 0.5711049991910694,
+      "grad_norm": 0.8664917924780835,
+      "learning_rate": 4.633189371124778e-06,
+      "loss": 0.8274,
+      "step": 3530
+    },
+    {
+      "epoch": 0.5719139297848245,
+      "grad_norm": 0.8421186404143896,
+      "learning_rate": 4.619108116145411e-06,
+      "loss": 0.8354,
+      "step": 3535
+    },
+    {
+      "epoch": 0.5727228603785796,
+      "grad_norm": 0.9394380176490387,
+      "learning_rate": 4.6050298991062045e-06,
+      "loss": 0.8269,
+      "step": 3540
+    },
+    {
+      "epoch": 0.5735317909723345,
+      "grad_norm": 0.8660091216092733,
+      "learning_rate": 4.590954832293039e-06,
+      "loss": 0.8431,
+      "step": 3545
+    },
+    {
+      "epoch": 0.5743407215660896,
+      "grad_norm": 0.8453775529626361,
+      "learning_rate": 4.57688302796667e-06,
+      "loss": 0.8189,
+      "step": 3550
+    },
+    {
+      "epoch": 0.5751496521598447,
+      "grad_norm": 0.9293223006762544,
+      "learning_rate": 4.562814598361834e-06,
+      "loss": 0.8272,
+      "step": 3555
+    },
+    {
+      "epoch": 0.5759585827535998,
+      "grad_norm": 0.8682075221715769,
+      "learning_rate": 4.548749655686346e-06,
+      "loss": 0.8345,
+      "step": 3560
+    },
+    {
+      "epoch": 0.5767675133473548,
+      "grad_norm": 0.8962654121926869,
+      "learning_rate": 4.534688312120216e-06,
+      "loss": 0.8328,
+      "step": 3565
+    },
+    {
+      "epoch": 0.5775764439411099,
+      "grad_norm": 0.8465460016413964,
+      "learning_rate": 4.520630679814742e-06,
+      "loss": 0.8203,
+      "step": 3570
+    },
+    {
+      "epoch": 0.5783853745348649,
+      "grad_norm": 0.8542030722056939,
+      "learning_rate": 4.506576870891628e-06,
+      "loss": 0.8356,
+      "step": 3575
+    },
+    {
+      "epoch": 0.5791943051286199,
+      "grad_norm": 0.8664072736373561,
+      "learning_rate": 4.49252699744208e-06,
+      "loss": 0.8282,
+      "step": 3580
+    },
+    {
+      "epoch": 0.580003235722375,
+      "grad_norm": 0.8908304692018135,
+      "learning_rate": 4.47848117152591e-06,
+      "loss": 0.8486,
+      "step": 3585
+    },
+    {
+      "epoch": 0.5808121663161301,
+      "grad_norm": 0.9604353292765517,
+      "learning_rate": 4.464439505170656e-06,
+      "loss": 0.828,
+      "step": 3590
+    },
+    {
+      "epoch": 0.5816210969098852,
+      "grad_norm": 0.8847638351175713,
+      "learning_rate": 4.450402110370677e-06,
+      "loss": 0.8282,
+      "step": 3595
+    },
+    {
+      "epoch": 0.5824300275036401,
+      "grad_norm": 0.8448125761174995,
+      "learning_rate": 4.43636909908626e-06,
+      "loss": 0.8381,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5832389580973952,
+      "grad_norm": 0.899925581945171,
+      "learning_rate": 4.422340583242733e-06,
+      "loss": 0.8316,
+      "step": 3605
+    },
+    {
+      "epoch": 0.5840478886911503,
+      "grad_norm": 0.8270612926284294,
+      "learning_rate": 4.408316674729569e-06,
+      "loss": 0.8369,
+      "step": 3610
+    },
+    {
+      "epoch": 0.5848568192849054,
+      "grad_norm": 0.8514310722916417,
+      "learning_rate": 4.394297485399488e-06,
+      "loss": 0.8123,
+      "step": 3615
+    },
+    {
+      "epoch": 0.5856657498786604,
+      "grad_norm": 0.847607926161957,
+      "learning_rate": 4.3802831270675785e-06,
+      "loss": 0.8125,
+      "step": 3620
+    },
+    {
+      "epoch": 0.5864746804724155,
+      "grad_norm": 0.8743645770925027,
+      "learning_rate": 4.3662737115103925e-06,
+      "loss": 0.8175,
+      "step": 3625
+    },
+    {
+      "epoch": 0.5872836110661706,
+      "grad_norm": 0.8872414789266562,
+      "learning_rate": 4.35226935046506e-06,
+      "loss": 0.8128,
+      "step": 3630
+    },
+    {
+      "epoch": 0.5880925416599255,
+      "grad_norm": 0.8748952049996549,
+      "learning_rate": 4.338270155628401e-06,
+      "loss": 0.8323,
+      "step": 3635
+    },
+    {
+      "epoch": 0.5889014722536806,
+      "grad_norm": 0.9131605811984947,
+      "learning_rate": 4.324276238656023e-06,
+      "loss": 0.8266,
+      "step": 3640
+    },
+    {
+      "epoch": 0.5897104028474357,
+      "grad_norm": 0.9378108739015335,
+      "learning_rate": 4.31028771116144e-06,
+      "loss": 0.8394,
+      "step": 3645
+    },
+    {
+      "epoch": 0.5905193334411908,
+      "grad_norm": 0.9270250400148877,
+      "learning_rate": 4.2963046847151875e-06,
+      "loss": 0.8415,
+      "step": 3650
+    },
+    {
+      "epoch": 0.5913282640349458,
+      "grad_norm": 0.8427644597969671,
+      "learning_rate": 4.282327270843919e-06,
+      "loss": 0.8206,
+      "step": 3655
+    },
+    {
+      "epoch": 0.5921371946287008,
+      "grad_norm": 0.9019710217551041,
+      "learning_rate": 4.268355581029523e-06,
+      "loss": 0.8414,
+      "step": 3660
+    },
+    {
+      "epoch": 0.5929461252224559,
+      "grad_norm": 0.9361559273358337,
+      "learning_rate": 4.254389726708234e-06,
+      "loss": 0.8388,
+      "step": 3665
+    },
+    {
+      "epoch": 0.593755055816211,
+      "grad_norm": 0.8479002824514841,
+      "learning_rate": 4.240429819269746e-06,
+      "loss": 0.8151,
+      "step": 3670
+    },
+    {
+      "epoch": 0.594563986409966,
+      "grad_norm": 0.8837530926010035,
+      "learning_rate": 4.226475970056318e-06,
+      "loss": 0.8212,
+      "step": 3675
+    },
+    {
+      "epoch": 0.5953729170037211,
+      "grad_norm": 0.8660636262138016,
+      "learning_rate": 4.212528290361893e-06,
+      "loss": 0.8571,
+      "step": 3680
+    },
+    {
+      "epoch": 0.5961818475974762,
+      "grad_norm": 0.9246667975571174,
+      "learning_rate": 4.198586891431203e-06,
+      "loss": 0.8397,
+      "step": 3685
+    },
+    {
+      "epoch": 0.5969907781912311,
+      "grad_norm": 0.90386970265181,
+      "learning_rate": 4.1846518844588906e-06,
+      "loss": 0.8365,
+      "step": 3690
+    },
+    {
+      "epoch": 0.5977997087849862,
+      "grad_norm": 0.9258635632218947,
+      "learning_rate": 4.170723380588609e-06,
+      "loss": 0.8403,
+      "step": 3695
+    },
+    {
+      "epoch": 0.5986086393787413,
+      "grad_norm": 0.8741091788263792,
+      "learning_rate": 4.156801490912153e-06,
+      "loss": 0.822,
+      "step": 3700
+    },
+    {
+      "epoch": 0.5994175699724964,
+      "grad_norm": 0.8552618808724114,
+      "learning_rate": 4.142886326468556e-06,
+      "loss": 0.8414,
+      "step": 3705
+    },
+    {
+      "epoch": 0.6002265005662514,
+      "grad_norm": 0.8577590928960753,
+      "learning_rate": 4.128977998243215e-06,
+      "loss": 0.8234,
+      "step": 3710
+    },
+    {
+      "epoch": 0.6010354311600065,
+      "grad_norm": 0.9640946415628675,
+      "learning_rate": 4.115076617167004e-06,
+      "loss": 0.8344,
+      "step": 3715
+    },
+    {
+      "epoch": 0.6018443617537615,
+      "grad_norm": 0.8840143210557655,
+      "learning_rate": 4.101182294115388e-06,
+      "loss": 0.8211,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6026532923475166,
+      "grad_norm": 0.8574994142653399,
+      "learning_rate": 4.087295139907528e-06,
+      "loss": 0.8166,
+      "step": 3725
+    },
+    {
+      "epoch": 0.6034622229412716,
+      "grad_norm": 0.8964092793127811,
+      "learning_rate": 4.073415265305422e-06,
+      "loss": 0.8171,
+      "step": 3730
+    },
+    {
+      "epoch": 0.6042711535350267,
+      "grad_norm": 0.963355011643523,
+      "learning_rate": 4.059542781012998e-06,
+      "loss": 0.8223,
+      "step": 3735
+    },
+    {
+      "epoch": 0.6050800841287818,
+      "grad_norm": 0.8678396011067857,
+      "learning_rate": 4.045677797675242e-06,
+      "loss": 0.8203,
+      "step": 3740
+    },
+    {
+      "epoch": 0.6058890147225368,
+      "grad_norm": 0.9121282453763787,
+      "learning_rate": 4.031820425877313e-06,
+      "loss": 0.8219,
+      "step": 3745
+    },
+    {
+      "epoch": 0.6066979453162918,
+      "grad_norm": 0.9216469820283191,
+      "learning_rate": 4.017970776143662e-06,
+      "loss": 0.8352,
+      "step": 3750
+    },
+    {
+      "epoch": 0.6075068759100469,
+      "grad_norm": 0.908885616786373,
+      "learning_rate": 4.004128958937149e-06,
+      "loss": 0.8241,
+      "step": 3755
+    },
+    {
+      "epoch": 0.608315806503802,
+      "grad_norm": 0.8942143952561401,
+      "learning_rate": 3.990295084658166e-06,
+      "loss": 0.8361,
+      "step": 3760
+    },
+    {
+      "epoch": 0.609124737097557,
+      "grad_norm": 0.8730532396217757,
+      "learning_rate": 3.976469263643748e-06,
+      "loss": 0.8384,
+      "step": 3765
+    },
+    {
+      "epoch": 0.6099336676913121,
+      "grad_norm": 0.9177096179325701,
+      "learning_rate": 3.962651606166703e-06,
+      "loss": 0.8184,
+      "step": 3770
+    },
+    {
+      "epoch": 0.6107425982850672,
+      "grad_norm": 0.9852181301325248,
+      "learning_rate": 3.948842222434728e-06,
+      "loss": 0.817,
+      "step": 3775
+    },
+    {
+      "epoch": 0.6115515288788222,
+      "grad_norm": 0.8956044101647403,
+      "learning_rate": 3.935041222589524e-06,
+      "loss": 0.818,
+      "step": 3780
+    },
+    {
+      "epoch": 0.6123604594725772,
+      "grad_norm": 0.9733581516182005,
+      "learning_rate": 3.9212487167059265e-06,
+      "loss": 0.8223,
+      "step": 3785
+    },
+    {
+      "epoch": 0.6131693900663323,
+      "grad_norm": 0.8650513303090718,
+      "learning_rate": 3.907464814791029e-06,
+      "loss": 0.8033,
+      "step": 3790
+    },
+    {
+      "epoch": 0.6139783206600874,
+      "grad_norm": 0.99672664240277,
+      "learning_rate": 3.893689626783294e-06,
+      "loss": 0.836,
+      "step": 3795
+    },
+    {
+      "epoch": 0.6147872512538424,
+      "grad_norm": 0.908345822113972,
+      "learning_rate": 3.8799232625516884e-06,
+      "loss": 0.8478,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6155961818475975,
+      "grad_norm": 0.8836902180965445,
+      "learning_rate": 3.866165831894796e-06,
+      "loss": 0.8235,
+      "step": 3805
+    },
+    {
+      "epoch": 0.6164051124413525,
+      "grad_norm": 0.9264471672860939,
+      "learning_rate": 3.852417444539953e-06,
+      "loss": 0.8247,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6172140430351076,
+      "grad_norm": 0.894173445093799,
+      "learning_rate": 3.8386782101423665e-06,
+      "loss": 0.8235,
+      "step": 3815
+    },
+    {
+      "epoch": 0.6180229736288626,
+      "grad_norm": 0.9438338724077253,
+      "learning_rate": 3.824948238284238e-06,
+      "loss": 0.8339,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6188319042226177,
+      "grad_norm": 0.9012049433189601,
+      "learning_rate": 3.8112276384738973e-06,
+      "loss": 0.8217,
+      "step": 3825
+    },
+    {
+      "epoch": 0.6196408348163728,
+      "grad_norm": 0.9631072787593842,
+      "learning_rate": 3.797516520144919e-06,
+      "loss": 0.821,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6204497654101279,
+      "grad_norm": 0.8562741768165036,
+      "learning_rate": 3.7838149926552565e-06,
+      "loss": 0.8254,
+      "step": 3835
+    },
+    {
+      "epoch": 0.6212586960038828,
+      "grad_norm": 0.9007576913836781,
+      "learning_rate": 3.770123165286369e-06,
+      "loss": 0.8187,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6220676265976379,
+      "grad_norm": 1.0271531293622427,
+      "learning_rate": 3.7564411472423467e-06,
+      "loss": 0.8358,
+      "step": 3845
+    },
+    {
+      "epoch": 0.622876557191393,
+      "grad_norm": 0.8914948323397264,
+      "learning_rate": 3.7427690476490462e-06,
+      "loss": 0.8275,
+      "step": 3850
+    },
+    {
+      "epoch": 0.623685487785148,
+      "grad_norm": 0.9328213480538451,
+      "learning_rate": 3.7291069755532146e-06,
+      "loss": 0.8167,
+      "step": 3855
+    },
+    {
+      "epoch": 0.6244944183789031,
+      "grad_norm": 0.8892937385561739,
+      "learning_rate": 3.7154550399216137e-06,
+      "loss": 0.8347,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6253033489726582,
+      "grad_norm": 0.8798858697832553,
+      "learning_rate": 3.7018133496401688e-06,
+      "loss": 0.8383,
+      "step": 3865
+    },
+    {
+      "epoch": 0.6261122795664132,
+      "grad_norm": 0.98439076252917,
+      "learning_rate": 3.688182013513085e-06,
+      "loss": 0.8274,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6269212101601682,
+      "grad_norm": 0.8370789210688374,
+      "learning_rate": 3.674561140261983e-06,
+      "loss": 0.8178,
+      "step": 3875
+    },
+    {
+      "epoch": 0.6277301407539233,
+      "grad_norm": 0.9153894187914345,
+      "learning_rate": 3.660950838525036e-06,
+      "loss": 0.8258,
+      "step": 3880
+    },
+    {
+      "epoch": 0.6285390713476784,
+      "grad_norm": 0.8922726078616626,
+      "learning_rate": 3.647351216856099e-06,
+      "loss": 0.8581,
+      "step": 3885
+    },
+    {
+      "epoch": 0.6293480019414335,
+      "grad_norm": 0.9058364873912355,
+      "learning_rate": 3.633762383723841e-06,
+      "loss": 0.827,
+      "step": 3890
+    },
+    {
+      "epoch": 0.6301569325351885,
+      "grad_norm": 0.8703352651889255,
+      "learning_rate": 3.6201844475108884e-06,
+      "loss": 0.8491,
+      "step": 3895
+    },
+    {
+      "epoch": 0.6309658631289435,
+      "grad_norm": 0.892101387058938,
+      "learning_rate": 3.606617516512953e-06,
+      "loss": 0.8175,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6317747937226986,
+      "grad_norm": 0.9365570155133393,
+      "learning_rate": 3.5930616989379697e-06,
+      "loss": 0.8388,
+      "step": 3905
+    },
+    {
+      "epoch": 0.6325837243164536,
+      "grad_norm": 0.8773819774765554,
+      "learning_rate": 3.5795171029052383e-06,
+      "loss": 0.8294,
+      "step": 3910
+    },
+    {
+      "epoch": 0.6333926549102087,
+      "grad_norm": 0.8714533378375239,
+      "learning_rate": 3.5659838364445505e-06,
+      "loss": 0.8346,
+      "step": 3915
+    },
+    {
+      "epoch": 0.6342015855039638,
+      "grad_norm": 0.9825642646199596,
+      "learning_rate": 3.552462007495338e-06,
+      "loss": 0.8376,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6350105160977189,
+      "grad_norm": 0.899943511678241,
+      "learning_rate": 3.5389517239058126e-06,
+      "loss": 0.8251,
+      "step": 3925
+    },
+    {
+      "epoch": 0.6358194466914738,
+      "grad_norm": 0.9817959591645147,
+      "learning_rate": 3.5254530934320956e-06,
+      "loss": 0.8332,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6366283772852289,
+      "grad_norm": 0.939309664119045,
+      "learning_rate": 3.5119662237373684e-06,
+      "loss": 0.8172,
+      "step": 3935
+    },
+    {
+      "epoch": 0.637437307878984,
+      "grad_norm": 0.9545145278463266,
+      "learning_rate": 3.4984912223910105e-06,
+      "loss": 0.8137,
+      "step": 3940
+    },
+    {
+      "epoch": 0.6382462384727391,
+      "grad_norm": 0.9284835788915782,
+      "learning_rate": 3.485028196867738e-06,
+      "loss": 0.825,
+      "step": 3945
+    },
+    {
+      "epoch": 0.6390551690664941,
+      "grad_norm": 1.0025398605998348,
+      "learning_rate": 3.4715772545467507e-06,
+      "loss": 0.8359,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6398640996602492,
+      "grad_norm": 0.9250686960694907,
+      "learning_rate": 3.458138502710876e-06,
+      "loss": 0.8445,
+      "step": 3955
+    },
+    {
+      "epoch": 0.6406730302540042,
+      "grad_norm": 0.885581537464598,
+      "learning_rate": 3.444712048545711e-06,
+      "loss": 0.8241,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6414819608477592,
+      "grad_norm": 0.9162743922784746,
+      "learning_rate": 3.431297999138768e-06,
+      "loss": 0.8178,
+      "step": 3965
+    },
+    {
+      "epoch": 0.6422908914415143,
+      "grad_norm": 0.9997360512776706,
+      "learning_rate": 3.417896461478619e-06,
+      "loss": 0.849,
+      "step": 3970
+    },
+    {
+      "epoch": 0.6430998220352694,
+      "grad_norm": 0.9230165688594162,
+      "learning_rate": 3.4045075424540484e-06,
+      "loss": 0.8048,
+      "step": 3975
+    },
+    {
+      "epoch": 0.6439087526290245,
+      "grad_norm": 0.8899064142475545,
+      "learning_rate": 3.3911313488531907e-06,
+      "loss": 0.8286,
+      "step": 3980
+    },
+    {
+      "epoch": 0.6447176832227794,
+      "grad_norm": 0.8669101247237674,
+      "learning_rate": 3.37776798736269e-06,
+      "loss": 0.8119,
+      "step": 3985
+    },
+    {
+      "epoch": 0.6455266138165345,
+      "grad_norm": 0.9167759750434173,
+      "learning_rate": 3.364417564566839e-06,
+      "loss": 0.8472,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6463355444102896,
+      "grad_norm": 0.9521570771011725,
+      "learning_rate": 3.351080186946736e-06,
+      "loss": 0.8251,
+      "step": 3995
+    },
+    {
+      "epoch": 0.6471444750040447,
+      "grad_norm": 0.8382337519442962,
+      "learning_rate": 3.3377559608794273e-06,
+      "loss": 0.8453,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6479534055977997,
+      "grad_norm": 0.9439756459989987,
+      "learning_rate": 3.3244449926370713e-06,
+      "loss": 0.8359,
+      "step": 4005
+    },
+    {
+      "epoch": 0.6487623361915548,
+      "grad_norm": 0.9306298379750102,
+      "learning_rate": 3.3111473883860813e-06,
+      "loss": 0.8196,
+      "step": 4010
+    },
+    {
+      "epoch": 0.6495712667853099,
+      "grad_norm": 0.8706086854228088,
+      "learning_rate": 3.2978632541862788e-06,
+      "loss": 0.8252,
+      "step": 4015
+    },
+    {
+      "epoch": 0.6503801973790648,
+      "grad_norm": 0.918543174205597,
+      "learning_rate": 3.2845926959900555e-06,
+      "loss": 0.8194,
+      "step": 4020
+    },
+    {
+      "epoch": 0.6511891279728199,
+      "grad_norm": 0.9355133643042608,
+      "learning_rate": 3.2713358196415147e-06,
+      "loss": 0.8279,
+      "step": 4025
+    },
+    {
+      "epoch": 0.651998058566575,
+      "grad_norm": 0.8875680347636398,
+      "learning_rate": 3.2580927308756426e-06,
+      "loss": 0.8336,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6528069891603301,
+      "grad_norm": 0.974049905246242,
+      "learning_rate": 3.2448635353174553e-06,
+      "loss": 0.8334,
+      "step": 4035
+    },
+    {
+      "epoch": 0.6536159197540851,
+      "grad_norm": 0.8679703990844894,
+      "learning_rate": 3.231648338481157e-06,
+      "loss": 0.8355,
+      "step": 4040
+    },
+    {
+      "epoch": 0.6544248503478401,
+      "grad_norm": 0.9299757764874559,
+      "learning_rate": 3.2184472457693005e-06,
+      "loss": 0.8401,
+      "step": 4045
+    },
+    {
+      "epoch": 0.6552337809415952,
+      "grad_norm": 0.8615524033230236,
+      "learning_rate": 3.2052603624719516e-06,
+      "loss": 0.8217,
+      "step": 4050
+    },
+    {
+      "epoch": 0.6560427115353503,
+      "grad_norm": 0.9060239627418345,
+      "learning_rate": 3.1920877937658325e-06,
+      "loss": 0.8173,
+      "step": 4055
+    },
+    {
+      "epoch": 0.6568516421291053,
+      "grad_norm": 0.9398327715146758,
+      "learning_rate": 3.178929644713504e-06,
+      "loss": 0.8412,
+      "step": 4060
+    },
+    {
+      "epoch": 0.6576605727228604,
+      "grad_norm": 0.8393737863171705,
+      "learning_rate": 3.1657860202625145e-06,
+      "loss": 0.8368,
+      "step": 4065
+    },
+    {
+      "epoch": 0.6584695033166155,
+      "grad_norm": 0.8802074487722802,
+      "learning_rate": 3.1526570252445665e-06,
+      "loss": 0.8203,
+      "step": 4070
+    },
+    {
+      "epoch": 0.6592784339103704,
+      "grad_norm": 0.9133104861467479,
+      "learning_rate": 3.1395427643746802e-06,
+      "loss": 0.8168,
+      "step": 4075
+    },
+    {
+      "epoch": 0.6600873645041255,
+      "grad_norm": 0.9090731300453533,
+      "learning_rate": 3.1264433422503564e-06,
+      "loss": 0.8466,
+      "step": 4080
+    },
+    {
+      "epoch": 0.6608962950978806,
+      "grad_norm": 0.9229902190782812,
+      "learning_rate": 3.113358863350747e-06,
+      "loss": 0.8334,
+      "step": 4085
+    },
+    {
+      "epoch": 0.6617052256916357,
+      "grad_norm": 0.8999096831432444,
+      "learning_rate": 3.100289432035818e-06,
+      "loss": 0.8145,
+      "step": 4090
+    },
+    {
+      "epoch": 0.6625141562853907,
+      "grad_norm": 0.9226844151049607,
+      "learning_rate": 3.087235152545517e-06,
+      "loss": 0.8245,
+      "step": 4095
+    },
+    {
+      "epoch": 0.6633230868791458,
+      "grad_norm": 0.9923339907314755,
+      "learning_rate": 3.074196128998943e-06,
+      "loss": 0.8067,
+      "step": 4100
+    },
+    {
+      "epoch": 0.6641320174729008,
+      "grad_norm": 0.8793470002054651,
+      "learning_rate": 3.0611724653935184e-06,
+      "loss": 0.8154,
+      "step": 4105
+    },
+    {
+      "epoch": 0.6649409480666559,
+      "grad_norm": 0.9952922559372246,
+      "learning_rate": 3.048164265604152e-06,
+      "loss": 0.8339,
+      "step": 4110
+    },
+    {
+      "epoch": 0.6657498786604109,
+      "grad_norm": 0.8934191696631508,
+      "learning_rate": 3.035171633382419e-06,
+      "loss": 0.8126,
+      "step": 4115
+    },
+    {
+      "epoch": 0.666558809254166,
+      "grad_norm": 0.9074584444838903,
+      "learning_rate": 3.0221946723557274e-06,
+      "loss": 0.814,
+      "step": 4120
+    },
+    {
+      "epoch": 0.6673677398479211,
+      "grad_norm": 0.8823615491806162,
+      "learning_rate": 3.009233486026497e-06,
+      "loss": 0.8231,
+      "step": 4125
+    },
+    {
+      "epoch": 0.6681766704416761,
+      "grad_norm": 0.8985312723090128,
+      "learning_rate": 2.9962881777713326e-06,
+      "loss": 0.8279,
+      "step": 4130
+    },
+    {
+      "epoch": 0.6689856010354311,
+      "grad_norm": 0.8759610512792171,
+      "learning_rate": 2.983358850840187e-06,
+      "loss": 0.8503,
+      "step": 4135
+    },
+    {
+      "epoch": 0.6697945316291862,
+      "grad_norm": 0.875080183012667,
+      "learning_rate": 2.97044560835556e-06,
+      "loss": 0.8361,
+      "step": 4140
+    },
+    {
+      "epoch": 0.6706034622229413,
+      "grad_norm": 0.9343434705098017,
+      "learning_rate": 2.9575485533116597e-06,
+      "loss": 0.8422,
+      "step": 4145
+    },
+    {
+      "epoch": 0.6714123928166963,
+      "grad_norm": 0.9509754443485635,
+      "learning_rate": 2.944667788573584e-06,
+      "loss": 0.824,
+      "step": 4150
+    },
+    {
+      "epoch": 0.6722213234104514,
+      "grad_norm": 0.8941436073858621,
+      "learning_rate": 2.9318034168765048e-06,
+      "loss": 0.8329,
+      "step": 4155
+    },
+    {
+      "epoch": 0.6730302540042065,
+      "grad_norm": 0.9216572906600096,
+      "learning_rate": 2.9189555408248436e-06,
+      "loss": 0.838,
+      "step": 4160
+    },
+    {
+      "epoch": 0.6738391845979615,
+      "grad_norm": 0.9049144748685041,
+      "learning_rate": 2.906124262891451e-06,
+      "loss": 0.8392,
+      "step": 4165
+    },
+    {
+      "epoch": 0.6746481151917165,
+      "grad_norm": 0.9088265471558502,
+      "learning_rate": 2.893309685416802e-06,
+      "loss": 0.8219,
+      "step": 4170
+    },
+    {
+      "epoch": 0.6754570457854716,
+      "grad_norm": 0.958093818584568,
+      "learning_rate": 2.880511910608164e-06,
+      "loss": 0.8361,
+      "step": 4175
+    },
+    {
+      "epoch": 0.6762659763792267,
+      "grad_norm": 0.9461864165794798,
+      "learning_rate": 2.8677310405387926e-06,
+      "loss": 0.8145,
+      "step": 4180
+    },
+    {
+      "epoch": 0.6770749069729817,
+      "grad_norm": 0.8862588950970776,
+      "learning_rate": 2.854967177147113e-06,
+      "loss": 0.8101,
+      "step": 4185
+    },
+    {
+      "epoch": 0.6778838375667368,
+      "grad_norm": 0.95284924784784,
+      "learning_rate": 2.8422204222359027e-06,
+      "loss": 0.8355,
+      "step": 4190
+    },
+    {
+      "epoch": 0.6786927681604918,
+      "grad_norm": 0.9473709868364139,
+      "learning_rate": 2.829490877471491e-06,
+      "loss": 0.8213,
+      "step": 4195
+    },
+    {
+      "epoch": 0.6795016987542469,
+      "grad_norm": 0.8870378301848979,
+      "learning_rate": 2.816778644382938e-06,
+      "loss": 0.8143,
+      "step": 4200
+    },
+    {
+      "epoch": 0.6803106293480019,
+      "grad_norm": 0.9018325718849391,
+      "learning_rate": 2.804083824361229e-06,
+      "loss": 0.8265,
+      "step": 4205
+    },
+    {
+      "epoch": 0.681119559941757,
+      "grad_norm": 0.8727955391410293,
+      "learning_rate": 2.7914065186584637e-06,
+      "loss": 0.8254,
+      "step": 4210
+    },
+    {
+      "epoch": 0.6819284905355121,
+      "grad_norm": 1.0739665165244483,
+      "learning_rate": 2.778746828387058e-06,
+      "loss": 0.82,
+      "step": 4215
+    },
+    {
+      "epoch": 0.6827374211292672,
+      "grad_norm": 0.9350732250295917,
+      "learning_rate": 2.766104854518915e-06,
+      "loss": 0.8391,
+      "step": 4220
+    },
+    {
+      "epoch": 0.6835463517230221,
+      "grad_norm": 1.0087459108347565,
+      "learning_rate": 2.753480697884647e-06,
+      "loss": 0.8191,
+      "step": 4225
+    },
+    {
+      "epoch": 0.6843552823167772,
+      "grad_norm": 0.9139563110633456,
+      "learning_rate": 2.740874459172752e-06,
+      "loss": 0.8148,
+      "step": 4230
+    },
+    {
+      "epoch": 0.6851642129105323,
+      "grad_norm": 0.8530060707424758,
+      "learning_rate": 2.728286238928821e-06,
+      "loss": 0.8053,
+      "step": 4235
+    },
+    {
+      "epoch": 0.6859731435042873,
+      "grad_norm": 0.8777686426009732,
+      "learning_rate": 2.7157161375547304e-06,
+      "loss": 0.8114,
+      "step": 4240
+    },
+    {
+      "epoch": 0.6867820740980424,
+      "grad_norm": 0.897912961167507,
+      "learning_rate": 2.7031642553078374e-06,
+      "loss": 0.8418,
+      "step": 4245
+    },
+    {
+      "epoch": 0.6875910046917975,
+      "grad_norm": 0.9279693202641751,
+      "learning_rate": 2.690630692300199e-06,
+      "loss": 0.8404,
+      "step": 4250
+    },
+    {
+      "epoch": 0.6883999352855525,
+      "grad_norm": 0.9248453334693622,
+      "learning_rate": 2.6781155484977495e-06,
+      "loss": 0.8494,
+      "step": 4255
+    },
+    {
+      "epoch": 0.6892088658793075,
+      "grad_norm": 0.9396606484969379,
+      "learning_rate": 2.6656189237195186e-06,
+      "loss": 0.8226,
+      "step": 4260
+    },
+    {
+      "epoch": 0.6900177964730626,
+      "grad_norm": 0.8634815816211927,
+      "learning_rate": 2.6531409176368296e-06,
+      "loss": 0.8265,
+      "step": 4265
+    },
+    {
+      "epoch": 0.6908267270668177,
+      "grad_norm": 0.9135686768769312,
+      "learning_rate": 2.6406816297725086e-06,
+      "loss": 0.8248,
+      "step": 4270
+    },
+    {
+      "epoch": 0.6916356576605728,
+      "grad_norm": 0.92902484349033,
+      "learning_rate": 2.628241159500081e-06,
+      "loss": 0.8173,
+      "step": 4275
+    },
+    {
+      "epoch": 0.6924445882543278,
+      "grad_norm": 0.8687123389790895,
+      "learning_rate": 2.6158196060429926e-06,
+      "loss": 0.8398,
+      "step": 4280
+    },
+    {
+      "epoch": 0.6932535188480828,
+      "grad_norm": 0.9217730078223995,
+      "learning_rate": 2.6034170684738065e-06,
+      "loss": 0.8309,
+      "step": 4285
+    },
+    {
+      "epoch": 0.6940624494418379,
+      "grad_norm": 0.9769331954136037,
+      "learning_rate": 2.591033645713424e-06,
+      "loss": 0.8223,
+      "step": 4290
+    },
+    {
+      "epoch": 0.6948713800355929,
+      "grad_norm": 0.908044715178041,
+      "learning_rate": 2.5786694365302855e-06,
+      "loss": 0.8281,
+      "step": 4295
+    },
+    {
+      "epoch": 0.695680310629348,
+      "grad_norm": 0.9111114311128146,
+      "learning_rate": 2.566324539539583e-06,
+      "loss": 0.8266,
+      "step": 4300
+    },
+    {
+      "epoch": 0.6964892412231031,
+      "grad_norm": 0.9303322655983225,
+      "learning_rate": 2.5539990532024827e-06,
+      "loss": 0.8111,
+      "step": 4305
+    },
+    {
+      "epoch": 0.6972981718168582,
+      "grad_norm": 0.8418692144813529,
+      "learning_rate": 2.5416930758253317e-06,
+      "loss": 0.8256,
+      "step": 4310
+    },
+    {
+      "epoch": 0.6981071024106131,
+      "grad_norm": 0.9992022955905648,
+      "learning_rate": 2.5294067055588765e-06,
+      "loss": 0.8032,
+      "step": 4315
+    },
+    {
+      "epoch": 0.6989160330043682,
+      "grad_norm": 0.8813775901696349,
+      "learning_rate": 2.517140040397482e-06,
+      "loss": 0.8168,
+      "step": 4320
+    },
+    {
+      "epoch": 0.6997249635981233,
+      "grad_norm": 0.9034982645101618,
+      "learning_rate": 2.5048931781783457e-06,
+      "loss": 0.8363,
+      "step": 4325
+    },
+    {
+      "epoch": 0.7005338941918783,
+      "grad_norm": 0.9291076912208419,
+      "learning_rate": 2.492666216580719e-06,
+      "loss": 0.8309,
+      "step": 4330
+    },
+    {
+      "epoch": 0.7013428247856334,
+      "grad_norm": 0.9136441839215993,
+      "learning_rate": 2.480459253125132e-06,
+      "loss": 0.8446,
+      "step": 4335
+    },
+    {
+      "epoch": 0.7021517553793885,
+      "grad_norm": 0.8945013641006528,
+      "learning_rate": 2.468272385172609e-06,
+      "loss": 0.8061,
+      "step": 4340
+    },
+    {
+      "epoch": 0.7029606859731435,
+      "grad_norm": 0.9843334357090001,
+      "learning_rate": 2.4561057099238973e-06,
+      "loss": 0.8245,
+      "step": 4345
+    },
+    {
+      "epoch": 0.7037696165668985,
+      "grad_norm": 0.9061201723841746,
+      "learning_rate": 2.4439593244186914e-06,
+      "loss": 0.8156,
+      "step": 4350
+    },
+    {
+      "epoch": 0.7045785471606536,
+      "grad_norm": 0.9595180789878346,
+      "learning_rate": 2.4318333255348524e-06,
+      "loss": 0.8362,
+      "step": 4355
+    },
+    {
+      "epoch": 0.7053874777544087,
+      "grad_norm": 0.8905904032697742,
+      "learning_rate": 2.4197278099876458e-06,
+      "loss": 0.8013,
+      "step": 4360
+    },
+    {
+      "epoch": 0.7061964083481638,
+      "grad_norm": 0.9061698973868614,
+      "learning_rate": 2.407642874328961e-06,
+      "loss": 0.805,
+      "step": 4365
+    },
+    {
+      "epoch": 0.7070053389419187,
+      "grad_norm": 0.9192998017239877,
+      "learning_rate": 2.3955786149465505e-06,
+      "loss": 0.8445,
+      "step": 4370
+    },
+    {
+      "epoch": 0.7078142695356738,
+      "grad_norm": 1.0236876517448235,
+      "learning_rate": 2.3835351280632514e-06,
+      "loss": 0.8484,
+      "step": 4375
+    },
+    {
+      "epoch": 0.7086232001294289,
+      "grad_norm": 0.9092749965225744,
+      "learning_rate": 2.3715125097362246e-06,
+      "loss": 0.8134,
+      "step": 4380
+    },
+    {
+      "epoch": 0.7094321307231839,
+      "grad_norm": 0.8913146454343571,
+      "learning_rate": 2.3595108558561814e-06,
+      "loss": 0.8231,
+      "step": 4385
+    },
+    {
+      "epoch": 0.710241061316939,
+      "grad_norm": 0.9163485544067049,
+      "learning_rate": 2.347530262146629e-06,
+      "loss": 0.8227,
+      "step": 4390
+    },
+    {
+      "epoch": 0.7110499919106941,
+      "grad_norm": 0.9692923072785117,
+      "learning_rate": 2.3355708241630998e-06,
+      "loss": 0.8114,
+      "step": 4395
+    },
+    {
+      "epoch": 0.7118589225044492,
+      "grad_norm": 0.9952307777966306,
+      "learning_rate": 2.3236326372923913e-06,
+      "loss": 0.8265,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7126678530982041,
+      "grad_norm": 0.8908142322793587,
+      "learning_rate": 2.3117157967518052e-06,
+      "loss": 0.8306,
+      "step": 4405
+    },
+    {
+      "epoch": 0.7134767836919592,
+      "grad_norm": 0.9192836496709301,
+      "learning_rate": 2.299820397588387e-06,
+      "loss": 0.8104,
+      "step": 4410
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 0.90827967789888,
+      "learning_rate": 2.2879465346781703e-06,
+      "loss": 0.8106,
+      "step": 4415
+    },
+    {
+      "epoch": 0.7150946448794694,
+      "grad_norm": 0.9019495653500154,
+      "learning_rate": 2.2760943027254168e-06,
+      "loss": 0.8162,
+      "step": 4420
+    },
+    {
+      "epoch": 0.7159035754732244,
+      "grad_norm": 0.9225860245555908,
+      "learning_rate": 2.264263796261864e-06,
+      "loss": 0.8385,
+      "step": 4425
+    },
+    {
+      "epoch": 0.7167125060669794,
+      "grad_norm": 0.9905684037464217,
+      "learning_rate": 2.2524551096459703e-06,
+      "loss": 0.836,
+      "step": 4430
+    },
+    {
+      "epoch": 0.7175214366607345,
+      "grad_norm": 0.9642215008382582,
+      "learning_rate": 2.240668337062162e-06,
+      "loss": 0.8382,
+      "step": 4435
+    },
+    {
+      "epoch": 0.7183303672544895,
+      "grad_norm": 1.02786094199622,
+      "learning_rate": 2.228903572520076e-06,
+      "loss": 0.8106,
+      "step": 4440
+    },
+    {
+      "epoch": 0.7191392978482446,
+      "grad_norm": 0.885661361219246,
+      "learning_rate": 2.2171609098538275e-06,
+      "loss": 0.8363,
+      "step": 4445
+    },
+    {
+      "epoch": 0.7199482284419997,
+      "grad_norm": 0.9432094404967053,
+      "learning_rate": 2.2054404427212427e-06,
+      "loss": 0.8295,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7207571590357548,
+      "grad_norm": 0.8867439139171059,
+      "learning_rate": 2.1937422646031216e-06,
+      "loss": 0.826,
+      "step": 4455
+    },
+    {
+      "epoch": 0.7215660896295097,
+      "grad_norm": 0.8799592291488351,
+      "learning_rate": 2.1820664688024885e-06,
+      "loss": 0.808,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7223750202232648,
+      "grad_norm": 0.9187111254961942,
+      "learning_rate": 2.1704131484438523e-06,
+      "loss": 0.8284,
+      "step": 4465
+    },
+    {
+      "epoch": 0.7231839508170199,
+      "grad_norm": 0.8585062008980405,
+      "learning_rate": 2.1587823964724564e-06,
+      "loss": 0.8333,
+      "step": 4470
+    },
+    {
+      "epoch": 0.723992881410775,
+      "grad_norm": 0.8979409403030056,
+      "learning_rate": 2.1471743056535455e-06,
+      "loss": 0.8401,
+      "step": 4475
+    },
+    {
+      "epoch": 0.72480181200453,
+      "grad_norm": 0.9181507434036461,
+      "learning_rate": 2.1355889685716225e-06,
+      "loss": 0.8209,
+      "step": 4480
+    },
+    {
+      "epoch": 0.7256107425982851,
+      "grad_norm": 0.942590732475379,
+      "learning_rate": 2.124026477629706e-06,
+      "loss": 0.8066,
+      "step": 4485
+    },
+    {
+      "epoch": 0.7264196731920401,
+      "grad_norm": 0.9429750289429594,
+      "learning_rate": 2.1124869250486053e-06,
+      "loss": 0.8224,
+      "step": 4490
+    },
+    {
+      "epoch": 0.7272286037857951,
+      "grad_norm": 0.9542611725247429,
+      "learning_rate": 2.1009704028661643e-06,
+      "loss": 0.8384,
+      "step": 4495
+    },
+    {
+      "epoch": 0.7280375343795502,
+      "grad_norm": 0.9040683126789256,
+      "learning_rate": 2.08947700293655e-06,
+      "loss": 0.8418,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7288464649733053,
+      "grad_norm": 0.9481483200578114,
+      "learning_rate": 2.078006816929503e-06,
+      "loss": 0.8144,
+      "step": 4505
+    },
+    {
+      "epoch": 0.7296553955670604,
+      "grad_norm": 0.8940640126195414,
+      "learning_rate": 2.066559936329618e-06,
+      "loss": 0.8059,
+      "step": 4510
+    },
+    {
+      "epoch": 0.7304643261608154,
+      "grad_norm": 0.9261592631490639,
+      "learning_rate": 2.0551364524356054e-06,
+      "loss": 0.8261,
+      "step": 4515
+    },
+    {
+      "epoch": 0.7312732567545704,
+      "grad_norm": 0.896600436452953,
+      "learning_rate": 2.0437364563595686e-06,
+      "loss": 0.8437,
+      "step": 4520
+    },
+    {
+      "epoch": 0.7320821873483255,
+      "grad_norm": 0.8919258636639074,
+      "learning_rate": 2.0323600390262743e-06,
+      "loss": 0.8226,
+      "step": 4525
+    },
+    {
+      "epoch": 0.7328911179420806,
+      "grad_norm": 0.9826507277779656,
+      "learning_rate": 2.0210072911724293e-06,
+      "loss": 0.8215,
+      "step": 4530
+    },
+    {
+      "epoch": 0.7337000485358356,
+      "grad_norm": 0.9182603432302395,
+      "learning_rate": 2.0096783033459564e-06,
+      "loss": 0.804,
+      "step": 4535
+    },
+    {
+      "epoch": 0.7345089791295907,
+      "grad_norm": 0.904398523841377,
+      "learning_rate": 1.99837316590527e-06,
+      "loss": 0.8385,
+      "step": 4540
+    },
+    {
+      "epoch": 0.7353179097233458,
+      "grad_norm": 0.8989331803375635,
+      "learning_rate": 1.987091969018561e-06,
+      "loss": 0.8225,
+      "step": 4545
+    },
+    {
+      "epoch": 0.7361268403171007,
+      "grad_norm": 0.8975780407900149,
+      "learning_rate": 1.9758348026630663e-06,
+      "loss": 0.8388,
+      "step": 4550
+    },
+    {
+      "epoch": 0.7369357709108558,
+      "grad_norm": 0.9121042226321563,
+      "learning_rate": 1.964601756624366e-06,
+      "loss": 0.8215,
+      "step": 4555
+    },
+    {
+      "epoch": 0.7377447015046109,
+      "grad_norm": 0.9214704840321436,
+      "learning_rate": 1.9533929204956553e-06,
+      "loss": 0.8142,
+      "step": 4560
+    },
+    {
+      "epoch": 0.738553632098366,
+      "grad_norm": 0.9189755165680765,
+      "learning_rate": 1.9422083836770405e-06,
+      "loss": 0.8307,
+      "step": 4565
+    },
+    {
+      "epoch": 0.739362562692121,
+      "grad_norm": 0.8679073250223568,
+      "learning_rate": 1.9310482353748146e-06,
+      "loss": 0.8118,
+      "step": 4570
+    },
+    {
+      "epoch": 0.7401714932858761,
+      "grad_norm": 0.9016222202190433,
+      "learning_rate": 1.9199125646007533e-06,
+      "loss": 0.8043,
+      "step": 4575
+    },
+    {
+      "epoch": 0.7409804238796311,
+      "grad_norm": 0.9105339155753437,
+      "learning_rate": 1.9088014601713993e-06,
+      "loss": 0.8376,
+      "step": 4580
+    },
+    {
+      "epoch": 0.7417893544733862,
+      "grad_norm": 1.005707657930975,
+      "learning_rate": 1.8977150107073632e-06,
+      "loss": 0.8219,
+      "step": 4585
+    },
+    {
+      "epoch": 0.7425982850671412,
+      "grad_norm": 0.8230187430989078,
+      "learning_rate": 1.8866533046326086e-06,
+      "loss": 0.839,
+      "step": 4590
+    },
+    {
+      "epoch": 0.7434072156608963,
+      "grad_norm": 0.9054157125777134,
+      "learning_rate": 1.8756164301737478e-06,
+      "loss": 0.8215,
+      "step": 4595
+    },
+    {
+      "epoch": 0.7442161462546514,
+      "grad_norm": 0.9386133499303263,
+      "learning_rate": 1.8646044753593429e-06,
+      "loss": 0.8139,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7450250768484064,
+      "grad_norm": 0.8974275889331189,
+      "learning_rate": 1.8536175280191971e-06,
+      "loss": 0.8269,
+      "step": 4605
+    },
+    {
+      "epoch": 0.7458340074421614,
+      "grad_norm": 0.9291872406905795,
+      "learning_rate": 1.8426556757836594e-06,
+      "loss": 0.8104,
+      "step": 4610
+    },
+    {
+      "epoch": 0.7466429380359165,
+      "grad_norm": 0.921360125713896,
+      "learning_rate": 1.8317190060829242e-06,
+      "loss": 0.8078,
+      "step": 4615
+    },
+    {
+      "epoch": 0.7474518686296716,
+      "grad_norm": 0.9154624659332359,
+      "learning_rate": 1.820807606146332e-06,
+      "loss": 0.8228,
+      "step": 4620
+    },
+    {
+      "epoch": 0.7482607992234266,
+      "grad_norm": 0.9910039258360209,
+      "learning_rate": 1.8099215630016759e-06,
+      "loss": 0.8162,
+      "step": 4625
+    },
+    {
+      "epoch": 0.7490697298171817,
+      "grad_norm": 0.9527574985037581,
+      "learning_rate": 1.7990609634745087e-06,
+      "loss": 0.8206,
+      "step": 4630
+    },
+    {
+      "epoch": 0.7498786604109368,
+      "grad_norm": 0.8617520277776372,
+      "learning_rate": 1.788225894187443e-06,
+      "loss": 0.82,
+      "step": 4635
+    },
+    {
+      "epoch": 0.7506875910046918,
+      "grad_norm": 1.0275065944999713,
+      "learning_rate": 1.7774164415594692e-06,
+      "loss": 0.8474,
+      "step": 4640
+    },
+    {
+      "epoch": 0.7514965215984468,
+      "grad_norm": 1.0046608971251627,
+      "learning_rate": 1.7666326918052667e-06,
+      "loss": 0.8144,
+      "step": 4645
+    },
+    {
+      "epoch": 0.7523054521922019,
+      "grad_norm": 0.9234744279179873,
+      "learning_rate": 1.7558747309345075e-06,
+      "loss": 0.8258,
+      "step": 4650
+    },
+    {
+      "epoch": 0.753114382785957,
+      "grad_norm": 0.987941691115384,
+      "learning_rate": 1.7451426447511771e-06,
+      "loss": 0.8161,
+      "step": 4655
+    },
+    {
+      "epoch": 0.753923313379712,
+      "grad_norm": 1.0005046357018823,
+      "learning_rate": 1.7344365188528844e-06,
+      "loss": 0.817,
+      "step": 4660
+    },
+    {
+      "epoch": 0.754732243973467,
+      "grad_norm": 0.9561730998580897,
+      "learning_rate": 1.7237564386301869e-06,
+      "loss": 0.8088,
+      "step": 4665
+    },
+    {
+      "epoch": 0.7555411745672221,
+      "grad_norm": 0.9447102620547789,
+      "learning_rate": 1.7131024892659048e-06,
+      "loss": 0.8118,
+      "step": 4670
+    },
+    {
+      "epoch": 0.7563501051609772,
+      "grad_norm": 0.9539633697057311,
+      "learning_rate": 1.7024747557344411e-06,
+      "loss": 0.844,
+      "step": 4675
+    },
+    {
+      "epoch": 0.7571590357547322,
+      "grad_norm": 1.0424885395489862,
+      "learning_rate": 1.6918733228011058e-06,
+      "loss": 0.8315,
+      "step": 4680
+    },
+    {
+      "epoch": 0.7579679663484873,
+      "grad_norm": 0.9458202947476442,
+      "learning_rate": 1.6812982750214385e-06,
+      "loss": 0.8304,
+      "step": 4685
+    },
+    {
+      "epoch": 0.7587768969422424,
+      "grad_norm": 0.918182742402279,
+      "learning_rate": 1.6707496967405347e-06,
+      "loss": 0.8115,
+      "step": 4690
+    },
+    {
+      "epoch": 0.7595858275359975,
+      "grad_norm": 0.9291030646393732,
+      "learning_rate": 1.6602276720923732e-06,
+      "loss": 0.8163,
+      "step": 4695
+    },
+    {
+      "epoch": 0.7603947581297524,
+      "grad_norm": 0.9588497352744614,
+      "learning_rate": 1.6497322849991443e-06,
+      "loss": 0.8266,
+      "step": 4700
+    },
+    {
+      "epoch": 0.7612036887235075,
+      "grad_norm": 0.9290169660288218,
+      "learning_rate": 1.6392636191705818e-06,
+      "loss": 0.8231,
+      "step": 4705
+    },
+    {
+      "epoch": 0.7620126193172626,
+      "grad_norm": 0.9469265731424403,
+      "learning_rate": 1.6288217581032945e-06,
+      "loss": 0.8283,
+      "step": 4710
+    },
+    {
+      "epoch": 0.7628215499110176,
+      "grad_norm": 0.9739835381552143,
+      "learning_rate": 1.618406785080095e-06,
+      "loss": 0.8358,
+      "step": 4715
+    },
+    {
+      "epoch": 0.7636304805047727,
+      "grad_norm": 0.9140268328463288,
+      "learning_rate": 1.60801878316935e-06,
+      "loss": 0.8173,
+      "step": 4720
+    },
+    {
+      "epoch": 0.7644394110985278,
+      "grad_norm": 0.9519079424417303,
+      "learning_rate": 1.5976578352243015e-06,
+      "loss": 0.8013,
+      "step": 4725
+    },
+    {
+      "epoch": 0.7652483416922828,
+      "grad_norm": 1.0216658413405784,
+      "learning_rate": 1.587324023882415e-06,
+      "loss": 0.827,
+      "step": 4730
+    },
+    {
+      "epoch": 0.7660572722860378,
+      "grad_norm": 0.981342150260489,
+      "learning_rate": 1.5770174315647185e-06,
+      "loss": 0.8172,
+      "step": 4735
+    },
+    {
+      "epoch": 0.7668662028797929,
+      "grad_norm": 0.9841540010061296,
+      "learning_rate": 1.566738140475146e-06,
+      "loss": 0.8298,
+      "step": 4740
+    },
+    {
+      "epoch": 0.767675133473548,
+      "grad_norm": 0.9647924689522289,
+      "learning_rate": 1.5564862325998754e-06,
+      "loss": 0.8349,
+      "step": 4745
+    },
+    {
+      "epoch": 0.7684840640673031,
+      "grad_norm": 0.9071693118539406,
+      "learning_rate": 1.5462617897066863e-06,
+      "loss": 0.8266,
+      "step": 4750
+    },
+    {
+      "epoch": 0.769292994661058,
+      "grad_norm": 0.9672024055908947,
+      "learning_rate": 1.536064893344298e-06,
+      "loss": 0.8062,
+      "step": 4755
+    },
+    {
+      "epoch": 0.7701019252548131,
+      "grad_norm": 0.9479852003456001,
+      "learning_rate": 1.5258956248417228e-06,
+      "loss": 0.8171,
+      "step": 4760
+    },
+    {
+      "epoch": 0.7709108558485682,
+      "grad_norm": 1.0847747231670404,
+      "learning_rate": 1.515754065307622e-06,
+      "loss": 0.8343,
+      "step": 4765
+    },
+    {
+      "epoch": 0.7717197864423232,
+      "grad_norm": 0.8930428867327512,
+      "learning_rate": 1.505640295629645e-06,
+      "loss": 0.8166,
+      "step": 4770
+    },
+    {
+      "epoch": 0.7725287170360783,
+      "grad_norm": 0.9689746816489254,
+      "learning_rate": 1.4955543964738e-06,
+      "loss": 0.8357,
+      "step": 4775
+    },
+    {
+      "epoch": 0.7733376476298334,
+      "grad_norm": 0.9559994524411303,
+      "learning_rate": 1.4854964482838014e-06,
+      "loss": 0.8497,
+      "step": 4780
+    },
+    {
+      "epoch": 0.7741465782235885,
+      "grad_norm": 0.9337905415707888,
+      "learning_rate": 1.4754665312804311e-06,
+      "loss": 0.8312,
+      "step": 4785
+    },
+    {
+      "epoch": 0.7749555088173434,
+      "grad_norm": 0.9273467544337363,
+      "learning_rate": 1.4654647254608988e-06,
+      "loss": 0.8234,
+      "step": 4790
+    },
+    {
+      "epoch": 0.7757644394110985,
+      "grad_norm": 0.9792032991670162,
+      "learning_rate": 1.4554911105982022e-06,
+      "loss": 0.8468,
+      "step": 4795
+    },
+    {
+      "epoch": 0.7765733700048536,
+      "grad_norm": 0.9045240623424784,
+      "learning_rate": 1.445545766240492e-06,
+      "loss": 0.8315,
+      "step": 4800
+    },
+    {
+      "epoch": 0.7773823005986087,
+      "grad_norm": 0.9547782673298132,
+      "learning_rate": 1.4356287717104384e-06,
+      "loss": 0.8259,
+      "step": 4805
+    },
+    {
+      "epoch": 0.7781912311923637,
+      "grad_norm": 1.0002987188424177,
+      "learning_rate": 1.4257402061045966e-06,
+      "loss": 0.8242,
+      "step": 4810
+    },
+    {
+      "epoch": 0.7790001617861187,
+      "grad_norm": 0.9750841170635594,
+      "learning_rate": 1.4158801482927764e-06,
+      "loss": 0.8253,
+      "step": 4815
+    },
+    {
+      "epoch": 0.7798090923798738,
+      "grad_norm": 1.0204763580485683,
+      "learning_rate": 1.4060486769174158e-06,
+      "loss": 0.8124,
+      "step": 4820
+    },
+    {
+      "epoch": 0.7806180229736288,
+      "grad_norm": 0.9140777892397043,
+      "learning_rate": 1.396245870392946e-06,
+      "loss": 0.8209,
+      "step": 4825
+    },
+    {
+      "epoch": 0.7814269535673839,
+      "grad_norm": 0.9326381455666519,
+      "learning_rate": 1.3864718069051765e-06,
+      "loss": 0.8085,
+      "step": 4830
+    },
+    {
+      "epoch": 0.782235884161139,
+      "grad_norm": 0.8927418792043738,
+      "learning_rate": 1.376726564410663e-06,
+      "loss": 0.8281,
+      "step": 4835
+    },
+    {
+      "epoch": 0.7830448147548941,
+      "grad_norm": 0.9878798053606094,
+      "learning_rate": 1.3670102206360936e-06,
+      "loss": 0.8393,
+      "step": 4840
+    },
+    {
+      "epoch": 0.783853745348649,
+      "grad_norm": 0.9345336173306545,
+      "learning_rate": 1.3573228530776605e-06,
+      "loss": 0.8044,
+      "step": 4845
+    },
+    {
+      "epoch": 0.7846626759424041,
+      "grad_norm": 0.9721628031873526,
+      "learning_rate": 1.3476645390004473e-06,
+      "loss": 0.8386,
+      "step": 4850
+    },
+    {
+      "epoch": 0.7854716065361592,
+      "grad_norm": 0.915209628149697,
+      "learning_rate": 1.3380353554378074e-06,
+      "loss": 0.8357,
+      "step": 4855
+    },
+    {
+      "epoch": 0.7862805371299143,
+      "grad_norm": 0.9820722069486887,
+      "learning_rate": 1.3284353791907584e-06,
+      "loss": 0.8333,
+      "step": 4860
+    },
+    {
+      "epoch": 0.7870894677236693,
+      "grad_norm": 0.9732881075592806,
+      "learning_rate": 1.3188646868273615e-06,
+      "loss": 0.8239,
+      "step": 4865
+    },
+    {
+      "epoch": 0.7878983983174244,
+      "grad_norm": 0.8961967498766817,
+      "learning_rate": 1.3093233546821143e-06,
+      "loss": 0.824,
+      "step": 4870
+    },
+    {
+      "epoch": 0.7887073289111795,
+      "grad_norm": 0.9052548621214841,
+      "learning_rate": 1.2998114588553429e-06,
+      "loss": 0.8267,
+      "step": 4875
+    },
+    {
+      "epoch": 0.7895162595049344,
+      "grad_norm": 1.0580362890932773,
+      "learning_rate": 1.2903290752125914e-06,
+      "loss": 0.8218,
+      "step": 4880
+    },
+    {
+      "epoch": 0.7903251900986895,
+      "grad_norm": 0.853800216222736,
+      "learning_rate": 1.28087627938402e-06,
+      "loss": 0.8166,
+      "step": 4885
+    },
+    {
+      "epoch": 0.7911341206924446,
+      "grad_norm": 0.9652580037959974,
+      "learning_rate": 1.2714531467637998e-06,
+      "loss": 0.8238,
+      "step": 4890
+    },
+    {
+      "epoch": 0.7919430512861997,
+      "grad_norm": 0.8842437766967849,
+      "learning_rate": 1.2620597525095135e-06,
+      "loss": 0.817,
+      "step": 4895
+    },
+    {
+      "epoch": 0.7927519818799547,
+      "grad_norm": 0.9258752585803977,
+      "learning_rate": 1.2526961715415542e-06,
+      "loss": 0.8279,
+      "step": 4900
+    },
+    {
+      "epoch": 0.7935609124737097,
+      "grad_norm": 0.8928563496158387,
+      "learning_rate": 1.2433624785425291e-06,
+      "loss": 0.8162,
+      "step": 4905
+    },
+    {
+      "epoch": 0.7943698430674648,
+      "grad_norm": 0.982000276972604,
+      "learning_rate": 1.2340587479566597e-06,
+      "loss": 0.8356,
+      "step": 4910
+    },
+    {
+      "epoch": 0.7951787736612199,
+      "grad_norm": 0.9282259377376256,
+      "learning_rate": 1.2247850539891947e-06,
+      "loss": 0.8253,
+      "step": 4915
+    },
+    {
+      "epoch": 0.7959877042549749,
+      "grad_norm": 0.9525465075992243,
+      "learning_rate": 1.2155414706058172e-06,
+      "loss": 0.8169,
+      "step": 4920
+    },
+    {
+      "epoch": 0.79679663484873,
+      "grad_norm": 0.9231085182420589,
+      "learning_rate": 1.206328071532048e-06,
+      "loss": 0.8138,
+      "step": 4925
+    },
+    {
+      "epoch": 0.7976055654424851,
+      "grad_norm": 0.9253319119591827,
+      "learning_rate": 1.197144930252665e-06,
+      "loss": 0.8225,
+      "step": 4930
+    },
+    {
+      "epoch": 0.79841449603624,
+      "grad_norm": 0.9002303313112969,
+      "learning_rate": 1.187992120011111e-06,
+      "loss": 0.8291,
+      "step": 4935
+    },
+    {
+      "epoch": 0.7992234266299951,
+      "grad_norm": 0.9376615616843144,
+      "learning_rate": 1.178869713808916e-06,
+      "loss": 0.8341,
+      "step": 4940
+    },
+    {
+      "epoch": 0.8000323572237502,
+      "grad_norm": 0.9698576047423542,
+      "learning_rate": 1.1697777844051105e-06,
+      "loss": 0.8147,
+      "step": 4945
+    },
+    {
+      "epoch": 0.8008412878175053,
+      "grad_norm": 0.9273363034177441,
+      "learning_rate": 1.1607164043156455e-06,
+      "loss": 0.8152,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8016502184112603,
+      "grad_norm": 1.0253181360844876,
+      "learning_rate": 1.1516856458128167e-06,
+      "loss": 0.8084,
+      "step": 4955
+    },
+    {
+      "epoch": 0.8024591490050154,
+      "grad_norm": 0.9400581889744581,
+      "learning_rate": 1.1426855809246846e-06,
+      "loss": 0.8232,
+      "step": 4960
+    },
+    {
+      "epoch": 0.8032680795987704,
+      "grad_norm": 0.9017439278181727,
+      "learning_rate": 1.133716281434502e-06,
+      "loss": 0.8212,
+      "step": 4965
+    },
+    {
+      "epoch": 0.8040770101925255,
+      "grad_norm": 0.9129823712372367,
+      "learning_rate": 1.1247778188801428e-06,
+      "loss": 0.8214,
+      "step": 4970
+    },
+    {
+      "epoch": 0.8048859407862805,
+      "grad_norm": 0.9288433560645074,
+      "learning_rate": 1.1158702645535285e-06,
+      "loss": 0.8474,
+      "step": 4975
+    },
+    {
+      "epoch": 0.8056948713800356,
+      "grad_norm": 0.9211800178241987,
+      "learning_rate": 1.1069936895000605e-06,
+      "loss": 0.8087,
+      "step": 4980
+    },
+    {
+      "epoch": 0.8065038019737907,
+      "grad_norm": 1.0542799666763112,
+      "learning_rate": 1.0981481645180563e-06,
+      "loss": 0.842,
+      "step": 4985
+    },
+    {
+      "epoch": 0.8073127325675457,
+      "grad_norm": 0.9078222972795535,
+      "learning_rate": 1.0893337601581766e-06,
+      "loss": 0.8171,
+      "step": 4990
+    },
+    {
+      "epoch": 0.8081216631613007,
+      "grad_norm": 0.8965670623934641,
+      "learning_rate": 1.0805505467228762e-06,
+      "loss": 0.8106,
+      "step": 4995
+    },
+    {
+      "epoch": 0.8089305937550558,
+      "grad_norm": 0.9618251738798234,
+      "learning_rate": 1.0717985942658299e-06,
+      "loss": 0.8146,
+      "step": 5000
+    },
+    {
+      "epoch": 0.8097395243488109,
+      "grad_norm": 0.950502760029193,
+      "learning_rate": 1.063077972591382e-06,
+      "loss": 0.8215,
+      "step": 5005
+    },
+    {
+      "epoch": 0.8105484549425659,
+      "grad_norm": 0.9493168241113547,
+      "learning_rate": 1.0543887512539851e-06,
+      "loss": 0.8194,
+      "step": 5010
+    },
+    {
+      "epoch": 0.811357385536321,
+      "grad_norm": 0.8823063979936068,
+      "learning_rate": 1.0457309995576498e-06,
+      "loss": 0.8081,
+      "step": 5015
+    },
+    {
+      "epoch": 0.8121663161300761,
+      "grad_norm": 0.9596843704171958,
+      "learning_rate": 1.0371047865553847e-06,
+      "loss": 0.7973,
+      "step": 5020
+    },
+    {
+      "epoch": 0.8129752467238311,
+      "grad_norm": 0.9284122972047167,
+      "learning_rate": 1.0285101810486535e-06,
+      "loss": 0.8468,
+      "step": 5025
+    },
+    {
+      "epoch": 0.8137841773175861,
+      "grad_norm": 0.922908519391859,
+      "learning_rate": 1.0199472515868235e-06,
+      "loss": 0.8602,
+      "step": 5030
+    },
+    {
+      "epoch": 0.8145931079113412,
+      "grad_norm": 0.9150347591949571,
+      "learning_rate": 1.0114160664666156e-06,
+      "loss": 0.8285,
+      "step": 5035
+    },
+    {
+      "epoch": 0.8154020385050963,
+      "grad_norm": 0.961458868733065,
+      "learning_rate": 1.0029166937315682e-06,
+      "loss": 0.8261,
+      "step": 5040
+    },
+    {
+      "epoch": 0.8162109690988513,
+      "grad_norm": 0.9795182791817205,
+      "learning_rate": 9.94449201171479e-07,
+      "loss": 0.8077,
+      "step": 5045
+    },
+    {
+      "epoch": 0.8170198996926064,
+      "grad_norm": 0.9712437822816266,
+      "learning_rate": 9.86013656321882e-07,
+      "loss": 0.8136,
+      "step": 5050
+    },
+    {
+      "epoch": 0.8178288302863614,
+      "grad_norm": 0.9243414819653508,
+      "learning_rate": 9.77610126463497e-07,
+      "loss": 0.8116,
+      "step": 5055
+    },
+    {
+      "epoch": 0.8186377608801165,
+      "grad_norm": 0.8953653000547861,
+      "learning_rate": 9.69238678621698e-07,
+      "loss": 0.8052,
+      "step": 5060
+    },
+    {
+      "epoch": 0.8194466914738715,
+      "grad_norm": 0.9571083425955552,
+      "learning_rate": 9.608993795659766e-07,
+      "loss": 0.8361,
+      "step": 5065
+    },
+    {
+      "epoch": 0.8202556220676266,
+      "grad_norm": 0.8812425656788687,
+      "learning_rate": 9.525922958094113e-07,
+      "loss": 0.8091,
+      "step": 5070
+    },
+    {
+      "epoch": 0.8210645526613817,
+      "grad_norm": 0.9769325986983122,
+      "learning_rate": 9.443174936081345e-07,
+      "loss": 0.8384,
+      "step": 5075
+    },
+    {
+      "epoch": 0.8218734832551368,
+      "grad_norm": 0.9429765500848418,
+      "learning_rate": 9.360750389608069e-07,
+      "loss": 0.823,
+      "step": 5080
+    },
+    {
+      "epoch": 0.8226824138488917,
+      "grad_norm": 0.921228115104811,
+      "learning_rate": 9.278649976080889e-07,
+      "loss": 0.8331,
+      "step": 5085
+    },
+    {
+      "epoch": 0.8234913444426468,
+      "grad_norm": 0.9819735566785222,
+      "learning_rate": 9.196874350321161e-07,
+      "loss": 0.829,
+      "step": 5090
+    },
+    {
+      "epoch": 0.8243002750364019,
+      "grad_norm": 0.8804316419490247,
+      "learning_rate": 9.11542416455981e-07,
+      "loss": 0.8253,
+      "step": 5095
+    },
+    {
+      "epoch": 0.8251092056301569,
+      "grad_norm": 0.9309408261176141,
+      "learning_rate": 9.034300068432045e-07,
+      "loss": 0.8294,
+      "step": 5100
+    },
+    {
+      "epoch": 0.825918136223912,
+      "grad_norm": 0.8366500179121683,
+      "learning_rate": 8.953502708972279e-07,
+      "loss": 0.8309,
+      "step": 5105
+    },
+    {
+      "epoch": 0.8267270668176671,
+      "grad_norm": 0.9984582201604612,
+      "learning_rate": 8.873032730608883e-07,
+      "loss": 0.8288,
+      "step": 5110
+    },
+    {
+      "epoch": 0.8275359974114221,
+      "grad_norm": 1.0257992816612576,
+      "learning_rate": 8.792890775159124e-07,
+      "loss": 0.832,
+      "step": 5115
+    },
+    {
+      "epoch": 0.8283449280051771,
+      "grad_norm": 0.9357127400993555,
+      "learning_rate": 8.713077481823978e-07,
+      "loss": 0.8377,
+      "step": 5120
+    },
+    {
+      "epoch": 0.8291538585989322,
+      "grad_norm": 1.042951179385166,
+      "learning_rate": 8.633593487183067e-07,
+      "loss": 0.8171,
+      "step": 5125
+    },
+    {
+      "epoch": 0.8299627891926873,
+      "grad_norm": 1.0012604552671969,
+      "learning_rate": 8.554439425189537e-07,
+      "loss": 0.8017,
+      "step": 5130
+    },
+    {
+      "epoch": 0.8307717197864424,
+      "grad_norm": 0.9393560978003819,
+      "learning_rate": 8.475615927165093e-07,
+      "loss": 0.8208,
+      "step": 5135
+    },
+    {
+      "epoch": 0.8315806503801974,
+      "grad_norm": 0.9829342370330283,
+      "learning_rate": 8.397123621794867e-07,
+      "loss": 0.8213,
+      "step": 5140
+    },
+    {
+      "epoch": 0.8323895809739524,
+      "grad_norm": 0.993397225397711,
+      "learning_rate": 8.318963135122471e-07,
+      "loss": 0.8386,
+      "step": 5145
+    },
+    {
+      "epoch": 0.8331985115677075,
+      "grad_norm": 0.9203262711299934,
+      "learning_rate": 8.241135090544966e-07,
+      "loss": 0.8135,
+      "step": 5150
+    },
+    {
+      "epoch": 0.8340074421614625,
+      "grad_norm": 0.9284177490644239,
+      "learning_rate": 8.163640108807897e-07,
+      "loss": 0.834,
+      "step": 5155
+    },
+    {
+      "epoch": 0.8348163727552176,
+      "grad_norm": 0.9517537835456206,
+      "learning_rate": 8.086478808000359e-07,
+      "loss": 0.7949,
+      "step": 5160
+    },
+    {
+      "epoch": 0.8356253033489727,
+      "grad_norm": 0.9372135921171859,
+      "learning_rate": 8.009651803550045e-07,
+      "loss": 0.8289,
+      "step": 5165
+    },
+    {
+      "epoch": 0.8364342339427278,
+      "grad_norm": 1.000548748366353,
+      "learning_rate": 7.933159708218347e-07,
+      "loss": 0.8268,
+      "step": 5170
+    },
+    {
+      "epoch": 0.8372431645364827,
+      "grad_norm": 0.9612535821852152,
+      "learning_rate": 7.857003132095481e-07,
+      "loss": 0.8182,
+      "step": 5175
+    },
+    {
+      "epoch": 0.8380520951302378,
+      "grad_norm": 0.9551451813093564,
+      "learning_rate": 7.7811826825956e-07,
+      "loss": 0.8399,
+      "step": 5180
+    },
+    {
+      "epoch": 0.8388610257239929,
+      "grad_norm": 0.9803361111190585,
+      "learning_rate": 7.705698964451941e-07,
+      "loss": 0.8345,
+      "step": 5185
+    },
+    {
+      "epoch": 0.839669956317748,
+      "grad_norm": 0.9615430440473038,
+      "learning_rate": 7.630552579712041e-07,
+      "loss": 0.8482,
+      "step": 5190
+    },
+    {
+      "epoch": 0.840478886911503,
+      "grad_norm": 0.8717103769599843,
+      "learning_rate": 7.555744127732922e-07,
+      "loss": 0.8133,
+      "step": 5195
+    },
+    {
+      "epoch": 0.841287817505258,
+      "grad_norm": 0.9559551946507219,
+      "learning_rate": 7.481274205176286e-07,
+      "loss": 0.8396,
+      "step": 5200
+    },
+    {
+      "epoch": 0.8420967480990131,
+      "grad_norm": 0.9386431006557889,
+      "learning_rate": 7.407143406003781e-07,
+      "loss": 0.8166,
+      "step": 5205
+    },
+    {
+      "epoch": 0.8429056786927681,
+      "grad_norm": 0.9453424188819014,
+      "learning_rate": 7.33335232147222e-07,
+      "loss": 0.8318,
+      "step": 5210
+    },
+    {
+      "epoch": 0.8437146092865232,
+      "grad_norm": 0.9335912363260552,
+      "learning_rate": 7.25990154012895e-07,
+      "loss": 0.8043,
+      "step": 5215
+    },
+    {
+      "epoch": 0.8445235398802783,
+      "grad_norm": 0.9242358250100629,
+      "learning_rate": 7.186791647807078e-07,
+      "loss": 0.8216,
+      "step": 5220
+    },
+    {
+      "epoch": 0.8453324704740334,
+      "grad_norm": 0.8917056191120556,
+      "learning_rate": 7.114023227620831e-07,
+      "loss": 0.8263,
+      "step": 5225
+    },
+    {
+      "epoch": 0.8461414010677883,
+      "grad_norm": 0.8756042557078504,
+      "learning_rate": 7.04159685996092e-07,
+      "loss": 0.8234,
+      "step": 5230
+    },
+    {
+      "epoch": 0.8469503316615434,
+      "grad_norm": 0.9669924510507611,
+      "learning_rate": 6.969513122489862e-07,
+      "loss": 0.8317,
+      "step": 5235
+    },
+    {
+      "epoch": 0.8477592622552985,
+      "grad_norm": 0.9690987783521561,
+      "learning_rate": 6.897772590137436e-07,
+      "loss": 0.8196,
+      "step": 5240
+    },
+    {
+      "epoch": 0.8485681928490536,
+      "grad_norm": 1.032895557175751,
+      "learning_rate": 6.826375835096038e-07,
+      "loss": 0.8017,
+      "step": 5245
+    },
+    {
+      "epoch": 0.8493771234428086,
+      "grad_norm": 1.0187566680457716,
+      "learning_rate": 6.755323426816169e-07,
+      "loss": 0.8354,
+      "step": 5250
+    },
+    {
+      "epoch": 0.8501860540365637,
+      "grad_norm": 0.8816437739999422,
+      "learning_rate": 6.684615932001847e-07,
+      "loss": 0.8581,
+      "step": 5255
+    },
+    {
+      "epoch": 0.8509949846303188,
+      "grad_norm": 0.9088385041537731,
+      "learning_rate": 6.614253914606134e-07,
+      "loss": 0.8036,
+      "step": 5260
+    },
+    {
+      "epoch": 0.8518039152240737,
+      "grad_norm": 0.9703072936941138,
+      "learning_rate": 6.544237935826558e-07,
+      "loss": 0.8277,
+      "step": 5265
+    },
+    {
+      "epoch": 0.8526128458178288,
+      "grad_norm": 0.9508642384505709,
+      "learning_rate": 6.474568554100768e-07,
+      "loss": 0.8243,
+      "step": 5270
+    },
+    {
+      "epoch": 0.8534217764115839,
+      "grad_norm": 0.9433685567993545,
+      "learning_rate": 6.405246325101955e-07,
+      "loss": 0.8056,
+      "step": 5275
+    },
+    {
+      "epoch": 0.854230707005339,
+      "grad_norm": 0.8978485349230313,
+      "learning_rate": 6.336271801734479e-07,
+      "loss": 0.8016,
+      "step": 5280
+    },
+    {
+      "epoch": 0.855039637599094,
+      "grad_norm": 0.9591854547149082,
+      "learning_rate": 6.267645534129446e-07,
+      "loss": 0.8184,
+      "step": 5285
+    },
+    {
+      "epoch": 0.855848568192849,
+      "grad_norm": 0.9368695693533301,
+      "learning_rate": 6.199368069640343e-07,
+      "loss": 0.8345,
+      "step": 5290
+    },
+    {
+      "epoch": 0.8566574987866041,
+      "grad_norm": 0.9363387262284574,
+      "learning_rate": 6.131439952838608e-07,
+      "loss": 0.8143,
+      "step": 5295
+    },
+    {
+      "epoch": 0.8574664293803592,
+      "grad_norm": 0.9449827101149645,
+      "learning_rate": 6.063861725509374e-07,
+      "loss": 0.801,
+      "step": 5300
+    },
+    {
+      "epoch": 0.8582753599741142,
+      "grad_norm": 0.9911552846071767,
+      "learning_rate": 5.996633926647083e-07,
+      "loss": 0.8248,
+      "step": 5305
+    },
+    {
+      "epoch": 0.8590842905678693,
+      "grad_norm": 0.9088842745097252,
+      "learning_rate": 5.92975709245121e-07,
+      "loss": 0.8047,
+      "step": 5310
+    },
+    {
+      "epoch": 0.8598932211616244,
+      "grad_norm": 0.8908028970825979,
+      "learning_rate": 5.863231756322019e-07,
+      "loss": 0.8242,
+      "step": 5315
+    },
+    {
+      "epoch": 0.8607021517553793,
+      "grad_norm": 0.9019135577959921,
+      "learning_rate": 5.797058448856213e-07,
+      "loss": 0.8182,
+      "step": 5320
+    },
+    {
+      "epoch": 0.8615110823491344,
+      "grad_norm": 0.9938100484918152,
+      "learning_rate": 5.731237697842811e-07,
+      "loss": 0.8381,
+      "step": 5325
+    },
+    {
+      "epoch": 0.8623200129428895,
+      "grad_norm": 0.9522671534271417,
+      "learning_rate": 5.665770028258876e-07,
+      "loss": 0.8103,
+      "step": 5330
+    },
+    {
+      "epoch": 0.8631289435366446,
+      "grad_norm": 1.0241489828374508,
+      "learning_rate": 5.600655962265345e-07,
+      "loss": 0.8253,
+      "step": 5335
+    },
+    {
+      "epoch": 0.8639378741303996,
+      "grad_norm": 0.8339409337033903,
+      "learning_rate": 5.535896019202853e-07,
+      "loss": 0.8091,
+      "step": 5340
+    },
+    {
+      "epoch": 0.8647468047241547,
+      "grad_norm": 0.953851217822305,
+      "learning_rate": 5.471490715587618e-07,
+      "loss": 0.8277,
+      "step": 5345
+    },
+    {
+      "epoch": 0.8655557353179097,
+      "grad_norm": 0.9253602828275306,
+      "learning_rate": 5.407440565107291e-07,
+      "loss": 0.8352,
+      "step": 5350
+    },
+    {
+      "epoch": 0.8663646659116648,
+      "grad_norm": 0.8979717645863395,
+      "learning_rate": 5.343746078616879e-07,
+      "loss": 0.8346,
+      "step": 5355
+    },
+    {
+      "epoch": 0.8671735965054198,
+      "grad_norm": 0.9052656857430195,
+      "learning_rate": 5.280407764134648e-07,
+      "loss": 0.8098,
+      "step": 5360
+    },
+    {
+      "epoch": 0.8679825270991749,
+      "grad_norm": 0.9265736364407307,
+      "learning_rate": 5.21742612683811e-07,
+      "loss": 0.8134,
+      "step": 5365
+    },
+    {
+      "epoch": 0.86879145769293,
+      "grad_norm": 0.9905238729956866,
+      "learning_rate": 5.154801669059961e-07,
+      "loss": 0.8057,
+      "step": 5370
+    },
+    {
+      "epoch": 0.869600388286685,
+      "grad_norm": 0.9346637770934627,
+      "learning_rate": 5.092534890284057e-07,
+      "loss": 0.8319,
+      "step": 5375
+    },
+    {
+      "epoch": 0.87040931888044,
+      "grad_norm": 0.9036166328056819,
+      "learning_rate": 5.030626287141488e-07,
+      "loss": 0.8242,
+      "step": 5380
+    },
+    {
+      "epoch": 0.8712182494741951,
+      "grad_norm": 0.957351416925992,
+      "learning_rate": 4.969076353406571e-07,
+      "loss": 0.81,
+      "step": 5385
+    },
+    {
+      "epoch": 0.8720271800679502,
+      "grad_norm": 0.9998030765172857,
+      "learning_rate": 4.907885579992943e-07,
+      "loss": 0.7931,
+      "step": 5390
+    },
+    {
+      "epoch": 0.8728361106617052,
+      "grad_norm": 0.9015423928862624,
+      "learning_rate": 4.847054454949617e-07,
+      "loss": 0.8321,
+      "step": 5395
+    },
+    {
+      "epoch": 0.8736450412554603,
+      "grad_norm": 0.8951302299296955,
+      "learning_rate": 4.786583463457106e-07,
+      "loss": 0.8175,
+      "step": 5400
+    },
+    {
+      "epoch": 0.8744539718492154,
+      "grad_norm": 0.987288936023476,
+      "learning_rate": 4.726473087823524e-07,
+      "loss": 0.8149,
+      "step": 5405
+    },
+    {
+      "epoch": 0.8752629024429704,
+      "grad_norm": 0.9607690476849372,
+      "learning_rate": 4.666723807480794e-07,
+      "loss": 0.8395,
+      "step": 5410
+    },
+    {
+      "epoch": 0.8760718330367254,
+      "grad_norm": 0.9412173085678744,
+      "learning_rate": 4.6073360989807805e-07,
+      "loss": 0.8131,
+      "step": 5415
+    },
+    {
+      "epoch": 0.8768807636304805,
+      "grad_norm": 0.9953808945790064,
+      "learning_rate": 4.5483104359915095e-07,
+      "loss": 0.8337,
+      "step": 5420
+    },
+    {
+      "epoch": 0.8776896942242356,
+      "grad_norm": 0.9224438419258796,
+      "learning_rate": 4.4896472892933693e-07,
+      "loss": 0.827,
+      "step": 5425
+    },
+    {
+      "epoch": 0.8784986248179906,
+      "grad_norm": 0.9067448515242419,
+      "learning_rate": 4.431347126775382e-07,
+      "loss": 0.8225,
+      "step": 5430
+    },
+    {
+      "epoch": 0.8793075554117457,
+      "grad_norm": 0.9160793517967267,
+      "learning_rate": 4.3734104134314505e-07,
+      "loss": 0.8093,
+      "step": 5435
+    },
+    {
+      "epoch": 0.8801164860055007,
+      "grad_norm": 0.8957867534440033,
+      "learning_rate": 4.3158376113566656e-07,
+      "loss": 0.8255,
+      "step": 5440
+    },
+    {
+      "epoch": 0.8809254165992558,
+      "grad_norm": 0.9414164683136745,
+      "learning_rate": 4.258629179743612e-07,
+      "loss": 0.8307,
+      "step": 5445
+    },
+    {
+      "epoch": 0.8817343471930108,
+      "grad_norm": 1.0412360371956955,
+      "learning_rate": 4.2017855748786975e-07,
+      "loss": 0.8167,
+      "step": 5450
+    },
+    {
+      "epoch": 0.8825432777867659,
+      "grad_norm": 0.9168749499249985,
+      "learning_rate": 4.1453072501385415e-07,
+      "loss": 0.807,
+      "step": 5455
+    },
+    {
+      "epoch": 0.883352208380521,
+      "grad_norm": 0.885047475917418,
+      "learning_rate": 4.089194655986306e-07,
+      "loss": 0.831,
+      "step": 5460
+    },
+    {
+      "epoch": 0.8841611389742761,
+      "grad_norm": 0.9248545364082547,
+      "learning_rate": 4.0334482399681684e-07,
+      "loss": 0.836,
+      "step": 5465
+    },
+    {
+      "epoch": 0.884970069568031,
+      "grad_norm": 0.9565631081942835,
+      "learning_rate": 3.978068446709721e-07,
+      "loss": 0.8206,
+      "step": 5470
+    },
+    {
+      "epoch": 0.8857790001617861,
+      "grad_norm": 0.9313667515936709,
+      "learning_rate": 3.923055717912411e-07,
+      "loss": 0.8315,
+      "step": 5475
+    },
+    {
+      "epoch": 0.8865879307555412,
+      "grad_norm": 0.9153735916152161,
+      "learning_rate": 3.868410492350044e-07,
+      "loss": 0.8315,
+      "step": 5480
+    },
+    {
+      "epoch": 0.8873968613492962,
+      "grad_norm": 0.8996467266796176,
+      "learning_rate": 3.8141332058652447e-07,
+      "loss": 0.8039,
+      "step": 5485
+    },
+    {
+      "epoch": 0.8882057919430513,
+      "grad_norm": 0.8739192973106119,
+      "learning_rate": 3.7602242913660325e-07,
+      "loss": 0.8196,
+      "step": 5490
+    },
+    {
+      "epoch": 0.8890147225368064,
+      "grad_norm": 1.0836849051236812,
+      "learning_rate": 3.7066841788223394e-07,
+      "loss": 0.8414,
+      "step": 5495
+    },
+    {
+      "epoch": 0.8898236531305614,
+      "grad_norm": 0.8886207081976889,
+      "learning_rate": 3.6535132952625784e-07,
+      "loss": 0.8041,
+      "step": 5500
+    },
+    {
+      "epoch": 0.8906325837243164,
+      "grad_norm": 0.8954848811354755,
+      "learning_rate": 3.6007120647702566e-07,
+      "loss": 0.836,
+      "step": 5505
+    },
+    {
+      "epoch": 0.8914415143180715,
+      "grad_norm": 0.9763361834402516,
+      "learning_rate": 3.548280908480556e-07,
+      "loss": 0.8193,
+      "step": 5510
+    },
+    {
+      "epoch": 0.8922504449118266,
+      "grad_norm": 0.9676090733901674,
+      "learning_rate": 3.496220244577025e-07,
+      "loss": 0.8247,
+      "step": 5515
+    },
+    {
+      "epoch": 0.8930593755055817,
+      "grad_norm": 0.9583454675377773,
+      "learning_rate": 3.4445304882882056e-07,
+      "loss": 0.8232,
+      "step": 5520
+    },
+    {
+      "epoch": 0.8938683060993367,
+      "grad_norm": 0.9149362551378993,
+      "learning_rate": 3.3932120518843315e-07,
+      "loss": 0.8378,
+      "step": 5525
+    },
+    {
+      "epoch": 0.8946772366930917,
+      "grad_norm": 0.9981131323135035,
+      "learning_rate": 3.342265344674034e-07,
+      "loss": 0.8364,
+      "step": 5530
+    },
+    {
+      "epoch": 0.8954861672868468,
+      "grad_norm": 0.9602463248675637,
+      "learning_rate": 3.291690773001116e-07,
+      "loss": 0.8148,
+      "step": 5535
+    },
+    {
+      "epoch": 0.8962950978806018,
+      "grad_norm": 0.9061694557646229,
+      "learning_rate": 3.241488740241222e-07,
+      "loss": 0.8135,
+      "step": 5540
+    },
+    {
+      "epoch": 0.8971040284743569,
+      "grad_norm": 0.9504624920131578,
+      "learning_rate": 3.1916596467987395e-07,
+      "loss": 0.8,
+      "step": 5545
+    },
+    {
+      "epoch": 0.897912959068112,
+      "grad_norm": 1.0083410313503633,
+      "learning_rate": 3.142203890103512e-07,
+      "loss": 0.8194,
+      "step": 5550
+    },
+    {
+      "epoch": 0.8987218896618671,
+      "grad_norm": 0.981599869718165,
+      "learning_rate": 3.093121864607707e-07,
+      "loss": 0.816,
+      "step": 5555
+    },
+    {
+      "epoch": 0.899530820255622,
+      "grad_norm": 0.927185199167883,
+      "learning_rate": 3.0444139617826605e-07,
+      "loss": 0.8133,
+      "step": 5560
+    },
+    {
+      "epoch": 0.9003397508493771,
+      "grad_norm": 1.0064121664416639,
+      "learning_rate": 2.996080570115778e-07,
+      "loss": 0.8262,
+      "step": 5565
+    },
+    {
+      "epoch": 0.9011486814431322,
+      "grad_norm": 0.9554574033151684,
+      "learning_rate": 2.9481220751073793e-07,
+      "loss": 0.8247,
+      "step": 5570
+    },
+    {
+      "epoch": 0.9019576120368872,
+      "grad_norm": 0.9588661920794765,
+      "learning_rate": 2.9005388592676987e-07,
+      "loss": 0.8273,
+      "step": 5575
+    },
+    {
+      "epoch": 0.9027665426306423,
+      "grad_norm": 0.885495966716363,
+      "learning_rate": 2.8533313021137765e-07,
+      "loss": 0.8326,
+      "step": 5580
+    },
+    {
+      "epoch": 0.9035754732243974,
+      "grad_norm": 0.9273868198730849,
+      "learning_rate": 2.806499780166455e-07,
+      "loss": 0.8073,
+      "step": 5585
+    },
+    {
+      "epoch": 0.9043844038181524,
+      "grad_norm": 1.0414377339942757,
+      "learning_rate": 2.760044666947387e-07,
+      "loss": 0.8324,
+      "step": 5590
+    },
+    {
+      "epoch": 0.9051933344119074,
+      "grad_norm": 0.9699497983834742,
+      "learning_rate": 2.7139663329760203e-07,
+      "loss": 0.8222,
+      "step": 5595
+    },
+    {
+      "epoch": 0.9060022650056625,
+      "grad_norm": 0.9269981190661996,
+      "learning_rate": 2.668265145766669e-07,
+      "loss": 0.8298,
+      "step": 5600
+    },
+    {
+      "epoch": 0.9068111955994176,
+      "grad_norm": 1.0012454133607382,
+      "learning_rate": 2.6229414698255907e-07,
+      "loss": 0.8293,
+      "step": 5605
+    },
+    {
+      "epoch": 0.9076201261931727,
+      "grad_norm": 0.9294225278070521,
+      "learning_rate": 2.577995666648053e-07,
+      "loss": 0.8228,
+      "step": 5610
+    },
+    {
+      "epoch": 0.9084290567869276,
+      "grad_norm": 0.940534897284778,
+      "learning_rate": 2.533428094715473e-07,
+      "loss": 0.8232,
+      "step": 5615
+    },
+    {
+      "epoch": 0.9092379873806827,
+      "grad_norm": 0.9310916767297108,
+      "learning_rate": 2.489239109492536e-07,
+      "loss": 0.8196,
+      "step": 5620
+    },
+    {
+      "epoch": 0.9100469179744378,
+      "grad_norm": 1.012544196651258,
+      "learning_rate": 2.4454290634243927e-07,
+      "loss": 0.8092,
+      "step": 5625
+    },
+    {
+      "epoch": 0.9108558485681928,
+      "grad_norm": 0.9941950459630348,
+      "learning_rate": 2.4019983059338005e-07,
+      "loss": 0.8219,
+      "step": 5630
+    },
+    {
+      "epoch": 0.9116647791619479,
+      "grad_norm": 1.0349621117968182,
+      "learning_rate": 2.3589471834183975e-07,
+      "loss": 0.7977,
+      "step": 5635
+    },
+    {
+      "epoch": 0.912473709755703,
+      "grad_norm": 0.9021187223816023,
+      "learning_rate": 2.3162760392478777e-07,
+      "loss": 0.8256,
+      "step": 5640
+    },
+    {
+      "epoch": 0.913282640349458,
+      "grad_norm": 0.9107025655941831,
+      "learning_rate": 2.2739852137612984e-07,
+      "loss": 0.8279,
+      "step": 5645
+    },
+    {
+      "epoch": 0.914091570943213,
+      "grad_norm": 0.9459797428393495,
+      "learning_rate": 2.2320750442643423e-07,
+      "loss": 0.8077,
+      "step": 5650
+    },
+    {
+      "epoch": 0.9149005015369681,
+      "grad_norm": 1.1288128324454283,
+      "learning_rate": 2.1905458650266276e-07,
+      "loss": 0.8407,
+      "step": 5655
+    },
+    {
+      "epoch": 0.9157094321307232,
+      "grad_norm": 0.9221316094387947,
+      "learning_rate": 2.149398007279052e-07,
+      "loss": 0.8209,
+      "step": 5660
+    },
+    {
+      "epoch": 0.9165183627244783,
+      "grad_norm": 0.9047592079692445,
+      "learning_rate": 2.108631799211158e-07,
+      "loss": 0.8037,
+      "step": 5665
+    },
+    {
+      "epoch": 0.9173272933182333,
+      "grad_norm": 0.9186500058678413,
+      "learning_rate": 2.0682475659684953e-07,
+      "loss": 0.836,
+      "step": 5670
+    },
+    {
+      "epoch": 0.9181362239119883,
+      "grad_norm": 0.8965926605186532,
+      "learning_rate": 2.0282456296500385e-07,
+      "loss": 0.8303,
+      "step": 5675
+    },
+    {
+      "epoch": 0.9189451545057434,
+      "grad_norm": 0.9119449898107598,
+      "learning_rate": 1.988626309305597e-07,
+      "loss": 0.838,
+      "step": 5680
+    },
+    {
+      "epoch": 0.9197540850994984,
+      "grad_norm": 0.9512693378379136,
+      "learning_rate": 1.9493899209333146e-07,
+      "loss": 0.8027,
+      "step": 5685
+    },
+    {
+      "epoch": 0.9205630156932535,
+      "grad_norm": 0.9690819188637002,
+      "learning_rate": 1.9105367774771122e-07,
+      "loss": 0.8125,
+      "step": 5690
+    },
+    {
+      "epoch": 0.9213719462870086,
+      "grad_norm": 0.9571795973045409,
+      "learning_rate": 1.8720671888242058e-07,
+      "loss": 0.8271,
+      "step": 5695
+    },
+    {
+      "epoch": 0.9221808768807637,
+      "grad_norm": 0.909548062163855,
+      "learning_rate": 1.8339814618026252e-07,
+      "loss": 0.833,
+      "step": 5700
+    },
+    {
+      "epoch": 0.9229898074745186,
+      "grad_norm": 0.9572976405315803,
+      "learning_rate": 1.7962799001787823e-07,
+      "loss": 0.8103,
+      "step": 5705
+    },
+    {
+      "epoch": 0.9237987380682737,
+      "grad_norm": 1.0064204020940664,
+      "learning_rate": 1.7589628046550343e-07,
+      "loss": 0.8369,
+      "step": 5710
+    },
+    {
+      "epoch": 0.9246076686620288,
+      "grad_norm": 0.9734981320914123,
+      "learning_rate": 1.7220304728672977e-07,
+      "loss": 0.8187,
+      "step": 5715
+    },
+    {
+      "epoch": 0.9254165992557839,
+      "grad_norm": 0.9779299096983186,
+      "learning_rate": 1.6854831993826591e-07,
+      "loss": 0.8212,
+      "step": 5720
+    },
+    {
+      "epoch": 0.9262255298495389,
+      "grad_norm": 0.8990160141351417,
+      "learning_rate": 1.6493212756970356e-07,
+      "loss": 0.8194,
+      "step": 5725
+    },
+    {
+      "epoch": 0.927034460443294,
+      "grad_norm": 0.9133770658973546,
+      "learning_rate": 1.6135449902328627e-07,
+      "loss": 0.8377,
+      "step": 5730
+    },
+    {
+      "epoch": 0.927843391037049,
+      "grad_norm": 0.8935752099105897,
+      "learning_rate": 1.578154628336753e-07,
+      "loss": 0.8247,
+      "step": 5735
+    },
+    {
+      "epoch": 0.928652321630804,
+      "grad_norm": 0.9599763188366742,
+      "learning_rate": 1.5431504722772605e-07,
+      "loss": 0.8259,
+      "step": 5740
+    },
+    {
+      "epoch": 0.9294612522245591,
+      "grad_norm": 0.9543468652422167,
+      "learning_rate": 1.5085328012426293e-07,
+      "loss": 0.8239,
+      "step": 5745
+    },
+    {
+      "epoch": 0.9302701828183142,
+      "grad_norm": 0.9093126800081541,
+      "learning_rate": 1.4743018913385308e-07,
+      "loss": 0.8313,
+      "step": 5750
+    },
+    {
+      "epoch": 0.9310791134120693,
+      "grad_norm": 1.002356581515829,
+      "learning_rate": 1.4404580155859106e-07,
+      "loss": 0.8288,
+      "step": 5755
+    },
+    {
+      "epoch": 0.9318880440058243,
+      "grad_norm": 0.900910009096251,
+      "learning_rate": 1.407001443918743e-07,
+      "loss": 0.824,
+      "step": 5760
+    },
+    {
+      "epoch": 0.9326969745995793,
+      "grad_norm": 0.9316638128288214,
+      "learning_rate": 1.373932443181958e-07,
+      "loss": 0.8376,
+      "step": 5765
+    },
+    {
+      "epoch": 0.9335059051933344,
+      "grad_norm": 0.8859091785962655,
+      "learning_rate": 1.3412512771292574e-07,
+      "loss": 0.8262,
+      "step": 5770
+    },
+    {
+      "epoch": 0.9343148357870895,
+      "grad_norm": 0.9517796063815803,
+      "learning_rate": 1.3089582064210293e-07,
+      "loss": 0.8332,
+      "step": 5775
+    },
+    {
+      "epoch": 0.9351237663808445,
+      "grad_norm": 0.956425135920899,
+      "learning_rate": 1.2770534886222709e-07,
+      "loss": 0.8441,
+      "step": 5780
+    },
+    {
+      "epoch": 0.9359326969745996,
+      "grad_norm": 0.9436260486622394,
+      "learning_rate": 1.2455373782005343e-07,
+      "loss": 0.8246,
+      "step": 5785
+    },
+    {
+      "epoch": 0.9367416275683547,
+      "grad_norm": 0.8907054197780977,
+      "learning_rate": 1.2144101265238795e-07,
+      "loss": 0.8226,
+      "step": 5790
+    },
+    {
+      "epoch": 0.9375505581621096,
+      "grad_norm": 0.9599764091397304,
+      "learning_rate": 1.1836719818588971e-07,
+      "loss": 0.8197,
+      "step": 5795
+    },
+    {
+      "epoch": 0.9383594887558647,
+      "grad_norm": 0.89616146112279,
+      "learning_rate": 1.1533231893687158e-07,
+      "loss": 0.8292,
+      "step": 5800
+    },
+    {
+      "epoch": 0.9391684193496198,
+      "grad_norm": 0.9254286869197753,
+      "learning_rate": 1.1233639911110317e-07,
+      "loss": 0.8106,
+      "step": 5805
+    },
+    {
+      "epoch": 0.9399773499433749,
+      "grad_norm": 0.9336133063869916,
+      "learning_rate": 1.0937946260362154e-07,
+      "loss": 0.8174,
+      "step": 5810
+    },
+    {
+      "epoch": 0.9407862805371299,
+      "grad_norm": 0.9193663783506368,
+      "learning_rate": 1.0646153299853523e-07,
+      "loss": 0.8219,
+      "step": 5815
+    },
+    {
+      "epoch": 0.941595211130885,
+      "grad_norm": 0.9922561636241934,
+      "learning_rate": 1.0358263356884223e-07,
+      "loss": 0.8192,
+      "step": 5820
+    },
+    {
+      "epoch": 0.94240414172464,
+      "grad_norm": 0.9269516184874731,
+      "learning_rate": 1.0074278727623955e-07,
+      "loss": 0.8342,
+      "step": 5825
+    },
+    {
+      "epoch": 0.9432130723183951,
+      "grad_norm": 0.9567075615786765,
+      "learning_rate": 9.794201677094162e-08,
+      "loss": 0.8456,
+      "step": 5830
+    },
+    {
+      "epoch": 0.9440220029121501,
+      "grad_norm": 0.900361140775962,
+      "learning_rate": 9.51803443915017e-08,
+      "loss": 0.8071,
+      "step": 5835
+    },
+    {
+      "epoch": 0.9448309335059052,
+      "grad_norm": 0.8928244797252599,
+      "learning_rate": 9.245779216463024e-08,
+      "loss": 0.8184,
+      "step": 5840
+    },
+    {
+      "epoch": 0.9456398640996603,
+      "grad_norm": 1.0034007579574744,
+      "learning_rate": 8.977438180502118e-08,
+      "loss": 0.8356,
+      "step": 5845
+    },
+    {
+      "epoch": 0.9464487946934153,
+      "grad_norm": 0.9678539855117049,
+      "learning_rate": 8.713013471517873e-08,
+      "loss": 0.8346,
+      "step": 5850
+    },
+    {
+      "epoch": 0.9472577252871703,
+      "grad_norm": 0.9529843175929286,
+      "learning_rate": 8.452507198524584e-08,
+      "loss": 0.8101,
+      "step": 5855
+    },
+    {
+      "epoch": 0.9480666558809254,
+      "grad_norm": 0.9262321483199513,
+      "learning_rate": 8.195921439283716e-08,
+      "loss": 0.8242,
+      "step": 5860
+    },
+    {
+      "epoch": 0.9488755864746805,
+      "grad_norm": 0.885777314821936,
+      "learning_rate": 7.943258240287355e-08,
+      "loss": 0.8201,
+      "step": 5865
+    },
+    {
+      "epoch": 0.9496845170684355,
+      "grad_norm": 0.9553802642628383,
+      "learning_rate": 7.694519616741503e-08,
+      "loss": 0.8331,
+      "step": 5870
+    },
+    {
+      "epoch": 0.9504934476621906,
+      "grad_norm": 0.9768522628826408,
+      "learning_rate": 7.449707552550533e-08,
+      "loss": 0.8147,
+      "step": 5875
+    },
+    {
+      "epoch": 0.9513023782559457,
+      "grad_norm": 0.9068927130823696,
+      "learning_rate": 7.208824000301151e-08,
+      "loss": 0.8302,
+      "step": 5880
+    },
+    {
+      "epoch": 0.9521113088497007,
+      "grad_norm": 0.9575224681944164,
+      "learning_rate": 6.971870881246678e-08,
+      "loss": 0.8145,
+      "step": 5885
+    },
+    {
+      "epoch": 0.9529202394434557,
+      "grad_norm": 0.8832000510088256,
+      "learning_rate": 6.738850085291904e-08,
+      "loss": 0.8177,
+      "step": 5890
+    },
+    {
+      "epoch": 0.9537291700372108,
+      "grad_norm": 0.9566260863289568,
+      "learning_rate": 6.509763470977925e-08,
+      "loss": 0.8321,
+      "step": 5895
+    },
+    {
+      "epoch": 0.9545381006309659,
+      "grad_norm": 0.9104128242370191,
+      "learning_rate": 6.284612865467499e-08,
+      "loss": 0.8271,
+      "step": 5900
+    },
+    {
+      "epoch": 0.9553470312247209,
+      "grad_norm": 0.9365563187928431,
+      "learning_rate": 6.063400064530155e-08,
+      "loss": 0.8251,
+      "step": 5905
+    },
+    {
+      "epoch": 0.956155961818476,
+      "grad_norm": 0.8777597042529512,
+      "learning_rate": 5.8461268325281096e-08,
+      "loss": 0.8021,
+      "step": 5910
+    },
+    {
+      "epoch": 0.956964892412231,
+      "grad_norm": 0.9899981204211554,
+      "learning_rate": 5.632794902402205e-08,
+      "loss": 0.8099,
+      "step": 5915
+    },
+    {
+      "epoch": 0.9577738230059861,
+      "grad_norm": 0.900474026342178,
+      "learning_rate": 5.423405975657936e-08,
+      "loss": 0.8367,
+      "step": 5920
+    },
+    {
+      "epoch": 0.9585827535997411,
+      "grad_norm": 1.0047500482028124,
+      "learning_rate": 5.217961722351894e-08,
+      "loss": 0.8343,
+      "step": 5925
+    },
+    {
+      "epoch": 0.9593916841934962,
+      "grad_norm": 0.9477198924309056,
+      "learning_rate": 5.016463781078618e-08,
+      "loss": 0.8317,
+      "step": 5930
+    },
+    {
+      "epoch": 0.9602006147872513,
+      "grad_norm": 0.8807307401976437,
+      "learning_rate": 4.818913758957378e-08,
+      "loss": 0.8239,
+      "step": 5935
+    },
+    {
+      "epoch": 0.9610095453810064,
+      "grad_norm": 0.9428322183465226,
+      "learning_rate": 4.6253132316194103e-08,
+      "loss": 0.8296,
+      "step": 5940
+    },
+    {
+      "epoch": 0.9618184759747613,
+      "grad_norm": 0.9762495130698056,
+      "learning_rate": 4.4356637431953734e-08,
+      "loss": 0.8035,
+      "step": 5945
+    },
+    {
+      "epoch": 0.9626274065685164,
+      "grad_norm": 0.9782271222260458,
+      "learning_rate": 4.2499668063029075e-08,
+      "loss": 0.8079,
+      "step": 5950
+    },
+    {
+      "epoch": 0.9634363371622715,
+      "grad_norm": 1.0490239258099892,
+      "learning_rate": 4.068223902034651e-08,
+      "loss": 0.827,
+      "step": 5955
+    },
+    {
+      "epoch": 0.9642452677560265,
+      "grad_norm": 0.9547918261422726,
+      "learning_rate": 3.89043647994658e-08,
+      "loss": 0.8309,
+      "step": 5960
+    },
+    {
+      "epoch": 0.9650541983497816,
+      "grad_norm": 0.9195603493515522,
+      "learning_rate": 3.716605958046071e-08,
+      "loss": 0.8324,
+      "step": 5965
+    },
+    {
+      "epoch": 0.9658631289435367,
+      "grad_norm": 0.9215008996007901,
+      "learning_rate": 3.546733722781026e-08,
+      "loss": 0.8202,
+      "step": 5970
+    },
+    {
+      "epoch": 0.9666720595372917,
+      "grad_norm": 0.9117492446000369,
+      "learning_rate": 3.3808211290284886e-08,
+      "loss": 0.8283,
+      "step": 5975
+    },
+    {
+      "epoch": 0.9674809901310467,
+      "grad_norm": 0.9238030724612377,
+      "learning_rate": 3.218869500084099e-08,
+      "loss": 0.8203,
+      "step": 5980
+    },
+    {
+      "epoch": 0.9682899207248018,
+      "grad_norm": 0.9683332126948087,
+      "learning_rate": 3.0608801276511556e-08,
+      "loss": 0.8063,
+      "step": 5985
+    },
+    {
+      "epoch": 0.9690988513185569,
+      "grad_norm": 0.9578244707163264,
+      "learning_rate": 2.9068542718307947e-08,
+      "loss": 0.8178,
+      "step": 5990
+    },
+    {
+      "epoch": 0.969907781912312,
+      "grad_norm": 1.0002466766621863,
+      "learning_rate": 2.7567931611116037e-08,
+      "loss": 0.8298,
+      "step": 5995
+    },
+    {
+      "epoch": 0.970716712506067,
+      "grad_norm": 0.9033323404105812,
+      "learning_rate": 2.6106979923599117e-08,
+      "loss": 0.8314,
+      "step": 6000
+    },
+    {
+      "epoch": 0.971525643099822,
+      "grad_norm": 0.9981727314544666,
+      "learning_rate": 2.4685699308102385e-08,
+      "loss": 0.8245,
+      "step": 6005
+    },
+    {
+      "epoch": 0.9723345736935771,
+      "grad_norm": 0.90324273915133,
+      "learning_rate": 2.330410110056025e-08,
+      "loss": 0.8128,
+      "step": 6010
+    },
+    {
+      "epoch": 0.9731435042873321,
+      "grad_norm": 0.9184000593665834,
+      "learning_rate": 2.1962196320406416e-08,
+      "loss": 0.8234,
+      "step": 6015
+    },
+    {
+      "epoch": 0.9739524348810872,
+      "grad_norm": 1.0208286644340638,
+      "learning_rate": 2.065999567048449e-08,
+      "loss": 0.8313,
+      "step": 6020
+    },
+    {
+      "epoch": 0.9747613654748423,
+      "grad_norm": 0.8848130385791,
+      "learning_rate": 1.9397509536964177e-08,
+      "loss": 0.8045,
+      "step": 6025
+    },
+    {
+      "epoch": 0.9755702960685974,
+      "grad_norm": 0.8868943198262542,
+      "learning_rate": 1.8174747989258002e-08,
+      "loss": 0.8181,
+      "step": 6030
+    },
+    {
+      "epoch": 0.9763792266623523,
+      "grad_norm": 0.9860697437983315,
+      "learning_rate": 1.6991720779939157e-08,
+      "loss": 0.8269,
+      "step": 6035
+    },
+    {
+      "epoch": 0.9771881572561074,
+      "grad_norm": 0.9169269167928082,
+      "learning_rate": 1.5848437344667124e-08,
+      "loss": 0.8155,
+      "step": 6040
+    },
+    {
+      "epoch": 0.9779970878498625,
+      "grad_norm": 0.8899129024501534,
+      "learning_rate": 1.4744906802110493e-08,
+      "loss": 0.814,
+      "step": 6045
+    },
+    {
+      "epoch": 0.9788060184436176,
+      "grad_norm": 0.9676626630215127,
+      "learning_rate": 1.3681137953872604e-08,
+      "loss": 0.8248,
+      "step": 6050
+    },
+    {
+      "epoch": 0.9796149490373726,
+      "grad_norm": 1.008805163098565,
+      "learning_rate": 1.2657139284425468e-08,
+      "loss": 0.8442,
+      "step": 6055
+    },
+    {
+      "epoch": 0.9804238796311276,
+      "grad_norm": 0.9275436235635999,
+      "learning_rate": 1.167291896103817e-08,
+      "loss": 0.8255,
+      "step": 6060
+    },
+    {
+      "epoch": 0.9812328102248827,
+      "grad_norm": 0.9859360057982008,
+      "learning_rate": 1.0728484833713581e-08,
+      "loss": 0.8231,
+      "step": 6065
+    },
+    {
+      "epoch": 0.9820417408186377,
+      "grad_norm": 0.9537378986780001,
+      "learning_rate": 9.823844435126184e-09,
+      "loss": 0.8204,
+      "step": 6070
+    },
+    {
+      "epoch": 0.9828506714123928,
+      "grad_norm": 0.9029585570220431,
+      "learning_rate": 8.959004980559905e-09,
+      "loss": 0.8351,
+      "step": 6075
+    },
+    {
+      "epoch": 0.9836596020061479,
+      "grad_norm": 0.9441776141647337,
+      "learning_rate": 8.133973367853708e-09,
+      "loss": 0.8401,
+      "step": 6080
+    },
+    {
+      "epoch": 0.984468532599903,
+      "grad_norm": 0.9139424255052578,
+      "learning_rate": 7.348756177343319e-09,
+      "loss": 0.8228,
+      "step": 6085
+    },
+    {
+      "epoch": 0.9852774631936579,
+      "grad_norm": 0.9582142736178327,
+      "learning_rate": 6.603359671810694e-09,
+      "loss": 0.8239,
+      "step": 6090
+    },
+    {
+      "epoch": 0.986086393787413,
+      "grad_norm": 0.9470852983866349,
+      "learning_rate": 5.8977897964335174e-09,
+      "loss": 0.8164,
+      "step": 6095
+    },
+    {
+      "epoch": 0.9868953243811681,
+      "grad_norm": 0.9016603805684872,
+      "learning_rate": 5.232052178738567e-09,
+      "loss": 0.8232,
+      "step": 6100
+    },
+    {
+      "epoch": 0.9877042549749232,
+      "grad_norm": 0.9549402446433066,
+      "learning_rate": 4.606152128555086e-09,
+      "loss": 0.8088,
+      "step": 6105
+    },
+    {
+      "epoch": 0.9885131855686782,
+      "grad_norm": 0.9529061460613366,
+      "learning_rate": 4.020094637973704e-09,
+      "loss": 0.8162,
+      "step": 6110
+    },
+    {
+      "epoch": 0.9893221161624333,
+      "grad_norm": 0.9500963750803435,
+      "learning_rate": 3.4738843813075795e-09,
+      "loss": 0.813,
+      "step": 6115
+    },
+    {
+      "epoch": 0.9901310467561883,
+      "grad_norm": 0.9329204750410419,
+      "learning_rate": 2.967525715052433e-09,
+      "loss": 0.839,
+      "step": 6120
+    },
+    {
+      "epoch": 0.9909399773499433,
+      "grad_norm": 0.9456290946732857,
+      "learning_rate": 2.5010226778537927e-09,
+      "loss": 0.8231,
+      "step": 6125
+    },
+    {
+      "epoch": 0.9917489079436984,
+      "grad_norm": 0.8612020082608306,
+      "learning_rate": 2.074378990474246e-09,
+      "loss": 0.8139,
+      "step": 6130
+    },
+    {
+      "epoch": 0.9925578385374535,
+      "grad_norm": 0.9267109066168538,
+      "learning_rate": 1.687598055764017e-09,
+      "loss": 0.8351,
+      "step": 6135
+    },
+    {
+      "epoch": 0.9933667691312086,
+      "grad_norm": 0.9928733972042116,
+      "learning_rate": 1.3406829586337656e-09,
+      "loss": 0.8033,
+      "step": 6140
+    },
+    {
+      "epoch": 0.9941756997249636,
+      "grad_norm": 0.882709413718118,
+      "learning_rate": 1.0336364660290532e-09,
+      "loss": 0.8098,
+      "step": 6145
+    },
+    {
+      "epoch": 0.9949846303187186,
+      "grad_norm": 0.9193633255003315,
+      "learning_rate": 7.664610269103589e-10,
+      "loss": 0.8285,
+      "step": 6150
+    },
+    {
+      "epoch": 0.9957935609124737,
+      "grad_norm": 0.9524125245282298,
+      "learning_rate": 5.391587722303194e-10,
+      "loss": 0.8087,
+      "step": 6155
+    },
+    {
+      "epoch": 0.9966024915062288,
+      "grad_norm": 0.9577558462345785,
+      "learning_rate": 3.5173151492096104e-10,
+      "loss": 0.8309,
+      "step": 6160
+    },
+    {
+      "epoch": 0.9974114220999838,
+      "grad_norm": 0.892739545774551,
+      "learning_rate": 2.0418074987538229e-10,
+      "loss": 0.8144,
+      "step": 6165
+    },
+    {
+      "epoch": 0.9982203526937389,
+      "grad_norm": 0.973028280918144,
+      "learning_rate": 9.650765393720563e-11,
+      "loss": 0.8147,
+      "step": 6170
+    },
+    {
+      "epoch": 0.999029283287494,
+      "grad_norm": 0.9043858725591138,
+      "learning_rate": 2.8713085892806415e-11,
+      "loss": 0.8312,
+      "step": 6175
+    },
+    {
+      "epoch": 0.9998382138812489,
+      "grad_norm": 0.9568219074455289,
+      "learning_rate": 7.975864613207762e-13,
+      "loss": 0.8409,
+      "step": 6180
+    },
     {
       "epoch": 1.0,
-      "eval_loss": 0.28806865215301514,
-      "eval_runtime": 1.191,
-      "eval_samples_per_second": 2.519,
-      "eval_steps_per_second": 0.84,
-      "step": 2414
+      "eval_loss": 0.8599082231521606,
+      "eval_runtime": 3.0154,
+      "eval_samples_per_second": 3.316,
+      "eval_steps_per_second": 0.995,
+      "step": 6181
     },
     {
       "epoch": 1.0,
-      "step": 2414,
-      "total_flos": 252721244405760.0,
-      "train_loss": 0.5156875643339054,
-      "train_runtime": 24515.4963,
-      "train_samples_per_second": 1.575,
-      "train_steps_per_second": 0.098
+      "step": 6181,
+      "total_flos": 2571089024581632.0,
+      "train_loss": 0.8638364601301117,
+      "train_runtime": 28571.7374,
+      "train_samples_per_second": 3.461,
+      "train_steps_per_second": 0.216
     }
   ],
   "logging_steps": 5,
-  "max_steps": 2414,
+  "max_steps": 6181,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 1,
   "save_steps": 100,
@@ -3424,7 +8702,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 252721244405760.0,
+  "total_flos": 2571089024581632.0,
   "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null