diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,3826 +1,3805 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 3.0,
+  "epoch": 2.988929889298893,
   "eval_steps": 500,
-  "global_step": 543,
+  "global_step": 540,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.0055248618784530384,
-      "grad_norm": 4.8501728653187755,
-      "learning_rate": 1.8181818181818183e-07,
-      "loss": 0.882,
+      "epoch": 0.005535055350553505,
+      "grad_norm": 4.88795566772026,
+      "learning_rate": 1.8518518518518518e-07,
+      "loss": 0.8842,
       "step": 1
     },
     {
-      "epoch": 0.011049723756906077,
-      "grad_norm": 5.284618547033595,
-      "learning_rate": 3.6363636363636366e-07,
-      "loss": 0.9499,
+      "epoch": 0.01107011070110701,
+      "grad_norm": 5.294943049287627,
+      "learning_rate": 3.7037037037037036e-07,
+      "loss": 0.9582,
       "step": 2
     },
     {
-      "epoch": 0.016574585635359115,
-      "grad_norm": 4.71655301867202,
-      "learning_rate": 5.454545454545455e-07,
-      "loss": 0.8375,
+      "epoch": 0.016605166051660517,
+      "grad_norm": 4.721450383705418,
+      "learning_rate": 5.555555555555555e-07,
+      "loss": 0.8326,
       "step": 3
     },
     {
-      "epoch": 0.022099447513812154,
-      "grad_norm": 4.691610018967258,
-      "learning_rate": 7.272727272727273e-07,
-      "loss": 0.8694,
+      "epoch": 0.02214022140221402,
+      "grad_norm": 4.702865900802319,
+      "learning_rate": 7.407407407407407e-07,
+      "loss": 0.8705,
       "step": 4
     },
     {
-      "epoch": 0.027624309392265192,
-      "grad_norm": 4.525811293283286,
-      "learning_rate": 9.090909090909091e-07,
-      "loss": 0.7922,
+      "epoch": 0.027675276752767528,
+      "grad_norm": 4.588661299548798,
+      "learning_rate": 9.259259259259259e-07,
+      "loss": 0.7979,
       "step": 5
     },
     {
-      "epoch": 0.03314917127071823,
-      "grad_norm": 4.515684677465201,
-      "learning_rate": 1.090909090909091e-06,
-      "loss": 0.8158,
+      "epoch": 0.033210332103321034,
+      "grad_norm": 4.514875582905285,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 0.8133,
       "step": 6
     },
     {
-      "epoch": 0.03867403314917127,
-      "grad_norm": 4.521464462348852,
-      "learning_rate": 1.2727272727272728e-06,
-      "loss": 0.8463,
+      "epoch": 0.03874538745387454,
+      "grad_norm": 4.3889371427334725,
+      "learning_rate": 1.2962962962962962e-06,
+      "loss": 0.8289,
       "step": 7
     },
     {
-      "epoch": 0.04419889502762431,
-      "grad_norm": 4.01317039922344,
-      "learning_rate": 1.4545454545454546e-06,
-      "loss": 0.7804,
+      "epoch": 0.04428044280442804,
+      "grad_norm": 3.681708725528932,
+      "learning_rate": 1.4814814814814815e-06,
+      "loss": 0.7494,
       "step": 8
     },
     {
-      "epoch": 0.049723756906077346,
-      "grad_norm": 3.725743569896934,
-      "learning_rate": 1.6363636363636365e-06,
-      "loss": 0.8386,
+      "epoch": 0.04981549815498155,
+      "grad_norm": 3.7696517468375483,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.8392,
       "step": 9
     },
     {
-      "epoch": 0.055248618784530384,
-      "grad_norm": 3.547034380075145,
-      "learning_rate": 1.8181818181818183e-06,
-      "loss": 0.7804,
+      "epoch": 0.055350553505535055,
+      "grad_norm": 3.5736119140088474,
+      "learning_rate": 1.8518518518518519e-06,
+      "loss": 0.7802,
       "step": 10
     },
     {
-      "epoch": 0.06077348066298342,
-      "grad_norm": 2.385734305081293,
-      "learning_rate": 2.0000000000000003e-06,
-      "loss": 0.7536,
+      "epoch": 0.06088560885608856,
+      "grad_norm": 2.352449321846944,
+      "learning_rate": 2.037037037037037e-06,
+      "loss": 0.7531,
       "step": 11
     },
     {
-      "epoch": 0.06629834254143646,
-      "grad_norm": 2.118903283570385,
-      "learning_rate": 2.181818181818182e-06,
-      "loss": 0.8003,
+      "epoch": 0.06642066420664207,
+      "grad_norm": 2.1396669501274728,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 0.8033,
       "step": 12
     },
     {
-      "epoch": 0.0718232044198895,
-      "grad_norm": 2.071463502085266,
-      "learning_rate": 2.363636363636364e-06,
-      "loss": 0.7728,
+      "epoch": 0.07195571955719557,
+      "grad_norm": 2.0686046715471744,
+      "learning_rate": 2.4074074074074075e-06,
+      "loss": 0.7739,
       "step": 13
     },
     {
-      "epoch": 0.07734806629834254,
-      "grad_norm": 2.11975896906294,
-      "learning_rate": 2.5454545454545456e-06,
-      "loss": 0.7816,
+      "epoch": 0.07749077490774908,
+      "grad_norm": 2.112065821246985,
+      "learning_rate": 2.5925925925925925e-06,
+      "loss": 0.7789,
       "step": 14
     },
     {
-      "epoch": 0.08287292817679558,
-      "grad_norm": 1.6109945627440467,
-      "learning_rate": 2.7272727272727272e-06,
-      "loss": 0.7138,
+      "epoch": 0.08302583025830258,
+      "grad_norm": 1.6986741641118395,
+      "learning_rate": 2.7777777777777783e-06,
+      "loss": 0.7151,
       "step": 15
     },
     {
-      "epoch": 0.08839779005524862,
-      "grad_norm": 1.946567883491615,
-      "learning_rate": 2.9090909090909093e-06,
-      "loss": 0.7254,
+      "epoch": 0.08856088560885608,
+      "grad_norm": 1.9671299367169393,
+      "learning_rate": 2.962962962962963e-06,
+      "loss": 0.7321,
       "step": 16
     },
     {
-      "epoch": 0.09392265193370165,
-      "grad_norm": 1.7266080007457147,
-      "learning_rate": 3.090909090909091e-06,
-      "loss": 0.7294,
+      "epoch": 0.0940959409594096,
+      "grad_norm": 1.727327433279914,
+      "learning_rate": 3.1481481481481483e-06,
+      "loss": 0.7381,
       "step": 17
     },
     {
-      "epoch": 0.09944751381215469,
-      "grad_norm": 1.6539369196579166,
-      "learning_rate": 3.272727272727273e-06,
-      "loss": 0.711,
+      "epoch": 0.0996309963099631,
+      "grad_norm": 1.65624237671333,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.7159,
       "step": 18
     },
     {
-      "epoch": 0.10497237569060773,
-      "grad_norm": 1.3760121270687458,
-      "learning_rate": 3.454545454545455e-06,
-      "loss": 0.6706,
+      "epoch": 0.10516605166051661,
+      "grad_norm": 1.3535125694848185,
+      "learning_rate": 3.5185185185185187e-06,
+      "loss": 0.6755,
       "step": 19
     },
     {
-      "epoch": 0.11049723756906077,
-      "grad_norm": 1.2322332262531337,
-      "learning_rate": 3.6363636363636366e-06,
-      "loss": 0.714,
+      "epoch": 0.11070110701107011,
+      "grad_norm": 1.2125088936587491,
+      "learning_rate": 3.7037037037037037e-06,
+      "loss": 0.7032,
       "step": 20
     },
     {
-      "epoch": 0.11602209944751381,
-      "grad_norm": 1.092460838605022,
-      "learning_rate": 3.818181818181819e-06,
-      "loss": 0.6221,
+      "epoch": 0.11623616236162361,
+      "grad_norm": 1.107066250339579,
+      "learning_rate": 3.88888888888889e-06,
+      "loss": 0.6217,
       "step": 21
     },
     {
-      "epoch": 0.12154696132596685,
-      "grad_norm": 1.3847134225952789,
-      "learning_rate": 4.000000000000001e-06,
-      "loss": 0.6874,
+      "epoch": 0.12177121771217712,
+      "grad_norm": 1.3653673904628112,
+      "learning_rate": 4.074074074074074e-06,
+      "loss": 0.6824,
       "step": 22
     },
     {
-      "epoch": 0.1270718232044199,
-      "grad_norm": 1.2895226334792846,
-      "learning_rate": 4.181818181818182e-06,
-      "loss": 0.6741,
+      "epoch": 0.12730627306273062,
+      "grad_norm": 1.289828643250104,
+      "learning_rate": 4.2592592592592596e-06,
+      "loss": 0.6796,
       "step": 23
     },
     {
-      "epoch": 0.13259668508287292,
-      "grad_norm": 1.3973784329721743,
-      "learning_rate": 4.363636363636364e-06,
-      "loss": 0.8114,
+      "epoch": 0.13284132841328414,
+      "grad_norm": 1.379866304591633,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.8102,
       "step": 24
     },
     {
-      "epoch": 0.13812154696132597,
-      "grad_norm": 1.1221564987656045,
-      "learning_rate": 4.5454545454545455e-06,
-      "loss": 0.6904,
+      "epoch": 0.13837638376383765,
+      "grad_norm": 1.1200972001221432,
+      "learning_rate": 4.62962962962963e-06,
+      "loss": 0.6953,
       "step": 25
     },
     {
-      "epoch": 0.143646408839779,
-      "grad_norm": 1.0813494118759077,
-      "learning_rate": 4.727272727272728e-06,
-      "loss": 0.6887,
+      "epoch": 0.14391143911439114,
+      "grad_norm": 1.071016299230562,
+      "learning_rate": 4.814814814814815e-06,
+      "loss": 0.6914,
       "step": 26
     },
     {
-      "epoch": 0.14917127071823205,
-      "grad_norm": 1.039002428401238,
-      "learning_rate": 4.90909090909091e-06,
-      "loss": 0.7305,
+      "epoch": 0.14944649446494465,
+      "grad_norm": 1.007816252436763,
+      "learning_rate": 5e-06,
+      "loss": 0.7266,
       "step": 27
     },
     {
-      "epoch": 0.15469613259668508,
-      "grad_norm": 0.9005740059397435,
-      "learning_rate": 5.090909090909091e-06,
-      "loss": 0.7216,
+      "epoch": 0.15498154981549817,
+      "grad_norm": 0.8824750556078126,
+      "learning_rate": 5.185185185185185e-06,
+      "loss": 0.716,
       "step": 28
     },
     {
-      "epoch": 0.16022099447513813,
-      "grad_norm": 1.0434084181940664,
-      "learning_rate": 5.272727272727273e-06,
-      "loss": 0.7262,
+      "epoch": 0.16051660516605165,
+      "grad_norm": 1.0612669135866166,
+      "learning_rate": 5.370370370370371e-06,
+      "loss": 0.7289,
       "step": 29
     },
     {
-      "epoch": 0.16574585635359115,
-      "grad_norm": 0.8640462038859854,
-      "learning_rate": 5.4545454545454545e-06,
-      "loss": 0.6818,
+      "epoch": 0.16605166051660517,
+      "grad_norm": 0.8626830347073103,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 0.6848,
       "step": 30
     },
     {
-      "epoch": 0.1712707182320442,
-      "grad_norm": 0.8711602353776259,
-      "learning_rate": 5.636363636363636e-06,
-      "loss": 0.6709,
+      "epoch": 0.17158671586715868,
+      "grad_norm": 0.8679382464145016,
+      "learning_rate": 5.740740740740741e-06,
+      "loss": 0.6615,
       "step": 31
     },
     {
-      "epoch": 0.17679558011049723,
-      "grad_norm": 0.7874062745697313,
-      "learning_rate": 5.8181818181818185e-06,
-      "loss": 0.7056,
+      "epoch": 0.17712177121771217,
+      "grad_norm": 0.8023350481646084,
+      "learning_rate": 5.925925925925926e-06,
+      "loss": 0.708,
       "step": 32
     },
     {
-      "epoch": 0.18232044198895028,
-      "grad_norm": 0.910787056447978,
-      "learning_rate": 6e-06,
-      "loss": 0.7321,
+      "epoch": 0.18265682656826568,
+      "grad_norm": 0.9095885057238613,
+      "learning_rate": 6.111111111111112e-06,
+      "loss": 0.7367,
       "step": 33
     },
     {
-      "epoch": 0.1878453038674033,
-      "grad_norm": 0.8834639779382965,
-      "learning_rate": 6.181818181818182e-06,
-      "loss": 0.6943,
+      "epoch": 0.1881918819188192,
+      "grad_norm": 0.8887858125212083,
+      "learning_rate": 6.296296296296297e-06,
+      "loss": 0.7039,
       "step": 34
     },
     {
-      "epoch": 0.19337016574585636,
-      "grad_norm": 0.9011491907153977,
-      "learning_rate": 6.363636363636364e-06,
-      "loss": 0.7474,
+      "epoch": 0.1937269372693727,
+      "grad_norm": 0.9119816001938974,
+      "learning_rate": 6.481481481481482e-06,
+      "loss": 0.7479,
       "step": 35
     },
     {
-      "epoch": 0.19889502762430938,
-      "grad_norm": 0.9863965225958002,
-      "learning_rate": 6.545454545454546e-06,
-      "loss": 0.7448,
+      "epoch": 0.1992619926199262,
+      "grad_norm": 0.9750213405228529,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.7354,
       "step": 36
     },
     {
-      "epoch": 0.20441988950276244,
-      "grad_norm": 0.804971042531312,
-      "learning_rate": 6.7272727272727275e-06,
-      "loss": 0.6026,
+      "epoch": 0.2047970479704797,
+      "grad_norm": 0.7893115852707117,
+      "learning_rate": 6.851851851851853e-06,
+      "loss": 0.5972,
       "step": 37
     },
     {
-      "epoch": 0.20994475138121546,
-      "grad_norm": 0.8495022527176822,
-      "learning_rate": 6.90909090909091e-06,
-      "loss": 0.686,
+      "epoch": 0.21033210332103322,
+      "grad_norm": 0.8249712148416332,
+      "learning_rate": 7.0370370370370375e-06,
+      "loss": 0.6766,
       "step": 38
     },
     {
-      "epoch": 0.2154696132596685,
-      "grad_norm": 0.8536642393909392,
-      "learning_rate": 7.0909090909090916e-06,
-      "loss": 0.6863,
+      "epoch": 0.2158671586715867,
+      "grad_norm": 0.8548009416827502,
+      "learning_rate": 7.222222222222223e-06,
+      "loss": 0.6979,
       "step": 39
     },
     {
-      "epoch": 0.22099447513812154,
-      "grad_norm": 0.7505296486129153,
-      "learning_rate": 7.272727272727273e-06,
-      "loss": 0.6495,
+      "epoch": 0.22140221402214022,
+      "grad_norm": 0.7430327142800517,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.6364,
       "step": 40
     },
     {
-      "epoch": 0.2265193370165746,
-      "grad_norm": 0.8613764113585566,
-      "learning_rate": 7.454545454545456e-06,
-      "loss": 0.6903,
+      "epoch": 0.22693726937269373,
+      "grad_norm": 0.8804985241917266,
+      "learning_rate": 7.592592592592594e-06,
+      "loss": 0.7009,
       "step": 41
     },
     {
-      "epoch": 0.23204419889502761,
-      "grad_norm": 0.9071577576208497,
-      "learning_rate": 7.636363636363638e-06,
-      "loss": 0.6681,
+      "epoch": 0.23247232472324722,
+      "grad_norm": 0.8685723187944967,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 0.665,
       "step": 42
     },
     {
-      "epoch": 0.23756906077348067,
-      "grad_norm": 0.7806943160679269,
-      "learning_rate": 7.81818181818182e-06,
-      "loss": 0.6454,
+      "epoch": 0.23800738007380073,
+      "grad_norm": 0.7863830014618772,
+      "learning_rate": 7.962962962962963e-06,
+      "loss": 0.6506,
       "step": 43
     },
     {
-      "epoch": 0.2430939226519337,
-      "grad_norm": 0.8542316753334585,
-      "learning_rate": 8.000000000000001e-06,
-      "loss": 0.6943,
+      "epoch": 0.24354243542435425,
+      "grad_norm": 0.89002673850871,
+      "learning_rate": 8.148148148148148e-06,
+      "loss": 0.704,
       "step": 44
     },
     {
-      "epoch": 0.24861878453038674,
-      "grad_norm": 0.8806155598957247,
-      "learning_rate": 8.181818181818183e-06,
-      "loss": 0.6376,
+      "epoch": 0.24907749077490776,
+      "grad_norm": 0.9353559189317285,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.6407,
       "step": 45
     },
     {
-      "epoch": 0.2541436464088398,
-      "grad_norm": 0.9140053110890461,
-      "learning_rate": 8.363636363636365e-06,
-      "loss": 0.6873,
+      "epoch": 0.25461254612546125,
+      "grad_norm": 0.8860584109544853,
+      "learning_rate": 8.518518518518519e-06,
+      "loss": 0.6877,
       "step": 46
     },
     {
-      "epoch": 0.2596685082872928,
-      "grad_norm": 0.9696816209283575,
-      "learning_rate": 8.545454545454546e-06,
-      "loss": 0.6275,
+      "epoch": 0.26014760147601473,
+      "grad_norm": 1.0345385393930384,
+      "learning_rate": 8.703703703703705e-06,
+      "loss": 0.6195,
       "step": 47
     },
     {
-      "epoch": 0.26519337016574585,
-      "grad_norm": 0.8011554339035484,
-      "learning_rate": 8.727272727272728e-06,
-      "loss": 0.7049,
+      "epoch": 0.2656826568265683,
+      "grad_norm": 0.8390822845074132,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 0.7143,
       "step": 48
     },
     {
-      "epoch": 0.27071823204419887,
-      "grad_norm": 0.8214450369964688,
-      "learning_rate": 8.90909090909091e-06,
-      "loss": 0.6914,
+      "epoch": 0.27121771217712176,
+      "grad_norm": 0.7829416679007993,
+      "learning_rate": 9.074074074074075e-06,
+      "loss": 0.6848,
       "step": 49
     },
     {
-      "epoch": 0.27624309392265195,
-      "grad_norm": 0.9615553631195245,
-      "learning_rate": 9.090909090909091e-06,
-      "loss": 0.6982,
+      "epoch": 0.2767527675276753,
+      "grad_norm": 1.0206245762170603,
+      "learning_rate": 9.25925925925926e-06,
+      "loss": 0.7,
       "step": 50
     },
     {
-      "epoch": 0.281767955801105,
-      "grad_norm": 0.8354942020446344,
-      "learning_rate": 9.272727272727273e-06,
-      "loss": 0.7023,
+      "epoch": 0.2822878228782288,
+      "grad_norm": 0.8735389450158357,
+      "learning_rate": 9.444444444444445e-06,
+      "loss": 0.7061,
       "step": 51
     },
     {
-      "epoch": 0.287292817679558,
-      "grad_norm": 0.8299384272303495,
-      "learning_rate": 9.454545454545456e-06,
-      "loss": 0.695,
+      "epoch": 0.2878228782287823,
+      "grad_norm": 0.813152245836957,
+      "learning_rate": 9.62962962962963e-06,
+      "loss": 0.6943,
       "step": 52
     },
     {
-      "epoch": 0.292817679558011,
-      "grad_norm": 0.919798873510162,
-      "learning_rate": 9.636363636363638e-06,
-      "loss": 0.7185,
+      "epoch": 0.2933579335793358,
+      "grad_norm": 0.9984159950425506,
+      "learning_rate": 9.814814814814815e-06,
+      "loss": 0.736,
       "step": 53
     },
     {
-      "epoch": 0.2983425414364641,
-      "grad_norm": 0.8232925581988623,
-      "learning_rate": 9.81818181818182e-06,
-      "loss": 0.6864,
+      "epoch": 0.2988929889298893,
+      "grad_norm": 0.8289001717630332,
+      "learning_rate": 1e-05,
+      "loss": 0.6843,
       "step": 54
     },
     {
-      "epoch": 0.30386740331491713,
-      "grad_norm": 0.7920084479497609,
-      "learning_rate": 1e-05,
-      "loss": 0.6191,
+      "epoch": 0.3044280442804428,
+      "grad_norm": 0.7804823508929742,
+      "learning_rate": 9.999895536228031e-06,
+      "loss": 0.6178,
       "step": 55
     },
     {
-      "epoch": 0.30939226519337015,
-      "grad_norm": 0.8758601266592275,
-      "learning_rate": 9.999896390730872e-06,
-      "loss": 0.6469,
+      "epoch": 0.30996309963099633,
+      "grad_norm": 0.8747433453571559,
+      "learning_rate": 9.999582149277188e-06,
+      "loss": 0.6482,
       "step": 56
     },
     {
-      "epoch": 0.3149171270718232,
-      "grad_norm": 1.0142576779420058,
-      "learning_rate": 9.99958556721744e-06,
-      "loss": 0.7147,
+      "epoch": 0.3154981549815498,
+      "grad_norm": 0.9957208410673493,
+      "learning_rate": 9.999059852242508e-06,
+      "loss": 0.7054,
       "step": 57
     },
     {
-      "epoch": 0.32044198895027626,
-      "grad_norm": 0.788012902837433,
-      "learning_rate": 9.99906754234138e-06,
-      "loss": 0.6935,
+      "epoch": 0.3210332103321033,
+      "grad_norm": 0.7818825969466411,
+      "learning_rate": 9.998328666948437e-06,
+      "loss": 0.6968,
       "step": 58
     },
     {
-      "epoch": 0.3259668508287293,
-      "grad_norm": 0.8574442174017383,
-      "learning_rate": 9.998342337571566e-06,
-      "loss": 0.6852,
+      "epoch": 0.32656826568265684,
+      "grad_norm": 0.8999544343447169,
+      "learning_rate": 9.997388623947927e-06,
+      "loss": 0.6979,
       "step": 59
     },
     {
-      "epoch": 0.3314917127071823,
-      "grad_norm": 0.8678503120132645,
-      "learning_rate": 9.997409982963173e-06,
-      "loss": 0.594,
+      "epoch": 0.33210332103321033,
+      "grad_norm": 0.877731722321392,
+      "learning_rate": 9.996239762521152e-06,
+      "loss": 0.5952,
       "step": 60
     },
     {
-      "epoch": 0.3370165745856354,
-      "grad_norm": 0.7829472641187216,
-      "learning_rate": 9.996270517156431e-06,
-      "loss": 0.6359,
+      "epoch": 0.3376383763837638,
+      "grad_norm": 0.7490764317219278,
+      "learning_rate": 9.994882130673869e-06,
+      "loss": 0.6225,
       "step": 61
     },
     {
-      "epoch": 0.3425414364640884,
-      "grad_norm": 0.9057620068251797,
-      "learning_rate": 9.994923987375029e-06,
-      "loss": 0.6948,
+      "epoch": 0.34317343173431736,
+      "grad_norm": 0.8743296067498224,
+      "learning_rate": 9.993315785135417e-06,
+      "loss": 0.6924,
       "step": 62
     },
     {
-      "epoch": 0.34806629834254144,
-      "grad_norm": 0.865674232498531,
-      "learning_rate": 9.993370449424153e-06,
-      "loss": 0.73,
+      "epoch": 0.34870848708487084,
+      "grad_norm": 0.8559368435923252,
+      "learning_rate": 9.991540791356342e-06,
+      "loss": 0.7266,
       "step": 63
     },
     {
-      "epoch": 0.35359116022099446,
-      "grad_norm": 0.9444683215382766,
-      "learning_rate": 9.991609967688177e-06,
-      "loss": 0.7042,
+      "epoch": 0.35424354243542433,
+      "grad_norm": 0.8920516831324726,
+      "learning_rate": 9.989557223505661e-06,
+      "loss": 0.705,
       "step": 64
     },
     {
-      "epoch": 0.35911602209944754,
-      "grad_norm": 0.8179346757831395,
-      "learning_rate": 9.98964261512799e-06,
-      "loss": 0.7034,
+      "epoch": 0.35977859778597787,
+      "grad_norm": 0.8193570293903212,
+      "learning_rate": 9.987365164467767e-06,
+      "loss": 0.6945,
       "step": 65
     },
     {
-      "epoch": 0.36464088397790057,
-      "grad_norm": 0.8424333161772681,
-      "learning_rate": 9.987468473277975e-06,
-      "loss": 0.6368,
+      "epoch": 0.36531365313653136,
+      "grad_norm": 0.8473504309187051,
+      "learning_rate": 9.98496470583896e-06,
+      "loss": 0.6412,
       "step": 66
     },
     {
-      "epoch": 0.3701657458563536,
-      "grad_norm": 1.1453016585540032,
-      "learning_rate": 9.985087632242634e-06,
-      "loss": 0.7059,
+      "epoch": 0.37084870848708484,
+      "grad_norm": 1.14566744243091,
+      "learning_rate": 9.98235594792363e-06,
+      "loss": 0.7258,
       "step": 67
     },
     {
-      "epoch": 0.3756906077348066,
-      "grad_norm": 0.8910424821188626,
-      "learning_rate": 9.982500190692846e-06,
-      "loss": 0.7011,
+      "epoch": 0.3763837638376384,
+      "grad_norm": 0.8763668658056738,
+      "learning_rate": 9.979538999730047e-06,
+      "loss": 0.7084,
       "step": 68
     },
     {
-      "epoch": 0.3812154696132597,
-      "grad_norm": 0.8430209673516222,
-      "learning_rate": 9.97970625586178e-06,
-      "loss": 0.7038,
+      "epoch": 0.38191881918819187,
+      "grad_norm": 0.8849767742762311,
+      "learning_rate": 9.976513978965829e-06,
+      "loss": 0.6984,
       "step": 69
     },
     {
-      "epoch": 0.3867403314917127,
-      "grad_norm": 1.0961889806167662,
-      "learning_rate": 9.976705943540458e-06,
-      "loss": 0.7167,
+      "epoch": 0.3874538745387454,
+      "grad_norm": 1.0687474160518997,
+      "learning_rate": 9.973281012033009e-06,
+      "loss": 0.7183,
       "step": 70
     },
     {
-      "epoch": 0.39226519337016574,
-      "grad_norm": 0.776673389768424,
-      "learning_rate": 9.973499378072947e-06,
-      "loss": 0.6383,
+      "epoch": 0.3929889298892989,
+      "grad_norm": 0.781213177981436,
+      "learning_rate": 9.96984023402275e-06,
+      "loss": 0.6504,
       "step": 71
     },
     {
-      "epoch": 0.39779005524861877,
-      "grad_norm": 0.7677036148013285,
-      "learning_rate": 9.970086692351204e-06,
-      "loss": 0.6711,
+      "epoch": 0.3985239852398524,
+      "grad_norm": 0.7379835559329746,
+      "learning_rate": 9.966191788709716e-06,
+      "loss": 0.6661,
       "step": 72
     },
     {
-      "epoch": 0.40331491712707185,
-      "grad_norm": 0.7769311932254055,
-      "learning_rate": 9.966468027809582e-06,
-      "loss": 0.6631,
+      "epoch": 0.4040590405904059,
+      "grad_norm": 0.7984703551016957,
+      "learning_rate": 9.962335828546049e-06,
+      "loss": 0.6613,
       "step": 73
     },
     {
-      "epoch": 0.4088397790055249,
-      "grad_norm": 1.345391434823482,
-      "learning_rate": 9.962643534418954e-06,
-      "loss": 0.7296,
+      "epoch": 0.4095940959409594,
+      "grad_norm": 1.2794348232710167,
+      "learning_rate": 9.958272514655006e-06,
+      "loss": 0.7307,
       "step": 74
     },
     {
-      "epoch": 0.4143646408839779,
-      "grad_norm": 0.8085666056467051,
-      "learning_rate": 9.958613370680507e-06,
-      "loss": 0.6619,
+      "epoch": 0.4151291512915129,
+      "grad_norm": 0.7630430424338107,
+      "learning_rate": 9.954002016824226e-06,
+      "loss": 0.665,
       "step": 75
     },
     {
-      "epoch": 0.4198895027624309,
-      "grad_norm": 0.726842812492243,
-      "learning_rate": 9.954377703619171e-06,
-      "loss": 0.6449,
+      "epoch": 0.42066420664206644,
+      "grad_norm": 0.7527869433219629,
+      "learning_rate": 9.949524513498636e-06,
+      "loss": 0.6424,
       "step": 76
     },
     {
-      "epoch": 0.425414364640884,
-      "grad_norm": 1.0071149507052566,
-      "learning_rate": 9.949936708776692e-06,
-      "loss": 0.761,
+      "epoch": 0.4261992619926199,
+      "grad_norm": 1.0051926115967615,
+      "learning_rate": 9.944840191772987e-06,
+      "loss": 0.7625,
       "step": 77
     },
     {
-      "epoch": 0.430939226519337,
-      "grad_norm": 0.8013700509400903,
-      "learning_rate": 9.945290570204361e-06,
-      "loss": 0.6456,
+      "epoch": 0.4317343173431734,
+      "grad_norm": 0.7754322119272198,
+      "learning_rate": 9.939949247384046e-06,
+      "loss": 0.648,
       "step": 78
     },
     {
-      "epoch": 0.43646408839779005,
-      "grad_norm": 0.7153613260752815,
-      "learning_rate": 9.940439480455386e-06,
-      "loss": 0.6883,
+      "epoch": 0.43726937269372695,
+      "grad_norm": 0.7146858924414144,
+      "learning_rate": 9.934851884702415e-06,
+      "loss": 0.6895,
       "step": 79
     },
     {
-      "epoch": 0.4419889502762431,
-      "grad_norm": 0.8042269533559475,
-      "learning_rate": 9.935383640576915e-06,
-      "loss": 0.7192,
+      "epoch": 0.44280442804428044,
+      "grad_norm": 0.8084465018612776,
+      "learning_rate": 9.929548316723983e-06,
+      "loss": 0.7189,
       "step": 80
     },
     {
-      "epoch": 0.44751381215469616,
-      "grad_norm": 0.8007389578297808,
-      "learning_rate": 9.930123260101697e-06,
-      "loss": 0.7022,
+      "epoch": 0.4483394833948339,
+      "grad_norm": 0.8066377981893613,
+      "learning_rate": 9.924038765061042e-06,
+      "loss": 0.6995,
       "step": 81
     },
     {
-      "epoch": 0.4530386740331492,
-      "grad_norm": 0.7451865462622602,
-      "learning_rate": 9.9246585570394e-06,
-      "loss": 0.653,
+      "epoch": 0.45387453874538747,
+      "grad_norm": 0.738107027929258,
+      "learning_rate": 9.918323459933006e-06,
+      "loss": 0.6529,
       "step": 82
     },
     {
-      "epoch": 0.4585635359116022,
-      "grad_norm": 0.7629901737313624,
-      "learning_rate": 9.918989757867584e-06,
-      "loss": 0.7316,
+      "epoch": 0.45940959409594095,
+      "grad_norm": 0.7741808681072159,
+      "learning_rate": 9.912402640156812e-06,
+      "loss": 0.7351,
       "step": 83
     },
     {
-      "epoch": 0.46408839779005523,
-      "grad_norm": 0.694181698027841,
-      "learning_rate": 9.9131170975223e-06,
-      "loss": 0.632,
+      "epoch": 0.46494464944649444,
+      "grad_norm": 0.7180376057707722,
+      "learning_rate": 9.906276553136924e-06,
+      "loss": 0.6305,
       "step": 84
     },
     {
-      "epoch": 0.4696132596685083,
-      "grad_norm": 0.8031579121005685,
-      "learning_rate": 9.907040819388372e-06,
-      "loss": 0.7125,
+      "epoch": 0.470479704797048,
+      "grad_norm": 0.8145618675350315,
+      "learning_rate": 9.899945454855007e-06,
+      "loss": 0.7026,
       "step": 85
     },
     {
-      "epoch": 0.47513812154696133,
-      "grad_norm": 0.8027715808171925,
-      "learning_rate": 9.90076117528929e-06,
-      "loss": 0.7054,
+      "epoch": 0.47601476014760147,
+      "grad_norm": 0.8245175153559692,
+      "learning_rate": 9.893409609859221e-06,
+      "loss": 0.7033,
       "step": 86
     },
     {
-      "epoch": 0.48066298342541436,
-      "grad_norm": 0.684436413306382,
-      "learning_rate": 9.89427842547679e-06,
-      "loss": 0.6908,
+      "epoch": 0.48154981549815495,
+      "grad_norm": 0.7026182893077407,
+      "learning_rate": 9.886669291253178e-06,
+      "loss": 0.6886,
       "step": 87
     },
     {
-      "epoch": 0.4861878453038674,
-      "grad_norm": 0.7492065837893762,
-      "learning_rate": 9.88759283862006e-06,
-      "loss": 0.6951,
+      "epoch": 0.4870848708487085,
+      "grad_norm": 0.7364819802578174,
+      "learning_rate": 9.879724780684518e-06,
+      "loss": 0.6942,
       "step": 88
     },
     {
-      "epoch": 0.49171270718232046,
-      "grad_norm": 0.815556456770822,
-      "learning_rate": 9.880704691794608e-06,
-      "loss": 0.6829,
+      "epoch": 0.492619926199262,
+      "grad_norm": 0.7981620114825012,
+      "learning_rate": 9.872576368333152e-06,
+      "loss": 0.6716,
       "step": 89
     },
     {
-      "epoch": 0.4972375690607735,
-      "grad_norm": 0.848254246152851,
-      "learning_rate": 9.873614270470778e-06,
-      "loss": 0.704,
+      "epoch": 0.4981549815498155,
+      "grad_norm": 0.8836384710800025,
+      "learning_rate": 9.86522435289912e-06,
+      "loss": 0.7237,
       "step": 90
     },
     {
-      "epoch": 0.5027624309392266,
-      "grad_norm": 0.7783965679247608,
-      "learning_rate": 9.866321868501914e-06,
-      "loss": 0.7535,
+      "epoch": 0.503690036900369,
+      "grad_norm": 0.7914423436808887,
+      "learning_rate": 9.857669041590135e-06,
+      "loss": 0.7616,
       "step": 91
     },
     {
-      "epoch": 0.5082872928176796,
-      "grad_norm": 0.7186199448098722,
-      "learning_rate": 9.858827788112195e-06,
-      "loss": 0.6771,
+      "epoch": 0.5092250922509225,
+      "grad_norm": 0.7232583216067142,
+      "learning_rate": 9.849910750108718e-06,
+      "loss": 0.6715,
       "step": 92
     },
     {
-      "epoch": 0.5138121546961326,
-      "grad_norm": 0.8382444185720348,
-      "learning_rate": 9.851132339884097e-06,
-      "loss": 0.6821,
+      "epoch": 0.514760147601476,
+      "grad_norm": 0.8941702082274658,
+      "learning_rate": 9.841949802639031e-06,
+      "loss": 0.6854,
       "step": 93
     },
     {
-      "epoch": 0.5193370165745856,
-      "grad_norm": 0.7135375433267422,
-      "learning_rate": 9.843235842745527e-06,
-      "loss": 0.6985,
+      "epoch": 0.5202952029520295,
+      "grad_norm": 0.7537078088279118,
+      "learning_rate": 9.833786531833311e-06,
+      "loss": 0.6934,
       "step": 94
     },
     {
-      "epoch": 0.5248618784530387,
-      "grad_norm": 0.7467452705314399,
-      "learning_rate": 9.835138623956603e-06,
-      "loss": 0.6875,
+      "epoch": 0.525830258302583,
+      "grad_norm": 0.7496363805903822,
+      "learning_rate": 9.825421278797984e-06,
+      "loss": 0.6757,
       "step": 95
     },
     {
-      "epoch": 0.5303867403314917,
-      "grad_norm": 0.6945947858932874,
-      "learning_rate": 9.826841019096095e-06,
-      "loss": 0.6488,
+      "epoch": 0.5313653136531366,
+      "grad_norm": 0.7811094162416035,
+      "learning_rate": 9.816854393079402e-06,
+      "loss": 0.6575,
       "step": 96
     },
     {
-      "epoch": 0.5359116022099447,
-      "grad_norm": 0.7817121402690996,
-      "learning_rate": 9.818343372047509e-06,
-      "loss": 0.6749,
+      "epoch": 0.5369003690036901,
+      "grad_norm": 0.80064135608355,
+      "learning_rate": 9.808086232649246e-06,
+      "loss": 0.6753,
       "step": 97
     },
     {
-      "epoch": 0.5414364640883977,
-      "grad_norm": 0.7141735093592115,
-      "learning_rate": 9.80964603498485e-06,
-      "loss": 0.6468,
+      "epoch": 0.5424354243542435,
+      "grad_norm": 0.7536916723498005,
+      "learning_rate": 9.79911716388956e-06,
+      "loss": 0.6473,
       "step": 98
     },
     {
-      "epoch": 0.5469613259668509,
-      "grad_norm": 0.7962920109310585,
-      "learning_rate": 9.80074936835801e-06,
-      "loss": 0.7068,
+      "epoch": 0.5479704797047971,
+      "grad_norm": 0.8008444136665656,
+      "learning_rate": 9.789947561577445e-06,
+      "loss": 0.6942,
       "step": 99
     },
     {
-      "epoch": 0.5524861878453039,
-      "grad_norm": 0.7317304477983744,
-      "learning_rate": 9.79165374087784e-06,
-      "loss": 0.7422,
+      "epoch": 0.5535055350553506,
+      "grad_norm": 0.7175301808160046,
+      "learning_rate": 9.7805778088694e-06,
+      "loss": 0.7231,
       "step": 100
     },
     {
-      "epoch": 0.5580110497237569,
-      "grad_norm": 0.7000495326190369,
-      "learning_rate": 9.782359529500867e-06,
-      "loss": 0.6541,
+      "epoch": 0.559040590405904,
+      "grad_norm": 0.6865056427054099,
+      "learning_rate": 9.771008297285307e-06,
+      "loss": 0.6522,
       "step": 101
     },
     {
-      "epoch": 0.56353591160221,
-      "grad_norm": 0.669926861217796,
-      "learning_rate": 9.772867119413667e-06,
-      "loss": 0.6509,
+      "epoch": 0.5645756457564576,
+      "grad_norm": 0.6945540437205031,
+      "learning_rate": 9.761239426692077e-06,
+      "loss": 0.659,
       "step": 102
     },
     {
-      "epoch": 0.569060773480663,
-      "grad_norm": 0.7233355042750117,
-      "learning_rate": 9.763176904016914e-06,
-      "loss": 0.7281,
+      "epoch": 0.5701107011070111,
+      "grad_norm": 0.7631536227470039,
+      "learning_rate": 9.75127160528694e-06,
+      "loss": 0.7251,
       "step": 103
     },
     {
-      "epoch": 0.574585635359116,
-      "grad_norm": 0.7297609103522801,
-      "learning_rate": 9.753289284909058e-06,
-      "loss": 0.6853,
+      "epoch": 0.5756457564575646,
+      "grad_norm": 0.7528731350152382,
+      "learning_rate": 9.741105249580383e-06,
+      "loss": 0.6988,
       "step": 104
     },
     {
-      "epoch": 0.580110497237569,
-      "grad_norm": 0.7837111907679791,
-      "learning_rate": 9.743204671869694e-06,
-      "loss": 0.7378,
+      "epoch": 0.5811808118081181,
+      "grad_norm": 0.7900410030089848,
+      "learning_rate": 9.730740784378755e-06,
+      "loss": 0.7346,
       "step": 105
     },
     {
-      "epoch": 0.585635359116022,
-      "grad_norm": 0.695348976767688,
-      "learning_rate": 9.73292348284258e-06,
-      "loss": 0.6548,
+      "epoch": 0.5867158671586716,
+      "grad_norm": 0.7179179917128015,
+      "learning_rate": 9.7201786427665e-06,
+      "loss": 0.6467,
       "step": 106
     },
     {
-      "epoch": 0.5911602209944752,
-      "grad_norm": 0.7062726297235884,
-      "learning_rate": 9.722446143918307e-06,
-      "loss": 0.7006,
+      "epoch": 0.5922509225092251,
+      "grad_norm": 0.754739804813151,
+      "learning_rate": 9.709419266088086e-06,
+      "loss": 0.7126,
       "step": 107
     },
     {
-      "epoch": 0.5966850828729282,
-      "grad_norm": 0.6884204999018269,
-      "learning_rate": 9.711773089316645e-06,
-      "loss": 0.6197,
+      "epoch": 0.5977859778597786,
+      "grad_norm": 0.7167700794769842,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.6259,
       "step": 108
     },
     {
-      "epoch": 0.6022099447513812,
-      "grad_norm": 0.7974626566246549,
-      "learning_rate": 9.70090476136855e-06,
-      "loss": 0.7452,
+      "epoch": 0.6033210332103321,
+      "grad_norm": 0.862393241288747,
+      "learning_rate": 9.687310614099676e-06,
+      "loss": 0.7462,
       "step": 109
     },
     {
-      "epoch": 0.6077348066298343,
-      "grad_norm": 0.7519249005086731,
-      "learning_rate": 9.689841610497828e-06,
-      "loss": 0.687,
+      "epoch": 0.6088560885608856,
+      "grad_norm": 0.7999243808596918,
+      "learning_rate": 9.67596226261095e-06,
+      "loss": 0.6879,
       "step": 110
     },
     {
-      "epoch": 0.6132596685082873,
-      "grad_norm": 0.7530757425183137,
-      "learning_rate": 9.678584095202468e-06,
-      "loss": 0.7223,
+      "epoch": 0.6143911439114391,
+      "grad_norm": 0.7608890684314967,
+      "learning_rate": 9.664418523660004e-06,
+      "loss": 0.7107,
       "step": 111
     },
     {
-      "epoch": 0.6187845303867403,
-      "grad_norm": 0.8292920525912827,
-      "learning_rate": 9.667132682035646e-06,
-      "loss": 0.6301,
+      "epoch": 0.6199261992619927,
+      "grad_norm": 0.8642130795032539,
+      "learning_rate": 9.652679879607843e-06,
+      "loss": 0.6217,
       "step": 112
     },
     {
-      "epoch": 0.6243093922651933,
-      "grad_norm": 0.7985728833994651,
-      "learning_rate": 9.655487845586378e-06,
-      "loss": 0.6117,
+      "epoch": 0.6254612546125461,
+      "grad_norm": 0.8331393122517589,
+      "learning_rate": 9.640746820959684e-06,
+      "loss": 0.621,
       "step": 113
     },
     {
-      "epoch": 0.6298342541436464,
-      "grad_norm": 0.675379570131735,
-      "learning_rate": 9.643650068459863e-06,
-      "loss": 0.5902,
+      "epoch": 0.6309963099630996,
+      "grad_norm": 0.6800899888090565,
+      "learning_rate": 9.628619846344453e-06,
+      "loss": 0.5866,
       "step": 114
     },
     {
-      "epoch": 0.6353591160220995,
-      "grad_norm": 0.7621705816706639,
-      "learning_rate": 9.631619841257477e-06,
-      "loss": 0.6144,
+      "epoch": 0.6365313653136532,
+      "grad_norm": 0.7683010600627519,
+      "learning_rate": 9.616299462493952e-06,
+      "loss": 0.6161,
       "step": 115
     },
     {
-      "epoch": 0.6408839779005525,
-      "grad_norm": 0.7372229704093334,
-      "learning_rate": 9.619397662556434e-06,
-      "loss": 0.6444,
+      "epoch": 0.6420664206642066,
+      "grad_norm": 0.8092930286332166,
+      "learning_rate": 9.603786184221693e-06,
+      "loss": 0.652,
       "step": 116
     },
     {
-      "epoch": 0.6464088397790055,
-      "grad_norm": 0.7171400121498231,
-      "learning_rate": 9.606984038889142e-06,
-      "loss": 0.6478,
+      "epoch": 0.6476014760147601,
+      "grad_norm": 0.75053691727633,
+      "learning_rate": 9.591080534401371e-06,
+      "loss": 0.6423,
       "step": 117
     },
     {
-      "epoch": 0.6519337016574586,
-      "grad_norm": 0.7162391776467386,
-      "learning_rate": 9.594379484722185e-06,
-      "loss": 0.6368,
+      "epoch": 0.6531365313653137,
+      "grad_norm": 0.713112544635858,
+      "learning_rate": 9.578183043945031e-06,
+      "loss": 0.6363,
       "step": 118
     },
     {
-      "epoch": 0.6574585635359116,
-      "grad_norm": 0.8119977538487404,
-      "learning_rate": 9.581584522435025e-06,
-      "loss": 0.6971,
+      "epoch": 0.6586715867158671,
+      "grad_norm": 0.8815831434837667,
+      "learning_rate": 9.565094251780872e-06,
+      "loss": 0.7031,
       "step": 119
     },
     {
-      "epoch": 0.6629834254143646,
-      "grad_norm": 0.7697231468364938,
-      "learning_rate": 9.568599682298337e-06,
-      "loss": 0.6323,
+      "epoch": 0.6642066420664207,
+      "grad_norm": 0.8001162241305896,
+      "learning_rate": 9.551814704830734e-06,
+      "loss": 0.6202,
       "step": 120
     },
     {
-      "epoch": 0.6685082872928176,
-      "grad_norm": 0.7222511824646166,
-      "learning_rate": 9.555425502452038e-06,
-      "loss": 0.7068,
+      "epoch": 0.6697416974169742,
+      "grad_norm": 0.7135742572247368,
+      "learning_rate": 9.538344957987245e-06,
+      "loss": 0.704,
       "step": 121
     },
     {
-      "epoch": 0.6740331491712708,
-      "grad_norm": 0.7059581814243575,
-      "learning_rate": 9.542062528882989e-06,
-      "loss": 0.6758,
+      "epoch": 0.6752767527675276,
+      "grad_norm": 0.7687650494748194,
+      "learning_rate": 9.524685574090627e-06,
+      "loss": 0.6892,
       "step": 122
     },
     {
-      "epoch": 0.6795580110497238,
-      "grad_norm": 0.7911262183983974,
-      "learning_rate": 9.528511315402358e-06,
-      "loss": 0.6844,
+      "epoch": 0.6808118081180812,
+      "grad_norm": 0.8827181678396219,
+      "learning_rate": 9.51083712390519e-06,
+      "loss": 0.685,
       "step": 123
     },
     {
-      "epoch": 0.6850828729281768,
-      "grad_norm": 0.9567455662688182,
-      "learning_rate": 9.514772423622675e-06,
-      "loss": 0.7468,
+      "epoch": 0.6863468634686347,
+      "grad_norm": 0.8882151233390665,
+      "learning_rate": 9.496800186095466e-06,
+      "loss": 0.7506,
       "step": 124
     },
     {
-      "epoch": 0.6906077348066298,
-      "grad_norm": 0.8075170793709282,
-      "learning_rate": 9.500846422934557e-06,
-      "loss": 0.6884,
+      "epoch": 0.6918819188191881,
+      "grad_norm": 0.7930642149741711,
+      "learning_rate": 9.482575347202047e-06,
+      "loss": 0.6835,
       "step": 125
     },
     {
-      "epoch": 0.6961325966850829,
-      "grad_norm": 0.6617434418531425,
-      "learning_rate": 9.4867338904831e-06,
-      "loss": 0.6807,
+      "epoch": 0.6974169741697417,
+      "grad_norm": 0.6933167090473802,
+      "learning_rate": 9.468163201617063e-06,
+      "loss": 0.6992,
       "step": 126
     },
     {
-      "epoch": 0.7016574585635359,
-      "grad_norm": 0.6852663876058741,
-      "learning_rate": 9.472435411143979e-06,
-      "loss": 0.631,
+      "epoch": 0.7029520295202952,
+      "grad_norm": 0.715214953281436,
+      "learning_rate": 9.453564351559348e-06,
+      "loss": 0.6442,
       "step": 127
     },
     {
-      "epoch": 0.7071823204419889,
-      "grad_norm": 0.8062476288399656,
-      "learning_rate": 9.457951577499187e-06,
-      "loss": 0.6678,
+      "epoch": 0.7084870848708487,
+      "grad_norm": 0.7554434464632439,
+      "learning_rate": 9.438779407049282e-06,
+      "loss": 0.6666,
       "step": 128
     },
     {
-      "epoch": 0.712707182320442,
-      "grad_norm": 0.6969129316123458,
-      "learning_rate": 9.443282989812494e-06,
-      "loss": 0.6413,
+      "epoch": 0.7140221402214022,
+      "grad_norm": 0.7346261351044595,
+      "learning_rate": 9.423808985883289e-06,
+      "loss": 0.6314,
       "step": 129
     },
     {
-      "epoch": 0.7182320441988951,
-      "grad_norm": 0.7139049384459147,
-      "learning_rate": 9.428430256004558e-06,
-      "loss": 0.6523,
+      "epoch": 0.7195571955719557,
+      "grad_norm": 0.7394564505170546,
+      "learning_rate": 9.40865371360804e-06,
+      "loss": 0.6532,
       "step": 130
     },
     {
-      "epoch": 0.7237569060773481,
-      "grad_norm": 0.7008590174042524,
-      "learning_rate": 9.413393991627737e-06,
-      "loss": 0.6826,
+      "epoch": 0.7250922509225092,
+      "grad_norm": 0.6920823973121119,
+      "learning_rate": 9.393314223494297e-06,
+      "loss": 0.6871,
       "step": 131
     },
     {
-      "epoch": 0.7292817679558011,
-      "grad_norm": 0.8833945053457845,
-      "learning_rate": 9.398174819840577e-06,
-      "loss": 0.6405,
+      "epoch": 0.7306273062730627,
+      "grad_norm": 0.8019568716365827,
+      "learning_rate": 9.377791156510456e-06,
+      "loss": 0.6313,
       "step": 132
     },
     {
-      "epoch": 0.7348066298342542,
-      "grad_norm": 0.7572193126349432,
-      "learning_rate": 9.382773371381986e-06,
-      "loss": 0.6689,
+      "epoch": 0.7361623616236163,
+      "grad_norm": 0.731865215320956,
+      "learning_rate": 9.362085161295768e-06,
+      "loss": 0.6757,
       "step": 133
     },
     {
-      "epoch": 0.7403314917127072,
-      "grad_norm": 0.7524950122671795,
-      "learning_rate": 9.367190284545087e-06,
-      "loss": 0.7109,
+      "epoch": 0.7416974169741697,
+      "grad_norm": 0.7331269825104397,
+      "learning_rate": 9.346196894133239e-06,
+      "loss": 0.7192,
       "step": 134
     },
     {
-      "epoch": 0.7458563535911602,
-      "grad_norm": 0.6769451685165733,
-      "learning_rate": 9.351426205150778e-06,
-      "loss": 0.66,
+      "epoch": 0.7472324723247232,
+      "grad_norm": 0.6689559430615435,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 0.6578,
       "step": 135
     },
     {
-      "epoch": 0.7513812154696132,
-      "grad_norm": 0.790773238492282,
-      "learning_rate": 9.335481786520955e-06,
-      "loss": 0.7216,
+      "epoch": 0.7527675276752768,
+      "grad_norm": 0.8011088236436561,
+      "learning_rate": 9.313876207150544e-06,
+      "loss": 0.7185,
       "step": 136
     },
     {
-      "epoch": 0.7569060773480663,
-      "grad_norm": 0.7100066068783747,
-      "learning_rate": 9.319357689451444e-06,
-      "loss": 0.6562,
+      "epoch": 0.7583025830258303,
+      "grad_norm": 0.7497062827280815,
+      "learning_rate": 9.297445137866726e-06,
+      "loss": 0.6652,
       "step": 137
     },
     {
-      "epoch": 0.7624309392265194,
-      "grad_norm": 0.7853719538210431,
-      "learning_rate": 9.30305458218461e-06,
-      "loss": 0.6882,
+      "epoch": 0.7638376383763837,
+      "grad_norm": 0.7999274641250289,
+      "learning_rate": 9.280834497651334e-06,
+      "loss": 0.6987,
       "step": 138
     },
     {
-      "epoch": 0.7679558011049724,
-      "grad_norm": 0.6638407553470856,
-      "learning_rate": 9.286573140381663e-06,
-      "loss": 0.6541,
+      "epoch": 0.7693726937269373,
+      "grad_norm": 0.6443825206939584,
+      "learning_rate": 9.264044980588415e-06,
+      "loss": 0.6537,
       "step": 139
     },
     {
-      "epoch": 0.7734806629834254,
-      "grad_norm": 0.6834239770243532,
-      "learning_rate": 9.26991404709466e-06,
-      "loss": 0.6289,
+      "epoch": 0.7749077490774908,
+      "grad_norm": 0.6830543309432857,
+      "learning_rate": 9.247077288236488e-06,
+      "loss": 0.6389,
       "step": 140
     },
     {
-      "epoch": 0.7790055248618785,
-      "grad_norm": 0.8713388179210387,
-      "learning_rate": 9.253077992738193e-06,
-      "loss": 0.73,
+      "epoch": 0.7804428044280443,
+      "grad_norm": 0.7912659097565806,
+      "learning_rate": 9.229932129599206e-06,
+      "loss": 0.7265,
       "step": 141
     },
     {
-      "epoch": 0.7845303867403315,
-      "grad_norm": 0.7629447190554447,
-      "learning_rate": 9.236065675060775e-06,
-      "loss": 0.693,
+      "epoch": 0.7859778597785978,
+      "grad_norm": 0.7608099518203012,
+      "learning_rate": 9.212610221095748e-06,
+      "loss": 0.6902,
       "step": 142
     },
     {
-      "epoch": 0.7900552486187845,
-      "grad_norm": 0.7544036306635357,
-      "learning_rate": 9.218877799115929e-06,
-      "loss": 0.7252,
+      "epoch": 0.7915129151291513,
+      "grad_norm": 0.7598999658417419,
+      "learning_rate": 9.195112286530874e-06,
+      "loss": 0.7237,
       "step": 143
     },
     {
-      "epoch": 0.7955801104972375,
-      "grad_norm": 0.6847405196212653,
-      "learning_rate": 9.201515077232958e-06,
-      "loss": 0.703,
+      "epoch": 0.7970479704797048,
+      "grad_norm": 0.6790407377272779,
+      "learning_rate": 9.177439057064684e-06,
+      "loss": 0.689,
       "step": 144
     },
     {
-      "epoch": 0.8011049723756906,
-      "grad_norm": 0.7322209407647546,
-      "learning_rate": 9.183978228987436e-06,
-      "loss": 0.6511,
+      "epoch": 0.8025830258302583,
+      "grad_norm": 0.7024267843652401,
+      "learning_rate": 9.159591271182058e-06,
+      "loss": 0.6356,
       "step": 145
     },
     {
-      "epoch": 0.8066298342541437,
-      "grad_norm": 0.7306655102140895,
-      "learning_rate": 9.166267981171369e-06,
-      "loss": 0.6477,
+      "epoch": 0.8081180811808119,
+      "grad_norm": 0.704433084066074,
+      "learning_rate": 9.141569674661816e-06,
+      "loss": 0.6453,
       "step": 146
     },
     {
-      "epoch": 0.8121546961325967,
-      "grad_norm": 0.7206398058794518,
-      "learning_rate": 9.148385067763094e-06,
-      "loss": 0.6407,
+      "epoch": 0.8136531365313653,
+      "grad_norm": 0.7169550452222242,
+      "learning_rate": 9.123375020545534e-06,
+      "loss": 0.6373,
       "step": 147
     },
     {
-      "epoch": 0.8176795580110497,
-      "grad_norm": 0.7036158657538799,
-      "learning_rate": 9.130330229896846e-06,
-      "loss": 0.6096,
+      "epoch": 0.8191881918819188,
+      "grad_norm": 0.694690233125659,
+      "learning_rate": 9.105008069106093e-06,
+      "loss": 0.6091,
       "step": 148
     },
     {
-      "epoch": 0.8232044198895028,
-      "grad_norm": 0.7807874970263721,
-      "learning_rate": 9.112104215832047e-06,
-      "loss": 0.658,
+      "epoch": 0.8247232472324724,
+      "grad_norm": 0.7335227368162949,
+      "learning_rate": 9.086469587815904e-06,
+      "loss": 0.6582,
       "step": 149
     },
     {
-      "epoch": 0.8287292817679558,
-      "grad_norm": 0.6887926406142375,
-      "learning_rate": 9.093707780922293e-06,
-      "loss": 0.6985,
+      "epoch": 0.8302583025830258,
+      "grad_norm": 0.6841111000042602,
+      "learning_rate": 9.067760351314838e-06,
+      "loss": 0.7072,
       "step": 150
     },
     {
-      "epoch": 0.8342541436464088,
-      "grad_norm": 0.7163709674127419,
-      "learning_rate": 9.075141687584056e-06,
-      "loss": 0.6801,
+      "epoch": 0.8357933579335793,
+      "grad_norm": 0.7083136755269921,
+      "learning_rate": 9.048881141377863e-06,
+      "loss": 0.6677,
       "step": 151
     },
     {
-      "epoch": 0.8397790055248618,
-      "grad_norm": 0.6837817476434366,
-      "learning_rate": 9.056406705265084e-06,
-      "loss": 0.6442,
+      "epoch": 0.8413284132841329,
+      "grad_norm": 0.6982890610414855,
+      "learning_rate": 9.029832746882372e-06,
+      "loss": 0.6526,
       "step": 152
     },
     {
-      "epoch": 0.8453038674033149,
-      "grad_norm": 0.6843857902663828,
-      "learning_rate": 9.037503610412502e-06,
-      "loss": 0.6446,
+      "epoch": 0.8468634686346863,
+      "grad_norm": 0.6860196076239153,
+      "learning_rate": 9.01061596377522e-06,
+      "loss": 0.6461,
       "step": 153
     },
     {
-      "epoch": 0.850828729281768,
-      "grad_norm": 0.7547759124038085,
-      "learning_rate": 9.018433186440648e-06,
-      "loss": 0.7106,
+      "epoch": 0.8523985239852399,
+      "grad_norm": 0.7376158689862853,
+      "learning_rate": 8.991231595039464e-06,
+      "loss": 0.7101,
       "step": 154
     },
     {
-      "epoch": 0.856353591160221,
-      "grad_norm": 0.6923617352622536,
-      "learning_rate": 8.999196223698599e-06,
-      "loss": 0.7418,
+      "epoch": 0.8579335793357934,
+      "grad_norm": 0.6975228556068289,
+      "learning_rate": 8.97168045066082e-06,
+      "loss": 0.7475,
       "step": 155
     },
     {
-      "epoch": 0.861878453038674,
-      "grad_norm": 0.7589388571367478,
-      "learning_rate": 8.979793519437413e-06,
-      "loss": 0.7064,
+      "epoch": 0.8634686346863468,
+      "grad_norm": 0.7486491127857194,
+      "learning_rate": 8.951963347593797e-06,
+      "loss": 0.6996,
       "step": 156
     },
     {
-      "epoch": 0.8674033149171271,
-      "grad_norm": 0.8111254921639891,
-      "learning_rate": 8.960225877777095e-06,
-      "loss": 0.7153,
+      "epoch": 0.8690036900369004,
+      "grad_norm": 0.7912254693906845,
+      "learning_rate": 8.932081109727582e-06,
+      "loss": 0.6985,
       "step": 157
     },
     {
-      "epoch": 0.8729281767955801,
-      "grad_norm": 0.6405807331541313,
-      "learning_rate": 8.940494109673266e-06,
-      "loss": 0.6121,
+      "epoch": 0.8745387453874539,
+      "grad_norm": 0.6439570726845703,
+      "learning_rate": 8.9120345678516e-06,
+      "loss": 0.625,
       "step": 158
     },
     {
-      "epoch": 0.8784530386740331,
-      "grad_norm": 0.6742570632281877,
-      "learning_rate": 8.920599032883553e-06,
-      "loss": 0.6781,
+      "epoch": 0.8800738007380073,
+      "grad_norm": 0.6572694602060432,
+      "learning_rate": 8.891824559620801e-06,
+      "loss": 0.6784,
       "step": 159
     },
     {
-      "epoch": 0.8839779005524862,
-      "grad_norm": 0.7843439776633434,
-      "learning_rate": 8.900541471933703e-06,
-      "loss": 0.6979,
+      "epoch": 0.8856088560885609,
+      "grad_norm": 0.7971159900336966,
+      "learning_rate": 8.871451929520662e-06,
+      "loss": 0.6988,
       "step": 160
     },
     {
-      "epoch": 0.8895027624309392,
-      "grad_norm": 0.725960679493056,
-      "learning_rate": 8.880322258083408e-06,
-      "loss": 0.6557,
+      "epoch": 0.8911439114391144,
+      "grad_norm": 0.7050522383893928,
+      "learning_rate": 8.8509175288319e-06,
+      "loss": 0.6445,
       "step": 161
     },
     {
-      "epoch": 0.8950276243093923,
-      "grad_norm": 0.6898678148433862,
-      "learning_rate": 8.859942229291856e-06,
-      "loss": 0.662,
+      "epoch": 0.8966789667896679,
+      "grad_norm": 0.6897534640630667,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.6698,
       "step": 162
     },
     {
-      "epoch": 0.9005524861878453,
-      "grad_norm": 0.7710968128854336,
-      "learning_rate": 8.839402230183e-06,
-      "loss": 0.662,
+      "epoch": 0.9022140221402214,
+      "grad_norm": 0.7286182755253351,
+      "learning_rate": 8.80936685457383e-06,
+      "loss": 0.6527,
       "step": 163
     },
     {
-      "epoch": 0.9060773480662984,
-      "grad_norm": 0.741382073637584,
-      "learning_rate": 8.818703112010562e-06,
-      "loss": 0.6935,
+      "epoch": 0.9077490774907749,
+      "grad_norm": 0.7250443386250167,
+      "learning_rate": 8.78835231722059e-06,
+      "loss": 0.6785,
       "step": 164
     },
     {
-      "epoch": 0.9116022099447514,
-      "grad_norm": 0.6831369278396727,
-      "learning_rate": 8.797845732622742e-06,
-      "loss": 0.7031,
+      "epoch": 0.9132841328413284,
+      "grad_norm": 0.6795202817087662,
+      "learning_rate": 8.767179481638303e-06,
+      "loss": 0.712,
       "step": 165
     },
     {
-      "epoch": 0.9171270718232044,
-      "grad_norm": 0.7492006743864014,
-      "learning_rate": 8.776830956426674e-06,
-      "loss": 0.708,
+      "epoch": 0.9188191881918819,
+      "grad_norm": 0.7673545156991856,
+      "learning_rate": 8.74584923254468e-06,
+      "loss": 0.7073,
       "step": 166
     },
     {
-      "epoch": 0.9226519337016574,
-      "grad_norm": 0.7613459925623692,
-      "learning_rate": 8.755659654352599e-06,
-      "loss": 0.7169,
+      "epoch": 0.9243542435424354,
+      "grad_norm": 0.7100681832870371,
+      "learning_rate": 8.72436246123503e-06,
+      "loss": 0.7097,
       "step": 167
     },
     {
-      "epoch": 0.9281767955801105,
-      "grad_norm": 0.6791670574304287,
-      "learning_rate": 8.734332703817771e-06,
-      "loss": 0.6991,
+      "epoch": 0.9298892988929889,
+      "grad_norm": 0.6750568200393077,
+      "learning_rate": 8.702720065545024e-06,
+      "loss": 0.6976,
       "step": 168
     },
     {
-      "epoch": 0.9337016574585635,
-      "grad_norm": 0.6681861340818701,
-      "learning_rate": 8.712850988690094e-06,
-      "loss": 0.7032,
+      "epoch": 0.9354243542435424,
+      "grad_norm": 0.6760810557529758,
+      "learning_rate": 8.680922949813177e-06,
+      "loss": 0.6883,
       "step": 169
     },
     {
-      "epoch": 0.9392265193370166,
-      "grad_norm": 0.6569011217941949,
-      "learning_rate": 8.691215399251489e-06,
-      "loss": 0.6439,
+      "epoch": 0.940959409594096,
+      "grad_norm": 0.6506834973761045,
+      "learning_rate": 8.658972024843063e-06,
+      "loss": 0.6434,
       "step": 170
     },
     {
-      "epoch": 0.9447513812154696,
-      "grad_norm": 0.7086161196938886,
-      "learning_rate": 8.669426832160997e-06,
-      "loss": 0.6193,
+      "epoch": 0.9464944649446494,
+      "grad_norm": 0.668877599120538,
+      "learning_rate": 8.636868207865244e-06,
+      "loss": 0.6198,
       "step": 171
     },
     {
-      "epoch": 0.9502762430939227,
-      "grad_norm": 0.654556823801651,
-      "learning_rate": 8.647486190417624e-06,
-      "loss": 0.6929,
+      "epoch": 0.9520295202952029,
+      "grad_norm": 0.6532330641471769,
+      "learning_rate": 8.614612422498965e-06,
+      "loss": 0.6806,
       "step": 172
     },
     {
-      "epoch": 0.9558011049723757,
-      "grad_norm": 0.7440213169432797,
-      "learning_rate": 8.625394383322914e-06,
-      "loss": 0.6779,
+      "epoch": 0.9575645756457565,
+      "grad_norm": 0.7178919179692898,
+      "learning_rate": 8.592205598713539e-06,
+      "loss": 0.6887,
       "step": 173
     },
     {
-      "epoch": 0.9613259668508287,
-      "grad_norm": 0.7345541693491946,
-      "learning_rate": 8.603152326443262e-06,
-      "loss": 0.6412,
+      "epoch": 0.9630996309963099,
+      "grad_norm": 0.6999250045861761,
+      "learning_rate": 8.569648672789496e-06,
+      "loss": 0.6442,
       "step": 174
     },
     {
-      "epoch": 0.9668508287292817,
-      "grad_norm": 0.7017136954703048,
-      "learning_rate": 8.580760941571968e-06,
-      "loss": 0.7249,
+      "epoch": 0.9686346863468634,
+      "grad_norm": 0.7075371495307748,
+      "learning_rate": 8.546942587279465e-06,
+      "loss": 0.7256,
       "step": 175
     },
     {
-      "epoch": 0.9723756906077348,
-      "grad_norm": 0.7533884250226044,
-      "learning_rate": 8.55822115669104e-06,
-      "loss": 0.6725,
+      "epoch": 0.974169741697417,
+      "grad_norm": 0.7139921209949371,
+      "learning_rate": 8.524088290968781e-06,
+      "loss": 0.669,
       "step": 176
     },
     {
-      "epoch": 0.9779005524861878,
-      "grad_norm": 0.8635896869993843,
-      "learning_rate": 8.535533905932739e-06,
-      "loss": 0.6778,
+      "epoch": 0.9797047970479705,
+      "grad_norm": 0.7843465216757487,
+      "learning_rate": 8.501086738835843e-06,
+      "loss": 0.679,
       "step": 177
     },
     {
-      "epoch": 0.9834254143646409,
-      "grad_norm": 0.7046176922723474,
-      "learning_rate": 8.512700129540847e-06,
-      "loss": 0.646,
+      "epoch": 0.985239852398524,
+      "grad_norm": 0.7078117075067463,
+      "learning_rate": 8.477938892012209e-06,
+      "loss": 0.6443,
       "step": 178
     },
     {
-      "epoch": 0.988950276243094,
-      "grad_norm": 0.7469626256845687,
-      "learning_rate": 8.489720773831717e-06,
-      "loss": 0.6742,
+      "epoch": 0.9907749077490775,
+      "grad_norm": 0.7191146480817366,
+      "learning_rate": 8.45464571774244e-06,
+      "loss": 0.6752,
       "step": 179
     },
     {
-      "epoch": 0.994475138121547,
-      "grad_norm": 0.8073184194204243,
-      "learning_rate": 8.466596791155055e-06,
-      "loss": 0.6745,
+      "epoch": 0.996309963099631,
+      "grad_norm": 0.7793217733052689,
+      "learning_rate": 8.43120818934367e-06,
+      "loss": 0.6809,
       "step": 180
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 0.6469115640516496,
-      "learning_rate": 8.443329139854434e-06,
-      "loss": 0.635,
+      "epoch": 1.0018450184501846,
+      "grad_norm": 0.9535816156617457,
+      "learning_rate": 8.407627286164948e-06,
+      "loss": 0.8581,
       "step": 181
     },
     {
-      "epoch": 1.0055248618784531,
-      "grad_norm": 0.7680249419301477,
-      "learning_rate": 8.419918784227592e-06,
-      "loss": 0.6638,
+      "epoch": 1.007380073800738,
+      "grad_norm": 0.7904687166212481,
+      "learning_rate": 8.38390399354631e-06,
+      "loss": 0.6589,
       "step": 182
     },
     {
-      "epoch": 1.011049723756906,
-      "grad_norm": 0.7714771127607805,
-      "learning_rate": 8.396366694486466e-06,
-      "loss": 0.6176,
+      "epoch": 1.0129151291512914,
+      "grad_norm": 0.7871113082768205,
+      "learning_rate": 8.360039302777614e-06,
+      "loss": 0.5974,
       "step": 183
     },
     {
-      "epoch": 1.0165745856353592,
-      "grad_norm": 0.6836633389900814,
-      "learning_rate": 8.372673846716977e-06,
-      "loss": 0.5742,
+      "epoch": 1.018450184501845,
+      "grad_norm": 0.7159872827730034,
+      "learning_rate": 8.336034211057098e-06,
+      "loss": 0.6009,
       "step": 184
     },
     {
-      "epoch": 1.022099447513812,
-      "grad_norm": 0.671339247626321,
-      "learning_rate": 8.348841222838579e-06,
-      "loss": 0.5965,
+      "epoch": 1.0239852398523985,
+      "grad_norm": 0.6616682950001487,
+      "learning_rate": 8.31188972144974e-06,
+      "loss": 0.5881,
       "step": 185
     },
     {
-      "epoch": 1.0276243093922652,
-      "grad_norm": 0.6818502606595346,
-      "learning_rate": 8.324869810563573e-06,
-      "loss": 0.6031,
+      "epoch": 1.029520295202952,
+      "grad_norm": 0.6850474409430557,
+      "learning_rate": 8.28760684284532e-06,
+      "loss": 0.5939,
       "step": 186
     },
     {
-      "epoch": 1.0331491712707181,
-      "grad_norm": 0.7679534342017695,
-      "learning_rate": 8.30076060335616e-06,
-      "loss": 0.6378,
+      "epoch": 1.0350553505535056,
+      "grad_norm": 0.7461860778494099,
+      "learning_rate": 8.263186589916273e-06,
+      "loss": 0.6383,
       "step": 187
     },
     {
-      "epoch": 1.0386740331491713,
-      "grad_norm": 0.6610817415773053,
-      "learning_rate": 8.276514600391272e-06,
-      "loss": 0.5828,
+      "epoch": 1.040590405904059,
+      "grad_norm": 0.7089074245925379,
+      "learning_rate": 8.238629983075296e-06,
+      "loss": 0.618,
       "step": 188
     },
     {
-      "epoch": 1.0441988950276242,
-      "grad_norm": 0.7265666171607357,
-      "learning_rate": 8.25213280651317e-06,
-      "loss": 0.5991,
+      "epoch": 1.0461254612546125,
+      "grad_norm": 0.6738261402996001,
+      "learning_rate": 8.213938048432697e-06,
+      "loss": 0.5554,
       "step": 189
     },
     {
-      "epoch": 1.0497237569060773,
-      "grad_norm": 0.7356973203535483,
-      "learning_rate": 8.227616232193794e-06,
-      "loss": 0.5877,
+      "epoch": 1.051660516605166,
+      "grad_norm": 0.7206220611502636,
+      "learning_rate": 8.18911181775353e-06,
+      "loss": 0.6136,
       "step": 190
     },
     {
-      "epoch": 1.0552486187845305,
-      "grad_norm": 0.7865650351374615,
-      "learning_rate": 8.202965893490877e-06,
-      "loss": 0.6476,
+      "epoch": 1.0571955719557196,
+      "grad_norm": 0.7352230176592326,
+      "learning_rate": 8.164152328414476e-06,
+      "loss": 0.6188,
       "step": 191
     },
     {
-      "epoch": 1.0607734806629834,
-      "grad_norm": 0.7194978525901626,
-      "learning_rate": 8.178182812005853e-06,
-      "loss": 0.5147,
+      "epoch": 1.062730627306273,
+      "grad_norm": 0.6943757541874687,
+      "learning_rate": 8.139060623360494e-06,
+      "loss": 0.5167,
       "step": 192
     },
     {
-      "epoch": 1.0662983425414365,
-      "grad_norm": 0.70366807919922,
-      "learning_rate": 8.153268014841507e-06,
-      "loss": 0.6128,
+      "epoch": 1.0682656826568266,
+      "grad_norm": 0.7778412839253838,
+      "learning_rate": 8.113837751061246e-06,
+      "loss": 0.6154,
       "step": 193
     },
     {
-      "epoch": 1.0718232044198894,
-      "grad_norm": 0.6751029207263034,
-      "learning_rate": 8.128222534559406e-06,
-      "loss": 0.5743,
+      "epoch": 1.07380073800738,
+      "grad_norm": 0.6551339832525483,
+      "learning_rate": 8.088484765467286e-06,
+      "loss": 0.5878,
       "step": 194
     },
     {
-      "epoch": 1.0773480662983426,
-      "grad_norm": 0.769584710702697,
-      "learning_rate": 8.103047409137114e-06,
-      "loss": 0.6736,
+      "epoch": 1.0793357933579335,
+      "grad_norm": 0.8149186205476209,
+      "learning_rate": 8.063002725966014e-06,
+      "loss": 0.7048,
       "step": 195
     },
     {
-      "epoch": 1.0828729281767955,
-      "grad_norm": 0.7906391664348668,
-      "learning_rate": 8.07774368192517e-06,
-      "loss": 0.6658,
+      "epoch": 1.084870848708487,
+      "grad_norm": 0.6880593641520117,
+      "learning_rate": 8.037392697337418e-06,
+      "loss": 0.6041,
       "step": 196
     },
     {
-      "epoch": 1.0883977900552486,
-      "grad_norm": 0.6510299748322192,
-      "learning_rate": 8.052312401603848e-06,
-      "loss": 0.5765,
+      "epoch": 1.0904059040590406,
+      "grad_norm": 0.7040695710194661,
+      "learning_rate": 8.011655749709575e-06,
+      "loss": 0.5979,
       "step": 197
     },
     {
-      "epoch": 1.0939226519337018,
-      "grad_norm": 0.8547974979381732,
-      "learning_rate": 8.026754622139691e-06,
-      "loss": 0.5903,
+      "epoch": 1.0959409594095941,
+      "grad_norm": 0.7965545246789066,
+      "learning_rate": 7.985792958513932e-06,
+      "loss": 0.5892,
       "step": 198
     },
     {
-      "epoch": 1.0994475138121547,
-      "grad_norm": 0.8338835673156341,
-      "learning_rate": 8.001071402741843e-06,
-      "loss": 0.6027,
+      "epoch": 1.1014760147601477,
+      "grad_norm": 0.7621766315488933,
+      "learning_rate": 7.95980540444038e-06,
+      "loss": 0.594,
       "step": 199
     },
     {
-      "epoch": 1.1049723756906078,
-      "grad_norm": 0.7681594836262089,
-      "learning_rate": 7.975263807818136e-06,
-      "loss": 0.6428,
+      "epoch": 1.1070110701107012,
+      "grad_norm": 0.7643147205973972,
+      "learning_rate": 7.93369417339209e-06,
+      "loss": 0.6597,
       "step": 200
     },
     {
-      "epoch": 1.1104972375690607,
-      "grad_norm": 0.7716286060753629,
-      "learning_rate": 7.949332906930995e-06,
-      "loss": 0.6502,
+      "epoch": 1.1125461254612545,
+      "grad_norm": 0.8216988466319578,
+      "learning_rate": 7.907460356440133e-06,
+      "loss": 0.6398,
       "step": 201
     },
     {
-      "epoch": 1.1160220994475138,
-      "grad_norm": 0.8404857249117871,
-      "learning_rate": 7.923279774753092e-06,
-      "loss": 0.5865,
+      "epoch": 1.118081180811808,
+      "grad_norm": 0.7718237762248213,
+      "learning_rate": 7.881105049777902e-06,
+      "loss": 0.5818,
       "step": 202
     },
     {
-      "epoch": 1.1215469613259668,
-      "grad_norm": 0.6784714640462896,
-      "learning_rate": 7.897105491022819e-06,
-      "loss": 0.5737,
+      "epoch": 1.1236162361623616,
+      "grad_norm": 0.6351521048330241,
+      "learning_rate": 7.854629354675292e-06,
+      "loss": 0.5774,
       "step": 203
     },
     {
-      "epoch": 1.12707182320442,
-      "grad_norm": 0.6611561272377011,
-      "learning_rate": 7.870811140499543e-06,
-      "loss": 0.5916,
+      "epoch": 1.1291512915129152,
+      "grad_norm": 0.6908992052540243,
+      "learning_rate": 7.828034377432694e-06,
+      "loss": 0.557,
       "step": 204
     },
     {
-      "epoch": 1.132596685082873,
-      "grad_norm": 0.6694407525681161,
-      "learning_rate": 7.844397812918637e-06,
-      "loss": 0.5629,
+      "epoch": 1.1346863468634687,
+      "grad_norm": 0.7232308696829104,
+      "learning_rate": 7.801321229334764e-06,
+      "loss": 0.5865,
       "step": 205
     },
     {
-      "epoch": 1.138121546961326,
-      "grad_norm": 0.6628808525691233,
-      "learning_rate": 7.817866602946326e-06,
-      "loss": 0.6174,
+      "epoch": 1.140221402214022,
+      "grad_norm": 0.6705064929908767,
+      "learning_rate": 7.774491026603985e-06,
+      "loss": 0.5934,
       "step": 206
     },
     {
-      "epoch": 1.143646408839779,
-      "grad_norm": 0.6433613204625658,
-      "learning_rate": 7.791218610134324e-06,
-      "loss": 0.5891,
+      "epoch": 1.1457564575645756,
+      "grad_norm": 0.6880567256154578,
+      "learning_rate": 7.747544890354031e-06,
+      "loss": 0.6281,
       "step": 207
     },
     {
-      "epoch": 1.149171270718232,
-      "grad_norm": 0.7083946072498601,
-      "learning_rate": 7.764454938874252e-06,
-      "loss": 0.5973,
+      "epoch": 1.151291512915129,
+      "grad_norm": 0.7043185315494912,
+      "learning_rate": 7.720483946542913e-06,
+      "loss": 0.5789,
       "step": 208
     },
     {
-      "epoch": 1.1546961325966851,
-      "grad_norm": 0.6920035749512741,
-      "learning_rate": 7.737576698351878e-06,
-      "loss": 0.6368,
+      "epoch": 1.1568265682656826,
+      "grad_norm": 0.749312030144366,
+      "learning_rate": 7.69330932592594e-06,
+      "loss": 0.6413,
       "step": 209
     },
     {
-      "epoch": 1.160220994475138,
-      "grad_norm": 0.77090065904851,
-      "learning_rate": 7.710585002501145e-06,
-      "loss": 0.6458,
+      "epoch": 1.1623616236162362,
+      "grad_norm": 0.8155713214886554,
+      "learning_rate": 7.666022164008458e-06,
+      "loss": 0.6249,
       "step": 210
     },
     {
-      "epoch": 1.1657458563535912,
-      "grad_norm": 0.7373547338916707,
-      "learning_rate": 7.683480969958005e-06,
-      "loss": 0.5674,
+      "epoch": 1.1678966789667897,
+      "grad_norm": 0.736980019422596,
+      "learning_rate": 7.638623600998409e-06,
+      "loss": 0.5835,
       "step": 211
     },
     {
-      "epoch": 1.1712707182320443,
-      "grad_norm": 0.6775847537528304,
-      "learning_rate": 7.656265724014054e-06,
-      "loss": 0.6039,
+      "epoch": 1.1734317343173433,
+      "grad_norm": 0.6987882968304525,
+      "learning_rate": 7.6111147817586925e-06,
+      "loss": 0.6169,
       "step": 212
     },
     {
-      "epoch": 1.1767955801104972,
-      "grad_norm": 0.7080491269295629,
-      "learning_rate": 7.628940392569995e-06,
-      "loss": 0.6395,
+      "epoch": 1.1789667896678966,
+      "grad_norm": 0.7052697317791319,
+      "learning_rate": 7.5834968557593155e-06,
+      "loss": 0.5993,
       "step": 213
     },
     {
-      "epoch": 1.1823204419889504,
-      "grad_norm": 0.6741064097723004,
-      "learning_rate": 7.601506108088874e-06,
-      "loss": 0.5882,
+      "epoch": 1.1845018450184501,
+      "grad_norm": 0.7070482187757918,
+      "learning_rate": 7.5557709770293664e-06,
+      "loss": 0.5973,
       "step": 214
     },
     {
-      "epoch": 1.1878453038674033,
-      "grad_norm": 0.657603362174331,
-      "learning_rate": 7.5739640075491546e-06,
-      "loss": 0.6184,
+      "epoch": 1.1900369003690037,
+      "grad_norm": 0.6930393067005186,
+      "learning_rate": 7.527938304108795e-06,
+      "loss": 0.6242,
       "step": 215
     },
     {
-      "epoch": 1.1933701657458564,
-      "grad_norm": 0.6791394246768295,
-      "learning_rate": 7.546315232397601e-06,
-      "loss": 0.6138,
+      "epoch": 1.1955719557195572,
+      "grad_norm": 0.6851172667466551,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.6435,
       "step": 216
     },
     {
-      "epoch": 1.1988950276243093,
-      "grad_norm": 0.6894132654756862,
-      "learning_rate": 7.518560928501969e-06,
-      "loss": 0.6252,
+      "epoch": 1.2011070110701108,
+      "grad_norm": 0.7259877682356197,
+      "learning_rate": 7.471957232119235e-06,
+      "loss": 0.6212,
       "step": 217
     },
     {
-      "epoch": 1.2044198895027625,
-      "grad_norm": 0.697165533998509,
-      "learning_rate": 7.4907022461035125e-06,
-      "loss": 0.6002,
+      "epoch": 1.2066420664206643,
+      "grad_norm": 0.640453551277484,
+      "learning_rate": 7.443811172247822e-06,
+      "loss": 0.5481,
       "step": 218
     },
     {
-      "epoch": 1.2099447513812154,
-      "grad_norm": 0.6444399979671767,
-      "learning_rate": 7.462740339769323e-06,
-      "loss": 0.5868,
+      "epoch": 1.2121771217712176,
+      "grad_norm": 0.6922785890990849,
+      "learning_rate": 7.415562996483193e-06,
+      "loss": 0.6075,
       "step": 219
     },
     {
-      "epoch": 1.2154696132596685,
-      "grad_norm": 0.7259698993482928,
-      "learning_rate": 7.434676368344469e-06,
-      "loss": 0.655,
+      "epoch": 1.2177121771217712,
+      "grad_norm": 0.7252156376468682,
+      "learning_rate": 7.387213885189746e-06,
+      "loss": 0.6357,
       "step": 220
     },
     {
-      "epoch": 1.2209944751381214,
-      "grad_norm": 0.7245469567674288,
-      "learning_rate": 7.406511494903982e-06,
-      "loss": 0.5975,
+      "epoch": 1.2232472324723247,
+      "grad_norm": 0.6646431209628268,
+      "learning_rate": 7.358765022949519e-06,
+      "loss": 0.6337,
       "step": 221
     },
     {
-      "epoch": 1.2265193370165746,
-      "grad_norm": 0.6743396575371344,
-      "learning_rate": 7.378246886704638e-06,
-      "loss": 0.6729,
+      "epoch": 1.2287822878228782,
+      "grad_norm": 0.6874785057119748,
+      "learning_rate": 7.330217598512696e-06,
+      "loss": 0.6359,
       "step": 222
     },
     {
-      "epoch": 1.2320441988950277,
-      "grad_norm": 0.7206318344181539,
-      "learning_rate": 7.349883715136601e-06,
-      "loss": 0.6123,
+      "epoch": 1.2343173431734318,
+      "grad_norm": 0.7165999600358128,
+      "learning_rate": 7.30157280474793e-06,
+      "loss": 0.5538,
       "step": 223
     },
     {
-      "epoch": 1.2375690607734806,
-      "grad_norm": 0.602101018107306,
-      "learning_rate": 7.321423155674858e-06,
-      "loss": 0.5287,
+      "epoch": 1.2398523985239853,
+      "grad_norm": 0.6714722516214744,
+      "learning_rate": 7.2728318385925035e-06,
+      "loss": 0.6,
       "step": 224
     },
     {
-      "epoch": 1.2430939226519337,
-      "grad_norm": 0.6466479963244331,
-      "learning_rate": 7.292866387830515e-06,
-      "loss": 0.5927,
+      "epoch": 1.2453874538745389,
+      "grad_norm": 0.7107004304709478,
+      "learning_rate": 7.243995901002312e-06,
+      "loss": 0.6217,
       "step": 225
     },
     {
-      "epoch": 1.2486187845303867,
-      "grad_norm": 0.7693210833084159,
-      "learning_rate": 7.264214595101913e-06,
-      "loss": 0.6321,
+      "epoch": 1.2509225092250922,
+      "grad_norm": 0.7303659741053162,
+      "learning_rate": 7.215066196901676e-06,
+      "loss": 0.6022,
       "step": 226
     },
     {
-      "epoch": 1.2541436464088398,
-      "grad_norm": 0.6789133468563491,
-      "learning_rate": 7.235468964925571e-06,
-      "loss": 0.5794,
+      "epoch": 1.2564575645756457,
+      "grad_norm": 0.6912084957079973,
+      "learning_rate": 7.186043935133005e-06,
+      "loss": 0.6017,
       "step": 227
     },
     {
-      "epoch": 1.2596685082872927,
-      "grad_norm": 0.7739752082859613,
-      "learning_rate": 7.206630688626981e-06,
-      "loss": 0.6563,
+      "epoch": 1.2619926199261993,
+      "grad_norm": 0.8016568540326536,
+      "learning_rate": 7.156930328406268e-06,
+      "loss": 0.6542,
       "step": 228
     },
     {
-      "epoch": 1.2651933701657458,
-      "grad_norm": 0.7370007253011229,
-      "learning_rate": 7.177700961371239e-06,
-      "loss": 0.6075,
+      "epoch": 1.2675276752767528,
+      "grad_norm": 0.743350369247089,
+      "learning_rate": 7.127726593248337e-06,
+      "loss": 0.5978,
       "step": 229
     },
     {
-      "epoch": 1.270718232044199,
-      "grad_norm": 0.7982040126875192,
-      "learning_rate": 7.148680982113502e-06,
-      "loss": 0.6422,
+      "epoch": 1.2730627306273063,
+      "grad_norm": 0.7404143977218907,
+      "learning_rate": 7.098433949952146e-06,
+      "loss": 0.6437,
       "step": 230
     },
     {
-      "epoch": 1.276243093922652,
-      "grad_norm": 0.6629426029965108,
-      "learning_rate": 7.119571953549305e-06,
-      "loss": 0.645,
+      "epoch": 1.2785977859778597,
+      "grad_norm": 0.684639141766216,
+      "learning_rate": 7.069053622525697e-06,
+      "loss": 0.6318,
       "step": 231
     },
     {
-      "epoch": 1.281767955801105,
-      "grad_norm": 0.6597217484497837,
-      "learning_rate": 7.0903750820647175e-06,
-      "loss": 0.6046,
+      "epoch": 1.2841328413284132,
+      "grad_norm": 0.6506218001874067,
+      "learning_rate": 7.039586838640918e-06,
+      "loss": 0.5983,
       "step": 232
     },
     {
-      "epoch": 1.287292817679558,
-      "grad_norm": 0.6171612521609076,
-      "learning_rate": 7.061091577686349e-06,
-      "loss": 0.5663,
+      "epoch": 1.2896678966789668,
+      "grad_norm": 0.6249292386050818,
+      "learning_rate": 7.0100348295823706e-06,
+      "loss": 0.5439,
       "step": 233
     },
     {
-      "epoch": 1.292817679558011,
-      "grad_norm": 0.6930974359608171,
-      "learning_rate": 7.031722654031192e-06,
-      "loss": 0.5901,
+      "epoch": 1.2952029520295203,
+      "grad_norm": 0.7671687660019585,
+      "learning_rate": 6.980398830195785e-06,
+      "loss": 0.6191,
       "step": 234
     },
     {
-      "epoch": 1.298342541436464,
-      "grad_norm": 0.6340608907226989,
-      "learning_rate": 7.002269528256334e-06,
-      "loss": 0.6084,
+      "epoch": 1.3007380073800738,
+      "grad_norm": 0.6567756554372846,
+      "learning_rate": 6.950680078836475e-06,
+      "loss": 0.6097,
       "step": 235
     },
     {
-      "epoch": 1.3038674033149171,
-      "grad_norm": 0.6695648424102065,
-      "learning_rate": 6.972733421008505e-06,
-      "loss": 0.5938,
+      "epoch": 1.3062730627306274,
+      "grad_norm": 0.7256041165241631,
+      "learning_rate": 6.920879817317588e-06,
+      "loss": 0.6237,
       "step": 236
     },
     {
-      "epoch": 1.3093922651933703,
-      "grad_norm": 0.7162730447160397,
-      "learning_rate": 6.943115556373503e-06,
-      "loss": 0.606,
+      "epoch": 1.311808118081181,
+      "grad_norm": 0.7062531126646335,
+      "learning_rate": 6.890999290858213e-06,
+      "loss": 0.5881,
       "step": 237
     },
     {
-      "epoch": 1.3149171270718232,
-      "grad_norm": 0.7136879200055026,
-      "learning_rate": 6.913417161825449e-06,
-      "loss": 0.6491,
+      "epoch": 1.3173431734317342,
+      "grad_norm": 0.7074515896572966,
+      "learning_rate": 6.861039748031351e-06,
+      "loss": 0.6399,
       "step": 238
     },
     {
-      "epoch": 1.3204419889502763,
-      "grad_norm": 0.696786721330877,
-      "learning_rate": 6.883639468175926e-06,
-      "loss": 0.6072,
+      "epoch": 1.3228782287822878,
+      "grad_norm": 0.7132720107486031,
+      "learning_rate": 6.8310024407117405e-06,
+      "loss": 0.6356,
       "step": 239
     },
     {
-      "epoch": 1.3259668508287292,
-      "grad_norm": 0.753245759137927,
-      "learning_rate": 6.853783709522963e-06,
-      "loss": 0.6018,
+      "epoch": 1.3284132841328413,
+      "grad_norm": 0.6998988120144802,
+      "learning_rate": 6.800888624023552e-06,
+      "loss": 0.6086,
       "step": 240
     },
     {
-      "epoch": 1.3314917127071824,
-      "grad_norm": 0.6968519374034413,
-      "learning_rate": 6.823851123199894e-06,
-      "loss": 0.6434,
+      "epoch": 1.3339483394833949,
+      "grad_norm": 0.7003663999513423,
+      "learning_rate": 6.770699556287939e-06,
+      "loss": 0.6242,
       "step": 241
     },
     {
-      "epoch": 1.3370165745856353,
-      "grad_norm": 0.6761342537400503,
-      "learning_rate": 6.793842949724074e-06,
-      "loss": 0.6225,
+      "epoch": 1.3394833948339484,
+      "grad_norm": 0.698672108839009,
+      "learning_rate": 6.740436498970453e-06,
+      "loss": 0.6583,
       "step": 242
     },
     {
-      "epoch": 1.3425414364640884,
-      "grad_norm": 0.7322253178916728,
-      "learning_rate": 6.763760432745475e-06,
-      "loss": 0.6131,
+      "epoch": 1.3450184501845017,
+      "grad_norm": 0.6864478737367914,
+      "learning_rate": 6.710100716628345e-06,
+      "loss": 0.5885,
       "step": 243
     },
     {
-      "epoch": 1.3480662983425415,
-      "grad_norm": 0.7699522982913194,
-      "learning_rate": 6.733604818995133e-06,
-      "loss": 0.6366,
+      "epoch": 1.3505535055350553,
+      "grad_norm": 0.7196144827279044,
+      "learning_rate": 6.679693476857712e-06,
+      "loss": 0.6075,
       "step": 244
     },
     {
-      "epoch": 1.3535911602209945,
-      "grad_norm": 0.6196932838702959,
-      "learning_rate": 6.703377358233489e-06,
-      "loss": 0.595,
+      "epoch": 1.3560885608856088,
+      "grad_norm": 0.6437948099966094,
+      "learning_rate": 6.649216050240539e-06,
+      "loss": 0.6097,
       "step": 245
     },
     {
-      "epoch": 1.3591160220994476,
-      "grad_norm": 0.7047536218041381,
-      "learning_rate": 6.673079303198591e-06,
-      "loss": 0.5941,
+      "epoch": 1.3616236162361623,
+      "grad_norm": 0.7620305140396721,
+      "learning_rate": 6.618669710291607e-06,
+      "loss": 0.5926,
       "step": 246
     },
     {
-      "epoch": 1.3646408839779005,
-      "grad_norm": 0.8335370354195214,
-      "learning_rate": 6.6427119095541745e-06,
-      "loss": 0.5967,
+      "epoch": 1.367158671586716,
+      "grad_norm": 0.7830266879556336,
+      "learning_rate": 6.588055733405266e-06,
+      "loss": 0.6033,
       "step": 247
     },
     {
-      "epoch": 1.3701657458563536,
-      "grad_norm": 0.7706739113368778,
-      "learning_rate": 6.612276435837622e-06,
-      "loss": 0.5917,
+      "epoch": 1.3726937269372694,
+      "grad_norm": 0.7123926132432568,
+      "learning_rate": 6.557375398802124e-06,
+      "loss": 0.6074,
       "step": 248
     },
     {
-      "epoch": 1.3756906077348066,
-      "grad_norm": 0.7446603300489363,
-      "learning_rate": 6.58177414340781e-06,
-      "loss": 0.6342,
+      "epoch": 1.378228782287823,
+      "grad_norm": 0.6724440991233431,
+      "learning_rate": 6.526629988475567e-06,
+      "loss": 0.6154,
       "step": 249
     },
     {
-      "epoch": 1.3812154696132597,
-      "grad_norm": 0.6592546584880928,
-      "learning_rate": 6.551206296392827e-06,
-      "loss": 0.5791,
+      "epoch": 1.3837638376383765,
+      "grad_norm": 0.705678216248485,
+      "learning_rate": 6.495820787138209e-06,
+      "loss": 0.5891,
       "step": 250
     },
     {
-      "epoch": 1.3867403314917128,
-      "grad_norm": 0.8080309844429614,
-      "learning_rate": 6.520574161637591e-06,
-      "loss": 0.6555,
+      "epoch": 1.3892988929889298,
+      "grad_norm": 0.7606972010978245,
+      "learning_rate": 6.4649490821682035e-06,
+      "loss": 0.6586,
       "step": 251
     },
     {
-      "epoch": 1.3922651933701657,
-      "grad_norm": 0.6757755218068,
-      "learning_rate": 6.4898790086513366e-06,
-      "loss": 0.6002,
+      "epoch": 1.3948339483394834,
+      "grad_norm": 0.627422187069653,
+      "learning_rate": 6.434016163555452e-06,
+      "loss": 0.5917,
       "step": 252
     },
     {
-      "epoch": 1.3977900552486187,
-      "grad_norm": 0.6540398994889906,
-      "learning_rate": 6.459122109555011e-06,
-      "loss": 0.5968,
+      "epoch": 1.400369003690037,
+      "grad_norm": 0.619815027726477,
+      "learning_rate": 6.403023323847695e-06,
+      "loss": 0.5878,
       "step": 253
     },
     {
-      "epoch": 1.4033149171270718,
-      "grad_norm": 0.6515476627729059,
-      "learning_rate": 6.42830473902855e-06,
-      "loss": 0.5915,
+      "epoch": 1.4059040590405905,
+      "grad_norm": 0.689330526944227,
+      "learning_rate": 6.371971858096509e-06,
+      "loss": 0.5972,
       "step": 254
     },
     {
-      "epoch": 1.408839779005525,
-      "grad_norm": 0.768393574101539,
-      "learning_rate": 6.397428174258048e-06,
-      "loss": 0.6129,
+      "epoch": 1.4114391143911438,
+      "grad_norm": 0.7173239755647481,
+      "learning_rate": 6.340863063803187e-06,
+      "loss": 0.6063,
       "step": 255
     },
     {
-      "epoch": 1.4143646408839778,
-      "grad_norm": 0.6602226522160092,
-      "learning_rate": 6.3664936948828296e-06,
-      "loss": 0.618,
+      "epoch": 1.4169741697416973,
+      "grad_norm": 0.6961778167779811,
+      "learning_rate": 6.30969824086453e-06,
+      "loss": 0.5807,
       "step": 256
     },
     {
-      "epoch": 1.419889502762431,
-      "grad_norm": 0.7186347942888126,
-      "learning_rate": 6.335502582942409e-06,
-      "loss": 0.58,
+      "epoch": 1.4225092250922509,
+      "grad_norm": 0.6879266863201914,
+      "learning_rate": 6.278478691518519e-06,
+      "loss": 0.5948,
       "step": 257
     },
     {
-      "epoch": 1.4254143646408841,
-      "grad_norm": 0.6867146462508941,
-      "learning_rate": 6.304456122823377e-06,
-      "loss": 0.6284,
+      "epoch": 1.4280442804428044,
+      "grad_norm": 0.7552082447717653,
+      "learning_rate": 6.247205720289907e-06,
+      "loss": 0.6406,
       "step": 258
     },
     {
-      "epoch": 1.430939226519337,
-      "grad_norm": 0.6640537281896691,
-      "learning_rate": 6.273355601206143e-06,
-      "loss": 0.628,
+      "epoch": 1.433579335793358,
+      "grad_norm": 0.6434001937152157,
+      "learning_rate": 6.215880633935709e-06,
+      "loss": 0.5952,
       "step": 259
     },
     {
-      "epoch": 1.43646408839779,
-      "grad_norm": 0.680561183563057,
-      "learning_rate": 6.24220230701164e-06,
-      "loss": 0.6601,
+      "epoch": 1.4391143911439115,
+      "grad_norm": 0.662148188293842,
+      "learning_rate": 6.184504741390596e-06,
+      "loss": 0.6319,
       "step": 260
     },
     {
-      "epoch": 1.441988950276243,
-      "grad_norm": 0.6266987287558203,
-      "learning_rate": 6.210997531347879e-06,
-      "loss": 0.5342,
+      "epoch": 1.444649446494465,
+      "grad_norm": 0.6153480539853726,
+      "learning_rate": 6.153079353712201e-06,
+      "loss": 0.5907,
       "step": 261
     },
     {
-      "epoch": 1.4475138121546962,
-      "grad_norm": 0.6970171405272507,
-      "learning_rate": 6.179742567456464e-06,
-      "loss": 0.6339,
+      "epoch": 1.4501845018450186,
+      "grad_norm": 0.6846732208662787,
+      "learning_rate": 6.121605784026339e-06,
+      "loss": 0.575,
       "step": 262
     },
     {
-      "epoch": 1.4530386740331491,
-      "grad_norm": 0.6699218536254724,
-      "learning_rate": 6.148438710658979e-06,
-      "loss": 0.5782,
+      "epoch": 1.455719557195572,
+      "grad_norm": 0.7039311923510397,
+      "learning_rate": 6.09008534747213e-06,
+      "loss": 0.6249,
       "step": 263
     },
     {
-      "epoch": 1.4585635359116023,
-      "grad_norm": 0.7406730671514048,
-      "learning_rate": 6.117087258303314e-06,
-      "loss": 0.6392,
+      "epoch": 1.4612546125461254,
+      "grad_norm": 0.7025974163318355,
+      "learning_rate": 6.058519361147055e-06,
+      "loss": 0.6097,
       "step": 264
     },
     {
-      "epoch": 1.4640883977900552,
-      "grad_norm": 0.6496784259742956,
-      "learning_rate": 6.085689509709893e-06,
-      "loss": 0.5848,
+      "epoch": 1.466789667896679,
+      "grad_norm": 0.6897172606752235,
+      "learning_rate": 6.02690914405191e-06,
+      "loss": 0.6459,
       "step": 265
     },
     {
-      "epoch": 1.4696132596685083,
-      "grad_norm": 0.6630461769430164,
-      "learning_rate": 6.0542467661178325e-06,
-      "loss": 0.6243,
+      "epoch": 1.4723247232472325,
+      "grad_norm": 0.6775134380343946,
+      "learning_rate": 5.995256017035703e-06,
+      "loss": 0.5369,
       "step": 266
     },
     {
-      "epoch": 1.4751381215469612,
-      "grad_norm": 0.7024087077842354,
-      "learning_rate": 6.022760330631006e-06,
-      "loss": 0.6081,
+      "epoch": 1.477859778597786,
+      "grad_norm": 0.7387800906697658,
+      "learning_rate": 5.9635613027404495e-06,
+      "loss": 0.6659,
       "step": 267
     },
     {
-      "epoch": 1.4806629834254144,
-      "grad_norm": 0.644335689792361,
-      "learning_rate": 5.991231508164037e-06,
-      "loss": 0.6315,
+      "epoch": 1.4833948339483394,
+      "grad_norm": 0.5992142954740902,
+      "learning_rate": 5.931826325545912e-06,
+      "loss": 0.6023,
       "step": 268
     },
     {
-      "epoch": 1.4861878453038675,
-      "grad_norm": 0.6011260710575211,
-      "learning_rate": 5.959661605388229e-06,
-      "loss": 0.5479,
+      "epoch": 1.488929889298893,
+      "grad_norm": 0.6869613178503449,
+      "learning_rate": 5.900052411514257e-06,
+      "loss": 0.5711,
       "step": 269
     },
     {
-      "epoch": 1.4917127071823204,
-      "grad_norm": 0.6735133008955199,
-      "learning_rate": 5.928051930677404e-06,
-      "loss": 0.6074,
+      "epoch": 1.4944649446494465,
+      "grad_norm": 0.6727011895310602,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 0.5675,
       "step": 270
     },
     {
-      "epoch": 1.4972375690607735,
-      "grad_norm": 0.724897423389795,
-      "learning_rate": 5.896403794053679e-06,
-      "loss": 0.6117,
+      "epoch": 1.5,
+      "grad_norm": 0.6728450572952694,
+      "learning_rate": 5.836393085267777e-06,
+      "loss": 0.6195,
       "step": 271
     },
     {
-      "epoch": 1.5027624309392267,
-      "grad_norm": 0.6432683628472128,
-      "learning_rate": 5.864718507133176e-06,
-      "loss": 0.6015,
+      "epoch": 1.5055350553505535,
+      "grad_norm": 0.6846152979341545,
+      "learning_rate": 5.804510333090287e-06,
+      "loss": 0.6213,
       "step": 272
     },
     {
-      "epoch": 1.5082872928176796,
-      "grad_norm": 0.6345537221217502,
-      "learning_rate": 5.83299738307166e-06,
-      "loss": 0.6491,
+      "epoch": 1.511070110701107,
+      "grad_norm": 0.6507562063592163,
+      "learning_rate": 5.772593964039203e-06,
+      "loss": 0.6403,
       "step": 273
     },
     {
-      "epoch": 1.5138121546961325,
-      "grad_norm": 0.6218808254180483,
-      "learning_rate": 5.801241736510128e-06,
-      "loss": 0.6134,
+      "epoch": 1.5166051660516606,
+      "grad_norm": 0.6432647166128416,
+      "learning_rate": 5.740645311756246e-06,
+      "loss": 0.6248,
       "step": 274
     },
     {
-      "epoch": 1.5193370165745856,
-      "grad_norm": 0.6782578788601987,
-      "learning_rate": 5.76945288352031e-06,
-      "loss": 0.6308,
+      "epoch": 1.5221402214022142,
+      "grad_norm": 0.6415893639295898,
+      "learning_rate": 5.708665711232103e-06,
+      "loss": 0.5917,
       "step": 275
     },
     {
-      "epoch": 1.5248618784530388,
-      "grad_norm": 0.5841892465281848,
-      "learning_rate": 5.7376321415501356e-06,
-      "loss": 0.5434,
+      "epoch": 1.5276752767527675,
+      "grad_norm": 0.6033176569216087,
+      "learning_rate": 5.6766564987506564e-06,
+      "loss": 0.5589,
       "step": 276
     },
     {
-      "epoch": 1.5303867403314917,
-      "grad_norm": 0.6650299495850774,
-      "learning_rate": 5.7057808293691305e-06,
-      "loss": 0.6293,
+      "epoch": 1.533210332103321,
+      "grad_norm": 0.7017492010981081,
+      "learning_rate": 5.644619011833134e-06,
+      "loss": 0.6301,
       "step": 277
     },
     {
-      "epoch": 1.5359116022099446,
-      "grad_norm": 0.752419119034265,
-      "learning_rate": 5.67390026701377e-06,
-      "loss": 0.6223,
+      "epoch": 1.5387453874538746,
+      "grad_norm": 0.7655426961892206,
+      "learning_rate": 5.612554589182228e-06,
+      "loss": 0.6593,
       "step": 278
     },
     {
-      "epoch": 1.5414364640883977,
-      "grad_norm": 0.8282947869336723,
-      "learning_rate": 5.641991775732756e-06,
-      "loss": 0.7056,
+      "epoch": 1.5442804428044279,
+      "grad_norm": 0.7237077018760665,
+      "learning_rate": 5.5804645706261515e-06,
+      "loss": 0.6946,
       "step": 279
     },
     {
-      "epoch": 1.5469613259668509,
-      "grad_norm": 0.7173511510171976,
-      "learning_rate": 5.610056677932274e-06,
-      "loss": 0.6027,
+      "epoch": 1.5498154981549814,
+      "grad_norm": 0.6848993007754327,
+      "learning_rate": 5.548350297062659e-06,
+      "loss": 0.6106,
       "step": 280
     },
     {
-      "epoch": 1.5524861878453038,
-      "grad_norm": 0.6564750666347287,
-      "learning_rate": 5.5780962971211795e-06,
-      "loss": 0.6608,
+      "epoch": 1.555350553505535,
+      "grad_norm": 0.6816964703469334,
+      "learning_rate": 5.516213110403009e-06,
+      "loss": 0.6959,
       "step": 281
     },
     {
-      "epoch": 1.558011049723757,
-      "grad_norm": 0.686798968992177,
-      "learning_rate": 5.546111957856155e-06,
-      "loss": 0.6637,
+      "epoch": 1.5608856088560885,
+      "grad_norm": 0.7121757712642759,
+      "learning_rate": 5.484054353515896e-06,
+      "loss": 0.6304,
       "step": 282
     },
     {
-      "epoch": 1.56353591160221,
-      "grad_norm": 0.8467673992557097,
-      "learning_rate": 5.514104985686802e-06,
-      "loss": 0.6916,
+      "epoch": 1.566420664206642,
+      "grad_norm": 0.785585840187818,
+      "learning_rate": 5.451875370171341e-06,
+      "loss": 0.6969,
       "step": 283
     },
     {
-      "epoch": 1.569060773480663,
-      "grad_norm": 0.7454256879275183,
-      "learning_rate": 5.482076707100723e-06,
-      "loss": 0.6741,
+      "epoch": 1.5719557195571956,
+      "grad_norm": 0.6929061668579762,
+      "learning_rate": 5.419677504984534e-06,
+      "loss": 0.6396,
       "step": 284
     },
     {
-      "epoch": 1.5745856353591159,
-      "grad_norm": 0.7068205882972726,
-      "learning_rate": 5.4500284494685275e-06,
-      "loss": 0.6049,
+      "epoch": 1.5774907749077491,
+      "grad_norm": 0.6711279274727765,
+      "learning_rate": 5.387462103359655e-06,
+      "loss": 0.6007,
       "step": 285
     },
     {
-      "epoch": 1.580110497237569,
-      "grad_norm": 0.6123366279350007,
-      "learning_rate": 5.417961540988837e-06,
-      "loss": 0.5804,
+      "epoch": 1.5830258302583027,
+      "grad_norm": 0.6361660730682954,
+      "learning_rate": 5.3552305114336515e-06,
+      "loss": 0.6057,
       "step": 286
     },
     {
-      "epoch": 1.5856353591160222,
-      "grad_norm": 0.6753783255596223,
-      "learning_rate": 5.385877310633233e-06,
-      "loss": 0.6144,
+      "epoch": 1.5885608856088562,
+      "grad_norm": 0.6937146439997242,
+      "learning_rate": 5.32298407601999e-06,
+      "loss": 0.5955,
       "step": 287
     },
     {
-      "epoch": 1.591160220994475,
-      "grad_norm": 0.6787303073540277,
-      "learning_rate": 5.353777088091177e-06,
-      "loss": 0.6109,
+      "epoch": 1.5940959409594095,
+      "grad_norm": 0.6641404122154957,
+      "learning_rate": 5.290724144552379e-06,
+      "loss": 0.5694,
       "step": 288
     },
     {
-      "epoch": 1.5966850828729282,
-      "grad_norm": 0.6689374877161456,
-      "learning_rate": 5.321662203714909e-06,
-      "loss": 0.5701,
+      "epoch": 1.599630996309963,
+      "grad_norm": 0.6925213117648781,
+      "learning_rate": 5.258452065028473e-06,
+      "loss": 0.578,
       "step": 289
     },
     {
-      "epoch": 1.6022099447513813,
-      "grad_norm": 0.7341945035228687,
-      "learning_rate": 5.289533988464307e-06,
-      "loss": 0.6451,
+      "epoch": 1.6051660516605166,
+      "grad_norm": 0.6749065542362322,
+      "learning_rate": 5.2261691859535325e-06,
+      "loss": 0.6517,
       "step": 290
     },
     {
-      "epoch": 1.6077348066298343,
-      "grad_norm": 0.7003036927886831,
-      "learning_rate": 5.257393773851733e-06,
-      "loss": 0.6009,
+      "epoch": 1.6107011070110702,
+      "grad_norm": 0.6569945077124647,
+      "learning_rate": 5.193876856284085e-06,
+      "loss": 0.5928,
       "step": 291
     },
     {
-      "epoch": 1.6132596685082872,
-      "grad_norm": 0.6798929226991296,
-      "learning_rate": 5.2252428918868446e-06,
-      "loss": 0.5811,
+      "epoch": 1.6162361623616235,
+      "grad_norm": 0.6611834974919497,
+      "learning_rate": 5.161576425371554e-06,
+      "loss": 0.5757,
       "step": 292
     },
     {
-      "epoch": 1.6187845303867403,
-      "grad_norm": 0.603212047810036,
-      "learning_rate": 5.193082675021393e-06,
-      "loss": 0.5346,
+      "epoch": 1.621771217712177,
+      "grad_norm": 0.6443267893510465,
+      "learning_rate": 5.1292692429058824e-06,
+      "loss": 0.567,
       "step": 293
     },
     {
-      "epoch": 1.6243093922651934,
-      "grad_norm": 0.6908051872037912,
-      "learning_rate": 5.160914456094005e-06,
-      "loss": 0.6487,
+      "epoch": 1.6273062730627306,
+      "grad_norm": 0.7370247594136543,
+      "learning_rate": 5.096956658859122e-06,
+      "loss": 0.6723,
       "step": 294
     },
     {
-      "epoch": 1.6298342541436464,
-      "grad_norm": 0.759496464512415,
-      "learning_rate": 5.1287395682749444e-06,
-      "loss": 0.6885,
+      "epoch": 1.632841328413284,
+      "grad_norm": 0.6633312016915573,
+      "learning_rate": 5.064640023429042e-06,
+      "loss": 0.6907,
       "step": 295
     },
     {
-      "epoch": 1.6353591160220995,
-      "grad_norm": 0.7182799104547218,
-      "learning_rate": 5.0965593450108495e-06,
-      "loss": 0.6172,
+      "epoch": 1.6383763837638377,
+      "grad_norm": 0.6796625855678173,
+      "learning_rate": 5.032320686982697e-06,
+      "loss": 0.6061,
       "step": 296
     },
     {
-      "epoch": 1.6408839779005526,
-      "grad_norm": 0.6463166442166366,
-      "learning_rate": 5.064375119969491e-06,
-      "loss": 0.6075,
+      "epoch": 1.6439114391143912,
+      "grad_norm": 0.6703488843908327,
+      "learning_rate": 5e-06,
+      "loss": 0.5949,
       "step": 297
     },
     {
-      "epoch": 1.6464088397790055,
-      "grad_norm": 0.7000976763577735,
-      "learning_rate": 5.03218822698448e-06,
-      "loss": 0.6416,
+      "epoch": 1.6494464944649447,
+      "grad_norm": 0.6879324468041449,
+      "learning_rate": 4.967679313017304e-06,
+      "loss": 0.6459,
       "step": 298
     },
     {
-      "epoch": 1.6519337016574585,
-      "grad_norm": 0.7020061467617188,
-      "learning_rate": 5e-06,
-      "loss": 0.5947,
+      "epoch": 1.6549815498154983,
+      "grad_norm": 0.6993913963211266,
+      "learning_rate": 4.9353599765709585e-06,
+      "loss": 0.5606,
       "step": 299
     },
     {
-      "epoch": 1.6574585635359116,
-      "grad_norm": 0.7488808415617535,
-      "learning_rate": 4.967811773015521e-06,
-      "loss": 0.5809,
+      "epoch": 1.6605166051660518,
+      "grad_norm": 0.6842342047928159,
+      "learning_rate": 4.903043341140879e-06,
+      "loss": 0.5859,
       "step": 300
     },
     {
-      "epoch": 1.6629834254143647,
-      "grad_norm": 0.634571743970242,
-      "learning_rate": 4.93562488003051e-06,
-      "loss": 0.5944,
+      "epoch": 1.6660516605166051,
+      "grad_norm": 0.6414031000352342,
+      "learning_rate": 4.870730757094121e-06,
+      "loss": 0.6028,
       "step": 301
     },
     {
-      "epoch": 1.6685082872928176,
-      "grad_norm": 0.6445574283924204,
-      "learning_rate": 4.90344065498915e-06,
-      "loss": 0.5984,
+      "epoch": 1.6715867158671587,
+      "grad_norm": 0.6730213374583364,
+      "learning_rate": 4.838423574628447e-06,
+      "loss": 0.5758,
       "step": 302
     },
     {
-      "epoch": 1.6740331491712708,
-      "grad_norm": 0.7137775145837433,
-      "learning_rate": 4.871260431725058e-06,
-      "loss": 0.5448,
+      "epoch": 1.6771217712177122,
+      "grad_norm": 0.685094126165947,
+      "learning_rate": 4.806123143715916e-06,
+      "loss": 0.5783,
       "step": 303
     },
     {
-      "epoch": 1.679558011049724,
-      "grad_norm": 0.6620727258550688,
-      "learning_rate": 4.8390855439059955e-06,
-      "loss": 0.5916,
+      "epoch": 1.6826568265682655,
+      "grad_norm": 0.6832171407137465,
+      "learning_rate": 4.773830814046469e-06,
+      "loss": 0.5928,
       "step": 304
     },
     {
-      "epoch": 1.6850828729281768,
-      "grad_norm": 0.648075981508642,
-      "learning_rate": 4.806917324978608e-06,
-      "loss": 0.5546,
+      "epoch": 1.688191881918819,
+      "grad_norm": 0.6333752235654893,
+      "learning_rate": 4.741547934971528e-06,
+      "loss": 0.5476,
       "step": 305
     },
     {
-      "epoch": 1.6906077348066297,
-      "grad_norm": 0.7077774255913108,
-      "learning_rate": 4.774757108113156e-06,
-      "loss": 0.5546,
+      "epoch": 1.6937269372693726,
+      "grad_norm": 0.7064803499084308,
+      "learning_rate": 4.7092758554476215e-06,
+      "loss": 0.6216,
       "step": 306
     },
     {
-      "epoch": 1.6961325966850829,
-      "grad_norm": 0.7237233387785816,
-      "learning_rate": 4.742606226148268e-06,
-      "loss": 0.6898,
+      "epoch": 1.6992619926199262,
+      "grad_norm": 0.6392420306364016,
+      "learning_rate": 4.677015923980012e-06,
+      "loss": 0.6679,
       "step": 307
     },
     {
-      "epoch": 1.701657458563536,
-      "grad_norm": 0.6284936804624925,
-      "learning_rate": 4.710466011535695e-06,
-      "loss": 0.6397,
+      "epoch": 1.7047970479704797,
+      "grad_norm": 0.6494731361510496,
+      "learning_rate": 4.644769488566351e-06,
+      "loss": 0.5886,
       "step": 308
     },
     {
-      "epoch": 1.707182320441989,
-      "grad_norm": 0.7927972591279655,
-      "learning_rate": 4.678337796285093e-06,
-      "loss": 0.6436,
+      "epoch": 1.7103321033210332,
+      "grad_norm": 0.6913018221886911,
+      "learning_rate": 4.6125378966403465e-06,
+      "loss": 0.6472,
       "step": 309
     },
     {
-      "epoch": 1.7127071823204418,
-      "grad_norm": 0.6592299730654411,
-      "learning_rate": 4.6462229119088234e-06,
-      "loss": 0.5714,
+      "epoch": 1.7158671586715868,
+      "grad_norm": 0.640802770083508,
+      "learning_rate": 4.580322495015466e-06,
+      "loss": 0.5561,
       "step": 310
     },
     {
-      "epoch": 1.7182320441988952,
-      "grad_norm": 0.6876331470178856,
-      "learning_rate": 4.614122689366769e-06,
-      "loss": 0.6147,
+      "epoch": 1.7214022140221403,
+      "grad_norm": 0.6960948975174029,
+      "learning_rate": 4.548124629828661e-06,
+      "loss": 0.6236,
       "step": 311
     },
     {
-      "epoch": 1.723756906077348,
-      "grad_norm": 0.6820420879873159,
-      "learning_rate": 4.582038459011165e-06,
-      "loss": 0.5625,
+      "epoch": 1.7269372693726939,
+      "grad_norm": 0.6389521774827648,
+      "learning_rate": 4.515945646484105e-06,
+      "loss": 0.5764,
       "step": 312
     },
     {
-      "epoch": 1.729281767955801,
-      "grad_norm": 0.7107222428969042,
-      "learning_rate": 4.549971550531474e-06,
-      "loss": 0.6362,
+      "epoch": 1.7324723247232472,
+      "grad_norm": 0.6505284331128637,
+      "learning_rate": 4.483786889596993e-06,
+      "loss": 0.6177,
       "step": 313
     },
     {
-      "epoch": 1.7348066298342542,
-      "grad_norm": 0.6991332889693773,
-      "learning_rate": 4.51792329289928e-06,
-      "loss": 0.6662,
+      "epoch": 1.7380073800738007,
+      "grad_norm": 0.7322008377502368,
+      "learning_rate": 4.451649702937343e-06,
+      "loss": 0.7031,
       "step": 314
     },
     {
-      "epoch": 1.7403314917127073,
-      "grad_norm": 0.6509307723816805,
-      "learning_rate": 4.485895014313198e-06,
-      "loss": 0.6239,
+      "epoch": 1.7435424354243543,
+      "grad_norm": 0.667856136908068,
+      "learning_rate": 4.4195354293738484e-06,
+      "loss": 0.6017,
       "step": 315
     },
     {
-      "epoch": 1.7458563535911602,
-      "grad_norm": 0.6475315801384613,
-      "learning_rate": 4.453888042143847e-06,
-      "loss": 0.6175,
+      "epoch": 1.7490774907749076,
+      "grad_norm": 0.6167961116394701,
+      "learning_rate": 4.387445410817774e-06,
+      "loss": 0.5948,
       "step": 316
     },
     {
-      "epoch": 1.7513812154696131,
-      "grad_norm": 0.6737553368323486,
-      "learning_rate": 4.421903702878822e-06,
-      "loss": 0.5807,
+      "epoch": 1.7546125461254611,
+      "grad_norm": 0.6488335556890084,
+      "learning_rate": 4.355380988166867e-06,
+      "loss": 0.5784,
       "step": 317
     },
     {
-      "epoch": 1.7569060773480663,
-      "grad_norm": 0.6579280198374517,
-      "learning_rate": 4.389943322067728e-06,
-      "loss": 0.5769,
+      "epoch": 1.7601476014760147,
+      "grad_norm": 0.6433975581483212,
+      "learning_rate": 4.323343501249346e-06,
+      "loss": 0.5302,
       "step": 318
     },
     {
-      "epoch": 1.7624309392265194,
-      "grad_norm": 0.6747867888004441,
-      "learning_rate": 4.358008224267245e-06,
-      "loss": 0.5521,
+      "epoch": 1.7656826568265682,
+      "grad_norm": 0.6706524043913139,
+      "learning_rate": 4.291334288767899e-06,
+      "loss": 0.5922,
       "step": 319
     },
     {
-      "epoch": 1.7679558011049723,
-      "grad_norm": 0.5916747700224341,
-      "learning_rate": 4.326099732986231e-06,
-      "loss": 0.533,
+      "epoch": 1.7712177121771218,
+      "grad_norm": 0.6591469335328922,
+      "learning_rate": 4.259354688243758e-06,
+      "loss": 0.5898,
       "step": 320
     },
     {
-      "epoch": 1.7734806629834254,
-      "grad_norm": 0.6766138160854878,
-      "learning_rate": 4.29421917063087e-06,
-      "loss": 0.6057,
+      "epoch": 1.7767527675276753,
+      "grad_norm": 0.6481007609954463,
+      "learning_rate": 4.227406035960798e-06,
+      "loss": 0.5922,
       "step": 321
     },
     {
-      "epoch": 1.7790055248618786,
-      "grad_norm": 0.6880698365575919,
-      "learning_rate": 4.262367858449867e-06,
-      "loss": 0.6024,
+      "epoch": 1.7822878228782288,
+      "grad_norm": 0.6650428282720707,
+      "learning_rate": 4.195489666909714e-06,
+      "loss": 0.6023,
       "step": 322
     },
     {
-      "epoch": 1.7845303867403315,
-      "grad_norm": 0.6506755980460356,
-      "learning_rate": 4.230547116479691e-06,
-      "loss": 0.5889,
+      "epoch": 1.7878228782287824,
+      "grad_norm": 0.636497135234757,
+      "learning_rate": 4.163606914732224e-06,
+      "loss": 0.5807,
       "step": 323
     },
     {
-      "epoch": 1.7900552486187844,
-      "grad_norm": 0.6597664588077232,
-      "learning_rate": 4.1987582634898724e-06,
-      "loss": 0.5664,
+      "epoch": 1.793357933579336,
+      "grad_norm": 0.6574211931288638,
+      "learning_rate": 4.131759111665349e-06,
+      "loss": 0.5459,
       "step": 324
     },
     {
-      "epoch": 1.7955801104972375,
-      "grad_norm": 0.6385623465686507,
-      "learning_rate": 4.167002616928341e-06,
-      "loss": 0.5943,
+      "epoch": 1.7988929889298892,
+      "grad_norm": 0.6684303900701675,
+      "learning_rate": 4.099947588485744e-06,
+      "loss": 0.6271,
       "step": 325
     },
     {
-      "epoch": 1.8011049723756907,
-      "grad_norm": 0.6967677614778646,
-      "learning_rate": 4.135281492866826e-06,
-      "loss": 0.6442,
+      "epoch": 1.8044280442804428,
+      "grad_norm": 0.7115325195567614,
+      "learning_rate": 4.06817367445409e-06,
+      "loss": 0.6444,
       "step": 326
     },
     {
-      "epoch": 1.8066298342541436,
-      "grad_norm": 0.6482473642073888,
-      "learning_rate": 4.103596205946323e-06,
-      "loss": 0.6201,
+      "epoch": 1.8099630996309963,
+      "grad_norm": 0.6896195168622968,
+      "learning_rate": 4.036438697259551e-06,
+      "loss": 0.5971,
       "step": 327
     },
     {
-      "epoch": 1.8121546961325967,
-      "grad_norm": 0.7193129213261211,
-      "learning_rate": 4.0719480693225964e-06,
-      "loss": 0.6096,
+      "epoch": 1.8154981549815496,
+      "grad_norm": 0.6687990794813583,
+      "learning_rate": 4.004743982964298e-06,
+      "loss": 0.6449,
       "step": 328
     },
     {
-      "epoch": 1.8176795580110499,
-      "grad_norm": 0.6888287876327789,
-      "learning_rate": 4.040338394611772e-06,
-      "loss": 0.6086,
+      "epoch": 1.8210332103321032,
+      "grad_norm": 0.6668553440261493,
+      "learning_rate": 3.9730908559480904e-06,
+      "loss": 0.5765,
       "step": 329
     },
     {
-      "epoch": 1.8232044198895028,
-      "grad_norm": 0.6553298757478724,
-      "learning_rate": 4.0087684918359646e-06,
-      "loss": 0.6059,
+      "epoch": 1.8265682656826567,
+      "grad_norm": 0.658154260938912,
+      "learning_rate": 3.941480638852948e-06,
+      "loss": 0.6424,
       "step": 330
     },
     {
-      "epoch": 1.8287292817679557,
-      "grad_norm": 0.6945457240360784,
-      "learning_rate": 3.977239669368998e-06,
-      "loss": 0.6322,
+      "epoch": 1.8321033210332103,
+      "grad_norm": 0.7676795703743602,
+      "learning_rate": 3.909914652527872e-06,
+      "loss": 0.5954,
       "step": 331
     },
     {
-      "epoch": 1.8342541436464088,
-      "grad_norm": 0.7470607646196792,
-      "learning_rate": 3.945753233882168e-06,
-      "loss": 0.6203,
+      "epoch": 1.8376383763837638,
+      "grad_norm": 0.753359036105102,
+      "learning_rate": 3.878394215973663e-06,
+      "loss": 0.6623,
       "step": 332
     },
     {
-      "epoch": 1.839779005524862,
-      "grad_norm": 0.6675969015071959,
-      "learning_rate": 3.9143104902901085e-06,
-      "loss": 0.6047,
+      "epoch": 1.8431734317343174,
+      "grad_norm": 0.6328329983343084,
+      "learning_rate": 3.8469206462878e-06,
+      "loss": 0.5646,
       "step": 333
     },
     {
-      "epoch": 1.8453038674033149,
-      "grad_norm": 0.631838443419436,
-      "learning_rate": 3.882912741696688e-06,
-      "loss": 0.5619,
+      "epoch": 1.848708487084871,
+      "grad_norm": 0.6522576574675761,
+      "learning_rate": 3.815495258609404e-06,
+      "loss": 0.5644,
       "step": 334
     },
     {
-      "epoch": 1.850828729281768,
-      "grad_norm": 0.6921273737693909,
-      "learning_rate": 3.851561289341023e-06,
-      "loss": 0.5852,
+      "epoch": 1.8542435424354244,
+      "grad_norm": 0.6478750849380562,
+      "learning_rate": 3.784119366064293e-06,
+      "loss": 0.5752,
       "step": 335
     },
     {
-      "epoch": 1.8563535911602211,
-      "grad_norm": 0.6212122428108089,
-      "learning_rate": 3.820257432543539e-06,
-      "loss": 0.5631,
+      "epoch": 1.859778597785978,
+      "grad_norm": 0.6468377900933637,
+      "learning_rate": 3.752794279710094e-06,
+      "loss": 0.6167,
       "step": 336
     },
     {
-      "epoch": 1.861878453038674,
-      "grad_norm": 0.7555498667849819,
-      "learning_rate": 3.789002468652121e-06,
-      "loss": 0.6727,
+      "epoch": 1.8653136531365315,
+      "grad_norm": 0.6738182100392553,
+      "learning_rate": 3.721521308481483e-06,
+      "loss": 0.6567,
       "step": 337
     },
     {
-      "epoch": 1.867403314917127,
-      "grad_norm": 0.6679491834699562,
-      "learning_rate": 3.7577976929883608e-06,
-      "loss": 0.5829,
+      "epoch": 1.8708487084870848,
+      "grad_norm": 0.6812077855923881,
+      "learning_rate": 3.690301759135471e-06,
+      "loss": 0.5609,
       "step": 338
     },
     {
-      "epoch": 1.87292817679558,
-      "grad_norm": 0.7841296013368897,
-      "learning_rate": 3.726644398793857e-06,
-      "loss": 0.6562,
+      "epoch": 1.8763837638376384,
+      "grad_norm": 0.739726141586659,
+      "learning_rate": 3.6591369361968127e-06,
+      "loss": 0.6721,
       "step": 339
     },
     {
-      "epoch": 1.8784530386740332,
-      "grad_norm": 0.6836490585487593,
-      "learning_rate": 3.695543877176626e-06,
-      "loss": 0.6354,
+      "epoch": 1.881918819188192,
+      "grad_norm": 0.6632323020436683,
+      "learning_rate": 3.6280281419034934e-06,
+      "loss": 0.6158,
       "step": 340
     },
     {
-      "epoch": 1.8839779005524862,
-      "grad_norm": 0.6394642182719062,
-      "learning_rate": 3.6644974170575907e-06,
-      "loss": 0.6078,
+      "epoch": 1.8874538745387452,
+      "grad_norm": 0.6554347656581475,
+      "learning_rate": 3.596976676152306e-06,
+      "loss": 0.6118,
       "step": 341
     },
     {
-      "epoch": 1.889502762430939,
-      "grad_norm": 0.7143267899941073,
-      "learning_rate": 3.6335063051171725e-06,
-      "loss": 0.6029,
+      "epoch": 1.8929889298892988,
+      "grad_norm": 0.6943502042805372,
+      "learning_rate": 3.5659838364445505e-06,
+      "loss": 0.6257,
       "step": 342
     },
     {
-      "epoch": 1.8950276243093924,
-      "grad_norm": 0.6909108022068383,
-      "learning_rate": 3.6025718257419532e-06,
-      "loss": 0.6233,
+      "epoch": 1.8985239852398523,
+      "grad_norm": 0.7344400676037952,
+      "learning_rate": 3.535050917831797e-06,
+      "loss": 0.6154,
       "step": 343
     },
     {
-      "epoch": 1.9005524861878453,
-      "grad_norm": 0.6329351448339191,
-      "learning_rate": 3.5716952609714517e-06,
-      "loss": 0.5878,
+      "epoch": 1.9040590405904059,
+      "grad_norm": 0.6960672172751073,
+      "learning_rate": 3.504179212861793e-06,
+      "loss": 0.6025,
       "step": 344
     },
     {
-      "epoch": 1.9060773480662982,
-      "grad_norm": 0.6478503333940434,
-      "learning_rate": 3.540877890444989e-06,
-      "loss": 0.5693,
+      "epoch": 1.9095940959409594,
+      "grad_norm": 0.6793261432569583,
+      "learning_rate": 3.473370011524435e-06,
+      "loss": 0.5547,
       "step": 345
     },
     {
-      "epoch": 1.9116022099447514,
-      "grad_norm": 0.676681543424856,
-      "learning_rate": 3.5101209913486655e-06,
-      "loss": 0.5788,
+      "epoch": 1.915129151291513,
+      "grad_norm": 0.6780588276365805,
+      "learning_rate": 3.442624601197877e-06,
+      "loss": 0.5933,
       "step": 346
     },
     {
-      "epoch": 1.9171270718232045,
-      "grad_norm": 0.6431181230720638,
-      "learning_rate": 3.4794258383624115e-06,
-      "loss": 0.6159,
+      "epoch": 1.9206642066420665,
+      "grad_norm": 0.6333360607546321,
+      "learning_rate": 3.4119442665947346e-06,
+      "loss": 0.6066,
       "step": 347
     },
     {
-      "epoch": 1.9226519337016574,
-      "grad_norm": 0.6037104641827318,
-      "learning_rate": 3.448793703607175e-06,
-      "loss": 0.5596,
+      "epoch": 1.92619926199262,
+      "grad_norm": 0.6297918625305783,
+      "learning_rate": 3.3813302897083955e-06,
+      "loss": 0.6139,
       "step": 348
     },
     {
-      "epoch": 1.9281767955801103,
-      "grad_norm": 0.6475020350302072,
-      "learning_rate": 3.4182258565921933e-06,
-      "loss": 0.6464,
+      "epoch": 1.9317343173431736,
+      "grad_norm": 0.65128547682773,
+      "learning_rate": 3.350783949759462e-06,
+      "loss": 0.6065,
       "step": 349
     },
     {
-      "epoch": 1.9337016574585635,
-      "grad_norm": 0.6509624192770844,
-      "learning_rate": 3.3877235641623797e-06,
-      "loss": 0.5798,
+      "epoch": 1.937269372693727,
+      "grad_norm": 0.663702410189263,
+      "learning_rate": 3.3203065231422904e-06,
+      "loss": 0.5595,
       "step": 350
     },
     {
-      "epoch": 1.9392265193370166,
-      "grad_norm": 0.73517544137987,
-      "learning_rate": 3.3572880904458267e-06,
-      "loss": 0.5513,
+      "epoch": 1.9428044280442804,
+      "grad_norm": 0.7456337908851745,
+      "learning_rate": 3.289899283371657e-06,
+      "loss": 0.5635,
       "step": 351
     },
     {
-      "epoch": 1.9447513812154695,
-      "grad_norm": 0.693346997717625,
-      "learning_rate": 3.32692069680141e-06,
-      "loss": 0.6178,
+      "epoch": 1.948339483394834,
+      "grad_norm": 0.6245107839051393,
+      "learning_rate": 3.259563501029548e-06,
+      "loss": 0.6149,
       "step": 352
     },
     {
-      "epoch": 1.9502762430939227,
-      "grad_norm": 0.6108811315736592,
-      "learning_rate": 3.2966226417665125e-06,
-      "loss": 0.6236,
+      "epoch": 1.9538745387453873,
+      "grad_norm": 0.6389202208791118,
+      "learning_rate": 3.2293004437120622e-06,
+      "loss": 0.6318,
       "step": 353
     },
     {
-      "epoch": 1.9558011049723758,
-      "grad_norm": 0.629537696210223,
-      "learning_rate": 3.2663951810048683e-06,
-      "loss": 0.6122,
+      "epoch": 1.9594095940959408,
+      "grad_norm": 0.6408635995037254,
+      "learning_rate": 3.1991113759764493e-06,
+      "loss": 0.6557,
       "step": 354
     },
     {
-      "epoch": 1.9613259668508287,
-      "grad_norm": 0.6789659150656049,
-      "learning_rate": 3.236239567254526e-06,
-      "loss": 0.6318,
+      "epoch": 1.9649446494464944,
+      "grad_norm": 0.6610364476788523,
+      "learning_rate": 3.1689975592882603e-06,
+      "loss": 0.6132,
       "step": 355
     },
     {
-      "epoch": 1.9668508287292816,
-      "grad_norm": 0.6281193326590523,
-      "learning_rate": 3.206157050275927e-06,
-      "loss": 0.5859,
+      "epoch": 1.970479704797048,
+      "grad_norm": 0.6162945914008929,
+      "learning_rate": 3.1389602519686515e-06,
+      "loss": 0.5645,
       "step": 356
     },
     {
-      "epoch": 1.9723756906077348,
-      "grad_norm": 0.6576491463900354,
-      "learning_rate": 3.176148876800109e-06,
-      "loss": 0.5849,
+      "epoch": 1.9760147601476015,
+      "grad_norm": 0.6343892270612764,
+      "learning_rate": 3.1090007091417884e-06,
+      "loss": 0.5662,
       "step": 357
     },
     {
-      "epoch": 1.977900552486188,
-      "grad_norm": 0.6853036340913029,
-      "learning_rate": 3.1462162904770376e-06,
-      "loss": 0.5779,
+      "epoch": 1.981549815498155,
+      "grad_norm": 0.6456644070716252,
+      "learning_rate": 3.0791201826824117e-06,
+      "loss": 0.5729,
       "step": 358
     },
     {
-      "epoch": 1.9834254143646408,
-      "grad_norm": 0.6363463477287988,
-      "learning_rate": 3.116360531824074e-06,
-      "loss": 0.5699,
+      "epoch": 1.9870848708487086,
+      "grad_norm": 0.6692278565767988,
+      "learning_rate": 3.049319921163526e-06,
+      "loss": 0.5692,
       "step": 359
     },
     {
-      "epoch": 1.988950276243094,
-      "grad_norm": 0.6709331445271008,
-      "learning_rate": 3.0865828381745515e-06,
-      "loss": 0.5765,
+      "epoch": 1.992619926199262,
+      "grad_norm": 0.6513390055411725,
+      "learning_rate": 3.019601169804216e-06,
+      "loss": 0.5702,
       "step": 360
     },
     {
-      "epoch": 1.994475138121547,
-      "grad_norm": 0.6288565179176786,
-      "learning_rate": 3.056884443626499e-06,
-      "loss": 0.5679,
+      "epoch": 1.9981549815498156,
+      "grad_norm": 0.8207020809923308,
+      "learning_rate": 2.9899651704176324e-06,
+      "loss": 0.712,
       "step": 361
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 0.6979469101641466,
-      "learning_rate": 3.027266578991497e-06,
-      "loss": 0.6038,
+      "epoch": 2.003690036900369,
+      "grad_norm": 0.8316443587754938,
+      "learning_rate": 2.9604131613590825e-06,
+      "loss": 0.6622,
       "step": 362
     },
     {
-      "epoch": 2.005524861878453,
-      "grad_norm": 0.6857162137141414,
-      "learning_rate": 2.997730471743667e-06,
-      "loss": 0.6094,
+      "epoch": 2.0092250922509227,
+      "grad_norm": 0.6321179010589463,
+      "learning_rate": 2.9309463774743047e-06,
+      "loss": 0.5349,
       "step": 363
     },
     {
-      "epoch": 2.0110497237569063,
-      "grad_norm": 0.6086171132783301,
-      "learning_rate": 2.9682773459688087e-06,
-      "loss": 0.4948,
+      "epoch": 2.014760147601476,
+      "grad_norm": 0.6623408052994709,
+      "learning_rate": 2.901566050047855e-06,
+      "loss": 0.558,
       "step": 364
     },
     {
-      "epoch": 2.016574585635359,
-      "grad_norm": 0.6681480376436151,
-      "learning_rate": 2.9389084223136523e-06,
-      "loss": 0.5459,
+      "epoch": 2.0202952029520294,
+      "grad_norm": 0.6148409552359754,
+      "learning_rate": 2.8722734067516637e-06,
+      "loss": 0.5131,
       "step": 365
     },
     {
-      "epoch": 2.022099447513812,
-      "grad_norm": 0.593654083071827,
-      "learning_rate": 2.9096249179352833e-06,
-      "loss": 0.5109,
+      "epoch": 2.025830258302583,
+      "grad_norm": 0.628811924344138,
+      "learning_rate": 2.843069671593734e-06,
+      "loss": 0.5053,
       "step": 366
     },
     {
-      "epoch": 2.027624309392265,
-      "grad_norm": 0.7007646517763502,
-      "learning_rate": 2.880428046450697e-06,
-      "loss": 0.5265,
+      "epoch": 2.0313653136531364,
+      "grad_norm": 0.7178060640719511,
+      "learning_rate": 2.813956064866996e-06,
+      "loss": 0.5215,
       "step": 367
     },
     {
-      "epoch": 2.0331491712707184,
-      "grad_norm": 0.702650158812983,
-      "learning_rate": 2.8513190178865004e-06,
-      "loss": 0.5257,
+      "epoch": 2.03690036900369,
+      "grad_norm": 0.64237638492648,
+      "learning_rate": 2.784933803098326e-06,
+      "loss": 0.5344,
       "step": 368
     },
     {
-      "epoch": 2.0386740331491713,
-      "grad_norm": 0.6531178861723401,
-      "learning_rate": 2.822299038628762e-06,
-      "loss": 0.5403,
+      "epoch": 2.0424354243542435,
+      "grad_norm": 0.660701484636481,
+      "learning_rate": 2.7560040989976894e-06,
+      "loss": 0.5522,
       "step": 369
     },
     {
-      "epoch": 2.044198895027624,
-      "grad_norm": 0.6472401575778369,
-      "learning_rate": 2.793369311373021e-06,
-      "loss": 0.5365,
+      "epoch": 2.047970479704797,
+      "grad_norm": 0.6442004508033177,
+      "learning_rate": 2.7271681614074973e-06,
+      "loss": 0.5692,
       "step": 370
     },
     {
-      "epoch": 2.0497237569060776,
-      "grad_norm": 0.6641615455285143,
-      "learning_rate": 2.7645310350744296e-06,
-      "loss": 0.5952,
+      "epoch": 2.0535055350553506,
+      "grad_norm": 0.701530505525817,
+      "learning_rate": 2.6984271952520723e-06,
+      "loss": 0.558,
       "step": 371
     },
     {
-      "epoch": 2.0552486187845305,
-      "grad_norm": 0.7123204501299103,
-      "learning_rate": 2.7357854048980893e-06,
-      "loss": 0.5631,
+      "epoch": 2.059040590405904,
+      "grad_norm": 0.6677439032763148,
+      "learning_rate": 2.6697824014873076e-06,
+      "loss": 0.5507,
       "step": 372
     },
     {
-      "epoch": 2.0607734806629834,
-      "grad_norm": 0.6900539716574728,
-      "learning_rate": 2.7071336121694856e-06,
-      "loss": 0.5452,
+      "epoch": 2.0645756457564577,
+      "grad_norm": 0.750156552596835,
+      "learning_rate": 2.641234977050484e-06,
+      "loss": 0.5605,
       "step": 373
     },
     {
-      "epoch": 2.0662983425414363,
-      "grad_norm": 0.7021705227314888,
-      "learning_rate": 2.6785768443251437e-06,
-      "loss": 0.5436,
+      "epoch": 2.0701107011070112,
+      "grad_norm": 0.7047824557323877,
+      "learning_rate": 2.6127861148102552e-06,
+      "loss": 0.5489,
       "step": 374
     },
     {
-      "epoch": 2.0718232044198897,
-      "grad_norm": 0.7605695654764374,
-      "learning_rate": 2.6501162848634023e-06,
-      "loss": 0.5488,
+      "epoch": 2.0756457564575648,
+      "grad_norm": 0.711760947816998,
+      "learning_rate": 2.5844370035168077e-06,
+      "loss": 0.5525,
       "step": 375
     },
     {
-      "epoch": 2.0773480662983426,
-      "grad_norm": 0.683125655831491,
-      "learning_rate": 2.621753113295361e-06,
-      "loss": 0.5851,
+      "epoch": 2.081180811808118,
+      "grad_norm": 0.6970505989934466,
+      "learning_rate": 2.5561888277521797e-06,
+      "loss": 0.533,
       "step": 376
     },
     {
-      "epoch": 2.0828729281767955,
-      "grad_norm": 0.6551843848513452,
-      "learning_rate": 2.5934885050960183e-06,
-      "loss": 0.5096,
+      "epoch": 2.0867158671586714,
+      "grad_norm": 0.6679637897186145,
+      "learning_rate": 2.528042767880766e-06,
+      "loss": 0.5421,
       "step": 377
     },
     {
-      "epoch": 2.0883977900552484,
-      "grad_norm": 0.7056721518164418,
-      "learning_rate": 2.565323631655532e-06,
-      "loss": 0.5273,
+      "epoch": 2.092250922509225,
+      "grad_norm": 0.6516580625237066,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.5511,
       "step": 378
     },
     {
-      "epoch": 2.0939226519337018,
-      "grad_norm": 0.6611500614916435,
-      "learning_rate": 2.537259660230679e-06,
-      "loss": 0.5454,
+      "epoch": 2.0977859778597785,
+      "grad_norm": 0.6626636324231238,
+      "learning_rate": 2.4720616958912054e-06,
+      "loss": 0.5732,
       "step": 379
     },
     {
-      "epoch": 2.0994475138121547,
-      "grad_norm": 0.7019525646170253,
-      "learning_rate": 2.5092977538964887e-06,
-      "loss": 0.5911,
+      "epoch": 2.103321033210332,
+      "grad_norm": 0.6723428723615049,
+      "learning_rate": 2.4442290229706344e-06,
+      "loss": 0.5727,
       "step": 380
     },
     {
-      "epoch": 2.1049723756906076,
-      "grad_norm": 0.6472748593975306,
-      "learning_rate": 2.4814390714980325e-06,
-      "loss": 0.5211,
+      "epoch": 2.1088560885608856,
+      "grad_norm": 0.6599974503924196,
+      "learning_rate": 2.4165031442406857e-06,
+      "loss": 0.5238,
       "step": 381
     },
     {
-      "epoch": 2.110497237569061,
-      "grad_norm": 0.7243217306106386,
-      "learning_rate": 2.4536847676023995e-06,
-      "loss": 0.5731,
+      "epoch": 2.114391143911439,
+      "grad_norm": 0.7008041331642095,
+      "learning_rate": 2.3888852182413087e-06,
+      "loss": 0.4859,
       "step": 382
     },
     {
-      "epoch": 2.116022099447514,
-      "grad_norm": 0.6881568022419648,
-      "learning_rate": 2.426035992450848e-06,
-      "loss": 0.4937,
+      "epoch": 2.1199261992619927,
+      "grad_norm": 0.7500710516966634,
+      "learning_rate": 2.361376399001592e-06,
+      "loss": 0.545,
       "step": 383
     },
     {
-      "epoch": 2.1215469613259668,
-      "grad_norm": 0.675968032373875,
-      "learning_rate": 2.398493891911127e-06,
-      "loss": 0.5293,
+      "epoch": 2.125461254612546,
+      "grad_norm": 0.6693296360494665,
+      "learning_rate": 2.333977835991545e-06,
+      "loss": 0.5467,
       "step": 384
     },
     {
-      "epoch": 2.12707182320442,
-      "grad_norm": 0.6793280674446409,
-      "learning_rate": 2.3710596074300045e-06,
-      "loss": 0.5502,
+      "epoch": 2.1309963099630997,
+      "grad_norm": 0.6478267617206592,
+      "learning_rate": 2.3066906740740626e-06,
+      "loss": 0.5347,
       "step": 385
     },
     {
-      "epoch": 2.132596685082873,
-      "grad_norm": 0.6827900429979371,
-      "learning_rate": 2.3437342759859472e-06,
-      "loss": 0.5774,
+      "epoch": 2.1365313653136533,
+      "grad_norm": 0.7177671429812185,
+      "learning_rate": 2.2795160534570866e-06,
+      "loss": 0.623,
       "step": 386
     },
     {
-      "epoch": 2.138121546961326,
-      "grad_norm": 0.6299741302424408,
-      "learning_rate": 2.316519030041998e-06,
-      "loss": 0.5355,
+      "epoch": 2.142066420664207,
+      "grad_norm": 0.6423839307947838,
+      "learning_rate": 2.2524551096459703e-06,
+      "loss": 0.5646,
       "step": 387
     },
     {
-      "epoch": 2.143646408839779,
-      "grad_norm": 0.679310688963614,
-      "learning_rate": 2.289414997498856e-06,
-      "loss": 0.579,
+      "epoch": 2.14760147601476,
+      "grad_norm": 0.6793354954895883,
+      "learning_rate": 2.2255089733960162e-06,
+      "loss": 0.5316,
       "step": 388
     },
     {
-      "epoch": 2.149171270718232,
-      "grad_norm": 0.6255767903887023,
-      "learning_rate": 2.2624233016481224e-06,
-      "loss": 0.5316,
+      "epoch": 2.1531365313653135,
+      "grad_norm": 0.663196223424689,
+      "learning_rate": 2.1986787706652377e-06,
+      "loss": 0.5788,
       "step": 389
     },
     {
-      "epoch": 2.154696132596685,
-      "grad_norm": 0.6592300712497258,
-      "learning_rate": 2.235545061125748e-06,
-      "loss": 0.5532,
+      "epoch": 2.158671586715867,
+      "grad_norm": 0.613244135525591,
+      "learning_rate": 2.171965622567308e-06,
+      "loss": 0.4808,
       "step": 390
     },
     {
-      "epoch": 2.160220994475138,
-      "grad_norm": 0.6610713391367827,
-      "learning_rate": 2.2087813898656775e-06,
-      "loss": 0.507,
+      "epoch": 2.1642066420664205,
+      "grad_norm": 0.6636202632875322,
+      "learning_rate": 2.1453706453247088e-06,
+      "loss": 0.5427,
       "step": 391
     },
     {
-      "epoch": 2.165745856353591,
-      "grad_norm": 0.6350291188218904,
-      "learning_rate": 2.182133397053675e-06,
-      "loss": 0.5256,
+      "epoch": 2.169741697416974,
+      "grad_norm": 0.7361578196639839,
+      "learning_rate": 2.1188949502220987e-06,
+      "loss": 0.555,
       "step": 392
     },
     {
-      "epoch": 2.1712707182320443,
-      "grad_norm": 0.7851921217380164,
-      "learning_rate": 2.1556021870813653e-06,
-      "loss": 0.5521,
+      "epoch": 2.1752767527675276,
+      "grad_norm": 0.7185291233253194,
+      "learning_rate": 2.0925396435598665e-06,
+      "loss": 0.5743,
       "step": 393
     },
     {
-      "epoch": 2.1767955801104972,
-      "grad_norm": 0.6866503114131093,
-      "learning_rate": 2.129188859500459e-06,
-      "loss": 0.5784,
+      "epoch": 2.180811808118081,
+      "grad_norm": 0.6519229097148226,
+      "learning_rate": 2.066305826607911e-06,
+      "loss": 0.5818,
       "step": 394
     },
     {
-      "epoch": 2.18232044198895,
-      "grad_norm": 0.6494728733892837,
-      "learning_rate": 2.102894508977182e-06,
-      "loss": 0.5601,
+      "epoch": 2.1863468634686347,
+      "grad_norm": 0.6790978922388056,
+      "learning_rate": 2.0401945955596206e-06,
+      "loss": 0.517,
       "step": 395
     },
     {
-      "epoch": 2.1878453038674035,
-      "grad_norm": 0.6678705573550681,
-      "learning_rate": 2.0767202252469113e-06,
-      "loss": 0.4986,
+      "epoch": 2.1918819188191883,
+      "grad_norm": 0.6753695790228323,
+      "learning_rate": 2.0142070414860704e-06,
+      "loss": 0.5176,
       "step": 396
     },
     {
-      "epoch": 2.1933701657458564,
-      "grad_norm": 0.7160045418734877,
-      "learning_rate": 2.0506670930690074e-06,
-      "loss": 0.5683,
+      "epoch": 2.197416974169742,
+      "grad_norm": 0.7265806499649696,
+      "learning_rate": 1.9883442502904284e-06,
+      "loss": 0.6055,
       "step": 397
     },
     {
-      "epoch": 2.1988950276243093,
-      "grad_norm": 0.6856778789385509,
-      "learning_rate": 2.0247361921818638e-06,
-      "loss": 0.5593,
+      "epoch": 2.2029520295202953,
+      "grad_norm": 0.6457410100738195,
+      "learning_rate": 1.962607302662582e-06,
+      "loss": 0.5406,
       "step": 398
     },
     {
-      "epoch": 2.2044198895027622,
-      "grad_norm": 0.6486089048078473,
-      "learning_rate": 1.9989285972581595e-06,
-      "loss": 0.5482,
+      "epoch": 2.208487084870849,
+      "grad_norm": 0.625524661837684,
+      "learning_rate": 1.936997274033986e-06,
+      "loss": 0.5076,
       "step": 399
     },
     {
-      "epoch": 2.2099447513812156,
-      "grad_norm": 0.6894974735994067,
-      "learning_rate": 1.9732453778603104e-06,
-      "loss": 0.5486,
+      "epoch": 2.2140221402214024,
+      "grad_norm": 0.637857740685298,
+      "learning_rate": 1.9115152345327154e-06,
+      "loss": 0.5578,
       "step": 400
     },
     {
-      "epoch": 2.2154696132596685,
-      "grad_norm": 0.6135965995020478,
-      "learning_rate": 1.947687598396154e-06,
-      "loss": 0.523,
+      "epoch": 2.2195571955719555,
+      "grad_norm": 0.6101593685630197,
+      "learning_rate": 1.8861622489387555e-06,
+      "loss": 0.4875,
       "step": 401
     },
     {
-      "epoch": 2.2209944751381214,
-      "grad_norm": 0.5824847944758283,
-      "learning_rate": 1.92225631807483e-06,
-      "loss": 0.4781,
+      "epoch": 2.225092250922509,
+      "grad_norm": 0.6518718553669549,
+      "learning_rate": 1.8609393766395083e-06,
+      "loss": 0.57,
       "step": 402
     },
     {
-      "epoch": 2.226519337016575,
-      "grad_norm": 0.6880197469905236,
-      "learning_rate": 1.896952590862886e-06,
-      "loss": 0.5563,
+      "epoch": 2.2306273062730626,
+      "grad_norm": 0.6624958993518161,
+      "learning_rate": 1.8358476715855262e-06,
+      "loss": 0.5176,
       "step": 403
     },
     {
-      "epoch": 2.2320441988950277,
-      "grad_norm": 0.6446697768546276,
-      "learning_rate": 1.8717774654405962e-06,
-      "loss": 0.5182,
+      "epoch": 2.236162361623616,
+      "grad_norm": 0.6719691100692955,
+      "learning_rate": 1.8108881822464697e-06,
+      "loss": 0.5313,
       "step": 404
     },
     {
-      "epoch": 2.2375690607734806,
-      "grad_norm": 0.6867656203180093,
-      "learning_rate": 1.8467319851584952e-06,
-      "loss": 0.5602,
+      "epoch": 2.2416974169741697,
+      "grad_norm": 0.6736416728036624,
+      "learning_rate": 1.7860619515673034e-06,
+      "loss": 0.5625,
       "step": 405
     },
     {
-      "epoch": 2.2430939226519335,
-      "grad_norm": 0.6379130888342771,
-      "learning_rate": 1.8218171879941465e-06,
-      "loss": 0.5657,
+      "epoch": 2.2472324723247232,
+      "grad_norm": 0.6422294559583493,
+      "learning_rate": 1.7613700169247055e-06,
+      "loss": 0.5705,
       "step": 406
     },
     {
-      "epoch": 2.248618784530387,
-      "grad_norm": 0.6345085950204016,
-      "learning_rate": 1.7970341065091246e-06,
-      "loss": 0.5266,
+      "epoch": 2.2527675276752768,
+      "grad_norm": 0.6603028481218244,
+      "learning_rate": 1.7368134100837286e-06,
+      "loss": 0.589,
       "step": 407
     },
     {
-      "epoch": 2.25414364640884,
-      "grad_norm": 0.6637996504745941,
-      "learning_rate": 1.7723837678062083e-06,
-      "loss": 0.5885,
+      "epoch": 2.2583025830258303,
+      "grad_norm": 0.5987866874874727,
+      "learning_rate": 1.7123931571546826e-06,
+      "loss": 0.4857,
       "step": 408
     },
     {
-      "epoch": 2.2596685082872927,
-      "grad_norm": 0.6029850494243252,
-      "learning_rate": 1.7478671934868302e-06,
-      "loss": 0.4729,
+      "epoch": 2.263837638376384,
+      "grad_norm": 0.6594047301752323,
+      "learning_rate": 1.6881102785502618e-06,
+      "loss": 0.5477,
       "step": 409
     },
     {
-      "epoch": 2.265193370165746,
-      "grad_norm": 0.6459276687581147,
-      "learning_rate": 1.7234853996087304e-06,
-      "loss": 0.5752,
+      "epoch": 2.2693726937269374,
+      "grad_norm": 0.6196530000721581,
+      "learning_rate": 1.6639657889429017e-06,
+      "loss": 0.5506,
       "step": 410
     },
     {
-      "epoch": 2.270718232044199,
-      "grad_norm": 0.6226965789608703,
-      "learning_rate": 1.699239396643841e-06,
-      "loss": 0.5411,
+      "epoch": 2.274907749077491,
+      "grad_norm": 0.6722537304232283,
+      "learning_rate": 1.639960697222388e-06,
+      "loss": 0.5797,
       "step": 411
     },
     {
-      "epoch": 2.276243093922652,
-      "grad_norm": 0.6397669563858227,
-      "learning_rate": 1.6751301894364274e-06,
-      "loss": 0.5467,
+      "epoch": 2.280442804428044,
+      "grad_norm": 0.615523426294855,
+      "learning_rate": 1.6160960064536907e-06,
+      "loss": 0.4981,
       "step": 412
     },
     {
-      "epoch": 2.281767955801105,
-      "grad_norm": 0.6608828543106198,
-      "learning_rate": 1.6511587771614208e-06,
-      "loss": 0.5103,
+      "epoch": 2.2859778597785976,
+      "grad_norm": 0.6568878327953758,
+      "learning_rate": 1.5923727138350548e-06,
+      "loss": 0.5263,
       "step": 413
     },
     {
-      "epoch": 2.287292817679558,
-      "grad_norm": 0.6785181930874952,
-      "learning_rate": 1.6273261532830242e-06,
-      "loss": 0.5667,
+      "epoch": 2.291512915129151,
+      "grad_norm": 0.6557176714698197,
+      "learning_rate": 1.5687918106563326e-06,
+      "loss": 0.5977,
       "step": 414
     },
     {
-      "epoch": 2.292817679558011,
-      "grad_norm": 0.6387931680506953,
-      "learning_rate": 1.6036333055135345e-06,
-      "loss": 0.5469,
+      "epoch": 2.2970479704797047,
+      "grad_norm": 0.6343521214286307,
+      "learning_rate": 1.5453542822575624e-06,
+      "loss": 0.4672,
       "step": 415
     },
     {
-      "epoch": 2.298342541436464,
-      "grad_norm": 0.6325771066839061,
-      "learning_rate": 1.5800812157724084e-06,
-      "loss": 0.4923,
+      "epoch": 2.302583025830258,
+      "grad_norm": 0.6160998844576888,
+      "learning_rate": 1.52206110798779e-06,
+      "loss": 0.5385,
       "step": 416
     },
     {
-      "epoch": 2.303867403314917,
-      "grad_norm": 0.6382423549193763,
-      "learning_rate": 1.556670860145567e-06,
-      "loss": 0.536,
+      "epoch": 2.3081180811808117,
+      "grad_norm": 0.6465234486304856,
+      "learning_rate": 1.4989132611641576e-06,
+      "loss": 0.5165,
       "step": 417
     },
     {
-      "epoch": 2.3093922651933703,
-      "grad_norm": 0.655241299309122,
-      "learning_rate": 1.533403208844947e-06,
-      "loss": 0.5437,
+      "epoch": 2.3136531365313653,
+      "grad_norm": 0.6620294899666387,
+      "learning_rate": 1.4759117090312197e-06,
+      "loss": 0.5605,
       "step": 418
     },
     {
-      "epoch": 2.314917127071823,
-      "grad_norm": 0.6805728068716513,
-      "learning_rate": 1.5102792261682813e-06,
-      "loss": 0.5418,
+      "epoch": 2.319188191881919,
+      "grad_norm": 0.6822424027974909,
+      "learning_rate": 1.453057412720536e-06,
+      "loss": 0.5921,
       "step": 419
     },
     {
-      "epoch": 2.320441988950276,
-      "grad_norm": 0.6836442326534425,
-      "learning_rate": 1.487299870459155e-06,
-      "loss": 0.6096,
+      "epoch": 2.3247232472324724,
+      "grad_norm": 0.6384847627809093,
+      "learning_rate": 1.4303513272105057e-06,
+      "loss": 0.5403,
       "step": 420
     },
     {
-      "epoch": 2.3259668508287294,
-      "grad_norm": 0.6284514006381783,
-      "learning_rate": 1.4644660940672628e-06,
-      "loss": 0.534,
+      "epoch": 2.330258302583026,
+      "grad_norm": 0.6719145192710272,
+      "learning_rate": 1.4077944012864636e-06,
+      "loss": 0.5854,
       "step": 421
     },
     {
-      "epoch": 2.3314917127071824,
-      "grad_norm": 0.6642371366553435,
-      "learning_rate": 1.4417788433089596e-06,
-      "loss": 0.5647,
+      "epoch": 2.3357933579335795,
+      "grad_norm": 0.640451661360814,
+      "learning_rate": 1.3853875775010355e-06,
+      "loss": 0.517,
       "step": 422
     },
     {
-      "epoch": 2.3370165745856353,
-      "grad_norm": 0.6541492771250856,
-      "learning_rate": 1.4192390584280347e-06,
-      "loss": 0.5209,
+      "epoch": 2.341328413284133,
+      "grad_norm": 0.6398126379640671,
+      "learning_rate": 1.3631317921347564e-06,
+      "loss": 0.5791,
       "step": 423
     },
     {
-      "epoch": 2.3425414364640886,
-      "grad_norm": 0.67885347112545,
-      "learning_rate": 1.3968476735567392e-06,
-      "loss": 0.5806,
+      "epoch": 2.3468634686346865,
+      "grad_norm": 0.6487742664028805,
+      "learning_rate": 1.3410279751569399e-06,
+      "loss": 0.5157,
       "step": 424
     },
     {
-      "epoch": 2.3480662983425415,
-      "grad_norm": 0.6716876869006629,
-      "learning_rate": 1.3746056166770872e-06,
-      "loss": 0.5247,
+      "epoch": 2.35239852398524,
+      "grad_norm": 0.6671465550992883,
+      "learning_rate": 1.3190770501868243e-06,
+      "loss": 0.5588,
       "step": 425
     },
     {
-      "epoch": 2.3535911602209945,
-      "grad_norm": 0.6934129162480718,
-      "learning_rate": 1.352513809582377e-06,
-      "loss": 0.5466,
+      "epoch": 2.357933579335793,
+      "grad_norm": 0.6454481634549197,
+      "learning_rate": 1.297279934454978e-06,
+      "loss": 0.5026,
       "step": 426
     },
     {
-      "epoch": 2.3591160220994474,
-      "grad_norm": 0.6400098878071714,
-      "learning_rate": 1.330573167839005e-06,
-      "loss": 0.4947,
+      "epoch": 2.3634686346863467,
+      "grad_norm": 0.6127427807997947,
+      "learning_rate": 1.2756375387649717e-06,
+      "loss": 0.5454,
       "step": 427
     },
     {
-      "epoch": 2.3646408839779007,
-      "grad_norm": 0.6163300040150086,
-      "learning_rate": 1.3087846007485134e-06,
-      "loss": 0.5675,
+      "epoch": 2.3690036900369003,
+      "grad_norm": 0.6438945248930303,
+      "learning_rate": 1.25415076745532e-06,
+      "loss": 0.545,
       "step": 428
     },
     {
-      "epoch": 2.3701657458563536,
-      "grad_norm": 0.6148189692207581,
-      "learning_rate": 1.2871490113099066e-06,
-      "loss": 0.5162,
+      "epoch": 2.374538745387454,
+      "grad_norm": 0.6124407692016695,
+      "learning_rate": 1.2328205183616964e-06,
+      "loss": 0.5189,
       "step": 429
     },
     {
-      "epoch": 2.3756906077348066,
-      "grad_norm": 0.6296512007771748,
-      "learning_rate": 1.2656672961822285e-06,
-      "loss": 0.5354,
+      "epoch": 2.3800738007380073,
+      "grad_norm": 0.6639137528285265,
+      "learning_rate": 1.2116476827794104e-06,
+      "loss": 0.549,
       "step": 430
     },
     {
-      "epoch": 2.3812154696132595,
-      "grad_norm": 0.6659353506949467,
-      "learning_rate": 1.2443403456474017e-06,
-      "loss": 0.5558,
+      "epoch": 2.385608856088561,
+      "grad_norm": 0.6472668638709654,
+      "learning_rate": 1.1906331454261704e-06,
+      "loss": 0.5616,
       "step": 431
     },
     {
-      "epoch": 2.386740331491713,
-      "grad_norm": 0.6250696714712421,
-      "learning_rate": 1.223169043573325e-06,
-      "loss": 0.5161,
+      "epoch": 2.3911439114391144,
+      "grad_norm": 0.6382738061476565,
+      "learning_rate": 1.1697777844051105e-06,
+      "loss": 0.5069,
       "step": 432
     },
     {
-      "epoch": 2.3922651933701657,
-      "grad_norm": 0.628285669477399,
-      "learning_rate": 1.2021542673772584e-06,
-      "loss": 0.5309,
+      "epoch": 2.396678966789668,
+      "grad_norm": 0.6095694583232508,
+      "learning_rate": 1.1490824711681026e-06,
+      "loss": 0.5096,
       "step": 433
     },
     {
-      "epoch": 2.3977900552486187,
-      "grad_norm": 0.6470067569677168,
-      "learning_rate": 1.1812968879894387e-06,
-      "loss": 0.5133,
+      "epoch": 2.4022140221402215,
+      "grad_norm": 0.6389566278508614,
+      "learning_rate": 1.1285480704793378e-06,
+      "loss": 0.5554,
       "step": 434
     },
     {
-      "epoch": 2.403314917127072,
-      "grad_norm": 0.6690982254466323,
-      "learning_rate": 1.1605977698170001e-06,
-      "loss": 0.5498,
+      "epoch": 2.407749077490775,
+      "grad_norm": 0.6565575087635012,
+      "learning_rate": 1.1081754403792e-06,
+      "loss": 0.5289,
       "step": 435
     },
     {
-      "epoch": 2.408839779005525,
-      "grad_norm": 0.6773043112171515,
-      "learning_rate": 1.1400577707081467e-06,
-      "loss": 0.5279,
+      "epoch": 2.4132841328413286,
+      "grad_norm": 0.6298283487396668,
+      "learning_rate": 1.0879654321484012e-06,
+      "loss": 0.5732,
       "step": 436
     },
     {
-      "epoch": 2.414364640883978,
-      "grad_norm": 0.6465508208097143,
-      "learning_rate": 1.1196777419165927e-06,
-      "loss": 0.5645,
+      "epoch": 2.4188191881918817,
+      "grad_norm": 0.6546518202275479,
+      "learning_rate": 1.067918890272419e-06,
+      "loss": 0.5408,
       "step": 437
     },
     {
-      "epoch": 2.4198895027624308,
-      "grad_norm": 0.6754843790206556,
-      "learning_rate": 1.0994585280662978e-06,
-      "loss": 0.5709,
+      "epoch": 2.4243542435424352,
+      "grad_norm": 0.6656591924546393,
+      "learning_rate": 1.0480366524062041e-06,
+      "loss": 0.551,
       "step": 438
     },
     {
-      "epoch": 2.425414364640884,
-      "grad_norm": 0.6373867979221823,
-      "learning_rate": 1.0794009671164484e-06,
-      "loss": 0.5301,
+      "epoch": 2.4298892988929888,
+      "grad_norm": 0.6298524026967492,
+      "learning_rate": 1.0283195493391823e-06,
+      "loss": 0.5011,
       "step": 439
     },
     {
-      "epoch": 2.430939226519337,
-      "grad_norm": 0.6529002066290224,
-      "learning_rate": 1.0595058903267357e-06,
-      "loss": 0.5283,
+      "epoch": 2.4354243542435423,
+      "grad_norm": 0.6525468703564422,
+      "learning_rate": 1.008768404960535e-06,
+      "loss": 0.5737,
       "step": 440
     },
     {
-      "epoch": 2.43646408839779,
-      "grad_norm": 0.6785490644572046,
-      "learning_rate": 1.0397741222229057e-06,
-      "loss": 0.5368,
+      "epoch": 2.440959409594096,
+      "grad_norm": 0.6544594096244687,
+      "learning_rate": 9.893840362247809e-07,
+      "loss": 0.5503,
       "step": 441
     },
     {
-      "epoch": 2.441988950276243,
-      "grad_norm": 0.6906247289237696,
-      "learning_rate": 1.0202064805625883e-06,
-      "loss": 0.5811,
+      "epoch": 2.4464944649446494,
+      "grad_norm": 0.6318395119148414,
+      "learning_rate": 9.701672531176287e-07,
+      "loss": 0.5224,
       "step": 442
     },
     {
-      "epoch": 2.447513812154696,
-      "grad_norm": 0.6332990091280457,
-      "learning_rate": 1.0008037763014033e-06,
-      "loss": 0.5324,
+      "epoch": 2.452029520295203,
+      "grad_norm": 0.6196921977020754,
+      "learning_rate": 9.511188586221376e-07,
+      "loss": 0.545,
       "step": 443
     },
     {
-      "epoch": 2.453038674033149,
-      "grad_norm": 0.5981606081277429,
-      "learning_rate": 9.815668135593548e-07,
-      "loss": 0.4981,
+      "epoch": 2.4575645756457565,
+      "grad_norm": 0.6327900918490961,
+      "learning_rate": 9.322396486851626e-07,
+      "loss": 0.5424,
       "step": 444
     },
     {
-      "epoch": 2.458563535911602,
-      "grad_norm": 0.6658535935861339,
-      "learning_rate": 9.624963895874995e-07,
-      "loss": 0.5804,
+      "epoch": 2.46309963099631,
+      "grad_norm": 0.6363476446254773,
+      "learning_rate": 9.135304121840976e-07,
+      "loss": 0.5538,
       "step": 445
     },
     {
-      "epoch": 2.4640883977900554,
-      "grad_norm": 0.6173994326467493,
-      "learning_rate": 9.435932947349169e-07,
-      "loss": 0.5223,
+      "epoch": 2.4686346863468636,
+      "grad_norm": 0.6313076689214548,
+      "learning_rate": 8.949919308939081e-07,
+      "loss": 0.5218,
       "step": 446
     },
     {
-      "epoch": 2.4696132596685083,
-      "grad_norm": 0.6696683287460099,
-      "learning_rate": 9.248583124159438e-07,
-      "loss": 0.5488,
+      "epoch": 2.474169741697417,
+      "grad_norm": 0.657993153154449,
+      "learning_rate": 8.766249794544662e-07,
+      "loss": 0.5321,
       "step": 447
     },
     {
-      "epoch": 2.4751381215469612,
-      "grad_norm": 0.681352463917548,
-      "learning_rate": 9.062922190777079e-07,
-      "loss": 0.5377,
+      "epoch": 2.4797047970479706,
+      "grad_norm": 0.6418969902696497,
+      "learning_rate": 8.584303253381848e-07,
+      "loss": 0.5194,
       "step": 448
     },
     {
-      "epoch": 2.4806629834254146,
-      "grad_norm": 0.6413299506586427,
-      "learning_rate": 8.878957841679542e-07,
-      "loss": 0.538,
+      "epoch": 2.485239852398524,
+      "grad_norm": 0.6043587008301916,
+      "learning_rate": 8.404087288179425e-07,
+      "loss": 0.5134,
       "step": 449
     },
     {
-      "epoch": 2.4861878453038675,
-      "grad_norm": 0.5976825742538746,
-      "learning_rate": 8.696697701031543e-07,
-      "loss": 0.4698,
+      "epoch": 2.4907749077490777,
+      "grad_norm": 0.6431540180087669,
+      "learning_rate": 8.225609429353187e-07,
+      "loss": 0.5348,
       "step": 450
     },
     {
-      "epoch": 2.4917127071823204,
-      "grad_norm": 0.6586286738339465,
-      "learning_rate": 8.516149322369055e-07,
-      "loss": 0.5529,
+      "epoch": 2.496309963099631,
+      "grad_norm": 0.6268189919104177,
+      "learning_rate": 8.048877134691269e-07,
+      "loss": 0.5043,
       "step": 451
     },
     {
-      "epoch": 2.4972375690607733,
-      "grad_norm": 0.6374084552099051,
-      "learning_rate": 8.337320188286318e-07,
-      "loss": 0.4934,
+      "epoch": 2.5018450184501844,
+      "grad_norm": 0.6178939835577518,
+      "learning_rate": 7.873897789042523e-07,
+      "loss": 0.5165,
       "step": 452
     },
     {
-      "epoch": 2.5027624309392267,
-      "grad_norm": 0.6429007341483025,
-      "learning_rate": 8.160217710125661e-07,
-      "loss": 0.5544,
+      "epoch": 2.507380073800738,
+      "grad_norm": 0.6747581351563309,
+      "learning_rate": 7.700678704007947e-07,
+      "loss": 0.6184,
       "step": 453
     },
     {
-      "epoch": 2.5082872928176796,
-      "grad_norm": 0.6664059102374921,
-      "learning_rate": 7.984849227670421e-07,
-      "loss": 0.6073,
+      "epoch": 2.5129151291512914,
+      "grad_norm": 0.6379894631440596,
+      "learning_rate": 7.529227117635135e-07,
+      "loss": 0.5297,
       "step": 454
     },
     {
-      "epoch": 2.5138121546961325,
-      "grad_norm": 0.6392153790680253,
-      "learning_rate": 7.811222008840719e-07,
-      "loss": 0.5542,
+      "epoch": 2.518450184501845,
+      "grad_norm": 0.6345560431082808,
+      "learning_rate": 7.35955019411585e-07,
+      "loss": 0.5723,
       "step": 455
     },
     {
-      "epoch": 2.5193370165745854,
-      "grad_norm": 0.6018857165135849,
-      "learning_rate": 7.639343249392256e-07,
-      "loss": 0.5285,
+      "epoch": 2.5239852398523985,
+      "grad_norm": 0.7041260690229593,
+      "learning_rate": 7.191655023486682e-07,
+      "loss": 0.5518,
       "step": 456
     },
     {
-      "epoch": 2.5248618784530388,
-      "grad_norm": 0.7674240660472489,
-      "learning_rate": 7.469220072618094e-07,
-      "loss": 0.5761,
+      "epoch": 2.529520295202952,
+      "grad_norm": 0.6596174352194049,
+      "learning_rate": 7.02554862133275e-07,
+      "loss": 0.5507,
       "step": 457
     },
     {
-      "epoch": 2.5303867403314917,
-      "grad_norm": 0.6496289775570502,
-      "learning_rate": 7.300859529053422e-07,
-      "loss": 0.5525,
+      "epoch": 2.5350553505535056,
+      "grad_norm": 0.6491422253781083,
+      "learning_rate": 6.86123792849458e-07,
+      "loss": 0.5146,
       "step": 458
     },
     {
-      "epoch": 2.5359116022099446,
-      "grad_norm": 0.6547899116663247,
-      "learning_rate": 7.13426859618338e-07,
-      "loss": 0.4973,
+      "epoch": 2.540590405904059,
+      "grad_norm": 0.6702481216161904,
+      "learning_rate": 6.698729810778065e-07,
+      "loss": 0.5837,
       "step": 459
     },
     {
-      "epoch": 2.541436464088398,
-      "grad_norm": 0.6505467220043211,
-      "learning_rate": 6.969454178153923e-07,
-      "loss": 0.5752,
+      "epoch": 2.5461254612546127,
+      "grad_norm": 0.6189552978352973,
+      "learning_rate": 6.53803105866761e-07,
+      "loss": 0.5343,
       "step": 460
     },
     {
-      "epoch": 2.546961325966851,
-      "grad_norm": 0.6735594708393216,
-      "learning_rate": 6.806423105485576e-07,
-      "loss": 0.5679,
+      "epoch": 2.551660516605166,
+      "grad_norm": 0.6835144543692675,
+      "learning_rate": 6.379148387042317e-07,
+      "loss": 0.5462,
       "step": 461
     },
     {
-      "epoch": 2.552486187845304,
-      "grad_norm": 0.7191676631486709,
-      "learning_rate": 6.645182134790467e-07,
-      "loss": 0.5415,
+      "epoch": 2.5571955719557193,
+      "grad_norm": 0.6464350965970691,
+      "learning_rate": 6.222088434895462e-07,
+      "loss": 0.5381,
       "step": 462
     },
     {
-      "epoch": 2.558011049723757,
-      "grad_norm": 0.6766548740002727,
-      "learning_rate": 6.485737948492237e-07,
-      "loss": 0.5656,
+      "epoch": 2.562730627306273,
+      "grad_norm": 0.6957103842105568,
+      "learning_rate": 6.066857765057055e-07,
+      "loss": 0.6151,
       "step": 463
     },
     {
-      "epoch": 2.56353591160221,
-      "grad_norm": 0.6610159478158815,
-      "learning_rate": 6.328097154549146e-07,
-      "loss": 0.576,
+      "epoch": 2.5682656826568264,
+      "grad_norm": 0.6030789961609527,
+      "learning_rate": 5.9134628639196e-07,
+      "loss": 0.5061,
       "step": 464
     },
     {
-      "epoch": 2.569060773480663,
-      "grad_norm": 0.6227047055678046,
-      "learning_rate": 6.172266286180162e-07,
-      "loss": 0.5068,
+      "epoch": 2.57380073800738,
+      "grad_norm": 0.6317615119983371,
+      "learning_rate": 5.76191014116711e-07,
+      "loss": 0.507,
       "step": 465
     },
     {
-      "epoch": 2.574585635359116,
-      "grad_norm": 0.6389412833805688,
-      "learning_rate": 6.018251801594232e-07,
-      "loss": 0.5258,
+      "epoch": 2.5793357933579335,
+      "grad_norm": 0.6435467983289933,
+      "learning_rate": 5.612205929507209e-07,
+      "loss": 0.5501,
       "step": 466
     },
     {
-      "epoch": 2.580110497237569,
-      "grad_norm": 0.6232609013170282,
-      "learning_rate": 5.866060083722624e-07,
-      "loss": 0.5355,
+      "epoch": 2.584870848708487,
+      "grad_norm": 0.6358149902010037,
+      "learning_rate": 5.464356484406535e-07,
+      "loss": 0.5211,
       "step": 467
     },
     {
-      "epoch": 2.585635359116022,
-      "grad_norm": 0.6278311488371855,
-      "learning_rate": 5.715697439954432e-07,
-      "loss": 0.5126,
+      "epoch": 2.5904059040590406,
+      "grad_norm": 0.6195286275874121,
+      "learning_rate": 5.318367983829393e-07,
+      "loss": 0.5032,
       "step": 468
     },
     {
-      "epoch": 2.591160220994475,
-      "grad_norm": 0.6077201096302113,
-      "learning_rate": 5.567170101875074e-07,
-      "loss": 0.4949,
+      "epoch": 2.595940959409594,
+      "grad_norm": 0.6692400503094423,
+      "learning_rate": 5.174246527979532e-07,
+      "loss": 0.5574,
       "step": 469
     },
     {
-      "epoch": 2.596685082872928,
-      "grad_norm": 0.6686676733895326,
-      "learning_rate": 5.420484225008138e-07,
-      "loss": 0.5645,
+      "epoch": 2.6014760147601477,
+      "grad_norm": 0.6342700435947363,
+      "learning_rate": 5.031998139045352e-07,
+      "loss": 0.5523,
       "step": 470
     },
     {
-      "epoch": 2.6022099447513813,
-      "grad_norm": 0.6427639590105313,
-      "learning_rate": 5.275645888560233e-07,
-      "loss": 0.5939,
+      "epoch": 2.607011070110701,
+      "grad_norm": 0.6412069996618961,
+      "learning_rate": 4.891628760948114e-07,
+      "loss": 0.5423,
       "step": 471
     },
     {
-      "epoch": 2.6077348066298343,
-      "grad_norm": 0.5974938612500432,
-      "learning_rate": 5.132661095168994e-07,
-      "loss": 0.5061,
+      "epoch": 2.6125461254612548,
+      "grad_norm": 0.628500100341704,
+      "learning_rate": 4.753144259093734e-07,
+      "loss": 0.5451,
       "step": 472
     },
     {
-      "epoch": 2.613259668508287,
-      "grad_norm": 0.6408484309671157,
-      "learning_rate": 4.991535770654449e-07,
-      "loss": 0.5004,
+      "epoch": 2.6180811808118083,
+      "grad_norm": 0.5904702310704353,
+      "learning_rate": 4.6165504201275635e-07,
+      "loss": 0.4915,
       "step": 473
     },
     {
-      "epoch": 2.6187845303867405,
-      "grad_norm": 0.6464185272075329,
-      "learning_rate": 4.852275763773251e-07,
-      "loss": 0.5614,
+      "epoch": 2.623616236162362,
+      "grad_norm": 0.7363244318199392,
+      "learning_rate": 4.481852951692672e-07,
+      "loss": 0.6614,
       "step": 474
     },
     {
-      "epoch": 2.6243093922651934,
-      "grad_norm": 0.7272955176468657,
-      "learning_rate": 4.71488684597643e-07,
-      "loss": 0.6543,
+      "epoch": 2.6291512915129154,
+      "grad_norm": 0.6437098900248531,
+      "learning_rate": 4.349057482191299e-07,
+      "loss": 0.5644,
       "step": 475
     },
     {
-      "epoch": 2.6298342541436464,
-      "grad_norm": 0.6731985657132716,
-      "learning_rate": 4.57937471117012e-07,
-      "loss": 0.5776,
+      "epoch": 2.6346863468634685,
+      "grad_norm": 0.6735179848745367,
+      "learning_rate": 4.2181695605497066e-07,
+      "loss": 0.6371,
       "step": 476
     },
     {
-      "epoch": 2.6353591160220997,
-      "grad_norm": 0.6348396347822451,
-      "learning_rate": 4.445744975479627e-07,
-      "loss": 0.6063,
+      "epoch": 2.640221402214022,
+      "grad_norm": 0.6166395492738032,
+      "learning_rate": 4.089194655986306e-07,
+      "loss": 0.5439,
       "step": 477
     },
     {
-      "epoch": 2.6408839779005526,
-      "grad_norm": 0.635930709870673,
-      "learning_rate": 4.3140031770166457e-07,
-      "loss": 0.5416,
+      "epoch": 2.6457564575645756,
+      "grad_norm": 0.5978015771749204,
+      "learning_rate": 3.9621381577830855e-07,
+      "loss": 0.4831,
       "step": 478
     },
     {
-      "epoch": 2.6464088397790055,
-      "grad_norm": 0.6098205107229787,
-      "learning_rate": 4.184154775649768e-07,
-      "loss": 0.5063,
+      "epoch": 2.651291512915129,
+      "grad_norm": 0.6462673792890513,
+      "learning_rate": 3.837005375060482e-07,
+      "loss": 0.597,
       "step": 479
     },
     {
-      "epoch": 2.6519337016574585,
-      "grad_norm": 0.6607556626481141,
-      "learning_rate": 4.0562051527781534e-07,
-      "loss": 0.5875,
+      "epoch": 2.6568265682656826,
+      "grad_norm": 0.6130295271007709,
+      "learning_rate": 3.7138015365554834e-07,
+      "loss": 0.5106,
       "step": 480
     },
     {
-      "epoch": 2.6574585635359114,
-      "grad_norm": 0.6324042696562345,
-      "learning_rate": 3.930159611108603e-07,
-      "loss": 0.4783,
+      "epoch": 2.662361623616236,
+      "grad_norm": 0.6671222024328364,
+      "learning_rate": 3.592531790403159e-07,
+      "loss": 0.5363,
       "step": 481
     },
     {
-      "epoch": 2.6629834254143647,
-      "grad_norm": 0.6938336791108509,
-      "learning_rate": 3.8060233744356634e-07,
-      "loss": 0.5537,
+      "epoch": 2.6678966789667897,
+      "grad_norm": 0.6138396843380316,
+      "learning_rate": 3.473201203921578e-07,
+      "loss": 0.5469,
       "step": 482
     },
     {
-      "epoch": 2.6685082872928176,
-      "grad_norm": 0.6325704993286224,
-      "learning_rate": 3.683801587425251e-07,
-      "loss": 0.5502,
+      "epoch": 2.6734317343173433,
+      "grad_norm": 0.6660598208965025,
+      "learning_rate": 3.355814763399973e-07,
+      "loss": 0.581,
       "step": 483
     },
     {
-      "epoch": 2.6740331491712706,
-      "grad_norm": 0.6934512585984158,
-      "learning_rate": 3.563499315401386e-07,
-      "loss": 0.5962,
+      "epoch": 2.678966789667897,
+      "grad_norm": 0.6108073985826324,
+      "learning_rate": 3.2403773738905185e-07,
+      "loss": 0.5273,
       "step": 484
     },
     {
-      "epoch": 2.679558011049724,
-      "grad_norm": 0.6128926230516574,
-      "learning_rate": 3.4451215441362264e-07,
-      "loss": 0.5282,
+      "epoch": 2.6845018450184504,
+      "grad_norm": 0.7379112157721128,
+      "learning_rate": 3.1268938590032495e-07,
+      "loss": 0.5626,
       "step": 485
     },
     {
-      "epoch": 2.685082872928177,
-      "grad_norm": 0.6812281139244459,
-      "learning_rate": 3.328673179643555e-07,
-      "loss": 0.5655,
+      "epoch": 2.6900369003690034,
+      "grad_norm": 0.6732052466349552,
+      "learning_rate": 3.015368960704584e-07,
+      "loss": 0.5891,
       "step": 486
     },
     {
-      "epoch": 2.6906077348066297,
-      "grad_norm": 0.6773585472582779,
-      "learning_rate": 3.214159047975324e-07,
-      "loss": 0.5818,
+      "epoch": 2.695571955719557,
+      "grad_norm": 0.6659946641749191,
+      "learning_rate": 2.905807339119138e-07,
+      "loss": 0.5258,
       "step": 487
     },
     {
-      "epoch": 2.696132596685083,
-      "grad_norm": 0.6516795165171537,
-      "learning_rate": 3.101583895021731e-07,
-      "loss": 0.5326,
+      "epoch": 2.7011070110701105,
+      "grad_norm": 0.6952105961310837,
+      "learning_rate": 2.798213572335001e-07,
+      "loss": 0.5581,
       "step": 488
     },
     {
-      "epoch": 2.701657458563536,
-      "grad_norm": 0.6662378644824801,
-      "learning_rate": 2.990952386314505e-07,
-      "loss": 0.5417,
+      "epoch": 2.706642066420664,
+      "grad_norm": 0.6365348737573899,
+      "learning_rate": 2.6925921562124867e-07,
+      "loss": 0.5311,
       "step": 489
     },
     {
-      "epoch": 2.707182320441989,
-      "grad_norm": 0.6210703639589708,
-      "learning_rate": 2.8822691068335515e-07,
-      "loss": 0.4857,
+      "epoch": 2.7121771217712176,
+      "grad_norm": 0.5766266659794174,
+      "learning_rate": 2.5889475041961767e-07,
+      "loss": 0.4352,
       "step": 490
     },
     {
-      "epoch": 2.712707182320442,
-      "grad_norm": 0.6196505032930314,
-      "learning_rate": 2.7755385608169374e-07,
-      "loss": 0.4956,
+      "epoch": 2.717712177121771,
+      "grad_norm": 0.7319731694850389,
+      "learning_rate": 2.487283947130609e-07,
+      "loss": 0.5061,
       "step": 491
     },
     {
-      "epoch": 2.718232044198895,
-      "grad_norm": 0.713069198760707,
-      "learning_rate": 2.6707651715742075e-07,
-      "loss": 0.5266,
+      "epoch": 2.7232472324723247,
+      "grad_norm": 0.6859080435441426,
+      "learning_rate": 2.3876057330792344e-07,
+      "loss": 0.55,
       "step": 492
     },
     {
-      "epoch": 2.723756906077348,
-      "grad_norm": 0.6649253442646119,
-      "learning_rate": 2.567953281303059e-07,
-      "loss": 0.5354,
+      "epoch": 2.7287822878228782,
+      "grad_norm": 0.6487814667155516,
+      "learning_rate": 2.289917027146943e-07,
+      "loss": 0.5418,
       "step": 493
     },
     {
-      "epoch": 2.729281767955801,
-      "grad_norm": 0.6407239217065679,
-      "learning_rate": 2.4671071509094367e-07,
-      "loss": 0.5376,
+      "epoch": 2.734317343173432,
+      "grad_norm": 0.6761313635926454,
+      "learning_rate": 2.1942219113060215e-07,
+      "loss": 0.5819,
       "step": 494
     },
     {
-      "epoch": 2.734806629834254,
-      "grad_norm": 0.6804096600469481,
-      "learning_rate": 2.368230959830875e-07,
-      "loss": 0.6262,
+      "epoch": 2.7398523985239853,
+      "grad_norm": 0.6242690303129593,
+      "learning_rate": 2.1005243842255552e-07,
+      "loss": 0.5655,
       "step": 495
     },
     {
-      "epoch": 2.7403314917127073,
-      "grad_norm": 0.6396360467624123,
-      "learning_rate": 2.2713288058633321e-07,
-      "loss": 0.5281,
+      "epoch": 2.745387453874539,
+      "grad_norm": 0.6222359002332329,
+      "learning_rate": 2.0088283611044034e-07,
+      "loss": 0.5279,
       "step": 496
     },
     {
-      "epoch": 2.74585635359116,
-      "grad_norm": 0.6476227999392966,
-      "learning_rate": 2.1764047049913528e-07,
-      "loss": 0.5318,
+      "epoch": 2.7509225092250924,
+      "grad_norm": 0.6148480903080046,
+      "learning_rate": 1.919137673507543e-07,
+      "loss": 0.5385,
       "step": 497
     },
     {
-      "epoch": 2.751381215469613,
-      "grad_norm": 0.6122643116968306,
-      "learning_rate": 2.0834625912216133e-07,
-      "loss": 0.4926,
+      "epoch": 2.756457564575646,
+      "grad_norm": 0.6004801018266677,
+      "learning_rate": 1.8314560692059836e-07,
+      "loss": 0.4995,
       "step": 498
     },
     {
-      "epoch": 2.7569060773480665,
-      "grad_norm": 0.6528467839639731,
-      "learning_rate": 1.992506316419912e-07,
-      "loss": 0.546,
+      "epoch": 2.7619926199261995,
+      "grad_norm": 0.6362099075257641,
+      "learning_rate": 1.745787212020178e-07,
+      "loss": 0.5248,
       "step": 499
     },
     {
-      "epoch": 2.7624309392265194,
-      "grad_norm": 0.61759278073022,
-      "learning_rate": 1.9035396501515148e-07,
-      "loss": 0.4984,
+      "epoch": 2.767527675276753,
+      "grad_norm": 0.6405183086515418,
+      "learning_rate": 1.6621346816668993e-07,
+      "loss": 0.5456,
       "step": 500
     },
     {
-      "epoch": 2.7679558011049723,
-      "grad_norm": 0.6443700794544784,
-      "learning_rate": 1.8165662795249172e-07,
-      "loss": 0.5885,
+      "epoch": 2.773062730627306,
+      "grad_norm": 0.6383996319685834,
+      "learning_rate": 1.5805019736097105e-07,
+      "loss": 0.5484,
       "step": 501
     },
     {
-      "epoch": 2.7734806629834257,
-      "grad_norm": 0.6316205180357609,
-      "learning_rate": 1.7315898090390748e-07,
-      "loss": 0.5291,
+      "epoch": 2.7785977859778597,
+      "grad_norm": 0.6484780190882176,
+      "learning_rate": 1.500892498912826e-07,
+      "loss": 0.5614,
       "step": 502
     },
     {
-      "epoch": 2.7790055248618786,
-      "grad_norm": 0.6628524099587767,
-      "learning_rate": 1.6486137604339813e-07,
-      "loss": 0.569,
+      "epoch": 2.784132841328413,
+      "grad_norm": 0.6104614408300383,
+      "learning_rate": 1.4233095840986756e-07,
+      "loss": 0.53,
       "step": 503
     },
     {
-      "epoch": 2.7845303867403315,
-      "grad_norm": 0.6149733294537271,
-      "learning_rate": 1.5676415725447424e-07,
-      "loss": 0.4872,
+      "epoch": 2.7896678966789668,
+      "grad_norm": 0.602450823572722,
+      "learning_rate": 1.3477564710088097e-07,
+      "loss": 0.4869,
       "step": 504
     },
     {
-      "epoch": 2.7900552486187844,
-      "grad_norm": 0.6199230062873984,
-      "learning_rate": 1.4886766011590449e-07,
-      "loss": 0.5333,
+      "epoch": 2.7952029520295203,
+      "grad_norm": 0.6213278708400987,
+      "learning_rate": 1.2742363166685035e-07,
+      "loss": 0.5867,
       "step": 505
     },
     {
-      "epoch": 2.7955801104972373,
-      "grad_norm": 0.65667152629127,
-      "learning_rate": 1.4117221188780616e-07,
-      "loss": 0.597,
+      "epoch": 2.800738007380074,
+      "grad_norm": 0.6572722468206588,
+      "learning_rate": 1.2027521931548214e-07,
+      "loss": 0.5681,
       "step": 506
     },
     {
-      "epoch": 2.8011049723756907,
-      "grad_norm": 0.6818138783253694,
-      "learning_rate": 1.3367813149808728e-07,
-      "loss": 0.5724,
+      "epoch": 2.8062730627306274,
+      "grad_norm": 0.6458832457141599,
+      "learning_rate": 1.1333070874682217e-07,
+      "loss": 0.5781,
       "step": 507
     },
     {
-      "epoch": 2.8066298342541436,
-      "grad_norm": 0.6178280944589913,
-      "learning_rate": 1.2638572952922478e-07,
-      "loss": 0.5352,
+      "epoch": 2.811808118081181,
+      "grad_norm": 0.6173074533807742,
+      "learning_rate": 1.0659039014077943e-07,
+      "loss": 0.5485,
       "step": 508
     },
     {
-      "epoch": 2.8121546961325965,
-      "grad_norm": 0.6134027399269469,
-      "learning_rate": 1.192953082053927e-07,
-      "loss": 0.5361,
+      "epoch": 2.8173431734317345,
+      "grad_norm": 0.576088238404349,
+      "learning_rate": 1.0005454514499413e-07,
+      "loss": 0.4937,
       "step": 509
     },
     {
-      "epoch": 2.81767955801105,
-      "grad_norm": 0.5711150946348813,
-      "learning_rate": 1.1240716137994045e-07,
-      "loss": 0.4995,
+      "epoch": 2.8228782287822876,
+      "grad_norm": 0.6463368067791495,
+      "learning_rate": 9.372344686307655e-08,
+      "loss": 0.5686,
       "step": 510
     },
     {
-      "epoch": 2.8232044198895028,
-      "grad_norm": 0.6559442348032254,
-      "learning_rate": 1.0572157452321097e-07,
-      "loss": 0.5625,
+      "epoch": 2.828413284132841,
+      "grad_norm": 0.6380928953673762,
+      "learning_rate": 8.759735984318896e-08,
+      "loss": 0.4994,
       "step": 511
     },
     {
-      "epoch": 2.8287292817679557,
-      "grad_norm": 0.6552526778451414,
-      "learning_rate": 9.923882471071123e-08,
-      "loss": 0.5439,
+      "epoch": 2.8339483394833946,
+      "grad_norm": 0.6305602061857106,
+      "learning_rate": 8.167654006699444e-08,
+      "loss": 0.5506,
       "step": 512
     },
     {
-      "epoch": 2.834254143646409,
-      "grad_norm": 0.6605300263231985,
-      "learning_rate": 9.295918061163034e-08,
-      "loss": 0.5461,
+      "epoch": 2.839483394833948,
+      "grad_norm": 0.6225751887694984,
+      "learning_rate": 7.59612349389599e-08,
+      "loss": 0.5042,
       "step": 513
     },
     {
-      "epoch": 2.839779005524862,
-      "grad_norm": 0.6591460854201797,
-      "learning_rate": 8.688290247770071e-08,
-      "loss": 0.4975,
+      "epoch": 2.8450184501845017,
+      "grad_norm": 0.6644219206184678,
+      "learning_rate": 7.04516832760177e-08,
+      "loss": 0.5969,
       "step": 514
     },
     {
-      "epoch": 2.845303867403315,
-      "grad_norm": 0.6367498607019844,
-      "learning_rate": 8.101024213241826e-08,
-      "loss": 0.5556,
+      "epoch": 2.8505535055350553,
+      "grad_norm": 0.6453523828993342,
+      "learning_rate": 6.514811529758747e-08,
+      "loss": 0.5817,
       "step": 515
     },
     {
-      "epoch": 2.8508287292817682,
-      "grad_norm": 0.654902042221914,
-      "learning_rate": 7.534144296060142e-08,
-      "loss": 0.5925,
+      "epoch": 2.856088560885609,
+      "grad_norm": 0.6229940991961801,
+      "learning_rate": 6.005075261595495e-08,
+      "loss": 0.537,
       "step": 516
     },
     {
-      "epoch": 2.856353591160221,
-      "grad_norm": 0.6837595818853758,
-      "learning_rate": 6.987673989830523e-08,
-      "loss": 0.5811,
+      "epoch": 2.8616236162361623,
+      "grad_norm": 0.6614904718651109,
+      "learning_rate": 5.515980822701439e-08,
+      "loss": 0.5757,
       "step": 517
     },
     {
-      "epoch": 2.861878453038674,
-      "grad_norm": 0.6227316374587448,
-      "learning_rate": 6.461635942308641e-08,
-      "loss": 0.5371,
+      "epoch": 2.867158671586716,
+      "grad_norm": 0.6494054996388359,
+      "learning_rate": 5.047548650136513e-08,
+      "loss": 0.5035,
       "step": 518
     },
     {
-      "epoch": 2.867403314917127,
-      "grad_norm": 0.6780391903150306,
-      "learning_rate": 5.9560519544614725e-08,
-      "loss": 0.4954,
+      "epoch": 2.8726937269372694,
+      "grad_norm": 0.5994286550317681,
+      "learning_rate": 4.599798317577342e-08,
+      "loss": 0.4803,
       "step": 519
     },
     {
-      "epoch": 2.87292817679558,
-      "grad_norm": 0.6322452524518853,
-      "learning_rate": 5.47094297956402e-08,
-      "loss": 0.5109,
+      "epoch": 2.878228782287823,
+      "grad_norm": 0.6745239619202922,
+      "learning_rate": 4.172748534499449e-08,
+      "loss": 0.5271,
       "step": 520
     },
     {
-      "epoch": 2.8784530386740332,
-      "grad_norm": 0.6320105555650557,
-      "learning_rate": 5.0063291223308993e-08,
-      "loss": 0.4752,
+      "epoch": 2.8837638376383765,
+      "grad_norm": 0.626628102926196,
+      "learning_rate": 3.766417145395218e-08,
+      "loss": 0.5359,
       "step": 521
     },
     {
-      "epoch": 2.883977900552486,
-      "grad_norm": 0.6772494967739778,
-      "learning_rate": 4.5622296380828936e-08,
-      "loss": 0.5787,
+      "epoch": 2.88929889298893,
+      "grad_norm": 0.6508218671654497,
+      "learning_rate": 3.3808211290284886e-08,
+      "loss": 0.5114,
       "step": 522
     },
     {
-      "epoch": 2.889502762430939,
-      "grad_norm": 0.6123147921965099,
-      "learning_rate": 4.138662931949255e-08,
-      "loss": 0.4779,
+      "epoch": 2.8948339483394836,
+      "grad_norm": 0.5951817188696978,
+      "learning_rate": 3.015976597725068e-08,
+      "loss": 0.5294,
       "step": 523
     },
     {
-      "epoch": 2.8950276243093924,
-      "grad_norm": 0.6387696683129898,
-      "learning_rate": 3.7356465581047105e-08,
-      "loss": 0.5329,
+      "epoch": 2.900369003690037,
+      "grad_norm": 0.671303151438165,
+      "learning_rate": 2.6718987966992683e-08,
+      "loss": 0.5213,
       "step": 524
     },
     {
-      "epoch": 2.9005524861878453,
-      "grad_norm": 0.6541159566548991,
-      "learning_rate": 3.353197219041981e-08,
-      "loss": 0.5193,
+      "epoch": 2.9059040590405907,
+      "grad_norm": 0.676031871740552,
+      "learning_rate": 2.3486021034170857e-08,
+      "loss": 0.5459,
       "step": 525
     },
     {
-      "epoch": 2.9060773480662982,
-      "grad_norm": 0.6447424767896685,
-      "learning_rate": 2.9913307648797293e-08,
-      "loss": 0.5528,
+      "epoch": 2.911439114391144,
+      "grad_norm": 0.6208018052133741,
+      "learning_rate": 2.0461000269953457e-08,
+      "loss": 0.5358,
       "step": 526
     },
     {
-      "epoch": 2.9116022099447516,
-      "grad_norm": 0.6455122275019286,
-      "learning_rate": 2.6500621927054716e-08,
-      "loss": 0.5462,
+      "epoch": 2.9169741697416973,
+      "grad_norm": 0.6381166866678656,
+      "learning_rate": 1.7644052076371544e-08,
+      "loss": 0.5146,
       "step": 527
     },
     {
-      "epoch": 2.9171270718232045,
-      "grad_norm": 0.6329939876148144,
-      "learning_rate": 2.3294056459541302e-08,
-      "loss": 0.4936,
+      "epoch": 2.922509225092251,
+      "grad_norm": 0.6198869071038576,
+      "learning_rate": 1.5035294161039882e-08,
+      "loss": 0.5176,
       "step": 528
     },
     {
-      "epoch": 2.9226519337016574,
-      "grad_norm": 0.6557883836151688,
-      "learning_rate": 2.0293744138219495e-08,
-      "loss": 0.5273,
+      "epoch": 2.9280442804428044,
+      "grad_norm": 0.6701848507204047,
+      "learning_rate": 1.2634835532233658e-08,
+      "loss": 0.5598,
       "step": 529
     },
     {
-      "epoch": 2.9281767955801103,
-      "grad_norm": 0.6686955747117752,
-      "learning_rate": 1.7499809307154892e-08,
-      "loss": 0.5506,
+      "epoch": 2.933579335793358,
+      "grad_norm": 0.6207812803170829,
+      "learning_rate": 1.044277649433989e-08,
+      "loss": 0.4992,
       "step": 530
     },
     {
-      "epoch": 2.9337016574585633,
-      "grad_norm": 0.5955974502813394,
-      "learning_rate": 1.4912367757366485e-08,
-      "loss": 0.487,
+      "epoch": 2.9391143911439115,
+      "grad_norm": 0.6255257686553812,
+      "learning_rate": 8.459208643659122e-09,
+      "loss": 0.474,
       "step": 531
     },
     {
-      "epoch": 2.9392265193370166,
-      "grad_norm": 0.6114812296017272,
-      "learning_rate": 1.2531526722026067e-08,
-      "loss": 0.4649,
+      "epoch": 2.944649446494465,
+      "grad_norm": 0.6636870576719867,
+      "learning_rate": 6.6842148645840374e-09,
+      "loss": 0.5284,
       "step": 532
     },
     {
-      "epoch": 2.9447513812154695,
-      "grad_norm": 0.6612511110875763,
-      "learning_rate": 1.0357384872011767e-08,
-      "loss": 0.5328,
+      "epoch": 2.9501845018450186,
+      "grad_norm": 0.6474860377915387,
+      "learning_rate": 5.11786932613223e-09,
+      "loss": 0.5327,
       "step": 533
     },
     {
-      "epoch": 2.9502762430939224,
-      "grad_norm": 0.6280473844471933,
-      "learning_rate": 8.390032311824115e-09,
-      "loss": 0.5157,
+      "epoch": 2.955719557195572,
+      "grad_norm": 0.6112767808361741,
+      "learning_rate": 3.760237478849793e-09,
+      "loss": 0.5189,
       "step": 534
     },
     {
-      "epoch": 2.955801104972376,
-      "grad_norm": 0.6485700698849401,
-      "learning_rate": 6.629550575847355e-09,
-      "loss": 0.5375,
+      "epoch": 2.961254612546125,
+      "grad_norm": 0.680320025195111,
+      "learning_rate": 2.611376052073511e-09,
+      "loss": 0.55,
       "step": 535
     },
     {
-      "epoch": 2.9613259668508287,
-      "grad_norm": 0.6621455408098162,
-      "learning_rate": 5.0760126249715935e-09,
-      "loss": 0.6062,
+      "epoch": 2.9667896678966788,
+      "grad_norm": 0.6580418332010902,
+      "learning_rate": 1.6713330515627512e-09,
+      "loss": 0.5855,
       "step": 536
     },
     {
-      "epoch": 2.9668508287292816,
-      "grad_norm": 0.6340833044767035,
-      "learning_rate": 3.729482843569665e-09,
-      "loss": 0.5398,
+      "epoch": 2.9723247232472323,
+      "grad_norm": 0.6263030989787438,
+      "learning_rate": 9.401477574932927e-10,
+      "loss": 0.5423,
       "step": 537
     },
     {
-      "epoch": 2.972375690607735,
-      "grad_norm": 0.6235121401681177,
-      "learning_rate": 2.5900170368281517e-09,
-      "loss": 0.5235,
+      "epoch": 2.977859778597786,
+      "grad_norm": 0.6000757044114218,
+      "learning_rate": 4.178507228136397e-10,
+      "loss": 0.4987,
       "step": 538
     },
     {
-      "epoch": 2.977900552486188,
-      "grad_norm": 0.626406886227414,
-      "learning_rate": 1.657662428434792e-09,
-      "loss": 0.5245,
+      "epoch": 2.9833948339483394,
+      "grad_norm": 0.6436284868599454,
+      "learning_rate": 1.0446377197104174e-10,
+      "loss": 0.5854,
       "step": 539
     },
     {
-      "epoch": 2.983425414364641,
-      "grad_norm": 0.6517990473785358,
-      "learning_rate": 9.324576586211553e-10,
-      "loss": 0.5615,
-      "step": 540
-    },
-    {
-      "epoch": 2.988950276243094,
-      "grad_norm": 0.6588091394692884,
-      "learning_rate": 4.1443278256170227e-10,
-      "loss": 0.5629,
-      "step": 541
-    },
-    {
-      "epoch": 2.994475138121547,
-      "grad_norm": 0.6095291946882728,
-      "learning_rate": 1.0360926912866831e-10,
-      "loss": 0.4717,
-      "step": 542
-    },
-    {
-      "epoch": 3.0,
-      "grad_norm": 0.6878897999106169,
+      "epoch": 2.988929889298893,
+      "grad_norm": 0.6261144170087621,
       "learning_rate": 0.0,
-      "loss": 0.5853,
-      "step": 543
+      "loss": 0.5354,
+      "step": 540
     },
     {
-      "epoch": 3.0,
-      "step": 543,
-      "total_flos": 136664820957184.0,
-      "train_loss": 0.613760003727444,
-      "train_runtime": 1147.7371,
-      "train_samples_per_second": 45.3,
-      "train_steps_per_second": 0.473
+      "epoch": 2.988929889298893,
+      "step": 540,
+      "total_flos": 110732096700416.0,
+      "train_loss": 0.6148645067104587,
+      "train_runtime": 2111.1531,
+      "train_samples_per_second": 24.628,
+      "train_steps_per_second": 0.256
     }
   ],
   "logging_steps": 1,
-  "max_steps": 543,
+  "max_steps": 540,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 3,
   "save_steps": 500,
@@ -3836,8 +3815,8 @@
       "attributes": {}
     }
   },
-  "total_flos": 136664820957184.0,
-  "train_batch_size": 3,
+  "total_flos": 110732096700416.0,
+  "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null
 }