diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,8850 +1,3026 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 14.981905910735826,
+  "epoch": 9.975903614457831,
   "eval_steps": 500,
-  "global_step": 6210,
+  "global_step": 2070,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.0024125452352231603,
-      "grad_norm": 6.09375,
-      "learning_rate": 3.22061191626409e-07,
-      "loss": 3.0105,
+      "epoch": 0.004819277108433735,
+      "grad_norm": 7.09375,
+      "learning_rate": 9.66183574879227e-07,
+      "loss": 3.0067,
       "step": 1
     },
     {
-      "epoch": 0.012062726176115802,
-      "grad_norm": 2.21875,
-      "learning_rate": 1.610305958132045e-06,
-      "loss": 3.0058,
+      "epoch": 0.024096385542168676,
+      "grad_norm": 6.625,
+      "learning_rate": 4.830917874396135e-06,
+      "loss": 3.0384,
       "step": 5
     },
     {
-      "epoch": 0.024125452352231604,
-      "grad_norm": 2.125,
-      "learning_rate": 3.22061191626409e-06,
-      "loss": 3.059,
+      "epoch": 0.04819277108433735,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.66183574879227e-06,
+      "loss": 2.9932,
       "step": 10
     },
     {
-      "epoch": 0.03618817852834741,
-      "grad_norm": 2.015625,
-      "learning_rate": 4.830917874396135e-06,
-      "loss": 3.011,
+      "epoch": 0.07228915662650602,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.4492753623188407e-05,
+      "loss": 2.8805,
       "step": 15
     },
     {
-      "epoch": 0.04825090470446321,
-      "grad_norm": 2.03125,
-      "learning_rate": 6.44122383252818e-06,
-      "loss": 3.0197,
+      "epoch": 0.0963855421686747,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.932367149758454e-05,
+      "loss": 2.7091,
       "step": 20
     },
     {
-      "epoch": 0.06031363088057901,
-      "grad_norm": 2.953125,
-      "learning_rate": 8.051529790660225e-06,
-      "loss": 3.004,
+      "epoch": 0.12048192771084337,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.4154589371980676e-05,
+      "loss": 2.5046,
       "step": 25
     },
     {
-      "epoch": 0.07237635705669482,
-      "grad_norm": 2.625,
-      "learning_rate": 9.66183574879227e-06,
-      "loss": 2.9712,
+      "epoch": 0.14457831325301204,
+      "grad_norm": 1.359375,
+      "learning_rate": 2.8985507246376814e-05,
+      "loss": 2.3296,
       "step": 30
     },
     {
-      "epoch": 0.08443908323281062,
-      "grad_norm": 3.984375,
-      "learning_rate": 1.1272141706924317e-05,
-      "loss": 2.9778,
+      "epoch": 0.1686746987951807,
+      "grad_norm": 1.53125,
+      "learning_rate": 3.381642512077295e-05,
+      "loss": 2.1499,
       "step": 35
     },
     {
-      "epoch": 0.09650180940892641,
-      "grad_norm": 2.34375,
-      "learning_rate": 1.288244766505636e-05,
-      "loss": 2.881,
+      "epoch": 0.1927710843373494,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.864734299516908e-05,
+      "loss": 1.9713,
       "step": 40
     },
     {
-      "epoch": 0.10856453558504221,
-      "grad_norm": 2.0625,
-      "learning_rate": 1.4492753623188407e-05,
-      "loss": 2.8379,
+      "epoch": 0.21686746987951808,
+      "grad_norm": 1.2421875,
+      "learning_rate": 4.347826086956522e-05,
+      "loss": 1.8202,
       "step": 45
     },
     {
-      "epoch": 0.12062726176115803,
-      "grad_norm": 2.96875,
-      "learning_rate": 1.610305958132045e-05,
-      "loss": 2.738,
+      "epoch": 0.24096385542168675,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.830917874396135e-05,
+      "loss": 1.6894,
       "step": 50
     },
     {
-      "epoch": 0.13268998793727382,
-      "grad_norm": 1.8359375,
-      "learning_rate": 1.77133655394525e-05,
-      "loss": 2.6525,
+      "epoch": 0.26506024096385544,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.3140096618357496e-05,
+      "loss": 1.571,
       "step": 55
     },
     {
-      "epoch": 0.14475271411338964,
-      "grad_norm": 2.609375,
-      "learning_rate": 1.932367149758454e-05,
-      "loss": 2.5732,
+      "epoch": 0.2891566265060241,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.797101449275363e-05,
+      "loss": 1.4798,
       "step": 60
     },
     {
-      "epoch": 0.15681544028950542,
-      "grad_norm": 1.4609375,
-      "learning_rate": 2.0933977455716587e-05,
-      "loss": 2.4813,
+      "epoch": 0.3132530120481928,
+      "grad_norm": 0.4375,
+      "learning_rate": 6.280193236714976e-05,
+      "loss": 1.4004,
       "step": 65
     },
     {
-      "epoch": 0.16887816646562123,
-      "grad_norm": 1.0546875,
-      "learning_rate": 2.2544283413848633e-05,
-      "loss": 2.4334,
+      "epoch": 0.3373493975903614,
+      "grad_norm": 0.470703125,
+      "learning_rate": 6.76328502415459e-05,
+      "loss": 1.3436,
       "step": 70
     },
     {
-      "epoch": 0.18094089264173704,
-      "grad_norm": 2.625,
-      "learning_rate": 2.4154589371980676e-05,
-      "loss": 2.3332,
+      "epoch": 0.3614457831325301,
+      "grad_norm": 0.265625,
+      "learning_rate": 7.246376811594203e-05,
+      "loss": 1.2867,
       "step": 75
     },
     {
-      "epoch": 0.19300361881785283,
-      "grad_norm": 1.1015625,
-      "learning_rate": 2.576489533011272e-05,
-      "loss": 2.2571,
+      "epoch": 0.3855421686746988,
+      "grad_norm": 0.2890625,
+      "learning_rate": 7.729468599033817e-05,
+      "loss": 1.2529,
       "step": 80
     },
     {
-      "epoch": 0.20506634499396864,
-      "grad_norm": 5.34375,
-      "learning_rate": 2.7375201288244768e-05,
-      "loss": 2.1755,
+      "epoch": 0.40963855421686746,
+      "grad_norm": 0.404296875,
+      "learning_rate": 8.21256038647343e-05,
+      "loss": 1.215,
       "step": 85
     },
     {
-      "epoch": 0.21712907117008443,
-      "grad_norm": 1.0390625,
-      "learning_rate": 2.8985507246376814e-05,
-      "loss": 2.089,
+      "epoch": 0.43373493975903615,
+      "grad_norm": 0.1796875,
+      "learning_rate": 8.695652173913044e-05,
+      "loss": 1.1938,
       "step": 90
     },
     {
-      "epoch": 0.22919179734620024,
-      "grad_norm": 3.328125,
-      "learning_rate": 3.059581320450886e-05,
-      "loss": 2.0223,
+      "epoch": 0.4578313253012048,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.178743961352657e-05,
+      "loss": 1.1765,
       "step": 95
     },
     {
-      "epoch": 0.24125452352231605,
-      "grad_norm": 1.0390625,
-      "learning_rate": 3.22061191626409e-05,
-      "loss": 1.9332,
+      "epoch": 0.4819277108433735,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.66183574879227e-05,
+      "loss": 1.1637,
       "step": 100
     },
     {
-      "epoch": 0.25331724969843183,
-      "grad_norm": 1.046875,
-      "learning_rate": 3.381642512077295e-05,
-      "loss": 1.8597,
+      "epoch": 0.5060240963855421,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00010144927536231885,
+      "loss": 1.1525,
       "step": 105
     },
     {
-      "epoch": 0.26537997587454765,
-      "grad_norm": 1.0859375,
-      "learning_rate": 3.5426731078905e-05,
-      "loss": 1.8142,
+      "epoch": 0.5301204819277109,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00010628019323671499,
+      "loss": 1.1378,
       "step": 110
     },
     {
-      "epoch": 0.27744270205066346,
-      "grad_norm": 0.625,
-      "learning_rate": 3.7037037037037037e-05,
-      "loss": 1.7501,
+      "epoch": 0.5542168674698795,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00011111111111111112,
+      "loss": 1.125,
       "step": 115
     },
     {
-      "epoch": 0.28950542822677927,
-      "grad_norm": 0.59375,
-      "learning_rate": 3.864734299516908e-05,
-      "loss": 1.6938,
+      "epoch": 0.5783132530120482,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00011594202898550725,
+      "loss": 1.1077,
       "step": 120
     },
     {
-      "epoch": 0.30156815440289503,
-      "grad_norm": 0.478515625,
-      "learning_rate": 4.025764895330113e-05,
-      "loss": 1.6289,
+      "epoch": 0.6024096385542169,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00012077294685990339,
+      "loss": 1.1027,
       "step": 125
     },
     {
-      "epoch": 0.31363088057901084,
-      "grad_norm": 0.671875,
-      "learning_rate": 4.1867954911433174e-05,
-      "loss": 1.5807,
+      "epoch": 0.6265060240963856,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00012560386473429953,
+      "loss": 1.1005,
       "step": 130
     },
     {
-      "epoch": 0.32569360675512665,
-      "grad_norm": 0.423828125,
-      "learning_rate": 4.347826086956522e-05,
-      "loss": 1.5408,
+      "epoch": 0.6506024096385542,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00013043478260869567,
+      "loss": 1.0961,
       "step": 135
     },
     {
-      "epoch": 0.33775633293124246,
-      "grad_norm": 0.494140625,
-      "learning_rate": 4.5088566827697266e-05,
-      "loss": 1.4961,
+      "epoch": 0.6746987951807228,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.0001352657004830918,
+      "loss": 1.0835,
       "step": 140
     },
     {
-      "epoch": 0.3498190591073583,
-      "grad_norm": 0.40234375,
-      "learning_rate": 4.669887278582931e-05,
-      "loss": 1.4528,
+      "epoch": 0.6987951807228916,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00014009661835748792,
+      "loss": 1.0731,
       "step": 145
     },
     {
-      "epoch": 0.3618817852834741,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.830917874396135e-05,
-      "loss": 1.4274,
+      "epoch": 0.7228915662650602,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00014492753623188405,
+      "loss": 1.0655,
       "step": 150
     },
     {
-      "epoch": 0.37394451145958985,
-      "grad_norm": 0.345703125,
-      "learning_rate": 4.99194847020934e-05,
-      "loss": 1.3942,
+      "epoch": 0.7469879518072289,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0001497584541062802,
+      "loss": 1.064,
       "step": 155
     },
     {
-      "epoch": 0.38600723763570566,
-      "grad_norm": 0.455078125,
-      "learning_rate": 5.152979066022544e-05,
-      "loss": 1.387,
+      "epoch": 0.7710843373493976,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015458937198067633,
+      "loss": 1.0539,
       "step": 160
     },
     {
-      "epoch": 0.39806996381182147,
-      "grad_norm": 0.404296875,
-      "learning_rate": 5.3140096618357496e-05,
-      "loss": 1.3418,
+      "epoch": 0.7951807228915663,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00015942028985507247,
+      "loss": 1.057,
       "step": 165
     },
     {
-      "epoch": 0.4101326899879373,
-      "grad_norm": 0.3046875,
-      "learning_rate": 5.4750402576489535e-05,
-      "loss": 1.3263,
+      "epoch": 0.8192771084337349,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001642512077294686,
+      "loss": 1.052,
       "step": 170
     },
     {
-      "epoch": 0.4221954161640531,
-      "grad_norm": 0.279296875,
-      "learning_rate": 5.6360708534621574e-05,
-      "loss": 1.3058,
+      "epoch": 0.8433734939759037,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00016908212560386474,
+      "loss": 1.0539,
       "step": 175
     },
     {
-      "epoch": 0.43425814234016885,
-      "grad_norm": 0.53125,
-      "learning_rate": 5.797101449275363e-05,
-      "loss": 1.2854,
+      "epoch": 0.8674698795180723,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00017391304347826088,
+      "loss": 1.0393,
       "step": 180
     },
     {
-      "epoch": 0.44632086851628466,
-      "grad_norm": 0.490234375,
-      "learning_rate": 5.9581320450885666e-05,
-      "loss": 1.2769,
+      "epoch": 0.891566265060241,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00017874396135265702,
+      "loss": 1.0354,
       "step": 185
     },
     {
-      "epoch": 0.4583835946924005,
-      "grad_norm": 0.359375,
-      "learning_rate": 6.119162640901772e-05,
-      "loss": 1.2572,
+      "epoch": 0.9156626506024096,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00018357487922705313,
+      "loss": 1.0443,
       "step": 190
     },
     {
-      "epoch": 0.4704463208685163,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 6.280193236714976e-05,
-      "loss": 1.2506,
+      "epoch": 0.9397590361445783,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00018840579710144927,
+      "loss": 1.0303,
       "step": 195
     },
     {
-      "epoch": 0.4825090470446321,
-      "grad_norm": 0.2734375,
-      "learning_rate": 6.44122383252818e-05,
-      "loss": 1.2466,
+      "epoch": 0.963855421686747,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0001932367149758454,
+      "loss": 1.0281,
       "step": 200
     },
     {
-      "epoch": 0.4945717732207479,
-      "grad_norm": 0.380859375,
-      "learning_rate": 6.602254428341386e-05,
-      "loss": 1.2347,
+      "epoch": 0.9879518072289156,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019806763285024154,
+      "loss": 1.0246,
       "step": 205
     },
     {
-      "epoch": 0.5066344993968637,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 6.76328502415459e-05,
-      "loss": 1.2245,
+      "epoch": 0.9975903614457832,
+      "eval_loss": 2.4549527168273926,
+      "eval_runtime": 0.4905,
+      "eval_samples_per_second": 20.386,
+      "eval_steps_per_second": 2.039,
+      "step": 207
+    },
+    {
+      "epoch": 1.0120481927710843,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019999872036643513,
+      "loss": 1.0207,
       "step": 210
     },
     {
-      "epoch": 0.5186972255729795,
-      "grad_norm": 0.2294921875,
-      "learning_rate": 6.924315619967794e-05,
-      "loss": 1.2147,
+      "epoch": 1.036144578313253,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00019999090050213636,
+      "loss": 1.0131,
       "step": 215
     },
     {
-      "epoch": 0.5307599517490953,
-      "grad_norm": 0.322265625,
-      "learning_rate": 7.085346215781e-05,
-      "loss": 1.2062,
+      "epoch": 1.0602409638554218,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019997597223631895,
+      "loss": 1.0098,
       "step": 220
     },
     {
-      "epoch": 0.5428226779252111,
-      "grad_norm": 0.416015625,
-      "learning_rate": 7.246376811594203e-05,
-      "loss": 1.2057,
+      "epoch": 1.0843373493975903,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00019995393663024054,
+      "loss": 1.0082,
       "step": 225
     },
     {
-      "epoch": 0.5548854041013269,
-      "grad_norm": 0.357421875,
-      "learning_rate": 7.407407407407407e-05,
-      "loss": 1.1828,
+      "epoch": 1.108433734939759,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00019992479525042303,
+      "loss": 1.0136,
       "step": 230
     },
     {
-      "epoch": 0.5669481302774427,
-      "grad_norm": 0.279296875,
-      "learning_rate": 7.568438003220612e-05,
-      "loss": 1.1799,
+      "epoch": 1.1325301204819278,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0001998885501685412,
+      "loss": 1.004,
       "step": 235
     },
     {
-      "epoch": 0.5790108564535585,
-      "grad_norm": 0.390625,
-      "learning_rate": 7.729468599033817e-05,
-      "loss": 1.1689,
+      "epoch": 1.1566265060240963,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00019984520396127553,
+      "loss": 1.0012,
       "step": 240
     },
     {
-      "epoch": 0.5910735826296744,
-      "grad_norm": 0.69921875,
-      "learning_rate": 7.890499194847021e-05,
-      "loss": 1.1764,
+      "epoch": 1.180722891566265,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001997947597101288,
+      "loss": 0.9956,
       "step": 245
     },
     {
-      "epoch": 0.6031363088057901,
-      "grad_norm": 0.46484375,
-      "learning_rate": 8.051529790660226e-05,
-      "loss": 1.1552,
+      "epoch": 1.2048192771084336,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0001997372210012073,
+      "loss": 0.9968,
       "step": 250
     },
     {
-      "epoch": 0.6151990349819059,
-      "grad_norm": 0.2421875,
-      "learning_rate": 8.21256038647343e-05,
-      "loss": 1.1596,
+      "epoch": 1.2289156626506024,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0001996725919249657,
+      "loss": 0.997,
       "step": 255
     },
     {
-      "epoch": 0.6272617611580217,
-      "grad_norm": 0.58984375,
-      "learning_rate": 8.373590982286635e-05,
-      "loss": 1.1637,
+      "epoch": 1.2530120481927711,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00019960087707591626,
+      "loss": 1.0023,
       "step": 260
     },
     {
-      "epoch": 0.6393244873341375,
-      "grad_norm": 0.578125,
-      "learning_rate": 8.53462157809984e-05,
-      "loss": 1.15,
+      "epoch": 1.2771084337349397,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019952208155230234,
+      "loss": 0.9934,
       "step": 265
     },
     {
-      "epoch": 0.6513872135102533,
-      "grad_norm": 0.41015625,
-      "learning_rate": 8.695652173913044e-05,
-      "loss": 1.1566,
+      "epoch": 1.3012048192771084,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019943621095573586,
+      "loss": 0.9971,
       "step": 270
     },
     {
-      "epoch": 0.6634499396863691,
-      "grad_norm": 0.80859375,
-      "learning_rate": 8.856682769726249e-05,
-      "loss": 1.1402,
+      "epoch": 1.3253012048192772,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00019934327139079915,
+      "loss": 0.9945,
       "step": 275
     },
     {
-      "epoch": 0.6755126658624849,
-      "grad_norm": 0.8984375,
-      "learning_rate": 9.017713365539453e-05,
-      "loss": 1.1368,
+      "epoch": 1.3493975903614457,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00019924326946461074,
+      "loss": 0.9864,
       "step": 280
     },
     {
-      "epoch": 0.6875753920386007,
-      "grad_norm": 0.3125,
-      "learning_rate": 9.178743961352657e-05,
-      "loss": 1.1269,
+      "epoch": 1.3734939759036144,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0001991362122863561,
+      "loss": 0.9804,
       "step": 285
     },
     {
-      "epoch": 0.6996381182147166,
-      "grad_norm": 0.341796875,
-      "learning_rate": 9.339774557165862e-05,
-      "loss": 1.1273,
+      "epoch": 1.3975903614457832,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001990221074667818,
+      "loss": 0.9826,
       "step": 290
     },
     {
-      "epoch": 0.7117008443908324,
-      "grad_norm": 0.26953125,
-      "learning_rate": 9.500805152979067e-05,
-      "loss": 1.1166,
+      "epoch": 1.4216867469879517,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00019890096311765465,
+      "loss": 0.9858,
       "step": 295
     },
     {
-      "epoch": 0.7237635705669482,
-      "grad_norm": 0.2578125,
-      "learning_rate": 9.66183574879227e-05,
-      "loss": 1.1147,
+      "epoch": 1.4457831325301205,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00019877278785118517,
+      "loss": 0.9808,
       "step": 300
     },
     {
-      "epoch": 0.7358262967430639,
-      "grad_norm": 0.310546875,
-      "learning_rate": 9.822866344605476e-05,
-      "loss": 1.1155,
+      "epoch": 1.4698795180722892,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019863759077941504,
+      "loss": 0.9733,
       "step": 305
     },
     {
-      "epoch": 0.7478890229191797,
-      "grad_norm": 0.41015625,
-      "learning_rate": 9.98389694041868e-05,
-      "loss": 1.1081,
+      "epoch": 1.4939759036144578,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00019849538151356955,
+      "loss": 0.9788,
       "step": 310
     },
     {
-      "epoch": 0.7599517490952955,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00010144927536231885,
-      "loss": 1.1083,
+      "epoch": 1.5180722891566265,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0001983461701633742,
+      "loss": 0.981,
       "step": 315
     },
     {
-      "epoch": 0.7720144752714113,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00010305958132045089,
-      "loss": 1.0981,
+      "epoch": 1.5421686746987953,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00019818996733633618,
+      "loss": 0.9806,
       "step": 320
     },
     {
-      "epoch": 0.7840772014475271,
-      "grad_norm": 0.478515625,
-      "learning_rate": 0.00010466988727858293,
-      "loss": 1.1001,
+      "epoch": 1.5662650602409638,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00019802678413699006,
+      "loss": 0.9681,
       "step": 325
     },
     {
-      "epoch": 0.7961399276236429,
-      "grad_norm": 0.5078125,
-      "learning_rate": 0.00010628019323671499,
-      "loss": 1.1048,
+      "epoch": 1.5903614457831325,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00019785663216610844,
+      "loss": 0.9758,
       "step": 330
     },
     {
-      "epoch": 0.8082026537997588,
-      "grad_norm": 0.71875,
-      "learning_rate": 0.00010789049919484701,
-      "loss": 1.0983,
+      "epoch": 1.6144578313253013,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0001976795235198773,
+      "loss": 0.9711,
       "step": 335
     },
     {
-      "epoch": 0.8202653799758746,
-      "grad_norm": 0.388671875,
-      "learning_rate": 0.00010950080515297907,
-      "loss": 1.0982,
+      "epoch": 1.6385542168674698,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019749547078903604,
+      "loss": 0.9666,
       "step": 340
     },
     {
-      "epoch": 0.8323281061519904,
-      "grad_norm": 0.69921875,
-      "learning_rate": 0.00011111111111111112,
-      "loss": 1.0982,
+      "epoch": 1.6626506024096386,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00019730448705798239,
+      "loss": 0.9673,
       "step": 345
     },
     {
-      "epoch": 0.8443908323281062,
-      "grad_norm": 0.53515625,
-      "learning_rate": 0.00011272141706924315,
-      "loss": 1.0968,
+      "epoch": 1.6867469879518073,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00019710658590384227,
+      "loss": 0.9631,
       "step": 350
     },
     {
-      "epoch": 0.856453558504222,
-      "grad_norm": 0.609375,
-      "learning_rate": 0.00011433172302737521,
-      "loss": 1.084,
+      "epoch": 1.7108433734939759,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00019690178139550443,
+      "loss": 0.9646,
       "step": 355
     },
     {
-      "epoch": 0.8685162846803377,
-      "grad_norm": 0.6015625,
-      "learning_rate": 0.00011594202898550725,
-      "loss": 1.0815,
+      "epoch": 1.7349397590361446,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00019669008809262062,
+      "loss": 0.9699,
       "step": 360
     },
     {
-      "epoch": 0.8805790108564535,
-      "grad_norm": 0.373046875,
-      "learning_rate": 0.00011755233494363929,
-      "loss": 1.083,
+      "epoch": 1.7590361445783134,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00019647152104457013,
+      "loss": 0.9639,
       "step": 365
     },
     {
-      "epoch": 0.8926417370325693,
-      "grad_norm": 0.68359375,
-      "learning_rate": 0.00011916264090177133,
-      "loss": 1.0735,
+      "epoch": 1.783132530120482,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00019624609578939027,
+      "loss": 0.9597,
       "step": 370
     },
     {
-      "epoch": 0.9047044632086851,
-      "grad_norm": 1.09375,
-      "learning_rate": 0.00012077294685990339,
-      "loss": 1.0852,
+      "epoch": 1.8072289156626506,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001960138283526715,
+      "loss": 0.962,
       "step": 375
     },
     {
-      "epoch": 0.916767189384801,
-      "grad_norm": 0.4765625,
-      "learning_rate": 0.00012238325281803544,
-      "loss": 1.0852,
+      "epoch": 1.8313253012048194,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001957747352464184,
+      "loss": 0.9628,
       "step": 380
     },
     {
-      "epoch": 0.9288299155609168,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00012399355877616748,
-      "loss": 1.0689,
+      "epoch": 1.855421686746988,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00019552883346787552,
+      "loss": 0.9582,
       "step": 385
     },
     {
-      "epoch": 0.9408926417370326,
-      "grad_norm": 0.48828125,
-      "learning_rate": 0.00012560386473429953,
-      "loss": 1.0771,
+      "epoch": 1.8795180722891565,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001952761404983194,
+      "loss": 0.9566,
       "step": 390
     },
     {
-      "epoch": 0.9529553679131484,
-      "grad_norm": 0.58203125,
-      "learning_rate": 0.00012721417069243158,
-      "loss": 1.0701,
+      "epoch": 1.9036144578313254,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0001950166743018156,
+      "loss": 0.9614,
       "step": 395
     },
     {
-      "epoch": 0.9650180940892642,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.0001288244766505636,
-      "loss": 1.07,
+      "epoch": 1.927710843373494,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00019475045332394153,
+      "loss": 0.9607,
       "step": 400
     },
     {
-      "epoch": 0.97708082026538,
-      "grad_norm": 0.423828125,
-      "learning_rate": 0.00013043478260869567,
-      "loss": 1.0665,
+      "epoch": 1.9518072289156625,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00019447749649047542,
+      "loss": 0.9511,
       "step": 405
     },
     {
-      "epoch": 0.9891435464414958,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00013204508856682771,
-      "loss": 1.0614,
+      "epoch": 1.9759036144578315,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0001941978232060507,
+      "loss": 0.9516,
       "step": 410
     },
     {
-      "epoch": 0.9987937273823885,
-      "eval_loss": 2.4759693145751953,
-      "eval_runtime": 0.6381,
-      "eval_samples_per_second": 15.671,
-      "eval_steps_per_second": 1.567,
-      "step": 414
+      "epoch": 2.0,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00019391145335277655,
+      "loss": 0.9556,
+      "step": 415
     },
     {
-      "epoch": 1.0012062726176116,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00013365539452495973,
-      "loss": 1.0719,
+      "epoch": 2.0,
+      "eval_loss": 2.453037977218628,
+      "eval_runtime": 0.4785,
+      "eval_samples_per_second": 20.897,
+      "eval_steps_per_second": 2.09,
       "step": 415
     },
     {
-      "epoch": 1.0132689987937273,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.0001352657004830918,
-      "loss": 1.0513,
+      "epoch": 2.0240963855421685,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00019361840728882447,
+      "loss": 0.9377,
       "step": 420
     },
     {
-      "epoch": 1.0253317249698433,
-      "grad_norm": 0.63671875,
-      "learning_rate": 0.00013687600644122385,
-      "loss": 1.0561,
+      "epoch": 2.0481927710843375,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00019331870584698093,
+      "loss": 0.9384,
       "step": 425
     },
     {
-      "epoch": 1.037394451145959,
-      "grad_norm": 0.41796875,
-      "learning_rate": 0.00013848631239935587,
-      "loss": 1.0549,
+      "epoch": 2.072289156626506,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019301237033316659,
+      "loss": 0.9315,
       "step": 430
     },
     {
-      "epoch": 1.0494571773220749,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00014009661835748792,
-      "loss": 1.052,
+      "epoch": 2.0963855421686746,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00019269942252492133,
+      "loss": 0.9328,
       "step": 435
     },
     {
-      "epoch": 1.0615199034981906,
-      "grad_norm": 0.482421875,
-      "learning_rate": 0.00014170692431562,
-      "loss": 1.0525,
+      "epoch": 2.1204819277108435,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001923798846698564,
+      "loss": 0.9374,
       "step": 440
     },
     {
-      "epoch": 1.0735826296743065,
-      "grad_norm": 0.7421875,
-      "learning_rate": 0.000143317230273752,
-      "loss": 1.0521,
+      "epoch": 2.144578313253012,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00019205377948407258,
+      "loss": 0.9304,
       "step": 445
     },
     {
-      "epoch": 1.0856453558504222,
-      "grad_norm": 0.50390625,
-      "learning_rate": 0.00014492753623188405,
-      "loss": 1.0538,
+      "epoch": 2.1686746987951806,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00019172113015054532,
+      "loss": 0.9319,
       "step": 450
     },
     {
-      "epoch": 1.097708082026538,
-      "grad_norm": 0.6484375,
-      "learning_rate": 0.00014653784219001613,
-      "loss": 1.0593,
+      "epoch": 2.1927710843373496,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00019138196031747681,
+      "loss": 0.9386,
       "step": 455
     },
     {
-      "epoch": 1.1097708082026538,
-      "grad_norm": 0.482421875,
-      "learning_rate": 0.00014814814814814815,
-      "loss": 1.0443,
+      "epoch": 2.216867469879518,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0001910362940966147,
+      "loss": 0.9383,
       "step": 460
     },
     {
-      "epoch": 1.1218335343787695,
-      "grad_norm": 0.373046875,
-      "learning_rate": 0.0001497584541062802,
-      "loss": 1.044,
+      "epoch": 2.2409638554216866,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00019068415606153787,
+      "loss": 0.9369,
       "step": 465
     },
     {
-      "epoch": 1.1338962605548855,
-      "grad_norm": 0.48828125,
-      "learning_rate": 0.00015136876006441224,
-      "loss": 1.0496,
+      "epoch": 2.2650602409638556,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00019032557124590974,
+      "loss": 0.9336,
       "step": 470
     },
     {
-      "epoch": 1.1459589867310012,
-      "grad_norm": 0.49609375,
-      "learning_rate": 0.00015297906602254428,
-      "loss": 1.0491,
+      "epoch": 2.289156626506024,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00018996056514169844,
+      "loss": 0.9292,
       "step": 475
     },
     {
-      "epoch": 1.158021712907117,
-      "grad_norm": 0.625,
-      "learning_rate": 0.00015458937198067633,
-      "loss": 1.0455,
+      "epoch": 2.3132530120481927,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0001895891636973646,
+      "loss": 0.9386,
       "step": 480
     },
     {
-      "epoch": 1.1700844390832328,
-      "grad_norm": 0.546875,
-      "learning_rate": 0.00015619967793880838,
-      "loss": 1.0366,
+      "epoch": 2.337349397590361,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00018921139331601667,
+      "loss": 0.9404,
       "step": 485
     },
     {
-      "epoch": 1.1821471652593487,
-      "grad_norm": 0.59765625,
-      "learning_rate": 0.00015780998389694042,
-      "loss": 1.0428,
+      "epoch": 2.36144578313253,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00018882728085353392,
+      "loss": 0.9267,
       "step": 490
     },
     {
-      "epoch": 1.1942098914354644,
-      "grad_norm": 1.4140625,
-      "learning_rate": 0.00015942028985507247,
-      "loss": 1.0449,
+      "epoch": 2.3855421686746987,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00018843685361665723,
+      "loss": 0.9285,
       "step": 495
     },
     {
-      "epoch": 1.2062726176115803,
-      "grad_norm": 0.6953125,
-      "learning_rate": 0.00016103059581320451,
-      "loss": 1.0363,
+      "epoch": 2.4096385542168672,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00018804013936104792,
+      "loss": 0.9396,
       "step": 500
     },
     {
-      "epoch": 1.218335343787696,
-      "grad_norm": 0.7890625,
-      "learning_rate": 0.00016264090177133656,
-      "loss": 1.0377,
+      "epoch": 2.433734939759036,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00018763716628931437,
+      "loss": 0.929,
       "step": 505
     },
     {
-      "epoch": 1.2303980699638117,
-      "grad_norm": 0.62890625,
-      "learning_rate": 0.0001642512077294686,
-      "loss": 1.0404,
+      "epoch": 2.4578313253012047,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0001872279630490074,
+      "loss": 0.9277,
       "step": 510
     },
     {
-      "epoch": 1.2424607961399277,
-      "grad_norm": 0.69921875,
-      "learning_rate": 0.00016586151368760065,
-      "loss": 1.0448,
+      "epoch": 2.4819277108433733,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00018681255873058338,
+      "loss": 0.9293,
       "step": 515
     },
     {
-      "epoch": 1.2545235223160434,
-      "grad_norm": 1.28125,
-      "learning_rate": 0.0001674718196457327,
-      "loss": 1.0441,
+      "epoch": 2.5060240963855422,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00018639098286533644,
+      "loss": 0.9253,
       "step": 520
     },
     {
-      "epoch": 1.2665862484921593,
-      "grad_norm": 0.68359375,
-      "learning_rate": 0.00016908212560386474,
-      "loss": 1.0349,
+      "epoch": 2.5301204819277108,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00018596326542329888,
+      "loss": 0.9231,
       "step": 525
     },
     {
-      "epoch": 1.278648974668275,
-      "grad_norm": 1.203125,
-      "learning_rate": 0.0001706924315619968,
-      "loss": 1.0504,
+      "epoch": 2.5542168674698793,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018552943681111067,
+      "loss": 0.9241,
       "step": 530
     },
     {
-      "epoch": 1.290711700844391,
-      "grad_norm": 0.8046875,
-      "learning_rate": 0.00017230273752012884,
-      "loss": 1.0362,
+      "epoch": 2.5783132530120483,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0001850895278698579,
+      "loss": 0.9234,
       "step": 535
     },
     {
-      "epoch": 1.3027744270205066,
-      "grad_norm": 0.80859375,
-      "learning_rate": 0.00017391304347826088,
-      "loss": 1.0454,
+      "epoch": 2.602409638554217,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00018464356987288013,
+      "loss": 0.9284,
       "step": 540
     },
     {
-      "epoch": 1.3148371531966223,
-      "grad_norm": 1.0078125,
-      "learning_rate": 0.00017552334943639293,
-      "loss": 1.0466,
+      "epoch": 2.6265060240963853,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001841915945235472,
+      "loss": 0.9247,
       "step": 545
     },
     {
-      "epoch": 1.3268998793727382,
-      "grad_norm": 0.6171875,
-      "learning_rate": 0.00017713365539452497,
-      "loss": 1.0319,
+      "epoch": 2.6506024096385543,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00018373363395300554,
+      "loss": 0.9314,
       "step": 550
     },
     {
-      "epoch": 1.3389626055488542,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.00017874396135265702,
-      "loss": 1.038,
+      "epoch": 2.674698795180723,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0001832697207178938,
+      "loss": 0.9308,
       "step": 555
     },
     {
-      "epoch": 1.3510253317249699,
-      "grad_norm": 0.349609375,
-      "learning_rate": 0.00018035426731078907,
-      "loss": 1.0181,
+      "epoch": 2.6987951807228914,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00018279988779802833,
+      "loss": 0.9275,
       "step": 560
     },
     {
-      "epoch": 1.3630880579010856,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.0001819645732689211,
-      "loss": 1.0236,
+      "epoch": 2.7228915662650603,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00018232416859405895,
+      "loss": 0.9261,
       "step": 565
     },
     {
-      "epoch": 1.3751507840772015,
-      "grad_norm": 0.36328125,
-      "learning_rate": 0.00018357487922705313,
-      "loss": 1.0274,
+      "epoch": 2.746987951807229,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00018184259692509406,
+      "loss": 0.925,
       "step": 570
     },
     {
-      "epoch": 1.3872135102533172,
-      "grad_norm": 0.40234375,
-      "learning_rate": 0.0001851851851851852,
-      "loss": 1.0281,
+      "epoch": 2.7710843373493974,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00018135520702629675,
+      "loss": 0.9211,
       "step": 575
     },
     {
-      "epoch": 1.399276236429433,
-      "grad_norm": 0.4765625,
-      "learning_rate": 0.00018679549114331725,
-      "loss": 1.0221,
+      "epoch": 2.7951807228915664,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00018086203354645089,
+      "loss": 0.9219,
       "step": 580
     },
     {
-      "epoch": 1.4113389626055488,
-      "grad_norm": 0.474609375,
-      "learning_rate": 0.00018840579710144927,
-      "loss": 1.0354,
+      "epoch": 2.819277108433735,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00018036311154549784,
+      "loss": 0.9304,
       "step": 585
     },
     {
-      "epoch": 1.4234016887816647,
-      "grad_norm": 0.3671875,
-      "learning_rate": 0.00019001610305958134,
-      "loss": 1.0257,
+      "epoch": 2.8433734939759034,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00017985847649204417,
+      "loss": 0.926,
       "step": 590
     },
     {
-      "epoch": 1.4354644149577804,
-      "grad_norm": 0.3515625,
-      "learning_rate": 0.0001916264090177134,
-      "loss": 1.0246,
+      "epoch": 2.8674698795180724,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00017934816426084008,
+      "loss": 0.9136,
       "step": 595
     },
     {
-      "epoch": 1.4475271411338961,
-      "grad_norm": 0.439453125,
-      "learning_rate": 0.0001932367149758454,
-      "loss": 1.0234,
+      "epoch": 2.891566265060241,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00017883221113022916,
+      "loss": 0.9189,
       "step": 600
     },
     {
-      "epoch": 1.459589867310012,
-      "grad_norm": 0.39453125,
-      "learning_rate": 0.00019484702093397745,
-      "loss": 1.0161,
+      "epoch": 2.9156626506024095,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001783106537795692,
+      "loss": 0.9138,
       "step": 605
     },
     {
-      "epoch": 1.471652593486128,
-      "grad_norm": 0.57421875,
-      "learning_rate": 0.00019645732689210952,
-      "loss": 1.017,
+      "epoch": 2.9397590361445785,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00017778352928662474,
+      "loss": 0.923,
       "step": 610
     },
     {
-      "epoch": 1.4837153196622437,
-      "grad_norm": 0.72265625,
-      "learning_rate": 0.00019806763285024154,
-      "loss": 1.0218,
+      "epoch": 2.963855421686747,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0001772508751249311,
+      "loss": 0.9259,
       "step": 615
     },
     {
-      "epoch": 1.4957780458383594,
-      "grad_norm": 0.609375,
-      "learning_rate": 0.0001996779388083736,
-      "loss": 1.0194,
+      "epoch": 2.9879518072289155,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00017671272916113052,
+      "loss": 0.9114,
       "step": 620
     },
     {
-      "epoch": 1.5078407720144753,
-      "grad_norm": 0.515625,
-      "learning_rate": 0.00019999974723244348,
-      "loss": 1.0206,
+      "epoch": 2.9975903614457833,
+      "eval_loss": 2.46407151222229,
+      "eval_runtime": 0.6669,
+      "eval_samples_per_second": 14.995,
+      "eval_steps_per_second": 1.5,
+      "step": 622
+    },
+    {
+      "epoch": 3.0120481927710845,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00017616912965228001,
+      "loss": 0.9017,
       "step": 625
     },
     {
-      "epoch": 1.5199034981905912,
-      "grad_norm": 0.3828125,
-      "learning_rate": 0.00019999872036643513,
-      "loss": 1.025,
+      "epoch": 3.036144578313253,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00017562011524313185,
+      "loss": 0.897,
       "step": 630
     },
     {
-      "epoch": 1.531966224366707,
-      "grad_norm": 0.66015625,
-      "learning_rate": 0.0001999969036121076,
-      "loss": 1.0246,
+      "epoch": 3.0602409638554215,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0001750657249633861,
+      "loss": 0.896,
       "step": 635
     },
     {
-      "epoch": 1.5440289505428226,
-      "grad_norm": 0.93359375,
-      "learning_rate": 0.00019999429698381143,
-      "loss": 1.0232,
+      "epoch": 3.0843373493975905,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00017450599822491615,
+      "loss": 0.9036,
       "step": 640
     },
     {
-      "epoch": 1.5560916767189386,
-      "grad_norm": 0.70703125,
-      "learning_rate": 0.00019999090050213636,
-      "loss": 1.0167,
+      "epoch": 3.108433734939759,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00017394097481896676,
+      "loss": 0.9018,
       "step": 645
     },
     {
-      "epoch": 1.5681544028950543,
-      "grad_norm": 0.62109375,
-      "learning_rate": 0.00019998671419391108,
-      "loss": 1.0087,
+      "epoch": 3.1325301204819276,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00017337069491332537,
+      "loss": 0.8983,
       "step": 650
     },
     {
-      "epoch": 1.58021712907117,
-      "grad_norm": 0.357421875,
-      "learning_rate": 0.00019998173809220325,
-      "loss": 1.0215,
+      "epoch": 3.1566265060240966,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00017279519904946647,
+      "loss": 0.896,
       "step": 655
     },
     {
-      "epoch": 1.5922798552472859,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00019997597223631895,
-      "loss": 1.0208,
+      "epoch": 3.180722891566265,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001722145281396697,
+      "loss": 0.907,
       "step": 660
     },
     {
-      "epoch": 1.6043425814234018,
-      "grad_norm": 0.390625,
-      "learning_rate": 0.0001999694166718026,
-      "loss": 1.01,
+      "epoch": 3.2048192771084336,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00017162872346411102,
+      "loss": 0.8986,
       "step": 665
     },
     {
-      "epoch": 1.6164053075995175,
-      "grad_norm": 0.37109375,
-      "learning_rate": 0.00019996207145043653,
-      "loss": 1.0138,
+      "epoch": 3.2289156626506026,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00017103782666792844,
+      "loss": 0.9046,
       "step": 670
     },
     {
-      "epoch": 1.6284680337756332,
-      "grad_norm": 0.375,
-      "learning_rate": 0.00019995393663024054,
-      "loss": 1.0069,
+      "epoch": 3.253012048192771,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00017044187975826124,
+      "loss": 0.9018,
       "step": 675
     },
     {
-      "epoch": 1.6405307599517491,
-      "grad_norm": 0.76171875,
-      "learning_rate": 0.00019994501227547137,
-      "loss": 1.0115,
+      "epoch": 3.2771084337349397,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00016984092510126367,
+      "loss": 0.903,
       "step": 680
     },
     {
-      "epoch": 1.652593486127865,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.00019993529845662244,
-      "loss": 1.0071,
+      "epoch": 3.3012048192771086,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001692350054190932,
+      "loss": 0.9077,
       "step": 685
     },
     {
-      "epoch": 1.6646562123039808,
-      "grad_norm": 0.515625,
-      "learning_rate": 0.00019992479525042303,
-      "loss": 1.0058,
+      "epoch": 3.325301204819277,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0001686241637868734,
+      "loss": 0.9033,
       "step": 690
     },
     {
-      "epoch": 1.6767189384800965,
-      "grad_norm": 0.6171875,
-      "learning_rate": 0.00019991350273983776,
-      "loss": 1.006,
+      "epoch": 3.3493975903614457,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00016800844362963147,
+      "loss": 0.9051,
       "step": 695
     },
     {
-      "epoch": 1.6887816646562124,
-      "grad_norm": 0.51171875,
-      "learning_rate": 0.00019990142101406603,
-      "loss": 1.0051,
+      "epoch": 3.3734939759036147,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00016738788871921152,
+      "loss": 0.8974,
       "step": 700
     },
     {
-      "epoch": 1.700844390832328,
-      "grad_norm": 0.466796875,
-      "learning_rate": 0.0001998885501685412,
-      "loss": 1.0037,
+      "epoch": 3.397590361445783,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00016676254317116252,
+      "loss": 0.9016,
       "step": 705
     },
     {
-      "epoch": 1.7129071170084438,
-      "grad_norm": 0.5546875,
-      "learning_rate": 0.0001998748903049299,
-      "loss": 1.0072,
+      "epoch": 3.4216867469879517,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0001661324514416022,
+      "loss": 0.9006,
       "step": 710
     },
     {
-      "epoch": 1.7249698431845597,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.0001998604415311311,
-      "loss": 1.0173,
+      "epoch": 3.4457831325301207,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00016549765832405653,
+      "loss": 0.9005,
       "step": 715
     },
     {
-      "epoch": 1.7370325693606756,
-      "grad_norm": 0.7109375,
-      "learning_rate": 0.00019984520396127553,
-      "loss": 1.0094,
+      "epoch": 3.4698795180722892,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0001648582089462756,
+      "loss": 0.9024,
       "step": 720
     },
     {
-      "epoch": 1.7490952955367913,
-      "grad_norm": 0.54296875,
-      "learning_rate": 0.00019982917771572443,
-      "loss": 1.0048,
+      "epoch": 3.4939759036144578,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00016421414876702518,
+      "loss": 0.8996,
       "step": 725
     },
     {
-      "epoch": 1.761158021712907,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.00019981236292106896,
-      "loss": 1.0063,
+      "epoch": 3.5180722891566267,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00016356552357285522,
+      "loss": 0.8988,
       "step": 730
     },
     {
-      "epoch": 1.773220747889023,
-      "grad_norm": 0.3515625,
-      "learning_rate": 0.0001997947597101288,
-      "loss": 1.0042,
+      "epoch": 3.5421686746987953,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0001629123794748447,
+      "loss": 0.8975,
       "step": 735
     },
     {
-      "epoch": 1.7852834740651389,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.00019977636822195153,
-      "loss": 0.9947,
+      "epoch": 3.566265060240964,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00016225476290532374,
+      "loss": 0.9,
       "step": 740
     },
     {
-      "epoch": 1.7973462002412546,
-      "grad_norm": 0.50390625,
-      "learning_rate": 0.0001997571886018112,
-      "loss": 1.0042,
+      "epoch": 3.5903614457831328,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00016159272061457255,
+      "loss": 0.8948,
       "step": 745
     },
     {
-      "epoch": 1.8094089264173703,
-      "grad_norm": 0.578125,
-      "learning_rate": 0.0001997372210012073,
-      "loss": 1.0002,
+      "epoch": 3.6144578313253013,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001609262996674981,
+      "loss": 0.8854,
       "step": 750
     },
     {
-      "epoch": 1.8214716525934862,
-      "grad_norm": 0.5625,
-      "learning_rate": 0.00019971646557786363,
-      "loss": 1.0061,
+      "epoch": 3.63855421686747,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0001602555474402881,
+      "loss": 0.8981,
       "step": 755
     },
     {
-      "epoch": 1.833534378769602,
-      "grad_norm": 0.408203125,
-      "learning_rate": 0.00019969492249572693,
-      "loss": 1.0024,
+      "epoch": 3.662650602409639,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00015958051161704307,
+      "loss": 0.8997,
       "step": 760
     },
     {
-      "epoch": 1.8455971049457176,
-      "grad_norm": 0.40625,
-      "learning_rate": 0.0001996725919249657,
-      "loss": 1.001,
+      "epoch": 3.6867469879518073,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00015890124018638638,
+      "loss": 0.8909,
       "step": 765
     },
     {
-      "epoch": 1.8576598311218335,
-      "grad_norm": 0.396484375,
-      "learning_rate": 0.0001996494740419687,
-      "loss": 1.0007,
+      "epoch": 3.710843373493976,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00015821778143805296,
+      "loss": 0.8904,
       "step": 770
     },
     {
-      "epoch": 1.8697225572979495,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.00019962556902934379,
-      "loss": 0.998,
+      "epoch": 3.734939759036145,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00015753018395945598,
+      "loss": 0.896,
       "step": 775
     },
     {
-      "epoch": 1.8817852834740652,
-      "grad_norm": 0.64453125,
-      "learning_rate": 0.00019960087707591626,
-      "loss": 0.9913,
+      "epoch": 3.7590361445783134,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00015683849663223308,
+      "loss": 0.9004,
       "step": 780
     },
     {
-      "epoch": 1.8938480096501809,
-      "grad_norm": 1.140625,
-      "learning_rate": 0.00019957539837672745,
-      "loss": 1.0082,
+      "epoch": 3.783132530120482,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00015614276862877113,
+      "loss": 0.8979,
       "step": 785
     },
     {
-      "epoch": 1.9059107358262968,
-      "grad_norm": 0.62890625,
-      "learning_rate": 0.00019954913313303323,
-      "loss": 1.0001,
+      "epoch": 3.807228915662651,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001554430494087107,
+      "loss": 0.893,
       "step": 790
     },
     {
-      "epoch": 1.9179734620024127,
-      "grad_norm": 0.68359375,
-      "learning_rate": 0.00019952208155230234,
-      "loss": 1.0091,
+      "epoch": 3.8313253012048194,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00015473938871542986,
+      "loss": 0.8977,
       "step": 795
     },
     {
-      "epoch": 1.9300361881785284,
-      "grad_norm": 0.5859375,
-      "learning_rate": 0.0001994942438482148,
-      "loss": 0.9919,
+      "epoch": 3.855421686746988,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00015403183657250788,
+      "loss": 0.8969,
       "step": 800
     },
     {
-      "epoch": 1.942098914354644,
-      "grad_norm": 0.419921875,
-      "learning_rate": 0.00019946562024066014,
-      "loss": 0.982,
+      "epoch": 3.8795180722891565,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00015332044328016914,
+      "loss": 0.8921,
       "step": 805
     },
     {
-      "epoch": 1.95416164053076,
-      "grad_norm": 0.46484375,
-      "learning_rate": 0.00019943621095573586,
-      "loss": 1.0035,
+      "epoch": 3.9036144578313254,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00015260525941170712,
+      "loss": 0.8913,
       "step": 810
     },
     {
-      "epoch": 1.9662243667068757,
-      "grad_norm": 0.373046875,
-      "learning_rate": 0.00019940601622574542,
-      "loss": 0.9854,
+      "epoch": 3.927710843373494,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00015188633580988926,
+      "loss": 0.896,
       "step": 815
     },
     {
-      "epoch": 1.9782870928829914,
-      "grad_norm": 0.49609375,
-      "learning_rate": 0.00019937503628919648,
-      "loss": 0.9947,
+      "epoch": 3.9518072289156625,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00015116372358334233,
+      "loss": 0.892,
       "step": 820
     },
     {
-      "epoch": 1.9903498190591074,
-      "grad_norm": 0.546875,
-      "learning_rate": 0.00019934327139079915,
-      "loss": 1.0004,
+      "epoch": 3.9759036144578315,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00015043747410291945,
+      "loss": 0.8948,
       "step": 825
     },
     {
-      "epoch": 2.0,
-      "eval_loss": 2.448092460632324,
-      "eval_runtime": 0.534,
-      "eval_samples_per_second": 18.726,
-      "eval_steps_per_second": 1.873,
-      "step": 829
+      "epoch": 4.0,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00014970763899804763,
+      "loss": 0.8927,
+      "step": 830
     },
     {
-      "epoch": 2.0024125452352233,
-      "grad_norm": 0.57421875,
-      "learning_rate": 0.00019931072178146374,
-      "loss": 0.9899,
+      "epoch": 4.0,
+      "eval_loss": 2.4882054328918457,
+      "eval_runtime": 0.482,
+      "eval_samples_per_second": 20.748,
+      "eval_steps_per_second": 2.075,
       "step": 830
     },
     {
-      "epoch": 2.0144752714113388,
-      "grad_norm": 0.423828125,
-      "learning_rate": 0.0001992773877182992,
-      "loss": 0.9899,
+      "epoch": 4.024096385542169,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001489742701530578,
+      "loss": 0.8742,
       "step": 835
     },
     {
-      "epoch": 2.0265379975874547,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00019924326946461074,
-      "loss": 0.9803,
+      "epoch": 4.048192771084337,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00014823741970349606,
+      "loss": 0.8702,
       "step": 840
     },
     {
-      "epoch": 2.0386007237635706,
-      "grad_norm": 0.291015625,
-      "learning_rate": 0.00019920836728989794,
-      "loss": 0.9873,
+      "epoch": 4.072289156626506,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0001474971400324177,
+      "loss": 0.8777,
       "step": 845
     },
     {
-      "epoch": 2.0506634499396865,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00019917268146985245,
-      "loss": 0.9799,
+      "epoch": 4.096385542168675,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00014675348376666278,
+      "loss": 0.8776,
       "step": 850
     },
     {
-      "epoch": 2.062726176115802,
-      "grad_norm": 0.3515625,
-      "learning_rate": 0.0001991362122863561,
-      "loss": 0.9773,
+      "epoch": 4.120481927710843,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00014600650377311522,
+      "loss": 0.875,
       "step": 855
     },
     {
-      "epoch": 2.074788902291918,
-      "grad_norm": 0.412109375,
-      "learning_rate": 0.00019909896002747831,
-      "loss": 0.9818,
+      "epoch": 4.144578313253012,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00014525625315494435,
+      "loss": 0.8817,
       "step": 860
     },
     {
-      "epoch": 2.086851628468034,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00019906092498747416,
-      "loss": 0.9779,
+      "epoch": 4.168674698795181,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00014450278524782986,
+      "loss": 0.8808,
       "step": 865
     },
     {
-      "epoch": 2.0989143546441498,
-      "grad_norm": 0.478515625,
-      "learning_rate": 0.0001990221074667818,
-      "loss": 0.9807,
+      "epoch": 4.192771084337349,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00014374615361616985,
+      "loss": 0.8775,
       "step": 870
     },
     {
-      "epoch": 2.1109770808202653,
-      "grad_norm": 0.408203125,
-      "learning_rate": 0.00019898250777202018,
-      "loss": 0.9841,
+      "epoch": 4.216867469879518,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00014298641204927342,
+      "loss": 0.8796,
       "step": 875
     },
     {
-      "epoch": 2.123039806996381,
-      "grad_norm": 0.470703125,
-      "learning_rate": 0.0001989421262159867,
-      "loss": 0.9812,
+      "epoch": 4.240963855421687,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0001422236145575362,
+      "loss": 0.8753,
       "step": 880
     },
     {
-      "epoch": 2.135102533172497,
-      "grad_norm": 0.65625,
-      "learning_rate": 0.00019890096311765465,
-      "loss": 0.9709,
+      "epoch": 4.265060240963855,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00014145781536860122,
+      "loss": 0.8735,
       "step": 885
     },
     {
-      "epoch": 2.147165259348613,
-      "grad_norm": 0.349609375,
-      "learning_rate": 0.0001988590188021707,
-      "loss": 0.981,
+      "epoch": 4.289156626506024,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00014068906892350343,
+      "loss": 0.8753,
       "step": 890
     },
     {
-      "epoch": 2.1592279855247285,
-      "grad_norm": 0.47265625,
-      "learning_rate": 0.0001988162936008523,
-      "loss": 0.9715,
+      "epoch": 4.313253012048193,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001399174298727998,
+      "loss": 0.8752,
       "step": 895
     },
     {
-      "epoch": 2.1712907117008444,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00019877278785118517,
-      "loss": 0.9842,
+      "epoch": 4.337349397590361,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00013914295307268396,
+      "loss": 0.8734,
       "step": 900
     },
     {
-      "epoch": 2.1833534378769603,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00019872850189682053,
-      "loss": 0.9776,
+      "epoch": 4.36144578313253,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00013836569358108647,
+      "loss": 0.8815,
       "step": 905
     },
     {
-      "epoch": 2.195416164053076,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.00019868343608757245,
-      "loss": 0.9921,
+      "epoch": 4.385542168674699,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00013758570665376086,
+      "loss": 0.8834,
       "step": 910
     },
     {
-      "epoch": 2.2074788902291917,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.00019863759077941504,
-      "loss": 0.9835,
+      "epoch": 4.409638554216867,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00013680304774035538,
+      "loss": 0.8787,
       "step": 915
     },
     {
-      "epoch": 2.2195416164053077,
-      "grad_norm": 0.48828125,
-      "learning_rate": 0.00019859096633447965,
-      "loss": 0.9828,
+      "epoch": 4.433734939759036,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00013601777248047105,
+      "loss": 0.8785,
       "step": 920
     },
     {
-      "epoch": 2.2316043425814236,
-      "grad_norm": 0.4765625,
-      "learning_rate": 0.00019854356312105207,
-      "loss": 0.9767,
+      "epoch": 4.457831325301205,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0001352299366997062,
+      "loss": 0.8729,
       "step": 925
     },
     {
-      "epoch": 2.243667068757539,
-      "grad_norm": 0.423828125,
-      "learning_rate": 0.00019849538151356955,
-      "loss": 0.982,
+      "epoch": 4.481927710843373,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001344395964056878,
+      "loss": 0.8778,
       "step": 930
     },
     {
-      "epoch": 2.255729794933655,
-      "grad_norm": 0.384765625,
-      "learning_rate": 0.00019844642189261778,
-      "loss": 0.9848,
+      "epoch": 4.506024096385542,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00013364680778409,
+      "loss": 0.8816,
       "step": 935
     },
     {
-      "epoch": 2.267792521109771,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00019839668464492811,
-      "loss": 0.9735,
+      "epoch": 4.530120481927711,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00013285162719463961,
+      "loss": 0.8762,
       "step": 940
     },
     {
-      "epoch": 2.2798552472858864,
-      "grad_norm": 0.365234375,
-      "learning_rate": 0.0001983461701633742,
-      "loss": 0.977,
+      "epoch": 4.554216867469879,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00013205411116710972,
+      "loss": 0.8776,
       "step": 945
     },
     {
-      "epoch": 2.2919179734620023,
-      "grad_norm": 0.46875,
-      "learning_rate": 0.00019829487884696925,
-      "loss": 0.9785,
+      "epoch": 4.578313253012048,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0001312543163973007,
+      "loss": 0.8674,
       "step": 950
     },
     {
-      "epoch": 2.3039806996381182,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.00019824281110086248,
-      "loss": 0.9816,
+      "epoch": 4.602409638554217,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00013045229974300993,
+      "loss": 0.8759,
       "step": 955
     },
     {
-      "epoch": 2.316043425814234,
-      "grad_norm": 0.416015625,
-      "learning_rate": 0.00019818996733633618,
-      "loss": 0.988,
+      "epoch": 4.626506024096385,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0001296481182199896,
+      "loss": 0.8727,
       "step": 960
     },
     {
-      "epoch": 2.3281061519903496,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.0001981363479708025,
-      "loss": 0.9873,
+      "epoch": 4.650602409638554,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00012884182899789343,
+      "loss": 0.8739,
       "step": 965
     },
     {
-      "epoch": 2.3401688781664656,
-      "grad_norm": 0.349609375,
-      "learning_rate": 0.00019808195342779992,
-      "loss": 0.9812,
+      "epoch": 4.674698795180722,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00012803348939621252,
+      "loss": 0.874,
       "step": 970
     },
     {
-      "epoch": 2.3522316043425815,
-      "grad_norm": 0.396484375,
-      "learning_rate": 0.00019802678413699006,
-      "loss": 0.9688,
+      "epoch": 4.698795180722891,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00012722315688020047,
+      "loss": 0.8745,
       "step": 975
     },
     {
-      "epoch": 2.3642943305186974,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.0001979708405341543,
-      "loss": 0.9744,
+      "epoch": 4.72289156626506,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00012641088905678802,
+      "loss": 0.8738,
       "step": 980
     },
     {
-      "epoch": 2.376357056694813,
-      "grad_norm": 0.36328125,
-      "learning_rate": 0.00019791412306119027,
-      "loss": 0.9682,
+      "epoch": 4.746987951807229,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.000125596743670488,
+      "loss": 0.8721,
       "step": 985
     },
     {
-      "epoch": 2.388419782870929,
-      "grad_norm": 0.3515625,
-      "learning_rate": 0.00019785663216610844,
-      "loss": 0.9854,
+      "epoch": 4.771084337349397,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00012478077859929,
+      "loss": 0.8734,
       "step": 990
     },
     {
-      "epoch": 2.4004825090470447,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00019779836830302843,
-      "loss": 0.9793,
+      "epoch": 4.795180722891566,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0001239630518505459,
+      "loss": 0.876,
       "step": 995
     },
     {
-      "epoch": 2.4125452352231607,
-      "grad_norm": 0.359375,
-      "learning_rate": 0.0001977393319321756,
-      "loss": 0.9852,
+      "epoch": 4.8192771084337345,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00012314362155684612,
+      "loss": 0.874,
       "step": 1000
     },
     {
-      "epoch": 2.424607961399276,
-      "grad_norm": 0.58203125,
-      "learning_rate": 0.0001976795235198773,
-      "loss": 0.9759,
+      "epoch": 4.843373493975903,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00012232254597188688,
+      "loss": 0.8752,
       "step": 1005
     },
     {
-      "epoch": 2.436670687575392,
-      "grad_norm": 0.42578125,
-      "learning_rate": 0.00019761894353855926,
-      "loss": 0.9689,
+      "epoch": 4.867469879518072,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00012149988346632894,
+      "loss": 0.8689,
       "step": 1010
     },
     {
-      "epoch": 2.448733413751508,
-      "grad_norm": 0.55859375,
-      "learning_rate": 0.0001975575924667418,
-      "loss": 0.9717,
+      "epoch": 4.891566265060241,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00012067569252364809,
+      "loss": 0.8704,
       "step": 1015
     },
     {
-      "epoch": 2.4607961399276235,
-      "grad_norm": 0.37890625,
-      "learning_rate": 0.00019749547078903604,
-      "loss": 0.9769,
+      "epoch": 4.9156626506024095,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0001198500317359774,
+      "loss": 0.8785,
       "step": 1020
     },
     {
-      "epoch": 2.4728588661037394,
-      "grad_norm": 0.451171875,
-      "learning_rate": 0.00019743257899614017,
-      "loss": 0.9699,
+      "epoch": 4.9397590361445785,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00011902295979994192,
+      "loss": 0.8732,
       "step": 1025
     },
     {
-      "epoch": 2.4849215922798553,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00019736891758483544,
-      "loss": 0.9818,
+      "epoch": 4.9638554216867465,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00011819453551248592,
+      "loss": 0.8796,
       "step": 1030
     },
     {
-      "epoch": 2.4969843184559712,
-      "grad_norm": 0.37109375,
-      "learning_rate": 0.00019730448705798239,
-      "loss": 0.9716,
+      "epoch": 4.9879518072289155,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00011736481776669306,
+      "loss": 0.8752,
       "step": 1035
     },
     {
-      "epoch": 2.5090470446320867,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.00019723928792451668,
-      "loss": 0.963,
+      "epoch": 4.997590361445783,
+      "eval_loss": 2.508066177368164,
+      "eval_runtime": 0.5,
+      "eval_samples_per_second": 20.0,
+      "eval_steps_per_second": 2.0,
+      "step": 1037
+    },
+    {
+      "epoch": 5.0120481927710845,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00011653386554759946,
+      "loss": 0.8612,
       "step": 1040
     },
     {
-      "epoch": 2.5211097708082026,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00019717332069944528,
-      "loss": 0.9629,
+      "epoch": 5.036144578313253,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00011570173792800066,
+      "loss": 0.8559,
       "step": 1045
     },
     {
-      "epoch": 2.5331724969843186,
-      "grad_norm": 0.353515625,
-      "learning_rate": 0.00019710658590384227,
-      "loss": 0.9741,
+      "epoch": 5.0602409638554215,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00011486849406425188,
+      "loss": 0.8507,
       "step": 1050
     },
     {
-      "epoch": 2.545235223160434,
-      "grad_norm": 0.59375,
-      "learning_rate": 0.00019703908406484467,
-      "loss": 0.9674,
+      "epoch": 5.0843373493975905,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00011403419319206284,
+      "loss": 0.8603,
       "step": 1055
     },
     {
-      "epoch": 2.55729794933655,
-      "grad_norm": 0.392578125,
-      "learning_rate": 0.00019697081571564854,
-      "loss": 0.9714,
+      "epoch": 5.108433734939759,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0001131988946222863,
+      "loss": 0.8592,
       "step": 1060
     },
     {
-      "epoch": 2.569360675512666,
-      "grad_norm": 0.39453125,
-      "learning_rate": 0.00019690178139550443,
-      "loss": 0.9656,
+      "epoch": 5.132530120481928,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00011236265773670196,
+      "loss": 0.855,
       "step": 1065
     },
     {
-      "epoch": 2.581423401688782,
-      "grad_norm": 0.35546875,
-      "learning_rate": 0.0001968319816497134,
-      "loss": 0.968,
+      "epoch": 5.156626506024097,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00011152554198379484,
+      "loss": 0.8566,
       "step": 1070
     },
     {
-      "epoch": 2.5934861278648977,
-      "grad_norm": 0.53515625,
-      "learning_rate": 0.00019676141702962253,
-      "loss": 0.9676,
+      "epoch": 5.180722891566265,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00011068760687452895,
+      "loss": 0.8521,
       "step": 1075
     },
     {
-      "epoch": 2.605548854041013,
-      "grad_norm": 0.359375,
-      "learning_rate": 0.00019669008809262062,
-      "loss": 0.9733,
+      "epoch": 5.204819277108434,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00010984891197811687,
+      "loss": 0.8605,
       "step": 1080
     },
     {
-      "epoch": 2.617611580217129,
-      "grad_norm": 0.486328125,
-      "learning_rate": 0.00019661799540213382,
-      "loss": 0.967,
+      "epoch": 5.228915662650603,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00010900951691778481,
+      "loss": 0.8623,
       "step": 1085
     },
     {
-      "epoch": 2.6296743063932446,
-      "grad_norm": 0.4375,
-      "learning_rate": 0.0001965451395276212,
-      "loss": 0.9751,
+      "epoch": 5.253012048192771,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00010816948136653386,
+      "loss": 0.8583,
       "step": 1090
     },
     {
-      "epoch": 2.6417370325693605,
-      "grad_norm": 0.62109375,
-      "learning_rate": 0.00019647152104457013,
-      "loss": 0.9745,
+      "epoch": 5.27710843373494,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.000107328865042898,
+      "loss": 0.8571,
       "step": 1095
     },
     {
-      "epoch": 2.6537997587454765,
-      "grad_norm": 0.39453125,
-      "learning_rate": 0.0001963971405344919,
-      "loss": 0.9751,
+      "epoch": 5.301204819277109,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00010648772770669861,
+      "loss": 0.862,
       "step": 1100
     },
     {
-      "epoch": 2.6658624849215924,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.0001963219985849169,
-      "loss": 0.9724,
+      "epoch": 5.325301204819277,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00010564612915479612,
+      "loss": 0.8531,
       "step": 1105
     },
     {
-      "epoch": 2.6779252110977083,
-      "grad_norm": 0.376953125,
-      "learning_rate": 0.00019624609578939027,
-      "loss": 0.973,
+      "epoch": 5.349397590361446,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00010480412921683888,
+      "loss": 0.8646,
       "step": 1110
     },
     {
-      "epoch": 2.689987937273824,
-      "grad_norm": 0.37890625,
-      "learning_rate": 0.000196169432747467,
-      "loss": 0.9771,
+      "epoch": 5.373493975903615,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00010396178775101014,
+      "loss": 0.858,
       "step": 1115
     },
     {
-      "epoch": 2.7020506634499397,
-      "grad_norm": 0.46484375,
-      "learning_rate": 0.00019609201006470724,
-      "loss": 0.9671,
+      "epoch": 5.397590361445783,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00010311916463977242,
+      "loss": 0.863,
       "step": 1120
     },
     {
-      "epoch": 2.7141133896260556,
-      "grad_norm": 0.443359375,
-      "learning_rate": 0.0001960138283526715,
-      "loss": 0.9698,
+      "epoch": 5.421686746987952,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00010227631978561056,
+      "loss": 0.8601,
       "step": 1125
     },
     {
-      "epoch": 2.726176115802171,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.000195934888228916,
-      "loss": 0.9645,
+      "epoch": 5.445783132530121,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00010143331310677331,
+      "loss": 0.8566,
       "step": 1130
     },
     {
-      "epoch": 2.738238841978287,
-      "grad_norm": 0.435546875,
-      "learning_rate": 0.0001958551903169874,
-      "loss": 0.9665,
+      "epoch": 5.469879518072289,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00010059020453301345,
+      "loss": 0.8538,
       "step": 1135
     },
     {
-      "epoch": 2.750301568154403,
-      "grad_norm": 0.365234375,
-      "learning_rate": 0.0001957747352464184,
-      "loss": 0.9719,
+      "epoch": 5.493975903614458,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.974705400132764e-05,
+      "loss": 0.8591,
       "step": 1140
     },
     {
-      "epoch": 2.762364294330519,
-      "grad_norm": 0.4140625,
-      "learning_rate": 0.0001956935236527222,
-      "loss": 0.9624,
+      "epoch": 5.518072289156627,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.890392145169531e-05,
+      "loss": 0.8576,
       "step": 1145
     },
     {
-      "epoch": 2.7744270205066344,
-      "grad_norm": 0.357421875,
-      "learning_rate": 0.00019561155617738797,
-      "loss": 0.9686,
+      "epoch": 5.542168674698795,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 9.806086682281758e-05,
+      "loss": 0.8615,
       "step": 1150
     },
     {
-      "epoch": 2.7864897466827503,
-      "grad_norm": 0.421875,
-      "learning_rate": 0.00019552883346787552,
-      "loss": 0.962,
+      "epoch": 5.566265060240964,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.721795004785605e-05,
+      "loss": 0.8572,
       "step": 1155
     },
     {
-      "epoch": 2.798552472858866,
-      "grad_norm": 0.423828125,
-      "learning_rate": 0.0001954453561776102,
-      "loss": 0.9684,
+      "epoch": 5.590361445783133,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.637523105017229e-05,
+      "loss": 0.8644,
       "step": 1160
     },
     {
-      "epoch": 2.8106151990349817,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.00019536112496597782,
-      "loss": 0.9759,
+      "epoch": 5.614457831325301,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 9.553276973906786e-05,
+      "loss": 0.854,
       "step": 1165
     },
     {
-      "epoch": 2.8226779252110976,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.0001952761404983194,
-      "loss": 0.9744,
+      "epoch": 5.63855421686747,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.469062600552509e-05,
+      "loss": 0.8526,
       "step": 1170
     },
     {
-      "epoch": 2.8347406513872135,
-      "grad_norm": 0.37890625,
-      "learning_rate": 0.00019519040344592593,
-      "loss": 0.9714,
+      "epoch": 5.662650602409639,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 9.384885971794961e-05,
+      "loss": 0.8563,
       "step": 1175
     },
     {
-      "epoch": 2.8468033775633295,
-      "grad_norm": 0.388671875,
-      "learning_rate": 0.000195103914486033,
-      "loss": 0.965,
+      "epoch": 5.686746987951807,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.300753071791434e-05,
+      "loss": 0.8614,
       "step": 1180
     },
     {
-      "epoch": 2.8588661037394454,
-      "grad_norm": 0.55078125,
-      "learning_rate": 0.0001950166743018156,
-      "loss": 0.9572,
+      "epoch": 5.710843373493976,
+      "grad_norm": 0.22265625,
+      "learning_rate": 9.216669881590515e-05,
+      "loss": 0.8633,
       "step": 1185
     },
     {
-      "epoch": 2.870928829915561,
-      "grad_norm": 0.44140625,
-      "learning_rate": 0.00019492868358238249,
-      "loss": 0.9563,
+      "epoch": 5.734939759036145,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.132642378706894e-05,
+      "loss": 0.8548,
       "step": 1190
     },
     {
-      "epoch": 2.882991556091677,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00019483994302277102,
-      "loss": 0.9597,
+      "epoch": 5.759036144578313,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.048676536696425e-05,
+      "loss": 0.8603,
       "step": 1195
     },
     {
-      "epoch": 2.8950542822677923,
-      "grad_norm": 0.42578125,
-      "learning_rate": 0.00019475045332394153,
-      "loss": 0.965,
+      "epoch": 5.783132530120482,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.964778324731467e-05,
+      "loss": 0.8643,
       "step": 1200
     },
     {
-      "epoch": 2.907117008443908,
-      "grad_norm": 0.443359375,
-      "learning_rate": 0.00019466021519277165,
-      "loss": 0.9562,
+      "epoch": 5.807228915662651,
+      "grad_norm": 0.330078125,
+      "learning_rate": 8.880953707176514e-05,
+      "loss": 0.8562,
       "step": 1205
     },
     {
-      "epoch": 2.919179734620024,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.000194569229342051,
-      "loss": 0.9573,
+      "epoch": 5.831325301204819,
+      "grad_norm": 0.25,
+      "learning_rate": 8.797208643164212e-05,
+      "loss": 0.8565,
       "step": 1210
     },
     {
-      "epoch": 2.93124246079614,
-      "grad_norm": 0.474609375,
-      "learning_rate": 0.00019447749649047542,
-      "loss": 0.9629,
+      "epoch": 5.855421686746988,
+      "grad_norm": 0.234375,
+      "learning_rate": 8.713549086171691e-05,
+      "loss": 0.8566,
       "step": 1215
     },
     {
-      "epoch": 2.943305186972256,
-      "grad_norm": 0.6015625,
-      "learning_rate": 0.0001943850173626412,
-      "loss": 0.978,
+      "epoch": 5.879518072289157,
+      "grad_norm": 0.279296875,
+      "learning_rate": 8.629980983597358e-05,
+      "loss": 0.8596,
       "step": 1220
     },
     {
-      "epoch": 2.9553679131483714,
-      "grad_norm": 0.50390625,
-      "learning_rate": 0.00019429179268903957,
-      "loss": 0.9696,
+      "epoch": 5.903614457831325,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.546510276338078e-05,
+      "loss": 0.8522,
       "step": 1225
     },
     {
-      "epoch": 2.9674306393244874,
-      "grad_norm": 0.3515625,
-      "learning_rate": 0.0001941978232060507,
-      "loss": 0.9682,
+      "epoch": 5.927710843373494,
+      "grad_norm": 0.263671875,
+      "learning_rate": 8.463142898366834e-05,
+      "loss": 0.8561,
       "step": 1230
     },
     {
-      "epoch": 2.9794933655006033,
-      "grad_norm": 0.3828125,
-      "learning_rate": 0.00019410310965593813,
-      "loss": 0.9498,
+      "epoch": 5.951807228915663,
+      "grad_norm": 0.275390625,
+      "learning_rate": 8.37988477631088e-05,
+      "loss": 0.8611,
       "step": 1235
     },
     {
-      "epoch": 2.9915560916767188,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.00019400765278684262,
-      "loss": 0.9586,
+      "epoch": 5.975903614457831,
+      "grad_norm": 0.267578125,
+      "learning_rate": 8.296741829030418e-05,
+      "loss": 0.8552,
       "step": 1240
     },
     {
-      "epoch": 2.9987937273823886,
-      "eval_loss": 2.4426419734954834,
-      "eval_runtime": 0.66,
-      "eval_samples_per_second": 15.151,
-      "eval_steps_per_second": 1.515,
-      "step": 1243
+      "epoch": 6.0,
+      "grad_norm": 0.248046875,
+      "learning_rate": 8.213719967197817e-05,
+      "loss": 0.8602,
+      "step": 1245
     },
     {
-      "epoch": 3.0036188178528347,
-      "grad_norm": 0.41015625,
-      "learning_rate": 0.00019391145335277655,
-      "loss": 0.9537,
+      "epoch": 6.0,
+      "eval_loss": 2.5277345180511475,
+      "eval_runtime": 0.4824,
+      "eval_samples_per_second": 20.732,
+      "eval_steps_per_second": 2.073,
       "step": 1245
     },
     {
-      "epoch": 3.0156815440289506,
-      "grad_norm": 0.423828125,
-      "learning_rate": 0.00019381451211361773,
-      "loss": 0.9384,
+      "epoch": 6.024096385542169,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 8.130825092877418e-05,
+      "loss": 0.8432,
       "step": 1250
     },
     {
-      "epoch": 3.0277442702050665,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00019371682983510344,
-      "loss": 0.9493,
+      "epoch": 6.048192771084337,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 8.04806309910597e-05,
+      "loss": 0.8387,
       "step": 1255
     },
     {
-      "epoch": 3.039806996381182,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00019361840728882447,
-      "loss": 0.9441,
+      "epoch": 6.072289156626506,
+      "grad_norm": 0.21484375,
+      "learning_rate": 7.965439869473664e-05,
+      "loss": 0.8463,
       "step": 1260
     },
     {
-      "epoch": 3.051869722557298,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.000193519245252219,
-      "loss": 0.9457,
+      "epoch": 6.096385542168675,
+      "grad_norm": 0.21875,
+      "learning_rate": 7.882961277705895e-05,
+      "loss": 0.8402,
       "step": 1265
     },
     {
-      "epoch": 3.063932448733414,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00019341934450856633,
-      "loss": 0.9498,
+      "epoch": 6.120481927710843,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 7.800633187245673e-05,
+      "loss": 0.8472,
       "step": 1270
     },
     {
-      "epoch": 3.0759951749095293,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.00019331870584698093,
-      "loss": 0.9539,
+      "epoch": 6.144578313253012,
+      "grad_norm": 0.271484375,
+      "learning_rate": 7.718461450836804e-05,
+      "loss": 0.8457,
       "step": 1275
     },
     {
-      "epoch": 3.0880579010856453,
-      "grad_norm": 0.39453125,
-      "learning_rate": 0.000193217330062406,
-      "loss": 0.9564,
+      "epoch": 6.168674698795181,
+      "grad_norm": 0.27734375,
+      "learning_rate": 7.636451910107806e-05,
+      "loss": 0.8381,
       "step": 1280
     },
     {
-      "epoch": 3.100120627261761,
-      "grad_norm": 0.296875,
-      "learning_rate": 0.00019311521795560723,
-      "loss": 0.9492,
+      "epoch": 6.192771084337349,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 7.554610395156624e-05,
+      "loss": 0.8384,
       "step": 1285
     },
     {
-      "epoch": 3.112183353437877,
-      "grad_norm": 0.359375,
-      "learning_rate": 0.00019301237033316659,
-      "loss": 0.9549,
+      "epoch": 6.216867469879518,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 7.472942724136174e-05,
+      "loss": 0.8468,
       "step": 1290
     },
     {
-      "epoch": 3.1242460796139926,
-      "grad_norm": 0.38671875,
-      "learning_rate": 0.00019290878800747578,
-      "loss": 0.9511,
+      "epoch": 6.240963855421687,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 7.391454702840722e-05,
+      "loss": 0.8472,
       "step": 1295
     },
     {
-      "epoch": 3.1363088057901085,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00019280447179673,
-      "loss": 0.9395,
+      "epoch": 6.265060240963855,
+      "grad_norm": 0.251953125,
+      "learning_rate": 7.310152124293146e-05,
+      "loss": 0.8441,
       "step": 1300
     },
     {
-      "epoch": 3.1483715319662244,
-      "grad_norm": 0.388671875,
-      "learning_rate": 0.00019269942252492133,
-      "loss": 0.951,
+      "epoch": 6.289156626506024,
+      "grad_norm": 0.251953125,
+      "learning_rate": 7.229040768333115e-05,
+      "loss": 0.8424,
       "step": 1305
     },
     {
-      "epoch": 3.1604342581423404,
-      "grad_norm": 0.3828125,
-      "learning_rate": 0.00019259364102183234,
-      "loss": 0.9471,
+      "epoch": 6.313253012048193,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 7.14812640120618e-05,
+      "loss": 0.8377,
       "step": 1310
     },
     {
-      "epoch": 3.172496984318456,
-      "grad_norm": 0.3515625,
-      "learning_rate": 0.00019248712812302947,
-      "loss": 0.9587,
+      "epoch": 6.337349397590361,
+      "grad_norm": 0.25390625,
+      "learning_rate": 7.067414775153871e-05,
+      "loss": 0.8428,
       "step": 1315
     },
     {
-      "epoch": 3.1845597104945718,
-      "grad_norm": 0.39453125,
-      "learning_rate": 0.0001923798846698564,
-      "loss": 0.9591,
+      "epoch": 6.36144578313253,
+      "grad_norm": 0.31640625,
+      "learning_rate": 6.986911628004753e-05,
+      "loss": 0.8411,
       "step": 1320
     },
     {
-      "epoch": 3.1966224366706877,
-      "grad_norm": 0.44140625,
-      "learning_rate": 0.0001922719115094275,
-      "loss": 0.9437,
+      "epoch": 6.385542168674699,
+      "grad_norm": 0.302734375,
+      "learning_rate": 6.906622682766526e-05,
+      "loss": 0.8451,
       "step": 1325
     },
     {
-      "epoch": 3.2086851628468036,
-      "grad_norm": 0.6171875,
-      "learning_rate": 0.00019216320949462103,
-      "loss": 0.9494,
+      "epoch": 6.409638554216867,
+      "grad_norm": 0.263671875,
+      "learning_rate": 6.826553647219175e-05,
+      "loss": 0.8468,
       "step": 1330
     },
     {
-      "epoch": 3.220747889022919,
-      "grad_norm": 0.8125,
-      "learning_rate": 0.00019205377948407258,
-      "loss": 0.9532,
+      "epoch": 6.433734939759036,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 6.74671021350919e-05,
+      "loss": 0.8499,
       "step": 1335
     },
     {
-      "epoch": 3.232810615199035,
-      "grad_norm": 0.361328125,
-      "learning_rate": 0.00019194362234216795,
-      "loss": 0.9568,
+      "epoch": 6.457831325301205,
+      "grad_norm": 0.234375,
+      "learning_rate": 6.667098057744927e-05,
+      "loss": 0.8447,
       "step": 1340
     },
     {
-      "epoch": 3.244873341375151,
-      "grad_norm": 0.94921875,
-      "learning_rate": 0.0001918327389390368,
-      "loss": 0.9531,
+      "epoch": 6.481927710843373,
+      "grad_norm": 0.31640625,
+      "learning_rate": 6.587722839593073e-05,
+      "loss": 0.8462,
       "step": 1345
     },
     {
-      "epoch": 3.2569360675512664,
-      "grad_norm": 0.384765625,
-      "learning_rate": 0.00019172113015054532,
-      "loss": 0.9561,
+      "epoch": 6.506024096385542,
+      "grad_norm": 0.3125,
+      "learning_rate": 6.508590201876317e-05,
+      "loss": 0.8477,
       "step": 1350
     },
     {
-      "epoch": 3.2689987937273823,
-      "grad_norm": 0.5703125,
-      "learning_rate": 0.0001916087968582896,
-      "loss": 0.9554,
+      "epoch": 6.530120481927711,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 6.429705770172168e-05,
+      "loss": 0.8473,
       "step": 1355
     },
     {
-      "epoch": 3.2810615199034983,
-      "grad_norm": 0.419921875,
-      "learning_rate": 0.0001914957399495885,
-      "loss": 0.95,
+      "epoch": 6.554216867469879,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 6.351075152413068e-05,
+      "loss": 0.8411,
       "step": 1360
     },
     {
-      "epoch": 3.293124246079614,
-      "grad_norm": 0.64453125,
-      "learning_rate": 0.00019138196031747681,
-      "loss": 0.9599,
+      "epoch": 6.578313253012048,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 6.272703938487694e-05,
+      "loss": 0.8514,
       "step": 1365
     },
     {
-      "epoch": 3.3051869722557297,
-      "grad_norm": 0.5859375,
-      "learning_rate": 0.00019126745886069807,
-      "loss": 0.9546,
+      "epoch": 6.602409638554217,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 6.194597699843581e-05,
+      "loss": 0.8415,
       "step": 1370
     },
     {
-      "epoch": 3.3172496984318456,
-      "grad_norm": 0.59375,
-      "learning_rate": 0.0001911522364836974,
-      "loss": 0.9521,
+      "epoch": 6.626506024096385,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 6.116761989091042e-05,
+      "loss": 0.8429,
       "step": 1375
     },
     {
-      "epoch": 3.3293124246079615,
-      "grad_norm": 0.365234375,
-      "learning_rate": 0.0001910362940966147,
-      "loss": 0.9571,
+      "epoch": 6.650602409638554,
+      "grad_norm": 0.228515625,
+      "learning_rate": 6.039202339608432e-05,
+      "loss": 0.8435,
       "step": 1380
     },
     {
-      "epoch": 3.341375150784077,
-      "grad_norm": 0.5078125,
-      "learning_rate": 0.0001909196326152769,
-      "loss": 0.9514,
+      "epoch": 6.674698795180722,
+      "grad_norm": 0.220703125,
+      "learning_rate": 5.961924265148777e-05,
+      "loss": 0.8455,
       "step": 1385
     },
     {
-      "epoch": 3.353437876960193,
-      "grad_norm": 0.66796875,
-      "learning_rate": 0.00019080225296119125,
-      "loss": 0.9482,
+      "epoch": 6.698795180722891,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 5.884933259447798e-05,
+      "loss": 0.8457,
       "step": 1390
     },
     {
-      "epoch": 3.365500603136309,
-      "grad_norm": 0.494140625,
-      "learning_rate": 0.00019068415606153787,
-      "loss": 0.9489,
+      "epoch": 6.72289156626506,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 5.8082347958333625e-05,
+      "loss": 0.8481,
       "step": 1395
     },
     {
-      "epoch": 3.3775633293124248,
-      "grad_norm": 0.462890625,
-      "learning_rate": 0.00019056534284916224,
-      "loss": 0.944,
+      "epoch": 6.746987951807229,
+      "grad_norm": 0.298828125,
+      "learning_rate": 5.731834326836366e-05,
+      "loss": 0.8469,
       "step": 1400
     },
     {
-      "epoch": 3.3896260554885402,
-      "grad_norm": 0.40625,
-      "learning_rate": 0.00019044581426256804,
-      "loss": 0.9536,
+      "epoch": 6.771084337349397,
+      "grad_norm": 0.21875,
+      "learning_rate": 5.6557372838031384e-05,
+      "loss": 0.8459,
       "step": 1405
     },
     {
-      "epoch": 3.401688781664656,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00019032557124590974,
-      "loss": 0.9492,
+      "epoch": 6.795180722891566,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 5.579949076509305e-05,
+      "loss": 0.8455,
       "step": 1410
     },
     {
-      "epoch": 3.413751507840772,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00019020461474898503,
-      "loss": 0.945,
+      "epoch": 6.8192771084337345,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 5.5044750927752106e-05,
+      "loss": 0.8444,
       "step": 1415
     },
     {
-      "epoch": 3.425814234016888,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00019008294572722738,
-      "loss": 0.955,
+      "epoch": 6.843373493975903,
+      "grad_norm": 0.251953125,
+      "learning_rate": 5.429320698082887e-05,
+      "loss": 0.8427,
       "step": 1420
     },
     {
-      "epoch": 3.4378769601930035,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00018996056514169844,
-      "loss": 0.9466,
+      "epoch": 6.867469879518072,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 5.354491235194635e-05,
+      "loss": 0.8466,
       "step": 1425
     },
     {
-      "epoch": 3.4499396863691194,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00018983747395908058,
-      "loss": 0.9436,
+      "epoch": 6.891566265060241,
+      "grad_norm": 0.228515625,
+      "learning_rate": 5.279992023773195e-05,
+      "loss": 0.8419,
       "step": 1430
     },
     {
-      "epoch": 3.4620024125452353,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.0001897136731516691,
-      "loss": 0.949,
+      "epoch": 6.9156626506024095,
+      "grad_norm": 0.232421875,
+      "learning_rate": 5.205828360003568e-05,
+      "loss": 0.8468,
       "step": 1435
     },
     {
-      "epoch": 3.4740651387213513,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.0001895891636973646,
-      "loss": 0.9629,
+      "epoch": 6.9397590361445785,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 5.1320055162165115e-05,
+      "loss": 0.8435,
       "step": 1440
     },
     {
-      "epoch": 3.4861278648974667,
-      "grad_norm": 0.359375,
-      "learning_rate": 0.00018946394657966536,
-      "loss": 0.9439,
+      "epoch": 6.9638554216867465,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 5.0585287405137305e-05,
+      "loss": 0.8444,
       "step": 1445
     },
     {
-      "epoch": 3.4981905910735827,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00018933802278765934,
-      "loss": 0.9455,
+      "epoch": 6.9879518072289155,
+      "grad_norm": 0.228515625,
+      "learning_rate": 4.9854032563947714e-05,
+      "loss": 0.8464,
       "step": 1450
     },
     {
-      "epoch": 3.5102533172496986,
-      "grad_norm": 0.365234375,
-      "learning_rate": 0.00018921139331601667,
-      "loss": 0.9462,
+      "epoch": 6.997590361445783,
+      "eval_loss": 2.5512688159942627,
+      "eval_runtime": 0.4955,
+      "eval_samples_per_second": 20.183,
+      "eval_steps_per_second": 2.018,
+      "step": 1452
+    },
+    {
+      "epoch": 7.0120481927710845,
+      "grad_norm": 0.20703125,
+      "learning_rate": 4.912634262385695e-05,
+      "loss": 0.8349,
       "step": 1455
     },
     {
-      "epoch": 3.522316043425814,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00018908405916498154,
-      "loss": 0.9499,
+      "epoch": 7.036144578313253,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 4.8402269316695134e-05,
+      "loss": 0.8372,
       "step": 1460
     },
     {
-      "epoch": 3.53437876960193,
-      "grad_norm": 0.484375,
-      "learning_rate": 0.00018895602134036434,
-      "loss": 0.9433,
+      "epoch": 7.0602409638554215,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 4.768186411718417e-05,
+      "loss": 0.8283,
       "step": 1465
     },
     {
-      "epoch": 3.546441495778046,
-      "grad_norm": 0.5078125,
-      "learning_rate": 0.00018882728085353392,
-      "loss": 0.9431,
+      "epoch": 7.0843373493975905,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 4.696517823927842e-05,
+      "loss": 0.842,
       "step": 1470
     },
     {
-      "epoch": 3.558504221954162,
-      "grad_norm": 0.45703125,
-      "learning_rate": 0.0001886978387214093,
-      "loss": 0.9494,
+      "epoch": 7.108433734939759,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 4.625226263252386e-05,
+      "loss": 0.8375,
       "step": 1475
     },
     {
-      "epoch": 3.5705669481302773,
-      "grad_norm": 0.484375,
-      "learning_rate": 0.0001885676959664519,
-      "loss": 0.9515,
+      "epoch": 7.132530120481928,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 4.554316797843609e-05,
+      "loss": 0.8257,
       "step": 1480
     },
     {
-      "epoch": 3.5826296743063932,
-      "grad_norm": 0.5546875,
-      "learning_rate": 0.00018843685361665723,
-      "loss": 0.9413,
+      "epoch": 7.156626506024097,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 4.483794468689728e-05,
+      "loss": 0.8347,
       "step": 1485
     },
     {
-      "epoch": 3.594692400482509,
-      "grad_norm": 0.6171875,
-      "learning_rate": 0.000188305312705547,
-      "loss": 0.9398,
+      "epoch": 7.180722891566265,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 4.413664289257265e-05,
+      "loss": 0.8354,
       "step": 1490
     },
     {
-      "epoch": 3.6067551266586246,
-      "grad_norm": 0.380859375,
-      "learning_rate": 0.0001881730742721608,
-      "loss": 0.9314,
+      "epoch": 7.204819277108434,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 4.343931245134616e-05,
+      "loss": 0.8382,
       "step": 1495
     },
     {
-      "epoch": 3.6188178528347406,
-      "grad_norm": 0.53515625,
-      "learning_rate": 0.00018804013936104792,
-      "loss": 0.9364,
+      "epoch": 7.228915662650603,
+      "grad_norm": 0.21875,
+      "learning_rate": 4.274600293677647e-05,
+      "loss": 0.8359,
       "step": 1500
     },
     {
-      "epoch": 3.6308805790108565,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.0001879065090222591,
-      "loss": 0.9502,
+      "epoch": 7.253012048192771,
+      "grad_norm": 0.2109375,
+      "learning_rate": 4.2056763636572574e-05,
+      "loss": 0.8351,
       "step": 1505
     },
     {
-      "epoch": 3.6429433051869724,
-      "grad_norm": 0.37890625,
-      "learning_rate": 0.00018777218431133833,
-      "loss": 0.9477,
+      "epoch": 7.27710843373494,
+      "grad_norm": 0.2265625,
+      "learning_rate": 4.137164354908999e-05,
+      "loss": 0.8377,
       "step": 1510
     },
     {
-      "epoch": 3.6550060313630883,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00018763716628931437,
-      "loss": 0.9462,
+      "epoch": 7.301204819277109,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 4.069069137984731e-05,
+      "loss": 0.8313,
       "step": 1515
     },
     {
-      "epoch": 3.667068757539204,
-      "grad_norm": 0.431640625,
-      "learning_rate": 0.00018750145602269247,
-      "loss": 0.9442,
+      "epoch": 7.325301204819277,
+      "grad_norm": 0.224609375,
+      "learning_rate": 4.001395553806391e-05,
+      "loss": 0.8382,
       "step": 1520
     },
     {
-      "epoch": 3.6791314837153197,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.00018736505458344586,
-      "loss": 0.9378,
+      "epoch": 7.349397590361446,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 3.9341484133218366e-05,
+      "loss": 0.8368,
       "step": 1525
     },
     {
-      "epoch": 3.691194209891435,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.0001872279630490074,
-      "loss": 0.9423,
+      "epoch": 7.373493975903615,
+      "grad_norm": 0.21484375,
+      "learning_rate": 3.8673324971628357e-05,
+      "loss": 0.8419,
       "step": 1530
     },
     {
-      "epoch": 3.703256936067551,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.00018709018250226094,
-      "loss": 0.9389,
+      "epoch": 7.397590361445783,
+      "grad_norm": 0.212890625,
+      "learning_rate": 3.800952555305216e-05,
+      "loss": 0.8357,
       "step": 1535
     },
     {
-      "epoch": 3.715319662243667,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00018695171403153285,
-      "loss": 0.9356,
+      "epoch": 7.421686746987952,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 3.7350133067311686e-05,
+      "loss": 0.8393,
       "step": 1540
     },
     {
-      "epoch": 3.727382388419783,
-      "grad_norm": 0.52734375,
-      "learning_rate": 0.00018681255873058338,
-      "loss": 0.9447,
+      "epoch": 7.445783132530121,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 3.669519439093801e-05,
+      "loss": 0.8319,
       "step": 1545
     },
     {
-      "epoch": 3.739445114595899,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00018667271769859812,
-      "loss": 0.9409,
+      "epoch": 7.469879518072289,
+      "grad_norm": 0.23828125,
+      "learning_rate": 3.604475608383858e-05,
+      "loss": 0.8358,
       "step": 1550
     },
     {
-      "epoch": 3.7515078407720144,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00018653219204017916,
-      "loss": 0.9457,
+      "epoch": 7.493975903614458,
+      "grad_norm": 0.212890625,
+      "learning_rate": 3.539886438598756e-05,
+      "loss": 0.8369,
       "step": 1555
     },
     {
-      "epoch": 3.7635705669481303,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.00018639098286533644,
-      "loss": 0.954,
+      "epoch": 7.518072289156627,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 3.475756521413839e-05,
+      "loss": 0.8355,
       "step": 1560
     },
     {
-      "epoch": 3.7756332931242462,
-      "grad_norm": 0.375,
-      "learning_rate": 0.00018624909128947903,
-      "loss": 0.9467,
+      "epoch": 7.542168674698795,
+      "grad_norm": 0.212890625,
+      "learning_rate": 3.412090415855963e-05,
+      "loss": 0.8374,
       "step": 1565
     },
     {
-      "epoch": 3.7876960193003617,
-      "grad_norm": 0.55078125,
-      "learning_rate": 0.00018610651843340627,
-      "loss": 0.9436,
+      "epoch": 7.566265060240964,
+      "grad_norm": 0.2265625,
+      "learning_rate": 3.348892647979389e-05,
+      "loss": 0.8353,
       "step": 1570
     },
     {
-      "epoch": 3.7997587454764776,
-      "grad_norm": 0.392578125,
-      "learning_rate": 0.00018596326542329888,
-      "loss": 0.9381,
+      "epoch": 7.590361445783133,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 3.2861677105440336e-05,
+      "loss": 0.8368,
       "step": 1575
     },
     {
-      "epoch": 3.8118214716525936,
-      "grad_norm": 0.41015625,
-      "learning_rate": 0.0001858193333907101,
-      "loss": 0.9408,
+      "epoch": 7.614457831325301,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 3.223920062696052e-05,
+      "loss": 0.8315,
       "step": 1580
     },
     {
-      "epoch": 3.8238841978287095,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.0001856747234725569,
-      "loss": 0.9456,
+      "epoch": 7.63855421686747,
+      "grad_norm": 0.212890625,
+      "learning_rate": 3.1621541296508695e-05,
+      "loss": 0.8373,
       "step": 1585
     },
     {
-      "epoch": 3.835946924004825,
-      "grad_norm": 0.357421875,
-      "learning_rate": 0.00018552943681111067,
-      "loss": 0.9405,
+      "epoch": 7.662650602409639,
+      "grad_norm": 0.21484375,
+      "learning_rate": 3.100874302378559e-05,
+      "loss": 0.8315,
       "step": 1590
     },
     {
-      "epoch": 3.848009650180941,
-      "grad_norm": 0.384765625,
-      "learning_rate": 0.00018538347455398846,
-      "loss": 0.9551,
+      "epoch": 7.686746987951807,
+      "grad_norm": 0.208984375,
+      "learning_rate": 3.0400849372917073e-05,
+      "loss": 0.839,
       "step": 1595
     },
     {
-      "epoch": 3.860072376357057,
-      "grad_norm": 0.375,
-      "learning_rate": 0.00018523683785414387,
-      "loss": 0.9401,
+      "epoch": 7.710843373493976,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 2.979790355935703e-05,
+      "loss": 0.8368,
       "step": 1600
     },
     {
-      "epoch": 3.8721351025331723,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.0001850895278698579,
-      "loss": 0.9374,
+      "epoch": 7.734939759036145,
+      "grad_norm": 0.205078125,
+      "learning_rate": 2.919994844681524e-05,
+      "loss": 0.8348,
       "step": 1605
     },
     {
-      "epoch": 3.884197828709288,
-      "grad_norm": 0.494140625,
-      "learning_rate": 0.00018494154576472976,
-      "loss": 0.937,
+      "epoch": 7.759036144578313,
+      "grad_norm": 0.205078125,
+      "learning_rate": 2.8607026544210114e-05,
+      "loss": 0.8323,
       "step": 1610
     },
     {
-      "epoch": 3.896260554885404,
-      "grad_norm": 0.357421875,
-      "learning_rate": 0.00018479289270766777,
-      "loss": 0.9398,
+      "epoch": 7.783132530120482,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.801918000264665e-05,
+      "loss": 0.8344,
       "step": 1615
     },
     {
-      "epoch": 3.90832328106152,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.00018464356987288013,
-      "loss": 0.9402,
+      "epoch": 7.807228915662651,
+      "grad_norm": 0.22265625,
+      "learning_rate": 2.7436450612420095e-05,
+      "loss": 0.8343,
       "step": 1620
     },
     {
-      "epoch": 3.920386007237636,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.0001844935784398655,
-      "loss": 0.9443,
+      "epoch": 7.831325301204819,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.6858879800044866e-05,
+      "loss": 0.8329,
       "step": 1625
     },
     {
-      "epoch": 3.9324487334137515,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00018434291959340394,
-      "loss": 0.9441,
+      "epoch": 7.855421686746988,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 2.6286508625309624e-05,
+      "loss": 0.8354,
       "step": 1630
     },
     {
-      "epoch": 3.9445114595898674,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.0001841915945235472,
-      "loss": 0.939,
+      "epoch": 7.879518072289157,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.57193777783582e-05,
+      "loss": 0.8399,
       "step": 1635
     },
     {
-      "epoch": 3.956574185765983,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00018403960442560973,
-      "loss": 0.94,
+      "epoch": 7.903614457831325,
+      "grad_norm": 0.2109375,
+      "learning_rate": 2.515752757679707e-05,
+      "loss": 0.8325,
       "step": 1640
     },
     {
-      "epoch": 3.9686369119420988,
-      "grad_norm": 0.3515625,
-      "learning_rate": 0.00018388695050015885,
-      "loss": 0.939,
+      "epoch": 7.927710843373494,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.4600997962828987e-05,
+      "loss": 0.8338,
       "step": 1645
     },
     {
-      "epoch": 3.9806996381182147,
-      "grad_norm": 0.408203125,
-      "learning_rate": 0.00018373363395300554,
-      "loss": 0.9399,
+      "epoch": 7.951807228915663,
+      "grad_norm": 0.2109375,
+      "learning_rate": 2.404982850041363e-05,
+      "loss": 0.8371,
       "step": 1650
     },
     {
-      "epoch": 3.9927623642943306,
-      "grad_norm": 0.455078125,
-      "learning_rate": 0.0001835796559951948,
-      "loss": 0.9412,
+      "epoch": 7.975903614457831,
+      "grad_norm": 0.205078125,
+      "learning_rate": 2.3504058372454884e-05,
+      "loss": 0.8319,
       "step": 1655
     },
     {
-      "epoch": 4.0,
-      "eval_loss": 2.449599504470825,
-      "eval_runtime": 0.535,
-      "eval_samples_per_second": 18.691,
-      "eval_steps_per_second": 1.869,
-      "step": 1658
+      "epoch": 8.0,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.2963726378015327e-05,
+      "loss": 0.8353,
+      "step": 1660
     },
     {
-      "epoch": 4.0048250904704465,
-      "grad_norm": 0.353515625,
-      "learning_rate": 0.0001834250178429961,
-      "loss": 0.9353,
+      "epoch": 8.0,
+      "eval_loss": 2.5614659786224365,
+      "eval_runtime": 0.4882,
+      "eval_samples_per_second": 20.483,
+      "eval_steps_per_second": 2.048,
       "step": 1660
     },
     {
-      "epoch": 4.0168878166465625,
-      "grad_norm": 0.390625,
-      "learning_rate": 0.0001832697207178938,
-      "loss": 0.9234,
+      "epoch": 8.024096385542169,
+      "grad_norm": 0.228515625,
+      "learning_rate": 2.242887092955801e-05,
+      "loss": 0.8313,
       "step": 1665
     },
     {
-      "epoch": 4.0289505428226775,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00018311376584657736,
-      "loss": 0.9264,
+      "epoch": 8.048192771084338,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 2.189953005021569e-05,
+      "loss": 0.8278,
       "step": 1670
     },
     {
-      "epoch": 4.041013268998793,
-      "grad_norm": 0.4375,
-      "learning_rate": 0.00018295715446093192,
-      "loss": 0.9233,
+      "epoch": 8.072289156626505,
+      "grad_norm": 0.22265625,
+      "learning_rate": 2.1375741371087677e-05,
+      "loss": 0.8321,
       "step": 1675
     },
     {
-      "epoch": 4.053075995174909,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00018279988779802833,
-      "loss": 0.9299,
+      "epoch": 8.096385542168674,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.085754212856471e-05,
+      "loss": 0.8368,
       "step": 1680
     },
     {
-      "epoch": 4.065138721351025,
-      "grad_norm": 0.416015625,
-      "learning_rate": 0.00018264196710011352,
-      "loss": 0.9359,
+      "epoch": 8.120481927710843,
+      "grad_norm": 0.21875,
+      "learning_rate": 2.0344969161681792e-05,
+      "loss": 0.8258,
       "step": 1685
     },
     {
-      "epoch": 4.077201447527141,
-      "grad_norm": 0.37109375,
-      "learning_rate": 0.00018248339361460057,
-      "loss": 0.9266,
+      "epoch": 8.144578313253012,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.983805890949927e-05,
+      "loss": 0.8273,
       "step": 1690
     },
     {
-      "epoch": 4.089264173703257,
-      "grad_norm": 0.36328125,
-      "learning_rate": 0.00018232416859405895,
-      "loss": 0.935,
+      "epoch": 8.168674698795181,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.9336847408512328e-05,
+      "loss": 0.8272,
       "step": 1695
     },
     {
-      "epoch": 4.101326899879373,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.00018216429329620456,
-      "loss": 0.9294,
+      "epoch": 8.19277108433735,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.884137029008921e-05,
+      "loss": 0.8334,
       "step": 1700
     },
     {
-      "epoch": 4.113389626055489,
-      "grad_norm": 0.359375,
-      "learning_rate": 0.00018200376898388996,
-      "loss": 0.9288,
+      "epoch": 8.216867469879517,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.8351662777938127e-05,
+      "loss": 0.8337,
       "step": 1705
     },
     {
-      "epoch": 4.125452352231604,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00018184259692509406,
-      "loss": 0.9264,
+      "epoch": 8.240963855421686,
+      "grad_norm": 0.2109375,
+      "learning_rate": 1.7867759685603114e-05,
+      "loss": 0.8306,
       "step": 1710
     },
     {
-      "epoch": 4.13751507840772,
-      "grad_norm": 0.490234375,
-      "learning_rate": 0.00018168077839291247,
-      "loss": 0.9421,
+      "epoch": 8.265060240963855,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.738969541398926e-05,
+      "loss": 0.8306,
       "step": 1715
     },
     {
-      "epoch": 4.149577804583836,
-      "grad_norm": 0.67578125,
-      "learning_rate": 0.00018151831466554726,
-      "loss": 0.9355,
+      "epoch": 8.289156626506024,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.691750394891707e-05,
+      "loss": 0.8299,
       "step": 1720
     },
     {
-      "epoch": 4.161640530759952,
-      "grad_norm": 0.365234375,
-      "learning_rate": 0.00018135520702629675,
-      "loss": 0.9349,
+      "epoch": 8.313253012048193,
+      "grad_norm": 0.205078125,
+      "learning_rate": 1.6451218858706374e-05,
+      "loss": 0.829,
       "step": 1725
     },
     {
-      "epoch": 4.173703256936068,
-      "grad_norm": 0.458984375,
-      "learning_rate": 0.00018119145676354575,
-      "loss": 0.9345,
+      "epoch": 8.337349397590362,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 1.59908732917899e-05,
+      "loss": 0.831,
       "step": 1730
     },
     {
-      "epoch": 4.185765983112184,
-      "grad_norm": 0.5625,
-      "learning_rate": 0.00018102706517075491,
-      "loss": 0.9319,
+      "epoch": 8.36144578313253,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.5536499974356866e-05,
+      "loss": 0.8269,
       "step": 1735
     },
     {
-      "epoch": 4.1978287092882995,
-      "grad_norm": 0.404296875,
-      "learning_rate": 0.00018086203354645089,
-      "loss": 0.9309,
+      "epoch": 8.385542168674698,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.5088131208026367e-05,
+      "loss": 0.834,
       "step": 1740
     },
     {
-      "epoch": 4.209891435464415,
-      "grad_norm": 0.369140625,
-      "learning_rate": 0.00018069636319421588,
-      "loss": 0.9378,
+      "epoch": 8.409638554216867,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 1.4645798867551008e-05,
+      "loss": 0.8324,
       "step": 1745
     },
     {
-      "epoch": 4.2219541616405305,
-      "grad_norm": 0.380859375,
-      "learning_rate": 0.00018053005542267736,
-      "loss": 0.9285,
+      "epoch": 8.433734939759036,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 1.4209534398551016e-05,
+      "loss": 0.8325,
       "step": 1750
     },
     {
-      "epoch": 4.234016887816646,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00018036311154549784,
-      "loss": 0.9314,
+      "epoch": 8.457831325301205,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 1.3779368815278647e-05,
+      "loss": 0.8392,
       "step": 1755
     },
     {
-      "epoch": 4.246079613992762,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00018019553288136435,
-      "loss": 0.9234,
+      "epoch": 8.481927710843374,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.335533269841347e-05,
+      "loss": 0.8284,
       "step": 1760
     },
     {
-      "epoch": 4.258142340168878,
-      "grad_norm": 0.41015625,
-      "learning_rate": 0.00018002732075397812,
-      "loss": 0.9286,
+      "epoch": 8.506024096385541,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.2937456192888309e-05,
+      "loss": 0.8332,
       "step": 1765
     },
     {
-      "epoch": 4.270205066344994,
-      "grad_norm": 0.390625,
-      "learning_rate": 0.00017985847649204417,
-      "loss": 0.9267,
+      "epoch": 8.53012048192771,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.252576900574618e-05,
+      "loss": 0.8269,
       "step": 1770
     },
     {
-      "epoch": 4.28226779252111,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00017968900142926057,
-      "loss": 0.9275,
+      "epoch": 8.55421686746988,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.2120300404028507e-05,
+      "loss": 0.8346,
       "step": 1775
     },
     {
-      "epoch": 4.294330518697226,
-      "grad_norm": 0.37890625,
-      "learning_rate": 0.00017951889690430824,
-      "loss": 0.9312,
+      "epoch": 8.578313253012048,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.1721079212694452e-05,
+      "loss": 0.8331,
       "step": 1780
     },
     {
-      "epoch": 4.306393244873341,
-      "grad_norm": 0.3984375,
-      "learning_rate": 0.00017934816426084008,
-      "loss": 0.9242,
+      "epoch": 8.602409638554217,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.1328133812571784e-05,
+      "loss": 0.8324,
       "step": 1785
     },
     {
-      "epoch": 4.318455971049457,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00017917680484747065,
-      "loss": 0.9297,
+      "epoch": 8.626506024096386,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 1.0941492138339183e-05,
+      "loss": 0.8285,
       "step": 1790
     },
     {
-      "epoch": 4.330518697225573,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.00017900482001776517,
-      "loss": 0.9206,
+      "epoch": 8.650602409638553,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 1.0561181676540444e-05,
+      "loss": 0.838,
       "step": 1795
     },
     {
-      "epoch": 4.342581423401689,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00017883221113022916,
-      "loss": 0.927,
+      "epoch": 8.674698795180722,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.01872294636304e-05,
+      "loss": 0.827,
       "step": 1800
     },
     {
-      "epoch": 4.354644149577805,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00017865897954829748,
-      "loss": 0.9324,
+      "epoch": 8.698795180722891,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.81966208405285e-06,
+      "loss": 0.8331,
       "step": 1805
     },
     {
-      "epoch": 4.366706875753921,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.0001784851266403237,
-      "loss": 0.9422,
+      "epoch": 8.72289156626506,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.458505668350759e-06,
+      "loss": 0.8304,
       "step": 1810
     },
     {
-      "epoch": 4.378769601930037,
-      "grad_norm": 0.353515625,
-      "learning_rate": 0.0001783106537795692,
-      "loss": 0.9326,
+      "epoch": 8.74698795180723,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.103785891308547e-06,
+      "loss": 0.8333,
       "step": 1815
     },
     {
-      "epoch": 4.390832328106152,
-      "grad_norm": 0.470703125,
-      "learning_rate": 0.00017813556234419234,
-      "loss": 0.9366,
+      "epoch": 8.771084337349398,
+      "grad_norm": 0.205078125,
+      "learning_rate": 8.755527970126853e-06,
+      "loss": 0.8325,
       "step": 1820
     },
     {
-      "epoch": 4.402895054282268,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00017795985371723763,
-      "loss": 0.9328,
+      "epoch": 8.795180722891565,
+      "grad_norm": 0.21484375,
+      "learning_rate": 8.413756662629879e-06,
+      "loss": 0.8326,
       "step": 1825
     },
     {
-      "epoch": 4.4149577804583835,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00017778352928662474,
-      "loss": 0.9288,
+      "epoch": 8.819277108433734,
+      "grad_norm": 0.208984375,
+      "learning_rate": 8.07849626550531e-06,
+      "loss": 0.8315,
       "step": 1830
     },
     {
-      "epoch": 4.427020506634499,
-      "grad_norm": 0.357421875,
-      "learning_rate": 0.00017760659044513757,
-      "loss": 0.9311,
+      "epoch": 8.843373493975903,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 7.74977061257709e-06,
+      "loss": 0.833,
       "step": 1835
     },
     {
-      "epoch": 4.439083232810615,
-      "grad_norm": 0.36328125,
-      "learning_rate": 0.00017742903859041325,
-      "loss": 0.9263,
+      "epoch": 8.867469879518072,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 7.427603073110967e-06,
+      "loss": 0.8299,
       "step": 1840
     },
     {
-      "epoch": 4.451145958986731,
-      "grad_norm": 0.390625,
-      "learning_rate": 0.0001772508751249311,
-      "loss": 0.9248,
+      "epoch": 8.891566265060241,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 7.1120165501533e-06,
+      "loss": 0.8282,
       "step": 1845
     },
     {
-      "epoch": 4.463208685162847,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.0001770721014560015,
-      "loss": 0.9322,
+      "epoch": 8.91566265060241,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 6.803033478902765e-06,
+      "loss": 0.8267,
       "step": 1850
     },
     {
-      "epoch": 4.475271411338962,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.00017689271899575491,
-      "loss": 0.9284,
+      "epoch": 8.939759036144578,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 6.500675825115454e-06,
+      "loss": 0.83,
       "step": 1855
     },
     {
-      "epoch": 4.487334137515078,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.00017671272916113052,
-      "loss": 0.9323,
+      "epoch": 8.963855421686747,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 6.204965083543368e-06,
+      "loss": 0.8267,
       "step": 1860
     },
     {
-      "epoch": 4.499396863691194,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00017653213337386517,
-      "loss": 0.9302,
+      "epoch": 8.987951807228916,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 5.915922276406249e-06,
+      "loss": 0.8267,
       "step": 1865
     },
     {
-      "epoch": 4.51145958986731,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.0001763509330604822,
-      "loss": 0.9298,
+      "epoch": 8.997590361445782,
+      "eval_loss": 2.5673654079437256,
+      "eval_runtime": 0.7609,
+      "eval_samples_per_second": 13.142,
+      "eval_steps_per_second": 1.314,
+      "step": 1867
+    },
+    {
+      "epoch": 9.012048192771084,
+      "grad_norm": 0.197265625,
+      "learning_rate": 5.633567951897145e-06,
+      "loss": 0.8245,
       "step": 1870
     },
     {
-      "epoch": 4.523522316043426,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.00017616912965228001,
-      "loss": 0.9278,
+      "epoch": 9.036144578313253,
+      "grad_norm": 0.205078125,
+      "learning_rate": 5.357922182721687e-06,
+      "loss": 0.8271,
       "step": 1875
     },
     {
-      "epoch": 4.535585042219542,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.00017598672458532088,
-      "loss": 0.9329,
+      "epoch": 9.060240963855422,
+      "grad_norm": 0.2109375,
+      "learning_rate": 5.08900456467103e-06,
+      "loss": 0.8373,
       "step": 1880
     },
     {
-      "epoch": 4.547647768395658,
-      "grad_norm": 0.400390625,
-      "learning_rate": 0.00017580371930041953,
-      "loss": 0.9251,
+      "epoch": 9.08433734939759,
+      "grad_norm": 0.20703125,
+      "learning_rate": 4.826834215228826e-06,
+      "loss": 0.8242,
       "step": 1885
     },
     {
-      "epoch": 4.559710494571773,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.00017562011524313185,
-      "loss": 0.9278,
+      "epoch": 9.108433734939759,
+      "grad_norm": 0.212890625,
+      "learning_rate": 4.5714297722121106e-06,
+      "loss": 0.8283,
       "step": 1890
     },
     {
-      "epoch": 4.571773220747889,
-      "grad_norm": 0.392578125,
-      "learning_rate": 0.0001754359138637434,
-      "loss": 0.9164,
+      "epoch": 9.132530120481928,
+      "grad_norm": 0.19921875,
+      "learning_rate": 4.322809392446392e-06,
+      "loss": 0.8313,
       "step": 1895
     },
     {
-      "epoch": 4.583835946924005,
-      "grad_norm": 0.349609375,
-      "learning_rate": 0.00017525111661725797,
-      "loss": 0.9184,
+      "epoch": 9.156626506024097,
+      "grad_norm": 0.208984375,
+      "learning_rate": 4.080990750474778e-06,
+      "loss": 0.8287,
       "step": 1900
     },
     {
-      "epoch": 4.595898673100121,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.0001750657249633861,
-      "loss": 0.9345,
+      "epoch": 9.180722891566266,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 3.84599103730161e-06,
+      "loss": 0.8265,
       "step": 1905
     },
     {
-      "epoch": 4.6079613992762365,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.00017487974036653361,
-      "loss": 0.9285,
+      "epoch": 9.204819277108435,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 3.617826959170256e-06,
+      "loss": 0.8307,
       "step": 1910
     },
     {
-      "epoch": 4.620024125452352,
-      "grad_norm": 0.52734375,
-      "learning_rate": 0.00017469316429578977,
-      "loss": 0.9219,
+      "epoch": 9.228915662650602,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 3.3965147363754555e-06,
+      "loss": 0.8343,
       "step": 1915
     },
     {
-      "epoch": 4.632086851628468,
-      "grad_norm": 0.365234375,
-      "learning_rate": 0.00017450599822491615,
-      "loss": 0.9192,
+      "epoch": 9.25301204819277,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 3.182070102110257e-06,
+      "loss": 0.8313,
       "step": 1920
     },
     {
-      "epoch": 4.644149577804583,
-      "grad_norm": 0.447265625,
-      "learning_rate": 0.00017431824363233457,
-      "loss": 0.9301,
+      "epoch": 9.27710843373494,
+      "grad_norm": 0.203125,
+      "learning_rate": 2.974508301347534e-06,
+      "loss": 0.8319,
       "step": 1925
     },
     {
-      "epoch": 4.656212303980699,
-      "grad_norm": 0.427734375,
-      "learning_rate": 0.00017412990200111556,
-      "loss": 0.9293,
+      "epoch": 9.301204819277109,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 2.7738440897561723e-06,
+      "loss": 0.8307,
       "step": 1930
     },
     {
-      "epoch": 4.668275030156815,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.00017394097481896676,
-      "loss": 0.9215,
+      "epoch": 9.325301204819278,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 2.580091732652101e-06,
+      "loss": 0.8316,
       "step": 1935
     },
     {
-      "epoch": 4.680337756332931,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.00017375146357822096,
-      "loss": 0.927,
+      "epoch": 9.349397590361447,
+      "grad_norm": 0.205078125,
+      "learning_rate": 2.3932650039841687e-06,
+      "loss": 0.8309,
       "step": 1940
     },
     {
-      "epoch": 4.692400482509047,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.00017356136977582454,
-      "loss": 0.9251,
+      "epoch": 9.373493975903614,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.213377185354959e-06,
+      "loss": 0.8335,
       "step": 1945
     },
     {
-      "epoch": 4.704463208685163,
-      "grad_norm": 0.412109375,
-      "learning_rate": 0.00017337069491332537,
-      "loss": 0.9291,
+      "epoch": 9.397590361445783,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 2.0404410650765817e-06,
+      "loss": 0.8307,
       "step": 1950
     },
     {
-      "epoch": 4.716525934861279,
-      "grad_norm": 0.3984375,
-      "learning_rate": 0.00017317944049686124,
-      "loss": 0.9288,
+      "epoch": 9.421686746987952,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.874468937261531e-06,
+      "loss": 0.8301,
       "step": 1955
     },
     {
-      "epoch": 4.728588661037395,
-      "grad_norm": 0.408203125,
-      "learning_rate": 0.00017298760803714775,
-      "loss": 0.9165,
+      "epoch": 9.44578313253012,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.71547260094872e-06,
+      "loss": 0.8262,
       "step": 1960
     },
     {
-      "epoch": 4.74065138721351,
-      "grad_norm": 0.55078125,
-      "learning_rate": 0.00017279519904946647,
-      "loss": 0.9309,
+      "epoch": 9.46987951807229,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 1.5634633592646609e-06,
+      "loss": 0.8284,
       "step": 1965
     },
     {
-      "epoch": 4.752714113389626,
-      "grad_norm": 0.447265625,
-      "learning_rate": 0.00017260221505365303,
-      "loss": 0.919,
+      "epoch": 9.493975903614459,
+      "grad_norm": 0.205078125,
+      "learning_rate": 1.4184520186199202e-06,
+      "loss": 0.8309,
       "step": 1970
     },
     {
-      "epoch": 4.764776839565742,
-      "grad_norm": 0.37109375,
-      "learning_rate": 0.00017240865757408495,
-      "loss": 0.9239,
+      "epoch": 9.518072289156626,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 1.2804488879408993e-06,
+      "loss": 0.8288,
       "step": 1975
     },
     {
-      "epoch": 4.776839565741858,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.0001722145281396697,
-      "loss": 0.9296,
+      "epoch": 9.542168674698795,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.1494637779369766e-06,
+      "loss": 0.8309,
       "step": 1980
     },
     {
-      "epoch": 4.788902291917974,
-      "grad_norm": 0.373046875,
-      "learning_rate": 0.00017201982828383264,
-      "loss": 0.9306,
-      "step": 1985
-    },
-    {
-      "epoch": 4.8009650180940895,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00017182455954450486,
-      "loss": 0.9283,
-      "step": 1990
-    },
-    {
-      "epoch": 4.813027744270205,
-      "grad_norm": 0.4140625,
-      "learning_rate": 0.00017162872346411102,
-      "loss": 0.9237,
-      "step": 1995
-    },
-    {
-      "epoch": 4.825090470446321,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00017143232158955732,
-      "loss": 0.9293,
-      "step": 2000
-    },
-    {
-      "epoch": 4.837153196622436,
-      "grad_norm": 0.40234375,
-      "learning_rate": 0.00017123535547221907,
-      "loss": 0.9295,
-      "step": 2005
-    },
-    {
-      "epoch": 4.849215922798552,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.00017103782666792844,
-      "loss": 0.9154,
-      "step": 2010
-    },
-    {
-      "epoch": 4.861278648974668,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00017083973673696247,
-      "loss": 0.9207,
-      "step": 2015
-    },
-    {
-      "epoch": 4.873341375150784,
-      "grad_norm": 0.359375,
-      "learning_rate": 0.00017064108724403033,
-      "loss": 0.9272,
-      "step": 2020
-    },
-    {
-      "epoch": 4.8854041013269,
-      "grad_norm": 0.35546875,
-      "learning_rate": 0.00017044187975826124,
-      "loss": 0.9224,
-      "step": 2025
-    },
-    {
-      "epoch": 4.897466827503016,
-      "grad_norm": 0.408203125,
-      "learning_rate": 0.00017024211585319203,
-      "loss": 0.9241,
-      "step": 2030
-    },
-    {
-      "epoch": 4.909529553679132,
-      "grad_norm": 0.357421875,
-      "learning_rate": 0.00017004179710675463,
-      "loss": 0.933,
-      "step": 2035
-    },
-    {
-      "epoch": 4.921592279855247,
-      "grad_norm": 0.294921875,
-      "learning_rate": 0.00016984092510126367,
-      "loss": 0.916,
-      "step": 2040
-    },
-    {
-      "epoch": 4.933655006031363,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00016963950142340396,
-      "loss": 0.9278,
-      "step": 2045
-    },
-    {
-      "epoch": 4.945717732207479,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00016943752766421794,
-      "loss": 0.9284,
-      "step": 2050
-    },
-    {
-      "epoch": 4.957780458383595,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.0001692350054190932,
-      "loss": 0.9358,
-      "step": 2055
-    },
-    {
-      "epoch": 4.969843184559711,
-      "grad_norm": 0.515625,
-      "learning_rate": 0.00016903193628774977,
-      "loss": 0.925,
-      "step": 2060
-    },
-    {
-      "epoch": 4.981905910735827,
-      "grad_norm": 0.359375,
-      "learning_rate": 0.00016882832187422743,
-      "loss": 0.9235,
-      "step": 2065
-    },
-    {
-      "epoch": 4.9939686369119425,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.0001686241637868734,
-      "loss": 0.9325,
-      "step": 2070
-    },
-    {
-      "epoch": 4.998793727382388,
-      "eval_loss": 2.4599645137786865,
-      "eval_runtime": 0.6669,
-      "eval_samples_per_second": 14.995,
-      "eval_steps_per_second": 1.499,
-      "step": 2072
-    },
-    {
-      "epoch": 5.0060313630880575,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.000168419463638329,
-      "loss": 0.9096,
-      "step": 2075
-    },
-    {
-      "epoch": 5.018094089264173,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00016821422304551766,
-      "loss": 0.9202,
-      "step": 2080
-    },
-    {
-      "epoch": 5.030156815440289,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00016800844362963147,
-      "loss": 0.9134,
-      "step": 2085
-    },
-    {
-      "epoch": 5.042219541616405,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00016780212701611886,
-      "loss": 0.909,
-      "step": 2090
-    },
-    {
-      "epoch": 5.054282267792521,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.0001675952748346715,
-      "loss": 0.9011,
-      "step": 2095
-    },
-    {
-      "epoch": 5.066344993968637,
-      "grad_norm": 0.36328125,
-      "learning_rate": 0.00016738788871921152,
-      "loss": 0.9104,
-      "step": 2100
-    },
-    {
-      "epoch": 5.078407720144753,
-      "grad_norm": 0.35546875,
-      "learning_rate": 0.0001671799703078786,
-      "loss": 0.9222,
-      "step": 2105
-    },
-    {
-      "epoch": 5.090470446320868,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00016697152124301695,
-      "loss": 0.9157,
-      "step": 2110
-    },
-    {
-      "epoch": 5.102533172496984,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.00016676254317116252,
-      "loss": 0.922,
-      "step": 2115
-    },
-    {
-      "epoch": 5.1145958986731,
-      "grad_norm": 0.361328125,
-      "learning_rate": 0.00016655303774302976,
-      "loss": 0.9036,
-      "step": 2120
-    },
-    {
-      "epoch": 5.126658624849216,
-      "grad_norm": 0.49609375,
-      "learning_rate": 0.0001663430066134988,
-      "loss": 0.9224,
-      "step": 2125
-    },
-    {
-      "epoch": 5.138721351025332,
-      "grad_norm": 0.50390625,
-      "learning_rate": 0.0001661324514416022,
-      "loss": 0.9104,
-      "step": 2130
-    },
-    {
-      "epoch": 5.150784077201448,
-      "grad_norm": 0.4296875,
-      "learning_rate": 0.0001659213738905119,
-      "loss": 0.9145,
-      "step": 2135
-    },
-    {
-      "epoch": 5.162846803377564,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00016570977562752623,
-      "loss": 0.9142,
-      "step": 2140
-    },
-    {
-      "epoch": 5.1749095295536796,
-      "grad_norm": 0.36328125,
-      "learning_rate": 0.00016549765832405653,
-      "loss": 0.9082,
-      "step": 2145
-    },
-    {
-      "epoch": 5.186972255729795,
-      "grad_norm": 0.50390625,
-      "learning_rate": 0.000165285023655614,
-      "loss": 0.9093,
-      "step": 2150
-    },
-    {
-      "epoch": 5.1990349819059105,
-      "grad_norm": 0.55859375,
-      "learning_rate": 0.00016507187330179663,
-      "loss": 0.9211,
-      "step": 2155
-    },
-    {
-      "epoch": 5.211097708082026,
-      "grad_norm": 0.42578125,
-      "learning_rate": 0.0001648582089462756,
-      "loss": 0.9133,
-      "step": 2160
-    },
-    {
-      "epoch": 5.223160434258142,
-      "grad_norm": 0.365234375,
-      "learning_rate": 0.00016464403227678238,
-      "loss": 0.9211,
-      "step": 2165
-    },
-    {
-      "epoch": 5.235223160434258,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00016442934498509505,
-      "loss": 0.9197,
-      "step": 2170
-    },
-    {
-      "epoch": 5.247285886610374,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00016421414876702518,
-      "loss": 0.909,
-      "step": 2175
-    },
-    {
-      "epoch": 5.25934861278649,
-      "grad_norm": 0.390625,
-      "learning_rate": 0.00016399844532240433,
-      "loss": 0.9158,
-      "step": 2180
-    },
-    {
-      "epoch": 5.271411338962605,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.0001637822363550706,
-      "loss": 0.914,
-      "step": 2185
-    },
-    {
-      "epoch": 5.283474065138721,
-      "grad_norm": 0.439453125,
-      "learning_rate": 0.00016356552357285522,
-      "loss": 0.9123,
-      "step": 2190
-    },
-    {
-      "epoch": 5.295536791314837,
-      "grad_norm": 0.361328125,
-      "learning_rate": 0.00016334830868756906,
-      "loss": 0.9221,
-      "step": 2195
-    },
-    {
-      "epoch": 5.307599517490953,
-      "grad_norm": 0.41015625,
-      "learning_rate": 0.00016313059341498901,
-      "loss": 0.9146,
-      "step": 2200
-    },
-    {
-      "epoch": 5.319662243667069,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.0001629123794748447,
-      "loss": 0.9072,
-      "step": 2205
-    },
-    {
-      "epoch": 5.331724969843185,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00016269366859080451,
-      "loss": 0.9155,
-      "step": 2210
-    },
-    {
-      "epoch": 5.343787696019301,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00016247446249046237,
-      "loss": 0.9152,
-      "step": 2215
-    },
-    {
-      "epoch": 5.355850422195417,
-      "grad_norm": 0.45703125,
-      "learning_rate": 0.00016225476290532374,
-      "loss": 0.9229,
-      "step": 2220
-    },
-    {
-      "epoch": 5.367913148371532,
-      "grad_norm": 0.40234375,
-      "learning_rate": 0.0001620345715707922,
-      "loss": 0.9192,
-      "step": 2225
-    },
-    {
-      "epoch": 5.379975874547648,
-      "grad_norm": 0.3828125,
-      "learning_rate": 0.00016181389022615564,
-      "loss": 0.9131,
-      "step": 2230
-    },
-    {
-      "epoch": 5.3920386007237635,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.00016159272061457255,
-      "loss": 0.9173,
-      "step": 2235
-    },
-    {
-      "epoch": 5.404101326899879,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.0001613710644830582,
-      "loss": 0.9159,
-      "step": 2240
-    },
-    {
-      "epoch": 5.416164053075995,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.0001611489235824709,
-      "loss": 0.9114,
-      "step": 2245
-    },
-    {
-      "epoch": 5.428226779252111,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.0001609262996674981,
-      "loss": 0.926,
-      "step": 2250
-    },
-    {
-      "epoch": 5.440289505428227,
-      "grad_norm": 0.404296875,
-      "learning_rate": 0.00016070319449664264,
-      "loss": 0.9121,
-      "step": 2255
-    },
-    {
-      "epoch": 5.452352231604342,
-      "grad_norm": 0.390625,
-      "learning_rate": 0.00016047960983220873,
-      "loss": 0.9099,
-      "step": 2260
-    },
-    {
-      "epoch": 5.464414957780458,
-      "grad_norm": 0.5703125,
-      "learning_rate": 0.0001602555474402881,
-      "loss": 0.9105,
-      "step": 2265
-    },
-    {
-      "epoch": 5.476477683956574,
-      "grad_norm": 0.61328125,
-      "learning_rate": 0.0001600310090907461,
-      "loss": 0.913,
-      "step": 2270
-    },
-    {
-      "epoch": 5.48854041013269,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00015980599655720758,
-      "loss": 0.9153,
-      "step": 2275
-    },
-    {
-      "epoch": 5.500603136308806,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00015958051161704307,
-      "loss": 0.9116,
-      "step": 2280
-    },
-    {
-      "epoch": 5.512665862484922,
-      "grad_norm": 0.50390625,
-      "learning_rate": 0.00015935455605135446,
-      "loss": 0.9138,
-      "step": 2285
-    },
-    {
-      "epoch": 5.524728588661038,
-      "grad_norm": 0.41015625,
-      "learning_rate": 0.00015912813164496131,
-      "loss": 0.9166,
-      "step": 2290
-    },
-    {
-      "epoch": 5.536791314837153,
-      "grad_norm": 0.380859375,
-      "learning_rate": 0.00015890124018638638,
-      "loss": 0.9174,
-      "step": 2295
-    },
-    {
-      "epoch": 5.548854041013269,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.0001586738834678418,
-      "loss": 0.9165,
-      "step": 2300
-    },
-    {
-      "epoch": 5.560916767189385,
-      "grad_norm": 0.291015625,
-      "learning_rate": 0.00015844606328521463,
-      "loss": 0.912,
-      "step": 2305
-    },
-    {
-      "epoch": 5.572979493365501,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00015821778143805296,
-      "loss": 0.9125,
-      "step": 2310
-    },
-    {
-      "epoch": 5.5850422195416165,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.0001579890397295515,
-      "loss": 0.9264,
-      "step": 2315
-    },
-    {
-      "epoch": 5.597104945717732,
-      "grad_norm": 0.298828125,
-      "learning_rate": 0.00015775983996653737,
-      "loss": 0.9129,
-      "step": 2320
-    },
-    {
-      "epoch": 5.609167671893848,
-      "grad_norm": 0.4375,
-      "learning_rate": 0.00015753018395945598,
-      "loss": 0.9139,
-      "step": 2325
-    },
-    {
-      "epoch": 5.621230398069963,
-      "grad_norm": 0.341796875,
-      "learning_rate": 0.00015730007352235644,
-      "loss": 0.9078,
-      "step": 2330
-    },
-    {
-      "epoch": 5.633293124246079,
-      "grad_norm": 0.3515625,
-      "learning_rate": 0.0001570695104728775,
-      "loss": 0.905,
-      "step": 2335
-    },
-    {
-      "epoch": 5.645355850422195,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00015683849663223308,
-      "loss": 0.906,
-      "step": 2340
-    },
-    {
-      "epoch": 5.657418576598311,
-      "grad_norm": 0.337890625,
-      "learning_rate": 0.0001566070338251978,
-      "loss": 0.9144,
-      "step": 2345
-    },
-    {
-      "epoch": 5.669481302774427,
-      "grad_norm": 0.376953125,
-      "learning_rate": 0.00015637512388009284,
-      "loss": 0.9171,
-      "step": 2350
-    },
-    {
-      "epoch": 5.681544028950543,
-      "grad_norm": 0.349609375,
-      "learning_rate": 0.00015614276862877113,
-      "loss": 0.9169,
-      "step": 2355
-    },
-    {
-      "epoch": 5.693606755126659,
-      "grad_norm": 0.353515625,
-      "learning_rate": 0.00015590996990660317,
-      "loss": 0.9131,
-      "step": 2360
-    },
-    {
-      "epoch": 5.705669481302774,
-      "grad_norm": 0.37109375,
-      "learning_rate": 0.0001556767295524624,
-      "loss": 0.9248,
-      "step": 2365
-    },
-    {
-      "epoch": 5.71773220747889,
-      "grad_norm": 0.38671875,
-      "learning_rate": 0.0001554430494087107,
-      "loss": 0.9108,
-      "step": 2370
-    },
-    {
-      "epoch": 5.729794933655006,
-      "grad_norm": 0.359375,
-      "learning_rate": 0.00015520893132118385,
-      "loss": 0.909,
-      "step": 2375
-    },
-    {
-      "epoch": 5.741857659831122,
-      "grad_norm": 0.578125,
-      "learning_rate": 0.000154974377139177,
-      "loss": 0.9094,
-      "step": 2380
-    },
-    {
-      "epoch": 5.753920386007238,
-      "grad_norm": 0.61328125,
-      "learning_rate": 0.00015473938871542986,
-      "loss": 0.9155,
-      "step": 2385
-    },
-    {
-      "epoch": 5.765983112183354,
-      "grad_norm": 0.5546875,
-      "learning_rate": 0.00015450396790611234,
-      "loss": 0.9167,
-      "step": 2390
-    },
-    {
-      "epoch": 5.7780458383594695,
-      "grad_norm": 0.3984375,
-      "learning_rate": 0.00015426811657080967,
-      "loss": 0.9265,
-      "step": 2395
-    },
-    {
-      "epoch": 5.790108564535585,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00015403183657250788,
-      "loss": 0.9131,
-      "step": 2400
-    },
-    {
-      "epoch": 5.8021712907117005,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00015379512977757896,
-      "loss": 0.9088,
-      "step": 2405
-    },
-    {
-      "epoch": 5.814234016887816,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.00015355799805576612,
-      "loss": 0.9101,
-      "step": 2410
-    },
-    {
-      "epoch": 5.826296743063932,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00015332044328016914,
-      "loss": 0.9118,
-      "step": 2415
-    },
-    {
-      "epoch": 5.838359469240048,
-      "grad_norm": 0.3515625,
-      "learning_rate": 0.00015308246732722943,
-      "loss": 0.9107,
-      "step": 2420
-    },
-    {
-      "epoch": 5.850422195416164,
-      "grad_norm": 0.376953125,
-      "learning_rate": 0.0001528440720767153,
-      "loss": 0.9167,
-      "step": 2425
-    },
-    {
-      "epoch": 5.86248492159228,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.00015260525941170712,
-      "loss": 0.9059,
-      "step": 2430
-    },
-    {
-      "epoch": 5.874547647768396,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00015236603121858232,
-      "loss": 0.918,
-      "step": 2435
-    },
-    {
-      "epoch": 5.886610373944512,
-      "grad_norm": 0.35546875,
-      "learning_rate": 0.00015212638938700068,
-      "loss": 0.9084,
-      "step": 2440
-    },
-    {
-      "epoch": 5.898673100120627,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00015188633580988926,
-      "loss": 0.9093,
-      "step": 2445
-    },
-    {
-      "epoch": 5.910735826296743,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00015164587238342744,
-      "loss": 0.9016,
-      "step": 2450
-    },
-    {
-      "epoch": 5.922798552472859,
-      "grad_norm": 0.68359375,
-      "learning_rate": 0.00015140500100703207,
-      "loss": 0.9129,
-      "step": 2455
-    },
-    {
-      "epoch": 5.934861278648975,
-      "grad_norm": 0.37109375,
-      "learning_rate": 0.00015116372358334233,
-      "loss": 0.922,
-      "step": 2460
-    },
-    {
-      "epoch": 5.946924004825091,
-      "grad_norm": 0.396484375,
-      "learning_rate": 0.00015092204201820492,
-      "loss": 0.915,
-      "step": 2465
-    },
-    {
-      "epoch": 5.958986731001207,
-      "grad_norm": 0.365234375,
-      "learning_rate": 0.00015067995822065858,
-      "loss": 0.9097,
-      "step": 2470
-    },
-    {
-      "epoch": 5.9710494571773225,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00015043747410291945,
-      "loss": 0.9011,
-      "step": 2475
-    },
-    {
-      "epoch": 5.9831121833534375,
-      "grad_norm": 0.38671875,
-      "learning_rate": 0.0001501945915803658,
-      "loss": 0.919,
-      "step": 2480
-    },
-    {
-      "epoch": 5.9951749095295535,
-      "grad_norm": 0.447265625,
-      "learning_rate": 0.0001499513125715228,
-      "loss": 0.9129,
-      "step": 2485
-    },
-    {
-      "epoch": 6.0,
-      "eval_loss": 2.4628818035125732,
-      "eval_runtime": 0.5365,
-      "eval_samples_per_second": 18.639,
-      "eval_steps_per_second": 1.864,
-      "step": 2487
-    },
-    {
-      "epoch": 6.007237635705669,
-      "grad_norm": 0.376953125,
-      "learning_rate": 0.00014970763899804763,
-      "loss": 0.9123,
-      "step": 2490
-    },
-    {
-      "epoch": 6.019300361881785,
-      "grad_norm": 0.37109375,
-      "learning_rate": 0.00014946357278471388,
-      "loss": 0.899,
-      "step": 2495
-    },
-    {
-      "epoch": 6.031363088057901,
-      "grad_norm": 0.482421875,
-      "learning_rate": 0.0001492191158593968,
-      "loss": 0.8955,
-      "step": 2500
-    },
-    {
-      "epoch": 6.043425814234017,
-      "grad_norm": 0.478515625,
-      "learning_rate": 0.0001489742701530578,
-      "loss": 0.9005,
-      "step": 2505
-    },
-    {
-      "epoch": 6.055488540410133,
-      "grad_norm": 0.41796875,
-      "learning_rate": 0.0001487290375997292,
-      "loss": 0.8971,
-      "step": 2510
-    },
-    {
-      "epoch": 6.067551266586248,
-      "grad_norm": 0.40625,
-      "learning_rate": 0.00014848342013649914,
-      "loss": 0.9056,
-      "step": 2515
-    },
-    {
-      "epoch": 6.079613992762364,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00014823741970349606,
-      "loss": 0.904,
-      "step": 2520
-    },
-    {
-      "epoch": 6.09167671893848,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00014799103824387358,
-      "loss": 0.8914,
-      "step": 2525
-    },
-    {
-      "epoch": 6.103739445114596,
-      "grad_norm": 0.361328125,
-      "learning_rate": 0.0001477442777037949,
-      "loss": 0.9086,
-      "step": 2530
-    },
-    {
-      "epoch": 6.115802171290712,
-      "grad_norm": 0.369140625,
-      "learning_rate": 0.0001474971400324177,
-      "loss": 0.9118,
-      "step": 2535
-    },
-    {
-      "epoch": 6.127864897466828,
-      "grad_norm": 0.36328125,
-      "learning_rate": 0.00014724962718187852,
-      "loss": 0.9044,
-      "step": 2540
-    },
-    {
-      "epoch": 6.139927623642944,
-      "grad_norm": 0.380859375,
-      "learning_rate": 0.00014700174110727747,
-      "loss": 0.9075,
-      "step": 2545
-    },
-    {
-      "epoch": 6.151990349819059,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00014675348376666278,
-      "loss": 0.8888,
-      "step": 2550
-    },
-    {
-      "epoch": 6.164053075995175,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.00014650485712101524,
-      "loss": 0.8976,
-      "step": 2555
-    },
-    {
-      "epoch": 6.1761158021712905,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00014625586313423287,
-      "loss": 0.9052,
-      "step": 2560
-    },
-    {
-      "epoch": 6.1881785283474064,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.00014600650377311522,
-      "loss": 0.8917,
-      "step": 2565
-    },
-    {
-      "epoch": 6.200241254523522,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00014575678100734796,
-      "loss": 0.9049,
-      "step": 2570
-    },
-    {
-      "epoch": 6.212303980699638,
-      "grad_norm": 0.390625,
-      "learning_rate": 0.0001455066968094873,
-      "loss": 0.9041,
-      "step": 2575
-    },
-    {
-      "epoch": 6.224366706875754,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00014525625315494435,
-      "loss": 0.9048,
-      "step": 2580
-    },
-    {
-      "epoch": 6.23642943305187,
-      "grad_norm": 0.375,
-      "learning_rate": 0.00014500545202196968,
-      "loss": 0.9093,
-      "step": 2585
-    },
-    {
-      "epoch": 6.248492159227985,
-      "grad_norm": 0.4296875,
-      "learning_rate": 0.00014475429539163742,
-      "loss": 0.9003,
-      "step": 2590
-    },
-    {
-      "epoch": 6.260554885404101,
-      "grad_norm": 0.39453125,
-      "learning_rate": 0.00014450278524782986,
-      "loss": 0.9062,
-      "step": 2595
-    },
-    {
-      "epoch": 6.272617611580217,
-      "grad_norm": 0.4609375,
-      "learning_rate": 0.00014425092357722168,
-      "loss": 0.9032,
-      "step": 2600
-    },
-    {
-      "epoch": 6.284680337756333,
-      "grad_norm": 0.3828125,
-      "learning_rate": 0.00014399871236926422,
-      "loss": 0.9009,
-      "step": 2605
-    },
-    {
-      "epoch": 6.296743063932449,
-      "grad_norm": 0.41796875,
-      "learning_rate": 0.00014374615361616985,
-      "loss": 0.8952,
-      "step": 2610
-    },
-    {
-      "epoch": 6.308805790108565,
-      "grad_norm": 0.376953125,
-      "learning_rate": 0.00014349324931289627,
-      "loss": 0.8959,
-      "step": 2615
-    },
-    {
-      "epoch": 6.320868516284681,
-      "grad_norm": 0.302734375,
-      "learning_rate": 0.0001432400014571305,
-      "loss": 0.906,
-      "step": 2620
-    },
-    {
-      "epoch": 6.332931242460796,
-      "grad_norm": 0.373046875,
-      "learning_rate": 0.00014298641204927342,
-      "loss": 0.899,
-      "step": 2625
-    },
-    {
-      "epoch": 6.344993968636912,
-      "grad_norm": 0.400390625,
-      "learning_rate": 0.00014273248309242372,
-      "loss": 0.895,
-      "step": 2630
-    },
-    {
-      "epoch": 6.357056694813028,
-      "grad_norm": 0.384765625,
-      "learning_rate": 0.0001424782165923623,
-      "loss": 0.8986,
-      "step": 2635
-    },
-    {
-      "epoch": 6.3691194209891435,
-      "grad_norm": 0.41796875,
-      "learning_rate": 0.0001422236145575362,
-      "loss": 0.9039,
-      "step": 2640
-    },
-    {
-      "epoch": 6.381182147165259,
-      "grad_norm": 0.443359375,
-      "learning_rate": 0.0001419686789990429,
-      "loss": 0.9086,
-      "step": 2645
-    },
-    {
-      "epoch": 6.393244873341375,
-      "grad_norm": 0.369140625,
-      "learning_rate": 0.0001417134119306144,
-      "loss": 0.8945,
-      "step": 2650
-    },
-    {
-      "epoch": 6.405307599517491,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.00014145781536860122,
-      "loss": 0.9083,
-      "step": 2655
-    },
-    {
-      "epoch": 6.417370325693607,
-      "grad_norm": 0.35546875,
-      "learning_rate": 0.00014120189133195657,
-      "loss": 0.9029,
-      "step": 2660
-    },
-    {
-      "epoch": 6.429433051869722,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00014094564184222042,
-      "loss": 0.9124,
-      "step": 2665
-    },
-    {
-      "epoch": 6.441495778045838,
-      "grad_norm": 0.361328125,
-      "learning_rate": 0.00014068906892350343,
-      "loss": 0.9106,
-      "step": 2670
-    },
-    {
-      "epoch": 6.453558504221954,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.00014043217460247109,
-      "loss": 0.9021,
-      "step": 2675
-    },
-    {
-      "epoch": 6.46562123039807,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.0001401749609083276,
-      "loss": 0.907,
-      "step": 2680
-    },
-    {
-      "epoch": 6.477683956574186,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.0001399174298727998,
-      "loss": 0.8993,
-      "step": 2685
-    },
-    {
-      "epoch": 6.489746682750302,
-      "grad_norm": 0.390625,
-      "learning_rate": 0.0001396595835301214,
-      "loss": 0.9022,
-      "step": 2690
-    },
-    {
-      "epoch": 6.501809408926418,
-      "grad_norm": 0.384765625,
-      "learning_rate": 0.00013940142391701652,
-      "loss": 0.9056,
-      "step": 2695
-    },
-    {
-      "epoch": 6.513872135102533,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00013914295307268396,
-      "loss": 0.9099,
-      "step": 2700
-    },
-    {
-      "epoch": 6.525934861278649,
-      "grad_norm": 0.365234375,
-      "learning_rate": 0.00013888417303878077,
-      "loss": 0.9055,
-      "step": 2705
-    },
-    {
-      "epoch": 6.537997587454765,
-      "grad_norm": 0.36328125,
-      "learning_rate": 0.00013862508585940644,
-      "loss": 0.9037,
-      "step": 2710
-    },
-    {
-      "epoch": 6.550060313630881,
-      "grad_norm": 0.353515625,
-      "learning_rate": 0.00013836569358108647,
-      "loss": 0.8909,
-      "step": 2715
-    },
-    {
-      "epoch": 6.5621230398069965,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.00013810599825275638,
-      "loss": 0.9075,
-      "step": 2720
-    },
-    {
-      "epoch": 6.574185765983112,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00013784600192574557,
-      "loss": 0.9112,
-      "step": 2725
-    },
-    {
-      "epoch": 6.586248492159228,
-      "grad_norm": 0.3828125,
-      "learning_rate": 0.00013758570665376086,
-      "loss": 0.9033,
-      "step": 2730
-    },
-    {
-      "epoch": 6.598311218335343,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.00013732511449287055,
-      "loss": 0.8993,
-      "step": 2735
-    },
-    {
-      "epoch": 6.610373944511459,
-      "grad_norm": 0.361328125,
-      "learning_rate": 0.00013706422750148797,
-      "loss": 0.8954,
-      "step": 2740
-    },
-    {
-      "epoch": 6.622436670687575,
-      "grad_norm": 0.396484375,
-      "learning_rate": 0.00013680304774035538,
-      "loss": 0.9054,
-      "step": 2745
-    },
-    {
-      "epoch": 6.634499396863691,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00013654157727252766,
-      "loss": 0.9018,
-      "step": 2750
-    },
-    {
-      "epoch": 6.646562123039807,
-      "grad_norm": 0.50390625,
-      "learning_rate": 0.00013627981816335583,
-      "loss": 0.9036,
-      "step": 2755
-    },
-    {
-      "epoch": 6.658624849215923,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00013601777248047105,
-      "loss": 0.9067,
-      "step": 2760
-    },
-    {
-      "epoch": 6.670687575392039,
-      "grad_norm": 0.3125,
-      "learning_rate": 0.000135755442293768,
-      "loss": 0.8984,
-      "step": 2765
-    },
-    {
-      "epoch": 6.682750301568154,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.00013549282967538872,
-      "loss": 0.907,
-      "step": 2770
-    },
-    {
-      "epoch": 6.69481302774427,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.0001352299366997062,
-      "loss": 0.9024,
-      "step": 2775
-    },
-    {
-      "epoch": 6.706875753920386,
-      "grad_norm": 0.375,
-      "learning_rate": 0.0001349667654433078,
-      "loss": 0.9069,
-      "step": 2780
-    },
-    {
-      "epoch": 6.718938480096502,
-      "grad_norm": 0.365234375,
-      "learning_rate": 0.00013470331798497926,
-      "loss": 0.9075,
-      "step": 2785
-    },
-    {
-      "epoch": 6.731001206272618,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.0001344395964056878,
-      "loss": 0.9034,
-      "step": 2790
-    },
-    {
-      "epoch": 6.743063932448734,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.0001341756027885661,
-      "loss": 0.9037,
-      "step": 2795
-    },
-    {
-      "epoch": 6.7551266586248495,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00013391133921889558,
-      "loss": 0.9075,
-      "step": 2800
-    },
-    {
-      "epoch": 6.7671893848009645,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00013364680778409,
-      "loss": 0.9046,
-      "step": 2805
-    },
-    {
-      "epoch": 6.7792521109770805,
-      "grad_norm": 0.35546875,
-      "learning_rate": 0.000133382010573679,
-      "loss": 0.9002,
-      "step": 2810
-    },
-    {
-      "epoch": 6.791314837153196,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.00013311694967929158,
-      "loss": 0.9066,
-      "step": 2815
-    },
-    {
-      "epoch": 6.803377563329312,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00013285162719463961,
-      "loss": 0.9057,
-      "step": 2820
-    },
-    {
-      "epoch": 6.815440289505428,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.0001325860452155012,
-      "loss": 0.8974,
-      "step": 2825
-    },
-    {
-      "epoch": 6.827503015681544,
-      "grad_norm": 0.30859375,
-      "learning_rate": 0.0001323202058397042,
-      "loss": 0.9006,
-      "step": 2830
-    },
-    {
-      "epoch": 6.83956574185766,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00013205411116710972,
-      "loss": 0.9023,
-      "step": 2835
-    },
-    {
-      "epoch": 6.851628468033776,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00013178776329959531,
-      "loss": 0.903,
-      "step": 2840
-    },
-    {
-      "epoch": 6.863691194209892,
-      "grad_norm": 0.380859375,
-      "learning_rate": 0.00013152116434103867,
-      "loss": 0.9027,
-      "step": 2845
-    },
-    {
-      "epoch": 6.875753920386007,
-      "grad_norm": 0.3828125,
-      "learning_rate": 0.0001312543163973007,
-      "loss": 0.9087,
-      "step": 2850
-    },
-    {
-      "epoch": 6.887816646562123,
-      "grad_norm": 0.359375,
-      "learning_rate": 0.00013098722157620917,
-      "loss": 0.8917,
-      "step": 2855
-    },
-    {
-      "epoch": 6.899879372738239,
-      "grad_norm": 0.37109375,
-      "learning_rate": 0.00013071988198754185,
-      "loss": 0.902,
-      "step": 2860
-    },
-    {
-      "epoch": 6.911942098914355,
-      "grad_norm": 0.357421875,
-      "learning_rate": 0.00013045229974300993,
-      "loss": 0.905,
-      "step": 2865
-    },
-    {
-      "epoch": 6.924004825090471,
-      "grad_norm": 0.3671875,
-      "learning_rate": 0.0001301844769562414,
-      "loss": 0.9072,
-      "step": 2870
-    },
-    {
-      "epoch": 6.936067551266587,
-      "grad_norm": 0.35546875,
-      "learning_rate": 0.00012991641574276418,
-      "loss": 0.9002,
-      "step": 2875
-    },
-    {
-      "epoch": 6.9481302774427025,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.0001296481182199896,
-      "loss": 0.8994,
-      "step": 2880
-    },
-    {
-      "epoch": 6.9601930036188175,
-      "grad_norm": 0.3046875,
-      "learning_rate": 0.0001293795865071956,
-      "loss": 0.9076,
-      "step": 2885
-    },
-    {
-      "epoch": 6.9722557297949335,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.0001291108227255099,
-      "loss": 0.9044,
-      "step": 2890
-    },
-    {
-      "epoch": 6.984318455971049,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00012884182899789343,
-      "loss": 0.9003,
-      "step": 2895
-    },
-    {
-      "epoch": 6.996381182147165,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.0001285726074491234,
-      "loss": 0.8995,
-      "step": 2900
-    },
-    {
-      "epoch": 6.998793727382388,
-      "eval_loss": 2.4702937602996826,
-      "eval_runtime": 0.6137,
-      "eval_samples_per_second": 16.294,
-      "eval_steps_per_second": 1.629,
-      "step": 2901
-    },
-    {
-      "epoch": 7.008443908323281,
-      "grad_norm": 0.349609375,
-      "learning_rate": 0.00012830316020577656,
-      "loss": 0.8934,
-      "step": 2905
-    },
-    {
-      "epoch": 7.020506634499397,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00012803348939621252,
-      "loss": 0.8901,
-      "step": 2910
-    },
-    {
-      "epoch": 7.032569360675513,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00012776359715055668,
-      "loss": 0.8963,
-      "step": 2915
-    },
-    {
-      "epoch": 7.044632086851628,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.0001274934856006837,
-      "loss": 0.8909,
-      "step": 2920
-    },
-    {
-      "epoch": 7.056694813027744,
-      "grad_norm": 0.353515625,
-      "learning_rate": 0.00012722315688020047,
-      "loss": 0.8836,
-      "step": 2925
-    },
-    {
-      "epoch": 7.06875753920386,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.0001269526131244292,
-      "loss": 0.8914,
-      "step": 2930
-    },
-    {
-      "epoch": 7.080820265379976,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.0001266818564703909,
-      "loss": 0.9006,
-      "step": 2935
-    },
-    {
-      "epoch": 7.092882991556092,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00012641088905678802,
-      "loss": 0.9071,
-      "step": 2940
-    },
-    {
-      "epoch": 7.104945717732208,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00012613971302398794,
-      "loss": 0.8939,
-      "step": 2945
-    },
-    {
-      "epoch": 7.117008443908324,
-      "grad_norm": 0.361328125,
-      "learning_rate": 0.00012586833051400588,
-      "loss": 0.8813,
-      "step": 2950
-    },
-    {
-      "epoch": 7.129071170084439,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.000125596743670488,
-      "loss": 0.8852,
-      "step": 2955
-    },
-    {
-      "epoch": 7.141133896260555,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00012532495463869452,
-      "loss": 0.8924,
-      "step": 2960
-    },
-    {
-      "epoch": 7.1531966224366705,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00012505296556548272,
-      "loss": 0.891,
-      "step": 2965
-    },
-    {
-      "epoch": 7.1652593486127865,
-      "grad_norm": 0.3515625,
-      "learning_rate": 0.00012478077859929,
-      "loss": 0.8914,
-      "step": 2970
-    },
-    {
-      "epoch": 7.177322074788902,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00012450839589011695,
-      "loss": 0.8947,
-      "step": 2975
-    },
-    {
-      "epoch": 7.189384800965018,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00012423581958951026,
-      "loss": 0.8955,
-      "step": 2980
-    },
-    {
-      "epoch": 7.201447527141134,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.0001239630518505459,
-      "loss": 0.898,
-      "step": 2985
-    },
-    {
-      "epoch": 7.213510253317249,
-      "grad_norm": 0.376953125,
-      "learning_rate": 0.00012369009482781192,
-      "loss": 0.8951,
-      "step": 2990
-    },
-    {
-      "epoch": 7.225572979493365,
-      "grad_norm": 0.53125,
-      "learning_rate": 0.00012341695067739154,
-      "loss": 0.8944,
-      "step": 2995
-    },
-    {
-      "epoch": 7.237635705669481,
-      "grad_norm": 0.421875,
-      "learning_rate": 0.00012314362155684612,
-      "loss": 0.8936,
-      "step": 3000
-    },
-    {
-      "epoch": 7.249698431845597,
-      "grad_norm": 0.43359375,
-      "learning_rate": 0.00012287010962519807,
-      "loss": 0.8998,
-      "step": 3005
-    },
-    {
-      "epoch": 7.261761158021713,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00012259641704291384,
-      "loss": 0.8959,
-      "step": 3010
-    },
-    {
-      "epoch": 7.273823884197829,
-      "grad_norm": 0.421875,
-      "learning_rate": 0.00012232254597188688,
-      "loss": 0.8942,
-      "step": 3015
-    },
-    {
-      "epoch": 7.285886610373945,
-      "grad_norm": 0.37890625,
-      "learning_rate": 0.00012204849857542038,
-      "loss": 0.8886,
-      "step": 3020
-    },
-    {
-      "epoch": 7.297949336550061,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00012177427701821051,
-      "loss": 0.8859,
-      "step": 3025
-    },
-    {
-      "epoch": 7.310012062726176,
-      "grad_norm": 0.361328125,
-      "learning_rate": 0.00012149988346632894,
-      "loss": 0.8988,
-      "step": 3030
-    },
-    {
-      "epoch": 7.322074788902292,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00012122532008720611,
-      "loss": 0.9011,
-      "step": 3035
-    },
-    {
-      "epoch": 7.334137515078408,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00012095058904961379,
-      "loss": 0.8903,
-      "step": 3040
-    },
-    {
-      "epoch": 7.3462002412545235,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00012067569252364809,
-      "loss": 0.9006,
-      "step": 3045
-    },
-    {
-      "epoch": 7.3582629674306395,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00012040063268071243,
-      "loss": 0.8968,
-      "step": 3050
-    },
-    {
-      "epoch": 7.370325693606755,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00012012541169350011,
-      "loss": 0.9044,
-      "step": 3055
-    },
-    {
-      "epoch": 7.382388419782871,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.0001198500317359774,
-      "loss": 0.8968,
-      "step": 3060
-    },
-    {
-      "epoch": 7.394451145958986,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00011957449498336626,
-      "loss": 0.8956,
-      "step": 3065
-    },
-    {
-      "epoch": 7.406513872135102,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.00011929880361212717,
-      "loss": 0.8967,
-      "step": 3070
-    },
-    {
-      "epoch": 7.418576598311218,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00011902295979994192,
-      "loss": 0.8995,
-      "step": 3075
-    },
-    {
-      "epoch": 7.430639324487334,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00011874696572569645,
-      "loss": 0.8923,
-      "step": 3080
-    },
-    {
-      "epoch": 7.44270205066345,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.0001184708235694636,
-      "loss": 0.8915,
-      "step": 3085
-    },
-    {
-      "epoch": 7.454764776839566,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00011819453551248592,
-      "loss": 0.8951,
-      "step": 3090
-    },
-    {
-      "epoch": 7.466827503015682,
-      "grad_norm": 0.365234375,
-      "learning_rate": 0.00011791810373715846,
-      "loss": 0.8924,
-      "step": 3095
-    },
-    {
-      "epoch": 7.478890229191798,
-      "grad_norm": 0.443359375,
-      "learning_rate": 0.0001176415304270114,
-      "loss": 0.895,
-      "step": 3100
-    },
-    {
-      "epoch": 7.490952955367913,
-      "grad_norm": 0.41796875,
-      "learning_rate": 0.00011736481776669306,
-      "loss": 0.8965,
-      "step": 3105
-    },
-    {
-      "epoch": 7.503015681544029,
-      "grad_norm": 0.36328125,
-      "learning_rate": 0.00011708796794195227,
-      "loss": 0.9014,
-      "step": 3110
-    },
-    {
-      "epoch": 7.515078407720145,
-      "grad_norm": 0.373046875,
-      "learning_rate": 0.00011681098313962158,
-      "loss": 0.8904,
-      "step": 3115
-    },
-    {
-      "epoch": 7.527141133896261,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00011653386554759946,
-      "loss": 0.8932,
-      "step": 3120
-    },
-    {
-      "epoch": 7.5392038600723765,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00011625661735483348,
-      "loss": 0.8999,
-      "step": 3125
-    },
-    {
-      "epoch": 7.5512665862484925,
-      "grad_norm": 0.37890625,
-      "learning_rate": 0.00011597924075130273,
-      "loss": 0.8913,
-      "step": 3130
-    },
-    {
-      "epoch": 7.563329312424608,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00011570173792800066,
-      "loss": 0.8957,
-      "step": 3135
-    },
-    {
-      "epoch": 7.575392038600723,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.00011542411107691767,
-      "loss": 0.8927,
-      "step": 3140
-    },
-    {
-      "epoch": 7.587454764776839,
-      "grad_norm": 0.376953125,
-      "learning_rate": 0.00011514636239102393,
-      "loss": 0.897,
-      "step": 3145
-    },
-    {
-      "epoch": 7.599517490952955,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00011486849406425188,
-      "loss": 0.8962,
-      "step": 3150
-    },
-    {
-      "epoch": 7.611580217129071,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00011459050829147915,
-      "loss": 0.8853,
-      "step": 3155
-    },
-    {
-      "epoch": 7.623642943305187,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.00011431240726851091,
-      "loss": 0.8978,
-      "step": 3160
-    },
-    {
-      "epoch": 7.635705669481303,
-      "grad_norm": 0.326171875,
-      "learning_rate": 0.00011403419319206284,
-      "loss": 0.894,
-      "step": 3165
-    },
-    {
-      "epoch": 7.647768395657419,
-      "grad_norm": 0.37890625,
-      "learning_rate": 0.00011375586825974349,
-      "loss": 0.8913,
-      "step": 3170
-    },
-    {
-      "epoch": 7.659831121833534,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.0001134774346700371,
-      "loss": 0.8915,
-      "step": 3175
-    },
-    {
-      "epoch": 7.67189384800965,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.0001131988946222863,
-      "loss": 0.8962,
-      "step": 3180
-    },
-    {
-      "epoch": 7.683956574185766,
-      "grad_norm": 0.322265625,
-      "learning_rate": 0.00011292025031667448,
-      "loss": 0.8949,
-      "step": 3185
-    },
-    {
-      "epoch": 7.696019300361882,
-      "grad_norm": 0.33203125,
-      "learning_rate": 0.00011264150395420865,
-      "loss": 0.8974,
-      "step": 3190
-    },
-    {
-      "epoch": 7.708082026537998,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00011236265773670196,
-      "loss": 0.8933,
-      "step": 3195
-    },
-    {
-      "epoch": 7.720144752714114,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00011208371386675624,
-      "loss": 0.8996,
-      "step": 3200
-    },
-    {
-      "epoch": 7.7322074788902295,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00011180467454774481,
-      "loss": 0.892,
-      "step": 3205
-    },
-    {
-      "epoch": 7.744270205066345,
-      "grad_norm": 0.35546875,
-      "learning_rate": 0.00011152554198379484,
-      "loss": 0.8943,
-      "step": 3210
-    },
-    {
-      "epoch": 7.7563329312424605,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00011124631837976997,
-      "loss": 0.8921,
-      "step": 3215
-    },
-    {
-      "epoch": 7.768395657418576,
-      "grad_norm": 0.373046875,
-      "learning_rate": 0.00011096700594125318,
-      "loss": 0.891,
-      "step": 3220
-    },
-    {
-      "epoch": 7.780458383594692,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00011068760687452895,
-      "loss": 0.8956,
-      "step": 3225
-    },
-    {
-      "epoch": 7.792521109770808,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00011040812338656615,
-      "loss": 0.8986,
-      "step": 3230
-    },
-    {
-      "epoch": 7.804583835946924,
-      "grad_norm": 0.3203125,
-      "learning_rate": 0.0001101285576850004,
-      "loss": 0.8905,
-      "step": 3235
-    },
-    {
-      "epoch": 7.81664656212304,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.00010984891197811687,
-      "loss": 0.8861,
-      "step": 3240
-    },
-    {
-      "epoch": 7.828709288299155,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00010956918847483255,
-      "loss": 0.8908,
-      "step": 3245
-    },
-    {
-      "epoch": 7.840772014475271,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00010928938938467896,
-      "loss": 0.8972,
-      "step": 3250
-    },
-    {
-      "epoch": 7.852834740651387,
-      "grad_norm": 0.310546875,
-      "learning_rate": 0.00010900951691778481,
-      "loss": 0.8977,
-      "step": 3255
-    },
-    {
-      "epoch": 7.864897466827503,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00010872957328485819,
-      "loss": 0.8974,
-      "step": 3260
-    },
-    {
-      "epoch": 7.876960193003619,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00010844956069716957,
-      "loss": 0.8964,
-      "step": 3265
-    },
-    {
-      "epoch": 7.889022919179735,
-      "grad_norm": 0.30078125,
-      "learning_rate": 0.00010816948136653386,
-      "loss": 0.889,
-      "step": 3270
-    },
-    {
-      "epoch": 7.901085645355851,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.00010788933750529335,
-      "loss": 0.8914,
-      "step": 3275
-    },
-    {
-      "epoch": 7.913148371531967,
-      "grad_norm": 0.314453125,
-      "learning_rate": 0.00010760913132630002,
-      "loss": 0.8923,
-      "step": 3280
-    },
-    {
-      "epoch": 7.9252110977080825,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.000107328865042898,
-      "loss": 0.8954,
-      "step": 3285
-    },
-    {
-      "epoch": 7.9372738238841976,
-      "grad_norm": 0.34375,
-      "learning_rate": 0.00010704854086890633,
-      "loss": 0.8992,
-      "step": 3290
-    },
-    {
-      "epoch": 7.9493365500603135,
-      "grad_norm": 0.330078125,
-      "learning_rate": 0.00010676816101860118,
-      "loss": 0.8954,
-      "step": 3295
-    },
-    {
-      "epoch": 7.961399276236429,
-      "grad_norm": 0.31640625,
-      "learning_rate": 0.00010648772770669861,
-      "loss": 0.8952,
-      "step": 3300
-    },
-    {
-      "epoch": 7.973462002412545,
-      "grad_norm": 0.353515625,
-      "learning_rate": 0.00010620724314833696,
-      "loss": 0.8903,
-      "step": 3305
-    },
-    {
-      "epoch": 7.985524728588661,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.00010592670955905935,
-      "loss": 0.8909,
-      "step": 3310
-    },
-    {
-      "epoch": 7.997587454764777,
-      "grad_norm": 0.357421875,
-      "learning_rate": 0.00010564612915479612,
-      "loss": 0.8999,
-      "step": 3315
-    },
-    {
-      "epoch": 8.0,
-      "eval_loss": 2.4829864501953125,
-      "eval_runtime": 0.5356,
-      "eval_samples_per_second": 18.67,
-      "eval_steps_per_second": 1.867,
-      "step": 3316
-    },
-    {
-      "epoch": 8.009650180940893,
-      "grad_norm": 0.306640625,
-      "learning_rate": 0.00010536550415184751,
-      "loss": 0.889,
-      "step": 3320
-    },
-    {
-      "epoch": 8.021712907117008,
-      "grad_norm": 0.3359375,
-      "learning_rate": 0.000105084836766866,
-      "loss": 0.8866,
-      "step": 3325
-    },
-    {
-      "epoch": 8.033775633293125,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00010480412921683888,
-      "loss": 0.8776,
-      "step": 3330
-    },
-    {
-      "epoch": 8.04583835946924,
-      "grad_norm": 0.361328125,
-      "learning_rate": 0.00010452338371907064,
-      "loss": 0.8824,
-      "step": 3335
-    },
-    {
-      "epoch": 8.057901085645355,
-      "grad_norm": 0.361328125,
-      "learning_rate": 0.00010424260249116557,
-      "loss": 0.8906,
-      "step": 3340
-    },
-    {
-      "epoch": 8.069963811821472,
-      "grad_norm": 0.328125,
-      "learning_rate": 0.00010396178775101014,
-      "loss": 0.8918,
-      "step": 3345
-    },
-    {
-      "epoch": 8.082026537997587,
-      "grad_norm": 0.359375,
-      "learning_rate": 0.0001036809417167556,
-      "loss": 0.8811,
-      "step": 3350
-    },
-    {
-      "epoch": 8.094089264173704,
-      "grad_norm": 0.359375,
-      "learning_rate": 0.00010340006660680032,
-      "loss": 0.896,
-      "step": 3355
-    },
-    {
-      "epoch": 8.106151990349819,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00010311916463977242,
-      "loss": 0.8858,
-      "step": 3360
-    },
-    {
-      "epoch": 8.118214716525936,
-      "grad_norm": 0.392578125,
-      "learning_rate": 0.00010283823803451208,
-      "loss": 0.8834,
-      "step": 3365
-    },
-    {
-      "epoch": 8.13027744270205,
-      "grad_norm": 0.32421875,
-      "learning_rate": 0.00010255728901005417,
-      "loss": 0.882,
-      "step": 3370
-    },
-    {
-      "epoch": 8.142340168878167,
-      "grad_norm": 0.33984375,
-      "learning_rate": 0.00010227631978561056,
-      "loss": 0.8819,
-      "step": 3375
-    },
-    {
-      "epoch": 8.154402895054282,
-      "grad_norm": 0.34765625,
-      "learning_rate": 0.00010199533258055282,
-      "loss": 0.8755,
-      "step": 3380
-    },
-    {
-      "epoch": 8.166465621230397,
-      "grad_norm": 0.333984375,
-      "learning_rate": 0.00010171432961439438,
-      "loss": 0.8847,
-      "step": 3385
-    },
-    {
-      "epoch": 8.178528347406514,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00010143331310677331,
-      "loss": 0.8951,
-      "step": 3390
-    },
-    {
-      "epoch": 8.19059107358263,
-      "grad_norm": 0.373046875,
-      "learning_rate": 0.00010115228527743452,
-      "loss": 0.888,
-      "step": 3395
-    },
-    {
-      "epoch": 8.202653799758746,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.00010087124834621248,
-      "loss": 0.8875,
-      "step": 3400
-    },
-    {
-      "epoch": 8.214716525934861,
-      "grad_norm": 0.345703125,
-      "learning_rate": 0.00010059020453301345,
-      "loss": 0.8902,
-      "step": 3405
-    },
-    {
-      "epoch": 8.226779252110978,
-      "grad_norm": 0.431640625,
-      "learning_rate": 0.00010030915605779809,
-      "loss": 0.8871,
-      "step": 3410
-    },
-    {
-      "epoch": 8.238841978287093,
-      "grad_norm": 0.318359375,
-      "learning_rate": 0.00010002810514056391,
-      "loss": 0.8884,
-      "step": 3415
-    },
-    {
-      "epoch": 8.250904704463208,
-      "grad_norm": 0.31640625,
-      "learning_rate": 9.974705400132764e-05,
-      "loss": 0.8859,
-      "step": 3420
-    },
-    {
-      "epoch": 8.262967430639325,
-      "grad_norm": 0.3125,
-      "learning_rate": 9.946600486010785e-05,
-      "loss": 0.8897,
-      "step": 3425
-    },
-    {
-      "epoch": 8.27503015681544,
-      "grad_norm": 0.326171875,
-      "learning_rate": 9.918495993690724e-05,
-      "loss": 0.8877,
-      "step": 3430
-    },
-    {
-      "epoch": 8.287092882991557,
-      "grad_norm": 0.31640625,
-      "learning_rate": 9.890392145169531e-05,
-      "loss": 0.8841,
-      "step": 3435
-    },
-    {
-      "epoch": 8.299155609167672,
-      "grad_norm": 0.447265625,
-      "learning_rate": 9.862289162439059e-05,
-      "loss": 0.8854,
-      "step": 3440
-    },
-    {
-      "epoch": 8.311218335343789,
-      "grad_norm": 0.341796875,
-      "learning_rate": 9.834187267484326e-05,
-      "loss": 0.8843,
-      "step": 3445
-    },
-    {
-      "epoch": 8.323281061519904,
-      "grad_norm": 0.361328125,
-      "learning_rate": 9.806086682281758e-05,
-      "loss": 0.8886,
-      "step": 3450
-    },
-    {
-      "epoch": 8.335343787696019,
-      "grad_norm": 0.33984375,
-      "learning_rate": 9.777987628797437e-05,
-      "loss": 0.8878,
-      "step": 3455
-    },
-    {
-      "epoch": 8.347406513872135,
-      "grad_norm": 0.33203125,
-      "learning_rate": 9.749890328985345e-05,
-      "loss": 0.8811,
-      "step": 3460
-    },
-    {
-      "epoch": 8.35946924004825,
-      "grad_norm": 0.40234375,
-      "learning_rate": 9.721795004785605e-05,
-      "loss": 0.8902,
-      "step": 3465
-    },
-    {
-      "epoch": 8.371531966224367,
-      "grad_norm": 0.380859375,
-      "learning_rate": 9.693701878122749e-05,
-      "loss": 0.8854,
-      "step": 3470
-    },
-    {
-      "epoch": 8.383594692400482,
-      "grad_norm": 0.3671875,
-      "learning_rate": 9.665611170903942e-05,
-      "loss": 0.8934,
-      "step": 3475
-    },
-    {
-      "epoch": 8.395657418576599,
-      "grad_norm": 0.421875,
-      "learning_rate": 9.637523105017229e-05,
-      "loss": 0.8905,
-      "step": 3480
-    },
-    {
-      "epoch": 8.407720144752714,
-      "grad_norm": 0.4140625,
-      "learning_rate": 9.609437902329814e-05,
-      "loss": 0.8884,
-      "step": 3485
-    },
-    {
-      "epoch": 8.41978287092883,
-      "grad_norm": 0.314453125,
-      "learning_rate": 9.581355784686266e-05,
-      "loss": 0.8864,
-      "step": 3490
-    },
-    {
-      "epoch": 8.431845597104946,
-      "grad_norm": 0.35546875,
-      "learning_rate": 9.553276973906786e-05,
-      "loss": 0.8924,
-      "step": 3495
-    },
-    {
-      "epoch": 8.443908323281061,
-      "grad_norm": 0.318359375,
-      "learning_rate": 9.525201691785462e-05,
-      "loss": 0.8938,
-      "step": 3500
-    },
-    {
-      "epoch": 8.455971049457178,
-      "grad_norm": 0.357421875,
-      "learning_rate": 9.497130160088508e-05,
-      "loss": 0.8937,
-      "step": 3505
-    },
-    {
-      "epoch": 8.468033775633293,
-      "grad_norm": 0.32421875,
-      "learning_rate": 9.469062600552509e-05,
-      "loss": 0.8847,
-      "step": 3510
-    },
-    {
-      "epoch": 8.48009650180941,
-      "grad_norm": 0.306640625,
-      "learning_rate": 9.44099923488267e-05,
-      "loss": 0.8882,
-      "step": 3515
-    },
-    {
-      "epoch": 8.492159227985525,
-      "grad_norm": 0.37890625,
-      "learning_rate": 9.412940284751089e-05,
-      "loss": 0.8885,
-      "step": 3520
-    },
-    {
-      "epoch": 8.50422195416164,
-      "grad_norm": 0.33984375,
-      "learning_rate": 9.384885971794961e-05,
-      "loss": 0.8938,
-      "step": 3525
-    },
-    {
-      "epoch": 8.516284680337757,
-      "grad_norm": 0.318359375,
-      "learning_rate": 9.356836517614863e-05,
-      "loss": 0.8938,
-      "step": 3530
-    },
-    {
-      "epoch": 8.528347406513872,
-      "grad_norm": 0.400390625,
-      "learning_rate": 9.328792143773e-05,
-      "loss": 0.8765,
-      "step": 3535
-    },
-    {
-      "epoch": 8.540410132689988,
-      "grad_norm": 0.38671875,
-      "learning_rate": 9.300753071791434e-05,
-      "loss": 0.8879,
-      "step": 3540
-    },
-    {
-      "epoch": 8.552472858866103,
-      "grad_norm": 0.361328125,
-      "learning_rate": 9.272719523150355e-05,
-      "loss": 0.8927,
-      "step": 3545
-    },
-    {
-      "epoch": 8.56453558504222,
-      "grad_norm": 0.35546875,
-      "learning_rate": 9.24469171928632e-05,
-      "loss": 0.8867,
-      "step": 3550
-    },
-    {
-      "epoch": 8.576598311218335,
-      "grad_norm": 0.494140625,
-      "learning_rate": 9.216669881590515e-05,
-      "loss": 0.8951,
-      "step": 3555
-    },
-    {
-      "epoch": 8.588661037394452,
-      "grad_norm": 0.349609375,
-      "learning_rate": 9.188654231406993e-05,
-      "loss": 0.8918,
-      "step": 3560
-    },
-    {
-      "epoch": 8.600723763570567,
-      "grad_norm": 0.345703125,
-      "learning_rate": 9.160644990030931e-05,
-      "loss": 0.8894,
-      "step": 3565
-    },
-    {
-      "epoch": 8.612786489746682,
-      "grad_norm": 0.376953125,
-      "learning_rate": 9.132642378706894e-05,
-      "loss": 0.8868,
-      "step": 3570
-    },
-    {
-      "epoch": 8.624849215922799,
-      "grad_norm": 0.322265625,
-      "learning_rate": 9.104646618627063e-05,
-      "loss": 0.8848,
-      "step": 3575
-    },
-    {
-      "epoch": 8.636911942098914,
-      "grad_norm": 0.369140625,
-      "learning_rate": 9.076657930929507e-05,
-      "loss": 0.8997,
-      "step": 3580
-    },
-    {
-      "epoch": 8.64897466827503,
-      "grad_norm": 0.341796875,
-      "learning_rate": 9.048676536696425e-05,
-      "loss": 0.8922,
-      "step": 3585
-    },
-    {
-      "epoch": 8.661037394451146,
-      "grad_norm": 0.365234375,
-      "learning_rate": 9.020702656952421e-05,
-      "loss": 0.8889,
-      "step": 3590
-    },
-    {
-      "epoch": 8.67310012062726,
-      "grad_norm": 0.373046875,
-      "learning_rate": 8.992736512662724e-05,
-      "loss": 0.8801,
-      "step": 3595
-    },
-    {
-      "epoch": 8.685162846803378,
-      "grad_norm": 0.3125,
-      "learning_rate": 8.964778324731467e-05,
-      "loss": 0.8891,
-      "step": 3600
-    },
-    {
-      "epoch": 8.697225572979493,
-      "grad_norm": 0.353515625,
-      "learning_rate": 8.93682831399994e-05,
-      "loss": 0.8902,
-      "step": 3605
-    },
-    {
-      "epoch": 8.70928829915561,
-      "grad_norm": 0.318359375,
-      "learning_rate": 8.908886701244838e-05,
-      "loss": 0.8901,
-      "step": 3610
-    },
-    {
-      "epoch": 8.721351025331725,
-      "grad_norm": 0.326171875,
-      "learning_rate": 8.880953707176514e-05,
-      "loss": 0.8915,
-      "step": 3615
-    },
-    {
-      "epoch": 8.733413751507841,
-      "grad_norm": 0.369140625,
-      "learning_rate": 8.853029552437259e-05,
-      "loss": 0.8906,
-      "step": 3620
-    },
-    {
-      "epoch": 8.745476477683956,
-      "grad_norm": 0.34765625,
-      "learning_rate": 8.825114457599527e-05,
-      "loss": 0.8869,
-      "step": 3625
-    },
-    {
-      "epoch": 8.757539203860073,
-      "grad_norm": 0.32421875,
-      "learning_rate": 8.797208643164212e-05,
-      "loss": 0.8998,
-      "step": 3630
-    },
-    {
-      "epoch": 8.769601930036188,
-      "grad_norm": 0.330078125,
-      "learning_rate": 8.7693123295589e-05,
-      "loss": 0.8776,
-      "step": 3635
-    },
-    {
-      "epoch": 8.781664656212303,
-      "grad_norm": 0.337890625,
-      "learning_rate": 8.741425737136143e-05,
-      "loss": 0.8951,
-      "step": 3640
-    },
-    {
-      "epoch": 8.79372738238842,
-      "grad_norm": 0.337890625,
-      "learning_rate": 8.713549086171691e-05,
-      "loss": 0.8865,
-      "step": 3645
-    },
-    {
-      "epoch": 8.805790108564535,
-      "grad_norm": 0.3203125,
-      "learning_rate": 8.685682596862768e-05,
-      "loss": 0.8861,
-      "step": 3650
-    },
-    {
-      "epoch": 8.817852834740652,
-      "grad_norm": 0.35546875,
-      "learning_rate": 8.65782648932634e-05,
-      "loss": 0.8918,
-      "step": 3655
-    },
-    {
-      "epoch": 8.829915560916767,
-      "grad_norm": 0.330078125,
-      "learning_rate": 8.629980983597358e-05,
-      "loss": 0.893,
-      "step": 3660
-    },
-    {
-      "epoch": 8.841978287092884,
-      "grad_norm": 0.337890625,
-      "learning_rate": 8.602146299627035e-05,
-      "loss": 0.8871,
-      "step": 3665
-    },
-    {
-      "epoch": 8.854041013268999,
-      "grad_norm": 0.326171875,
-      "learning_rate": 8.574322657281105e-05,
-      "loss": 0.8899,
-      "step": 3670
-    },
-    {
-      "epoch": 8.866103739445114,
-      "grad_norm": 0.326171875,
-      "learning_rate": 8.546510276338078e-05,
-      "loss": 0.884,
-      "step": 3675
-    },
-    {
-      "epoch": 8.87816646562123,
-      "grad_norm": 0.44140625,
-      "learning_rate": 8.518709376487515e-05,
-      "loss": 0.8889,
-      "step": 3680
-    },
-    {
-      "epoch": 8.890229191797346,
-      "grad_norm": 0.333984375,
-      "learning_rate": 8.490920177328282e-05,
-      "loss": 0.8859,
-      "step": 3685
-    },
-    {
-      "epoch": 8.902291917973463,
-      "grad_norm": 0.3515625,
-      "learning_rate": 8.463142898366834e-05,
-      "loss": 0.8816,
-      "step": 3690
-    },
-    {
-      "epoch": 8.914354644149578,
-      "grad_norm": 0.330078125,
-      "learning_rate": 8.435377759015455e-05,
-      "loss": 0.8858,
-      "step": 3695
-    },
-    {
-      "epoch": 8.926417370325694,
-      "grad_norm": 0.326171875,
-      "learning_rate": 8.407624978590543e-05,
-      "loss": 0.8767,
-      "step": 3700
-    },
-    {
-      "epoch": 8.93848009650181,
-      "grad_norm": 0.3984375,
-      "learning_rate": 8.37988477631088e-05,
-      "loss": 0.8946,
-      "step": 3705
-    },
-    {
-      "epoch": 8.950542822677924,
-      "grad_norm": 0.3125,
-      "learning_rate": 8.352157371295884e-05,
-      "loss": 0.883,
-      "step": 3710
-    },
-    {
-      "epoch": 8.962605548854041,
-      "grad_norm": 0.326171875,
-      "learning_rate": 8.324442982563887e-05,
-      "loss": 0.8866,
-      "step": 3715
-    },
-    {
-      "epoch": 8.974668275030156,
-      "grad_norm": 0.330078125,
-      "learning_rate": 8.296741829030418e-05,
-      "loss": 0.8844,
-      "step": 3720
-    },
-    {
-      "epoch": 8.986731001206273,
-      "grad_norm": 0.34375,
-      "learning_rate": 8.269054129506449e-05,
-      "loss": 0.88,
-      "step": 3725
-    },
-    {
-      "epoch": 8.998793727382388,
-      "grad_norm": 0.341796875,
-      "learning_rate": 8.241380102696681e-05,
-      "loss": 0.8762,
-      "step": 3730
-    },
-    {
-      "epoch": 8.998793727382388,
-      "eval_loss": 2.493363618850708,
-      "eval_runtime": 0.66,
-      "eval_samples_per_second": 15.152,
-      "eval_steps_per_second": 1.515,
-      "step": 3730
-    },
-    {
-      "epoch": 9.010856453558505,
-      "grad_norm": 0.36328125,
-      "learning_rate": 8.213719967197817e-05,
-      "loss": 0.8809,
-      "step": 3735
-    },
-    {
-      "epoch": 9.02291917973462,
-      "grad_norm": 0.322265625,
-      "learning_rate": 8.186073941496835e-05,
-      "loss": 0.8795,
-      "step": 3740
-    },
-    {
-      "epoch": 9.034981905910735,
-      "grad_norm": 0.33203125,
-      "learning_rate": 8.158442243969256e-05,
-      "loss": 0.8773,
-      "step": 3745
-    },
-    {
-      "epoch": 9.047044632086852,
-      "grad_norm": 0.3359375,
-      "learning_rate": 8.130825092877418e-05,
-      "loss": 0.8885,
-      "step": 3750
-    },
-    {
-      "epoch": 9.059107358262967,
-      "grad_norm": 0.345703125,
-      "learning_rate": 8.103222706368774e-05,
-      "loss": 0.8857,
-      "step": 3755
-    },
-    {
-      "epoch": 9.071170084439084,
-      "grad_norm": 0.322265625,
-      "learning_rate": 8.075635302474134e-05,
-      "loss": 0.8747,
-      "step": 3760
-    },
-    {
-      "epoch": 9.083232810615199,
-      "grad_norm": 0.34765625,
-      "learning_rate": 8.04806309910597e-05,
-      "loss": 0.879,
-      "step": 3765
-    },
-    {
-      "epoch": 9.095295536791316,
-      "grad_norm": 0.35546875,
-      "learning_rate": 8.020506314056677e-05,
-      "loss": 0.878,
-      "step": 3770
-    },
-    {
-      "epoch": 9.10735826296743,
-      "grad_norm": 0.322265625,
-      "learning_rate": 7.992965164996875e-05,
-      "loss": 0.8809,
-      "step": 3775
-    },
-    {
-      "epoch": 9.119420989143546,
-      "grad_norm": 0.35546875,
-      "learning_rate": 7.965439869473664e-05,
-      "loss": 0.8815,
-      "step": 3780
-    },
-    {
-      "epoch": 9.131483715319662,
-      "grad_norm": 0.35546875,
-      "learning_rate": 7.937930644908917e-05,
-      "loss": 0.8826,
-      "step": 3785
-    },
-    {
-      "epoch": 9.143546441495777,
-      "grad_norm": 0.345703125,
-      "learning_rate": 7.91043770859757e-05,
-      "loss": 0.8863,
-      "step": 3790
-    },
-    {
-      "epoch": 9.155609167671894,
-      "grad_norm": 0.40234375,
-      "learning_rate": 7.882961277705895e-05,
-      "loss": 0.8792,
-      "step": 3795
-    },
-    {
-      "epoch": 9.16767189384801,
-      "grad_norm": 0.326171875,
-      "learning_rate": 7.855501569269782e-05,
-      "loss": 0.8743,
-      "step": 3800
-    },
-    {
-      "epoch": 9.179734620024126,
-      "grad_norm": 0.314453125,
-      "learning_rate": 7.82805880019304e-05,
-      "loss": 0.881,
-      "step": 3805
-    },
-    {
-      "epoch": 9.191797346200241,
-      "grad_norm": 0.3203125,
-      "learning_rate": 7.800633187245673e-05,
-      "loss": 0.8873,
-      "step": 3810
-    },
-    {
-      "epoch": 9.203860072376358,
-      "grad_norm": 0.341796875,
-      "learning_rate": 7.773224947062163e-05,
-      "loss": 0.8766,
-      "step": 3815
-    },
-    {
-      "epoch": 9.215922798552473,
-      "grad_norm": 0.384765625,
-      "learning_rate": 7.745834296139763e-05,
-      "loss": 0.8818,
-      "step": 3820
-    },
-    {
-      "epoch": 9.227985524728588,
-      "grad_norm": 0.328125,
-      "learning_rate": 7.718461450836804e-05,
-      "loss": 0.8894,
-      "step": 3825
-    },
-    {
-      "epoch": 9.240048250904705,
-      "grad_norm": 0.314453125,
-      "learning_rate": 7.691106627370955e-05,
-      "loss": 0.8866,
-      "step": 3830
-    },
-    {
-      "epoch": 9.25211097708082,
-      "grad_norm": 0.341796875,
-      "learning_rate": 7.663770041817534e-05,
-      "loss": 0.8854,
-      "step": 3835
-    },
-    {
-      "epoch": 9.264173703256937,
-      "grad_norm": 0.328125,
-      "learning_rate": 7.636451910107806e-05,
-      "loss": 0.8832,
-      "step": 3840
-    },
-    {
-      "epoch": 9.276236429433052,
-      "grad_norm": 0.33984375,
-      "learning_rate": 7.609152448027261e-05,
-      "loss": 0.8853,
-      "step": 3845
-    },
-    {
-      "epoch": 9.288299155609169,
-      "grad_norm": 0.33984375,
-      "learning_rate": 7.581871871213913e-05,
-      "loss": 0.8826,
-      "step": 3850
-    },
-    {
-      "epoch": 9.300361881785284,
-      "grad_norm": 0.384765625,
-      "learning_rate": 7.554610395156624e-05,
-      "loss": 0.8851,
-      "step": 3855
-    },
-    {
-      "epoch": 9.312424607961399,
-      "grad_norm": 0.32421875,
-      "learning_rate": 7.527368235193355e-05,
-      "loss": 0.8812,
-      "step": 3860
-    },
-    {
-      "epoch": 9.324487334137515,
-      "grad_norm": 0.412109375,
-      "learning_rate": 7.500145606509501e-05,
-      "loss": 0.8883,
-      "step": 3865
-    },
-    {
-      "epoch": 9.33655006031363,
-      "grad_norm": 0.337890625,
-      "learning_rate": 7.472942724136174e-05,
-      "loss": 0.885,
-      "step": 3870
-    },
-    {
-      "epoch": 9.348612786489747,
-      "grad_norm": 0.337890625,
-      "learning_rate": 7.44575980294852e-05,
-      "loss": 0.8857,
-      "step": 3875
-    },
-    {
-      "epoch": 9.360675512665862,
-      "grad_norm": 0.369140625,
-      "learning_rate": 7.418597057664005e-05,
-      "loss": 0.8872,
-      "step": 3880
-    },
-    {
-      "epoch": 9.372738238841979,
-      "grad_norm": 0.3125,
-      "learning_rate": 7.391454702840722e-05,
-      "loss": 0.8826,
-      "step": 3885
-    },
-    {
-      "epoch": 9.384800965018094,
-      "grad_norm": 0.314453125,
-      "learning_rate": 7.36433295287571e-05,
-      "loss": 0.8822,
-      "step": 3890
-    },
-    {
-      "epoch": 9.39686369119421,
-      "grad_norm": 0.34765625,
-      "learning_rate": 7.337232022003244e-05,
-      "loss": 0.8833,
-      "step": 3895
-    },
-    {
-      "epoch": 9.408926417370326,
-      "grad_norm": 0.359375,
-      "learning_rate": 7.310152124293146e-05,
-      "loss": 0.8782,
-      "step": 3900
-    },
-    {
-      "epoch": 9.420989143546441,
-      "grad_norm": 0.314453125,
-      "learning_rate": 7.28309347364911e-05,
-      "loss": 0.8879,
-      "step": 3905
-    },
-    {
-      "epoch": 9.433051869722558,
-      "grad_norm": 0.326171875,
-      "learning_rate": 7.256056283806986e-05,
-      "loss": 0.8806,
-      "step": 3910
-    },
-    {
-      "epoch": 9.445114595898673,
-      "grad_norm": 0.33984375,
-      "learning_rate": 7.229040768333115e-05,
-      "loss": 0.8841,
-      "step": 3915
-    },
-    {
-      "epoch": 9.45717732207479,
-      "grad_norm": 0.314453125,
-      "learning_rate": 7.202047140622621e-05,
-      "loss": 0.8718,
-      "step": 3920
-    },
-    {
-      "epoch": 9.469240048250905,
-      "grad_norm": 0.341796875,
-      "learning_rate": 7.175075613897756e-05,
-      "loss": 0.8843,
-      "step": 3925
-    },
-    {
-      "epoch": 9.48130277442702,
-      "grad_norm": 0.34765625,
-      "learning_rate": 7.14812640120618e-05,
-      "loss": 0.8788,
-      "step": 3930
-    },
-    {
-      "epoch": 9.493365500603137,
-      "grad_norm": 0.33984375,
-      "learning_rate": 7.121199715419295e-05,
-      "loss": 0.8891,
-      "step": 3935
-    },
-    {
-      "epoch": 9.505428226779252,
-      "grad_norm": 0.3359375,
-      "learning_rate": 7.094295769230577e-05,
-      "loss": 0.8808,
-      "step": 3940
-    },
-    {
-      "epoch": 9.517490952955368,
-      "grad_norm": 0.306640625,
-      "learning_rate": 7.067414775153871e-05,
-      "loss": 0.8817,
-      "step": 3945
-    },
-    {
-      "epoch": 9.529553679131483,
-      "grad_norm": 0.359375,
-      "learning_rate": 7.040556945521721e-05,
-      "loss": 0.8837,
-      "step": 3950
-    },
-    {
-      "epoch": 9.5416164053076,
-      "grad_norm": 0.30859375,
-      "learning_rate": 7.013722492483708e-05,
-      "loss": 0.8867,
-      "step": 3955
-    },
-    {
-      "epoch": 9.553679131483715,
-      "grad_norm": 0.31640625,
-      "learning_rate": 6.986911628004753e-05,
-      "loss": 0.8807,
-      "step": 3960
-    },
-    {
-      "epoch": 9.56574185765983,
-      "grad_norm": 0.306640625,
-      "learning_rate": 6.96012456386345e-05,
-      "loss": 0.873,
-      "step": 3965
-    },
-    {
-      "epoch": 9.577804583835947,
-      "grad_norm": 0.330078125,
-      "learning_rate": 6.933361511650395e-05,
-      "loss": 0.8852,
-      "step": 3970
-    },
-    {
-      "epoch": 9.589867310012062,
-      "grad_norm": 0.326171875,
-      "learning_rate": 6.906622682766526e-05,
-      "loss": 0.8914,
-      "step": 3975
-    },
-    {
-      "epoch": 9.601930036188179,
-      "grad_norm": 0.357421875,
-      "learning_rate": 6.879908288421425e-05,
-      "loss": 0.8691,
-      "step": 3980
-    },
-    {
-      "epoch": 9.613992762364294,
-      "grad_norm": 0.341796875,
-      "learning_rate": 6.853218539631672e-05,
-      "loss": 0.8839,
-      "step": 3985
-    },
-    {
-      "epoch": 9.62605548854041,
-      "grad_norm": 0.310546875,
-      "learning_rate": 6.826553647219175e-05,
-      "loss": 0.8879,
-      "step": 3990
-    },
-    {
-      "epoch": 9.638118214716526,
-      "grad_norm": 0.349609375,
-      "learning_rate": 6.799913821809501e-05,
-      "loss": 0.8807,
-      "step": 3995
-    },
-    {
-      "epoch": 9.650180940892643,
-      "grad_norm": 0.34375,
-      "learning_rate": 6.773299273830207e-05,
-      "loss": 0.8771,
-      "step": 4000
-    },
-    {
-      "epoch": 9.662243667068758,
-      "grad_norm": 0.341796875,
-      "learning_rate": 6.74671021350919e-05,
-      "loss": 0.8814,
-      "step": 4005
-    },
-    {
-      "epoch": 9.674306393244873,
-      "grad_norm": 0.318359375,
-      "learning_rate": 6.720146850873025e-05,
-      "loss": 0.8844,
-      "step": 4010
-    },
-    {
-      "epoch": 9.68636911942099,
-      "grad_norm": 0.296875,
-      "learning_rate": 6.69360939574529e-05,
-      "loss": 0.8801,
-      "step": 4015
-    },
-    {
-      "epoch": 9.698431845597105,
-      "grad_norm": 0.31640625,
-      "learning_rate": 6.667098057744927e-05,
-      "loss": 0.8837,
-      "step": 4020
-    },
-    {
-      "epoch": 9.710494571773221,
-      "grad_norm": 0.314453125,
-      "learning_rate": 6.640613046284581e-05,
-      "loss": 0.8821,
-      "step": 4025
-    },
-    {
-      "epoch": 9.722557297949336,
-      "grad_norm": 0.310546875,
-      "learning_rate": 6.614154570568934e-05,
-      "loss": 0.89,
-      "step": 4030
-    },
-    {
-      "epoch": 9.734620024125451,
-      "grad_norm": 0.322265625,
-      "learning_rate": 6.587722839593073e-05,
-      "loss": 0.8775,
-      "step": 4035
-    },
-    {
-      "epoch": 9.746682750301568,
-      "grad_norm": 0.31640625,
-      "learning_rate": 6.56131806214083e-05,
-      "loss": 0.8802,
-      "step": 4040
-    },
-    {
-      "epoch": 9.758745476477683,
-      "grad_norm": 0.373046875,
-      "learning_rate": 6.534940446783122e-05,
-      "loss": 0.8763,
-      "step": 4045
-    },
-    {
-      "epoch": 9.7708082026538,
-      "grad_norm": 0.435546875,
-      "learning_rate": 6.508590201876317e-05,
-      "loss": 0.8889,
-      "step": 4050
-    },
-    {
-      "epoch": 9.782870928829915,
-      "grad_norm": 0.33984375,
-      "learning_rate": 6.482267535560583e-05,
-      "loss": 0.8794,
-      "step": 4055
-    },
-    {
-      "epoch": 9.794933655006032,
-      "grad_norm": 0.3203125,
-      "learning_rate": 6.455972655758253e-05,
-      "loss": 0.8911,
-      "step": 4060
-    },
-    {
-      "epoch": 9.806996381182147,
-      "grad_norm": 0.322265625,
-      "learning_rate": 6.429705770172168e-05,
-      "loss": 0.8822,
-      "step": 4065
-    },
-    {
-      "epoch": 9.819059107358264,
-      "grad_norm": 0.32421875,
-      "learning_rate": 6.403467086284041e-05,
-      "loss": 0.881,
-      "step": 4070
-    },
-    {
-      "epoch": 9.831121833534379,
-      "grad_norm": 0.30078125,
-      "learning_rate": 6.377256811352825e-05,
-      "loss": 0.8814,
-      "step": 4075
-    },
-    {
-      "epoch": 9.843184559710494,
-      "grad_norm": 0.328125,
-      "learning_rate": 6.351075152413068e-05,
-      "loss": 0.8883,
-      "step": 4080
-    },
-    {
-      "epoch": 9.85524728588661,
-      "grad_norm": 0.322265625,
-      "learning_rate": 6.324922316273282e-05,
-      "loss": 0.8806,
-      "step": 4085
-    },
-    {
-      "epoch": 9.867310012062726,
-      "grad_norm": 0.31640625,
-      "learning_rate": 6.298798509514312e-05,
-      "loss": 0.8843,
-      "step": 4090
-    },
-    {
-      "epoch": 9.879372738238843,
-      "grad_norm": 0.353515625,
-      "learning_rate": 6.272703938487694e-05,
-      "loss": 0.8865,
-      "step": 4095
-    },
-    {
-      "epoch": 9.891435464414958,
-      "grad_norm": 0.3515625,
-      "learning_rate": 6.246638809314036e-05,
-      "loss": 0.882,
-      "step": 4100
-    },
-    {
-      "epoch": 9.903498190591074,
-      "grad_norm": 0.322265625,
-      "learning_rate": 6.220603327881375e-05,
-      "loss": 0.8865,
-      "step": 4105
-    },
-    {
-      "epoch": 9.91556091676719,
-      "grad_norm": 0.318359375,
-      "learning_rate": 6.194597699843581e-05,
-      "loss": 0.8814,
-      "step": 4110
-    },
-    {
-      "epoch": 9.927623642943304,
-      "grad_norm": 0.3515625,
-      "learning_rate": 6.168622130618694e-05,
-      "loss": 0.8799,
-      "step": 4115
-    },
-    {
-      "epoch": 9.939686369119421,
-      "grad_norm": 0.3046875,
-      "learning_rate": 6.142676825387328e-05,
-      "loss": 0.8835,
-      "step": 4120
-    },
-    {
-      "epoch": 9.951749095295536,
-      "grad_norm": 0.32421875,
-      "learning_rate": 6.116761989091042e-05,
-      "loss": 0.882,
-      "step": 4125
-    },
-    {
-      "epoch": 9.963811821471653,
-      "grad_norm": 0.326171875,
-      "learning_rate": 6.090877826430721e-05,
-      "loss": 0.8727,
-      "step": 4130
-    },
-    {
-      "epoch": 9.975874547647768,
-      "grad_norm": 0.404296875,
-      "learning_rate": 6.065024541864952e-05,
-      "loss": 0.8867,
-      "step": 4135
-    },
-    {
-      "epoch": 9.987937273823885,
-      "grad_norm": 0.3515625,
-      "learning_rate": 6.039202339608432e-05,
-      "loss": 0.8825,
-      "step": 4140
-    },
-    {
-      "epoch": 10.0,
-      "grad_norm": 0.333984375,
-      "learning_rate": 6.0134114236303275e-05,
-      "loss": 0.8821,
-      "step": 4145
-    },
-    {
-      "epoch": 10.0,
-      "eval_loss": 2.4973604679107666,
-      "eval_runtime": 0.5303,
-      "eval_samples_per_second": 18.857,
-      "eval_steps_per_second": 1.886,
-      "step": 4145
-    },
-    {
-      "epoch": 10.012062726176115,
-      "grad_norm": 0.337890625,
-      "learning_rate": 5.987651997652677e-05,
-      "loss": 0.8757,
-      "step": 4150
-    },
-    {
-      "epoch": 10.024125452352232,
-      "grad_norm": 0.3125,
-      "learning_rate": 5.961924265148777e-05,
-      "loss": 0.8787,
-      "step": 4155
-    },
-    {
-      "epoch": 10.036188178528347,
-      "grad_norm": 0.326171875,
-      "learning_rate": 5.93622842934159e-05,
-      "loss": 0.8815,
-      "step": 4160
-    },
-    {
-      "epoch": 10.048250904704464,
-      "grad_norm": 0.3515625,
-      "learning_rate": 5.9105646932021155e-05,
-      "loss": 0.8806,
-      "step": 4165
-    },
-    {
-      "epoch": 10.060313630880579,
-      "grad_norm": 0.314453125,
-      "learning_rate": 5.884933259447798e-05,
-      "loss": 0.8848,
-      "step": 4170
-    },
-    {
-      "epoch": 10.072376357056696,
-      "grad_norm": 0.31640625,
-      "learning_rate": 5.8593343305409355e-05,
-      "loss": 0.8788,
-      "step": 4175
-    },
-    {
-      "epoch": 10.08443908323281,
-      "grad_norm": 0.310546875,
-      "learning_rate": 5.833768108687063e-05,
-      "loss": 0.8794,
-      "step": 4180
-    },
-    {
-      "epoch": 10.096501809408926,
-      "grad_norm": 0.31640625,
-      "learning_rate": 5.8082347958333625e-05,
-      "loss": 0.8835,
-      "step": 4185
-    },
-    {
-      "epoch": 10.108564535585042,
-      "grad_norm": 0.359375,
-      "learning_rate": 5.7827345936670674e-05,
-      "loss": 0.872,
-      "step": 4190
-    },
-    {
-      "epoch": 10.120627261761157,
-      "grad_norm": 0.333984375,
-      "learning_rate": 5.7572677036138826e-05,
-      "loss": 0.8782,
-      "step": 4195
-    },
-    {
-      "epoch": 10.132689987937274,
-      "grad_norm": 0.330078125,
-      "learning_rate": 5.731834326836366e-05,
-      "loss": 0.8814,
-      "step": 4200
-    },
-    {
-      "epoch": 10.14475271411339,
-      "grad_norm": 0.369140625,
-      "learning_rate": 5.7064346642323584e-05,
-      "loss": 0.8783,
-      "step": 4205
-    },
-    {
-      "epoch": 10.156815440289506,
-      "grad_norm": 0.357421875,
-      "learning_rate": 5.681068916433403e-05,
-      "loss": 0.8805,
-      "step": 4210
-    },
-    {
-      "epoch": 10.168878166465621,
-      "grad_norm": 0.314453125,
-      "learning_rate": 5.6557372838031384e-05,
-      "loss": 0.8792,
-      "step": 4215
-    },
-    {
-      "epoch": 10.180940892641736,
-      "grad_norm": 0.34765625,
-      "learning_rate": 5.6304399664357275e-05,
-      "loss": 0.8759,
-      "step": 4220
-    },
-    {
-      "epoch": 10.193003618817853,
-      "grad_norm": 0.3125,
-      "learning_rate": 5.605177164154293e-05,
-      "loss": 0.8775,
-      "step": 4225
-    },
-    {
-      "epoch": 10.205066344993968,
-      "grad_norm": 0.3203125,
-      "learning_rate": 5.579949076509305e-05,
-      "loss": 0.8815,
-      "step": 4230
-    },
-    {
-      "epoch": 10.217129071170085,
-      "grad_norm": 0.310546875,
-      "learning_rate": 5.5547559027770325e-05,
-      "loss": 0.863,
-      "step": 4235
-    },
-    {
-      "epoch": 10.2291917973462,
-      "grad_norm": 0.32421875,
-      "learning_rate": 5.529597841957959e-05,
-      "loss": 0.8743,
-      "step": 4240
-    },
-    {
-      "epoch": 10.241254523522317,
-      "grad_norm": 0.3515625,
-      "learning_rate": 5.5044750927752106e-05,
-      "loss": 0.8785,
-      "step": 4245
-    },
-    {
-      "epoch": 10.253317249698432,
-      "grad_norm": 0.3203125,
-      "learning_rate": 5.479387853672988e-05,
-      "loss": 0.8758,
-      "step": 4250
-    },
-    {
-      "epoch": 10.265379975874549,
-      "grad_norm": 0.31640625,
-      "learning_rate": 5.4543363228149946e-05,
-      "loss": 0.8833,
-      "step": 4255
-    },
-    {
-      "epoch": 10.277442702050664,
-      "grad_norm": 0.349609375,
-      "learning_rate": 5.429320698082887e-05,
-      "loss": 0.88,
-      "step": 4260
-    },
-    {
-      "epoch": 10.289505428226779,
-      "grad_norm": 0.326171875,
-      "learning_rate": 5.404341177074686e-05,
-      "loss": 0.876,
-      "step": 4265
-    },
-    {
-      "epoch": 10.301568154402895,
-      "grad_norm": 0.322265625,
-      "learning_rate": 5.379397957103231e-05,
-      "loss": 0.873,
-      "step": 4270
-    },
-    {
-      "epoch": 10.31363088057901,
-      "grad_norm": 0.337890625,
-      "learning_rate": 5.354491235194635e-05,
-      "loss": 0.8816,
-      "step": 4275
-    },
-    {
-      "epoch": 10.325693606755127,
-      "grad_norm": 0.3359375,
-      "learning_rate": 5.3296212080866967e-05,
-      "loss": 0.8828,
-      "step": 4280
-    },
-    {
-      "epoch": 10.337756332931242,
-      "grad_norm": 0.330078125,
-      "learning_rate": 5.304788072227367e-05,
-      "loss": 0.8825,
-      "step": 4285
-    },
-    {
-      "epoch": 10.349819059107359,
-      "grad_norm": 0.302734375,
-      "learning_rate": 5.279992023773195e-05,
-      "loss": 0.8735,
-      "step": 4290
-    },
-    {
-      "epoch": 10.361881785283474,
-      "grad_norm": 0.31640625,
-      "learning_rate": 5.255233258587784e-05,
-      "loss": 0.881,
-      "step": 4295
-    },
-    {
-      "epoch": 10.37394451145959,
-      "grad_norm": 0.31640625,
-      "learning_rate": 5.2305119722402254e-05,
-      "loss": 0.8755,
-      "step": 4300
-    },
-    {
-      "epoch": 10.386007237635706,
-      "grad_norm": 0.328125,
-      "learning_rate": 5.205828360003568e-05,
-      "loss": 0.8776,
-      "step": 4305
-    },
-    {
-      "epoch": 10.398069963811821,
-      "grad_norm": 0.3125,
-      "learning_rate": 5.181182616853286e-05,
-      "loss": 0.8838,
-      "step": 4310
-    },
-    {
-      "epoch": 10.410132689987938,
-      "grad_norm": 0.314453125,
-      "learning_rate": 5.15657493746571e-05,
-      "loss": 0.8817,
-      "step": 4315
-    },
-    {
-      "epoch": 10.422195416164053,
-      "grad_norm": 0.34765625,
-      "learning_rate": 5.1320055162165115e-05,
-      "loss": 0.8752,
-      "step": 4320
-    },
-    {
-      "epoch": 10.43425814234017,
-      "grad_norm": 0.337890625,
-      "learning_rate": 5.10747454717917e-05,
-      "loss": 0.8791,
-      "step": 4325
-    },
-    {
-      "epoch": 10.446320868516285,
-      "grad_norm": 0.318359375,
-      "learning_rate": 5.082982224123418e-05,
-      "loss": 0.8745,
-      "step": 4330
-    },
-    {
-      "epoch": 10.4583835946924,
-      "grad_norm": 0.349609375,
-      "learning_rate": 5.0585287405137305e-05,
-      "loss": 0.8712,
-      "step": 4335
-    },
-    {
-      "epoch": 10.470446320868517,
-      "grad_norm": 0.322265625,
-      "learning_rate": 5.034114289507792e-05,
-      "loss": 0.871,
-      "step": 4340
-    },
-    {
-      "epoch": 10.482509047044632,
-      "grad_norm": 0.31640625,
-      "learning_rate": 5.009739063954964e-05,
-      "loss": 0.8759,
-      "step": 4345
-    },
-    {
-      "epoch": 10.494571773220748,
-      "grad_norm": 0.33203125,
-      "learning_rate": 4.9854032563947714e-05,
-      "loss": 0.8772,
-      "step": 4350
-    },
-    {
-      "epoch": 10.506634499396863,
-      "grad_norm": 0.306640625,
-      "learning_rate": 4.9611070590553724e-05,
-      "loss": 0.8886,
-      "step": 4355
-    },
-    {
-      "epoch": 10.51869722557298,
-      "grad_norm": 0.341796875,
-      "learning_rate": 4.936850663852053e-05,
-      "loss": 0.8729,
-      "step": 4360
-    },
-    {
-      "epoch": 10.530759951749095,
-      "grad_norm": 0.337890625,
-      "learning_rate": 4.912634262385695e-05,
-      "loss": 0.8781,
-      "step": 4365
-    },
-    {
-      "epoch": 10.54282267792521,
-      "grad_norm": 0.3125,
-      "learning_rate": 4.888458045941269e-05,
-      "loss": 0.8788,
-      "step": 4370
-    },
-    {
-      "epoch": 10.554885404101327,
-      "grad_norm": 0.330078125,
-      "learning_rate": 4.864322205486335e-05,
-      "loss": 0.8897,
-      "step": 4375
-    },
-    {
-      "epoch": 10.566948130277442,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.8402269316695134e-05,
-      "loss": 0.878,
-      "step": 4380
-    },
-    {
-      "epoch": 10.579010856453559,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.8161724148189934e-05,
-      "loss": 0.8773,
-      "step": 4385
-    },
-    {
-      "epoch": 10.591073582629674,
-      "grad_norm": 0.330078125,
-      "learning_rate": 4.792158844941023e-05,
-      "loss": 0.8813,
-      "step": 4390
-    },
-    {
-      "epoch": 10.60313630880579,
-      "grad_norm": 0.328125,
-      "learning_rate": 4.768186411718417e-05,
-      "loss": 0.8865,
-      "step": 4395
-    },
-    {
-      "epoch": 10.615199034981906,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.7442553045090474e-05,
-      "loss": 0.8789,
-      "step": 4400
-    },
-    {
-      "epoch": 10.627261761158021,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.720365712344349e-05,
-      "loss": 0.8774,
-      "step": 4405
-    },
-    {
-      "epoch": 10.639324487334138,
-      "grad_norm": 0.373046875,
-      "learning_rate": 4.696517823927842e-05,
-      "loss": 0.8797,
-      "step": 4410
-    },
-    {
-      "epoch": 10.651387213510253,
-      "grad_norm": 0.3359375,
-      "learning_rate": 4.672711827633618e-05,
-      "loss": 0.877,
-      "step": 4415
-    },
-    {
-      "epoch": 10.66344993968637,
-      "grad_norm": 0.353515625,
-      "learning_rate": 4.6489479115048676e-05,
-      "loss": 0.8766,
-      "step": 4420
-    },
-    {
-      "epoch": 10.675512665862485,
-      "grad_norm": 0.33984375,
-      "learning_rate": 4.625226263252386e-05,
-      "loss": 0.8804,
-      "step": 4425
-    },
-    {
-      "epoch": 10.687575392038601,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.601547070253107e-05,
-      "loss": 0.8889,
-      "step": 4430
-    },
-    {
-      "epoch": 10.699638118214716,
-      "grad_norm": 0.341796875,
-      "learning_rate": 4.577910519548602e-05,
-      "loss": 0.8747,
-      "step": 4435
-    },
-    {
-      "epoch": 10.711700844390833,
-      "grad_norm": 0.3515625,
-      "learning_rate": 4.554316797843609e-05,
-      "loss": 0.8816,
-      "step": 4440
-    },
-    {
-      "epoch": 10.723763570566948,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.53076609150457e-05,
-      "loss": 0.8796,
-      "step": 4445
-    },
-    {
-      "epoch": 10.735826296743063,
-      "grad_norm": 0.328125,
-      "learning_rate": 4.5072585865581395e-05,
-      "loss": 0.8836,
-      "step": 4450
-    },
-    {
-      "epoch": 10.74788902291918,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.483794468689728e-05,
-      "loss": 0.8779,
-      "step": 4455
-    },
-    {
-      "epoch": 10.759951749095295,
-      "grad_norm": 0.326171875,
-      "learning_rate": 4.460373923242041e-05,
-      "loss": 0.8766,
-      "step": 4460
-    },
-    {
-      "epoch": 10.772014475271412,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.436997135213592e-05,
-      "loss": 0.8831,
-      "step": 4465
-    },
-    {
-      "epoch": 10.784077201447527,
-      "grad_norm": 0.3046875,
-      "learning_rate": 4.413664289257265e-05,
-      "loss": 0.8775,
-      "step": 4470
-    },
-    {
-      "epoch": 10.796139927623642,
-      "grad_norm": 0.3203125,
-      "learning_rate": 4.3903755696788386e-05,
-      "loss": 0.8744,
-      "step": 4475
-    },
-    {
-      "epoch": 10.808202653799759,
-      "grad_norm": 0.306640625,
-      "learning_rate": 4.3671311604355494e-05,
-      "loss": 0.8816,
-      "step": 4480
-    },
-    {
-      "epoch": 10.820265379975874,
-      "grad_norm": 0.3359375,
-      "learning_rate": 4.343931245134616e-05,
-      "loss": 0.8757,
-      "step": 4485
-    },
-    {
-      "epoch": 10.83232810615199,
-      "grad_norm": 0.34375,
-      "learning_rate": 4.3207760070318026e-05,
-      "loss": 0.8827,
-      "step": 4490
-    },
-    {
-      "epoch": 10.844390832328106,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.2976656290299764e-05,
-      "loss": 0.8823,
-      "step": 4495
-    },
-    {
-      "epoch": 10.856453558504223,
-      "grad_norm": 0.326171875,
-      "learning_rate": 4.274600293677647e-05,
-      "loss": 0.8856,
-      "step": 4500
-    },
-    {
-      "epoch": 10.868516284680338,
-      "grad_norm": 0.326171875,
-      "learning_rate": 4.2515801831675305e-05,
-      "loss": 0.8786,
-      "step": 4505
-    },
-    {
-      "epoch": 10.880579010856454,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.228605479335127e-05,
-      "loss": 0.8808,
-      "step": 4510
-    },
-    {
-      "epoch": 10.89264173703257,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.2056763636572574e-05,
-      "loss": 0.8746,
-      "step": 4515
-    },
-    {
-      "epoch": 10.904704463208684,
-      "grad_norm": 0.330078125,
-      "learning_rate": 4.182793017250643e-05,
-      "loss": 0.8731,
-      "step": 4520
-    },
-    {
-      "epoch": 10.916767189384801,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.1599556208704785e-05,
-      "loss": 0.8791,
-      "step": 4525
-    },
-    {
-      "epoch": 10.928829915560916,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.137164354908999e-05,
-      "loss": 0.8785,
-      "step": 4530
-    },
-    {
-      "epoch": 10.940892641737033,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.114419399394052e-05,
-      "loss": 0.88,
-      "step": 4535
-    },
-    {
-      "epoch": 10.952955367913148,
-      "grad_norm": 0.32421875,
-      "learning_rate": 4.0917209339876896e-05,
-      "loss": 0.8801,
-      "step": 4540
-    },
-    {
-      "epoch": 10.965018094089265,
-      "grad_norm": 0.3359375,
-      "learning_rate": 4.069069137984731e-05,
-      "loss": 0.8809,
-      "step": 4545
-    },
-    {
-      "epoch": 10.97708082026538,
-      "grad_norm": 0.318359375,
-      "learning_rate": 4.046464190311355e-05,
-      "loss": 0.8911,
-      "step": 4550
-    },
-    {
-      "epoch": 10.989143546441495,
-      "grad_norm": 0.333984375,
-      "learning_rate": 4.023906269523686e-05,
-      "loss": 0.8697,
-      "step": 4555
-    },
-    {
-      "epoch": 10.998793727382388,
-      "eval_loss": 2.5013487339019775,
-      "eval_runtime": 0.6548,
-      "eval_samples_per_second": 15.273,
-      "eval_steps_per_second": 1.527,
-      "step": 4559
-    },
-    {
-      "epoch": 11.001206272617612,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.001395553806391e-05,
-      "loss": 0.8789,
-      "step": 4560
-    },
-    {
-      "epoch": 11.013268998793727,
-      "grad_norm": 0.33203125,
-      "learning_rate": 3.9789322209712566e-05,
-      "loss": 0.8779,
-      "step": 4565
-    },
-    {
-      "epoch": 11.025331724969844,
-      "grad_norm": 0.34375,
-      "learning_rate": 3.956516448455794e-05,
-      "loss": 0.8665,
-      "step": 4570
-    },
-    {
-      "epoch": 11.037394451145959,
-      "grad_norm": 0.33203125,
-      "learning_rate": 3.9341484133218366e-05,
-      "loss": 0.8736,
-      "step": 4575
-    },
-    {
-      "epoch": 11.049457177322076,
-      "grad_norm": 0.3359375,
-      "learning_rate": 3.9118282922541474e-05,
-      "loss": 0.8686,
-      "step": 4580
-    },
-    {
-      "epoch": 11.06151990349819,
-      "grad_norm": 0.314453125,
-      "learning_rate": 3.8895562615590075e-05,
-      "loss": 0.8758,
-      "step": 4585
-    },
-    {
-      "epoch": 11.073582629674306,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.8673324971628357e-05,
-      "loss": 0.8722,
-      "step": 4590
-    },
-    {
-      "epoch": 11.085645355850422,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.8451571746108027e-05,
-      "loss": 0.8769,
-      "step": 4595
-    },
-    {
-      "epoch": 11.097708082026537,
-      "grad_norm": 0.326171875,
-      "learning_rate": 3.8230304690654304e-05,
-      "loss": 0.8696,
-      "step": 4600
-    },
-    {
-      "epoch": 11.109770808202654,
-      "grad_norm": 0.328125,
-      "learning_rate": 3.800952555305216e-05,
-      "loss": 0.8767,
-      "step": 4605
-    },
-    {
-      "epoch": 11.12183353437877,
-      "grad_norm": 0.337890625,
-      "learning_rate": 3.778923607723252e-05,
-      "loss": 0.8717,
-      "step": 4610
-    },
-    {
-      "epoch": 11.133896260554886,
-      "grad_norm": 0.341796875,
-      "learning_rate": 3.756943800325852e-05,
-      "loss": 0.8716,
-      "step": 4615
-    },
-    {
-      "epoch": 11.145958986731001,
-      "grad_norm": 0.345703125,
-      "learning_rate": 3.7350133067311686e-05,
-      "loss": 0.8779,
-      "step": 4620
-    },
-    {
-      "epoch": 11.158021712907116,
-      "grad_norm": 0.34375,
-      "learning_rate": 3.7131323001678254e-05,
-      "loss": 0.8674,
-      "step": 4625
-    },
-    {
-      "epoch": 11.170084439083233,
-      "grad_norm": 0.326171875,
-      "learning_rate": 3.691300953473548e-05,
-      "loss": 0.8784,
-      "step": 4630
-    },
-    {
-      "epoch": 11.182147165259348,
-      "grad_norm": 0.328125,
-      "learning_rate": 3.669519439093801e-05,
-      "loss": 0.8854,
-      "step": 4635
-    },
-    {
-      "epoch": 11.194209891435465,
-      "grad_norm": 0.314453125,
-      "learning_rate": 3.6477879290804206e-05,
-      "loss": 0.8706,
-      "step": 4640
-    },
-    {
-      "epoch": 11.20627261761158,
-      "grad_norm": 0.326171875,
-      "learning_rate": 3.626106595090268e-05,
-      "loss": 0.8776,
-      "step": 4645
-    },
-    {
-      "epoch": 11.218335343787697,
-      "grad_norm": 0.357421875,
-      "learning_rate": 3.604475608383858e-05,
-      "loss": 0.8725,
-      "step": 4650
-    },
-    {
-      "epoch": 11.230398069963812,
-      "grad_norm": 0.314453125,
-      "learning_rate": 3.582895139824014e-05,
-      "loss": 0.8751,
-      "step": 4655
-    },
-    {
-      "epoch": 11.242460796139927,
-      "grad_norm": 0.33203125,
-      "learning_rate": 3.561365359874513e-05,
-      "loss": 0.8789,
-      "step": 4660
-    },
-    {
-      "epoch": 11.254523522316044,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.539886438598756e-05,
-      "loss": 0.8773,
-      "step": 4665
-    },
-    {
-      "epoch": 11.266586248492159,
-      "grad_norm": 0.314453125,
-      "learning_rate": 3.518458545658401e-05,
-      "loss": 0.877,
-      "step": 4670
-    },
-    {
-      "epoch": 11.278648974668275,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.497081850312033e-05,
-      "loss": 0.8754,
-      "step": 4675
-    },
-    {
-      "epoch": 11.29071170084439,
-      "grad_norm": 0.322265625,
-      "learning_rate": 3.475756521413839e-05,
-      "loss": 0.8763,
-      "step": 4680
-    },
-    {
-      "epoch": 11.302774427020507,
-      "grad_norm": 0.31640625,
-      "learning_rate": 3.454482727412254e-05,
-      "loss": 0.8778,
-      "step": 4685
-    },
-    {
-      "epoch": 11.314837153196622,
-      "grad_norm": 0.330078125,
-      "learning_rate": 3.433260636348639e-05,
-      "loss": 0.8784,
-      "step": 4690
-    },
-    {
-      "epoch": 11.32689987937274,
-      "grad_norm": 0.3125,
-      "learning_rate": 3.412090415855963e-05,
-      "loss": 0.8745,
-      "step": 4695
-    },
-    {
-      "epoch": 11.338962605548854,
-      "grad_norm": 0.32421875,
-      "learning_rate": 3.3909722331574623e-05,
-      "loss": 0.8756,
-      "step": 4700
-    },
-    {
-      "epoch": 11.35102533172497,
-      "grad_norm": 0.318359375,
-      "learning_rate": 3.369906255065328e-05,
-      "loss": 0.8827,
-      "step": 4705
-    },
-    {
-      "epoch": 11.363088057901086,
-      "grad_norm": 0.328125,
-      "learning_rate": 3.348892647979389e-05,
-      "loss": 0.8763,
-      "step": 4710
-    },
-    {
-      "epoch": 11.375150784077201,
-      "grad_norm": 0.31640625,
-      "learning_rate": 3.3279315778858036e-05,
-      "loss": 0.8815,
-      "step": 4715
-    },
-    {
-      "epoch": 11.387213510253318,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.307023210355729e-05,
-      "loss": 0.8787,
-      "step": 4720
-    },
-    {
-      "epoch": 11.399276236429433,
-      "grad_norm": 0.318359375,
-      "learning_rate": 3.2861677105440336e-05,
-      "loss": 0.8794,
-      "step": 4725
-    },
-    {
-      "epoch": 11.41133896260555,
-      "grad_norm": 0.314453125,
-      "learning_rate": 3.265365243187981e-05,
-      "loss": 0.8753,
-      "step": 4730
-    },
-    {
-      "epoch": 11.423401688781665,
-      "grad_norm": 0.349609375,
-      "learning_rate": 3.2446159726059345e-05,
-      "loss": 0.8819,
-      "step": 4735
-    },
-    {
-      "epoch": 11.43546441495778,
-      "grad_norm": 0.318359375,
-      "learning_rate": 3.223920062696052e-05,
-      "loss": 0.8727,
-      "step": 4740
-    },
-    {
-      "epoch": 11.447527141133897,
-      "grad_norm": 0.322265625,
-      "learning_rate": 3.203277676935007e-05,
-      "loss": 0.8743,
-      "step": 4745
-    },
-    {
-      "epoch": 11.459589867310012,
-      "grad_norm": 0.314453125,
-      "learning_rate": 3.182688978376678e-05,
-      "loss": 0.8777,
-      "step": 4750
-    },
-    {
-      "epoch": 11.471652593486128,
-      "grad_norm": 0.31640625,
-      "learning_rate": 3.1621541296508695e-05,
-      "loss": 0.8759,
-      "step": 4755
-    },
-    {
-      "epoch": 11.483715319662243,
-      "grad_norm": 0.318359375,
-      "learning_rate": 3.141673292962026e-05,
-      "loss": 0.8693,
-      "step": 4760
-    },
-    {
-      "epoch": 11.49577804583836,
-      "grad_norm": 0.337890625,
-      "learning_rate": 3.121246630087961e-05,
-      "loss": 0.8747,
-      "step": 4765
-    },
-    {
-      "epoch": 11.507840772014475,
-      "grad_norm": 0.314453125,
-      "learning_rate": 3.100874302378559e-05,
-      "loss": 0.8789,
-      "step": 4770
-    },
-    {
-      "epoch": 11.51990349819059,
-      "grad_norm": 0.3203125,
-      "learning_rate": 3.080556470754513e-05,
-      "loss": 0.8858,
-      "step": 4775
-    },
-    {
-      "epoch": 11.531966224366707,
-      "grad_norm": 0.318359375,
-      "learning_rate": 3.0602932957060635e-05,
-      "loss": 0.8751,
-      "step": 4780
-    },
-    {
-      "epoch": 11.544028950542822,
-      "grad_norm": 0.337890625,
-      "learning_rate": 3.0400849372917073e-05,
-      "loss": 0.8828,
-      "step": 4785
-    },
-    {
-      "epoch": 11.556091676718939,
-      "grad_norm": 0.302734375,
-      "learning_rate": 3.019931555136949e-05,
-      "loss": 0.8751,
-      "step": 4790
-    },
-    {
-      "epoch": 11.568154402895054,
-      "grad_norm": 0.31640625,
-      "learning_rate": 2.9998333084330333e-05,
-      "loss": 0.8835,
-      "step": 4795
-    },
-    {
-      "epoch": 11.58021712907117,
-      "grad_norm": 0.322265625,
-      "learning_rate": 2.979790355935703e-05,
-      "loss": 0.8705,
-      "step": 4800
-    },
-    {
-      "epoch": 11.592279855247286,
-      "grad_norm": 0.322265625,
-      "learning_rate": 2.9598028559639212e-05,
-      "loss": 0.8677,
-      "step": 4805
-    },
-    {
-      "epoch": 11.604342581423401,
-      "grad_norm": 0.318359375,
-      "learning_rate": 2.9398709663986324e-05,
-      "loss": 0.8772,
-      "step": 4810
-    },
-    {
-      "epoch": 11.616405307599518,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.919994844681524e-05,
-      "loss": 0.8788,
-      "step": 4815
-    },
-    {
-      "epoch": 11.628468033775633,
-      "grad_norm": 0.322265625,
-      "learning_rate": 2.900174647813767e-05,
-      "loss": 0.8767,
-      "step": 4820
-    },
-    {
-      "epoch": 11.64053075995175,
-      "grad_norm": 0.314453125,
-      "learning_rate": 2.8804105323547837e-05,
-      "loss": 0.8813,
-      "step": 4825
-    },
-    {
-      "epoch": 11.652593486127865,
-      "grad_norm": 0.318359375,
-      "learning_rate": 2.8607026544210114e-05,
-      "loss": 0.8797,
-      "step": 4830
-    },
-    {
-      "epoch": 11.664656212303981,
-      "grad_norm": 0.330078125,
-      "learning_rate": 2.8410511696846676e-05,
-      "loss": 0.8802,
-      "step": 4835
-    },
-    {
-      "epoch": 11.676718938480096,
-      "grad_norm": 0.328125,
-      "learning_rate": 2.8214562333725227e-05,
-      "loss": 0.8825,
-      "step": 4840
-    },
-    {
-      "epoch": 11.688781664656211,
-      "grad_norm": 0.328125,
-      "learning_rate": 2.801918000264665e-05,
-      "loss": 0.8741,
-      "step": 4845
-    },
-    {
-      "epoch": 11.700844390832328,
-      "grad_norm": 0.314453125,
-      "learning_rate": 2.7824366246932985e-05,
-      "loss": 0.8843,
-      "step": 4850
-    },
-    {
-      "epoch": 11.712907117008443,
-      "grad_norm": 0.314453125,
-      "learning_rate": 2.7630122605414988e-05,
-      "loss": 0.8783,
-      "step": 4855
-    },
-    {
-      "epoch": 11.72496984318456,
-      "grad_norm": 0.32421875,
-      "learning_rate": 2.7436450612420095e-05,
-      "loss": 0.8717,
-      "step": 4860
-    },
-    {
-      "epoch": 11.737032569360675,
-      "grad_norm": 0.33203125,
-      "learning_rate": 2.7243351797760397e-05,
-      "loss": 0.8807,
-      "step": 4865
-    },
-    {
-      "epoch": 11.749095295536792,
-      "grad_norm": 0.3125,
-      "learning_rate": 2.705082768672036e-05,
-      "loss": 0.8715,
-      "step": 4870
-    },
-    {
-      "epoch": 11.761158021712907,
-      "grad_norm": 0.349609375,
-      "learning_rate": 2.6858879800044866e-05,
-      "loss": 0.8765,
-      "step": 4875
-    },
-    {
-      "epoch": 11.773220747889024,
-      "grad_norm": 0.31640625,
-      "learning_rate": 2.6667509653927334e-05,
-      "loss": 0.8744,
-      "step": 4880
-    },
-    {
-      "epoch": 11.785283474065139,
-      "grad_norm": 0.3125,
-      "learning_rate": 2.64767187599975e-05,
-      "loss": 0.8793,
-      "step": 4885
-    },
-    {
-      "epoch": 11.797346200241254,
-      "grad_norm": 0.31640625,
-      "learning_rate": 2.6286508625309624e-05,
-      "loss": 0.8762,
-      "step": 4890
-    },
-    {
-      "epoch": 11.80940892641737,
-      "grad_norm": 0.322265625,
-      "learning_rate": 2.6096880752330553e-05,
-      "loss": 0.8839,
-      "step": 4895
-    },
-    {
-      "epoch": 11.821471652593486,
-      "grad_norm": 0.33984375,
-      "learning_rate": 2.5907836638927928e-05,
-      "loss": 0.8828,
-      "step": 4900
-    },
-    {
-      "epoch": 11.833534378769603,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.57193777783582e-05,
-      "loss": 0.8782,
-      "step": 4905
-    },
-    {
-      "epoch": 11.845597104945718,
-      "grad_norm": 0.322265625,
-      "learning_rate": 2.553150565925493e-05,
-      "loss": 0.8708,
-      "step": 4910
-    },
-    {
-      "epoch": 11.857659831121833,
-      "grad_norm": 0.318359375,
-      "learning_rate": 2.534422176561705e-05,
-      "loss": 0.879,
-      "step": 4915
-    },
-    {
-      "epoch": 11.86972255729795,
-      "grad_norm": 0.314453125,
-      "learning_rate": 2.515752757679707e-05,
-      "loss": 0.8808,
-      "step": 4920
-    },
-    {
-      "epoch": 11.881785283474064,
-      "grad_norm": 0.353515625,
-      "learning_rate": 2.4971424567489434e-05,
-      "loss": 0.8777,
-      "step": 4925
-    },
-    {
-      "epoch": 11.893848009650181,
-      "grad_norm": 0.318359375,
-      "learning_rate": 2.478591420771894e-05,
-      "loss": 0.875,
-      "step": 4930
-    },
-    {
-      "epoch": 11.905910735826296,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.4600997962828987e-05,
-      "loss": 0.8767,
-      "step": 4935
-    },
-    {
-      "epoch": 11.917973462002413,
-      "grad_norm": 0.357421875,
-      "learning_rate": 2.441667729347006e-05,
-      "loss": 0.8761,
-      "step": 4940
-    },
-    {
-      "epoch": 11.930036188178528,
-      "grad_norm": 0.33984375,
-      "learning_rate": 2.423295365558821e-05,
-      "loss": 0.8821,
-      "step": 4945
-    },
-    {
-      "epoch": 11.942098914354645,
-      "grad_norm": 0.3125,
-      "learning_rate": 2.404982850041363e-05,
-      "loss": 0.8829,
-      "step": 4950
-    },
-    {
-      "epoch": 11.95416164053076,
-      "grad_norm": 0.3203125,
-      "learning_rate": 2.3867303274448994e-05,
-      "loss": 0.8794,
-      "step": 4955
-    },
-    {
-      "epoch": 11.966224366706875,
-      "grad_norm": 0.3203125,
-      "learning_rate": 2.3685379419458164e-05,
-      "loss": 0.882,
-      "step": 4960
-    },
-    {
-      "epoch": 11.978287092882992,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.3504058372454884e-05,
-      "loss": 0.8693,
-      "step": 4965
-    },
-    {
-      "epoch": 11.990349819059107,
-      "grad_norm": 0.337890625,
-      "learning_rate": 2.3323341565691204e-05,
-      "loss": 0.8729,
-      "step": 4970
-    },
-    {
-      "epoch": 12.0,
-      "eval_loss": 2.5031228065490723,
-      "eval_runtime": 0.5349,
-      "eval_samples_per_second": 18.696,
-      "eval_steps_per_second": 1.87,
-      "step": 4974
-    },
-    {
-      "epoch": 12.002412545235224,
-      "grad_norm": 0.310546875,
-      "learning_rate": 2.3143230426646312e-05,
-      "loss": 0.873,
-      "step": 4975
-    },
-    {
-      "epoch": 12.014475271411339,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.2963726378015327e-05,
-      "loss": 0.8753,
-      "step": 4980
-    },
-    {
-      "epoch": 12.026537997587456,
-      "grad_norm": 0.32421875,
-      "learning_rate": 2.278483083769787e-05,
-      "loss": 0.8806,
-      "step": 4985
-    },
-    {
-      "epoch": 12.03860072376357,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.2606545218787012e-05,
-      "loss": 0.8825,
-      "step": 4990
-    },
-    {
-      "epoch": 12.050663449939686,
-      "grad_norm": 0.31640625,
-      "learning_rate": 2.242887092955801e-05,
-      "loss": 0.8671,
-      "step": 4995
-    },
-    {
-      "epoch": 12.062726176115802,
-      "grad_norm": 0.298828125,
-      "learning_rate": 2.2251809373457345e-05,
-      "loss": 0.8707,
-      "step": 5000
-    },
-    {
-      "epoch": 12.074788902291917,
-      "grad_norm": 0.314453125,
-      "learning_rate": 2.207536194909142e-05,
-      "loss": 0.8738,
-      "step": 5005
-    },
-    {
-      "epoch": 12.086851628468034,
-      "grad_norm": 0.318359375,
-      "learning_rate": 2.189953005021569e-05,
-      "loss": 0.8725,
-      "step": 5010
-    },
-    {
-      "epoch": 12.09891435464415,
-      "grad_norm": 0.314453125,
-      "learning_rate": 2.1724315065723532e-05,
-      "loss": 0.8718,
-      "step": 5015
-    },
-    {
-      "epoch": 12.110977080820266,
-      "grad_norm": 0.314453125,
-      "learning_rate": 2.1549718379635377e-05,
-      "loss": 0.8764,
-      "step": 5020
-    },
-    {
-      "epoch": 12.123039806996381,
-      "grad_norm": 0.32421875,
-      "learning_rate": 2.1375741371087677e-05,
-      "loss": 0.8717,
-      "step": 5025
-    },
-    {
-      "epoch": 12.135102533172496,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.1202385414322078e-05,
-      "loss": 0.8799,
-      "step": 5030
-    },
-    {
-      "epoch": 12.147165259348613,
-      "grad_norm": 0.318359375,
-      "learning_rate": 2.1029651878674604e-05,
-      "loss": 0.8683,
-      "step": 5035
-    },
-    {
-      "epoch": 12.159227985524728,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.085754212856471e-05,
-      "loss": 0.8839,
-      "step": 5040
-    },
-    {
-      "epoch": 12.171290711700845,
-      "grad_norm": 0.314453125,
-      "learning_rate": 2.068605752348457e-05,
-      "loss": 0.8786,
-      "step": 5045
-    },
-    {
-      "epoch": 12.18335343787696,
-      "grad_norm": 0.318359375,
-      "learning_rate": 2.0515199417988452e-05,
-      "loss": 0.8729,
-      "step": 5050
-    },
-    {
-      "epoch": 12.195416164053077,
-      "grad_norm": 0.3203125,
-      "learning_rate": 2.0344969161681792e-05,
-      "loss": 0.8784,
-      "step": 5055
-    },
-    {
-      "epoch": 12.207478890229192,
-      "grad_norm": 0.310546875,
-      "learning_rate": 2.01753680992107e-05,
-      "loss": 0.8831,
-      "step": 5060
-    },
-    {
-      "epoch": 12.219541616405307,
-      "grad_norm": 0.3203125,
-      "learning_rate": 2.0006397570251356e-05,
-      "loss": 0.872,
-      "step": 5065
-    },
-    {
-      "epoch": 12.231604342581424,
-      "grad_norm": 0.318359375,
-      "learning_rate": 1.983805890949927e-05,
-      "loss": 0.8756,
-      "step": 5070
-    },
-    {
-      "epoch": 12.243667068757539,
-      "grad_norm": 0.32421875,
-      "learning_rate": 1.9670353446658873e-05,
-      "loss": 0.8795,
-      "step": 5075
-    },
-    {
-      "epoch": 12.255729794933655,
-      "grad_norm": 0.30078125,
-      "learning_rate": 1.9503282506432974e-05,
-      "loss": 0.8744,
-      "step": 5080
-    },
-    {
-      "epoch": 12.26779252110977,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.9336847408512328e-05,
-      "loss": 0.8728,
-      "step": 5085
-    },
-    {
-      "epoch": 12.279855247285887,
-      "grad_norm": 0.318359375,
-      "learning_rate": 1.917104946756515e-05,
-      "loss": 0.867,
-      "step": 5090
-    },
-    {
-      "epoch": 12.291917973462002,
-      "grad_norm": 0.318359375,
-      "learning_rate": 1.9005889993226735e-05,
-      "loss": 0.8756,
-      "step": 5095
-    },
-    {
-      "epoch": 12.303980699638117,
-      "grad_norm": 0.314453125,
-      "learning_rate": 1.884137029008921e-05,
-      "loss": 0.8761,
-      "step": 5100
-    },
-    {
-      "epoch": 12.316043425814234,
-      "grad_norm": 0.31640625,
-      "learning_rate": 1.8677491657691116e-05,
-      "loss": 0.8738,
-      "step": 5105
-    },
-    {
-      "epoch": 12.32810615199035,
-      "grad_norm": 0.322265625,
-      "learning_rate": 1.8514255390507183e-05,
-      "loss": 0.8793,
-      "step": 5110
-    },
-    {
-      "epoch": 12.340168878166466,
-      "grad_norm": 0.30859375,
-      "learning_rate": 1.8351662777938127e-05,
-      "loss": 0.8741,
-      "step": 5115
-    },
-    {
-      "epoch": 12.352231604342581,
-      "grad_norm": 0.322265625,
-      "learning_rate": 1.818971510430042e-05,
-      "loss": 0.876,
-      "step": 5120
-    },
-    {
-      "epoch": 12.364294330518698,
-      "grad_norm": 0.330078125,
-      "learning_rate": 1.802841364881621e-05,
-      "loss": 0.87,
-      "step": 5125
-    },
-    {
-      "epoch": 12.376357056694813,
-      "grad_norm": 0.318359375,
-      "learning_rate": 1.7867759685603114e-05,
-      "loss": 0.8818,
-      "step": 5130
-    },
-    {
-      "epoch": 12.38841978287093,
-      "grad_norm": 0.318359375,
-      "learning_rate": 1.7707754483664308e-05,
-      "loss": 0.8782,
-      "step": 5135
-    },
-    {
-      "epoch": 12.400482509047045,
-      "grad_norm": 0.3125,
-      "learning_rate": 1.754839930687836e-05,
-      "loss": 0.8743,
-      "step": 5140
-    },
-    {
-      "epoch": 12.41254523522316,
-      "grad_norm": 0.30859375,
-      "learning_rate": 1.738969541398926e-05,
-      "loss": 0.8783,
-      "step": 5145
-    },
-    {
-      "epoch": 12.424607961399277,
-      "grad_norm": 0.3203125,
-      "learning_rate": 1.723164405859663e-05,
-      "loss": 0.872,
-      "step": 5150
-    },
-    {
-      "epoch": 12.436670687575392,
-      "grad_norm": 0.341796875,
-      "learning_rate": 1.7074246489145595e-05,
-      "loss": 0.8739,
-      "step": 5155
-    },
-    {
-      "epoch": 12.448733413751508,
-      "grad_norm": 0.3125,
-      "learning_rate": 1.691750394891707e-05,
-      "loss": 0.8779,
-      "step": 5160
-    },
-    {
-      "epoch": 12.460796139927623,
-      "grad_norm": 0.310546875,
-      "learning_rate": 1.6761417676017956e-05,
-      "loss": 0.8758,
-      "step": 5165
-    },
-    {
-      "epoch": 12.47285886610374,
-      "grad_norm": 0.30859375,
-      "learning_rate": 1.660598890337124e-05,
-      "loss": 0.881,
-      "step": 5170
-    },
-    {
-      "epoch": 12.484921592279855,
-      "grad_norm": 0.318359375,
-      "learning_rate": 1.6451218858706374e-05,
-      "loss": 0.8806,
-      "step": 5175
-    },
-    {
-      "epoch": 12.49698431845597,
-      "grad_norm": 0.3125,
-      "learning_rate": 1.629710876454945e-05,
-      "loss": 0.8771,
-      "step": 5180
-    },
-    {
-      "epoch": 12.509047044632087,
-      "grad_norm": 0.322265625,
-      "learning_rate": 1.6143659838213752e-05,
-      "loss": 0.8784,
-      "step": 5185
-    },
-    {
-      "epoch": 12.521109770808202,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.59908732917899e-05,
-      "loss": 0.863,
-      "step": 5190
-    },
-    {
-      "epoch": 12.533172496984319,
-      "grad_norm": 0.310546875,
-      "learning_rate": 1.583875033213642e-05,
-      "loss": 0.87,
-      "step": 5195
-    },
-    {
-      "epoch": 12.545235223160434,
-      "grad_norm": 0.3203125,
-      "learning_rate": 1.568729216087017e-05,
-      "loss": 0.8741,
-      "step": 5200
-    },
-    {
-      "epoch": 12.55729794933655,
-      "grad_norm": 0.33203125,
-      "learning_rate": 1.5536499974356866e-05,
-      "loss": 0.8759,
-      "step": 5205
-    },
-    {
-      "epoch": 12.569360675512666,
-      "grad_norm": 0.310546875,
-      "learning_rate": 1.5386374963701633e-05,
-      "loss": 0.8841,
-      "step": 5210
-    },
-    {
-      "epoch": 12.581423401688781,
-      "grad_norm": 0.3203125,
-      "learning_rate": 1.5236918314739568e-05,
-      "loss": 0.8843,
-      "step": 5215
-    },
-    {
-      "epoch": 12.593486127864898,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.5088131208026367e-05,
-      "loss": 0.8694,
-      "step": 5220
-    },
-    {
-      "epoch": 12.605548854041013,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.4940014818829074e-05,
-      "loss": 0.8738,
-      "step": 5225
-    },
-    {
-      "epoch": 12.61761158021713,
-      "grad_norm": 0.322265625,
-      "learning_rate": 1.4792570317116671e-05,
-      "loss": 0.8694,
-      "step": 5230
-    },
-    {
-      "epoch": 12.629674306393245,
-      "grad_norm": 0.31640625,
-      "learning_rate": 1.4645798867551008e-05,
-      "loss": 0.8774,
-      "step": 5235
-    },
-    {
-      "epoch": 12.641737032569361,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.4499701629477446e-05,
-      "loss": 0.8805,
-      "step": 5240
-    },
-    {
-      "epoch": 12.653799758745476,
-      "grad_norm": 0.31640625,
-      "learning_rate": 1.4354279756915735e-05,
-      "loss": 0.8815,
-      "step": 5245
-    },
-    {
-      "epoch": 12.665862484921592,
-      "grad_norm": 0.310546875,
-      "learning_rate": 1.4209534398551016e-05,
-      "loss": 0.875,
-      "step": 5250
-    },
-    {
-      "epoch": 12.677925211097708,
-      "grad_norm": 0.31640625,
-      "learning_rate": 1.406546669772456e-05,
-      "loss": 0.8757,
-      "step": 5255
-    },
-    {
-      "epoch": 12.689987937273823,
-      "grad_norm": 0.333984375,
-      "learning_rate": 1.3922077792424881e-05,
-      "loss": 0.8773,
-      "step": 5260
-    },
-    {
-      "epoch": 12.70205066344994,
-      "grad_norm": 0.32421875,
-      "learning_rate": 1.3779368815278647e-05,
-      "loss": 0.873,
-      "step": 5265
-    },
-    {
-      "epoch": 12.714113389626055,
-      "grad_norm": 0.31640625,
-      "learning_rate": 1.3637340893541895e-05,
-      "loss": 0.8811,
-      "step": 5270
-    },
-    {
-      "epoch": 12.726176115802172,
-      "grad_norm": 0.314453125,
-      "learning_rate": 1.3495995149090911e-05,
-      "loss": 0.8726,
-      "step": 5275
-    },
-    {
-      "epoch": 12.738238841978287,
-      "grad_norm": 0.31640625,
-      "learning_rate": 1.335533269841347e-05,
-      "loss": 0.8703,
-      "step": 5280
-    },
-    {
-      "epoch": 12.750301568154402,
-      "grad_norm": 0.314453125,
-      "learning_rate": 1.321535465260012e-05,
-      "loss": 0.8764,
-      "step": 5285
-    },
-    {
-      "epoch": 12.762364294330519,
-      "grad_norm": 0.33984375,
-      "learning_rate": 1.307606211733522e-05,
-      "loss": 0.8762,
-      "step": 5290
-    },
-    {
-      "epoch": 12.774427020506634,
-      "grad_norm": 0.318359375,
-      "learning_rate": 1.2937456192888309e-05,
-      "loss": 0.8707,
-      "step": 5295
-    },
-    {
-      "epoch": 12.78648974668275,
-      "grad_norm": 0.337890625,
-      "learning_rate": 1.2799537974105402e-05,
-      "loss": 0.8822,
-      "step": 5300
-    },
-    {
-      "epoch": 12.798552472858866,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.2662308550400336e-05,
-      "loss": 0.8761,
-      "step": 5305
-    },
-    {
-      "epoch": 12.810615199034983,
-      "grad_norm": 0.365234375,
-      "learning_rate": 1.252576900574618e-05,
-      "loss": 0.8743,
-      "step": 5310
-    },
-    {
-      "epoch": 12.822677925211098,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.2389920418666633e-05,
-      "loss": 0.8689,
-      "step": 5315
-    },
-    {
-      "epoch": 12.834740651387214,
-      "grad_norm": 0.322265625,
-      "learning_rate": 1.225476386222757e-05,
-      "loss": 0.8698,
-      "step": 5320
-    },
-    {
-      "epoch": 12.84680337756333,
-      "grad_norm": 0.328125,
-      "learning_rate": 1.2120300404028507e-05,
-      "loss": 0.8802,
-      "step": 5325
-    },
-    {
-      "epoch": 12.858866103739444,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.1986531106194143e-05,
-      "loss": 0.8739,
-      "step": 5330
-    },
-    {
-      "epoch": 12.870928829915561,
-      "grad_norm": 0.322265625,
-      "learning_rate": 1.1853457025366111e-05,
-      "loss": 0.8814,
-      "step": 5335
-    },
-    {
-      "epoch": 12.882991556091676,
-      "grad_norm": 0.333984375,
-      "learning_rate": 1.1721079212694452e-05,
-      "loss": 0.8802,
-      "step": 5340
-    },
-    {
-      "epoch": 12.895054282267793,
-      "grad_norm": 0.33984375,
-      "learning_rate": 1.1589398713829424e-05,
-      "loss": 0.8746,
-      "step": 5345
-    },
-    {
-      "epoch": 12.907117008443908,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.1458416568913232e-05,
-      "loss": 0.8724,
-      "step": 5350
-    },
-    {
-      "epoch": 12.919179734620023,
-      "grad_norm": 0.3203125,
-      "learning_rate": 1.1328133812571784e-05,
-      "loss": 0.8836,
-      "step": 5355
-    },
-    {
-      "epoch": 12.93124246079614,
-      "grad_norm": 0.330078125,
-      "learning_rate": 1.1198551473906493e-05,
-      "loss": 0.8663,
-      "step": 5360
-    },
-    {
-      "epoch": 12.943305186972255,
-      "grad_norm": 0.3203125,
-      "learning_rate": 1.1069670576486224e-05,
-      "loss": 0.8722,
-      "step": 5365
-    },
-    {
-      "epoch": 12.955367913148372,
-      "grad_norm": 0.32421875,
-      "learning_rate": 1.0941492138339183e-05,
-      "loss": 0.8718,
-      "step": 5370
-    },
-    {
-      "epoch": 12.967430639324487,
-      "grad_norm": 0.3125,
-      "learning_rate": 1.0814017171944812e-05,
-      "loss": 0.875,
-      "step": 5375
-    },
-    {
-      "epoch": 12.979493365500604,
-      "grad_norm": 0.33984375,
-      "learning_rate": 1.068724668422586e-05,
-      "loss": 0.8768,
-      "step": 5380
-    },
-    {
-      "epoch": 12.991556091676719,
-      "grad_norm": 0.314453125,
-      "learning_rate": 1.0561181676540444e-05,
-      "loss": 0.8779,
-      "step": 5385
-    },
-    {
-      "epoch": 12.998793727382388,
-      "eval_loss": 2.5023179054260254,
-      "eval_runtime": 0.6415,
-      "eval_samples_per_second": 15.589,
-      "eval_steps_per_second": 1.559,
-      "step": 5388
-    },
-    {
-      "epoch": 13.003618817852836,
-      "grad_norm": 0.310546875,
-      "learning_rate": 1.0435823144674051e-05,
-      "loss": 0.8808,
-      "step": 5390
-    },
-    {
-      "epoch": 13.01568154402895,
-      "grad_norm": 0.30859375,
-      "learning_rate": 1.0311172078831766e-05,
-      "loss": 0.882,
-      "step": 5395
-    },
-    {
-      "epoch": 13.027744270205066,
-      "grad_norm": 0.3125,
-      "learning_rate": 1.01872294636304e-05,
-      "loss": 0.8872,
-      "step": 5400
-    },
-    {
-      "epoch": 13.039806996381182,
-      "grad_norm": 0.314453125,
-      "learning_rate": 1.0063996278090704e-05,
-      "loss": 0.8692,
-      "step": 5405
-    },
-    {
-      "epoch": 13.051869722557297,
-      "grad_norm": 0.333984375,
-      "learning_rate": 9.941473495629683e-06,
-      "loss": 0.8738,
-      "step": 5410
-    },
-    {
-      "epoch": 13.063932448733414,
-      "grad_norm": 0.30859375,
-      "learning_rate": 9.81966208405285e-06,
-      "loss": 0.8858,
-      "step": 5415
-    },
-    {
-      "epoch": 13.07599517490953,
-      "grad_norm": 0.310546875,
-      "learning_rate": 9.698563005546668e-06,
-      "loss": 0.878,
-      "step": 5420
-    },
-    {
-      "epoch": 13.088057901085646,
-      "grad_norm": 0.306640625,
-      "learning_rate": 9.578177216670824e-06,
-      "loss": 0.8625,
-      "step": 5425
-    },
-    {
-      "epoch": 13.100120627261761,
-      "grad_norm": 0.32421875,
-      "learning_rate": 9.458505668350759e-06,
-      "loss": 0.8794,
-      "step": 5430
-    },
-    {
-      "epoch": 13.112183353437876,
-      "grad_norm": 0.3125,
-      "learning_rate": 9.339549305870187e-06,
-      "loss": 0.8684,
-      "step": 5435
-    },
-    {
-      "epoch": 13.124246079613993,
-      "grad_norm": 0.314453125,
-      "learning_rate": 9.221309068863527e-06,
-      "loss": 0.874,
-      "step": 5440
-    },
-    {
-      "epoch": 13.136308805790108,
-      "grad_norm": 0.3203125,
-      "learning_rate": 9.103785891308547e-06,
-      "loss": 0.872,
-      "step": 5445
-    },
-    {
-      "epoch": 13.148371531966225,
-      "grad_norm": 0.326171875,
-      "learning_rate": 8.986980701518966e-06,
-      "loss": 0.8779,
-      "step": 5450
-    },
-    {
-      "epoch": 13.16043425814234,
-      "grad_norm": 0.31640625,
-      "learning_rate": 8.870894422137177e-06,
-      "loss": 0.8714,
-      "step": 5455
-    },
-    {
-      "epoch": 13.172496984318457,
-      "grad_norm": 0.3125,
-      "learning_rate": 8.755527970126853e-06,
-      "loss": 0.8753,
-      "step": 5460
-    },
-    {
-      "epoch": 13.184559710494572,
-      "grad_norm": 0.326171875,
-      "learning_rate": 8.640882256765759e-06,
-      "loss": 0.8712,
-      "step": 5465
-    },
-    {
-      "epoch": 13.196622436670687,
-      "grad_norm": 0.31640625,
-      "learning_rate": 8.526958187638635e-06,
-      "loss": 0.8703,
-      "step": 5470
-    },
-    {
-      "epoch": 13.208685162846804,
-      "grad_norm": 0.33203125,
-      "learning_rate": 8.413756662629879e-06,
-      "loss": 0.874,
-      "step": 5475
-    },
-    {
-      "epoch": 13.220747889022919,
-      "grad_norm": 0.33203125,
-      "learning_rate": 8.301278575916538e-06,
-      "loss": 0.8721,
-      "step": 5480
-    },
-    {
-      "epoch": 13.232810615199035,
-      "grad_norm": 0.310546875,
-      "learning_rate": 8.189524815961291e-06,
-      "loss": 0.8761,
-      "step": 5485
-    },
-    {
-      "epoch": 13.24487334137515,
-      "grad_norm": 0.32421875,
-      "learning_rate": 8.07849626550531e-06,
-      "loss": 0.8616,
-      "step": 5490
-    },
-    {
-      "epoch": 13.256936067551267,
-      "grad_norm": 0.3125,
-      "learning_rate": 7.968193801561363e-06,
-      "loss": 0.881,
-      "step": 5495
-    },
-    {
-      "epoch": 13.268998793727382,
-      "grad_norm": 0.31640625,
-      "learning_rate": 7.858618295406895e-06,
-      "loss": 0.8724,
-      "step": 5500
-    },
-    {
-      "epoch": 13.281061519903497,
-      "grad_norm": 0.314453125,
-      "learning_rate": 7.74977061257709e-06,
-      "loss": 0.8826,
-      "step": 5505
-    },
-    {
-      "epoch": 13.293124246079614,
-      "grad_norm": 0.310546875,
-      "learning_rate": 7.641651612858081e-06,
-      "loss": 0.8864,
-      "step": 5510
-    },
-    {
-      "epoch": 13.30518697225573,
-      "grad_norm": 0.31640625,
-      "learning_rate": 7.534262150280136e-06,
-      "loss": 0.8781,
-      "step": 5515
-    },
-    {
-      "epoch": 13.317249698431846,
-      "grad_norm": 0.34375,
-      "learning_rate": 7.427603073110967e-06,
-      "loss": 0.8721,
-      "step": 5520
-    },
-    {
-      "epoch": 13.329312424607961,
-      "grad_norm": 0.310546875,
-      "learning_rate": 7.321675223848923e-06,
-      "loss": 0.8746,
-      "step": 5525
-    },
-    {
-      "epoch": 13.341375150784078,
-      "grad_norm": 0.3125,
-      "learning_rate": 7.216479439216406e-06,
-      "loss": 0.8657,
-      "step": 5530
-    },
-    {
-      "epoch": 13.353437876960193,
-      "grad_norm": 0.326171875,
-      "learning_rate": 7.1120165501533e-06,
-      "loss": 0.8668,
-      "step": 5535
-    },
-    {
-      "epoch": 13.365500603136308,
-      "grad_norm": 0.318359375,
-      "learning_rate": 7.0082873818102964e-06,
-      "loss": 0.8716,
-      "step": 5540
-    },
-    {
-      "epoch": 13.377563329312425,
-      "grad_norm": 0.326171875,
-      "learning_rate": 6.90529275354247e-06,
-      "loss": 0.8693,
-      "step": 5545
-    },
-    {
-      "epoch": 13.38962605548854,
-      "grad_norm": 0.3125,
-      "learning_rate": 6.803033478902765e-06,
-      "loss": 0.8786,
-      "step": 5550
-    },
-    {
-      "epoch": 13.401688781664657,
-      "grad_norm": 0.328125,
-      "learning_rate": 6.7015103656355885e-06,
-      "loss": 0.8846,
-      "step": 5555
-    },
-    {
-      "epoch": 13.413751507840772,
-      "grad_norm": 0.314453125,
-      "learning_rate": 6.600724215670429e-06,
-      "loss": 0.8735,
-      "step": 5560
-    },
-    {
-      "epoch": 13.425814234016888,
-      "grad_norm": 0.33203125,
-      "learning_rate": 6.500675825115454e-06,
-      "loss": 0.8701,
-      "step": 5565
-    },
-    {
-      "epoch": 13.437876960193003,
-      "grad_norm": 0.30859375,
-      "learning_rate": 6.401365984251384e-06,
-      "loss": 0.8764,
-      "step": 5570
-    },
-    {
-      "epoch": 13.44993968636912,
-      "grad_norm": 0.322265625,
-      "learning_rate": 6.302795477525058e-06,
-      "loss": 0.8824,
-      "step": 5575
-    },
-    {
-      "epoch": 13.462002412545235,
-      "grad_norm": 0.322265625,
-      "learning_rate": 6.204965083543368e-06,
-      "loss": 0.8773,
-      "step": 5580
-    },
-    {
-      "epoch": 13.47406513872135,
-      "grad_norm": 0.310546875,
-      "learning_rate": 6.107875575067057e-06,
-      "loss": 0.868,
-      "step": 5585
-    },
-    {
-      "epoch": 13.486127864897467,
-      "grad_norm": 0.31640625,
-      "learning_rate": 6.011527719004617e-06,
-      "loss": 0.8706,
-      "step": 5590
-    },
-    {
-      "epoch": 13.498190591073582,
-      "grad_norm": 0.330078125,
-      "learning_rate": 5.915922276406249e-06,
-      "loss": 0.878,
-      "step": 5595
-    },
-    {
-      "epoch": 13.510253317249699,
-      "grad_norm": 0.314453125,
-      "learning_rate": 5.8210600024578234e-06,
-      "loss": 0.8739,
-      "step": 5600
-    },
-    {
-      "epoch": 13.522316043425814,
-      "grad_norm": 0.3125,
-      "learning_rate": 5.726941646474992e-06,
-      "loss": 0.8768,
-      "step": 5605
-    },
-    {
-      "epoch": 13.534378769601929,
-      "grad_norm": 0.30859375,
-      "learning_rate": 5.633567951897145e-06,
-      "loss": 0.8785,
-      "step": 5610
-    },
-    {
-      "epoch": 13.546441495778046,
-      "grad_norm": 0.326171875,
-      "learning_rate": 5.5409396562816076e-06,
-      "loss": 0.8751,
-      "step": 5615
-    },
-    {
-      "epoch": 13.558504221954161,
-      "grad_norm": 0.30859375,
-      "learning_rate": 5.449057491297871e-06,
-      "loss": 0.8861,
-      "step": 5620
-    },
-    {
-      "epoch": 13.570566948130278,
-      "grad_norm": 0.314453125,
-      "learning_rate": 5.357922182721687e-06,
-      "loss": 0.8782,
-      "step": 5625
-    },
-    {
-      "epoch": 13.582629674306393,
-      "grad_norm": 0.30859375,
-      "learning_rate": 5.267534450429423e-06,
-      "loss": 0.8705,
-      "step": 5630
-    },
-    {
-      "epoch": 13.59469240048251,
-      "grad_norm": 0.3125,
-      "learning_rate": 5.177895008392353e-06,
-      "loss": 0.8755,
-      "step": 5635
-    },
-    {
-      "epoch": 13.606755126658625,
-      "grad_norm": 0.302734375,
-      "learning_rate": 5.08900456467103e-06,
-      "loss": 0.8744,
-      "step": 5640
-    },
-    {
-      "epoch": 13.618817852834741,
-      "grad_norm": 0.3125,
-      "learning_rate": 5.000863821409674e-06,
-      "loss": 0.8811,
-      "step": 5645
-    },
-    {
-      "epoch": 13.630880579010856,
-      "grad_norm": 0.30859375,
-      "learning_rate": 4.9134734748305986e-06,
-      "loss": 0.8718,
-      "step": 5650
-    },
-    {
-      "epoch": 13.642943305186972,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.826834215228826e-06,
-      "loss": 0.8773,
-      "step": 5655
-    },
-    {
-      "epoch": 13.655006031363088,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.740946726966466e-06,
-      "loss": 0.8811,
-      "step": 5660
-    },
-    {
-      "epoch": 13.667068757539203,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.655811688467448e-06,
-      "loss": 0.8729,
-      "step": 5665
-    },
-    {
-      "epoch": 13.67913148371532,
-      "grad_norm": 0.337890625,
-      "learning_rate": 4.5714297722121106e-06,
-      "loss": 0.8811,
-      "step": 5670
-    },
-    {
-      "epoch": 13.691194209891435,
-      "grad_norm": 0.3046875,
-      "learning_rate": 4.4878016447318704e-06,
-      "loss": 0.8661,
-      "step": 5675
-    },
-    {
-      "epoch": 13.703256936067552,
-      "grad_norm": 0.345703125,
-      "learning_rate": 4.404927966604011e-06,
-      "loss": 0.8745,
-      "step": 5680
-    },
-    {
-      "epoch": 13.715319662243667,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.322809392446392e-06,
-      "loss": 0.8771,
-      "step": 5685
-    },
-    {
-      "epoch": 13.727382388419782,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.241446570912344e-06,
-      "loss": 0.874,
-      "step": 5690
-    },
-    {
-      "epoch": 13.739445114595899,
-      "grad_norm": 0.306640625,
-      "learning_rate": 4.160840144685507e-06,
-      "loss": 0.8768,
-      "step": 5695
-    },
-    {
-      "epoch": 13.751507840772014,
-      "grad_norm": 0.31640625,
-      "learning_rate": 4.080990750474778e-06,
-      "loss": 0.8706,
-      "step": 5700
-    },
-    {
-      "epoch": 13.76357056694813,
-      "grad_norm": 0.306640625,
-      "learning_rate": 4.001899019009281e-06,
-      "loss": 0.8727,
-      "step": 5705
-    },
-    {
-      "epoch": 13.775633293124246,
-      "grad_norm": 0.322265625,
-      "learning_rate": 3.92356557503335e-06,
-      "loss": 0.8792,
-      "step": 5710
-    },
-    {
-      "epoch": 13.787696019300363,
-      "grad_norm": 0.3203125,
-      "learning_rate": 3.84599103730161e-06,
-      "loss": 0.864,
-      "step": 5715
-    },
-    {
-      "epoch": 13.799758745476478,
-      "grad_norm": 0.31640625,
-      "learning_rate": 3.7691760185741387e-06,
-      "loss": 0.8733,
-      "step": 5720
-    },
-    {
-      "epoch": 13.811821471652593,
-      "grad_norm": 0.337890625,
-      "learning_rate": 3.6931211256115537e-06,
-      "loss": 0.8753,
-      "step": 5725
-    },
-    {
-      "epoch": 13.82388419782871,
-      "grad_norm": 0.32421875,
-      "learning_rate": 3.617826959170256e-06,
-      "loss": 0.8714,
-      "step": 5730
-    },
-    {
-      "epoch": 13.835946924004825,
-      "grad_norm": 0.318359375,
-      "learning_rate": 3.543294113997664e-06,
-      "loss": 0.8741,
-      "step": 5735
-    },
-    {
-      "epoch": 13.848009650180941,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.469523178827583e-06,
-      "loss": 0.881,
-      "step": 5740
-    },
-    {
-      "epoch": 13.860072376357056,
-      "grad_norm": 0.306640625,
-      "learning_rate": 3.3965147363754555e-06,
-      "loss": 0.8778,
-      "step": 5745
-    },
-    {
-      "epoch": 13.872135102533173,
-      "grad_norm": 0.310546875,
-      "learning_rate": 3.3242693633337983e-06,
-      "loss": 0.8804,
-      "step": 5750
-    },
-    {
-      "epoch": 13.884197828709288,
-      "grad_norm": 0.314453125,
-      "learning_rate": 3.252787630367715e-06,
-      "loss": 0.882,
-      "step": 5755
-    },
-    {
-      "epoch": 13.896260554885405,
-      "grad_norm": 0.3046875,
-      "learning_rate": 3.182070102110257e-06,
-      "loss": 0.8672,
-      "step": 5760
-    },
-    {
-      "epoch": 13.90832328106152,
-      "grad_norm": 0.314453125,
-      "learning_rate": 3.1121173371580825e-06,
-      "loss": 0.8772,
-      "step": 5765
-    },
-    {
-      "epoch": 13.920386007237635,
-      "grad_norm": 0.314453125,
-      "learning_rate": 3.042929888066992e-06,
-      "loss": 0.8807,
-      "step": 5770
-    },
-    {
-      "epoch": 13.932448733413752,
-      "grad_norm": 0.322265625,
-      "learning_rate": 2.974508301347534e-06,
-      "loss": 0.8796,
-      "step": 5775
-    },
-    {
-      "epoch": 13.944511459589867,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.9068531174607615e-06,
-      "loss": 0.8712,
-      "step": 5780
-    },
-    {
-      "epoch": 13.956574185765984,
-      "grad_norm": 0.318359375,
-      "learning_rate": 2.839964870813916e-06,
-      "loss": 0.8744,
-      "step": 5785
-    },
-    {
-      "epoch": 13.968636911942099,
-      "grad_norm": 0.314453125,
-      "learning_rate": 2.7738440897561723e-06,
-      "loss": 0.8754,
-      "step": 5790
-    },
-    {
-      "epoch": 13.980699638118214,
-      "grad_norm": 0.3125,
-      "learning_rate": 2.7084912965745778e-06,
-      "loss": 0.8753,
-      "step": 5795
-    },
-    {
-      "epoch": 13.99276236429433,
-      "grad_norm": 0.314453125,
-      "learning_rate": 2.6439070074897874e-06,
-      "loss": 0.8743,
-      "step": 5800
-    },
-    {
-      "epoch": 14.0,
-      "eval_loss": 2.5032882690429688,
-      "eval_runtime": 0.5357,
-      "eval_samples_per_second": 18.667,
-      "eval_steps_per_second": 1.867,
-      "step": 5803
-    },
-    {
-      "epoch": 14.004825090470446,
-      "grad_norm": 0.31640625,
-      "learning_rate": 2.580091732652101e-06,
-      "loss": 0.8799,
-      "step": 5805
-    },
-    {
-      "epoch": 14.016887816646562,
-      "grad_norm": 0.31640625,
-      "learning_rate": 2.5170459761373664e-06,
-      "loss": 0.8781,
-      "step": 5810
-    },
-    {
-      "epoch": 14.028950542822678,
-      "grad_norm": 0.31640625,
-      "learning_rate": 2.454770235943027e-06,
-      "loss": 0.8751,
-      "step": 5815
-    },
-    {
-      "epoch": 14.041013268998794,
-      "grad_norm": 0.30859375,
-      "learning_rate": 2.3932650039841687e-06,
-      "loss": 0.887,
-      "step": 5820
-    },
-    {
-      "epoch": 14.05307599517491,
-      "grad_norm": 0.31640625,
-      "learning_rate": 2.3325307660896577e-06,
-      "loss": 0.8755,
-      "step": 5825
-    },
-    {
-      "epoch": 14.065138721351026,
-      "grad_norm": 0.318359375,
-      "learning_rate": 2.2725680019982765e-06,
-      "loss": 0.8792,
-      "step": 5830
-    },
-    {
-      "epoch": 14.077201447527141,
-      "grad_norm": 0.322265625,
-      "learning_rate": 2.213377185354959e-06,
-      "loss": 0.8719,
-      "step": 5835
-    },
-    {
-      "epoch": 14.089264173703256,
-      "grad_norm": 0.306640625,
-      "learning_rate": 2.1549587837070394e-06,
-      "loss": 0.8779,
-      "step": 5840
-    },
-    {
-      "epoch": 14.101326899879373,
-      "grad_norm": 0.314453125,
-      "learning_rate": 2.0973132585005552e-06,
-      "loss": 0.8774,
-      "step": 5845
-    },
-    {
-      "epoch": 14.113389626055488,
-      "grad_norm": 0.318359375,
-      "learning_rate": 2.0404410650765817e-06,
-      "loss": 0.8718,
-      "step": 5850
-    },
-    {
-      "epoch": 14.125452352231605,
-      "grad_norm": 0.330078125,
-      "learning_rate": 1.9843426526677033e-06,
-      "loss": 0.8853,
-      "step": 5855
-    },
-    {
-      "epoch": 14.13751507840772,
-      "grad_norm": 0.326171875,
-      "learning_rate": 1.9290184643943944e-06,
-      "loss": 0.8779,
-      "step": 5860
-    },
-    {
-      "epoch": 14.149577804583837,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.874468937261531e-06,
-      "loss": 0.8692,
-      "step": 5865
-    },
-    {
-      "epoch": 14.161640530759952,
-      "grad_norm": 0.310546875,
-      "learning_rate": 1.820694502154996e-06,
-      "loss": 0.8756,
-      "step": 5870
-    },
-    {
-      "epoch": 14.173703256936067,
-      "grad_norm": 0.296875,
-      "learning_rate": 1.7676955838381804e-06,
-      "loss": 0.8697,
-      "step": 5875
-    },
-    {
-      "epoch": 14.185765983112184,
-      "grad_norm": 0.318359375,
-      "learning_rate": 1.71547260094872e-06,
-      "loss": 0.8832,
-      "step": 5880
-    },
-    {
-      "epoch": 14.197828709288299,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.6640259659951419e-06,
-      "loss": 0.8761,
-      "step": 5885
-    },
-    {
-      "epoch": 14.209891435464415,
-      "grad_norm": 0.333984375,
-      "learning_rate": 1.61335608535359e-06,
-      "loss": 0.8785,
-      "step": 5890
-    },
-    {
-      "epoch": 14.22195416164053,
-      "grad_norm": 0.32421875,
-      "learning_rate": 1.5634633592646609e-06,
-      "loss": 0.8726,
-      "step": 5895
-    },
-    {
-      "epoch": 14.234016887816647,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.5143481818302164e-06,
-      "loss": 0.8694,
-      "step": 5900
-    },
-    {
-      "epoch": 14.246079613992762,
-      "grad_norm": 0.333984375,
-      "learning_rate": 1.4660109410102652e-06,
-      "loss": 0.8852,
-      "step": 5905
-    },
-    {
-      "epoch": 14.258142340168877,
-      "grad_norm": 0.314453125,
-      "learning_rate": 1.4184520186199202e-06,
-      "loss": 0.8746,
-      "step": 5910
-    },
-    {
-      "epoch": 14.270205066344994,
-      "grad_norm": 0.3046875,
-      "learning_rate": 1.3716717903263566e-06,
-      "loss": 0.8767,
-      "step": 5915
-    },
-    {
-      "epoch": 14.28226779252111,
-      "grad_norm": 0.3203125,
-      "learning_rate": 1.3256706256458585e-06,
-      "loss": 0.8694,
-      "step": 5920
-    },
-    {
-      "epoch": 14.294330518697226,
-      "grad_norm": 0.310546875,
-      "learning_rate": 1.2804488879408993e-06,
-      "loss": 0.8709,
-      "step": 5925
-    },
-    {
-      "epoch": 14.306393244873341,
-      "grad_norm": 0.322265625,
-      "learning_rate": 1.2360069344172887e-06,
-      "loss": 0.8722,
-      "step": 5930
-    },
-    {
-      "epoch": 14.318455971049458,
-      "grad_norm": 0.30859375,
-      "learning_rate": 1.1923451161213074e-06,
-      "loss": 0.8754,
-      "step": 5935
-    },
-    {
-      "epoch": 14.330518697225573,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.1494637779369766e-06,
-      "loss": 0.8737,
-      "step": 5940
-    },
-    {
-      "epoch": 14.342581423401688,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.107363258583305e-06,
-      "loss": 0.8713,
-      "step": 5945
-    },
-    {
-      "epoch": 14.354644149577805,
-      "grad_norm": 0.302734375,
-      "learning_rate": 1.0660438906116342e-06,
-      "loss": 0.8744,
-      "step": 5950
-    },
-    {
-      "epoch": 14.36670687575392,
-      "grad_norm": 0.328125,
+      "epoch": 9.566265060240964,
+      "grad_norm": 0.2021484375,
       "learning_rate": 1.0255060004030093e-06,
-      "loss": 0.8805,
-      "step": 5955
-    },
-    {
-      "epoch": 14.378769601930037,
-      "grad_norm": 0.3046875,
-      "learning_rate": 9.857499081655786e-07,
-      "loss": 0.8786,
-      "step": 5960
-    },
-    {
-      "epoch": 14.390832328106152,
-      "grad_norm": 0.30078125,
-      "learning_rate": 9.467759279320976e-07,
-      "loss": 0.882,
-      "step": 5965
+      "loss": 0.8251,
+      "step": 1985
     },
     {
-      "epoch": 14.402895054282268,
-      "grad_norm": 0.3046875,
+      "epoch": 9.590361445783133,
+      "grad_norm": 0.2119140625,
       "learning_rate": 9.085843675574079e-07,
-      "loss": 0.8756,
-      "step": 5970
-    },
-    {
-      "epoch": 14.414957780458383,
-      "grad_norm": 0.3125,
-      "learning_rate": 8.711755287160727e-07,
-      "loss": 0.8743,
-      "step": 5975
-    },
-    {
-      "epoch": 14.427020506634499,
-      "grad_norm": 0.310546875,
-      "learning_rate": 8.345497068998897e-07,
-      "loss": 0.8792,
-      "step": 5980
+      "loss": 0.8351,
+      "step": 1990
     },
     {
-      "epoch": 14.439083232810615,
-      "grad_norm": 0.30859375,
+      "epoch": 9.614457831325302,
+      "grad_norm": 0.201171875,
       "learning_rate": 7.987071914156596e-07,
-      "loss": 0.8664,
-      "step": 5985
-    },
-    {
-      "epoch": 14.45114595898673,
-      "grad_norm": 0.3046875,
-      "learning_rate": 7.636482653828658e-07,
-      "loss": 0.8744,
-      "step": 5990
-    },
-    {
-      "epoch": 14.463208685162847,
-      "grad_norm": 0.310546875,
-      "learning_rate": 7.293732057313874e-07,
-      "loss": 0.8674,
-      "step": 5995
+      "loss": 0.8259,
+      "step": 1995
     },
     {
-      "epoch": 14.475271411338962,
-      "grad_norm": 0.322265625,
+      "epoch": 9.638554216867469,
+      "grad_norm": 0.2099609375,
       "learning_rate": 6.958822831994005e-07,
-      "loss": 0.8744,
-      "step": 6000
-    },
-    {
-      "epoch": 14.487334137515079,
-      "grad_norm": 0.30078125,
-      "learning_rate": 6.631757623311586e-07,
-      "loss": 0.872,
-      "step": 6005
-    },
-    {
-      "epoch": 14.499396863691194,
-      "grad_norm": 0.3046875,
-      "learning_rate": 6.312539014749818e-07,
-      "loss": 0.8838,
-      "step": 6010
+      "loss": 0.8283,
+      "step": 2000
     },
     {
-      "epoch": 14.51145958986731,
-      "grad_norm": 0.326171875,
+      "epoch": 9.662650602409638,
+      "grad_norm": 0.19921875,
       "learning_rate": 6.001169527811268e-07,
-      "loss": 0.8684,
-      "step": 6015
-    },
-    {
-      "epoch": 14.523522316043426,
-      "grad_norm": 0.33203125,
-      "learning_rate": 5.697651621998535e-07,
-      "loss": 0.8776,
-      "step": 6020
-    },
-    {
-      "epoch": 14.535585042219541,
-      "grad_norm": 0.3125,
-      "learning_rate": 5.401987694794941e-07,
-      "loss": 0.8694,
-      "step": 6025
+      "loss": 0.8274,
+      "step": 2005
     },
     {
-      "epoch": 14.547647768395658,
-      "grad_norm": 0.3203125,
+      "epoch": 9.686746987951807,
+      "grad_norm": 0.20703125,
       "learning_rate": 5.114180081645214e-07,
-      "loss": 0.8698,
-      "step": 6030
-    },
-    {
-      "epoch": 14.559710494571773,
-      "grad_norm": 0.310546875,
-      "learning_rate": 4.834231055937055e-07,
-      "loss": 0.8816,
-      "step": 6035
-    },
-    {
-      "epoch": 14.57177322074789,
-      "grad_norm": 0.314453125,
-      "learning_rate": 4.5621428289834843e-07,
-      "loss": 0.8839,
-      "step": 6040
+      "loss": 0.8303,
+      "step": 2010
     },
     {
-      "epoch": 14.583835946924005,
-      "grad_norm": 0.318359375,
+      "epoch": 9.710843373493976,
+      "grad_norm": 0.212890625,
       "learning_rate": 4.2979175500050817e-07,
-      "loss": 0.8699,
-      "step": 6045
-    },
-    {
-      "epoch": 14.595898673100121,
-      "grad_norm": 0.306640625,
-      "learning_rate": 4.0415573061133307e-07,
-      "loss": 0.8657,
-      "step": 6050
-    },
-    {
-      "epoch": 14.607961399276236,
-      "grad_norm": 0.345703125,
-      "learning_rate": 3.7930641222934103e-07,
-      "loss": 0.8832,
-      "step": 6055
+      "loss": 0.8301,
+      "step": 2015
     },
     {
-      "epoch": 14.620024125452352,
-      "grad_norm": 0.30859375,
+      "epoch": 9.734939759036145,
+      "grad_norm": 0.201171875,
       "learning_rate": 3.552439961389431e-07,
-      "loss": 0.8762,
-      "step": 6060
-    },
-    {
-      "epoch": 14.632086851628468,
-      "grad_norm": 0.322265625,
-      "learning_rate": 3.3196867240876675e-07,
-      "loss": 0.8634,
-      "step": 6065
-    },
-    {
-      "epoch": 14.644149577804583,
-      "grad_norm": 0.32421875,
-      "learning_rate": 3.094806248902349e-07,
-      "loss": 0.8749,
-      "step": 6070
+      "loss": 0.829,
+      "step": 2020
     },
     {
-      "epoch": 14.6562123039807,
-      "grad_norm": 0.3046875,
+      "epoch": 9.759036144578314,
+      "grad_norm": 0.212890625,
       "learning_rate": 2.877800312160783e-07,
-      "loss": 0.8708,
-      "step": 6075
-    },
-    {
-      "epoch": 14.668275030156815,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.668670627989478e-07,
-      "loss": 0.874,
-      "step": 6080
-    },
-    {
-      "epoch": 14.680337756332932,
-      "grad_norm": 0.31640625,
-      "learning_rate": 2.4674188483003734e-07,
-      "loss": 0.881,
-      "step": 6085
+      "loss": 0.8267,
+      "step": 2025
     },
     {
-      "epoch": 14.692400482509047,
-      "grad_norm": 0.310546875,
+      "epoch": 9.783132530120483,
+      "grad_norm": 0.20703125,
       "learning_rate": 2.274046562778409e-07,
-      "loss": 0.864,
-      "step": 6090
-    },
-    {
-      "epoch": 14.704463208685162,
-      "grad_norm": 0.302734375,
-      "learning_rate": 2.088555298867978e-07,
-      "loss": 0.8669,
-      "step": 6095
-    },
-    {
-      "epoch": 14.716525934861279,
-      "grad_norm": 0.310546875,
-      "learning_rate": 1.910946521761714e-07,
-      "loss": 0.8725,
-      "step": 6100
+      "loss": 0.8285,
+      "step": 2030
     },
     {
-      "epoch": 14.728588661037394,
-      "grad_norm": 0.30859375,
+      "epoch": 9.80722891566265,
+      "grad_norm": 0.20703125,
       "learning_rate": 1.7412216343885014e-07,
-      "loss": 0.8664,
-      "step": 6105
-    },
-    {
-      "epoch": 14.74065138721351,
-      "grad_norm": 0.310546875,
-      "learning_rate": 1.5793819774027053e-07,
-      "loss": 0.875,
-      "step": 6110
-    },
-    {
-      "epoch": 14.752714113389626,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.425428829173181e-07,
-      "loss": 0.8681,
-      "step": 6115
+      "loss": 0.8311,
+      "step": 2035
     },
     {
-      "epoch": 14.764776839565743,
-      "grad_norm": 0.302734375,
+      "epoch": 9.831325301204819,
+      "grad_norm": 0.2119140625,
       "learning_rate": 1.2793634057732818e-07,
-      "loss": 0.8831,
-      "step": 6120
-    },
-    {
-      "epoch": 14.776839565741858,
-      "grad_norm": 0.310546875,
-      "learning_rate": 1.1411868609717546e-07,
-      "loss": 0.8769,
-      "step": 6125
-    },
-    {
-      "epoch": 14.788902291917973,
-      "grad_norm": 0.314453125,
-      "learning_rate": 1.010900286222638e-07,
-      "loss": 0.8717,
-      "step": 6130
+      "loss": 0.8311,
+      "step": 2040
     },
     {
-      "epoch": 14.80096501809409,
-      "grad_norm": 0.3203125,
+      "epoch": 9.855421686746988,
+      "grad_norm": 0.2109375,
       "learning_rate": 8.885047106578227e-08,
-      "loss": 0.8743,
-      "step": 6135
-    },
-    {
-      "epoch": 14.813027744270205,
-      "grad_norm": 0.318359375,
-      "learning_rate": 7.740011010778369e-08,
-      "loss": 0.867,
-      "step": 6140
-    },
-    {
-      "epoch": 14.825090470446321,
-      "grad_norm": 0.31640625,
-      "learning_rate": 6.673903619449639e-08,
-      "loss": 0.875,
-      "step": 6145
+      "loss": 0.83,
+      "step": 2045
     },
     {
-      "epoch": 14.837153196622436,
-      "grad_norm": 0.330078125,
+      "epoch": 9.879518072289157,
+      "grad_norm": 0.205078125,
       "learning_rate": 5.6867333537580226e-08,
-      "loss": 0.8846,
-      "step": 6150
-    },
-    {
-      "epoch": 14.849215922798553,
-      "grad_norm": 0.322265625,
-      "learning_rate": 4.7785080113449396e-08,
-      "loss": 0.8832,
-      "step": 6155
-    },
-    {
-      "epoch": 14.861278648974668,
-      "grad_norm": 0.32421875,
-      "learning_rate": 3.9492347662684057e-08,
-      "loss": 0.8764,
-      "step": 6160
+      "loss": 0.8322,
+      "step": 2050
     },
     {
-      "epoch": 14.873341375150783,
-      "grad_norm": 0.3203125,
+      "epoch": 9.903614457831326,
+      "grad_norm": 0.21484375,
       "learning_rate": 3.1989201689452967e-08,
-      "loss": 0.8753,
-      "step": 6165
-    },
-    {
-      "epoch": 14.8854041013269,
-      "grad_norm": 0.3046875,
-      "learning_rate": 2.5275701460991674e-08,
-      "loss": 0.8682,
-      "step": 6170
-    },
-    {
-      "epoch": 14.897466827503015,
-      "grad_norm": 0.306640625,
-      "learning_rate": 1.9351900007114066e-08,
-      "loss": 0.8719,
-      "step": 6175
+      "loss": 0.8305,
+      "step": 2055
     },
     {
-      "epoch": 14.909529553679132,
-      "grad_norm": 0.30078125,
+      "epoch": 9.927710843373493,
+      "grad_norm": 0.2080078125,
       "learning_rate": 1.4217844119857048e-08,
-      "loss": 0.8759,
-      "step": 6180
-    },
-    {
-      "epoch": 14.921592279855247,
-      "grad_norm": 0.30859375,
-      "learning_rate": 9.873574353025384e-09,
-      "loss": 0.8732,
-      "step": 6185
-    },
-    {
-      "epoch": 14.933655006031364,
-      "grad_norm": 0.31640625,
-      "learning_rate": 6.3191250219474255e-09,
-      "loss": 0.8793,
-      "step": 6190
+      "loss": 0.8293,
+      "step": 2060
     },
     {
-      "epoch": 14.945717732207479,
-      "grad_norm": 0.298828125,
+      "epoch": 9.951807228915662,
+      "grad_norm": 0.2041015625,
       "learning_rate": 3.554524203175369e-09,
-      "loss": 0.876,
-      "step": 6195
-    },
-    {
-      "epoch": 14.957780458383596,
-      "grad_norm": 0.318359375,
-      "learning_rate": 1.5797937342298952e-09,
-      "loss": 0.8788,
-      "step": 6200
-    },
-    {
-      "epoch": 14.96984318455971,
-      "grad_norm": 0.30859375,
-      "learning_rate": 3.949492134780464e-10,
-      "loss": 0.8771,
-      "step": 6205
+      "loss": 0.8287,
+      "step": 2065
     },
     {
-      "epoch": 14.981905910735826,
-      "grad_norm": 0.31640625,
+      "epoch": 9.975903614457831,
+      "grad_norm": 0.2177734375,
       "learning_rate": 0.0,
-      "loss": 0.8746,
-      "step": 6210
+      "loss": 0.827,
+      "step": 2070
     },
     {
-      "epoch": 14.981905910735826,
-      "eval_loss": 2.503845453262329,
-      "eval_runtime": 0.5267,
-      "eval_samples_per_second": 18.985,
-      "eval_steps_per_second": 1.899,
-      "step": 6210
+      "epoch": 9.975903614457831,
+      "eval_loss": 2.566892147064209,
+      "eval_runtime": 0.4844,
+      "eval_samples_per_second": 20.643,
+      "eval_steps_per_second": 2.064,
+      "step": 2070
     },
     {
-      "epoch": 14.981905910735826,
-      "step": 6210,
-      "total_flos": 4.863451355047526e+18,
-      "train_loss": 0.9534509487582098,
-      "train_runtime": 21285.3674,
-      "train_samples_per_second": 18.683,
-      "train_steps_per_second": 0.292
+      "epoch": 9.975903614457831,
+      "step": 2070,
+      "total_flos": 3.290190024938619e+18,
+      "train_loss": 0.9333097650233099,
+      "train_runtime": 14306.303,
+      "train_samples_per_second": 18.532,
+      "train_steps_per_second": 0.145
     }
   ],
   "logging_steps": 5,
-  "max_steps": 6210,
+  "max_steps": 2070,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 15,
+  "num_train_epochs": 10,
   "save_steps": 100,
-  "total_flos": 4.863451355047526e+18,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.290190024938619e+18,
   "train_batch_size": 8,
   "trial_name": null,
   "trial_params": null