diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,21000 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 14968,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 6.680919294494922e-05,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.3360053440213764e-07,
+      "loss": 1.5739,
+      "step": 1
+    },
+    {
+      "epoch": 0.00033404596472474615,
+      "grad_norm": 0.431640625,
+      "learning_rate": 6.680026720106881e-07,
+      "loss": 1.6599,
+      "step": 5
+    },
+    {
+      "epoch": 0.0006680919294494923,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.3360053440213763e-06,
+      "loss": 1.6221,
+      "step": 10
+    },
+    {
+      "epoch": 0.0010021378941742383,
+      "grad_norm": 0.361328125,
+      "learning_rate": 2.004008016032064e-06,
+      "loss": 1.5389,
+      "step": 15
+    },
+    {
+      "epoch": 0.0013361838588989846,
+      "grad_norm": 0.318359375,
+      "learning_rate": 2.6720106880427525e-06,
+      "loss": 1.5856,
+      "step": 20
+    },
+    {
+      "epoch": 0.0016702298236237307,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.3400133600534405e-06,
+      "loss": 1.5469,
+      "step": 25
+    },
+    {
+      "epoch": 0.0020042757883484766,
+      "grad_norm": 0.361328125,
+      "learning_rate": 4.008016032064128e-06,
+      "loss": 1.5633,
+      "step": 30
+    },
+    {
+      "epoch": 0.0023383217530732227,
+      "grad_norm": 0.279296875,
+      "learning_rate": 4.676018704074817e-06,
+      "loss": 1.5587,
+      "step": 35
+    },
+    {
+      "epoch": 0.002672367717797969,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5.344021376085505e-06,
+      "loss": 1.6141,
+      "step": 40
+    },
+    {
+      "epoch": 0.0030064136825227153,
+      "grad_norm": 0.318359375,
+      "learning_rate": 6.012024048096193e-06,
+      "loss": 1.5299,
+      "step": 45
+    },
+    {
+      "epoch": 0.0033404596472474614,
+      "grad_norm": 0.283203125,
+      "learning_rate": 6.680026720106881e-06,
+      "loss": 1.5561,
+      "step": 50
+    },
+    {
+      "epoch": 0.0036745056119722075,
+      "grad_norm": 0.35546875,
+      "learning_rate": 7.348029392117569e-06,
+      "loss": 1.5794,
+      "step": 55
+    },
+    {
+      "epoch": 0.004008551576696953,
+      "grad_norm": 0.310546875,
+      "learning_rate": 8.016032064128256e-06,
+      "loss": 1.5619,
+      "step": 60
+    },
+    {
+      "epoch": 0.004342597541421699,
+      "grad_norm": 0.2265625,
+      "learning_rate": 8.684034736138945e-06,
+      "loss": 1.5249,
+      "step": 65
+    },
+    {
+      "epoch": 0.004676643506146445,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.352037408149633e-06,
+      "loss": 1.5814,
+      "step": 70
+    },
+    {
+      "epoch": 0.0050106894708711915,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.0020040080160322e-05,
+      "loss": 1.483,
+      "step": 75
+    },
+    {
+      "epoch": 0.005344735435595938,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.068804275217101e-05,
+      "loss": 1.3885,
+      "step": 80
+    },
+    {
+      "epoch": 0.0056787814003206845,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.1356045424181697e-05,
+      "loss": 1.4409,
+      "step": 85
+    },
+    {
+      "epoch": 0.006012827365045431,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2024048096192385e-05,
+      "loss": 1.4832,
+      "step": 90
+    },
+    {
+      "epoch": 0.006346873329770177,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.2692050768203074e-05,
+      "loss": 1.4454,
+      "step": 95
+    },
+    {
+      "epoch": 0.006680919294494923,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.3360053440213762e-05,
+      "loss": 1.5162,
+      "step": 100
+    },
+    {
+      "epoch": 0.007014965259219669,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.4028056112224449e-05,
+      "loss": 1.4367,
+      "step": 105
+    },
+    {
+      "epoch": 0.007349011223944415,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.4696058784235137e-05,
+      "loss": 1.4453,
+      "step": 110
+    },
+    {
+      "epoch": 0.007683057188669161,
+      "grad_norm": 0.23046875,
+      "learning_rate": 1.5364061456245828e-05,
+      "loss": 1.4518,
+      "step": 115
+    },
+    {
+      "epoch": 0.008017103153393906,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 1.6032064128256513e-05,
+      "loss": 1.4095,
+      "step": 120
+    },
+    {
+      "epoch": 0.008351149118118652,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 1.6700066800267204e-05,
+      "loss": 1.3303,
+      "step": 125
+    },
+    {
+      "epoch": 0.008685195082843399,
+      "grad_norm": 0.2265625,
+      "learning_rate": 1.736806947227789e-05,
+      "loss": 1.5248,
+      "step": 130
+    },
+    {
+      "epoch": 0.009019241047568145,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.8036072144288578e-05,
+      "loss": 1.3274,
+      "step": 135
+    },
+    {
+      "epoch": 0.00935328701229289,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.8704074816299266e-05,
+      "loss": 1.4265,
+      "step": 140
+    },
+    {
+      "epoch": 0.009687332977017637,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9372077488309955e-05,
+      "loss": 1.3648,
+      "step": 145
+    },
+    {
+      "epoch": 0.010021378941742383,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.0040080160320643e-05,
+      "loss": 1.2991,
+      "step": 150
+    },
+    {
+      "epoch": 0.010355424906467129,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.070808283233133e-05,
+      "loss": 1.4394,
+      "step": 155
+    },
+    {
+      "epoch": 0.010689470871191877,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.137608550434202e-05,
+      "loss": 1.3224,
+      "step": 160
+    },
+    {
+      "epoch": 0.011023516835916623,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.2044088176352705e-05,
+      "loss": 1.3659,
+      "step": 165
+    },
+    {
+      "epoch": 0.011357562800641369,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.2712090848363394e-05,
+      "loss": 1.3688,
+      "step": 170
+    },
+    {
+      "epoch": 0.011691608765366115,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.3380093520374082e-05,
+      "loss": 1.3416,
+      "step": 175
+    },
+    {
+      "epoch": 0.012025654730090861,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.404809619238477e-05,
+      "loss": 1.3394,
+      "step": 180
+    },
+    {
+      "epoch": 0.012359700694815607,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 2.471609886439546e-05,
+      "loss": 1.2949,
+      "step": 185
+    },
+    {
+      "epoch": 0.012693746659540353,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.5384101536406147e-05,
+      "loss": 1.4109,
+      "step": 190
+    },
+    {
+      "epoch": 0.0130277926242651,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.6052104208416833e-05,
+      "loss": 1.3706,
+      "step": 195
+    },
+    {
+      "epoch": 0.013361838588989846,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.6720106880427524e-05,
+      "loss": 1.3242,
+      "step": 200
+    },
+    {
+      "epoch": 0.013695884553714592,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.7388109552438213e-05,
+      "loss": 1.3187,
+      "step": 205
+    },
+    {
+      "epoch": 0.014029930518439338,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.8056112224448898e-05,
+      "loss": 1.4374,
+      "step": 210
+    },
+    {
+      "epoch": 0.014363976483164084,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 2.8724114896459586e-05,
+      "loss": 1.2938,
+      "step": 215
+    },
+    {
+      "epoch": 0.01469802244788883,
+      "grad_norm": 0.21875,
+      "learning_rate": 2.9392117568470275e-05,
+      "loss": 1.3603,
+      "step": 220
+    },
+    {
+      "epoch": 0.015032068412613576,
+      "grad_norm": 0.20703125,
+      "learning_rate": 3.0060120240480967e-05,
+      "loss": 1.3989,
+      "step": 225
+    },
+    {
+      "epoch": 0.015366114377338322,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 3.0728122912491655e-05,
+      "loss": 1.3631,
+      "step": 230
+    },
+    {
+      "epoch": 0.015700160342063067,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.139612558450234e-05,
+      "loss": 1.318,
+      "step": 235
+    },
+    {
+      "epoch": 0.016034206306787813,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 3.2064128256513025e-05,
+      "loss": 1.344,
+      "step": 240
+    },
+    {
+      "epoch": 0.01636825227151256,
+      "grad_norm": 0.1875,
+      "learning_rate": 3.273213092852371e-05,
+      "loss": 1.3326,
+      "step": 245
+    },
+    {
+      "epoch": 0.016702298236237305,
+      "grad_norm": 0.216796875,
+      "learning_rate": 3.340013360053441e-05,
+      "loss": 1.2978,
+      "step": 250
+    },
+    {
+      "epoch": 0.01703634420096205,
+      "grad_norm": 0.228515625,
+      "learning_rate": 3.4068136272545094e-05,
+      "loss": 1.2572,
+      "step": 255
+    },
+    {
+      "epoch": 0.017370390165686797,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.473613894455578e-05,
+      "loss": 1.2778,
+      "step": 260
+    },
+    {
+      "epoch": 0.017704436130411543,
+      "grad_norm": 0.30078125,
+      "learning_rate": 3.5404141616566464e-05,
+      "loss": 1.2495,
+      "step": 265
+    },
+    {
+      "epoch": 0.01803848209513629,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.6072144288577156e-05,
+      "loss": 1.3069,
+      "step": 270
+    },
+    {
+      "epoch": 0.018372528059861035,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 3.674014696058785e-05,
+      "loss": 1.3499,
+      "step": 275
+    },
+    {
+      "epoch": 0.01870657402458578,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 3.740814963259853e-05,
+      "loss": 1.3404,
+      "step": 280
+    },
+    {
+      "epoch": 0.019040619989310528,
+      "grad_norm": 0.21875,
+      "learning_rate": 3.807615230460922e-05,
+      "loss": 1.2487,
+      "step": 285
+    },
+    {
+      "epoch": 0.019374665954035274,
+      "grad_norm": 0.212890625,
+      "learning_rate": 3.874415497661991e-05,
+      "loss": 1.3379,
+      "step": 290
+    },
+    {
+      "epoch": 0.01970871191876002,
+      "grad_norm": 0.1953125,
+      "learning_rate": 3.94121576486306e-05,
+      "loss": 1.2742,
+      "step": 295
+    },
+    {
+      "epoch": 0.020042757883484766,
+      "grad_norm": 0.193359375,
+      "learning_rate": 4.0080160320641287e-05,
+      "loss": 1.3057,
+      "step": 300
+    },
+    {
+      "epoch": 0.020376803848209512,
+      "grad_norm": 0.185546875,
+      "learning_rate": 4.074816299265197e-05,
+      "loss": 1.2915,
+      "step": 305
+    },
+    {
+      "epoch": 0.020710849812934258,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 4.141616566466266e-05,
+      "loss": 1.2924,
+      "step": 310
+    },
+    {
+      "epoch": 0.021044895777659008,
+      "grad_norm": 0.19140625,
+      "learning_rate": 4.208416833667335e-05,
+      "loss": 1.2266,
+      "step": 315
+    },
+    {
+      "epoch": 0.021378941742383754,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 4.275217100868404e-05,
+      "loss": 1.3178,
+      "step": 320
+    },
+    {
+      "epoch": 0.0217129877071085,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 4.3420173680694725e-05,
+      "loss": 1.2772,
+      "step": 325
+    },
+    {
+      "epoch": 0.022047033671833246,
+      "grad_norm": 0.189453125,
+      "learning_rate": 4.408817635270541e-05,
+      "loss": 1.3171,
+      "step": 330
+    },
+    {
+      "epoch": 0.022381079636557992,
+      "grad_norm": 0.236328125,
+      "learning_rate": 4.47561790247161e-05,
+      "loss": 1.3287,
+      "step": 335
+    },
+    {
+      "epoch": 0.022715125601282738,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 4.542418169672679e-05,
+      "loss": 1.3499,
+      "step": 340
+    },
+    {
+      "epoch": 0.023049171566007484,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 4.609218436873748e-05,
+      "loss": 1.2967,
+      "step": 345
+    },
+    {
+      "epoch": 0.02338321753073223,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 4.6760187040748164e-05,
+      "loss": 1.2266,
+      "step": 350
+    },
+    {
+      "epoch": 0.023717263495456976,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 4.742818971275885e-05,
+      "loss": 1.265,
+      "step": 355
+    },
+    {
+      "epoch": 0.024051309460181722,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 4.809619238476954e-05,
+      "loss": 1.2426,
+      "step": 360
+    },
+    {
+      "epoch": 0.02438535542490647,
+      "grad_norm": 0.244140625,
+      "learning_rate": 4.876419505678023e-05,
+      "loss": 1.3144,
+      "step": 365
+    },
+    {
+      "epoch": 0.024719401389631215,
+      "grad_norm": 0.208984375,
+      "learning_rate": 4.943219772879092e-05,
+      "loss": 1.2892,
+      "step": 370
+    },
+    {
+      "epoch": 0.02505344735435596,
+      "grad_norm": 0.216796875,
+      "learning_rate": 5.01002004008016e-05,
+      "loss": 1.1596,
+      "step": 375
+    },
+    {
+      "epoch": 0.025387493319080707,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 5.0768203072812295e-05,
+      "loss": 1.3151,
+      "step": 380
+    },
+    {
+      "epoch": 0.025721539283805453,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 5.143620574482299e-05,
+      "loss": 1.2643,
+      "step": 385
+    },
+    {
+      "epoch": 0.0260555852485302,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 5.2104208416833665e-05,
+      "loss": 1.2695,
+      "step": 390
+    },
+    {
+      "epoch": 0.026389631213254945,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 5.277221108884436e-05,
+      "loss": 1.2472,
+      "step": 395
+    },
+    {
+      "epoch": 0.02672367717797969,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 5.344021376085505e-05,
+      "loss": 1.3066,
+      "step": 400
+    },
+    {
+      "epoch": 0.027057723142704437,
+      "grad_norm": 0.251953125,
+      "learning_rate": 5.4108216432865734e-05,
+      "loss": 1.2664,
+      "step": 405
+    },
+    {
+      "epoch": 0.027391769107429183,
+      "grad_norm": 0.251953125,
+      "learning_rate": 5.4776219104876426e-05,
+      "loss": 1.221,
+      "step": 410
+    },
+    {
+      "epoch": 0.02772581507215393,
+      "grad_norm": 0.201171875,
+      "learning_rate": 5.5444221776887104e-05,
+      "loss": 1.3246,
+      "step": 415
+    },
+    {
+      "epoch": 0.028059861036878676,
+      "grad_norm": 0.22265625,
+      "learning_rate": 5.6112224448897796e-05,
+      "loss": 1.2621,
+      "step": 420
+    },
+    {
+      "epoch": 0.028393907001603422,
+      "grad_norm": 0.2578125,
+      "learning_rate": 5.6780227120908494e-05,
+      "loss": 1.3305,
+      "step": 425
+    },
+    {
+      "epoch": 0.028727952966328168,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 5.744822979291917e-05,
+      "loss": 1.2464,
+      "step": 430
+    },
+    {
+      "epoch": 0.029061998931052914,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 5.8116232464929865e-05,
+      "loss": 1.284,
+      "step": 435
+    },
+    {
+      "epoch": 0.02939604489577766,
+      "grad_norm": 0.2109375,
+      "learning_rate": 5.878423513694055e-05,
+      "loss": 1.3277,
+      "step": 440
+    },
+    {
+      "epoch": 0.029730090860502406,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 5.945223780895124e-05,
+      "loss": 1.2383,
+      "step": 445
+    },
+    {
+      "epoch": 0.030064136825227152,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 6.012024048096193e-05,
+      "loss": 1.2604,
+      "step": 450
+    },
+    {
+      "epoch": 0.0303981827899519,
+      "grad_norm": 0.228515625,
+      "learning_rate": 6.078824315297261e-05,
+      "loss": 1.302,
+      "step": 455
+    },
+    {
+      "epoch": 0.030732228754676644,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 6.145624582498331e-05,
+      "loss": 1.2275,
+      "step": 460
+    },
+    {
+      "epoch": 0.03106627471940139,
+      "grad_norm": 0.21484375,
+      "learning_rate": 6.212424849699398e-05,
+      "loss": 1.3504,
+      "step": 465
+    },
+    {
+      "epoch": 0.03140032068412613,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 6.279225116900468e-05,
+      "loss": 1.2612,
+      "step": 470
+    },
+    {
+      "epoch": 0.03173436664885088,
+      "grad_norm": 0.23828125,
+      "learning_rate": 6.346025384101537e-05,
+      "loss": 1.2555,
+      "step": 475
+    },
+    {
+      "epoch": 0.032068412613575625,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 6.412825651302605e-05,
+      "loss": 1.2519,
+      "step": 480
+    },
+    {
+      "epoch": 0.032402458578300375,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 6.479625918503675e-05,
+      "loss": 1.2233,
+      "step": 485
+    },
+    {
+      "epoch": 0.03273650454302512,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 6.546426185704742e-05,
+      "loss": 1.2148,
+      "step": 490
+    },
+    {
+      "epoch": 0.03307055050774987,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 6.613226452905812e-05,
+      "loss": 1.2288,
+      "step": 495
+    },
+    {
+      "epoch": 0.03340459647247461,
+      "grad_norm": 0.224609375,
+      "learning_rate": 6.680026720106882e-05,
+      "loss": 1.3175,
+      "step": 500
+    },
+    {
+      "epoch": 0.03373864243719936,
+      "grad_norm": 0.193359375,
+      "learning_rate": 6.746826987307949e-05,
+      "loss": 1.2848,
+      "step": 505
+    },
+    {
+      "epoch": 0.0340726884019241,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 6.813627254509019e-05,
+      "loss": 1.2759,
+      "step": 510
+    },
+    {
+      "epoch": 0.03440673436664885,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 6.880427521710087e-05,
+      "loss": 1.3355,
+      "step": 515
+    },
+    {
+      "epoch": 0.034740780331373594,
+      "grad_norm": 0.201171875,
+      "learning_rate": 6.947227788911156e-05,
+      "loss": 1.2873,
+      "step": 520
+    },
+    {
+      "epoch": 0.035074826296098344,
+      "grad_norm": 0.21484375,
+      "learning_rate": 7.014028056112226e-05,
+      "loss": 1.27,
+      "step": 525
+    },
+    {
+      "epoch": 0.035408872260823086,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 7.080828323313293e-05,
+      "loss": 1.1939,
+      "step": 530
+    },
+    {
+      "epoch": 0.035742918225547836,
+      "grad_norm": 0.212890625,
+      "learning_rate": 7.147628590514363e-05,
+      "loss": 1.2536,
+      "step": 535
+    },
+    {
+      "epoch": 0.03607696419027258,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 7.214428857715431e-05,
+      "loss": 1.276,
+      "step": 540
+    },
+    {
+      "epoch": 0.03641101015499733,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 7.2812291249165e-05,
+      "loss": 1.2621,
+      "step": 545
+    },
+    {
+      "epoch": 0.03674505611972207,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 7.34802939211757e-05,
+      "loss": 1.2361,
+      "step": 550
+    },
+    {
+      "epoch": 0.03707910208444682,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 7.414829659318637e-05,
+      "loss": 1.3944,
+      "step": 555
+    },
+    {
+      "epoch": 0.03741314804917156,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 7.481629926519707e-05,
+      "loss": 1.2375,
+      "step": 560
+    },
+    {
+      "epoch": 0.03774719401389631,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 7.548430193720776e-05,
+      "loss": 1.2653,
+      "step": 565
+    },
+    {
+      "epoch": 0.038081239978621055,
+      "grad_norm": 0.208984375,
+      "learning_rate": 7.615230460921844e-05,
+      "loss": 1.2265,
+      "step": 570
+    },
+    {
+      "epoch": 0.038415285943345805,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 7.682030728122913e-05,
+      "loss": 1.263,
+      "step": 575
+    },
+    {
+      "epoch": 0.03874933190807055,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 7.748830995323982e-05,
+      "loss": 1.1832,
+      "step": 580
+    },
+    {
+      "epoch": 0.0390833778727953,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 7.81563126252505e-05,
+      "loss": 1.2563,
+      "step": 585
+    },
+    {
+      "epoch": 0.03941742383752004,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.88243152972612e-05,
+      "loss": 1.2236,
+      "step": 590
+    },
+    {
+      "epoch": 0.03975146980224479,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 7.949231796927187e-05,
+      "loss": 1.2766,
+      "step": 595
+    },
+    {
+      "epoch": 0.04008551576696953,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 8.016032064128257e-05,
+      "loss": 1.2377,
+      "step": 600
+    },
+    {
+      "epoch": 0.04041956173169428,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 8.082832331329326e-05,
+      "loss": 1.2083,
+      "step": 605
+    },
+    {
+      "epoch": 0.040753607696419024,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.149632598530394e-05,
+      "loss": 1.3272,
+      "step": 610
+    },
+    {
+      "epoch": 0.04108765366114377,
+      "grad_norm": 0.1953125,
+      "learning_rate": 8.216432865731464e-05,
+      "loss": 1.2428,
+      "step": 615
+    },
+    {
+      "epoch": 0.041421699625868516,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.283233132932531e-05,
+      "loss": 1.3088,
+      "step": 620
+    },
+    {
+      "epoch": 0.041755745590593266,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 8.350033400133601e-05,
+      "loss": 1.2174,
+      "step": 625
+    },
+    {
+      "epoch": 0.042089791555318015,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 8.41683366733467e-05,
+      "loss": 1.1424,
+      "step": 630
+    },
+    {
+      "epoch": 0.04242383752004276,
+      "grad_norm": 0.234375,
+      "learning_rate": 8.483633934535738e-05,
+      "loss": 1.2562,
+      "step": 635
+    },
+    {
+      "epoch": 0.04275788348476751,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 8.550434201736808e-05,
+      "loss": 1.2747,
+      "step": 640
+    },
+    {
+      "epoch": 0.04309192944949225,
+      "grad_norm": 0.1953125,
+      "learning_rate": 8.617234468937875e-05,
+      "loss": 1.2727,
+      "step": 645
+    },
+    {
+      "epoch": 0.043425975414217,
+      "grad_norm": 0.1953125,
+      "learning_rate": 8.684034736138945e-05,
+      "loss": 1.2413,
+      "step": 650
+    },
+    {
+      "epoch": 0.04376002137894174,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 8.750835003340014e-05,
+      "loss": 1.2405,
+      "step": 655
+    },
+    {
+      "epoch": 0.04409406734366649,
+      "grad_norm": 0.1875,
+      "learning_rate": 8.817635270541082e-05,
+      "loss": 1.2667,
+      "step": 660
+    },
+    {
+      "epoch": 0.044428113308391234,
+      "grad_norm": 0.23828125,
+      "learning_rate": 8.884435537742152e-05,
+      "loss": 1.2387,
+      "step": 665
+    },
+    {
+      "epoch": 0.044762159273115984,
+      "grad_norm": 0.193359375,
+      "learning_rate": 8.95123580494322e-05,
+      "loss": 1.1747,
+      "step": 670
+    },
+    {
+      "epoch": 0.04509620523784073,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 9.018036072144289e-05,
+      "loss": 1.1909,
+      "step": 675
+    },
+    {
+      "epoch": 0.045430251202565476,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 9.084836339345357e-05,
+      "loss": 1.2968,
+      "step": 680
+    },
+    {
+      "epoch": 0.04576429716729022,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.151636606546426e-05,
+      "loss": 1.2266,
+      "step": 685
+    },
+    {
+      "epoch": 0.04609834313201497,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 9.218436873747496e-05,
+      "loss": 1.1976,
+      "step": 690
+    },
+    {
+      "epoch": 0.04643238909673971,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.285237140948564e-05,
+      "loss": 1.2633,
+      "step": 695
+    },
+    {
+      "epoch": 0.04676643506146446,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.352037408149633e-05,
+      "loss": 1.1995,
+      "step": 700
+    },
+    {
+      "epoch": 0.0471004810261892,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 9.418837675350703e-05,
+      "loss": 1.2037,
+      "step": 705
+    },
+    {
+      "epoch": 0.04743452699091395,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 9.48563794255177e-05,
+      "loss": 1.2941,
+      "step": 710
+    },
+    {
+      "epoch": 0.047768572955638695,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 9.55243820975284e-05,
+      "loss": 1.1902,
+      "step": 715
+    },
+    {
+      "epoch": 0.048102618920363445,
+      "grad_norm": 0.205078125,
+      "learning_rate": 9.619238476953908e-05,
+      "loss": 1.214,
+      "step": 720
+    },
+    {
+      "epoch": 0.04843666488508819,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 9.686038744154977e-05,
+      "loss": 1.1536,
+      "step": 725
+    },
+    {
+      "epoch": 0.04877071084981294,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.752839011356047e-05,
+      "loss": 1.2421,
+      "step": 730
+    },
+    {
+      "epoch": 0.04910475681453768,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.819639278557115e-05,
+      "loss": 1.2786,
+      "step": 735
+    },
+    {
+      "epoch": 0.04943880277926243,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 9.886439545758184e-05,
+      "loss": 1.2637,
+      "step": 740
+    },
+    {
+      "epoch": 0.04977284874398717,
+      "grad_norm": 0.197265625,
+      "learning_rate": 9.953239812959252e-05,
+      "loss": 1.2836,
+      "step": 745
+    },
+    {
+      "epoch": 0.05010689470871192,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001002004008016032,
+      "loss": 1.281,
+      "step": 750
+    },
+    {
+      "epoch": 0.050440940673436664,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001008684034736139,
+      "loss": 1.2277,
+      "step": 755
+    },
+    {
+      "epoch": 0.050774986638161414,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00010153640614562459,
+      "loss": 1.2501,
+      "step": 760
+    },
+    {
+      "epoch": 0.051109032602886156,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00010220440881763526,
+      "loss": 1.294,
+      "step": 765
+    },
+    {
+      "epoch": 0.051443078567610906,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00010287241148964597,
+      "loss": 1.2892,
+      "step": 770
+    },
+    {
+      "epoch": 0.05177712453233565,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00010354041416165665,
+      "loss": 1.2675,
+      "step": 775
+    },
+    {
+      "epoch": 0.0521111704970604,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00010420841683366733,
+      "loss": 1.2157,
+      "step": 780
+    },
+    {
+      "epoch": 0.05244521646178514,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00010487641950567804,
+      "loss": 1.2635,
+      "step": 785
+    },
+    {
+      "epoch": 0.05277926242650989,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00010554442217768871,
+      "loss": 1.2452,
+      "step": 790
+    },
+    {
+      "epoch": 0.05311330839123463,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0001062124248496994,
+      "loss": 1.1851,
+      "step": 795
+    },
+    {
+      "epoch": 0.05344735435595938,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0001068804275217101,
+      "loss": 1.2847,
+      "step": 800
+    },
+    {
+      "epoch": 0.053781400320684125,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00010754843019372078,
+      "loss": 1.2923,
+      "step": 805
+    },
+    {
+      "epoch": 0.054115446285408875,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00010821643286573147,
+      "loss": 1.1858,
+      "step": 810
+    },
+    {
+      "epoch": 0.05444949225013362,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00010888443553774215,
+      "loss": 1.2945,
+      "step": 815
+    },
+    {
+      "epoch": 0.05478353821485837,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00010955243820975285,
+      "loss": 1.3139,
+      "step": 820
+    },
+    {
+      "epoch": 0.05511758417958311,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00011022044088176354,
+      "loss": 1.1953,
+      "step": 825
+    },
+    {
+      "epoch": 0.05545163014430786,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00011088844355377421,
+      "loss": 1.2724,
+      "step": 830
+    },
+    {
+      "epoch": 0.0557856761090326,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00011155644622578492,
+      "loss": 1.3072,
+      "step": 835
+    },
+    {
+      "epoch": 0.05611972207375735,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00011222444889779559,
+      "loss": 1.2644,
+      "step": 840
+    },
+    {
+      "epoch": 0.056453768038482094,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00011289245156980628,
+      "loss": 1.2475,
+      "step": 845
+    },
+    {
+      "epoch": 0.056787814003206843,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00011356045424181699,
+      "loss": 1.1897,
+      "step": 850
+    },
+    {
+      "epoch": 0.057121859967931586,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00011422845691382766,
+      "loss": 1.2661,
+      "step": 855
+    },
+    {
+      "epoch": 0.057455905932656336,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00011489645958583835,
+      "loss": 1.2381,
+      "step": 860
+    },
+    {
+      "epoch": 0.05778995189738108,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00011556446225784903,
+      "loss": 1.2789,
+      "step": 865
+    },
+    {
+      "epoch": 0.05812399786210583,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00011623246492985973,
+      "loss": 1.2097,
+      "step": 870
+    },
+    {
+      "epoch": 0.05845804382683057,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00011690046760187041,
+      "loss": 1.2372,
+      "step": 875
+    },
+    {
+      "epoch": 0.05879208979155532,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0001175684702738811,
+      "loss": 1.2229,
+      "step": 880
+    },
+    {
+      "epoch": 0.05912613575628006,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001182364729458918,
+      "loss": 1.2605,
+      "step": 885
+    },
+    {
+      "epoch": 0.05946018172100481,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00011890447561790248,
+      "loss": 1.2223,
+      "step": 890
+    },
+    {
+      "epoch": 0.059794227685729555,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00011957247828991315,
+      "loss": 1.2466,
+      "step": 895
+    },
+    {
+      "epoch": 0.060128273650454304,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00012024048096192387,
+      "loss": 1.2771,
+      "step": 900
+    },
+    {
+      "epoch": 0.06046231961517905,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00012090848363393454,
+      "loss": 1.2373,
+      "step": 905
+    },
+    {
+      "epoch": 0.0607963655799038,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00012157648630594522,
+      "loss": 1.2916,
+      "step": 910
+    },
+    {
+      "epoch": 0.06113041154462854,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0001222444889779559,
+      "loss": 1.2504,
+      "step": 915
+    },
+    {
+      "epoch": 0.06146445750935329,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00012291249164996662,
+      "loss": 1.2714,
+      "step": 920
+    },
+    {
+      "epoch": 0.06179850347407803,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001235804943219773,
+      "loss": 1.2904,
+      "step": 925
+    },
+    {
+      "epoch": 0.06213254943880278,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00012424849699398796,
+      "loss": 1.254,
+      "step": 930
+    },
+    {
+      "epoch": 0.062466595403527524,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0001249164996659987,
+      "loss": 1.3173,
+      "step": 935
+    },
+    {
+      "epoch": 0.06280064136825227,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00012558450233800936,
+      "loss": 1.2083,
+      "step": 940
+    },
+    {
+      "epoch": 0.06313468733297702,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00012625250501002003,
+      "loss": 1.2594,
+      "step": 945
+    },
+    {
+      "epoch": 0.06346873329770177,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00012692050768203073,
+      "loss": 1.2027,
+      "step": 950
+    },
+    {
+      "epoch": 0.06380277926242651,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00012758851035404143,
+      "loss": 1.1886,
+      "step": 955
+    },
+    {
+      "epoch": 0.06413682522715125,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001282565130260521,
+      "loss": 1.1855,
+      "step": 960
+    },
+    {
+      "epoch": 0.064470871191876,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001289245156980628,
+      "loss": 1.2326,
+      "step": 965
+    },
+    {
+      "epoch": 0.06480491715660075,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001295925183700735,
+      "loss": 1.2502,
+      "step": 970
+    },
+    {
+      "epoch": 0.0651389631213255,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00013026052104208417,
+      "loss": 1.1563,
+      "step": 975
+    },
+    {
+      "epoch": 0.06547300908605024,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00013092852371409484,
+      "loss": 1.2203,
+      "step": 980
+    },
+    {
+      "epoch": 0.06580705505077498,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00013159652638610557,
+      "loss": 1.3099,
+      "step": 985
+    },
+    {
+      "epoch": 0.06614110101549973,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00013226452905811624,
+      "loss": 1.2904,
+      "step": 990
+    },
+    {
+      "epoch": 0.06647514698022448,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0001329325317301269,
+      "loss": 1.2602,
+      "step": 995
+    },
+    {
+      "epoch": 0.06680919294494922,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00013360053440213764,
+      "loss": 1.1857,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06714323890967397,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001342685370741483,
+      "loss": 1.2519,
+      "step": 1005
+    },
+    {
+      "epoch": 0.06747728487439872,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00013493653974615898,
+      "loss": 1.1755,
+      "step": 1010
+    },
+    {
+      "epoch": 0.06781133083912347,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00013560454241816968,
+      "loss": 1.2496,
+      "step": 1015
+    },
+    {
+      "epoch": 0.0681453768038482,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00013627254509018038,
+      "loss": 1.2386,
+      "step": 1020
+    },
+    {
+      "epoch": 0.06847942276857295,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00013694054776219105,
+      "loss": 1.2811,
+      "step": 1025
+    },
+    {
+      "epoch": 0.0688134687332977,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00013760855043420175,
+      "loss": 1.2174,
+      "step": 1030
+    },
+    {
+      "epoch": 0.06914751469802245,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00013827655310621244,
+      "loss": 1.2521,
+      "step": 1035
+    },
+    {
+      "epoch": 0.06948156066274719,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00013894455577822312,
+      "loss": 1.2147,
+      "step": 1040
+    },
+    {
+      "epoch": 0.06981560662747194,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0001396125584502338,
+      "loss": 1.2223,
+      "step": 1045
+    },
+    {
+      "epoch": 0.07014965259219669,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001402805611222445,
+      "loss": 1.1894,
+      "step": 1050
+    },
+    {
+      "epoch": 0.07048369855692144,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00014094856379425518,
+      "loss": 1.2567,
+      "step": 1055
+    },
+    {
+      "epoch": 0.07081774452164617,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00014161656646626586,
+      "loss": 1.2306,
+      "step": 1060
+    },
+    {
+      "epoch": 0.07115179048637092,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00014228456913827658,
+      "loss": 1.19,
+      "step": 1065
+    },
+    {
+      "epoch": 0.07148583645109567,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00014295257181028725,
+      "loss": 1.2903,
+      "step": 1070
+    },
+    {
+      "epoch": 0.07181988241582042,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00014362057448229792,
+      "loss": 1.1799,
+      "step": 1075
+    },
+    {
+      "epoch": 0.07215392838054516,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00014428857715430862,
+      "loss": 1.2499,
+      "step": 1080
+    },
+    {
+      "epoch": 0.0724879743452699,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00014495657982631932,
+      "loss": 1.256,
+      "step": 1085
+    },
+    {
+      "epoch": 0.07282202030999466,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00014562458249833,
+      "loss": 1.3308,
+      "step": 1090
+    },
+    {
+      "epoch": 0.0731560662747194,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001462925851703407,
+      "loss": 1.2816,
+      "step": 1095
+    },
+    {
+      "epoch": 0.07349011223944414,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001469605878423514,
+      "loss": 1.2938,
+      "step": 1100
+    },
+    {
+      "epoch": 0.07382415820416889,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00014762859051436206,
+      "loss": 1.3056,
+      "step": 1105
+    },
+    {
+      "epoch": 0.07415820416889364,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00014829659318637273,
+      "loss": 1.2525,
+      "step": 1110
+    },
+    {
+      "epoch": 0.07449225013361839,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00014896459585838346,
+      "loss": 1.1993,
+      "step": 1115
+    },
+    {
+      "epoch": 0.07482629609834313,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00014963259853039413,
+      "loss": 1.2613,
+      "step": 1120
+    },
+    {
+      "epoch": 0.07516034206306788,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001503006012024048,
+      "loss": 1.2884,
+      "step": 1125
+    },
+    {
+      "epoch": 0.07549438802779262,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00015096860387441553,
+      "loss": 1.2162,
+      "step": 1130
+    },
+    {
+      "epoch": 0.07582843399251737,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001516366065464262,
+      "loss": 1.2131,
+      "step": 1135
+    },
+    {
+      "epoch": 0.07616247995724211,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00015230460921843687,
+      "loss": 1.2212,
+      "step": 1140
+    },
+    {
+      "epoch": 0.07649652592196686,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00015297261189044757,
+      "loss": 1.1528,
+      "step": 1145
+    },
+    {
+      "epoch": 0.07683057188669161,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00015364061456245827,
+      "loss": 1.125,
+      "step": 1150
+    },
+    {
+      "epoch": 0.07716461785141636,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00015430861723446894,
+      "loss": 1.2868,
+      "step": 1155
+    },
+    {
+      "epoch": 0.0774986638161411,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00015497661990647964,
+      "loss": 1.2277,
+      "step": 1160
+    },
+    {
+      "epoch": 0.07783270978086584,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00015564462257849034,
+      "loss": 1.2085,
+      "step": 1165
+    },
+    {
+      "epoch": 0.0781667557455906,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.000156312625250501,
+      "loss": 1.1623,
+      "step": 1170
+    },
+    {
+      "epoch": 0.07850080171031534,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00015698062792251168,
+      "loss": 1.2236,
+      "step": 1175
+    },
+    {
+      "epoch": 0.07883484767504008,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001576486305945224,
+      "loss": 1.2148,
+      "step": 1180
+    },
+    {
+      "epoch": 0.07916889363976483,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00015831663326653308,
+      "loss": 1.2213,
+      "step": 1185
+    },
+    {
+      "epoch": 0.07950293960448958,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00015898463593854375,
+      "loss": 1.1879,
+      "step": 1190
+    },
+    {
+      "epoch": 0.07983698556921433,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00015965263861055445,
+      "loss": 1.2233,
+      "step": 1195
+    },
+    {
+      "epoch": 0.08017103153393906,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00016032064128256515,
+      "loss": 1.2497,
+      "step": 1200
+    },
+    {
+      "epoch": 0.08050507749866381,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00016098864395457582,
+      "loss": 1.1905,
+      "step": 1205
+    },
+    {
+      "epoch": 0.08083912346338856,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00016165664662658652,
+      "loss": 1.1979,
+      "step": 1210
+    },
+    {
+      "epoch": 0.08117316942811331,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00016232464929859721,
+      "loss": 1.2394,
+      "step": 1215
+    },
+    {
+      "epoch": 0.08150721539283805,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001629926519706079,
+      "loss": 1.2193,
+      "step": 1220
+    },
+    {
+      "epoch": 0.0818412613575628,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00016366065464261859,
+      "loss": 1.169,
+      "step": 1225
+    },
+    {
+      "epoch": 0.08217530732228755,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00016432865731462928,
+      "loss": 1.2845,
+      "step": 1230
+    },
+    {
+      "epoch": 0.0825093532870123,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00016499665998663996,
+      "loss": 1.298,
+      "step": 1235
+    },
+    {
+      "epoch": 0.08284339925173703,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00016566466265865063,
+      "loss": 1.2208,
+      "step": 1240
+    },
+    {
+      "epoch": 0.08317744521646178,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00016633266533066135,
+      "loss": 1.2539,
+      "step": 1245
+    },
+    {
+      "epoch": 0.08351149118118653,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00016700066800267202,
+      "loss": 1.2138,
+      "step": 1250
+    },
+    {
+      "epoch": 0.08384553714591128,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001676686706746827,
+      "loss": 1.2358,
+      "step": 1255
+    },
+    {
+      "epoch": 0.08417958311063603,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001683366733466934,
+      "loss": 1.2247,
+      "step": 1260
+    },
+    {
+      "epoch": 0.08451362907536077,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001690046760187041,
+      "loss": 1.3043,
+      "step": 1265
+    },
+    {
+      "epoch": 0.08484767504008552,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00016967267869071476,
+      "loss": 1.1915,
+      "step": 1270
+    },
+    {
+      "epoch": 0.08518172100481027,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00017034068136272546,
+      "loss": 1.151,
+      "step": 1275
+    },
+    {
+      "epoch": 0.08551576696953501,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00017100868403473616,
+      "loss": 1.2174,
+      "step": 1280
+    },
+    {
+      "epoch": 0.08584981293425975,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00017167668670674683,
+      "loss": 1.2237,
+      "step": 1285
+    },
+    {
+      "epoch": 0.0861838588989845,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001723446893787575,
+      "loss": 1.2163,
+      "step": 1290
+    },
+    {
+      "epoch": 0.08651790486370925,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00017301269205076823,
+      "loss": 1.3327,
+      "step": 1295
+    },
+    {
+      "epoch": 0.086851950828434,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001736806947227789,
+      "loss": 1.2554,
+      "step": 1300
+    },
+    {
+      "epoch": 0.08718599679315873,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00017434869739478957,
+      "loss": 1.2164,
+      "step": 1305
+    },
+    {
+      "epoch": 0.08752004275788348,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00017501670006680027,
+      "loss": 1.1899,
+      "step": 1310
+    },
+    {
+      "epoch": 0.08785408872260823,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00017568470273881097,
+      "loss": 1.2598,
+      "step": 1315
+    },
+    {
+      "epoch": 0.08818813468733298,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00017635270541082164,
+      "loss": 1.2093,
+      "step": 1320
+    },
+    {
+      "epoch": 0.08852218065205772,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00017702070808283234,
+      "loss": 1.1936,
+      "step": 1325
+    },
+    {
+      "epoch": 0.08885622661678247,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00017768871075484304,
+      "loss": 1.3092,
+      "step": 1330
+    },
+    {
+      "epoch": 0.08919027258150722,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001783567134268537,
+      "loss": 1.2603,
+      "step": 1335
+    },
+    {
+      "epoch": 0.08952431854623197,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001790247160988644,
+      "loss": 1.1828,
+      "step": 1340
+    },
+    {
+      "epoch": 0.0898583645109567,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001796927187708751,
+      "loss": 1.2985,
+      "step": 1345
+    },
+    {
+      "epoch": 0.09019241047568145,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00018036072144288578,
+      "loss": 1.1845,
+      "step": 1350
+    },
+    {
+      "epoch": 0.0905264564404062,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00018102872411489645,
+      "loss": 1.1018,
+      "step": 1355
+    },
+    {
+      "epoch": 0.09086050240513095,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00018169672678690715,
+      "loss": 1.1901,
+      "step": 1360
+    },
+    {
+      "epoch": 0.09119454836985569,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00018236472945891785,
+      "loss": 1.2401,
+      "step": 1365
+    },
+    {
+      "epoch": 0.09152859433458044,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00018303273213092852,
+      "loss": 1.2479,
+      "step": 1370
+    },
+    {
+      "epoch": 0.09186264029930519,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00018370073480293922,
+      "loss": 1.1413,
+      "step": 1375
+    },
+    {
+      "epoch": 0.09219668626402994,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00018436873747494992,
+      "loss": 1.2119,
+      "step": 1380
+    },
+    {
+      "epoch": 0.09253073222875467,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001850367401469606,
+      "loss": 1.2831,
+      "step": 1385
+    },
+    {
+      "epoch": 0.09286477819347942,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001857047428189713,
+      "loss": 1.2524,
+      "step": 1390
+    },
+    {
+      "epoch": 0.09319882415820417,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00018637274549098199,
+      "loss": 1.2294,
+      "step": 1395
+    },
+    {
+      "epoch": 0.09353287012292892,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00018704074816299266,
+      "loss": 1.1901,
+      "step": 1400
+    },
+    {
+      "epoch": 0.09386691608765366,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00018770875083500336,
+      "loss": 1.2334,
+      "step": 1405
+    },
+    {
+      "epoch": 0.0942009620523784,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00018837675350701405,
+      "loss": 1.2637,
+      "step": 1410
+    },
+    {
+      "epoch": 0.09453500801710316,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00018904475617902473,
+      "loss": 1.2487,
+      "step": 1415
+    },
+    {
+      "epoch": 0.0948690539818279,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0001897127588510354,
+      "loss": 1.2222,
+      "step": 1420
+    },
+    {
+      "epoch": 0.09520309994655264,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001903807615230461,
+      "loss": 1.2418,
+      "step": 1425
+    },
+    {
+      "epoch": 0.09553714591127739,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001910487641950568,
+      "loss": 1.2197,
+      "step": 1430
+    },
+    {
+      "epoch": 0.09587119187600214,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00019171676686706747,
+      "loss": 1.1918,
+      "step": 1435
+    },
+    {
+      "epoch": 0.09620523784072689,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019238476953907816,
+      "loss": 1.2533,
+      "step": 1440
+    },
+    {
+      "epoch": 0.09653928380545163,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00019305277221108886,
+      "loss": 1.2039,
+      "step": 1445
+    },
+    {
+      "epoch": 0.09687332977017638,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019372077488309953,
+      "loss": 1.218,
+      "step": 1450
+    },
+    {
+      "epoch": 0.09720737573490112,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00019438877755511023,
+      "loss": 1.2209,
+      "step": 1455
+    },
+    {
+      "epoch": 0.09754142169962587,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019505678022712093,
+      "loss": 1.1822,
+      "step": 1460
+    },
+    {
+      "epoch": 0.09787546766435061,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001957247828991316,
+      "loss": 1.2394,
+      "step": 1465
+    },
+    {
+      "epoch": 0.09820951362907536,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0001963927855711423,
+      "loss": 1.2377,
+      "step": 1470
+    },
+    {
+      "epoch": 0.09854355959380011,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00019706078824315297,
+      "loss": 1.2419,
+      "step": 1475
+    },
+    {
+      "epoch": 0.09887760555852486,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019772879091516367,
+      "loss": 1.2013,
+      "step": 1480
+    },
+    {
+      "epoch": 0.0992116515232496,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019839679358717434,
+      "loss": 1.2789,
+      "step": 1485
+    },
+    {
+      "epoch": 0.09954569748797434,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019906479625918504,
+      "loss": 1.2002,
+      "step": 1490
+    },
+    {
+      "epoch": 0.0998797434526991,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019973279893119574,
+      "loss": 1.2699,
+      "step": 1495
+    },
+    {
+      "epoch": 0.10021378941742384,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00019999997552557096,
+      "loss": 1.2202,
+      "step": 1500
+    },
+    {
+      "epoch": 0.10054783538214858,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00019999982595965903,
+      "loss": 1.1978,
+      "step": 1505
+    },
+    {
+      "epoch": 0.10088188134687333,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00019999954042494334,
+      "loss": 1.1724,
+      "step": 1510
+    },
+    {
+      "epoch": 0.10121592731159808,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00019999911892181214,
+      "loss": 1.1925,
+      "step": 1515
+    },
+    {
+      "epoch": 0.10154997327632283,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0001999985614508385,
+      "loss": 1.194,
+      "step": 1520
+    },
+    {
+      "epoch": 0.10188401924104756,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001999978680127804,
+      "loss": 1.2429,
+      "step": 1525
+    },
+    {
+      "epoch": 0.10221806520577231,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00019999703860858073,
+      "loss": 1.2303,
+      "step": 1530
+    },
+    {
+      "epoch": 0.10255211117049706,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0001999960732393672,
+      "loss": 1.2445,
+      "step": 1535
+    },
+    {
+      "epoch": 0.10288615713522181,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001999949719064525,
+      "loss": 1.2214,
+      "step": 1540
+    },
+    {
+      "epoch": 0.10322020309994655,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019999373461133398,
+      "loss": 1.3003,
+      "step": 1545
+    },
+    {
+      "epoch": 0.1035542490646713,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019999236135569408,
+      "loss": 1.1941,
+      "step": 1550
+    },
+    {
+      "epoch": 0.10388829502939605,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00019999085214139994,
+      "loss": 1.1791,
+      "step": 1555
+    },
+    {
+      "epoch": 0.1042223409941208,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00019998920697050364,
+      "loss": 1.2118,
+      "step": 1560
+    },
+    {
+      "epoch": 0.10455638695884553,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0001999874258452421,
+      "loss": 1.252,
+      "step": 1565
+    },
+    {
+      "epoch": 0.10489043292357028,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019998550876803708,
+      "loss": 1.1598,
+      "step": 1570
+    },
+    {
+      "epoch": 0.10522447888829503,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019998345574149526,
+      "loss": 1.2936,
+      "step": 1575
+    },
+    {
+      "epoch": 0.10555852485301978,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001999812667684081,
+      "loss": 1.2732,
+      "step": 1580
+    },
+    {
+      "epoch": 0.10589257081774452,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001999789418517519,
+      "loss": 1.2097,
+      "step": 1585
+    },
+    {
+      "epoch": 0.10622661678246927,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019997648099468786,
+      "loss": 1.3343,
+      "step": 1590
+    },
+    {
+      "epoch": 0.10656066274719402,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.000199973884200562,
+      "loss": 1.2659,
+      "step": 1595
+    },
+    {
+      "epoch": 0.10689470871191876,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019997115147290506,
+      "loss": 1.2585,
+      "step": 1600
+    },
+    {
+      "epoch": 0.1072287546766435,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001999682828154328,
+      "loss": 1.2175,
+      "step": 1605
+    },
+    {
+      "epoch": 0.10756280064136825,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019996527823204567,
+      "loss": 1.208,
+      "step": 1610
+    },
+    {
+      "epoch": 0.107896846606093,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.000199962137726829,
+      "loss": 1.2164,
+      "step": 1615
+    },
+    {
+      "epoch": 0.10823089257081775,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019995886130405287,
+      "loss": 1.2604,
+      "step": 1620
+    },
+    {
+      "epoch": 0.10856493853554249,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00019995544896817222,
+      "loss": 1.276,
+      "step": 1625
+    },
+    {
+      "epoch": 0.10889898450026723,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00019995190072382677,
+      "loss": 1.2216,
+      "step": 1630
+    },
+    {
+      "epoch": 0.10923303046499198,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.000199948216575841,
+      "loss": 1.2231,
+      "step": 1635
+    },
+    {
+      "epoch": 0.10956707642971673,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001999443965292243,
+      "loss": 1.1956,
+      "step": 1640
+    },
+    {
+      "epoch": 0.10990112239444147,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019994044058917063,
+      "loss": 1.2923,
+      "step": 1645
+    },
+    {
+      "epoch": 0.11023516835916622,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019993634876105896,
+      "loss": 1.2428,
+      "step": 1650
+    },
+    {
+      "epoch": 0.11056921432389097,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001999321210504528,
+      "loss": 1.1977,
+      "step": 1655
+    },
+    {
+      "epoch": 0.11090326028861572,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019992775746310062,
+      "loss": 1.201,
+      "step": 1660
+    },
+    {
+      "epoch": 0.11123730625334045,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019992325800493547,
+      "loss": 1.2272,
+      "step": 1665
+    },
+    {
+      "epoch": 0.1115713522180652,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00019991862268207527,
+      "loss": 1.1801,
+      "step": 1670
+    },
+    {
+      "epoch": 0.11190539818278995,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00019991385150082265,
+      "loss": 1.2178,
+      "step": 1675
+    },
+    {
+      "epoch": 0.1122394441475147,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019990894446766485,
+      "loss": 1.2739,
+      "step": 1680
+    },
+    {
+      "epoch": 0.11257349011223944,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00019990390158927402,
+      "loss": 1.2091,
+      "step": 1685
+    },
+    {
+      "epoch": 0.11290753607696419,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019989872287250684,
+      "loss": 1.1765,
+      "step": 1690
+    },
+    {
+      "epoch": 0.11324158204168894,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019989340832440478,
+      "loss": 1.2185,
+      "step": 1695
+    },
+    {
+      "epoch": 0.11357562800641369,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019988795795219396,
+      "loss": 1.2335,
+      "step": 1700
+    },
+    {
+      "epoch": 0.11390967397113842,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019988237176328527,
+      "loss": 1.2053,
+      "step": 1705
+    },
+    {
+      "epoch": 0.11424371993586317,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019987664976527412,
+      "loss": 1.2055,
+      "step": 1710
+    },
+    {
+      "epoch": 0.11457776590058792,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019987079196594069,
+      "loss": 1.25,
+      "step": 1715
+    },
+    {
+      "epoch": 0.11491181186531267,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001998647983732498,
+      "loss": 1.2435,
+      "step": 1720
+    },
+    {
+      "epoch": 0.11524585783003741,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019985866899535079,
+      "loss": 1.16,
+      "step": 1725
+    },
+    {
+      "epoch": 0.11557990379476216,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001998524038405778,
+      "loss": 1.2465,
+      "step": 1730
+    },
+    {
+      "epoch": 0.1159139497594869,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019984600291744948,
+      "loss": 1.2855,
+      "step": 1735
+    },
+    {
+      "epoch": 0.11624799572421166,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001998394662346691,
+      "loss": 1.2028,
+      "step": 1740
+    },
+    {
+      "epoch": 0.11658204168893639,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019983279380112454,
+      "loss": 1.2496,
+      "step": 1745
+    },
+    {
+      "epoch": 0.11691608765366114,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00019982598562588822,
+      "loss": 1.2577,
+      "step": 1750
+    },
+    {
+      "epoch": 0.11725013361838589,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019981904171821716,
+      "loss": 1.2954,
+      "step": 1755
+    },
+    {
+      "epoch": 0.11758417958311064,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001998119620875529,
+      "loss": 1.2704,
+      "step": 1760
+    },
+    {
+      "epoch": 0.11791822554783538,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001998047467435216,
+      "loss": 1.2034,
+      "step": 1765
+    },
+    {
+      "epoch": 0.11825227151256013,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019979739569593385,
+      "loss": 1.2341,
+      "step": 1770
+    },
+    {
+      "epoch": 0.11858631747728487,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019978990895478483,
+      "loss": 1.2971,
+      "step": 1775
+    },
+    {
+      "epoch": 0.11892036344200962,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00019978228653025416,
+      "loss": 1.2409,
+      "step": 1780
+    },
+    {
+      "epoch": 0.11925440940673436,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.000199774528432706,
+      "loss": 1.2282,
+      "step": 1785
+    },
+    {
+      "epoch": 0.11958845537145911,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019976663467268893,
+      "loss": 1.2715,
+      "step": 1790
+    },
+    {
+      "epoch": 0.11992250133618386,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019975860526093604,
+      "loss": 1.2379,
+      "step": 1795
+    },
+    {
+      "epoch": 0.12025654730090861,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019975044020836485,
+      "loss": 1.2745,
+      "step": 1800
+    },
+    {
+      "epoch": 0.12059059326563334,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001997421395260773,
+      "loss": 1.2342,
+      "step": 1805
+    },
+    {
+      "epoch": 0.1209246392303581,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00019973370322535976,
+      "loss": 1.2811,
+      "step": 1810
+    },
+    {
+      "epoch": 0.12125868519508284,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00019972513131768298,
+      "loss": 1.2948,
+      "step": 1815
+    },
+    {
+      "epoch": 0.1215927311598076,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001997164238147021,
+      "loss": 1.2323,
+      "step": 1820
+    },
+    {
+      "epoch": 0.12192677712453233,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019970758072825658,
+      "loss": 1.2297,
+      "step": 1825
+    },
+    {
+      "epoch": 0.12226082308925708,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019969860207037034,
+      "loss": 1.1426,
+      "step": 1830
+    },
+    {
+      "epoch": 0.12259486905398183,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019968948785325158,
+      "loss": 1.3128,
+      "step": 1835
+    },
+    {
+      "epoch": 0.12292891501870658,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019968023808929276,
+      "loss": 1.247,
+      "step": 1840
+    },
+    {
+      "epoch": 0.12326296098343131,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019967085279107077,
+      "loss": 1.1878,
+      "step": 1845
+    },
+    {
+      "epoch": 0.12359700694815606,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00019966133197134664,
+      "loss": 1.1013,
+      "step": 1850
+    },
+    {
+      "epoch": 0.12393105291288081,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00019965167564306576,
+      "loss": 1.2437,
+      "step": 1855
+    },
+    {
+      "epoch": 0.12426509887760556,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019964188381935776,
+      "loss": 1.233,
+      "step": 1860
+    },
+    {
+      "epoch": 0.1245991448423303,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019963195651353645,
+      "loss": 1.2822,
+      "step": 1865
+    },
+    {
+      "epoch": 0.12493319080705505,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019962189373909996,
+      "loss": 1.2954,
+      "step": 1870
+    },
+    {
+      "epoch": 0.1252672367717798,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0001996116955097305,
+      "loss": 1.2188,
+      "step": 1875
+    },
+    {
+      "epoch": 0.12560128273650453,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001996013618392945,
+      "loss": 1.2498,
+      "step": 1880
+    },
+    {
+      "epoch": 0.12593532870122928,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00019959089274184256,
+      "loss": 1.1953,
+      "step": 1885
+    },
+    {
+      "epoch": 0.12626937466595403,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019958028823160946,
+      "loss": 1.2453,
+      "step": 1890
+    },
+    {
+      "epoch": 0.12660342063067878,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019956954832301397,
+      "loss": 1.1966,
+      "step": 1895
+    },
+    {
+      "epoch": 0.12693746659540353,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001995586730306591,
+      "loss": 1.2389,
+      "step": 1900
+    },
+    {
+      "epoch": 0.12727151256012828,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00019954766236933183,
+      "loss": 1.2609,
+      "step": 1905
+    },
+    {
+      "epoch": 0.12760555852485303,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019953651635400333,
+      "loss": 1.2263,
+      "step": 1910
+    },
+    {
+      "epoch": 0.12793960448957778,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019952523499982864,
+      "loss": 1.1892,
+      "step": 1915
+    },
+    {
+      "epoch": 0.1282736504543025,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00019951381832214698,
+      "loss": 1.1666,
+      "step": 1920
+    },
+    {
+      "epoch": 0.12860769641902725,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019950226633648144,
+      "loss": 1.2589,
+      "step": 1925
+    },
+    {
+      "epoch": 0.128941742383752,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00019949057905853919,
+      "loss": 1.1863,
+      "step": 1930
+    },
+    {
+      "epoch": 0.12927578834847675,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019947875650421127,
+      "loss": 1.1823,
+      "step": 1935
+    },
+    {
+      "epoch": 0.1296098343132015,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001994667986895727,
+      "loss": 1.2282,
+      "step": 1940
+    },
+    {
+      "epoch": 0.12994388027792625,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00019945470563088248,
+      "loss": 1.1948,
+      "step": 1945
+    },
+    {
+      "epoch": 0.130277926242651,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019944247734458333,
+      "loss": 1.2441,
+      "step": 1950
+    },
+    {
+      "epoch": 0.13061197220737575,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00019943011384730198,
+      "loss": 1.1913,
+      "step": 1955
+    },
+    {
+      "epoch": 0.13094601817210047,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019941761515584894,
+      "loss": 1.2447,
+      "step": 1960
+    },
+    {
+      "epoch": 0.13128006413682522,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00019940498128721856,
+      "loss": 1.2564,
+      "step": 1965
+    },
+    {
+      "epoch": 0.13161411010154997,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00019939221225858902,
+      "loss": 1.193,
+      "step": 1970
+    },
+    {
+      "epoch": 0.13194815606627472,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00019937930808732222,
+      "loss": 1.3114,
+      "step": 1975
+    },
+    {
+      "epoch": 0.13228220203099947,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019936626879096383,
+      "loss": 1.1698,
+      "step": 1980
+    },
+    {
+      "epoch": 0.13261624799572422,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019935309438724326,
+      "loss": 1.2549,
+      "step": 1985
+    },
+    {
+      "epoch": 0.13295029396044897,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019933978489407366,
+      "loss": 1.1987,
+      "step": 1990
+    },
+    {
+      "epoch": 0.13328433992517372,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019932634032955178,
+      "loss": 1.1868,
+      "step": 1995
+    },
+    {
+      "epoch": 0.13361838588989844,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019931276071195804,
+      "loss": 1.2345,
+      "step": 2000
+    },
+    {
+      "epoch": 0.1339524318546232,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019929904605975657,
+      "loss": 1.2149,
+      "step": 2005
+    },
+    {
+      "epoch": 0.13428647781934794,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00019928519639159507,
+      "loss": 1.2072,
+      "step": 2010
+    },
+    {
+      "epoch": 0.1346205237840727,
+      "grad_norm": 31.375,
+      "learning_rate": 0.00019927121172630473,
+      "loss": 1.2617,
+      "step": 2015
+    },
+    {
+      "epoch": 0.13495456974879744,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001992570920829004,
+      "loss": 1.2062,
+      "step": 2020
+    },
+    {
+      "epoch": 0.1352886157135222,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001992428374805804,
+      "loss": 1.1609,
+      "step": 2025
+    },
+    {
+      "epoch": 0.13562266167824694,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001992284479387266,
+      "loss": 1.2839,
+      "step": 2030
+    },
+    {
+      "epoch": 0.13595670764297169,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019921392347690435,
+      "loss": 1.1998,
+      "step": 2035
+    },
+    {
+      "epoch": 0.1362907536076964,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001991992641148624,
+      "loss": 1.2281,
+      "step": 2040
+    },
+    {
+      "epoch": 0.13662479957242116,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019918446987253287,
+      "loss": 1.1933,
+      "step": 2045
+    },
+    {
+      "epoch": 0.1369588455371459,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019916954077003147,
+      "loss": 1.1923,
+      "step": 2050
+    },
+    {
+      "epoch": 0.13729289150187066,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00019915447682765705,
+      "loss": 1.1975,
+      "step": 2055
+    },
+    {
+      "epoch": 0.1376269374665954,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.000199139278065892,
+      "loss": 1.23,
+      "step": 2060
+    },
+    {
+      "epoch": 0.13796098343132016,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0001991239445054019,
+      "loss": 1.2278,
+      "step": 2065
+    },
+    {
+      "epoch": 0.1382950293960449,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001991084761670356,
+      "loss": 1.1755,
+      "step": 2070
+    },
+    {
+      "epoch": 0.13862907536076965,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00019909287307182534,
+      "loss": 1.1468,
+      "step": 2075
+    },
+    {
+      "epoch": 0.13896312132549438,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019907713524098638,
+      "loss": 1.2159,
+      "step": 2080
+    },
+    {
+      "epoch": 0.13929716729021913,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001990612626959174,
+      "loss": 1.1975,
+      "step": 2085
+    },
+    {
+      "epoch": 0.13963121325494388,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0001990452554582001,
+      "loss": 1.2577,
+      "step": 2090
+    },
+    {
+      "epoch": 0.13996525921966863,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00019902911354959936,
+      "loss": 1.1977,
+      "step": 2095
+    },
+    {
+      "epoch": 0.14029930518439337,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019901283699206323,
+      "loss": 1.2434,
+      "step": 2100
+    },
+    {
+      "epoch": 0.14063335114911812,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00019899642580772274,
+      "loss": 1.2707,
+      "step": 2105
+    },
+    {
+      "epoch": 0.14096739711384287,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.000198979880018892,
+      "loss": 1.1902,
+      "step": 2110
+    },
+    {
+      "epoch": 0.14130144307856762,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00019896319964806823,
+      "loss": 1.215,
+      "step": 2115
+    },
+    {
+      "epoch": 0.14163548904329235,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019894638471793153,
+      "loss": 1.2495,
+      "step": 2120
+    },
+    {
+      "epoch": 0.1419695350080171,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.000198929435251345,
+      "loss": 1.2628,
+      "step": 2125
+    },
+    {
+      "epoch": 0.14230358097274184,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019891235127135465,
+      "loss": 1.2156,
+      "step": 2130
+    },
+    {
+      "epoch": 0.1426376269374666,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019889513280118946,
+      "loss": 1.1396,
+      "step": 2135
+    },
+    {
+      "epoch": 0.14297167290219134,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019887777986426117,
+      "loss": 1.2356,
+      "step": 2140
+    },
+    {
+      "epoch": 0.1433057188669161,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00019886029248416441,
+      "loss": 1.2302,
+      "step": 2145
+    },
+    {
+      "epoch": 0.14363976483164084,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019884267068467662,
+      "loss": 1.199,
+      "step": 2150
+    },
+    {
+      "epoch": 0.1439738107963656,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019882491448975796,
+      "loss": 1.3024,
+      "step": 2155
+    },
+    {
+      "epoch": 0.14430785676109031,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019880702392355138,
+      "loss": 1.188,
+      "step": 2160
+    },
+    {
+      "epoch": 0.14464190272581506,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00019878899901038254,
+      "loss": 1.2895,
+      "step": 2165
+    },
+    {
+      "epoch": 0.1449759486905398,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00019877083977475968,
+      "loss": 1.2214,
+      "step": 2170
+    },
+    {
+      "epoch": 0.14530999465526456,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00019875254624137376,
+      "loss": 1.2429,
+      "step": 2175
+    },
+    {
+      "epoch": 0.1456440406199893,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00019873411843509832,
+      "loss": 1.2168,
+      "step": 2180
+    },
+    {
+      "epoch": 0.14597808658471406,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019871555638098954,
+      "loss": 1.2398,
+      "step": 2185
+    },
+    {
+      "epoch": 0.1463121325494388,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019869686010428597,
+      "loss": 1.2839,
+      "step": 2190
+    },
+    {
+      "epoch": 0.14664617851416356,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00019867802963040881,
+      "loss": 1.2139,
+      "step": 2195
+    },
+    {
+      "epoch": 0.14698022447888828,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019865906498496162,
+      "loss": 1.2241,
+      "step": 2200
+    },
+    {
+      "epoch": 0.14731427044361303,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019863996619373054,
+      "loss": 1.2731,
+      "step": 2205
+    },
+    {
+      "epoch": 0.14764831640833778,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019862073328268394,
+      "loss": 1.2639,
+      "step": 2210
+    },
+    {
+      "epoch": 0.14798236237306253,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019860136627797262,
+      "loss": 1.2519,
+      "step": 2215
+    },
+    {
+      "epoch": 0.14831640833778728,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001985818652059298,
+      "loss": 1.2553,
+      "step": 2220
+    },
+    {
+      "epoch": 0.14865045430251203,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001985622300930708,
+      "loss": 1.2233,
+      "step": 2225
+    },
+    {
+      "epoch": 0.14898450026723678,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001985424609660933,
+      "loss": 1.1913,
+      "step": 2230
+    },
+    {
+      "epoch": 0.14931854623196153,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019852255785187724,
+      "loss": 1.217,
+      "step": 2235
+    },
+    {
+      "epoch": 0.14965259219668625,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00019850252077748467,
+      "loss": 1.2496,
+      "step": 2240
+    },
+    {
+      "epoch": 0.149986638161411,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019848234977015984,
+      "loss": 1.251,
+      "step": 2245
+    },
+    {
+      "epoch": 0.15032068412613575,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019846204485732903,
+      "loss": 1.2931,
+      "step": 2250
+    },
+    {
+      "epoch": 0.1506547300908605,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001984416060666007,
+      "loss": 1.2315,
+      "step": 2255
+    },
+    {
+      "epoch": 0.15098877605558525,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019842103342576517,
+      "loss": 1.1459,
+      "step": 2260
+    },
+    {
+      "epoch": 0.15132282202031,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00019840032696279494,
+      "loss": 1.2563,
+      "step": 2265
+    },
+    {
+      "epoch": 0.15165686798503475,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019837948670584437,
+      "loss": 1.212,
+      "step": 2270
+    },
+    {
+      "epoch": 0.1519909139497595,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019835851268324982,
+      "loss": 1.3146,
+      "step": 2275
+    },
+    {
+      "epoch": 0.15232495991448422,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019833740492352934,
+      "loss": 1.2306,
+      "step": 2280
+    },
+    {
+      "epoch": 0.15265900587920897,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019831616345538305,
+      "loss": 1.2235,
+      "step": 2285
+    },
+    {
+      "epoch": 0.15299305184393372,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001982947883076927,
+      "loss": 1.1821,
+      "step": 2290
+    },
+    {
+      "epoch": 0.15332709780865847,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00019827327950952195,
+      "loss": 1.2706,
+      "step": 2295
+    },
+    {
+      "epoch": 0.15366114377338322,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019825163709011605,
+      "loss": 1.1829,
+      "step": 2300
+    },
+    {
+      "epoch": 0.15399518973810797,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.000198229861078902,
+      "loss": 1.194,
+      "step": 2305
+    },
+    {
+      "epoch": 0.15432923570283272,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019820795150548846,
+      "loss": 1.2001,
+      "step": 2310
+    },
+    {
+      "epoch": 0.15466328166755747,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019818590839966563,
+      "loss": 1.1742,
+      "step": 2315
+    },
+    {
+      "epoch": 0.1549973276322822,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019816373179140534,
+      "loss": 1.2324,
+      "step": 2320
+    },
+    {
+      "epoch": 0.15533137359700694,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019814142171086088,
+      "loss": 1.1786,
+      "step": 2325
+    },
+    {
+      "epoch": 0.1556654195617317,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001981189781883671,
+      "loss": 1.2989,
+      "step": 2330
+    },
+    {
+      "epoch": 0.15599946552645644,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019809640125444016,
+      "loss": 1.2714,
+      "step": 2335
+    },
+    {
+      "epoch": 0.1563335114911812,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00019807369093977778,
+      "loss": 1.2568,
+      "step": 2340
+    },
+    {
+      "epoch": 0.15666755745590594,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019805084727525895,
+      "loss": 1.2131,
+      "step": 2345
+    },
+    {
+      "epoch": 0.1570016034206307,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019802787029194393,
+      "loss": 1.1464,
+      "step": 2350
+    },
+    {
+      "epoch": 0.15733564938535544,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00019800476002107437,
+      "loss": 1.2525,
+      "step": 2355
+    },
+    {
+      "epoch": 0.15766969535008016,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.000197981516494073,
+      "loss": 1.2007,
+      "step": 2360
+    },
+    {
+      "epoch": 0.1580037413148049,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001979581397425439,
+      "loss": 1.2535,
+      "step": 2365
+    },
+    {
+      "epoch": 0.15833778727952966,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001979346297982722,
+      "loss": 1.235,
+      "step": 2370
+    },
+    {
+      "epoch": 0.1586718332442544,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001979109866932241,
+      "loss": 1.162,
+      "step": 2375
+    },
+    {
+      "epoch": 0.15900587920897916,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00019788721045954692,
+      "loss": 1.2224,
+      "step": 2380
+    },
+    {
+      "epoch": 0.1593399251737039,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.000197863301129569,
+      "loss": 1.3007,
+      "step": 2385
+    },
+    {
+      "epoch": 0.15967397113842866,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019783925873579966,
+      "loss": 1.2242,
+      "step": 2390
+    },
+    {
+      "epoch": 0.1600080171031534,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.000197815083310929,
+      "loss": 1.2477,
+      "step": 2395
+    },
+    {
+      "epoch": 0.16034206306787813,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019779077488782824,
+      "loss": 1.1975,
+      "step": 2400
+    },
+    {
+      "epoch": 0.16067610903260288,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001977663334995492,
+      "loss": 1.2646,
+      "step": 2405
+    },
+    {
+      "epoch": 0.16101015499732763,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001977417591793247,
+      "loss": 1.2167,
+      "step": 2410
+    },
+    {
+      "epoch": 0.16134420096205238,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019771705196056812,
+      "loss": 1.2048,
+      "step": 2415
+    },
+    {
+      "epoch": 0.16167824692677712,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00019769221187687368,
+      "loss": 1.183,
+      "step": 2420
+    },
+    {
+      "epoch": 0.16201229289150187,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001976672389620162,
+      "loss": 1.281,
+      "step": 2425
+    },
+    {
+      "epoch": 0.16234633885622662,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001976421332499511,
+      "loss": 1.2045,
+      "step": 2430
+    },
+    {
+      "epoch": 0.16268038482095137,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00019761689477481434,
+      "loss": 1.243,
+      "step": 2435
+    },
+    {
+      "epoch": 0.1630144307856761,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0001975915235709225,
+      "loss": 1.1954,
+      "step": 2440
+    },
+    {
+      "epoch": 0.16334847675040085,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00019756601967277256,
+      "loss": 1.1924,
+      "step": 2445
+    },
+    {
+      "epoch": 0.1636825227151256,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019754038311504187,
+      "loss": 1.1605,
+      "step": 2450
+    },
+    {
+      "epoch": 0.16401656867985034,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019751461393258829,
+      "loss": 1.3125,
+      "step": 2455
+    },
+    {
+      "epoch": 0.1643506146445751,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019748871216044984,
+      "loss": 1.2302,
+      "step": 2460
+    },
+    {
+      "epoch": 0.16468466060929984,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019746267783384496,
+      "loss": 1.2435,
+      "step": 2465
+    },
+    {
+      "epoch": 0.1650187065740246,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00019743651098817227,
+      "loss": 1.1917,
+      "step": 2470
+    },
+    {
+      "epoch": 0.16535275253874934,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019741021165901054,
+      "loss": 1.2182,
+      "step": 2475
+    },
+    {
+      "epoch": 0.16568679850347406,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00019738377988211877,
+      "loss": 1.2176,
+      "step": 2480
+    },
+    {
+      "epoch": 0.1660208444681988,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.000197357215693436,
+      "loss": 1.2962,
+      "step": 2485
+    },
+    {
+      "epoch": 0.16635489043292356,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00019733051912908126,
+      "loss": 1.2024,
+      "step": 2490
+    },
+    {
+      "epoch": 0.1666889363976483,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00019730369022535362,
+      "loss": 1.1819,
+      "step": 2495
+    },
+    {
+      "epoch": 0.16702298236237306,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001972767290187321,
+      "loss": 1.1487,
+      "step": 2500
+    },
+    {
+      "epoch": 0.1673570283270978,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001972496355458756,
+      "loss": 1.2591,
+      "step": 2505
+    },
+    {
+      "epoch": 0.16769107429182256,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019722240984362284,
+      "loss": 1.2167,
+      "step": 2510
+    },
+    {
+      "epoch": 0.1680251202565473,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00019719505194899233,
+      "loss": 1.2727,
+      "step": 2515
+    },
+    {
+      "epoch": 0.16835916622127206,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00019716756189918234,
+      "loss": 1.2353,
+      "step": 2520
+    },
+    {
+      "epoch": 0.16869321218599678,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001971399397315709,
+      "loss": 1.1215,
+      "step": 2525
+    },
+    {
+      "epoch": 0.16902725815072153,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019711218548371546,
+      "loss": 1.2287,
+      "step": 2530
+    },
+    {
+      "epoch": 0.16936130411544628,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019708429919335335,
+      "loss": 1.2297,
+      "step": 2535
+    },
+    {
+      "epoch": 0.16969535008017103,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019705628089840122,
+      "loss": 1.252,
+      "step": 2540
+    },
+    {
+      "epoch": 0.17002939604489578,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001970281306369553,
+      "loss": 1.2608,
+      "step": 2545
+    },
+    {
+      "epoch": 0.17036344200962053,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001969998484472912,
+      "loss": 1.1916,
+      "step": 2550
+    },
+    {
+      "epoch": 0.17069748797434528,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019697143436786397,
+      "loss": 1.1868,
+      "step": 2555
+    },
+    {
+      "epoch": 0.17103153393907003,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019694288843730796,
+      "loss": 1.1567,
+      "step": 2560
+    },
+    {
+      "epoch": 0.17136557990379475,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001969142106944368,
+      "loss": 1.2043,
+      "step": 2565
+    },
+    {
+      "epoch": 0.1716996258685195,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019688540117824332,
+      "loss": 1.1954,
+      "step": 2570
+    },
+    {
+      "epoch": 0.17203367183324425,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00019685645992789956,
+      "loss": 1.1889,
+      "step": 2575
+    },
+    {
+      "epoch": 0.172367717797969,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019682738698275663,
+      "loss": 1.1758,
+      "step": 2580
+    },
+    {
+      "epoch": 0.17270176376269375,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001967981823823448,
+      "loss": 1.2197,
+      "step": 2585
+    },
+    {
+      "epoch": 0.1730358097274185,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001967688461663732,
+      "loss": 1.1922,
+      "step": 2590
+    },
+    {
+      "epoch": 0.17336985569214325,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001967393783747301,
+      "loss": 1.2168,
+      "step": 2595
+    },
+    {
+      "epoch": 0.173703901656868,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019670977904748252,
+      "loss": 1.2387,
+      "step": 2600
+    },
+    {
+      "epoch": 0.17403794762159272,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019668004822487634,
+      "loss": 1.186,
+      "step": 2605
+    },
+    {
+      "epoch": 0.17437199358631747,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00019665018594733634,
+      "loss": 1.2346,
+      "step": 2610
+    },
+    {
+      "epoch": 0.17470603955104222,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019662019225546594,
+      "loss": 1.2375,
+      "step": 2615
+    },
+    {
+      "epoch": 0.17504008551576697,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019659006719004727,
+      "loss": 1.221,
+      "step": 2620
+    },
+    {
+      "epoch": 0.17537413148049172,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019655981079204113,
+      "loss": 1.1965,
+      "step": 2625
+    },
+    {
+      "epoch": 0.17570817744521647,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001965294231025868,
+      "loss": 1.1957,
+      "step": 2630
+    },
+    {
+      "epoch": 0.17604222340994122,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00019649890416300217,
+      "loss": 1.2214,
+      "step": 2635
+    },
+    {
+      "epoch": 0.17637626937466597,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019646825401478356,
+      "loss": 1.2766,
+      "step": 2640
+    },
+    {
+      "epoch": 0.1767103153393907,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00019643747269960566,
+      "loss": 1.241,
+      "step": 2645
+    },
+    {
+      "epoch": 0.17704436130411544,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001964065602593215,
+      "loss": 1.1926,
+      "step": 2650
+    },
+    {
+      "epoch": 0.1773784072688402,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001963755167359625,
+      "loss": 1.2595,
+      "step": 2655
+    },
+    {
+      "epoch": 0.17771245323356494,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019634434217173817,
+      "loss": 1.2557,
+      "step": 2660
+    },
+    {
+      "epoch": 0.1780464991982897,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001963130366090363,
+      "loss": 1.2221,
+      "step": 2665
+    },
+    {
+      "epoch": 0.17838054516301444,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019628160009042275,
+      "loss": 1.2297,
+      "step": 2670
+    },
+    {
+      "epoch": 0.1787145911277392,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019625003265864147,
+      "loss": 1.2172,
+      "step": 2675
+    },
+    {
+      "epoch": 0.17904863709246394,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00019621833435661437,
+      "loss": 1.2086,
+      "step": 2680
+    },
+    {
+      "epoch": 0.17938268305718866,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00019618650522744137,
+      "loss": 1.1887,
+      "step": 2685
+    },
+    {
+      "epoch": 0.1797167290219134,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019615454531440017,
+      "loss": 1.2667,
+      "step": 2690
+    },
+    {
+      "epoch": 0.18005077498663816,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00019612245466094641,
+      "loss": 1.2165,
+      "step": 2695
+    },
+    {
+      "epoch": 0.1803848209513629,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019609023331071344,
+      "loss": 1.2485,
+      "step": 2700
+    },
+    {
+      "epoch": 0.18071886691608766,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001960578813075123,
+      "loss": 1.2353,
+      "step": 2705
+    },
+    {
+      "epoch": 0.1810529128808124,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019602539869533167,
+      "loss": 1.2373,
+      "step": 2710
+    },
+    {
+      "epoch": 0.18138695884553716,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019599278551833788,
+      "loss": 1.1801,
+      "step": 2715
+    },
+    {
+      "epoch": 0.1817210048102619,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00019596004182087477,
+      "loss": 1.2516,
+      "step": 2720
+    },
+    {
+      "epoch": 0.18205505077498663,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00019592716764746363,
+      "loss": 1.2321,
+      "step": 2725
+    },
+    {
+      "epoch": 0.18238909673971138,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019589416304280314,
+      "loss": 1.2326,
+      "step": 2730
+    },
+    {
+      "epoch": 0.18272314270443613,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019586102805176932,
+      "loss": 1.2251,
+      "step": 2735
+    },
+    {
+      "epoch": 0.18305718866916088,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019582776271941557,
+      "loss": 1.1592,
+      "step": 2740
+    },
+    {
+      "epoch": 0.18339123463388562,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019579436709097237,
+      "loss": 1.2147,
+      "step": 2745
+    },
+    {
+      "epoch": 0.18372528059861037,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00019576084121184745,
+      "loss": 1.1742,
+      "step": 2750
+    },
+    {
+      "epoch": 0.18405932656333512,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019572718512762566,
+      "loss": 1.1736,
+      "step": 2755
+    },
+    {
+      "epoch": 0.18439337252805987,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00019569339888406883,
+      "loss": 1.1304,
+      "step": 2760
+    },
+    {
+      "epoch": 0.1847274184927846,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001956594825271158,
+      "loss": 1.2708,
+      "step": 2765
+    },
+    {
+      "epoch": 0.18506146445750934,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019562543610288232,
+      "loss": 1.1739,
+      "step": 2770
+    },
+    {
+      "epoch": 0.1853955104222341,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019559125965766096,
+      "loss": 1.2504,
+      "step": 2775
+    },
+    {
+      "epoch": 0.18572955638695884,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001955569532379211,
+      "loss": 1.1925,
+      "step": 2780
+    },
+    {
+      "epoch": 0.1860636023516836,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019552251689030893,
+      "loss": 1.2344,
+      "step": 2785
+    },
+    {
+      "epoch": 0.18639764831640834,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019548795066164713,
+      "loss": 1.1737,
+      "step": 2790
+    },
+    {
+      "epoch": 0.1867316942811331,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019545325459893512,
+      "loss": 1.2059,
+      "step": 2795
+    },
+    {
+      "epoch": 0.18706574024585784,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001954184287493488,
+      "loss": 1.1795,
+      "step": 2800
+    },
+    {
+      "epoch": 0.18739978621058256,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00019538347316024052,
+      "loss": 1.266,
+      "step": 2805
+    },
+    {
+      "epoch": 0.1877338321753073,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00019534838787913902,
+      "loss": 1.2198,
+      "step": 2810
+    },
+    {
+      "epoch": 0.18806787814003206,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001953131729537495,
+      "loss": 1.2272,
+      "step": 2815
+    },
+    {
+      "epoch": 0.1884019241047568,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019527782843195335,
+      "loss": 1.1629,
+      "step": 2820
+    },
+    {
+      "epoch": 0.18873597006948156,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019524235436180814,
+      "loss": 1.2099,
+      "step": 2825
+    },
+    {
+      "epoch": 0.1890700160342063,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019520675079154763,
+      "loss": 1.2457,
+      "step": 2830
+    },
+    {
+      "epoch": 0.18940406199893106,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019517101776958166,
+      "loss": 1.1824,
+      "step": 2835
+    },
+    {
+      "epoch": 0.1897381079636558,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019513515534449606,
+      "loss": 1.2036,
+      "step": 2840
+    },
+    {
+      "epoch": 0.19007215392838053,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00019509916356505268,
+      "loss": 1.1686,
+      "step": 2845
+    },
+    {
+      "epoch": 0.19040619989310528,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001950630424801891,
+      "loss": 1.1715,
+      "step": 2850
+    },
+    {
+      "epoch": 0.19074024585783003,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00019502679213901893,
+      "loss": 1.3139,
+      "step": 2855
+    },
+    {
+      "epoch": 0.19107429182255478,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019499041259083132,
+      "loss": 1.1451,
+      "step": 2860
+    },
+    {
+      "epoch": 0.19140833778727953,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00019495390388509122,
+      "loss": 1.2861,
+      "step": 2865
+    },
+    {
+      "epoch": 0.19174238375200428,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019491726607143918,
+      "loss": 1.2522,
+      "step": 2870
+    },
+    {
+      "epoch": 0.19207642971672903,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00019488049919969127,
+      "loss": 1.2232,
+      "step": 2875
+    },
+    {
+      "epoch": 0.19241047568145378,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00019484360331983907,
+      "loss": 1.265,
+      "step": 2880
+    },
+    {
+      "epoch": 0.1927445216461785,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019480657848204954,
+      "loss": 1.1885,
+      "step": 2885
+    },
+    {
+      "epoch": 0.19307856761090325,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019476942473666497,
+      "loss": 1.2748,
+      "step": 2890
+    },
+    {
+      "epoch": 0.193412613575628,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.000194732142134203,
+      "loss": 1.223,
+      "step": 2895
+    },
+    {
+      "epoch": 0.19374665954035275,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019469473072535642,
+      "loss": 1.1482,
+      "step": 2900
+    },
+    {
+      "epoch": 0.1940807055050775,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0001946571905609931,
+      "loss": 1.1468,
+      "step": 2905
+    },
+    {
+      "epoch": 0.19441475146980225,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00019461952169215615,
+      "loss": 1.334,
+      "step": 2910
+    },
+    {
+      "epoch": 0.194748797434527,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00019458172417006347,
+      "loss": 1.2264,
+      "step": 2915
+    },
+    {
+      "epoch": 0.19508284339925175,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00019454379804610805,
+      "loss": 1.2272,
+      "step": 2920
+    },
+    {
+      "epoch": 0.19541688936397647,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019450574337185765,
+      "loss": 1.235,
+      "step": 2925
+    },
+    {
+      "epoch": 0.19575093532870122,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019446756019905482,
+      "loss": 1.2242,
+      "step": 2930
+    },
+    {
+      "epoch": 0.19608498129342597,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019442924857961694,
+      "loss": 1.2805,
+      "step": 2935
+    },
+    {
+      "epoch": 0.19641902725815072,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019439080856563585,
+      "loss": 1.2176,
+      "step": 2940
+    },
+    {
+      "epoch": 0.19675307322287547,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019435224020937812,
+      "loss": 1.1892,
+      "step": 2945
+    },
+    {
+      "epoch": 0.19708711918760022,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001943135435632848,
+      "loss": 1.3424,
+      "step": 2950
+    },
+    {
+      "epoch": 0.19742116515232497,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019427471867997128,
+      "loss": 1.2189,
+      "step": 2955
+    },
+    {
+      "epoch": 0.19775521111704972,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019423576561222744,
+      "loss": 1.2175,
+      "step": 2960
+    },
+    {
+      "epoch": 0.19808925708177444,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019419668441301733,
+      "loss": 1.2744,
+      "step": 2965
+    },
+    {
+      "epoch": 0.1984233030464992,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019415747513547936,
+      "loss": 1.2161,
+      "step": 2970
+    },
+    {
+      "epoch": 0.19875734901122394,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019411813783292594,
+      "loss": 1.1341,
+      "step": 2975
+    },
+    {
+      "epoch": 0.1990913949759487,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019407867255884367,
+      "loss": 1.2168,
+      "step": 2980
+    },
+    {
+      "epoch": 0.19942544094067344,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001940390793668931,
+      "loss": 1.2508,
+      "step": 2985
+    },
+    {
+      "epoch": 0.1997594869053982,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019399935831090868,
+      "loss": 1.2179,
+      "step": 2990
+    },
+    {
+      "epoch": 0.20009353287012294,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00019395950944489876,
+      "loss": 1.292,
+      "step": 2995
+    },
+    {
+      "epoch": 0.20042757883484769,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001939195328230455,
+      "loss": 1.2248,
+      "step": 3000
+    },
+    {
+      "epoch": 0.2007616247995724,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019387942849970465,
+      "loss": 1.2645,
+      "step": 3005
+    },
+    {
+      "epoch": 0.20109567076429716,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001938391965294058,
+      "loss": 1.1764,
+      "step": 3010
+    },
+    {
+      "epoch": 0.2014297167290219,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00019379883696685183,
+      "loss": 1.1976,
+      "step": 3015
+    },
+    {
+      "epoch": 0.20176376269374666,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019375834986691933,
+      "loss": 1.2362,
+      "step": 3020
+    },
+    {
+      "epoch": 0.2020978086584714,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001937177352846582,
+      "loss": 1.2767,
+      "step": 3025
+    },
+    {
+      "epoch": 0.20243185462319616,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001936769932752917,
+      "loss": 1.2597,
+      "step": 3030
+    },
+    {
+      "epoch": 0.2027659005879209,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019363612389421638,
+      "loss": 1.2247,
+      "step": 3035
+    },
+    {
+      "epoch": 0.20309994655264565,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019359512719700192,
+      "loss": 1.1646,
+      "step": 3040
+    },
+    {
+      "epoch": 0.20343399251737038,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00019355400323939112,
+      "loss": 1.1893,
+      "step": 3045
+    },
+    {
+      "epoch": 0.20376803848209513,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00019351275207729984,
+      "loss": 1.2971,
+      "step": 3050
+    },
+    {
+      "epoch": 0.20410208444681988,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001934713737668169,
+      "loss": 1.2324,
+      "step": 3055
+    },
+    {
+      "epoch": 0.20443613041154463,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.000193429868364204,
+      "loss": 1.1936,
+      "step": 3060
+    },
+    {
+      "epoch": 0.20477017637626937,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001933882359258956,
+      "loss": 1.2243,
+      "step": 3065
+    },
+    {
+      "epoch": 0.20510422234099412,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00019334647650849897,
+      "loss": 1.1963,
+      "step": 3070
+    },
+    {
+      "epoch": 0.20543826830571887,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00019330459016879395,
+      "loss": 1.2023,
+      "step": 3075
+    },
+    {
+      "epoch": 0.20577231427044362,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019326257696373304,
+      "loss": 1.2018,
+      "step": 3080
+    },
+    {
+      "epoch": 0.20610636023516835,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019322043695044116,
+      "loss": 1.1572,
+      "step": 3085
+    },
+    {
+      "epoch": 0.2064404061998931,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001931781701862157,
+      "loss": 1.2123,
+      "step": 3090
+    },
+    {
+      "epoch": 0.20677445216461784,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019313577672852632,
+      "loss": 1.2391,
+      "step": 3095
+    },
+    {
+      "epoch": 0.2071084981293426,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019309325663501508,
+      "loss": 1.2305,
+      "step": 3100
+    },
+    {
+      "epoch": 0.20744254409406734,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00019305060996349606,
+      "loss": 1.1387,
+      "step": 3105
+    },
+    {
+      "epoch": 0.2077765900587921,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019300783677195563,
+      "loss": 1.2658,
+      "step": 3110
+    },
+    {
+      "epoch": 0.20811063602351684,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019296493711855198,
+      "loss": 1.2224,
+      "step": 3115
+    },
+    {
+      "epoch": 0.2084446819882416,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019292191106161542,
+      "loss": 1.1874,
+      "step": 3120
+    },
+    {
+      "epoch": 0.20877872795296634,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019287875865964808,
+      "loss": 1.1604,
+      "step": 3125
+    },
+    {
+      "epoch": 0.20911277391769106,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019283547997132381,
+      "loss": 1.2352,
+      "step": 3130
+    },
+    {
+      "epoch": 0.2094468198824158,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019279207505548825,
+      "loss": 1.2506,
+      "step": 3135
+    },
+    {
+      "epoch": 0.20978086584714056,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019274854397115866,
+      "loss": 1.2556,
+      "step": 3140
+    },
+    {
+      "epoch": 0.2101149118118653,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019270488677752387,
+      "loss": 1.2294,
+      "step": 3145
+    },
+    {
+      "epoch": 0.21044895777659006,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0001926611035339441,
+      "loss": 1.1891,
+      "step": 3150
+    },
+    {
+      "epoch": 0.2107830037413148,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019261719429995098,
+      "loss": 1.2341,
+      "step": 3155
+    },
+    {
+      "epoch": 0.21111704970603956,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00019257315913524754,
+      "loss": 1.2232,
+      "step": 3160
+    },
+    {
+      "epoch": 0.2114510956707643,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00019252899809970794,
+      "loss": 1.3083,
+      "step": 3165
+    },
+    {
+      "epoch": 0.21178514163548903,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019248471125337752,
+      "loss": 1.2538,
+      "step": 3170
+    },
+    {
+      "epoch": 0.21211918760021378,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019244029865647267,
+      "loss": 1.2017,
+      "step": 3175
+    },
+    {
+      "epoch": 0.21245323356493853,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019239576036938078,
+      "loss": 1.2138,
+      "step": 3180
+    },
+    {
+      "epoch": 0.21278727952966328,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019235109645266015,
+      "loss": 1.1828,
+      "step": 3185
+    },
+    {
+      "epoch": 0.21312132549438803,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019230630696703984,
+      "loss": 1.208,
+      "step": 3190
+    },
+    {
+      "epoch": 0.21345537145911278,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001922613919734197,
+      "loss": 1.141,
+      "step": 3195
+    },
+    {
+      "epoch": 0.21378941742383753,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019221635153287024,
+      "loss": 1.2047,
+      "step": 3200
+    },
+    {
+      "epoch": 0.21412346338856228,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019217118570663249,
+      "loss": 1.2994,
+      "step": 3205
+    },
+    {
+      "epoch": 0.214457509353287,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.000192125894556118,
+      "loss": 1.1787,
+      "step": 3210
+    },
+    {
+      "epoch": 0.21479155531801175,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001920804781429087,
+      "loss": 1.2072,
+      "step": 3215
+    },
+    {
+      "epoch": 0.2151256012827365,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019203493652875686,
+      "loss": 1.1978,
+      "step": 3220
+    },
+    {
+      "epoch": 0.21545964724746125,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019198926977558495,
+      "loss": 1.1938,
+      "step": 3225
+    },
+    {
+      "epoch": 0.215793693212186,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00019194347794548565,
+      "loss": 1.253,
+      "step": 3230
+    },
+    {
+      "epoch": 0.21612773917691075,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001918975611007217,
+      "loss": 1.198,
+      "step": 3235
+    },
+    {
+      "epoch": 0.2164617851416355,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019185151930372574,
+      "loss": 1.2049,
+      "step": 3240
+    },
+    {
+      "epoch": 0.21679583110636025,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00019180535261710037,
+      "loss": 1.2952,
+      "step": 3245
+    },
+    {
+      "epoch": 0.21712987707108497,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.000191759061103618,
+      "loss": 1.1782,
+      "step": 3250
+    },
+    {
+      "epoch": 0.21746392303580972,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00019171264482622076,
+      "loss": 1.1968,
+      "step": 3255
+    },
+    {
+      "epoch": 0.21779796900053447,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019166610384802038,
+      "loss": 1.2566,
+      "step": 3260
+    },
+    {
+      "epoch": 0.21813201496525922,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019161943823229824,
+      "loss": 1.2565,
+      "step": 3265
+    },
+    {
+      "epoch": 0.21846606092998397,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019157264804250506,
+      "loss": 1.1353,
+      "step": 3270
+    },
+    {
+      "epoch": 0.21880010689470872,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00019152573334226114,
+      "loss": 1.2901,
+      "step": 3275
+    },
+    {
+      "epoch": 0.21913415285943347,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019147869419535577,
+      "loss": 1.1798,
+      "step": 3280
+    },
+    {
+      "epoch": 0.21946819882415822,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001914315306657478,
+      "loss": 1.2607,
+      "step": 3285
+    },
+    {
+      "epoch": 0.21980224478888294,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001913842428175649,
+      "loss": 1.1978,
+      "step": 3290
+    },
+    {
+      "epoch": 0.2201362907536077,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00019133683071510395,
+      "loss": 1.1645,
+      "step": 3295
+    },
+    {
+      "epoch": 0.22047033671833244,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00019128929442283074,
+      "loss": 1.1069,
+      "step": 3300
+    },
+    {
+      "epoch": 0.2208043826830572,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0001912416340053799,
+      "loss": 1.279,
+      "step": 3305
+    },
+    {
+      "epoch": 0.22113842864778194,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00019119384952755484,
+      "loss": 1.2085,
+      "step": 3310
+    },
+    {
+      "epoch": 0.2214724746125067,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019114594105432766,
+      "loss": 1.1412,
+      "step": 3315
+    },
+    {
+      "epoch": 0.22180652057723144,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019109790865083905,
+      "loss": 1.1815,
+      "step": 3320
+    },
+    {
+      "epoch": 0.22214056654195619,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019104975238239818,
+      "loss": 1.238,
+      "step": 3325
+    },
+    {
+      "epoch": 0.2224746125066809,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019100147231448274,
+      "loss": 1.2348,
+      "step": 3330
+    },
+    {
+      "epoch": 0.22280865847140566,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001909530685127386,
+      "loss": 1.2156,
+      "step": 3335
+    },
+    {
+      "epoch": 0.2231427044361304,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019090454104298004,
+      "loss": 1.1559,
+      "step": 3340
+    },
+    {
+      "epoch": 0.22347675040085516,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019085588997118927,
+      "loss": 1.1645,
+      "step": 3345
+    },
+    {
+      "epoch": 0.2238107963655799,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00019080711536351676,
+      "loss": 1.2397,
+      "step": 3350
+    },
+    {
+      "epoch": 0.22414484233030466,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019075821728628087,
+      "loss": 1.2215,
+      "step": 3355
+    },
+    {
+      "epoch": 0.2244788882950294,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00019070919580596783,
+      "loss": 1.242,
+      "step": 3360
+    },
+    {
+      "epoch": 0.22481293425975415,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00019066005098923168,
+      "loss": 1.2694,
+      "step": 3365
+    },
+    {
+      "epoch": 0.22514698022447888,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019061078290289415,
+      "loss": 1.2075,
+      "step": 3370
+    },
+    {
+      "epoch": 0.22548102618920363,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019056139161394462,
+      "loss": 1.2271,
+      "step": 3375
+    },
+    {
+      "epoch": 0.22581507215392838,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00019051187718953992,
+      "loss": 1.2696,
+      "step": 3380
+    },
+    {
+      "epoch": 0.22614911811865313,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019046223969700434,
+      "loss": 1.2897,
+      "step": 3385
+    },
+    {
+      "epoch": 0.22648316408337787,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001904124792038295,
+      "loss": 1.1922,
+      "step": 3390
+    },
+    {
+      "epoch": 0.22681721004810262,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019036259577767426,
+      "loss": 1.2858,
+      "step": 3395
+    },
+    {
+      "epoch": 0.22715125601282737,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00019031258948636466,
+      "loss": 1.1695,
+      "step": 3400
+    },
+    {
+      "epoch": 0.22748530197755212,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019026246039789376,
+      "loss": 1.1508,
+      "step": 3405
+    },
+    {
+      "epoch": 0.22781934794227685,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001902122085804216,
+      "loss": 1.1518,
+      "step": 3410
+    },
+    {
+      "epoch": 0.2281533939070016,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001901618341022751,
+      "loss": 1.213,
+      "step": 3415
+    },
+    {
+      "epoch": 0.22848743987172634,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019011133703194797,
+      "loss": 1.2272,
+      "step": 3420
+    },
+    {
+      "epoch": 0.2288214858364511,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019006071743810055,
+      "loss": 1.2391,
+      "step": 3425
+    },
+    {
+      "epoch": 0.22915553180117584,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00019000997538955985,
+      "loss": 1.2277,
+      "step": 3430
+    },
+    {
+      "epoch": 0.2294895777659006,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0001899591109553193,
+      "loss": 1.2127,
+      "step": 3435
+    },
+    {
+      "epoch": 0.22982362373062534,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00018990812420453885,
+      "loss": 1.2484,
+      "step": 3440
+    },
+    {
+      "epoch": 0.2301576696953501,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00018985701520654466,
+      "loss": 1.262,
+      "step": 3445
+    },
+    {
+      "epoch": 0.23049171566007481,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00018980578403082917,
+      "loss": 1.1882,
+      "step": 3450
+    },
+    {
+      "epoch": 0.23082576162479956,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00018975443074705086,
+      "loss": 1.217,
+      "step": 3455
+    },
+    {
+      "epoch": 0.2311598075895243,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00018970295542503434,
+      "loss": 1.2297,
+      "step": 3460
+    },
+    {
+      "epoch": 0.23149385355424906,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001896513581347701,
+      "loss": 1.1437,
+      "step": 3465
+    },
+    {
+      "epoch": 0.2318278995189738,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001895996389464145,
+      "loss": 1.2221,
+      "step": 3470
+    },
+    {
+      "epoch": 0.23216194548369856,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001895477979302896,
+      "loss": 1.2822,
+      "step": 3475
+    },
+    {
+      "epoch": 0.2324959914484233,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00018949583515688313,
+      "loss": 1.2558,
+      "step": 3480
+    },
+    {
+      "epoch": 0.23283003741314806,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001894437506968484,
+      "loss": 1.2,
+      "step": 3485
+    },
+    {
+      "epoch": 0.23316408337787278,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018939154462100418,
+      "loss": 1.1973,
+      "step": 3490
+    },
+    {
+      "epoch": 0.23349812934259753,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001893392170003345,
+      "loss": 1.1924,
+      "step": 3495
+    },
+    {
+      "epoch": 0.23383217530732228,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001892867679059887,
+      "loss": 1.2833,
+      "step": 3500
+    },
+    {
+      "epoch": 0.23416622127204703,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001892341974092814,
+      "loss": 1.2072,
+      "step": 3505
+    },
+    {
+      "epoch": 0.23450026723677178,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018918150558169217,
+      "loss": 1.2041,
+      "step": 3510
+    },
+    {
+      "epoch": 0.23483431320149653,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00018912869249486556,
+      "loss": 1.1624,
+      "step": 3515
+    },
+    {
+      "epoch": 0.23516835916622128,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.000189075758220611,
+      "loss": 1.2527,
+      "step": 3520
+    },
+    {
+      "epoch": 0.23550240513094603,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001890227028309028,
+      "loss": 1.0995,
+      "step": 3525
+    },
+    {
+      "epoch": 0.23583645109567075,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00018896952639787978,
+      "loss": 1.2456,
+      "step": 3530
+    },
+    {
+      "epoch": 0.2361704970603955,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00018891622899384544,
+      "loss": 1.19,
+      "step": 3535
+    },
+    {
+      "epoch": 0.23650454302512025,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001888628106912678,
+      "loss": 1.2165,
+      "step": 3540
+    },
+    {
+      "epoch": 0.236838588989845,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00018880927156277914,
+      "loss": 1.1449,
+      "step": 3545
+    },
+    {
+      "epoch": 0.23717263495456975,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00018875561168117617,
+      "loss": 1.2019,
+      "step": 3550
+    },
+    {
+      "epoch": 0.2375066809192945,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00018870183111941965,
+      "loss": 1.2256,
+      "step": 3555
+    },
+    {
+      "epoch": 0.23784072688401925,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018864792995063455,
+      "loss": 1.141,
+      "step": 3560
+    },
+    {
+      "epoch": 0.238174772848744,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00018859390824810975,
+      "loss": 1.1647,
+      "step": 3565
+    },
+    {
+      "epoch": 0.23850881881346872,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00018853976608529803,
+      "loss": 1.2427,
+      "step": 3570
+    },
+    {
+      "epoch": 0.23884286477819347,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00018848550353581597,
+      "loss": 1.2271,
+      "step": 3575
+    },
+    {
+      "epoch": 0.23917691074291822,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00018843112067344387,
+      "loss": 1.1665,
+      "step": 3580
+    },
+    {
+      "epoch": 0.23951095670764297,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00018837661757212555,
+      "loss": 1.1786,
+      "step": 3585
+    },
+    {
+      "epoch": 0.23984500267236772,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00018832199430596835,
+      "loss": 1.2068,
+      "step": 3590
+    },
+    {
+      "epoch": 0.24017904863709247,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.000188267250949243,
+      "loss": 1.2176,
+      "step": 3595
+    },
+    {
+      "epoch": 0.24051309460181722,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00018821238757638358,
+      "loss": 1.2054,
+      "step": 3600
+    },
+    {
+      "epoch": 0.24084714056654197,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00018815740426198724,
+      "loss": 1.1923,
+      "step": 3605
+    },
+    {
+      "epoch": 0.2411811865312667,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00018810230108081425,
+      "loss": 1.18,
+      "step": 3610
+    },
+    {
+      "epoch": 0.24151523249599144,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00018804707810778792,
+      "loss": 1.1351,
+      "step": 3615
+    },
+    {
+      "epoch": 0.2418492784607162,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00018799173541799436,
+      "loss": 1.2466,
+      "step": 3620
+    },
+    {
+      "epoch": 0.24218332442544094,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00018793627308668248,
+      "loss": 1.2114,
+      "step": 3625
+    },
+    {
+      "epoch": 0.2425173703901657,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00018788069118926397,
+      "loss": 1.216,
+      "step": 3630
+    },
+    {
+      "epoch": 0.24285141635489044,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001878249898013129,
+      "loss": 1.2487,
+      "step": 3635
+    },
+    {
+      "epoch": 0.2431854623196152,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00018776916899856594,
+      "loss": 1.2802,
+      "step": 3640
+    },
+    {
+      "epoch": 0.24351950828433994,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00018771322885692213,
+      "loss": 1.2153,
+      "step": 3645
+    },
+    {
+      "epoch": 0.24385355424906466,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00018765716945244273,
+      "loss": 1.1273,
+      "step": 3650
+    },
+    {
+      "epoch": 0.2441876002137894,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00018760099086135115,
+      "loss": 1.1616,
+      "step": 3655
+    },
+    {
+      "epoch": 0.24452164617851416,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00018754469316003292,
+      "loss": 1.2522,
+      "step": 3660
+    },
+    {
+      "epoch": 0.2448556921432389,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00018748827642503545,
+      "loss": 1.2216,
+      "step": 3665
+    },
+    {
+      "epoch": 0.24518973810796366,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001874317407330681,
+      "loss": 1.1305,
+      "step": 3670
+    },
+    {
+      "epoch": 0.2455237840726884,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001873750861610018,
+      "loss": 1.179,
+      "step": 3675
+    },
+    {
+      "epoch": 0.24585783003741316,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00018731831278586932,
+      "loss": 1.2022,
+      "step": 3680
+    },
+    {
+      "epoch": 0.2461918760021379,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001872614206848648,
+      "loss": 1.1367,
+      "step": 3685
+    },
+    {
+      "epoch": 0.24652592196686263,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00018720440993534395,
+      "loss": 1.1631,
+      "step": 3690
+    },
+    {
+      "epoch": 0.24685996793158738,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001871472806148237,
+      "loss": 1.2256,
+      "step": 3695
+    },
+    {
+      "epoch": 0.24719401389631213,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00018709003280098225,
+      "loss": 1.2209,
+      "step": 3700
+    },
+    {
+      "epoch": 0.24752805986103688,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001870326665716589,
+      "loss": 1.1896,
+      "step": 3705
+    },
+    {
+      "epoch": 0.24786210582576162,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00018697518200485398,
+      "loss": 1.2845,
+      "step": 3710
+    },
+    {
+      "epoch": 0.24819615179048637,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00018691757917872867,
+      "loss": 1.1432,
+      "step": 3715
+    },
+    {
+      "epoch": 0.24853019775521112,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00018685985817160503,
+      "loss": 1.265,
+      "step": 3720
+    },
+    {
+      "epoch": 0.24886424371993587,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0001868020190619657,
+      "loss": 1.2452,
+      "step": 3725
+    },
+    {
+      "epoch": 0.2491982896846606,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00018674406192845398,
+      "loss": 1.2468,
+      "step": 3730
+    },
+    {
+      "epoch": 0.24953233564938535,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00018668598684987368,
+      "loss": 1.1719,
+      "step": 3735
+    },
+    {
+      "epoch": 0.2498663816141101,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00018662779390518885,
+      "loss": 1.1946,
+      "step": 3740
+    },
+    {
+      "epoch": 0.25020042757883487,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00018656948317352393,
+      "loss": 1.3211,
+      "step": 3745
+    },
+    {
+      "epoch": 0.2505344735435596,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00018651105473416345,
+      "loss": 1.2501,
+      "step": 3750
+    },
+    {
+      "epoch": 0.2508685195082843,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.000186452508666552,
+      "loss": 1.2048,
+      "step": 3755
+    },
+    {
+      "epoch": 0.25120256547300907,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001863938450502941,
+      "loss": 1.2069,
+      "step": 3760
+    },
+    {
+      "epoch": 0.2515366114377338,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00018633506396515407,
+      "loss": 1.1784,
+      "step": 3765
+    },
+    {
+      "epoch": 0.25187065740245856,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00018627616549105606,
+      "loss": 1.1697,
+      "step": 3770
+    },
+    {
+      "epoch": 0.2522047033671833,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018621714970808374,
+      "loss": 1.1564,
+      "step": 3775
+    },
+    {
+      "epoch": 0.25253874933190806,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00018615801669648026,
+      "loss": 1.1939,
+      "step": 3780
+    },
+    {
+      "epoch": 0.2528727952966328,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00018609876653664825,
+      "loss": 1.2784,
+      "step": 3785
+    },
+    {
+      "epoch": 0.25320684126135756,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00018603939930914956,
+      "loss": 1.1739,
+      "step": 3790
+    },
+    {
+      "epoch": 0.2535408872260823,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00018597991509470524,
+      "loss": 1.2513,
+      "step": 3795
+    },
+    {
+      "epoch": 0.25387493319080706,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00018592031397419545,
+      "loss": 1.1615,
+      "step": 3800
+    },
+    {
+      "epoch": 0.2542089791555318,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001858605960286592,
+      "loss": 1.2272,
+      "step": 3805
+    },
+    {
+      "epoch": 0.25454302512025656,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00018580076133929444,
+      "loss": 1.2155,
+      "step": 3810
+    },
+    {
+      "epoch": 0.2548770710849813,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00018574080998745785,
+      "loss": 1.2615,
+      "step": 3815
+    },
+    {
+      "epoch": 0.25521111704970606,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00018568074205466465,
+      "loss": 1.2998,
+      "step": 3820
+    },
+    {
+      "epoch": 0.2555451630144308,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00018562055762258862,
+      "loss": 1.1872,
+      "step": 3825
+    },
+    {
+      "epoch": 0.25587920897915556,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00018556025677306205,
+      "loss": 1.195,
+      "step": 3830
+    },
+    {
+      "epoch": 0.25621325494388025,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018549983958807533,
+      "loss": 1.1688,
+      "step": 3835
+    },
+    {
+      "epoch": 0.256547300908605,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00018543930614977719,
+      "loss": 1.2799,
+      "step": 3840
+    },
+    {
+      "epoch": 0.25688134687332975,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00018537865654047432,
+      "loss": 1.1605,
+      "step": 3845
+    },
+    {
+      "epoch": 0.2572153928380545,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00018531789084263143,
+      "loss": 1.2179,
+      "step": 3850
+    },
+    {
+      "epoch": 0.25754943880277925,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001852570091388711,
+      "loss": 1.1944,
+      "step": 3855
+    },
+    {
+      "epoch": 0.257883484767504,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00018519601151197352,
+      "loss": 1.2285,
+      "step": 3860
+    },
+    {
+      "epoch": 0.25821753073222875,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00018513489804487666,
+      "loss": 1.2673,
+      "step": 3865
+    },
+    {
+      "epoch": 0.2585515766969535,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00018507366882067583,
+      "loss": 1.2363,
+      "step": 3870
+    },
+    {
+      "epoch": 0.25888562266167825,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00018501232392262385,
+      "loss": 1.266,
+      "step": 3875
+    },
+    {
+      "epoch": 0.259219668626403,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00018495086343413083,
+      "loss": 1.2432,
+      "step": 3880
+    },
+    {
+      "epoch": 0.25955371459112775,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00018488928743876394,
+      "loss": 1.2437,
+      "step": 3885
+    },
+    {
+      "epoch": 0.2598877605558525,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018482759602024752,
+      "loss": 1.2451,
+      "step": 3890
+    },
+    {
+      "epoch": 0.26022180652057725,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001847657892624628,
+      "loss": 1.2399,
+      "step": 3895
+    },
+    {
+      "epoch": 0.260555852485302,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001847038672494478,
+      "loss": 1.2465,
+      "step": 3900
+    },
+    {
+      "epoch": 0.26088989845002675,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001846418300653973,
+      "loss": 1.2469,
+      "step": 3905
+    },
+    {
+      "epoch": 0.2612239444147515,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001845796777946627,
+      "loss": 1.1983,
+      "step": 3910
+    },
+    {
+      "epoch": 0.2615579903794762,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001845174105217518,
+      "loss": 1.1489,
+      "step": 3915
+    },
+    {
+      "epoch": 0.26189203634420094,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00018445502833132883,
+      "loss": 1.1875,
+      "step": 3920
+    },
+    {
+      "epoch": 0.2622260823089257,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00018439253130821427,
+      "loss": 1.1905,
+      "step": 3925
+    },
+    {
+      "epoch": 0.26256012827365044,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001843299195373847,
+      "loss": 1.2525,
+      "step": 3930
+    },
+    {
+      "epoch": 0.2628941742383752,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00018426719310397273,
+      "loss": 1.215,
+      "step": 3935
+    },
+    {
+      "epoch": 0.26322822020309994,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001842043520932669,
+      "loss": 1.2433,
+      "step": 3940
+    },
+    {
+      "epoch": 0.2635622661678247,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001841413965907116,
+      "loss": 1.1854,
+      "step": 3945
+    },
+    {
+      "epoch": 0.26389631213254944,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00018407832668190671,
+      "loss": 1.1176,
+      "step": 3950
+    },
+    {
+      "epoch": 0.2642303580972742,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00018401514245260783,
+      "loss": 1.2202,
+      "step": 3955
+    },
+    {
+      "epoch": 0.26456440406199894,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.000183951843988726,
+      "loss": 1.2002,
+      "step": 3960
+    },
+    {
+      "epoch": 0.2648984500267237,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00018388843137632748,
+      "loss": 1.2431,
+      "step": 3965
+    },
+    {
+      "epoch": 0.26523249599144844,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00018382490470163378,
+      "loss": 1.1313,
+      "step": 3970
+    },
+    {
+      "epoch": 0.2655665419561732,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018376126405102153,
+      "loss": 1.2414,
+      "step": 3975
+    },
+    {
+      "epoch": 0.26590058792089794,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00018369750951102232,
+      "loss": 1.2146,
+      "step": 3980
+    },
+    {
+      "epoch": 0.2662346338856227,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00018363364116832256,
+      "loss": 1.1797,
+      "step": 3985
+    },
+    {
+      "epoch": 0.26656867985034743,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00018356965910976348,
+      "loss": 1.236,
+      "step": 3990
+    },
+    {
+      "epoch": 0.26690272581507213,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018350556342234078,
+      "loss": 1.1745,
+      "step": 3995
+    },
+    {
+      "epoch": 0.2672367717797969,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00018344135419320483,
+      "loss": 1.1889,
+      "step": 4000
+    },
+    {
+      "epoch": 0.2675708177445216,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018337703150966025,
+      "loss": 1.2263,
+      "step": 4005
+    },
+    {
+      "epoch": 0.2679048637092464,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.000183312595459166,
+      "loss": 1.1455,
+      "step": 4010
+    },
+    {
+      "epoch": 0.2682389096739711,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001832480461293352,
+      "loss": 1.1681,
+      "step": 4015
+    },
+    {
+      "epoch": 0.2685729556386959,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001831833836079349,
+      "loss": 1.2668,
+      "step": 4020
+    },
+    {
+      "epoch": 0.2689070016034206,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00018311860798288609,
+      "loss": 1.2165,
+      "step": 4025
+    },
+    {
+      "epoch": 0.2692410475681454,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00018305371934226362,
+      "loss": 1.2629,
+      "step": 4030
+    },
+    {
+      "epoch": 0.2695750935328701,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00018298871777429595,
+      "loss": 1.1995,
+      "step": 4035
+    },
+    {
+      "epoch": 0.2699091394975949,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018292360336736506,
+      "loss": 1.1677,
+      "step": 4040
+    },
+    {
+      "epoch": 0.2702431854623196,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00018285837621000636,
+      "loss": 1.2907,
+      "step": 4045
+    },
+    {
+      "epoch": 0.2705772314270444,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00018279303639090865,
+      "loss": 1.2465,
+      "step": 4050
+    },
+    {
+      "epoch": 0.2709112773917691,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00018272758399891383,
+      "loss": 1.1909,
+      "step": 4055
+    },
+    {
+      "epoch": 0.2712453233564939,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00018266201912301688,
+      "loss": 1.2178,
+      "step": 4060
+    },
+    {
+      "epoch": 0.2715793693212186,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00018259634185236574,
+      "loss": 1.2175,
+      "step": 4065
+    },
+    {
+      "epoch": 0.27191341528594337,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00018253055227626116,
+      "loss": 1.1551,
+      "step": 4070
+    },
+    {
+      "epoch": 0.27224746125066807,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00018246465048415663,
+      "loss": 1.1781,
+      "step": 4075
+    },
+    {
+      "epoch": 0.2725815072153928,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00018239863656565813,
+      "loss": 1.2389,
+      "step": 4080
+    },
+    {
+      "epoch": 0.27291555318011756,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00018233251061052421,
+      "loss": 1.3091,
+      "step": 4085
+    },
+    {
+      "epoch": 0.2732495991448423,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018226627270866562,
+      "loss": 1.2638,
+      "step": 4090
+    },
+    {
+      "epoch": 0.27358364510956706,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00018219992295014548,
+      "loss": 1.194,
+      "step": 4095
+    },
+    {
+      "epoch": 0.2739176910742918,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00018213346142517884,
+      "loss": 1.2313,
+      "step": 4100
+    },
+    {
+      "epoch": 0.27425173703901656,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00018206688822413288,
+      "loss": 1.1716,
+      "step": 4105
+    },
+    {
+      "epoch": 0.2745857830037413,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00018200020343752646,
+      "loss": 1.2846,
+      "step": 4110
+    },
+    {
+      "epoch": 0.27491982896846606,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018193340715603033,
+      "loss": 1.154,
+      "step": 4115
+    },
+    {
+      "epoch": 0.2752538749331908,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00018186649947046668,
+      "loss": 1.211,
+      "step": 4120
+    },
+    {
+      "epoch": 0.27558792089791556,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001817994804718093,
+      "loss": 1.2262,
+      "step": 4125
+    },
+    {
+      "epoch": 0.2759219668626403,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00018173235025118324,
+      "loss": 1.2007,
+      "step": 4130
+    },
+    {
+      "epoch": 0.27625601282736506,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001816651088998649,
+      "loss": 1.228,
+      "step": 4135
+    },
+    {
+      "epoch": 0.2765900587920898,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00018159775650928155,
+      "loss": 1.2326,
+      "step": 4140
+    },
+    {
+      "epoch": 0.27692410475681456,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001815302931710117,
+      "loss": 1.1988,
+      "step": 4145
+    },
+    {
+      "epoch": 0.2772581507215393,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00018146271897678457,
+      "loss": 1.2058,
+      "step": 4150
+    },
+    {
+      "epoch": 0.277592196686264,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00018139503401848017,
+      "loss": 1.3223,
+      "step": 4155
+    },
+    {
+      "epoch": 0.27792624265098875,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00018132723838812907,
+      "loss": 1.2199,
+      "step": 4160
+    },
+    {
+      "epoch": 0.2782602886157135,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00018125933217791234,
+      "loss": 1.2513,
+      "step": 4165
+    },
+    {
+      "epoch": 0.27859433458043825,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018119131548016137,
+      "loss": 1.2097,
+      "step": 4170
+    },
+    {
+      "epoch": 0.278928380545163,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00018112318838735787,
+      "loss": 1.1601,
+      "step": 4175
+    },
+    {
+      "epoch": 0.27926242650988775,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00018105495099213353,
+      "loss": 1.2154,
+      "step": 4180
+    },
+    {
+      "epoch": 0.2795964724746125,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00018098660338727017,
+      "loss": 1.1971,
+      "step": 4185
+    },
+    {
+      "epoch": 0.27993051843933725,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001809181456656993,
+      "loss": 1.1371,
+      "step": 4190
+    },
+    {
+      "epoch": 0.280264564404062,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00018084957792050224,
+      "loss": 1.2547,
+      "step": 4195
+    },
+    {
+      "epoch": 0.28059861036878675,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001807809002449099,
+      "loss": 1.201,
+      "step": 4200
+    },
+    {
+      "epoch": 0.2809326563335115,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00018071211273230263,
+      "loss": 1.2014,
+      "step": 4205
+    },
+    {
+      "epoch": 0.28126670229823625,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00018064321547621022,
+      "loss": 1.2495,
+      "step": 4210
+    },
+    {
+      "epoch": 0.281600748262961,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00018057420857031157,
+      "loss": 1.1429,
+      "step": 4215
+    },
+    {
+      "epoch": 0.28193479422768575,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001805050921084347,
+      "loss": 1.1956,
+      "step": 4220
+    },
+    {
+      "epoch": 0.2822688401924105,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001804358661845566,
+      "loss": 1.1969,
+      "step": 4225
+    },
+    {
+      "epoch": 0.28260288615713525,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00018036653089280308,
+      "loss": 1.1448,
+      "step": 4230
+    },
+    {
+      "epoch": 0.28293693212185994,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00018029708632744871,
+      "loss": 1.2246,
+      "step": 4235
+    },
+    {
+      "epoch": 0.2832709780865847,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00018022753258291658,
+      "loss": 1.2453,
+      "step": 4240
+    },
+    {
+      "epoch": 0.28360502405130944,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00018015786975377824,
+      "loss": 1.1488,
+      "step": 4245
+    },
+    {
+      "epoch": 0.2839390700160342,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00018008809793475358,
+      "loss": 1.1986,
+      "step": 4250
+    },
+    {
+      "epoch": 0.28427311598075894,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00018001821722071068,
+      "loss": 1.2035,
+      "step": 4255
+    },
+    {
+      "epoch": 0.2846071619454837,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00017994822770666565,
+      "loss": 1.2267,
+      "step": 4260
+    },
+    {
+      "epoch": 0.28494120791020844,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001798781294877826,
+      "loss": 1.2708,
+      "step": 4265
+    },
+    {
+      "epoch": 0.2852752538749332,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00017980792265937336,
+      "loss": 1.2181,
+      "step": 4270
+    },
+    {
+      "epoch": 0.28560929983965794,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00017973760731689753,
+      "loss": 1.2263,
+      "step": 4275
+    },
+    {
+      "epoch": 0.2859433458043827,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00017966718355596218,
+      "loss": 1.1812,
+      "step": 4280
+    },
+    {
+      "epoch": 0.28627739176910744,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00017959665147232177,
+      "loss": 1.1684,
+      "step": 4285
+    },
+    {
+      "epoch": 0.2866114377338322,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00017952601116187823,
+      "loss": 1.1948,
+      "step": 4290
+    },
+    {
+      "epoch": 0.28694548369855694,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00017945526272068038,
+      "loss": 1.2083,
+      "step": 4295
+    },
+    {
+      "epoch": 0.2872795296632817,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00017938440624492427,
+      "loss": 1.2976,
+      "step": 4300
+    },
+    {
+      "epoch": 0.28761357562800643,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00017931344183095272,
+      "loss": 1.1737,
+      "step": 4305
+    },
+    {
+      "epoch": 0.2879476215927312,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00017924236957525544,
+      "loss": 1.3016,
+      "step": 4310
+    },
+    {
+      "epoch": 0.2882816675574559,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00017917118957446864,
+      "loss": 1.183,
+      "step": 4315
+    },
+    {
+      "epoch": 0.28861571352218063,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00017909990192537504,
+      "loss": 1.2075,
+      "step": 4320
+    },
+    {
+      "epoch": 0.2889497594869054,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00017902850672490387,
+      "loss": 1.1567,
+      "step": 4325
+    },
+    {
+      "epoch": 0.2892838054516301,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00017895700407013045,
+      "loss": 1.1603,
+      "step": 4330
+    },
+    {
+      "epoch": 0.2896178514163549,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00017888539405827624,
+      "loss": 1.1885,
+      "step": 4335
+    },
+    {
+      "epoch": 0.2899518973810796,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001788136767867087,
+      "loss": 1.2023,
+      "step": 4340
+    },
+    {
+      "epoch": 0.2902859433458044,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00017874185235294113,
+      "loss": 1.2503,
+      "step": 4345
+    },
+    {
+      "epoch": 0.2906199893105291,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001786699208546325,
+      "loss": 1.229,
+      "step": 4350
+    },
+    {
+      "epoch": 0.2909540352752539,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00017859788238958738,
+      "loss": 1.2023,
+      "step": 4355
+    },
+    {
+      "epoch": 0.2912880812399786,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00017852573705575583,
+      "loss": 1.2153,
+      "step": 4360
+    },
+    {
+      "epoch": 0.2916221272047034,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00017845348495123308,
+      "loss": 1.1974,
+      "step": 4365
+    },
+    {
+      "epoch": 0.2919561731694281,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00017838112617425968,
+      "loss": 1.1569,
+      "step": 4370
+    },
+    {
+      "epoch": 0.2922902191341529,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017830866082322116,
+      "loss": 1.2243,
+      "step": 4375
+    },
+    {
+      "epoch": 0.2926242650988776,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00017823608899664796,
+      "loss": 1.2616,
+      "step": 4380
+    },
+    {
+      "epoch": 0.2929583110636024,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001781634107932153,
+      "loss": 1.2427,
+      "step": 4385
+    },
+    {
+      "epoch": 0.2932923570283271,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.000178090626311743,
+      "loss": 1.2594,
+      "step": 4390
+    },
+    {
+      "epoch": 0.29362640299305187,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001780177356511955,
+      "loss": 1.1679,
+      "step": 4395
+    },
+    {
+      "epoch": 0.29396044895777657,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00017794473891068142,
+      "loss": 1.1783,
+      "step": 4400
+    },
+    {
+      "epoch": 0.2942944949225013,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001778716361894538,
+      "loss": 1.2501,
+      "step": 4405
+    },
+    {
+      "epoch": 0.29462854088722606,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00017779842758690973,
+      "loss": 1.2928,
+      "step": 4410
+    },
+    {
+      "epoch": 0.2949625868519508,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00017772511320259023,
+      "loss": 1.1966,
+      "step": 4415
+    },
+    {
+      "epoch": 0.29529663281667556,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001776516931361801,
+      "loss": 1.28,
+      "step": 4420
+    },
+    {
+      "epoch": 0.2956306787814003,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00017757816748750798,
+      "loss": 1.1887,
+      "step": 4425
+    },
+    {
+      "epoch": 0.29596472474612506,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00017750453635654591,
+      "loss": 1.2065,
+      "step": 4430
+    },
+    {
+      "epoch": 0.2962987707108498,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001774307998434095,
+      "loss": 1.2641,
+      "step": 4435
+    },
+    {
+      "epoch": 0.29663281667557456,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017735695804835757,
+      "loss": 1.1513,
+      "step": 4440
+    },
+    {
+      "epoch": 0.2969668626402993,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001772830110717921,
+      "loss": 1.2108,
+      "step": 4445
+    },
+    {
+      "epoch": 0.29730090860502406,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017720895901425805,
+      "loss": 1.1339,
+      "step": 4450
+    },
+    {
+      "epoch": 0.2976349545697488,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001771348019764433,
+      "loss": 1.1995,
+      "step": 4455
+    },
+    {
+      "epoch": 0.29796900053447356,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001770605400591785,
+      "loss": 1.2197,
+      "step": 4460
+    },
+    {
+      "epoch": 0.2983030464991983,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017698617336343685,
+      "loss": 1.226,
+      "step": 4465
+    },
+    {
+      "epoch": 0.29863709246392306,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.000176911701990334,
+      "loss": 1.1742,
+      "step": 4470
+    },
+    {
+      "epoch": 0.2989711384286478,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00017683712604112798,
+      "loss": 1.2294,
+      "step": 4475
+    },
+    {
+      "epoch": 0.2993051843933725,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00017676244561721905,
+      "loss": 1.1323,
+      "step": 4480
+    },
+    {
+      "epoch": 0.29963923035809725,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00017668766082014936,
+      "loss": 1.1976,
+      "step": 4485
+    },
+    {
+      "epoch": 0.299973276322822,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00017661277175160314,
+      "loss": 1.3004,
+      "step": 4490
+    },
+    {
+      "epoch": 0.30030732228754675,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00017653777851340635,
+      "loss": 1.2209,
+      "step": 4495
+    },
+    {
+      "epoch": 0.3006413682522715,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001764626812075266,
+      "loss": 1.3061,
+      "step": 4500
+    },
+    {
+      "epoch": 0.30097541421699625,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00017638747993607286,
+      "loss": 1.2313,
+      "step": 4505
+    },
+    {
+      "epoch": 0.301309460181721,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00017631217480129573,
+      "loss": 1.2196,
+      "step": 4510
+    },
+    {
+      "epoch": 0.30164350614644575,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00017623676590558675,
+      "loss": 1.2387,
+      "step": 4515
+    },
+    {
+      "epoch": 0.3019775521111705,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00017616125335147875,
+      "loss": 1.1669,
+      "step": 4520
+    },
+    {
+      "epoch": 0.30231159807589525,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00017608563724164536,
+      "loss": 1.2865,
+      "step": 4525
+    },
+    {
+      "epoch": 0.30264564404062,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001760099176789012,
+      "loss": 1.191,
+      "step": 4530
+    },
+    {
+      "epoch": 0.30297969000534475,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00017593409476620127,
+      "loss": 1.2484,
+      "step": 4535
+    },
+    {
+      "epoch": 0.3033137359700695,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00017585816860664135,
+      "loss": 1.2039,
+      "step": 4540
+    },
+    {
+      "epoch": 0.30364778193479425,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00017578213930345753,
+      "loss": 1.2493,
+      "step": 4545
+    },
+    {
+      "epoch": 0.303981827899519,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0001757060069600261,
+      "loss": 1.1464,
+      "step": 4550
+    },
+    {
+      "epoch": 0.30431587386424375,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017562977167986344,
+      "loss": 1.2524,
+      "step": 4555
+    },
+    {
+      "epoch": 0.30464991982896844,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00017555343356662597,
+      "loss": 1.1804,
+      "step": 4560
+    },
+    {
+      "epoch": 0.3049839657936932,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00017547699272410988,
+      "loss": 1.1869,
+      "step": 4565
+    },
+    {
+      "epoch": 0.30531801175841794,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017540044925625102,
+      "loss": 1.1751,
+      "step": 4570
+    },
+    {
+      "epoch": 0.3056520577231427,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00017532380326712487,
+      "loss": 1.2452,
+      "step": 4575
+    },
+    {
+      "epoch": 0.30598610368786744,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001752470548609462,
+      "loss": 1.2203,
+      "step": 4580
+    },
+    {
+      "epoch": 0.3063201496525922,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00017517020414206913,
+      "loss": 1.2208,
+      "step": 4585
+    },
+    {
+      "epoch": 0.30665419561731694,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017509325121498677,
+      "loss": 1.1297,
+      "step": 4590
+    },
+    {
+      "epoch": 0.3069882415820417,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017501619618433135,
+      "loss": 1.2869,
+      "step": 4595
+    },
+    {
+      "epoch": 0.30732228754676644,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00017493903915487377,
+      "loss": 1.2867,
+      "step": 4600
+    },
+    {
+      "epoch": 0.3076563335114912,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001748617802315238,
+      "loss": 1.1794,
+      "step": 4605
+    },
+    {
+      "epoch": 0.30799037947621594,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001747844195193296,
+      "loss": 1.1931,
+      "step": 4610
+    },
+    {
+      "epoch": 0.3083244254409407,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001747069571234778,
+      "loss": 1.1756,
+      "step": 4615
+    },
+    {
+      "epoch": 0.30865847140566544,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017462939314929327,
+      "loss": 1.2366,
+      "step": 4620
+    },
+    {
+      "epoch": 0.3089925173703902,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.000174551727702239,
+      "loss": 1.1803,
+      "step": 4625
+    },
+    {
+      "epoch": 0.30932656333511493,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00017447396088791597,
+      "loss": 1.2106,
+      "step": 4630
+    },
+    {
+      "epoch": 0.3096606092998397,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00017439609281206297,
+      "loss": 1.2554,
+      "step": 4635
+    },
+    {
+      "epoch": 0.3099946552645644,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00017431812358055645,
+      "loss": 1.1544,
+      "step": 4640
+    },
+    {
+      "epoch": 0.31032870122928913,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00017424005329941047,
+      "loss": 1.1944,
+      "step": 4645
+    },
+    {
+      "epoch": 0.3106627471940139,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00017416188207477638,
+      "loss": 1.2195,
+      "step": 4650
+    },
+    {
+      "epoch": 0.3109967931587386,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00017408361001294292,
+      "loss": 1.2621,
+      "step": 4655
+    },
+    {
+      "epoch": 0.3113308391234634,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001740052372203358,
+      "loss": 1.2301,
+      "step": 4660
+    },
+    {
+      "epoch": 0.3116648850881881,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00017392676380351775,
+      "loss": 1.1779,
+      "step": 4665
+    },
+    {
+      "epoch": 0.3119989310529129,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00017384818986918837,
+      "loss": 1.1954,
+      "step": 4670
+    },
+    {
+      "epoch": 0.3123329770176376,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00017376951552418386,
+      "loss": 1.2073,
+      "step": 4675
+    },
+    {
+      "epoch": 0.3126670229823624,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00017369074087547696,
+      "loss": 1.1381,
+      "step": 4680
+    },
+    {
+      "epoch": 0.3130010689470871,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00017361186603017685,
+      "loss": 1.2583,
+      "step": 4685
+    },
+    {
+      "epoch": 0.3133351149118119,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00017353289109552883,
+      "loss": 1.2078,
+      "step": 4690
+    },
+    {
+      "epoch": 0.3136691608765366,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00017345381617891442,
+      "loss": 1.2324,
+      "step": 4695
+    },
+    {
+      "epoch": 0.3140032068412614,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.000173374641387851,
+      "loss": 1.1969,
+      "step": 4700
+    },
+    {
+      "epoch": 0.3143372528059861,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001732953668299918,
+      "loss": 1.1406,
+      "step": 4705
+    },
+    {
+      "epoch": 0.31467129877071087,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001732159926131256,
+      "loss": 1.1391,
+      "step": 4710
+    },
+    {
+      "epoch": 0.3150053447354356,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00017313651884517684,
+      "loss": 1.1954,
+      "step": 4715
+    },
+    {
+      "epoch": 0.3153393907001603,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00017305694563420524,
+      "loss": 1.218,
+      "step": 4720
+    },
+    {
+      "epoch": 0.31567343666488507,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00017297727308840564,
+      "loss": 1.2108,
+      "step": 4725
+    },
+    {
+      "epoch": 0.3160074826296098,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00017289750131610813,
+      "loss": 1.2214,
+      "step": 4730
+    },
+    {
+      "epoch": 0.31634152859433456,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00017281763042577763,
+      "loss": 1.3101,
+      "step": 4735
+    },
+    {
+      "epoch": 0.3166755745590593,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017273766052601378,
+      "loss": 1.2593,
+      "step": 4740
+    },
+    {
+      "epoch": 0.31700962052378406,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00017265759172555085,
+      "loss": 1.2348,
+      "step": 4745
+    },
+    {
+      "epoch": 0.3173436664885088,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001725774241332577,
+      "loss": 1.1378,
+      "step": 4750
+    },
+    {
+      "epoch": 0.31767771245323356,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00017249715785813737,
+      "loss": 1.1839,
+      "step": 4755
+    },
+    {
+      "epoch": 0.3180117584179583,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00017241679300932717,
+      "loss": 1.2049,
+      "step": 4760
+    },
+    {
+      "epoch": 0.31834580438268306,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017233632969609842,
+      "loss": 1.2061,
+      "step": 4765
+    },
+    {
+      "epoch": 0.3186798503474078,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00017225576802785636,
+      "loss": 1.2253,
+      "step": 4770
+    },
+    {
+      "epoch": 0.31901389631213256,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001721751081141398,
+      "loss": 1.2239,
+      "step": 4775
+    },
+    {
+      "epoch": 0.3193479422768573,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00017209435006462136,
+      "loss": 1.1895,
+      "step": 4780
+    },
+    {
+      "epoch": 0.31968198824158206,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00017201349398910694,
+      "loss": 1.3404,
+      "step": 4785
+    },
+    {
+      "epoch": 0.3200160342063068,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017193253999753575,
+      "loss": 1.199,
+      "step": 4790
+    },
+    {
+      "epoch": 0.32035008017103156,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00017185148819998022,
+      "loss": 1.2542,
+      "step": 4795
+    },
+    {
+      "epoch": 0.32068412613575625,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001717703387066456,
+      "loss": 1.2918,
+      "step": 4800
+    },
+    {
+      "epoch": 0.321018172100481,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00017168909162787016,
+      "loss": 1.1468,
+      "step": 4805
+    },
+    {
+      "epoch": 0.32135221806520575,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00017160774707412476,
+      "loss": 1.2845,
+      "step": 4810
+    },
+    {
+      "epoch": 0.3216862640299305,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00017152630515601281,
+      "loss": 1.1376,
+      "step": 4815
+    },
+    {
+      "epoch": 0.32202030999465525,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001714447659842701,
+      "loss": 1.2282,
+      "step": 4820
+    },
+    {
+      "epoch": 0.32235435595938,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00017136312966976465,
+      "loss": 1.2559,
+      "step": 4825
+    },
+    {
+      "epoch": 0.32268840192410475,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00017128139632349658,
+      "loss": 1.1685,
+      "step": 4830
+    },
+    {
+      "epoch": 0.3230224478888295,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017119956605659792,
+      "loss": 1.1611,
+      "step": 4835
+    },
+    {
+      "epoch": 0.32335649385355425,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001711176389803325,
+      "loss": 1.2007,
+      "step": 4840
+    },
+    {
+      "epoch": 0.323690539818279,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001710356152060958,
+      "loss": 1.1714,
+      "step": 4845
+    },
+    {
+      "epoch": 0.32402458578300375,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00017095349484541478,
+      "loss": 1.1959,
+      "step": 4850
+    },
+    {
+      "epoch": 0.3243586317477285,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00017087127800994767,
+      "loss": 1.2585,
+      "step": 4855
+    },
+    {
+      "epoch": 0.32469267771245325,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00017078896481148388,
+      "loss": 1.1676,
+      "step": 4860
+    },
+    {
+      "epoch": 0.325026723677178,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00017070655536194397,
+      "loss": 1.1425,
+      "step": 4865
+    },
+    {
+      "epoch": 0.32536076964190275,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00017062404977337918,
+      "loss": 1.1537,
+      "step": 4870
+    },
+    {
+      "epoch": 0.3256948156066275,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00017054144815797164,
+      "loss": 1.1753,
+      "step": 4875
+    },
+    {
+      "epoch": 0.3260288615713522,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00017045875062803395,
+      "loss": 1.2631,
+      "step": 4880
+    },
+    {
+      "epoch": 0.32636290753607694,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00017037595729600913,
+      "loss": 1.2179,
+      "step": 4885
+    },
+    {
+      "epoch": 0.3266969535008017,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00017029306827447049,
+      "loss": 1.1933,
+      "step": 4890
+    },
+    {
+      "epoch": 0.32703099946552644,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00017021008367612144,
+      "loss": 1.2329,
+      "step": 4895
+    },
+    {
+      "epoch": 0.3273650454302512,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00017012700361379533,
+      "loss": 1.2341,
+      "step": 4900
+    },
+    {
+      "epoch": 0.32769909139497594,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00017004382820045533,
+      "loss": 1.1853,
+      "step": 4905
+    },
+    {
+      "epoch": 0.3280331373597007,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001699605575491943,
+      "loss": 1.1806,
+      "step": 4910
+    },
+    {
+      "epoch": 0.32836718332442544,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00016987719177323445,
+      "loss": 1.2325,
+      "step": 4915
+    },
+    {
+      "epoch": 0.3287012292891502,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001697937309859275,
+      "loss": 1.2619,
+      "step": 4920
+    },
+    {
+      "epoch": 0.32903527525387494,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00016971017530075427,
+      "loss": 1.2323,
+      "step": 4925
+    },
+    {
+      "epoch": 0.3293693212185997,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0001696265248313246,
+      "loss": 1.2028,
+      "step": 4930
+    },
+    {
+      "epoch": 0.32970336718332444,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00016954277969137723,
+      "loss": 1.1464,
+      "step": 4935
+    },
+    {
+      "epoch": 0.3300374131480492,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00016945893999477965,
+      "loss": 1.2066,
+      "step": 4940
+    },
+    {
+      "epoch": 0.33037145911277394,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00016937500585552785,
+      "loss": 1.1981,
+      "step": 4945
+    },
+    {
+      "epoch": 0.3307055050774987,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00016929097738774634,
+      "loss": 1.1917,
+      "step": 4950
+    },
+    {
+      "epoch": 0.33103955104222343,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00016920685470568777,
+      "loss": 1.1599,
+      "step": 4955
+    },
+    {
+      "epoch": 0.33137359700694813,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00016912263792373295,
+      "loss": 1.2997,
+      "step": 4960
+    },
+    {
+      "epoch": 0.3317076429716729,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00016903832715639062,
+      "loss": 1.1939,
+      "step": 4965
+    },
+    {
+      "epoch": 0.3320416889363976,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00016895392251829736,
+      "loss": 1.1861,
+      "step": 4970
+    },
+    {
+      "epoch": 0.3323757349011224,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00016886942412421734,
+      "loss": 1.282,
+      "step": 4975
+    },
+    {
+      "epoch": 0.3327097808658471,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00016878483208904217,
+      "loss": 1.2393,
+      "step": 4980
+    },
+    {
+      "epoch": 0.3330438268305719,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00016870014652779086,
+      "loss": 1.2781,
+      "step": 4985
+    },
+    {
+      "epoch": 0.3333778727952966,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00016861536755560956,
+      "loss": 1.242,
+      "step": 4990
+    },
+    {
+      "epoch": 0.3337119187600214,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001685304952877714,
+      "loss": 1.1841,
+      "step": 4995
+    },
+    {
+      "epoch": 0.3340459647247461,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001684455298396764,
+      "loss": 1.2658,
+      "step": 5000
+    },
+    {
+      "epoch": 0.3343800106894709,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00016836047132685132,
+      "loss": 1.3034,
+      "step": 5005
+    },
+    {
+      "epoch": 0.3347140566541956,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001682753198649493,
+      "loss": 1.1442,
+      "step": 5010
+    },
+    {
+      "epoch": 0.3350481026189204,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00016819007556975003,
+      "loss": 1.2386,
+      "step": 5015
+    },
+    {
+      "epoch": 0.3353821485836451,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001681047385571594,
+      "loss": 1.3074,
+      "step": 5020
+    },
+    {
+      "epoch": 0.3357161945483699,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001680193089432092,
+      "loss": 1.215,
+      "step": 5025
+    },
+    {
+      "epoch": 0.3360502405130946,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00016793378684405735,
+      "loss": 1.2023,
+      "step": 5030
+    },
+    {
+      "epoch": 0.33638428647781937,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00016784817237598744,
+      "loss": 1.1826,
+      "step": 5035
+    },
+    {
+      "epoch": 0.3367183324425441,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0001677624656554086,
+      "loss": 1.2126,
+      "step": 5040
+    },
+    {
+      "epoch": 0.3370523784072688,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00016767666679885546,
+      "loss": 1.2062,
+      "step": 5045
+    },
+    {
+      "epoch": 0.33738642437199357,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00016759077592298788,
+      "loss": 1.2673,
+      "step": 5050
+    },
+    {
+      "epoch": 0.3377204703367183,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00016750479314459087,
+      "loss": 1.2311,
+      "step": 5055
+    },
+    {
+      "epoch": 0.33805451630144306,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00016741871858057437,
+      "loss": 1.1168,
+      "step": 5060
+    },
+    {
+      "epoch": 0.3383885622661678,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00016733255234797318,
+      "loss": 1.1359,
+      "step": 5065
+    },
+    {
+      "epoch": 0.33872260823089256,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00016724629456394666,
+      "loss": 1.1967,
+      "step": 5070
+    },
+    {
+      "epoch": 0.3390566541956173,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001671599453457787,
+      "loss": 1.2424,
+      "step": 5075
+    },
+    {
+      "epoch": 0.33939070016034206,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00016707350481087754,
+      "loss": 1.1751,
+      "step": 5080
+    },
+    {
+      "epoch": 0.3397247461250668,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00016698697307677548,
+      "loss": 1.1932,
+      "step": 5085
+    },
+    {
+      "epoch": 0.34005879208979156,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00016690035026112893,
+      "loss": 1.2002,
+      "step": 5090
+    },
+    {
+      "epoch": 0.3403928380545163,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001668136364817181,
+      "loss": 1.2033,
+      "step": 5095
+    },
+    {
+      "epoch": 0.34072688401924106,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001667268318564469,
+      "loss": 1.1139,
+      "step": 5100
+    },
+    {
+      "epoch": 0.3410609299839658,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001666399365033427,
+      "loss": 1.2557,
+      "step": 5105
+    },
+    {
+      "epoch": 0.34139497594869056,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00016655295054055633,
+      "loss": 1.1648,
+      "step": 5110
+    },
+    {
+      "epoch": 0.3417290219134153,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0001664658740863617,
+      "loss": 1.2161,
+      "step": 5115
+    },
+    {
+      "epoch": 0.34206306787814006,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00016637870725915593,
+      "loss": 1.2772,
+      "step": 5120
+    },
+    {
+      "epoch": 0.34239711384286475,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00016629145017745878,
+      "loss": 1.2373,
+      "step": 5125
+    },
+    {
+      "epoch": 0.3427311598075895,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.000166204102959913,
+      "loss": 1.1672,
+      "step": 5130
+    },
+    {
+      "epoch": 0.34306520577231425,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00016611666572528372,
+      "loss": 1.2734,
+      "step": 5135
+    },
+    {
+      "epoch": 0.343399251737039,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00016602913859245847,
+      "loss": 1.2255,
+      "step": 5140
+    },
+    {
+      "epoch": 0.34373329770176375,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001659415216804471,
+      "loss": 1.2051,
+      "step": 5145
+    },
+    {
+      "epoch": 0.3440673436664885,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00016585381510838144,
+      "loss": 1.203,
+      "step": 5150
+    },
+    {
+      "epoch": 0.34440138963121325,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00016576601899551534,
+      "loss": 1.1811,
+      "step": 5155
+    },
+    {
+      "epoch": 0.344735435595938,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00016567813346122427,
+      "loss": 1.2409,
+      "step": 5160
+    },
+    {
+      "epoch": 0.34506948156066275,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00016559015862500538,
+      "loss": 1.1553,
+      "step": 5165
+    },
+    {
+      "epoch": 0.3454035275253875,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00016550209460647718,
+      "loss": 1.2214,
+      "step": 5170
+    },
+    {
+      "epoch": 0.34573757349011225,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00016541394152537952,
+      "loss": 1.2735,
+      "step": 5175
+    },
+    {
+      "epoch": 0.346071619454837,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00016532569950157325,
+      "loss": 1.2076,
+      "step": 5180
+    },
+    {
+      "epoch": 0.34640566541956175,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00016523736865504021,
+      "loss": 1.1431,
+      "step": 5185
+    },
+    {
+      "epoch": 0.3467397113842865,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00016514894910588305,
+      "loss": 1.2222,
+      "step": 5190
+    },
+    {
+      "epoch": 0.34707375734901125,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00016506044097432495,
+      "loss": 1.1915,
+      "step": 5195
+    },
+    {
+      "epoch": 0.347407803313736,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00016497184438070956,
+      "loss": 1.2848,
+      "step": 5200
+    },
+    {
+      "epoch": 0.3477418492784607,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0001648831594455008,
+      "loss": 1.1968,
+      "step": 5205
+    },
+    {
+      "epoch": 0.34807589524318544,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00016479438628928277,
+      "loss": 1.2443,
+      "step": 5210
+    },
+    {
+      "epoch": 0.3484099412079102,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00016470552503275947,
+      "loss": 1.1879,
+      "step": 5215
+    },
+    {
+      "epoch": 0.34874398717263494,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001646165757967547,
+      "loss": 1.1934,
+      "step": 5220
+    },
+    {
+      "epoch": 0.3490780331373597,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00016452753870221183,
+      "loss": 1.2376,
+      "step": 5225
+    },
+    {
+      "epoch": 0.34941207910208444,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001644384138701938,
+      "loss": 1.2077,
+      "step": 5230
+    },
+    {
+      "epoch": 0.3497461250668092,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00016434920142188278,
+      "loss": 1.2144,
+      "step": 5235
+    },
+    {
+      "epoch": 0.35008017103153394,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00016425990147858003,
+      "loss": 1.2423,
+      "step": 5240
+    },
+    {
+      "epoch": 0.3504142169962587,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00016417051416170594,
+      "loss": 1.1769,
+      "step": 5245
+    },
+    {
+      "epoch": 0.35074826296098344,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00016408103959279945,
+      "loss": 1.1844,
+      "step": 5250
+    },
+    {
+      "epoch": 0.3510823089257082,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00016399147789351837,
+      "loss": 1.1612,
+      "step": 5255
+    },
+    {
+      "epoch": 0.35141635489043294,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00016390182918563887,
+      "loss": 1.2393,
+      "step": 5260
+    },
+    {
+      "epoch": 0.3517504008551577,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001638120935910554,
+      "loss": 1.187,
+      "step": 5265
+    },
+    {
+      "epoch": 0.35208444681988244,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00016372227123178057,
+      "loss": 1.1734,
+      "step": 5270
+    },
+    {
+      "epoch": 0.3524184927846072,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00016363236222994505,
+      "loss": 1.1329,
+      "step": 5275
+    },
+    {
+      "epoch": 0.35275253874933193,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001635423667077972,
+      "loss": 1.2187,
+      "step": 5280
+    },
+    {
+      "epoch": 0.35308658471405663,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001634522847877031,
+      "loss": 1.2226,
+      "step": 5285
+    },
+    {
+      "epoch": 0.3534206306787814,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00016336211659214621,
+      "loss": 1.1586,
+      "step": 5290
+    },
+    {
+      "epoch": 0.3537546766435061,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00016327186224372747,
+      "loss": 1.1645,
+      "step": 5295
+    },
+    {
+      "epoch": 0.3540887226082309,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00016318152186516472,
+      "loss": 1.2019,
+      "step": 5300
+    },
+    {
+      "epoch": 0.3544227685729556,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00016309109557929302,
+      "loss": 1.2287,
+      "step": 5305
+    },
+    {
+      "epoch": 0.3547568145376804,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00016300058350906404,
+      "loss": 1.1239,
+      "step": 5310
+    },
+    {
+      "epoch": 0.3550908605024051,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00016290998577754622,
+      "loss": 1.1858,
+      "step": 5315
+    },
+    {
+      "epoch": 0.3554249064671299,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00016281930250792442,
+      "loss": 1.2656,
+      "step": 5320
+    },
+    {
+      "epoch": 0.3557589524318546,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00016272853382349979,
+      "loss": 1.1655,
+      "step": 5325
+    },
+    {
+      "epoch": 0.3560929983965794,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00016263767984768965,
+      "loss": 1.2054,
+      "step": 5330
+    },
+    {
+      "epoch": 0.3564270443613041,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00016254674070402731,
+      "loss": 1.1843,
+      "step": 5335
+    },
+    {
+      "epoch": 0.3567610903260289,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001624557165161618,
+      "loss": 1.2551,
+      "step": 5340
+    },
+    {
+      "epoch": 0.3570951362907536,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00016236460740785784,
+      "loss": 1.179,
+      "step": 5345
+    },
+    {
+      "epoch": 0.3574291822554784,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00016227341350299568,
+      "loss": 1.2391,
+      "step": 5350
+    },
+    {
+      "epoch": 0.3577632282202031,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00016218213492557072,
+      "loss": 1.1763,
+      "step": 5355
+    },
+    {
+      "epoch": 0.35809727418492787,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00016209077179969356,
+      "loss": 1.1736,
+      "step": 5360
+    },
+    {
+      "epoch": 0.35843132014965257,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00016199932424958984,
+      "loss": 1.1226,
+      "step": 5365
+    },
+    {
+      "epoch": 0.3587653661143773,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00016190779239959988,
+      "loss": 1.1357,
+      "step": 5370
+    },
+    {
+      "epoch": 0.35909941207910206,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00016181617637417862,
+      "loss": 1.232,
+      "step": 5375
+    },
+    {
+      "epoch": 0.3594334580438268,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00016172447629789555,
+      "loss": 1.1751,
+      "step": 5380
+    },
+    {
+      "epoch": 0.35976750400855156,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00016163269229543437,
+      "loss": 1.2229,
+      "step": 5385
+    },
+    {
+      "epoch": 0.3601015499732763,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00016154082449159284,
+      "loss": 1.2067,
+      "step": 5390
+    },
+    {
+      "epoch": 0.36043559593800106,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00016144887301128283,
+      "loss": 1.2289,
+      "step": 5395
+    },
+    {
+      "epoch": 0.3607696419027258,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00016135683797952982,
+      "loss": 1.1391,
+      "step": 5400
+    },
+    {
+      "epoch": 0.36110368786745056,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00016126471952147297,
+      "loss": 1.2043,
+      "step": 5405
+    },
+    {
+      "epoch": 0.3614377338321753,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00016117251776236492,
+      "loss": 1.2456,
+      "step": 5410
+    },
+    {
+      "epoch": 0.36177177979690006,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00016108023282757143,
+      "loss": 1.3066,
+      "step": 5415
+    },
+    {
+      "epoch": 0.3621058257616248,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00016098786484257147,
+      "loss": 1.1601,
+      "step": 5420
+    },
+    {
+      "epoch": 0.36243987172634956,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00016089541393295696,
+      "loss": 1.2307,
+      "step": 5425
+    },
+    {
+      "epoch": 0.3627739176910743,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00016080288022443241,
+      "loss": 1.2327,
+      "step": 5430
+    },
+    {
+      "epoch": 0.36310796365579906,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001607102638428151,
+      "loss": 1.1261,
+      "step": 5435
+    },
+    {
+      "epoch": 0.3634420096205238,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00016061756491403463,
+      "loss": 1.1447,
+      "step": 5440
+    },
+    {
+      "epoch": 0.3637760555852485,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00016052478356413282,
+      "loss": 1.2484,
+      "step": 5445
+    },
+    {
+      "epoch": 0.36411010154997325,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00016043191991926356,
+      "loss": 1.2338,
+      "step": 5450
+    },
+    {
+      "epoch": 0.364444147514698,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001603389741056927,
+      "loss": 1.1808,
+      "step": 5455
+    },
+    {
+      "epoch": 0.36477819347942275,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00016024594624979775,
+      "loss": 1.2165,
+      "step": 5460
+    },
+    {
+      "epoch": 0.3651122394441475,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001601528364780678,
+      "loss": 1.1568,
+      "step": 5465
+    },
+    {
+      "epoch": 0.36544628540887225,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00016005964491710328,
+      "loss": 1.2476,
+      "step": 5470
+    },
+    {
+      "epoch": 0.365780331373597,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00015996637169361593,
+      "loss": 1.2275,
+      "step": 5475
+    },
+    {
+      "epoch": 0.36611437733832175,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00015987301693442838,
+      "loss": 1.1998,
+      "step": 5480
+    },
+    {
+      "epoch": 0.3664484233030465,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00015977958076647428,
+      "loss": 1.2051,
+      "step": 5485
+    },
+    {
+      "epoch": 0.36678246926777125,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001596860633167978,
+      "loss": 1.2561,
+      "step": 5490
+    },
+    {
+      "epoch": 0.367116515232496,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001595924647125538,
+      "loss": 1.1694,
+      "step": 5495
+    },
+    {
+      "epoch": 0.36745056119722075,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00015949878508100733,
+      "loss": 1.2307,
+      "step": 5500
+    },
+    {
+      "epoch": 0.3677846071619455,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00015940502454953376,
+      "loss": 1.2637,
+      "step": 5505
+    },
+    {
+      "epoch": 0.36811865312667025,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00015931118324561837,
+      "loss": 1.1909,
+      "step": 5510
+    },
+    {
+      "epoch": 0.368452699091395,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00015921726129685624,
+      "loss": 1.2717,
+      "step": 5515
+    },
+    {
+      "epoch": 0.36878674505611975,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00015912325883095217,
+      "loss": 1.138,
+      "step": 5520
+    },
+    {
+      "epoch": 0.36912079102084444,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001590291759757204,
+      "loss": 1.1574,
+      "step": 5525
+    },
+    {
+      "epoch": 0.3694548369855692,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00015893501285908448,
+      "loss": 1.2592,
+      "step": 5530
+    },
+    {
+      "epoch": 0.36978888295029394,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00015884076960907711,
+      "loss": 1.2071,
+      "step": 5535
+    },
+    {
+      "epoch": 0.3701229289150187,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00015874644635383996,
+      "loss": 1.249,
+      "step": 5540
+    },
+    {
+      "epoch": 0.37045697487974344,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00015865204322162337,
+      "loss": 1.1818,
+      "step": 5545
+    },
+    {
+      "epoch": 0.3707910208444682,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00015855756034078647,
+      "loss": 1.2327,
+      "step": 5550
+    },
+    {
+      "epoch": 0.37112506680919294,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00015846299783979669,
+      "loss": 1.2458,
+      "step": 5555
+    },
+    {
+      "epoch": 0.3714591127739177,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001583683558472297,
+      "loss": 1.2344,
+      "step": 5560
+    },
+    {
+      "epoch": 0.37179315873864244,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001582736344917694,
+      "loss": 1.1543,
+      "step": 5565
+    },
+    {
+      "epoch": 0.3721272047033672,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00015817883390220746,
+      "loss": 1.223,
+      "step": 5570
+    },
+    {
+      "epoch": 0.37246125066809194,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00015808395420744334,
+      "loss": 1.1616,
+      "step": 5575
+    },
+    {
+      "epoch": 0.3727952966328167,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00015798899553648403,
+      "loss": 1.1649,
+      "step": 5580
+    },
+    {
+      "epoch": 0.37312934259754144,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00015789395801844397,
+      "loss": 1.1598,
+      "step": 5585
+    },
+    {
+      "epoch": 0.3734633885622662,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00015779884178254472,
+      "loss": 1.1505,
+      "step": 5590
+    },
+    {
+      "epoch": 0.37379743452699093,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00015770364695811493,
+      "loss": 1.2566,
+      "step": 5595
+    },
+    {
+      "epoch": 0.3741314804917157,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001576083736745901,
+      "loss": 1.1727,
+      "step": 5600
+    },
+    {
+      "epoch": 0.3744655264564404,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00015751302206151236,
+      "loss": 1.2458,
+      "step": 5605
+    },
+    {
+      "epoch": 0.37479957242116513,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0001574175922485304,
+      "loss": 1.1991,
+      "step": 5610
+    },
+    {
+      "epoch": 0.3751336183858899,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00015732208436539927,
+      "loss": 1.2558,
+      "step": 5615
+    },
+    {
+      "epoch": 0.3754676643506146,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00015722649854198005,
+      "loss": 1.1166,
+      "step": 5620
+    },
+    {
+      "epoch": 0.3758017103153394,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001571308349082399,
+      "loss": 1.152,
+      "step": 5625
+    },
+    {
+      "epoch": 0.3761357562800641,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00015703509359425176,
+      "loss": 1.2202,
+      "step": 5630
+    },
+    {
+      "epoch": 0.3764698022447889,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00015693927473019417,
+      "loss": 1.2383,
+      "step": 5635
+    },
+    {
+      "epoch": 0.3768038482095136,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001568433784463511,
+      "loss": 1.1833,
+      "step": 5640
+    },
+    {
+      "epoch": 0.3771378941742384,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001567474048731118,
+      "loss": 1.1411,
+      "step": 5645
+    },
+    {
+      "epoch": 0.3774719401389631,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00015665135414097065,
+      "loss": 1.1714,
+      "step": 5650
+    },
+    {
+      "epoch": 0.3778059861036879,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001565552263805269,
+      "loss": 1.2308,
+      "step": 5655
+    },
+    {
+      "epoch": 0.3781400320684126,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00015645902172248453,
+      "loss": 1.2254,
+      "step": 5660
+    },
+    {
+      "epoch": 0.3784740780331374,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00015636274029765207,
+      "loss": 1.2096,
+      "step": 5665
+    },
+    {
+      "epoch": 0.3788081239978621,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00015626638223694252,
+      "loss": 1.2076,
+      "step": 5670
+    },
+    {
+      "epoch": 0.3791421699625869,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00015616994767137294,
+      "loss": 1.2172,
+      "step": 5675
+    },
+    {
+      "epoch": 0.3794762159273116,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001560734367320645,
+      "loss": 1.2691,
+      "step": 5680
+    },
+    {
+      "epoch": 0.37981026189203637,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00015597684955024222,
+      "loss": 1.1919,
+      "step": 5685
+    },
+    {
+      "epoch": 0.38014430785676107,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00015588018625723477,
+      "loss": 1.2075,
+      "step": 5690
+    },
+    {
+      "epoch": 0.3804783538214858,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00015578344698447428,
+      "loss": 1.1873,
+      "step": 5695
+    },
+    {
+      "epoch": 0.38081239978621056,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001556866318634962,
+      "loss": 1.2008,
+      "step": 5700
+    },
+    {
+      "epoch": 0.3811464457509353,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00015558974102593913,
+      "loss": 1.1994,
+      "step": 5705
+    },
+    {
+      "epoch": 0.38148049171566006,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0001554927746035446,
+      "loss": 1.241,
+      "step": 5710
+    },
+    {
+      "epoch": 0.3818145376803848,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00015539573272815697,
+      "loss": 1.2538,
+      "step": 5715
+    },
+    {
+      "epoch": 0.38214858364510956,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00015529861553172314,
+      "loss": 1.1708,
+      "step": 5720
+    },
+    {
+      "epoch": 0.3824826296098343,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00015520142314629239,
+      "loss": 1.1842,
+      "step": 5725
+    },
+    {
+      "epoch": 0.38281667557455906,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00015510415570401626,
+      "loss": 1.1753,
+      "step": 5730
+    },
+    {
+      "epoch": 0.3831507215392838,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001550068133371484,
+      "loss": 1.2631,
+      "step": 5735
+    },
+    {
+      "epoch": 0.38348476750400856,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001549093961780443,
+      "loss": 1.2291,
+      "step": 5740
+    },
+    {
+      "epoch": 0.3838188134687333,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001548119043591611,
+      "loss": 1.228,
+      "step": 5745
+    },
+    {
+      "epoch": 0.38415285943345806,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00015471433801305756,
+      "loss": 1.1866,
+      "step": 5750
+    },
+    {
+      "epoch": 0.3844869053981828,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00015461669727239363,
+      "loss": 1.1626,
+      "step": 5755
+    },
+    {
+      "epoch": 0.38482095136290756,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001545189822699305,
+      "loss": 1.277,
+      "step": 5760
+    },
+    {
+      "epoch": 0.3851549973276323,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00015442119313853033,
+      "loss": 1.219,
+      "step": 5765
+    },
+    {
+      "epoch": 0.385489043292357,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001543233300111561,
+      "loss": 1.2212,
+      "step": 5770
+    },
+    {
+      "epoch": 0.38582308925708175,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001542253930208713,
+      "loss": 1.2303,
+      "step": 5775
+    },
+    {
+      "epoch": 0.3861571352218065,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00015412738230083993,
+      "loss": 1.2315,
+      "step": 5780
+    },
+    {
+      "epoch": 0.38649118118653125,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00015402929798432629,
+      "loss": 1.224,
+      "step": 5785
+    },
+    {
+      "epoch": 0.386825227151256,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00015393114020469462,
+      "loss": 1.1687,
+      "step": 5790
+    },
+    {
+      "epoch": 0.38715927311598075,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001538329090954091,
+      "loss": 1.1959,
+      "step": 5795
+    },
+    {
+      "epoch": 0.3874933190807055,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001537346047900337,
+      "loss": 1.2247,
+      "step": 5800
+    },
+    {
+      "epoch": 0.38782736504543025,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00015363622742223175,
+      "loss": 1.2288,
+      "step": 5805
+    },
+    {
+      "epoch": 0.388161411010155,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001535377771257661,
+      "loss": 1.2688,
+      "step": 5810
+    },
+    {
+      "epoch": 0.38849545697487975,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0001534392540344986,
+      "loss": 1.1998,
+      "step": 5815
+    },
+    {
+      "epoch": 0.3888295029396045,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00015334065828239023,
+      "loss": 1.1662,
+      "step": 5820
+    },
+    {
+      "epoch": 0.38916354890432925,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00015324199000350062,
+      "loss": 1.2877,
+      "step": 5825
+    },
+    {
+      "epoch": 0.389497594869054,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00015314324933198806,
+      "loss": 1.2341,
+      "step": 5830
+    },
+    {
+      "epoch": 0.38983164083377875,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001530444364021094,
+      "loss": 1.1665,
+      "step": 5835
+    },
+    {
+      "epoch": 0.3901656867985035,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00015294555134821956,
+      "loss": 1.2566,
+      "step": 5840
+    },
+    {
+      "epoch": 0.39049973276322825,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001528465943047716,
+      "loss": 1.1577,
+      "step": 5845
+    },
+    {
+      "epoch": 0.39083377872795294,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00015274756540631644,
+      "loss": 1.1293,
+      "step": 5850
+    },
+    {
+      "epoch": 0.3911678246926777,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00015264846478750278,
+      "loss": 1.1906,
+      "step": 5855
+    },
+    {
+      "epoch": 0.39150187065740244,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00015254929258307678,
+      "loss": 1.2896,
+      "step": 5860
+    },
+    {
+      "epoch": 0.3918359166221272,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001524500489278819,
+      "loss": 1.1557,
+      "step": 5865
+    },
+    {
+      "epoch": 0.39216996258685194,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00015235073395685877,
+      "loss": 1.1846,
+      "step": 5870
+    },
+    {
+      "epoch": 0.3925040085515767,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00015225134780504505,
+      "loss": 1.2353,
+      "step": 5875
+    },
+    {
+      "epoch": 0.39283805451630144,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00015215189060757507,
+      "loss": 1.1962,
+      "step": 5880
+    },
+    {
+      "epoch": 0.3931721004810262,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00015205236249967995,
+      "loss": 1.1898,
+      "step": 5885
+    },
+    {
+      "epoch": 0.39350614644575094,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.000151952763616687,
+      "loss": 1.2293,
+      "step": 5890
+    },
+    {
+      "epoch": 0.3938401924104757,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00015185309409401985,
+      "loss": 1.1423,
+      "step": 5895
+    },
+    {
+      "epoch": 0.39417423837520044,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00015175335406719827,
+      "loss": 1.1552,
+      "step": 5900
+    },
+    {
+      "epoch": 0.3945082843399252,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00015165354367183777,
+      "loss": 1.1193,
+      "step": 5905
+    },
+    {
+      "epoch": 0.39484233030464994,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00015155366304364962,
+      "loss": 1.1415,
+      "step": 5910
+    },
+    {
+      "epoch": 0.3951763762693747,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00015145371231844047,
+      "loss": 1.1684,
+      "step": 5915
+    },
+    {
+      "epoch": 0.39551042223409943,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00015135369163211252,
+      "loss": 1.1763,
+      "step": 5920
+    },
+    {
+      "epoch": 0.3958444681988242,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00015125360112066275,
+      "loss": 1.11,
+      "step": 5925
+    },
+    {
+      "epoch": 0.3961785141635489,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001511534409201834,
+      "loss": 1.1879,
+      "step": 5930
+    },
+    {
+      "epoch": 0.39651256012827363,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00015105321116686132,
+      "loss": 1.2056,
+      "step": 5935
+    },
+    {
+      "epoch": 0.3968466060929984,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00015095291199697784,
+      "loss": 1.187,
+      "step": 5940
+    },
+    {
+      "epoch": 0.3971806520577231,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001508525435469089,
+      "loss": 1.2731,
+      "step": 5945
+    },
+    {
+      "epoch": 0.3975146980224479,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00015075210595312448,
+      "loss": 1.1658,
+      "step": 5950
+    },
+    {
+      "epoch": 0.3978487439871726,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001506515993521886,
+      "loss": 1.154,
+      "step": 5955
+    },
+    {
+      "epoch": 0.3981827899518974,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001505510238807591,
+      "loss": 1.195,
+      "step": 5960
+    },
+    {
+      "epoch": 0.3985168359166221,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00015045037967558754,
+      "loss": 1.2198,
+      "step": 5965
+    },
+    {
+      "epoch": 0.3988508818813469,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00015034966687351884,
+      "loss": 1.1808,
+      "step": 5970
+    },
+    {
+      "epoch": 0.3991849278460716,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00015024888561149125,
+      "loss": 1.1722,
+      "step": 5975
+    },
+    {
+      "epoch": 0.3995189738107964,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00015014803602653607,
+      "loss": 1.1929,
+      "step": 5980
+    },
+    {
+      "epoch": 0.3998530197755211,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001500471182557775,
+      "loss": 1.1078,
+      "step": 5985
+    },
+    {
+      "epoch": 0.4001870657402459,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00014994613243643248,
+      "loss": 1.193,
+      "step": 5990
+    },
+    {
+      "epoch": 0.4005211117049706,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00014984507870581046,
+      "loss": 1.1894,
+      "step": 5995
+    },
+    {
+      "epoch": 0.40085515766969537,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00014974395720131328,
+      "loss": 1.2523,
+      "step": 6000
+    },
+    {
+      "epoch": 0.4011892036344201,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00014964276806043477,
+      "loss": 1.2658,
+      "step": 6005
+    },
+    {
+      "epoch": 0.4015232495991448,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001495415114207609,
+      "loss": 1.2058,
+      "step": 6010
+    },
+    {
+      "epoch": 0.40185729556386957,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001494401874199694,
+      "loss": 1.1817,
+      "step": 6015
+    },
+    {
+      "epoch": 0.4021913415285943,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00014933879619582943,
+      "loss": 1.1867,
+      "step": 6020
+    },
+    {
+      "epoch": 0.40252538749331906,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00014923733788620175,
+      "loss": 1.2503,
+      "step": 6025
+    },
+    {
+      "epoch": 0.4028594334580438,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001491358126290382,
+      "loss": 1.2104,
+      "step": 6030
+    },
+    {
+      "epoch": 0.40319347942276856,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001490342205623817,
+      "loss": 1.2023,
+      "step": 6035
+    },
+    {
+      "epoch": 0.4035275253874933,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00014893256182436609,
+      "loss": 1.154,
+      "step": 6040
+    },
+    {
+      "epoch": 0.40386157135221806,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00014883083655321567,
+      "loss": 1.0871,
+      "step": 6045
+    },
+    {
+      "epoch": 0.4041956173169428,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00014872904488724535,
+      "loss": 1.1745,
+      "step": 6050
+    },
+    {
+      "epoch": 0.40452966328166756,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001486271869648603,
+      "loss": 1.2076,
+      "step": 6055
+    },
+    {
+      "epoch": 0.4048637092463923,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00014852526292455576,
+      "loss": 1.2039,
+      "step": 6060
+    },
+    {
+      "epoch": 0.40519775521111706,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00014842327290491688,
+      "loss": 1.1722,
+      "step": 6065
+    },
+    {
+      "epoch": 0.4055318011758418,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00014832121704461848,
+      "loss": 1.1733,
+      "step": 6070
+    },
+    {
+      "epoch": 0.40586584714056656,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014821909548242497,
+      "loss": 1.2495,
+      "step": 6075
+    },
+    {
+      "epoch": 0.4061998931052913,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00014811690835718998,
+      "loss": 1.1307,
+      "step": 6080
+    },
+    {
+      "epoch": 0.40653393907001606,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00014801465580785648,
+      "loss": 1.2543,
+      "step": 6085
+    },
+    {
+      "epoch": 0.40686798503474075,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00014791233797345618,
+      "loss": 1.1583,
+      "step": 6090
+    },
+    {
+      "epoch": 0.4072020309994655,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00014780995499310973,
+      "loss": 1.2537,
+      "step": 6095
+    },
+    {
+      "epoch": 0.40753607696419025,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00014770750700602623,
+      "loss": 1.26,
+      "step": 6100
+    },
+    {
+      "epoch": 0.407870122928915,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00014760499415150327,
+      "loss": 1.2814,
+      "step": 6105
+    },
+    {
+      "epoch": 0.40820416889363975,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00014750241656892653,
+      "loss": 1.1918,
+      "step": 6110
+    },
+    {
+      "epoch": 0.4085382148583645,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014739977439776983,
+      "loss": 1.2569,
+      "step": 6115
+    },
+    {
+      "epoch": 0.40887226082308925,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00014729706777759474,
+      "loss": 1.2649,
+      "step": 6120
+    },
+    {
+      "epoch": 0.409206306787814,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00014719429684805041,
+      "loss": 1.2275,
+      "step": 6125
+    },
+    {
+      "epoch": 0.40954035275253875,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00014709146174887356,
+      "loss": 1.2521,
+      "step": 6130
+    },
+    {
+      "epoch": 0.4098743987172635,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00014698856261988804,
+      "loss": 1.2256,
+      "step": 6135
+    },
+    {
+      "epoch": 0.41020844468198825,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00014688559960100483,
+      "loss": 1.1749,
+      "step": 6140
+    },
+    {
+      "epoch": 0.410542490646713,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001467825728322217,
+      "loss": 1.1549,
+      "step": 6145
+    },
+    {
+      "epoch": 0.41087653661143775,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00014667948245362329,
+      "loss": 1.2315,
+      "step": 6150
+    },
+    {
+      "epoch": 0.4112105825761625,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00014657632860538047,
+      "loss": 1.1947,
+      "step": 6155
+    },
+    {
+      "epoch": 0.41154462854088725,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00014647311142775056,
+      "loss": 1.2204,
+      "step": 6160
+    },
+    {
+      "epoch": 0.411878674505612,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00014636983106107703,
+      "loss": 1.2413,
+      "step": 6165
+    },
+    {
+      "epoch": 0.4122127204703367,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00014626648764578916,
+      "loss": 1.1824,
+      "step": 6170
+    },
+    {
+      "epoch": 0.41254676643506144,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.000146163081322402,
+      "loss": 1.2953,
+      "step": 6175
+    },
+    {
+      "epoch": 0.4128808123997862,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00014605961223151614,
+      "loss": 1.1684,
+      "step": 6180
+    },
+    {
+      "epoch": 0.41321485836451094,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00014595608051381752,
+      "loss": 1.1756,
+      "step": 6185
+    },
+    {
+      "epoch": 0.4135489043292357,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001458524863100772,
+      "loss": 1.2342,
+      "step": 6190
+    },
+    {
+      "epoch": 0.41388295029396044,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00014574882976115124,
+      "loss": 1.2005,
+      "step": 6195
+    },
+    {
+      "epoch": 0.4142169962586852,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00014564511100798044,
+      "loss": 1.2599,
+      "step": 6200
+    },
+    {
+      "epoch": 0.41455104222340994,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00014554133019159022,
+      "loss": 1.2165,
+      "step": 6205
+    },
+    {
+      "epoch": 0.4148850881881347,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00014543748745309034,
+      "loss": 1.1746,
+      "step": 6210
+    },
+    {
+      "epoch": 0.41521913415285944,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00014533358293367481,
+      "loss": 1.2066,
+      "step": 6215
+    },
+    {
+      "epoch": 0.4155531801175842,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00014522961677462153,
+      "loss": 1.1775,
+      "step": 6220
+    },
+    {
+      "epoch": 0.41588722608230894,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001451255891172924,
+      "loss": 1.1441,
+      "step": 6225
+    },
+    {
+      "epoch": 0.4162212720470337,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001450215001031327,
+      "loss": 1.2242,
+      "step": 6230
+    },
+    {
+      "epoch": 0.41655531801175844,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00014491734987367137,
+      "loss": 1.2,
+      "step": 6235
+    },
+    {
+      "epoch": 0.4168893639764832,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00014481313857052044,
+      "loss": 1.2246,
+      "step": 6240
+    },
+    {
+      "epoch": 0.41722340994120793,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00014470886633537498,
+      "loss": 1.1961,
+      "step": 6245
+    },
+    {
+      "epoch": 0.4175574559059327,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.000144604533310013,
+      "loss": 1.2179,
+      "step": 6250
+    },
+    {
+      "epoch": 0.4178915018706574,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00014450013963629508,
+      "loss": 1.1481,
+      "step": 6255
+    },
+    {
+      "epoch": 0.4182255478353821,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00014439568545616437,
+      "loss": 1.1695,
+      "step": 6260
+    },
+    {
+      "epoch": 0.4185595938001069,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001442911709116461,
+      "loss": 1.1626,
+      "step": 6265
+    },
+    {
+      "epoch": 0.4188936397648316,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001441865961448478,
+      "loss": 1.1622,
+      "step": 6270
+    },
+    {
+      "epoch": 0.4192276857295564,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001440819612979587,
+      "loss": 1.2383,
+      "step": 6275
+    },
+    {
+      "epoch": 0.4195617316942811,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00014397726651324983,
+      "loss": 1.2379,
+      "step": 6280
+    },
+    {
+      "epoch": 0.4198957776590059,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00014387251193307367,
+      "loss": 1.1498,
+      "step": 6285
+    },
+    {
+      "epoch": 0.4202298236237306,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014376769769986405,
+      "loss": 1.2617,
+      "step": 6290
+    },
+    {
+      "epoch": 0.4205638695884554,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00014366282395613587,
+      "loss": 1.3093,
+      "step": 6295
+    },
+    {
+      "epoch": 0.4208979155531801,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001435578908444849,
+      "loss": 1.1134,
+      "step": 6300
+    },
+    {
+      "epoch": 0.4212319615179049,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00014345289850758777,
+      "loss": 1.2403,
+      "step": 6305
+    },
+    {
+      "epoch": 0.4215660074826296,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00014334784708820144,
+      "loss": 1.1466,
+      "step": 6310
+    },
+    {
+      "epoch": 0.4219000534473544,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00014324273672916343,
+      "loss": 1.1237,
+      "step": 6315
+    },
+    {
+      "epoch": 0.4222340994120791,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00014313756757339122,
+      "loss": 1.2737,
+      "step": 6320
+    },
+    {
+      "epoch": 0.42256814537680387,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00014303233976388236,
+      "loss": 1.1712,
+      "step": 6325
+    },
+    {
+      "epoch": 0.4229021913415286,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00014292705344371402,
+      "loss": 1.1599,
+      "step": 6330
+    },
+    {
+      "epoch": 0.4232362373062533,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00014282170875604307,
+      "loss": 1.163,
+      "step": 6335
+    },
+    {
+      "epoch": 0.42357028327097807,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00014271630584410558,
+      "loss": 1.1546,
+      "step": 6340
+    },
+    {
+      "epoch": 0.4239043292357028,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00014261084485121697,
+      "loss": 1.232,
+      "step": 6345
+    },
+    {
+      "epoch": 0.42423837520042756,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00014250532592077148,
+      "loss": 1.2507,
+      "step": 6350
+    },
+    {
+      "epoch": 0.4245724211651523,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00014239974919624224,
+      "loss": 1.216,
+      "step": 6355
+    },
+    {
+      "epoch": 0.42490646712987706,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00014229411482118083,
+      "loss": 1.2722,
+      "step": 6360
+    },
+    {
+      "epoch": 0.4252405130946018,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00014218842293921738,
+      "loss": 1.1772,
+      "step": 6365
+    },
+    {
+      "epoch": 0.42557455905932656,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00014208267369406012,
+      "loss": 1.2117,
+      "step": 6370
+    },
+    {
+      "epoch": 0.4259086050240513,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001419768672294952,
+      "loss": 1.1975,
+      "step": 6375
+    },
+    {
+      "epoch": 0.42624265098877606,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00014187100368938678,
+      "loss": 1.1905,
+      "step": 6380
+    },
+    {
+      "epoch": 0.4265766969535008,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00014176508321767637,
+      "loss": 1.1521,
+      "step": 6385
+    },
+    {
+      "epoch": 0.42691074291822556,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00014165910595838313,
+      "loss": 1.2083,
+      "step": 6390
+    },
+    {
+      "epoch": 0.4272447888829503,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00014155307205560323,
+      "loss": 1.2467,
+      "step": 6395
+    },
+    {
+      "epoch": 0.42757883484767506,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00014144698165351,
+      "loss": 1.2489,
+      "step": 6400
+    },
+    {
+      "epoch": 0.4279128808123998,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00014134083489635355,
+      "loss": 1.2708,
+      "step": 6405
+    },
+    {
+      "epoch": 0.42824692677712456,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00014123463192846058,
+      "loss": 1.1689,
+      "step": 6410
+    },
+    {
+      "epoch": 0.42858097274184925,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00014112837289423426,
+      "loss": 1.1623,
+      "step": 6415
+    },
+    {
+      "epoch": 0.428915018706574,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00014102205793815398,
+      "loss": 1.1605,
+      "step": 6420
+    },
+    {
+      "epoch": 0.42924906467129875,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00014091568720477518,
+      "loss": 1.288,
+      "step": 6425
+    },
+    {
+      "epoch": 0.4295831106360235,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00014080926083872907,
+      "loss": 1.1916,
+      "step": 6430
+    },
+    {
+      "epoch": 0.42991715660074825,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00014070277898472263,
+      "loss": 1.2131,
+      "step": 6435
+    },
+    {
+      "epoch": 0.430251202565473,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00014059624178753817,
+      "loss": 1.1224,
+      "step": 6440
+    },
+    {
+      "epoch": 0.43058524853019775,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001404896493920333,
+      "loss": 1.244,
+      "step": 6445
+    },
+    {
+      "epoch": 0.4309192944949225,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001403830019431407,
+      "loss": 1.232,
+      "step": 6450
+    },
+    {
+      "epoch": 0.43125334045964725,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00014027629958586788,
+      "loss": 1.2099,
+      "step": 6455
+    },
+    {
+      "epoch": 0.431587386424372,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00014016954246529696,
+      "loss": 1.229,
+      "step": 6460
+    },
+    {
+      "epoch": 0.43192143238909675,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00014006273072658462,
+      "loss": 1.239,
+      "step": 6465
+    },
+    {
+      "epoch": 0.4322554783538215,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00013995586451496177,
+      "loss": 1.2018,
+      "step": 6470
+    },
+    {
+      "epoch": 0.43258952431854625,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001398489439757333,
+      "loss": 1.1465,
+      "step": 6475
+    },
+    {
+      "epoch": 0.432923570283271,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00013974196925427816,
+      "loss": 1.2251,
+      "step": 6480
+    },
+    {
+      "epoch": 0.43325761624799575,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00013963494049604871,
+      "loss": 1.2515,
+      "step": 6485
+    },
+    {
+      "epoch": 0.4335916622127205,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00013952785784657106,
+      "loss": 1.2291,
+      "step": 6490
+    },
+    {
+      "epoch": 0.4339257081774452,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001394207214514444,
+      "loss": 1.2463,
+      "step": 6495
+    },
+    {
+      "epoch": 0.43425975414216994,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00013931353145634102,
+      "loss": 1.152,
+      "step": 6500
+    },
+    {
+      "epoch": 0.4345938001068947,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001392062880070062,
+      "loss": 1.2804,
+      "step": 6505
+    },
+    {
+      "epoch": 0.43492784607161944,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00013909899124925774,
+      "loss": 1.1706,
+      "step": 6510
+    },
+    {
+      "epoch": 0.4352618920363442,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001389916413289861,
+      "loss": 1.1676,
+      "step": 6515
+    },
+    {
+      "epoch": 0.43559593800106894,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00013888423839215395,
+      "loss": 1.1549,
+      "step": 6520
+    },
+    {
+      "epoch": 0.4359299839657937,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001387767825847959,
+      "loss": 1.2364,
+      "step": 6525
+    },
+    {
+      "epoch": 0.43626402993051844,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001386692740530187,
+      "loss": 1.244,
+      "step": 6530
+    },
+    {
+      "epoch": 0.4365980758952432,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00013856171294300066,
+      "loss": 1.1887,
+      "step": 6535
+    },
+    {
+      "epoch": 0.43693212185996794,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00013845409940099152,
+      "loss": 1.0936,
+      "step": 6540
+    },
+    {
+      "epoch": 0.4372661678246927,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00013834643357331245,
+      "loss": 1.2448,
+      "step": 6545
+    },
+    {
+      "epoch": 0.43760021378941744,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001382387156063556,
+      "loss": 1.1776,
+      "step": 6550
+    },
+    {
+      "epoch": 0.4379342597541422,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001381309456465841,
+      "loss": 1.2031,
+      "step": 6555
+    },
+    {
+      "epoch": 0.43826830571886694,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001380231238405317,
+      "loss": 1.2349,
+      "step": 6560
+    },
+    {
+      "epoch": 0.4386023516835917,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00013791525033480268,
+      "loss": 1.1826,
+      "step": 6565
+    },
+    {
+      "epoch": 0.43893639764831643,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00013780732527607156,
+      "loss": 1.2273,
+      "step": 6570
+    },
+    {
+      "epoch": 0.43927044361304113,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00013769934881108312,
+      "loss": 1.1712,
+      "step": 6575
+    },
+    {
+      "epoch": 0.4396044895777659,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00013759132108665182,
+      "loss": 1.1598,
+      "step": 6580
+    },
+    {
+      "epoch": 0.4399385355424906,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.000137483242249662,
+      "loss": 1.1881,
+      "step": 6585
+    },
+    {
+      "epoch": 0.4402725815072154,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00013737511244706733,
+      "loss": 1.1738,
+      "step": 6590
+    },
+    {
+      "epoch": 0.4406066274719401,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00013726693182589093,
+      "loss": 1.2403,
+      "step": 6595
+    },
+    {
+      "epoch": 0.4409406734366649,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00013715870053322492,
+      "loss": 1.1563,
+      "step": 6600
+    },
+    {
+      "epoch": 0.4412747194013896,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001370504187162304,
+      "loss": 1.1397,
+      "step": 6605
+    },
+    {
+      "epoch": 0.4416087653661144,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00013694208652213703,
+      "loss": 1.1837,
+      "step": 6610
+    },
+    {
+      "epoch": 0.4419428113308391,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00013683370409824317,
+      "loss": 1.196,
+      "step": 6615
+    },
+    {
+      "epoch": 0.4422768572955639,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00013672527159191525,
+      "loss": 1.2161,
+      "step": 6620
+    },
+    {
+      "epoch": 0.4426109032602886,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00013661678915058797,
+      "loss": 1.2789,
+      "step": 6625
+    },
+    {
+      "epoch": 0.4429449492250134,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00013650825692176387,
+      "loss": 1.2474,
+      "step": 6630
+    },
+    {
+      "epoch": 0.4432789951897381,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00013639967505301313,
+      "loss": 1.2226,
+      "step": 6635
+    },
+    {
+      "epoch": 0.4436130411544629,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00013629104369197351,
+      "loss": 1.2119,
+      "step": 6640
+    },
+    {
+      "epoch": 0.4439470871191876,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00013618236298635003,
+      "loss": 1.1872,
+      "step": 6645
+    },
+    {
+      "epoch": 0.44428113308391237,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0001360736330839148,
+      "loss": 1.2023,
+      "step": 6650
+    },
+    {
+      "epoch": 0.44461517904863707,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00013596485413250683,
+      "loss": 1.1973,
+      "step": 6655
+    },
+    {
+      "epoch": 0.4449492250133618,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001358560262800318,
+      "loss": 1.1778,
+      "step": 6660
+    },
+    {
+      "epoch": 0.44528327097808656,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00013574714967446192,
+      "loss": 1.1713,
+      "step": 6665
+    },
+    {
+      "epoch": 0.4456173169428113,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00013563822446383564,
+      "loss": 1.2628,
+      "step": 6670
+    },
+    {
+      "epoch": 0.44595136290753606,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00013552925079625755,
+      "loss": 1.2013,
+      "step": 6675
+    },
+    {
+      "epoch": 0.4462854088722608,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00013542022881989803,
+      "loss": 1.2194,
+      "step": 6680
+    },
+    {
+      "epoch": 0.44661945483698556,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00013531115868299336,
+      "loss": 1.1687,
+      "step": 6685
+    },
+    {
+      "epoch": 0.4469535008017103,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.000135202040533845,
+      "loss": 1.1385,
+      "step": 6690
+    },
+    {
+      "epoch": 0.44728754676643506,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00013509287452081995,
+      "loss": 1.225,
+      "step": 6695
+    },
+    {
+      "epoch": 0.4476215927311598,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00013498366079235015,
+      "loss": 1.2127,
+      "step": 6700
+    },
+    {
+      "epoch": 0.44795563869588456,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001348743994969325,
+      "loss": 1.2052,
+      "step": 6705
+    },
+    {
+      "epoch": 0.4482896846606093,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00013476509078312845,
+      "loss": 1.2059,
+      "step": 6710
+    },
+    {
+      "epoch": 0.44862373062533406,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0001346557347995641,
+      "loss": 1.1859,
+      "step": 6715
+    },
+    {
+      "epoch": 0.4489577765900588,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00013454633169492967,
+      "loss": 1.1002,
+      "step": 6720
+    },
+    {
+      "epoch": 0.44929182255478356,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00013443688161797953,
+      "loss": 1.1503,
+      "step": 6725
+    },
+    {
+      "epoch": 0.4496258685195083,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00013432738471753195,
+      "loss": 1.1687,
+      "step": 6730
+    },
+    {
+      "epoch": 0.449959914484233,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00013421784114246873,
+      "loss": 1.2168,
+      "step": 6735
+    },
+    {
+      "epoch": 0.45029396044895775,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00013410825104173528,
+      "loss": 1.2876,
+      "step": 6740
+    },
+    {
+      "epoch": 0.4506280064136825,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00013399861456434017,
+      "loss": 1.2127,
+      "step": 6745
+    },
+    {
+      "epoch": 0.45096205237840725,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00013388893185935512,
+      "loss": 1.229,
+      "step": 6750
+    },
+    {
+      "epoch": 0.451296098343132,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00013377920307591453,
+      "loss": 1.2186,
+      "step": 6755
+    },
+    {
+      "epoch": 0.45163014430785675,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00013366942836321575,
+      "loss": 1.1846,
+      "step": 6760
+    },
+    {
+      "epoch": 0.4519641902725815,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00013355960787051827,
+      "loss": 1.2155,
+      "step": 6765
+    },
+    {
+      "epoch": 0.45229823623730625,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.000133449741747144,
+      "loss": 1.2372,
+      "step": 6770
+    },
+    {
+      "epoch": 0.452632282202031,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00013333983014247687,
+      "loss": 1.2568,
+      "step": 6775
+    },
+    {
+      "epoch": 0.45296632816675575,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001332298732059626,
+      "loss": 1.1603,
+      "step": 6780
+    },
+    {
+      "epoch": 0.4533003741314805,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001331198710871086,
+      "loss": 1.309,
+      "step": 6785
+    },
+    {
+      "epoch": 0.45363442009620525,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00013300982393548368,
+      "loss": 1.1875,
+      "step": 6790
+    },
+    {
+      "epoch": 0.45396846606093,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00013289973190071797,
+      "loss": 1.2092,
+      "step": 6795
+    },
+    {
+      "epoch": 0.45430251202565475,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00013278959513250243,
+      "loss": 1.2222,
+      "step": 6800
+    },
+    {
+      "epoch": 0.4546365579903795,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001326794137805891,
+      "loss": 1.1878,
+      "step": 6805
+    },
+    {
+      "epoch": 0.45497060395510425,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001325691879947904,
+      "loss": 1.1878,
+      "step": 6810
+    },
+    {
+      "epoch": 0.45530464991982894,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001324589179249793,
+      "loss": 1.2343,
+      "step": 6815
+    },
+    {
+      "epoch": 0.4556386958845537,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.000132348603721089,
+      "loss": 1.2661,
+      "step": 6820
+    },
+    {
+      "epoch": 0.45597274184927844,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00013223824553311263,
+      "loss": 1.3178,
+      "step": 6825
+    },
+    {
+      "epoch": 0.4563067878140032,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00013212784351110312,
+      "loss": 1.1792,
+      "step": 6830
+    },
+    {
+      "epoch": 0.45664083377872794,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00013201739780517311,
+      "loss": 1.1898,
+      "step": 6835
+    },
+    {
+      "epoch": 0.4569748797434527,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00013190690856549456,
+      "loss": 1.2211,
+      "step": 6840
+    },
+    {
+      "epoch": 0.45730892570817744,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00013179637594229858,
+      "loss": 1.1925,
+      "step": 6845
+    },
+    {
+      "epoch": 0.4576429716729022,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00013168580008587536,
+      "loss": 1.2215,
+      "step": 6850
+    },
+    {
+      "epoch": 0.45797701763762694,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001315751811465738,
+      "loss": 1.1885,
+      "step": 6855
+    },
+    {
+      "epoch": 0.4583110636023517,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00013146451927480146,
+      "loss": 1.221,
+      "step": 6860
+    },
+    {
+      "epoch": 0.45864510956707644,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00013135381462102413,
+      "loss": 1.2224,
+      "step": 6865
+    },
+    {
+      "epoch": 0.4589791555318012,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.000131243067335766,
+      "loss": 1.1484,
+      "step": 6870
+    },
+    {
+      "epoch": 0.45931320149652594,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00013113227756960898,
+      "loss": 1.2461,
+      "step": 6875
+    },
+    {
+      "epoch": 0.4596472474612507,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00013102144547319286,
+      "loss": 1.2175,
+      "step": 6880
+    },
+    {
+      "epoch": 0.45998129342597543,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00013091057119721505,
+      "loss": 1.2129,
+      "step": 6885
+    },
+    {
+      "epoch": 0.4603153393907002,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00013079965489243015,
+      "loss": 1.1792,
+      "step": 6890
+    },
+    {
+      "epoch": 0.46064938535542493,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00013068869670965008,
+      "loss": 1.195,
+      "step": 6895
+    },
+    {
+      "epoch": 0.46098343132014963,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00013057769679974358,
+      "loss": 1.1676,
+      "step": 6900
+    },
+    {
+      "epoch": 0.4613174772848744,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00013046665531363615,
+      "loss": 1.1518,
+      "step": 6905
+    },
+    {
+      "epoch": 0.4616515232495991,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00013035557240230982,
+      "loss": 1.2451,
+      "step": 6910
+    },
+    {
+      "epoch": 0.4619855692143239,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00013024444821680304,
+      "loss": 1.2153,
+      "step": 6915
+    },
+    {
+      "epoch": 0.4623196151790486,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001301332829082102,
+      "loss": 1.2301,
+      "step": 6920
+    },
+    {
+      "epoch": 0.4626536611437734,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00013002207662768175,
+      "loss": 1.2094,
+      "step": 6925
+    },
+    {
+      "epoch": 0.4629877071084981,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001299108295264238,
+      "loss": 1.2514,
+      "step": 6930
+    },
+    {
+      "epoch": 0.4633217530732229,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00012979954175569797,
+      "loss": 1.2319,
+      "step": 6935
+    },
+    {
+      "epoch": 0.4636557990379476,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00012968821346682113,
+      "loss": 1.1269,
+      "step": 6940
+    },
+    {
+      "epoch": 0.4639898450026724,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00012957684481116537,
+      "loss": 1.1569,
+      "step": 6945
+    },
+    {
+      "epoch": 0.4643238909673971,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00012946543594015753,
+      "loss": 1.1848,
+      "step": 6950
+    },
+    {
+      "epoch": 0.4646579369321219,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00012935398700527915,
+      "loss": 1.1894,
+      "step": 6955
+    },
+    {
+      "epoch": 0.4649919828968466,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00012924249815806632,
+      "loss": 1.2352,
+      "step": 6960
+    },
+    {
+      "epoch": 0.4653260288615714,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00012913096955010937,
+      "loss": 1.2918,
+      "step": 6965
+    },
+    {
+      "epoch": 0.4656600748262961,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00012901940133305267,
+      "loss": 1.2738,
+      "step": 6970
+    },
+    {
+      "epoch": 0.46599412079102087,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00012890779365859443,
+      "loss": 1.0624,
+      "step": 6975
+    },
+    {
+      "epoch": 0.46632816675574557,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00012879614667848655,
+      "loss": 1.2576,
+      "step": 6980
+    },
+    {
+      "epoch": 0.4666622127204703,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00012868446054453434,
+      "loss": 1.2957,
+      "step": 6985
+    },
+    {
+      "epoch": 0.46699625868519506,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00012857273540859643,
+      "loss": 1.1472,
+      "step": 6990
+    },
+    {
+      "epoch": 0.4673303046499198,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0001284609714225843,
+      "loss": 1.1081,
+      "step": 6995
+    },
+    {
+      "epoch": 0.46766435061464456,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00012834916873846245,
+      "loss": 1.2406,
+      "step": 7000
+    },
+    {
+      "epoch": 0.4679983965793693,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00012823732750824794,
+      "loss": 1.2261,
+      "step": 7005
+    },
+    {
+      "epoch": 0.46833244254409406,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00012812544788401014,
+      "loss": 1.2323,
+      "step": 7010
+    },
+    {
+      "epoch": 0.4686664885088188,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00012801353001787072,
+      "loss": 1.1986,
+      "step": 7015
+    },
+    {
+      "epoch": 0.46900053447354356,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001279015740620034,
+      "loss": 1.1924,
+      "step": 7020
+    },
+    {
+      "epoch": 0.4693345804382683,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00012778958016863357,
+      "loss": 1.204,
+      "step": 7025
+    },
+    {
+      "epoch": 0.46966862640299306,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0001276775484900382,
+      "loss": 1.206,
+      "step": 7030
+    },
+    {
+      "epoch": 0.4700026723677178,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00012756547917854578,
+      "loss": 1.2013,
+      "step": 7035
+    },
+    {
+      "epoch": 0.47033671833244256,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0001274533723865358,
+      "loss": 1.2634,
+      "step": 7040
+    },
+    {
+      "epoch": 0.4706707642971673,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00012734122826643884,
+      "loss": 1.302,
+      "step": 7045
+    },
+    {
+      "epoch": 0.47100481026189206,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00012722904697073616,
+      "loss": 1.2133,
+      "step": 7050
+    },
+    {
+      "epoch": 0.4713388562266168,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00012711682865195964,
+      "loss": 1.1755,
+      "step": 7055
+    },
+    {
+      "epoch": 0.4716729021913415,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00012700457346269137,
+      "loss": 1.2221,
+      "step": 7060
+    },
+    {
+      "epoch": 0.47200694815606625,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00012689228155556373,
+      "loss": 1.2812,
+      "step": 7065
+    },
+    {
+      "epoch": 0.472340994120791,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00012677995308325887,
+      "loss": 1.2279,
+      "step": 7070
+    },
+    {
+      "epoch": 0.47267504008551575,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001266675881985088,
+      "loss": 1.1915,
+      "step": 7075
+    },
+    {
+      "epoch": 0.4730090860502405,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00012655518705409496,
+      "loss": 1.187,
+      "step": 7080
+    },
+    {
+      "epoch": 0.47334313201496525,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00012644274980284806,
+      "loss": 1.1714,
+      "step": 7085
+    },
+    {
+      "epoch": 0.47367717797969,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00012633027659764804,
+      "loss": 1.2443,
+      "step": 7090
+    },
+    {
+      "epoch": 0.47401122394441475,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00012621776759142356,
+      "loss": 1.2601,
+      "step": 7095
+    },
+    {
+      "epoch": 0.4743452699091395,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001261052229371521,
+      "loss": 1.2128,
+      "step": 7100
+    },
+    {
+      "epoch": 0.47467931587386425,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00012599264278785952,
+      "loss": 1.252,
+      "step": 7105
+    },
+    {
+      "epoch": 0.475013361838589,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00012588002729661994,
+      "loss": 1.1306,
+      "step": 7110
+    },
+    {
+      "epoch": 0.47534740780331375,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00012576737661655559,
+      "loss": 1.2137,
+      "step": 7115
+    },
+    {
+      "epoch": 0.4756814537680385,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00012565469090083651,
+      "loss": 1.2275,
+      "step": 7120
+    },
+    {
+      "epoch": 0.47601549973276325,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001255419703026804,
+      "loss": 1.1858,
+      "step": 7125
+    },
+    {
+      "epoch": 0.476349545697488,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00012542921497535233,
+      "loss": 1.1734,
+      "step": 7130
+    },
+    {
+      "epoch": 0.47668359166221275,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00012531642507216474,
+      "loss": 1.1865,
+      "step": 7135
+    },
+    {
+      "epoch": 0.47701763762693744,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00012520360074647687,
+      "loss": 1.2398,
+      "step": 7140
+    },
+    {
+      "epoch": 0.4773516835916622,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00012509074215169493,
+      "loss": 1.2366,
+      "step": 7145
+    },
+    {
+      "epoch": 0.47768572955638694,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00012497784944127166,
+      "loss": 1.1808,
+      "step": 7150
+    },
+    {
+      "epoch": 0.4780197755211117,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00012486492276870615,
+      "loss": 1.1907,
+      "step": 7155
+    },
+    {
+      "epoch": 0.47835382148583644,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00012475196228754374,
+      "loss": 1.1947,
+      "step": 7160
+    },
+    {
+      "epoch": 0.4786878674505612,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00012463896815137582,
+      "loss": 1.1483,
+      "step": 7165
+    },
+    {
+      "epoch": 0.47902191341528594,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00012452594051383923,
+      "loss": 1.1198,
+      "step": 7170
+    },
+    {
+      "epoch": 0.4793559593800107,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00012441287952861673,
+      "loss": 1.2169,
+      "step": 7175
+    },
+    {
+      "epoch": 0.47969000534473544,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00012429978534943617,
+      "loss": 1.2445,
+      "step": 7180
+    },
+    {
+      "epoch": 0.4800240513094602,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00012418665813007066,
+      "loss": 1.1686,
+      "step": 7185
+    },
+    {
+      "epoch": 0.48035809727418494,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00012407349802433818,
+      "loss": 1.2286,
+      "step": 7190
+    },
+    {
+      "epoch": 0.4806921432389097,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00012396030518610143,
+      "loss": 1.2835,
+      "step": 7195
+    },
+    {
+      "epoch": 0.48102618920363444,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00012384707976926767,
+      "loss": 1.2347,
+      "step": 7200
+    },
+    {
+      "epoch": 0.4813602351683592,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00012373382192778834,
+      "loss": 1.1946,
+      "step": 7205
+    },
+    {
+      "epoch": 0.48169428113308393,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00012362053181565912,
+      "loss": 1.1201,
+      "step": 7210
+    },
+    {
+      "epoch": 0.4820283270978087,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00012350720958691943,
+      "loss": 1.1542,
+      "step": 7215
+    },
+    {
+      "epoch": 0.4823623730625334,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00012339385539565244,
+      "loss": 1.1681,
+      "step": 7220
+    },
+    {
+      "epoch": 0.48269641902725813,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00012328046939598475,
+      "loss": 1.2992,
+      "step": 7225
+    },
+    {
+      "epoch": 0.4830304649919829,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00012316705174208624,
+      "loss": 1.1038,
+      "step": 7230
+    },
+    {
+      "epoch": 0.4833645109567076,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00012305360258816977,
+      "loss": 1.141,
+      "step": 7235
+    },
+    {
+      "epoch": 0.4836985569214324,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00012294012208849105,
+      "loss": 1.1671,
+      "step": 7240
+    },
+    {
+      "epoch": 0.4840326028861571,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00012282661039734848,
+      "loss": 1.2421,
+      "step": 7245
+    },
+    {
+      "epoch": 0.4843666488508819,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001227130676690828,
+      "loss": 1.2105,
+      "step": 7250
+    },
+    {
+      "epoch": 0.4847006948156066,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00012259949405807696,
+      "loss": 1.1635,
+      "step": 7255
+    },
+    {
+      "epoch": 0.4850347407803314,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00012248588971875587,
+      "loss": 1.2874,
+      "step": 7260
+    },
+    {
+      "epoch": 0.4853687867450561,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00012237225480558637,
+      "loss": 1.1925,
+      "step": 7265
+    },
+    {
+      "epoch": 0.4857028327097809,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00012225858947307664,
+      "loss": 1.208,
+      "step": 7270
+    },
+    {
+      "epoch": 0.4860368786745056,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001221448938757764,
+      "loss": 1.2509,
+      "step": 7275
+    },
+    {
+      "epoch": 0.4863709246392304,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00012203116816827646,
+      "loss": 1.1224,
+      "step": 7280
+    },
+    {
+      "epoch": 0.4867049706039551,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00012191741250520861,
+      "loss": 1.1747,
+      "step": 7285
+    },
+    {
+      "epoch": 0.48703901656867987,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00012180362704124523,
+      "loss": 1.1535,
+      "step": 7290
+    },
+    {
+      "epoch": 0.4873730625334046,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00012168981193109945,
+      "loss": 1.213,
+      "step": 7295
+    },
+    {
+      "epoch": 0.4877071084981293,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00012157596732952448,
+      "loss": 1.1681,
+      "step": 7300
+    },
+    {
+      "epoch": 0.48804115446285407,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00012146209339131376,
+      "loss": 1.1547,
+      "step": 7305
+    },
+    {
+      "epoch": 0.4883752004275788,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00012134819027130062,
+      "loss": 1.1921,
+      "step": 7310
+    },
+    {
+      "epoch": 0.48870924639230356,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00012123425812435803,
+      "loss": 1.2203,
+      "step": 7315
+    },
+    {
+      "epoch": 0.4890432923570283,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00012112029710539842,
+      "loss": 1.1512,
+      "step": 7320
+    },
+    {
+      "epoch": 0.48937733832175306,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00012100630736937348,
+      "loss": 1.192,
+      "step": 7325
+    },
+    {
+      "epoch": 0.4897113842864778,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00012089228907127403,
+      "loss": 1.1679,
+      "step": 7330
+    },
+    {
+      "epoch": 0.49004543025120256,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00012077824236612958,
+      "loss": 1.2825,
+      "step": 7335
+    },
+    {
+      "epoch": 0.4903794762159273,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00012066416740900836,
+      "loss": 1.1705,
+      "step": 7340
+    },
+    {
+      "epoch": 0.49071352218065206,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00012055006435501703,
+      "loss": 1.1813,
+      "step": 7345
+    },
+    {
+      "epoch": 0.4910475681453768,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001204359333593004,
+      "loss": 1.2041,
+      "step": 7350
+    },
+    {
+      "epoch": 0.49138161411010156,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00012032177457704124,
+      "loss": 1.2289,
+      "step": 7355
+    },
+    {
+      "epoch": 0.4917156600748263,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00012020758816346023,
+      "loss": 1.1914,
+      "step": 7360
+    },
+    {
+      "epoch": 0.49204970603955106,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00012009337427381549,
+      "loss": 1.1316,
+      "step": 7365
+    },
+    {
+      "epoch": 0.4923837520042758,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00011997913306340257,
+      "loss": 1.1829,
+      "step": 7370
+    },
+    {
+      "epoch": 0.49271779796900056,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00011986486468755412,
+      "loss": 1.2015,
+      "step": 7375
+    },
+    {
+      "epoch": 0.49305184393372525,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00011975056930163978,
+      "loss": 1.169,
+      "step": 7380
+    },
+    {
+      "epoch": 0.49338588989845,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00011963624706106589,
+      "loss": 1.2223,
+      "step": 7385
+    },
+    {
+      "epoch": 0.49371993586317475,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001195218981212753,
+      "loss": 1.2101,
+      "step": 7390
+    },
+    {
+      "epoch": 0.4940539818278995,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00011940752263774717,
+      "loss": 1.278,
+      "step": 7395
+    },
+    {
+      "epoch": 0.49438802779262425,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001192931207659967,
+      "loss": 1.173,
+      "step": 7400
+    },
+    {
+      "epoch": 0.494722073757349,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00011917869266157513,
+      "loss": 1.1914,
+      "step": 7405
+    },
+    {
+      "epoch": 0.49505611972207375,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00011906423848006913,
+      "loss": 1.2184,
+      "step": 7410
+    },
+    {
+      "epoch": 0.4953901656867985,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.000118949758377101,
+      "loss": 1.3098,
+      "step": 7415
+    },
+    {
+      "epoch": 0.49572421165152325,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00011883525250832828,
+      "loss": 1.2174,
+      "step": 7420
+    },
+    {
+      "epoch": 0.496058257616248,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00011872072102944346,
+      "loss": 1.2241,
+      "step": 7425
+    },
+    {
+      "epoch": 0.49639230358097275,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00011860616409617386,
+      "loss": 1.1798,
+      "step": 7430
+    },
+    {
+      "epoch": 0.4967263495456975,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001184915818642815,
+      "loss": 1.1564,
+      "step": 7435
+    },
+    {
+      "epoch": 0.49706039551042225,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00011837697448956275,
+      "loss": 1.2022,
+      "step": 7440
+    },
+    {
+      "epoch": 0.497394441475147,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001182623421278481,
+      "loss": 1.1389,
+      "step": 7445
+    },
+    {
+      "epoch": 0.49772848743987175,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00011814768493500213,
+      "loss": 1.1532,
+      "step": 7450
+    },
+    {
+      "epoch": 0.4980625334045965,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00011803300306692306,
+      "loss": 1.1738,
+      "step": 7455
+    },
+    {
+      "epoch": 0.4983965793693212,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00011791829667954277,
+      "loss": 1.243,
+      "step": 7460
+    },
+    {
+      "epoch": 0.49873062533404594,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00011780356592882645,
+      "loss": 1.1898,
+      "step": 7465
+    },
+    {
+      "epoch": 0.4990646712987707,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00011768881097077238,
+      "loss": 1.1964,
+      "step": 7470
+    },
+    {
+      "epoch": 0.49939871726349544,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00011757403196141172,
+      "loss": 1.1461,
+      "step": 7475
+    },
+    {
+      "epoch": 0.4997327632282202,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00011745922905680849,
+      "loss": 1.2334,
+      "step": 7480
+    },
+    {
+      "epoch": 0.500066809192945,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00011734440241305902,
+      "loss": 1.2266,
+      "step": 7485
+    },
+    {
+      "epoch": 0.5004008551576697,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00011722955218629204,
+      "loss": 1.2048,
+      "step": 7490
+    },
+    {
+      "epoch": 0.5007349011223945,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00011711467853266826,
+      "loss": 1.2477,
+      "step": 7495
+    },
+    {
+      "epoch": 0.5010689470871192,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011699978160838032,
+      "loss": 1.1918,
+      "step": 7500
+    },
+    {
+      "epoch": 0.5014029930518439,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00011688486156965246,
+      "loss": 1.1969,
+      "step": 7505
+    },
+    {
+      "epoch": 0.5017370390165686,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00011676991857274035,
+      "loss": 1.2102,
+      "step": 7510
+    },
+    {
+      "epoch": 0.5020710849812934,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001166549527739309,
+      "loss": 1.2321,
+      "step": 7515
+    },
+    {
+      "epoch": 0.5024051309460181,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00011653996432954192,
+      "loss": 1.1598,
+      "step": 7520
+    },
+    {
+      "epoch": 0.5027391769107429,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00011642495339592224,
+      "loss": 1.2661,
+      "step": 7525
+    },
+    {
+      "epoch": 0.5030732228754676,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00011630992012945099,
+      "loss": 1.1201,
+      "step": 7530
+    },
+    {
+      "epoch": 0.5034072688401924,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00011619486468653785,
+      "loss": 1.2572,
+      "step": 7535
+    },
+    {
+      "epoch": 0.5037413148049171,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00011607978722362258,
+      "loss": 1.1534,
+      "step": 7540
+    },
+    {
+      "epoch": 0.5040753607696419,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00011596468789717494,
+      "loss": 1.1787,
+      "step": 7545
+    },
+    {
+      "epoch": 0.5044094067343666,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001158495668636944,
+      "loss": 1.1895,
+      "step": 7550
+    },
+    {
+      "epoch": 0.5047434526990914,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00011573442427970982,
+      "loss": 1.2308,
+      "step": 7555
+    },
+    {
+      "epoch": 0.5050774986638161,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00011561926030177957,
+      "loss": 1.183,
+      "step": 7560
+    },
+    {
+      "epoch": 0.5054115446285409,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00011550407508649095,
+      "loss": 1.1378,
+      "step": 7565
+    },
+    {
+      "epoch": 0.5057455905932656,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00011538886879046023,
+      "loss": 1.2704,
+      "step": 7570
+    },
+    {
+      "epoch": 0.5060796365579904,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00011527364157033227,
+      "loss": 1.0969,
+      "step": 7575
+    },
+    {
+      "epoch": 0.5064136825227151,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00011515839358278045,
+      "loss": 1.2111,
+      "step": 7580
+    },
+    {
+      "epoch": 0.5067477284874399,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001150431249845063,
+      "loss": 1.2049,
+      "step": 7585
+    },
+    {
+      "epoch": 0.5070817744521646,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00011492783593223948,
+      "loss": 1.2238,
+      "step": 7590
+    },
+    {
+      "epoch": 0.5074158204168894,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00011481252658273737,
+      "loss": 1.1828,
+      "step": 7595
+    },
+    {
+      "epoch": 0.5077498663816141,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00011469719709278503,
+      "loss": 1.1336,
+      "step": 7600
+    },
+    {
+      "epoch": 0.5080839123463389,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00011458184761919481,
+      "loss": 1.2439,
+      "step": 7605
+    },
+    {
+      "epoch": 0.5084179583110636,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00011446647831880633,
+      "loss": 1.2414,
+      "step": 7610
+    },
+    {
+      "epoch": 0.5087520042757884,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00011435108934848608,
+      "loss": 1.1807,
+      "step": 7615
+    },
+    {
+      "epoch": 0.5090860502405131,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00011423568086512737,
+      "loss": 1.1715,
+      "step": 7620
+    },
+    {
+      "epoch": 0.5094200962052379,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00011412025302564999,
+      "loss": 1.2178,
+      "step": 7625
+    },
+    {
+      "epoch": 0.5097541421699626,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00011400480598700006,
+      "loss": 1.1287,
+      "step": 7630
+    },
+    {
+      "epoch": 0.5100881881346874,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00011388933990614982,
+      "loss": 1.1457,
+      "step": 7635
+    },
+    {
+      "epoch": 0.5104222340994121,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0001137738549400974,
+      "loss": 1.1473,
+      "step": 7640
+    },
+    {
+      "epoch": 0.5107562800641369,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00011365835124586657,
+      "loss": 1.2109,
+      "step": 7645
+    },
+    {
+      "epoch": 0.5110903260288616,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00011354282898050661,
+      "loss": 1.1932,
+      "step": 7650
+    },
+    {
+      "epoch": 0.5114243719935864,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00011342728830109209,
+      "loss": 1.2119,
+      "step": 7655
+    },
+    {
+      "epoch": 0.5117584179583111,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00011331172936472243,
+      "loss": 1.1444,
+      "step": 7660
+    },
+    {
+      "epoch": 0.5120924639230358,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011319615232852217,
+      "loss": 1.2466,
+      "step": 7665
+    },
+    {
+      "epoch": 0.5124265098877605,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00011308055734964018,
+      "loss": 1.1991,
+      "step": 7670
+    },
+    {
+      "epoch": 0.5127605558524853,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00011296494458524986,
+      "loss": 1.149,
+      "step": 7675
+    },
+    {
+      "epoch": 0.51309460181721,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00011284931419254883,
+      "loss": 1.2253,
+      "step": 7680
+    },
+    {
+      "epoch": 0.5134286477819348,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00011273366632875859,
+      "loss": 1.2265,
+      "step": 7685
+    },
+    {
+      "epoch": 0.5137626937466595,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00011261800115112441,
+      "loss": 1.2184,
+      "step": 7690
+    },
+    {
+      "epoch": 0.5140967397113843,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00011250231881691517,
+      "loss": 1.1543,
+      "step": 7695
+    },
+    {
+      "epoch": 0.514430785676109,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00011238661948342302,
+      "loss": 1.1452,
+      "step": 7700
+    },
+    {
+      "epoch": 0.5147648316408338,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00011227090330796317,
+      "loss": 1.1757,
+      "step": 7705
+    },
+    {
+      "epoch": 0.5150988776055585,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00011215517044787387,
+      "loss": 1.2306,
+      "step": 7710
+    },
+    {
+      "epoch": 0.5154329235702833,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00011203942106051595,
+      "loss": 1.1709,
+      "step": 7715
+    },
+    {
+      "epoch": 0.515766969535008,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00011192365530327275,
+      "loss": 1.2167,
+      "step": 7720
+    },
+    {
+      "epoch": 0.5161010154997328,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00011180787333354983,
+      "loss": 1.1912,
+      "step": 7725
+    },
+    {
+      "epoch": 0.5164350614644575,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00011169207530877486,
+      "loss": 1.2212,
+      "step": 7730
+    },
+    {
+      "epoch": 0.5167691074291823,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00011157626138639727,
+      "loss": 1.2778,
+      "step": 7735
+    },
+    {
+      "epoch": 0.517103153393907,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00011146043172388819,
+      "loss": 1.2223,
+      "step": 7740
+    },
+    {
+      "epoch": 0.5174371993586317,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00011134458647874007,
+      "loss": 1.2156,
+      "step": 7745
+    },
+    {
+      "epoch": 0.5177712453233565,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00011122872580846652,
+      "loss": 1.1062,
+      "step": 7750
+    },
+    {
+      "epoch": 0.5181052912880812,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00011111284987060228,
+      "loss": 1.1876,
+      "step": 7755
+    },
+    {
+      "epoch": 0.518439337252806,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00011099695882270272,
+      "loss": 1.1698,
+      "step": 7760
+    },
+    {
+      "epoch": 0.5187733832175307,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011088105282234376,
+      "loss": 1.1686,
+      "step": 7765
+    },
+    {
+      "epoch": 0.5191074291822555,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00011076513202712167,
+      "loss": 1.1526,
+      "step": 7770
+    },
+    {
+      "epoch": 0.5194414751469802,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011064919659465289,
+      "loss": 1.1648,
+      "step": 7775
+    },
+    {
+      "epoch": 0.519775521111705,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00011053324668257368,
+      "loss": 1.1674,
+      "step": 7780
+    },
+    {
+      "epoch": 0.5201095670764297,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00011041728244854004,
+      "loss": 1.1817,
+      "step": 7785
+    },
+    {
+      "epoch": 0.5204436130411545,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00011030130405022738,
+      "loss": 1.1222,
+      "step": 7790
+    },
+    {
+      "epoch": 0.5207776590058792,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00011018531164533048,
+      "loss": 1.1805,
+      "step": 7795
+    },
+    {
+      "epoch": 0.521111704970604,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011006930539156308,
+      "loss": 1.1889,
+      "step": 7800
+    },
+    {
+      "epoch": 0.5214457509353287,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00010995328544665775,
+      "loss": 1.1897,
+      "step": 7805
+    },
+    {
+      "epoch": 0.5217797969000535,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00010983725196836574,
+      "loss": 1.1815,
+      "step": 7810
+    },
+    {
+      "epoch": 0.5221138428647782,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00010972120511445656,
+      "loss": 1.1831,
+      "step": 7815
+    },
+    {
+      "epoch": 0.522447888829503,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00010960514504271813,
+      "loss": 1.1478,
+      "step": 7820
+    },
+    {
+      "epoch": 0.5227819347942276,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00010948907191095612,
+      "loss": 1.2321,
+      "step": 7825
+    },
+    {
+      "epoch": 0.5231159807589524,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00010937298587699409,
+      "loss": 1.1213,
+      "step": 7830
+    },
+    {
+      "epoch": 0.5234500267236771,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00010925688709867312,
+      "loss": 1.2076,
+      "step": 7835
+    },
+    {
+      "epoch": 0.5237840726884019,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001091407757338516,
+      "loss": 1.2197,
+      "step": 7840
+    },
+    {
+      "epoch": 0.5241181186531266,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00010902465194040501,
+      "loss": 1.2402,
+      "step": 7845
+    },
+    {
+      "epoch": 0.5244521646178514,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010890851587622577,
+      "loss": 1.2424,
+      "step": 7850
+    },
+    {
+      "epoch": 0.5247862105825761,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010879236769922301,
+      "loss": 1.2017,
+      "step": 7855
+    },
+    {
+      "epoch": 0.5251202565473009,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00010867620756732219,
+      "loss": 1.2269,
+      "step": 7860
+    },
+    {
+      "epoch": 0.5254543025120256,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010856003563846526,
+      "loss": 1.1909,
+      "step": 7865
+    },
+    {
+      "epoch": 0.5257883484767504,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00010844385207061001,
+      "loss": 1.1847,
+      "step": 7870
+    },
+    {
+      "epoch": 0.5261223944414751,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00010832765702173011,
+      "loss": 1.1651,
+      "step": 7875
+    },
+    {
+      "epoch": 0.5264564404061999,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00010821145064981487,
+      "loss": 1.1334,
+      "step": 7880
+    },
+    {
+      "epoch": 0.5267904863709246,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00010809523311286897,
+      "loss": 1.1816,
+      "step": 7885
+    },
+    {
+      "epoch": 0.5271245323356494,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00010797900456891228,
+      "loss": 1.2023,
+      "step": 7890
+    },
+    {
+      "epoch": 0.5274585783003741,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00010786276517597968,
+      "loss": 1.1492,
+      "step": 7895
+    },
+    {
+      "epoch": 0.5277926242650989,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001077465150921207,
+      "loss": 1.2383,
+      "step": 7900
+    },
+    {
+      "epoch": 0.5281266702298236,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00010763025447539948,
+      "loss": 1.1999,
+      "step": 7905
+    },
+    {
+      "epoch": 0.5284607161945484,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00010751398348389446,
+      "loss": 1.1763,
+      "step": 7910
+    },
+    {
+      "epoch": 0.5287947621592731,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00010739770227569821,
+      "loss": 1.2671,
+      "step": 7915
+    },
+    {
+      "epoch": 0.5291288081239979,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00010728141100891716,
+      "loss": 1.2537,
+      "step": 7920
+    },
+    {
+      "epoch": 0.5294628540887226,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010716510984167142,
+      "loss": 1.2583,
+      "step": 7925
+    },
+    {
+      "epoch": 0.5297969000534474,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00010704879893209463,
+      "loss": 1.177,
+      "step": 7930
+    },
+    {
+      "epoch": 0.5301309460181721,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00010693247843833352,
+      "loss": 1.2566,
+      "step": 7935
+    },
+    {
+      "epoch": 0.5304649919828969,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010681614851854802,
+      "loss": 1.1253,
+      "step": 7940
+    },
+    {
+      "epoch": 0.5307990379476216,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00010669980933091079,
+      "loss": 1.2334,
+      "step": 7945
+    },
+    {
+      "epoch": 0.5311330839123464,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0001065834610336071,
+      "loss": 1.2108,
+      "step": 7950
+    },
+    {
+      "epoch": 0.5314671298770711,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001064671037848346,
+      "loss": 1.2085,
+      "step": 7955
+    },
+    {
+      "epoch": 0.5318011758417959,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00010635073774280315,
+      "loss": 1.1995,
+      "step": 7960
+    },
+    {
+      "epoch": 0.5321352218065206,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00010623436306573455,
+      "loss": 1.162,
+      "step": 7965
+    },
+    {
+      "epoch": 0.5324692677712454,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00010611797991186229,
+      "loss": 1.2132,
+      "step": 7970
+    },
+    {
+      "epoch": 0.5328033137359701,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010600158843943149,
+      "loss": 1.2682,
+      "step": 7975
+    },
+    {
+      "epoch": 0.5331373597006949,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00010588518880669842,
+      "loss": 1.1729,
+      "step": 7980
+    },
+    {
+      "epoch": 0.5334714056654195,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00010576878117193066,
+      "loss": 1.1858,
+      "step": 7985
+    },
+    {
+      "epoch": 0.5338054516301443,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00010565236569340646,
+      "loss": 1.2354,
+      "step": 7990
+    },
+    {
+      "epoch": 0.534139497594869,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00010553594252941488,
+      "loss": 1.1875,
+      "step": 7995
+    },
+    {
+      "epoch": 0.5344735435595938,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00010541951183825536,
+      "loss": 1.1688,
+      "step": 8000
+    },
+    {
+      "epoch": 0.5348075895243185,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00010530307377823762,
+      "loss": 1.1839,
+      "step": 8005
+    },
+    {
+      "epoch": 0.5351416354890433,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00010518662850768133,
+      "loss": 1.1939,
+      "step": 8010
+    },
+    {
+      "epoch": 0.535475681453768,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00010507017618491603,
+      "loss": 1.2179,
+      "step": 8015
+    },
+    {
+      "epoch": 0.5358097274184928,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00010495371696828083,
+      "loss": 1.2316,
+      "step": 8020
+    },
+    {
+      "epoch": 0.5361437733832175,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010483725101612419,
+      "loss": 1.2384,
+      "step": 8025
+    },
+    {
+      "epoch": 0.5364778193479423,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00010472077848680378,
+      "loss": 1.1438,
+      "step": 8030
+    },
+    {
+      "epoch": 0.536811865312667,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010460429953868614,
+      "loss": 1.2008,
+      "step": 8035
+    },
+    {
+      "epoch": 0.5371459112773918,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010448781433014663,
+      "loss": 1.1596,
+      "step": 8040
+    },
+    {
+      "epoch": 0.5374799572421165,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00010437132301956897,
+      "loss": 1.156,
+      "step": 8045
+    },
+    {
+      "epoch": 0.5378140032068413,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00010425482576534545,
+      "loss": 1.1974,
+      "step": 8050
+    },
+    {
+      "epoch": 0.538148049171566,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00010413832272587609,
+      "loss": 1.091,
+      "step": 8055
+    },
+    {
+      "epoch": 0.5384820951362908,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00010402181405956906,
+      "loss": 1.262,
+      "step": 8060
+    },
+    {
+      "epoch": 0.5388161411010155,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00010390529992484004,
+      "loss": 1.169,
+      "step": 8065
+    },
+    {
+      "epoch": 0.5391501870657402,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00010378878048011218,
+      "loss": 1.1552,
+      "step": 8070
+    },
+    {
+      "epoch": 0.539484233030465,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00010367225588381584,
+      "loss": 1.195,
+      "step": 8075
+    },
+    {
+      "epoch": 0.5398182789951897,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00010355572629438846,
+      "loss": 1.239,
+      "step": 8080
+    },
+    {
+      "epoch": 0.5401523249599145,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010343919187027413,
+      "loss": 1.2839,
+      "step": 8085
+    },
+    {
+      "epoch": 0.5404863709246392,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00010332265276992362,
+      "loss": 1.203,
+      "step": 8090
+    },
+    {
+      "epoch": 0.540820416889364,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00010320610915179402,
+      "loss": 1.1622,
+      "step": 8095
+    },
+    {
+      "epoch": 0.5411544628540887,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00010308956117434858,
+      "loss": 1.1227,
+      "step": 8100
+    },
+    {
+      "epoch": 0.5414885088188135,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00010297300899605644,
+      "loss": 1.2179,
+      "step": 8105
+    },
+    {
+      "epoch": 0.5418225547835382,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00010285645277539252,
+      "loss": 1.1927,
+      "step": 8110
+    },
+    {
+      "epoch": 0.542156600748263,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00010273989267083717,
+      "loss": 1.1551,
+      "step": 8115
+    },
+    {
+      "epoch": 0.5424906467129877,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.000102623328840876,
+      "loss": 1.2515,
+      "step": 8120
+    },
+    {
+      "epoch": 0.5428246926777125,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00010250676144399984,
+      "loss": 1.1713,
+      "step": 8125
+    },
+    {
+      "epoch": 0.5431587386424372,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00010239019063870416,
+      "loss": 1.1947,
+      "step": 8130
+    },
+    {
+      "epoch": 0.543492784607162,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00010227361658348922,
+      "loss": 1.2157,
+      "step": 8135
+    },
+    {
+      "epoch": 0.5438268305718867,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00010215703943685964,
+      "loss": 1.1747,
+      "step": 8140
+    },
+    {
+      "epoch": 0.5441608765366115,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001020404593573242,
+      "loss": 1.2436,
+      "step": 8145
+    },
+    {
+      "epoch": 0.5444949225013361,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00010192387650339579,
+      "loss": 1.1607,
+      "step": 8150
+    },
+    {
+      "epoch": 0.5448289684660609,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00010180729103359094,
+      "loss": 1.1594,
+      "step": 8155
+    },
+    {
+      "epoch": 0.5451630144307856,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00010169070310642983,
+      "loss": 1.1701,
+      "step": 8160
+    },
+    {
+      "epoch": 0.5454970603955104,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001015741128804359,
+      "loss": 1.2794,
+      "step": 8165
+    },
+    {
+      "epoch": 0.5458311063602351,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00010145752051413584,
+      "loss": 1.2377,
+      "step": 8170
+    },
+    {
+      "epoch": 0.5461651523249599,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00010134092616605908,
+      "loss": 1.1754,
+      "step": 8175
+    },
+    {
+      "epoch": 0.5464991982896846,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001012243299947379,
+      "loss": 1.1771,
+      "step": 8180
+    },
+    {
+      "epoch": 0.5468332442544094,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00010110773215870695,
+      "loss": 1.2177,
+      "step": 8185
+    },
+    {
+      "epoch": 0.5471672902191341,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010099113281650325,
+      "loss": 1.2478,
+      "step": 8190
+    },
+    {
+      "epoch": 0.5475013361838589,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00010087453212666574,
+      "loss": 1.1897,
+      "step": 8195
+    },
+    {
+      "epoch": 0.5478353821485836,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001007579302477353,
+      "loss": 1.179,
+      "step": 8200
+    },
+    {
+      "epoch": 0.5481694281133084,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00010064132733825438,
+      "loss": 1.1982,
+      "step": 8205
+    },
+    {
+      "epoch": 0.5485034740780331,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00010052472355676683,
+      "loss": 1.1418,
+      "step": 8210
+    },
+    {
+      "epoch": 0.5488375200427579,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00010040811906181769,
+      "loss": 1.1776,
+      "step": 8215
+    },
+    {
+      "epoch": 0.5491715660074826,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00010029151401195298,
+      "loss": 1.133,
+      "step": 8220
+    },
+    {
+      "epoch": 0.5495056119722074,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00010017490856571945,
+      "loss": 1.2514,
+      "step": 8225
+    },
+    {
+      "epoch": 0.5498396579369321,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00010005830288166445,
+      "loss": 1.1875,
+      "step": 8230
+    },
+    {
+      "epoch": 0.5501737039016569,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.994169711833555e-05,
+      "loss": 1.2005,
+      "step": 8235
+    },
+    {
+      "epoch": 0.5505077498663816,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.982509143428054e-05,
+      "loss": 1.2274,
+      "step": 8240
+    },
+    {
+      "epoch": 0.5508417958311064,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 9.970848598804705e-05,
+      "loss": 1.2221,
+      "step": 8245
+    },
+    {
+      "epoch": 0.5511758417958311,
+      "grad_norm": 0.1796875,
+      "learning_rate": 9.959188093818234e-05,
+      "loss": 1.2734,
+      "step": 8250
+    },
+    {
+      "epoch": 0.5515098877605559,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.947527644323319e-05,
+      "loss": 1.1843,
+      "step": 8255
+    },
+    {
+      "epoch": 0.5518439337252806,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 9.935867266174566e-05,
+      "loss": 1.2689,
+      "step": 8260
+    },
+    {
+      "epoch": 0.5521779796900054,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.924206975226471e-05,
+      "loss": 1.3259,
+      "step": 8265
+    },
+    {
+      "epoch": 0.5525120256547301,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.912546787333427e-05,
+      "loss": 1.1586,
+      "step": 8270
+    },
+    {
+      "epoch": 0.5528460716194549,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.900886718349676e-05,
+      "loss": 1.1334,
+      "step": 8275
+    },
+    {
+      "epoch": 0.5531801175841796,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.889226784129306e-05,
+      "loss": 1.2147,
+      "step": 8280
+    },
+    {
+      "epoch": 0.5535141635489044,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.877567000526213e-05,
+      "loss": 1.1871,
+      "step": 8285
+    },
+    {
+      "epoch": 0.5538482095136291,
+      "grad_norm": 0.1640625,
+      "learning_rate": 9.865907383394096e-05,
+      "loss": 1.2581,
+      "step": 8290
+    },
+    {
+      "epoch": 0.5541822554783539,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.854247948586417e-05,
+      "loss": 1.1312,
+      "step": 8295
+    },
+    {
+      "epoch": 0.5545163014430786,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 9.84258871195641e-05,
+      "loss": 1.1728,
+      "step": 8300
+    },
+    {
+      "epoch": 0.5548503474078034,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.830929689357019e-05,
+      "loss": 1.2742,
+      "step": 8305
+    },
+    {
+      "epoch": 0.555184393372528,
+      "grad_norm": 0.166015625,
+      "learning_rate": 9.819270896640908e-05,
+      "loss": 1.2569,
+      "step": 8310
+    },
+    {
+      "epoch": 0.5555184393372528,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.807612349660423e-05,
+      "loss": 1.2754,
+      "step": 8315
+    },
+    {
+      "epoch": 0.5558524853019775,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.795954064267581e-05,
+      "loss": 1.1742,
+      "step": 8320
+    },
+    {
+      "epoch": 0.5561865312667023,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.784296056314037e-05,
+      "loss": 1.1708,
+      "step": 8325
+    },
+    {
+      "epoch": 0.556520577231427,
+      "grad_norm": 0.44140625,
+      "learning_rate": 9.772638341651079e-05,
+      "loss": 1.2346,
+      "step": 8330
+    },
+    {
+      "epoch": 0.5568546231961518,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 9.760980936129585e-05,
+      "loss": 1.1983,
+      "step": 8335
+    },
+    {
+      "epoch": 0.5571886691608765,
+      "grad_norm": 0.166015625,
+      "learning_rate": 9.749323855600017e-05,
+      "loss": 1.2003,
+      "step": 8340
+    },
+    {
+      "epoch": 0.5575227151256013,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 9.737667115912402e-05,
+      "loss": 1.2915,
+      "step": 8345
+    },
+    {
+      "epoch": 0.557856761090326,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.726010732916288e-05,
+      "loss": 1.2653,
+      "step": 8350
+    },
+    {
+      "epoch": 0.5581908070550508,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 9.714354722460753e-05,
+      "loss": 1.2065,
+      "step": 8355
+    },
+    {
+      "epoch": 0.5585248530197755,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 9.702699100394355e-05,
+      "loss": 1.1583,
+      "step": 8360
+    },
+    {
+      "epoch": 0.5588588989845003,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 9.691043882565145e-05,
+      "loss": 1.189,
+      "step": 8365
+    },
+    {
+      "epoch": 0.559192944949225,
+      "grad_norm": 0.162109375,
+      "learning_rate": 9.6793890848206e-05,
+      "loss": 1.2118,
+      "step": 8370
+    },
+    {
+      "epoch": 0.5595269909139498,
+      "grad_norm": 0.1640625,
+      "learning_rate": 9.66773472300764e-05,
+      "loss": 1.1771,
+      "step": 8375
+    },
+    {
+      "epoch": 0.5598610368786745,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.656080812972591e-05,
+      "loss": 1.287,
+      "step": 8380
+    },
+    {
+      "epoch": 0.5601950828433993,
+      "grad_norm": 0.162109375,
+      "learning_rate": 9.644427370561157e-05,
+      "loss": 1.1679,
+      "step": 8385
+    },
+    {
+      "epoch": 0.560529128808124,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 9.632774411618414e-05,
+      "loss": 1.1996,
+      "step": 8390
+    },
+    {
+      "epoch": 0.5608631747728487,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.621121951988783e-05,
+      "loss": 1.2401,
+      "step": 8395
+    },
+    {
+      "epoch": 0.5611972207375735,
+      "grad_norm": 0.1640625,
+      "learning_rate": 9.609470007516e-05,
+      "loss": 1.1879,
+      "step": 8400
+    },
+    {
+      "epoch": 0.5615312667022982,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 9.597818594043096e-05,
+      "loss": 1.2456,
+      "step": 8405
+    },
+    {
+      "epoch": 0.561865312667023,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.586167727412395e-05,
+      "loss": 1.1403,
+      "step": 8410
+    },
+    {
+      "epoch": 0.5621993586317477,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 9.574517423465462e-05,
+      "loss": 1.2011,
+      "step": 8415
+    },
+    {
+      "epoch": 0.5625334045964725,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 9.562867698043101e-05,
+      "loss": 1.1682,
+      "step": 8420
+    },
+    {
+      "epoch": 0.5628674505611972,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.55121856698534e-05,
+      "loss": 1.1735,
+      "step": 8425
+    },
+    {
+      "epoch": 0.563201496525922,
+      "grad_norm": 0.162109375,
+      "learning_rate": 9.539570046131389e-05,
+      "loss": 1.1531,
+      "step": 8430
+    },
+    {
+      "epoch": 0.5635355424906467,
+      "grad_norm": 0.1640625,
+      "learning_rate": 9.527922151319626e-05,
+      "loss": 1.1076,
+      "step": 8435
+    },
+    {
+      "epoch": 0.5638695884553715,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.516274898387582e-05,
+      "loss": 1.1787,
+      "step": 8440
+    },
+    {
+      "epoch": 0.5642036344200962,
+      "grad_norm": 0.1640625,
+      "learning_rate": 9.504628303171922e-05,
+      "loss": 1.2271,
+      "step": 8445
+    },
+    {
+      "epoch": 0.564537680384821,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.492982381508398e-05,
+      "loss": 1.1947,
+      "step": 8450
+    },
+    {
+      "epoch": 0.5648717263495457,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.481337149231868e-05,
+      "loss": 1.1849,
+      "step": 8455
+    },
+    {
+      "epoch": 0.5652057723142705,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 9.469692622176239e-05,
+      "loss": 1.2134,
+      "step": 8460
+    },
+    {
+      "epoch": 0.5655398182789952,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 9.458048816174465e-05,
+      "loss": 1.2114,
+      "step": 8465
+    },
+    {
+      "epoch": 0.5658738642437199,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 9.446405747058513e-05,
+      "loss": 1.2596,
+      "step": 8470
+    },
+    {
+      "epoch": 0.5662079102084446,
+      "grad_norm": 0.169921875,
+      "learning_rate": 9.434763430659357e-05,
+      "loss": 1.1842,
+      "step": 8475
+    },
+    {
+      "epoch": 0.5665419561731694,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 9.423121882806934e-05,
+      "loss": 1.1554,
+      "step": 8480
+    },
+    {
+      "epoch": 0.5668760021378941,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 9.411481119330156e-05,
+      "loss": 1.1946,
+      "step": 8485
+    },
+    {
+      "epoch": 0.5672100481026189,
+      "grad_norm": 0.158203125,
+      "learning_rate": 9.399841156056853e-05,
+      "loss": 1.134,
+      "step": 8490
+    },
+    {
+      "epoch": 0.5675440940673436,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.388202008813772e-05,
+      "loss": 1.1399,
+      "step": 8495
+    },
+    {
+      "epoch": 0.5678781400320684,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.376563693426548e-05,
+      "loss": 1.1826,
+      "step": 8500
+    },
+    {
+      "epoch": 0.5682121859967931,
+      "grad_norm": 0.1640625,
+      "learning_rate": 9.364926225719686e-05,
+      "loss": 1.1893,
+      "step": 8505
+    },
+    {
+      "epoch": 0.5685462319615179,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 9.353289621516539e-05,
+      "loss": 1.158,
+      "step": 8510
+    },
+    {
+      "epoch": 0.5688802779262426,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.341653896639293e-05,
+      "loss": 1.1754,
+      "step": 8515
+    },
+    {
+      "epoch": 0.5692143238909674,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.330019066908923e-05,
+      "loss": 1.2279,
+      "step": 8520
+    },
+    {
+      "epoch": 0.5695483698556921,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.318385148145199e-05,
+      "loss": 1.2369,
+      "step": 8525
+    },
+    {
+      "epoch": 0.5698824158204169,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 9.306752156166652e-05,
+      "loss": 1.1594,
+      "step": 8530
+    },
+    {
+      "epoch": 0.5702164617851416,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 9.295120106790542e-05,
+      "loss": 1.2326,
+      "step": 8535
+    },
+    {
+      "epoch": 0.5705505077498664,
+      "grad_norm": 0.166015625,
+      "learning_rate": 9.283489015832857e-05,
+      "loss": 1.2411,
+      "step": 8540
+    },
+    {
+      "epoch": 0.5708845537145911,
+      "grad_norm": 0.169921875,
+      "learning_rate": 9.271858899108285e-05,
+      "loss": 1.1679,
+      "step": 8545
+    },
+    {
+      "epoch": 0.5712185996793159,
+      "grad_norm": 0.1640625,
+      "learning_rate": 9.260229772430181e-05,
+      "loss": 1.194,
+      "step": 8550
+    },
+    {
+      "epoch": 0.5715526456440406,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 9.248601651610556e-05,
+      "loss": 1.2654,
+      "step": 8555
+    },
+    {
+      "epoch": 0.5718866916087654,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.236974552460055e-05,
+      "loss": 1.2777,
+      "step": 8560
+    },
+    {
+      "epoch": 0.5722207375734901,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 9.225348490787935e-05,
+      "loss": 1.2713,
+      "step": 8565
+    },
+    {
+      "epoch": 0.5725547835382149,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 9.213723482402035e-05,
+      "loss": 1.1932,
+      "step": 8570
+    },
+    {
+      "epoch": 0.5728888295029396,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 9.202099543108771e-05,
+      "loss": 1.2309,
+      "step": 8575
+    },
+    {
+      "epoch": 0.5732228754676644,
+      "grad_norm": 0.1796875,
+      "learning_rate": 9.190476688713103e-05,
+      "loss": 1.1724,
+      "step": 8580
+    },
+    {
+      "epoch": 0.5735569214323891,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 9.178854935018516e-05,
+      "loss": 1.1986,
+      "step": 8585
+    },
+    {
+      "epoch": 0.5738909673971139,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.167234297826992e-05,
+      "loss": 1.1978,
+      "step": 8590
+    },
+    {
+      "epoch": 0.5742250133618386,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 9.155614792939004e-05,
+      "loss": 1.1005,
+      "step": 8595
+    },
+    {
+      "epoch": 0.5745590593265634,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.143996436153476e-05,
+      "loss": 1.2731,
+      "step": 8600
+    },
+    {
+      "epoch": 0.5748931052912881,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 9.13237924326778e-05,
+      "loss": 1.2199,
+      "step": 8605
+    },
+    {
+      "epoch": 0.5752271512560129,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.120763230077703e-05,
+      "loss": 1.1892,
+      "step": 8610
+    },
+    {
+      "epoch": 0.5755611972207376,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.109148412377426e-05,
+      "loss": 1.261,
+      "step": 8615
+    },
+    {
+      "epoch": 0.5758952431854624,
+      "grad_norm": 0.16015625,
+      "learning_rate": 9.097534805959502e-05,
+      "loss": 1.2014,
+      "step": 8620
+    },
+    {
+      "epoch": 0.5762292891501871,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 9.085922426614844e-05,
+      "loss": 1.2072,
+      "step": 8625
+    },
+    {
+      "epoch": 0.5765633351149118,
+      "grad_norm": 0.166015625,
+      "learning_rate": 9.074311290132693e-05,
+      "loss": 1.154,
+      "step": 8630
+    },
+    {
+      "epoch": 0.5768973810796365,
+      "grad_norm": 0.1796875,
+      "learning_rate": 9.062701412300592e-05,
+      "loss": 1.2357,
+      "step": 8635
+    },
+    {
+      "epoch": 0.5772314270443613,
+      "grad_norm": 0.201171875,
+      "learning_rate": 9.051092808904389e-05,
+      "loss": 1.1894,
+      "step": 8640
+    },
+    {
+      "epoch": 0.577565473009086,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 9.03948549572819e-05,
+      "loss": 1.2215,
+      "step": 8645
+    },
+    {
+      "epoch": 0.5778995189738108,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.027879488554346e-05,
+      "loss": 1.2391,
+      "step": 8650
+    },
+    {
+      "epoch": 0.5782335649385355,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 9.016274803163431e-05,
+      "loss": 1.2081,
+      "step": 8655
+    },
+    {
+      "epoch": 0.5785676109032603,
+      "grad_norm": 0.16015625,
+      "learning_rate": 9.004671455334228e-05,
+      "loss": 1.1283,
+      "step": 8660
+    },
+    {
+      "epoch": 0.578901656867985,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.993069460843693e-05,
+      "loss": 1.2581,
+      "step": 8665
+    },
+    {
+      "epoch": 0.5792357028327098,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.981468835466953e-05,
+      "loss": 1.1975,
+      "step": 8670
+    },
+    {
+      "epoch": 0.5795697487974345,
+      "grad_norm": 0.181640625,
+      "learning_rate": 8.969869594977263e-05,
+      "loss": 1.2193,
+      "step": 8675
+    },
+    {
+      "epoch": 0.5799037947621593,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 8.958271755145999e-05,
+      "loss": 1.1285,
+      "step": 8680
+    },
+    {
+      "epoch": 0.580237840726884,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.946675331742636e-05,
+      "loss": 1.1692,
+      "step": 8685
+    },
+    {
+      "epoch": 0.5805718866916088,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 8.935080340534715e-05,
+      "loss": 1.14,
+      "step": 8690
+    },
+    {
+      "epoch": 0.5809059326563335,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.923486797287834e-05,
+      "loss": 1.2068,
+      "step": 8695
+    },
+    {
+      "epoch": 0.5812399786210583,
+      "grad_norm": 0.16796875,
+      "learning_rate": 8.911894717765625e-05,
+      "loss": 1.2347,
+      "step": 8700
+    },
+    {
+      "epoch": 0.581574024585783,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 8.90030411772973e-05,
+      "loss": 1.1138,
+      "step": 8705
+    },
+    {
+      "epoch": 0.5819080705505077,
+      "grad_norm": 0.16015625,
+      "learning_rate": 8.888715012939773e-05,
+      "loss": 1.2134,
+      "step": 8710
+    },
+    {
+      "epoch": 0.5822421165152325,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.87712741915335e-05,
+      "loss": 1.1751,
+      "step": 8715
+    },
+    {
+      "epoch": 0.5825761624799572,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.865541352125998e-05,
+      "loss": 1.2041,
+      "step": 8720
+    },
+    {
+      "epoch": 0.582910208444682,
+      "grad_norm": 0.388671875,
+      "learning_rate": 8.853956827611182e-05,
+      "loss": 1.2135,
+      "step": 8725
+    },
+    {
+      "epoch": 0.5832442544094067,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 8.842373861360271e-05,
+      "loss": 1.2284,
+      "step": 8730
+    },
+    {
+      "epoch": 0.5835783003741315,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 8.830792469122517e-05,
+      "loss": 1.2092,
+      "step": 8735
+    },
+    {
+      "epoch": 0.5839123463388562,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 8.819212666645018e-05,
+      "loss": 1.2435,
+      "step": 8740
+    },
+    {
+      "epoch": 0.584246392303581,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 8.807634469672727e-05,
+      "loss": 1.1988,
+      "step": 8745
+    },
+    {
+      "epoch": 0.5845804382683057,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 8.796057893948409e-05,
+      "loss": 1.264,
+      "step": 8750
+    },
+    {
+      "epoch": 0.5849144842330305,
+      "grad_norm": 0.171875,
+      "learning_rate": 8.784482955212614e-05,
+      "loss": 1.1828,
+      "step": 8755
+    },
+    {
+      "epoch": 0.5852485301977552,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 8.772909669203684e-05,
+      "loss": 1.1714,
+      "step": 8760
+    },
+    {
+      "epoch": 0.58558257616248,
+      "grad_norm": 0.1640625,
+      "learning_rate": 8.7613380516577e-05,
+      "loss": 1.2169,
+      "step": 8765
+    },
+    {
+      "epoch": 0.5859166221272047,
+      "grad_norm": 0.158203125,
+      "learning_rate": 8.749768118308485e-05,
+      "loss": 1.1714,
+      "step": 8770
+    },
+    {
+      "epoch": 0.5862506680919295,
+      "grad_norm": 0.16796875,
+      "learning_rate": 8.73819988488756e-05,
+      "loss": 1.1609,
+      "step": 8775
+    },
+    {
+      "epoch": 0.5865847140566542,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 8.726633367124146e-05,
+      "loss": 1.1455,
+      "step": 8780
+    },
+    {
+      "epoch": 0.586918760021379,
+      "grad_norm": 0.1640625,
+      "learning_rate": 8.715068580745116e-05,
+      "loss": 1.2395,
+      "step": 8785
+    },
+    {
+      "epoch": 0.5872528059861037,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.703505541475015e-05,
+      "loss": 1.2282,
+      "step": 8790
+    },
+    {
+      "epoch": 0.5875868519508284,
+      "grad_norm": 0.171875,
+      "learning_rate": 8.691944265035985e-05,
+      "loss": 1.2212,
+      "step": 8795
+    },
+    {
+      "epoch": 0.5879208979155531,
+      "grad_norm": 0.16796875,
+      "learning_rate": 8.680384767147785e-05,
+      "loss": 1.2372,
+      "step": 8800
+    },
+    {
+      "epoch": 0.5882549438802779,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.668827063527758e-05,
+      "loss": 1.1653,
+      "step": 8805
+    },
+    {
+      "epoch": 0.5885889898450026,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 8.657271169890797e-05,
+      "loss": 1.1536,
+      "step": 8810
+    },
+    {
+      "epoch": 0.5889230358097274,
+      "grad_norm": 0.162109375,
+      "learning_rate": 8.645717101949338e-05,
+      "loss": 1.1778,
+      "step": 8815
+    },
+    {
+      "epoch": 0.5892570817744521,
+      "grad_norm": 0.162109375,
+      "learning_rate": 8.634164875413343e-05,
+      "loss": 1.2575,
+      "step": 8820
+    },
+    {
+      "epoch": 0.5895911277391769,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 8.622614505990263e-05,
+      "loss": 1.1312,
+      "step": 8825
+    },
+    {
+      "epoch": 0.5899251737039016,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.61106600938502e-05,
+      "loss": 1.1868,
+      "step": 8830
+    },
+    {
+      "epoch": 0.5902592196686264,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.599519401299997e-05,
+      "loss": 1.2483,
+      "step": 8835
+    },
+    {
+      "epoch": 0.5905932656333511,
+      "grad_norm": 0.20703125,
+      "learning_rate": 8.587974697435005e-05,
+      "loss": 1.2095,
+      "step": 8840
+    },
+    {
+      "epoch": 0.5909273115980759,
+      "grad_norm": 0.1640625,
+      "learning_rate": 8.576431913487265e-05,
+      "loss": 1.1909,
+      "step": 8845
+    },
+    {
+      "epoch": 0.5912613575628006,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.564891065151392e-05,
+      "loss": 1.2462,
+      "step": 8850
+    },
+    {
+      "epoch": 0.5915954035275254,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 8.553352168119368e-05,
+      "loss": 1.2182,
+      "step": 8855
+    },
+    {
+      "epoch": 0.5919294494922501,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 8.54181523808052e-05,
+      "loss": 1.1494,
+      "step": 8860
+    },
+    {
+      "epoch": 0.5922634954569749,
+      "grad_norm": 0.16796875,
+      "learning_rate": 8.530280290721499e-05,
+      "loss": 1.1958,
+      "step": 8865
+    },
+    {
+      "epoch": 0.5925975414216996,
+      "grad_norm": 0.171875,
+      "learning_rate": 8.518747341726265e-05,
+      "loss": 1.2156,
+      "step": 8870
+    },
+    {
+      "epoch": 0.5929315873864244,
+      "grad_norm": 0.171875,
+      "learning_rate": 8.507216406776056e-05,
+      "loss": 1.1586,
+      "step": 8875
+    },
+    {
+      "epoch": 0.5932656333511491,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.495687501549371e-05,
+      "loss": 1.1507,
+      "step": 8880
+    },
+    {
+      "epoch": 0.5935996793158739,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.484160641721958e-05,
+      "loss": 1.1466,
+      "step": 8885
+    },
+    {
+      "epoch": 0.5939337252805986,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 8.472635842966776e-05,
+      "loss": 1.178,
+      "step": 8890
+    },
+    {
+      "epoch": 0.5942677712453234,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.46111312095398e-05,
+      "loss": 1.2257,
+      "step": 8895
+    },
+    {
+      "epoch": 0.5946018172100481,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 8.449592491350909e-05,
+      "loss": 1.1655,
+      "step": 8900
+    },
+    {
+      "epoch": 0.5949358631747729,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 8.438073969822047e-05,
+      "loss": 1.1703,
+      "step": 8905
+    },
+    {
+      "epoch": 0.5952699091394976,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 8.426557572029019e-05,
+      "loss": 1.3285,
+      "step": 8910
+    },
+    {
+      "epoch": 0.5956039551042224,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 8.415043313630563e-05,
+      "loss": 1.1821,
+      "step": 8915
+    },
+    {
+      "epoch": 0.5959380010689471,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.403531210282507e-05,
+      "loss": 1.2104,
+      "step": 8920
+    },
+    {
+      "epoch": 0.5962720470336719,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 8.392021277637743e-05,
+      "loss": 1.1807,
+      "step": 8925
+    },
+    {
+      "epoch": 0.5966060929983966,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 8.380513531346218e-05,
+      "loss": 1.2005,
+      "step": 8930
+    },
+    {
+      "epoch": 0.5969401389631214,
+      "grad_norm": 0.171875,
+      "learning_rate": 8.369007987054906e-05,
+      "loss": 1.2966,
+      "step": 8935
+    },
+    {
+      "epoch": 0.5972741849278461,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 8.357504660407779e-05,
+      "loss": 1.16,
+      "step": 8940
+    },
+    {
+      "epoch": 0.5976082308925709,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 8.346003567045806e-05,
+      "loss": 1.1801,
+      "step": 8945
+    },
+    {
+      "epoch": 0.5979422768572956,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.334504722606912e-05,
+      "loss": 1.2457,
+      "step": 8950
+    },
+    {
+      "epoch": 0.5982763228220203,
+      "grad_norm": 0.15234375,
+      "learning_rate": 8.323008142725967e-05,
+      "loss": 1.1636,
+      "step": 8955
+    },
+    {
+      "epoch": 0.598610368786745,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 8.311513843034755e-05,
+      "loss": 1.1566,
+      "step": 8960
+    },
+    {
+      "epoch": 0.5989444147514698,
+      "grad_norm": 0.1640625,
+      "learning_rate": 8.300021839161969e-05,
+      "loss": 1.2084,
+      "step": 8965
+    },
+    {
+      "epoch": 0.5992784607161945,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.288532146733173e-05,
+      "loss": 1.1723,
+      "step": 8970
+    },
+    {
+      "epoch": 0.5996125066809193,
+      "grad_norm": 0.181640625,
+      "learning_rate": 8.277044781370799e-05,
+      "loss": 1.2041,
+      "step": 8975
+    },
+    {
+      "epoch": 0.599946552645644,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 8.265559758694099e-05,
+      "loss": 1.2412,
+      "step": 8980
+    },
+    {
+      "epoch": 0.6002805986103688,
+      "grad_norm": 0.181640625,
+      "learning_rate": 8.254077094319153e-05,
+      "loss": 1.1691,
+      "step": 8985
+    },
+    {
+      "epoch": 0.6006146445750935,
+      "grad_norm": 0.162109375,
+      "learning_rate": 8.24259680385883e-05,
+      "loss": 1.1641,
+      "step": 8990
+    },
+    {
+      "epoch": 0.6009486905398183,
+      "grad_norm": 0.16015625,
+      "learning_rate": 8.231118902922767e-05,
+      "loss": 1.1753,
+      "step": 8995
+    },
+    {
+      "epoch": 0.601282736504543,
+      "grad_norm": 0.1640625,
+      "learning_rate": 8.219643407117356e-05,
+      "loss": 1.1704,
+      "step": 9000
+    },
+    {
+      "epoch": 0.6016167824692678,
+      "grad_norm": 0.232421875,
+      "learning_rate": 8.208170332045723e-05,
+      "loss": 1.1438,
+      "step": 9005
+    },
+    {
+      "epoch": 0.6019508284339925,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 8.196699693307695e-05,
+      "loss": 1.1879,
+      "step": 9010
+    },
+    {
+      "epoch": 0.6022848743987173,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.18523150649979e-05,
+      "loss": 1.2226,
+      "step": 9015
+    },
+    {
+      "epoch": 0.602618920363442,
+      "grad_norm": 0.166015625,
+      "learning_rate": 8.173765787215193e-05,
+      "loss": 1.2468,
+      "step": 9020
+    },
+    {
+      "epoch": 0.6029529663281668,
+      "grad_norm": 0.181640625,
+      "learning_rate": 8.162302551043728e-05,
+      "loss": 1.1547,
+      "step": 9025
+    },
+    {
+      "epoch": 0.6032870122928915,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.150841813571849e-05,
+      "loss": 1.1479,
+      "step": 9030
+    },
+    {
+      "epoch": 0.6036210582576162,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 8.139383590382614e-05,
+      "loss": 1.2236,
+      "step": 9035
+    },
+    {
+      "epoch": 0.603955104222341,
+      "grad_norm": 0.16015625,
+      "learning_rate": 8.127927897055658e-05,
+      "loss": 1.2452,
+      "step": 9040
+    },
+    {
+      "epoch": 0.6042891501870657,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.116474749167175e-05,
+      "loss": 1.1888,
+      "step": 9045
+    },
+    {
+      "epoch": 0.6046231961517905,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.105024162289901e-05,
+      "loss": 1.1423,
+      "step": 9050
+    },
+    {
+      "epoch": 0.6049572421165152,
+      "grad_norm": 0.193359375,
+      "learning_rate": 8.093576151993092e-05,
+      "loss": 1.1972,
+      "step": 9055
+    },
+    {
+      "epoch": 0.60529128808124,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 8.082130733842492e-05,
+      "loss": 1.1641,
+      "step": 9060
+    },
+    {
+      "epoch": 0.6056253340459647,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 8.070687923400328e-05,
+      "loss": 1.2438,
+      "step": 9065
+    },
+    {
+      "epoch": 0.6059593800106895,
+      "grad_norm": 0.166015625,
+      "learning_rate": 8.059247736225285e-05,
+      "loss": 1.2421,
+      "step": 9070
+    },
+    {
+      "epoch": 0.6062934259754142,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.047810187872473e-05,
+      "loss": 1.2011,
+      "step": 9075
+    },
+    {
+      "epoch": 0.606627471940139,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.036375293893413e-05,
+      "loss": 1.1614,
+      "step": 9080
+    },
+    {
+      "epoch": 0.6069615179048637,
+      "grad_norm": 0.1640625,
+      "learning_rate": 8.024943069836023e-05,
+      "loss": 1.121,
+      "step": 9085
+    },
+    {
+      "epoch": 0.6072955638695885,
+      "grad_norm": 0.15625,
+      "learning_rate": 8.013513531244593e-05,
+      "loss": 1.1455,
+      "step": 9090
+    },
+    {
+      "epoch": 0.6076296098343132,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.002086693659746e-05,
+      "loss": 1.157,
+      "step": 9095
+    },
+    {
+      "epoch": 0.607963655799038,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.990662572618452e-05,
+      "loss": 1.1843,
+      "step": 9100
+    },
+    {
+      "epoch": 0.6082977017637627,
+      "grad_norm": 0.20703125,
+      "learning_rate": 7.979241183653978e-05,
+      "loss": 1.2587,
+      "step": 9105
+    },
+    {
+      "epoch": 0.6086317477284875,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 7.967822542295877e-05,
+      "loss": 1.2498,
+      "step": 9110
+    },
+    {
+      "epoch": 0.6089657936932121,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.956406664069964e-05,
+      "loss": 1.2265,
+      "step": 9115
+    },
+    {
+      "epoch": 0.6092998396579369,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.9449935644983e-05,
+      "loss": 1.1527,
+      "step": 9120
+    },
+    {
+      "epoch": 0.6096338856226616,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.933583259099162e-05,
+      "loss": 1.1783,
+      "step": 9125
+    },
+    {
+      "epoch": 0.6099679315873864,
+      "grad_norm": 0.16015625,
+      "learning_rate": 7.922175763387043e-05,
+      "loss": 1.2366,
+      "step": 9130
+    },
+    {
+      "epoch": 0.6103019775521111,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.910771092872599e-05,
+      "loss": 1.1738,
+      "step": 9135
+    },
+    {
+      "epoch": 0.6106360235168359,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.899369263062654e-05,
+      "loss": 1.1757,
+      "step": 9140
+    },
+    {
+      "epoch": 0.6109700694815606,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 7.887970289460161e-05,
+      "loss": 1.28,
+      "step": 9145
+    },
+    {
+      "epoch": 0.6113041154462854,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.876574187564201e-05,
+      "loss": 1.1575,
+      "step": 9150
+    },
+    {
+      "epoch": 0.6116381614110101,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.865180972869938e-05,
+      "loss": 1.1488,
+      "step": 9155
+    },
+    {
+      "epoch": 0.6119722073757349,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.853790660868625e-05,
+      "loss": 1.2271,
+      "step": 9160
+    },
+    {
+      "epoch": 0.6123062533404596,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 7.842403267047556e-05,
+      "loss": 1.1493,
+      "step": 9165
+    },
+    {
+      "epoch": 0.6126402993051844,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 7.831018806890058e-05,
+      "loss": 1.1521,
+      "step": 9170
+    },
+    {
+      "epoch": 0.6129743452699091,
+      "grad_norm": 0.232421875,
+      "learning_rate": 7.819637295875478e-05,
+      "loss": 1.0877,
+      "step": 9175
+    },
+    {
+      "epoch": 0.6133083912346339,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.808258749479142e-05,
+      "loss": 1.2084,
+      "step": 9180
+    },
+    {
+      "epoch": 0.6136424371993586,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.796883183172352e-05,
+      "loss": 1.2174,
+      "step": 9185
+    },
+    {
+      "epoch": 0.6139764831640834,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.785510612422359e-05,
+      "loss": 1.2459,
+      "step": 9190
+    },
+    {
+      "epoch": 0.6143105291288081,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.774141052692337e-05,
+      "loss": 1.232,
+      "step": 9195
+    },
+    {
+      "epoch": 0.6146445750935329,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.762774519441366e-05,
+      "loss": 1.1734,
+      "step": 9200
+    },
+    {
+      "epoch": 0.6149786210582576,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.751411028124414e-05,
+      "loss": 1.1733,
+      "step": 9205
+    },
+    {
+      "epoch": 0.6153126670229824,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.740050594192308e-05,
+      "loss": 1.2005,
+      "step": 9210
+    },
+    {
+      "epoch": 0.6156467129877071,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.728693233091721e-05,
+      "loss": 1.2093,
+      "step": 9215
+    },
+    {
+      "epoch": 0.6159807589524319,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.717338960265152e-05,
+      "loss": 1.1545,
+      "step": 9220
+    },
+    {
+      "epoch": 0.6163148049171566,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.705987791150895e-05,
+      "loss": 1.2138,
+      "step": 9225
+    },
+    {
+      "epoch": 0.6166488508818814,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.694639741183027e-05,
+      "loss": 1.2281,
+      "step": 9230
+    },
+    {
+      "epoch": 0.6169828968466061,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.68329482579138e-05,
+      "loss": 1.2119,
+      "step": 9235
+    },
+    {
+      "epoch": 0.6173169428113309,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.671953060401528e-05,
+      "loss": 1.2096,
+      "step": 9240
+    },
+    {
+      "epoch": 0.6176509887760556,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.660614460434754e-05,
+      "loss": 1.2063,
+      "step": 9245
+    },
+    {
+      "epoch": 0.6179850347407804,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 7.649279041308057e-05,
+      "loss": 1.2487,
+      "step": 9250
+    },
+    {
+      "epoch": 0.6183190807055051,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.637946818434087e-05,
+      "loss": 1.2365,
+      "step": 9255
+    },
+    {
+      "epoch": 0.6186531266702299,
+      "grad_norm": 0.193359375,
+      "learning_rate": 7.626617807221166e-05,
+      "loss": 1.1823,
+      "step": 9260
+    },
+    {
+      "epoch": 0.6189871726349546,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.615292023073235e-05,
+      "loss": 1.1834,
+      "step": 9265
+    },
+    {
+      "epoch": 0.6193212185996794,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.603969481389856e-05,
+      "loss": 1.248,
+      "step": 9270
+    },
+    {
+      "epoch": 0.619655264564404,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.592650197566181e-05,
+      "loss": 1.2045,
+      "step": 9275
+    },
+    {
+      "epoch": 0.6199893105291288,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.581334186992934e-05,
+      "loss": 1.184,
+      "step": 9280
+    },
+    {
+      "epoch": 0.6203233564938535,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.570021465056384e-05,
+      "loss": 1.1881,
+      "step": 9285
+    },
+    {
+      "epoch": 0.6206574024585783,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.558712047138328e-05,
+      "loss": 1.2263,
+      "step": 9290
+    },
+    {
+      "epoch": 0.620991448423303,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.547405948616079e-05,
+      "loss": 1.2007,
+      "step": 9295
+    },
+    {
+      "epoch": 0.6213254943880278,
+      "grad_norm": 0.189453125,
+      "learning_rate": 7.536103184862424e-05,
+      "loss": 1.1391,
+      "step": 9300
+    },
+    {
+      "epoch": 0.6216595403527525,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.524803771245628e-05,
+      "loss": 1.2094,
+      "step": 9305
+    },
+    {
+      "epoch": 0.6219935863174773,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.513507723129386e-05,
+      "loss": 1.2207,
+      "step": 9310
+    },
+    {
+      "epoch": 0.622327632282202,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.502215055872838e-05,
+      "loss": 1.2251,
+      "step": 9315
+    },
+    {
+      "epoch": 0.6226616782469268,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.49092578483051e-05,
+      "loss": 1.163,
+      "step": 9320
+    },
+    {
+      "epoch": 0.6229957242116515,
+      "grad_norm": 0.205078125,
+      "learning_rate": 7.479639925352318e-05,
+      "loss": 1.2131,
+      "step": 9325
+    },
+    {
+      "epoch": 0.6233297701763763,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.468357492783531e-05,
+      "loss": 1.1521,
+      "step": 9330
+    },
+    {
+      "epoch": 0.623663816141101,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.457078502464768e-05,
+      "loss": 1.1685,
+      "step": 9335
+    },
+    {
+      "epoch": 0.6239978621058258,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.44580296973196e-05,
+      "loss": 1.2071,
+      "step": 9340
+    },
+    {
+      "epoch": 0.6243319080705505,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.434530909916351e-05,
+      "loss": 1.2186,
+      "step": 9345
+    },
+    {
+      "epoch": 0.6246659540352753,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 7.423262338344444e-05,
+      "loss": 1.1182,
+      "step": 9350
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.411997270338008e-05,
+      "loss": 1.149,
+      "step": 9355
+    },
+    {
+      "epoch": 0.6253340459647247,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.400735721214052e-05,
+      "loss": 1.15,
+      "step": 9360
+    },
+    {
+      "epoch": 0.6256680919294495,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.389477706284793e-05,
+      "loss": 1.1869,
+      "step": 9365
+    },
+    {
+      "epoch": 0.6260021378941742,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.378223240857643e-05,
+      "loss": 1.2278,
+      "step": 9370
+    },
+    {
+      "epoch": 0.626336183858899,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.366972340235197e-05,
+      "loss": 1.1563,
+      "step": 9375
+    },
+    {
+      "epoch": 0.6266702298236237,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.355725019715195e-05,
+      "loss": 1.1964,
+      "step": 9380
+    },
+    {
+      "epoch": 0.6270042757883485,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 7.344481294590509e-05,
+      "loss": 1.1799,
+      "step": 9385
+    },
+    {
+      "epoch": 0.6273383217530732,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.333241180149123e-05,
+      "loss": 1.2004,
+      "step": 9390
+    },
+    {
+      "epoch": 0.627672367717798,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.322004691674118e-05,
+      "loss": 1.3072,
+      "step": 9395
+    },
+    {
+      "epoch": 0.6280064136825227,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.310771844443631e-05,
+      "loss": 1.3167,
+      "step": 9400
+    },
+    {
+      "epoch": 0.6283404596472475,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 7.299542653730865e-05,
+      "loss": 1.1872,
+      "step": 9405
+    },
+    {
+      "epoch": 0.6286745056119722,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.288317134804038e-05,
+      "loss": 1.1386,
+      "step": 9410
+    },
+    {
+      "epoch": 0.629008551576697,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 7.277095302926385e-05,
+      "loss": 1.1987,
+      "step": 9415
+    },
+    {
+      "epoch": 0.6293425975414217,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.265877173356116e-05,
+      "loss": 1.1932,
+      "step": 9420
+    },
+    {
+      "epoch": 0.6296766435061465,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.254662761346423e-05,
+      "loss": 1.1853,
+      "step": 9425
+    },
+    {
+      "epoch": 0.6300106894708712,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.243452082145423e-05,
+      "loss": 1.1307,
+      "step": 9430
+    },
+    {
+      "epoch": 0.630344735435596,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.232245150996181e-05,
+      "loss": 1.1789,
+      "step": 9435
+    },
+    {
+      "epoch": 0.6306787814003206,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 7.221041983136646e-05,
+      "loss": 1.2618,
+      "step": 9440
+    },
+    {
+      "epoch": 0.6310128273650454,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.209842593799662e-05,
+      "loss": 1.2339,
+      "step": 9445
+    },
+    {
+      "epoch": 0.6313468733297701,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.198646998212928e-05,
+      "loss": 1.1588,
+      "step": 9450
+    },
+    {
+      "epoch": 0.6316809192944949,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.187455211598989e-05,
+      "loss": 1.2706,
+      "step": 9455
+    },
+    {
+      "epoch": 0.6320149652592196,
+      "grad_norm": 0.162109375,
+      "learning_rate": 7.176267249175209e-05,
+      "loss": 1.2431,
+      "step": 9460
+    },
+    {
+      "epoch": 0.6323490112239444,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.165083126153756e-05,
+      "loss": 1.1589,
+      "step": 9465
+    },
+    {
+      "epoch": 0.6326830571886691,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.153902857741571e-05,
+      "loss": 1.3123,
+      "step": 9470
+    },
+    {
+      "epoch": 0.6330171031533939,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 7.14272645914036e-05,
+      "loss": 1.2838,
+      "step": 9475
+    },
+    {
+      "epoch": 0.6333511491181186,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.131553945546568e-05,
+      "loss": 1.2397,
+      "step": 9480
+    },
+    {
+      "epoch": 0.6336851950828434,
+      "grad_norm": 0.189453125,
+      "learning_rate": 7.120385332151348e-05,
+      "loss": 1.1422,
+      "step": 9485
+    },
+    {
+      "epoch": 0.6340192410475681,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.109220634140558e-05,
+      "loss": 1.2258,
+      "step": 9490
+    },
+    {
+      "epoch": 0.6343532870122929,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.098059866694733e-05,
+      "loss": 1.1345,
+      "step": 9495
+    },
+    {
+      "epoch": 0.6346873329770176,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.086903044989064e-05,
+      "loss": 1.1749,
+      "step": 9500
+    },
+    {
+      "epoch": 0.6350213789417424,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.075750184193368e-05,
+      "loss": 1.1437,
+      "step": 9505
+    },
+    {
+      "epoch": 0.6353554249064671,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.064601299472087e-05,
+      "loss": 1.1846,
+      "step": 9510
+    },
+    {
+      "epoch": 0.6356894708711919,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.053456405984253e-05,
+      "loss": 1.2066,
+      "step": 9515
+    },
+    {
+      "epoch": 0.6360235168359166,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.042315518883467e-05,
+      "loss": 1.203,
+      "step": 9520
+    },
+    {
+      "epoch": 0.6363575628006414,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.031178653317886e-05,
+      "loss": 1.2101,
+      "step": 9525
+    },
+    {
+      "epoch": 0.6366916087653661,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 7.020045824430205e-05,
+      "loss": 1.1811,
+      "step": 9530
+    },
+    {
+      "epoch": 0.6370256547300909,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.008917047357624e-05,
+      "loss": 1.2073,
+      "step": 9535
+    },
+    {
+      "epoch": 0.6373597006948156,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.997792337231827e-05,
+      "loss": 1.1933,
+      "step": 9540
+    },
+    {
+      "epoch": 0.6376937466595404,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.986671709178985e-05,
+      "loss": 1.1735,
+      "step": 9545
+    },
+    {
+      "epoch": 0.6380277926242651,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.975555178319701e-05,
+      "loss": 1.2503,
+      "step": 9550
+    },
+    {
+      "epoch": 0.6383618385889899,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.964442759769017e-05,
+      "loss": 1.2005,
+      "step": 9555
+    },
+    {
+      "epoch": 0.6386958845537146,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 6.953334468636386e-05,
+      "loss": 1.2382,
+      "step": 9560
+    },
+    {
+      "epoch": 0.6390299305184394,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.942230320025645e-05,
+      "loss": 1.0983,
+      "step": 9565
+    },
+    {
+      "epoch": 0.6393639764831641,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.931130329034993e-05,
+      "loss": 1.1712,
+      "step": 9570
+    },
+    {
+      "epoch": 0.6396980224478889,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.920034510756986e-05,
+      "loss": 1.2003,
+      "step": 9575
+    },
+    {
+      "epoch": 0.6400320684126136,
+      "grad_norm": 0.162109375,
+      "learning_rate": 6.9089428802785e-05,
+      "loss": 1.15,
+      "step": 9580
+    },
+    {
+      "epoch": 0.6403661143773384,
+      "grad_norm": 0.18359375,
+      "learning_rate": 6.897855452680715e-05,
+      "loss": 1.2444,
+      "step": 9585
+    },
+    {
+      "epoch": 0.6407001603420631,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.886772243039105e-05,
+      "loss": 1.2818,
+      "step": 9590
+    },
+    {
+      "epoch": 0.6410342063067879,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.875693266423404e-05,
+      "loss": 1.1941,
+      "step": 9595
+    },
+    {
+      "epoch": 0.6413682522715125,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.864618537897588e-05,
+      "loss": 1.184,
+      "step": 9600
+    },
+    {
+      "epoch": 0.6417022982362373,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 6.853548072519859e-05,
+      "loss": 1.114,
+      "step": 9605
+    },
+    {
+      "epoch": 0.642036344200962,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.842481885342625e-05,
+      "loss": 1.2692,
+      "step": 9610
+    },
+    {
+      "epoch": 0.6423703901656868,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.831419991412464e-05,
+      "loss": 1.1408,
+      "step": 9615
+    },
+    {
+      "epoch": 0.6427044361304115,
+      "grad_norm": 0.158203125,
+      "learning_rate": 6.820362405770143e-05,
+      "loss": 1.2433,
+      "step": 9620
+    },
+    {
+      "epoch": 0.6430384820951363,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.809309143450545e-05,
+      "loss": 1.2076,
+      "step": 9625
+    },
+    {
+      "epoch": 0.643372528059861,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.798260219482691e-05,
+      "loss": 1.2467,
+      "step": 9630
+    },
+    {
+      "epoch": 0.6437065740245858,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.787215648889689e-05,
+      "loss": 1.2093,
+      "step": 9635
+    },
+    {
+      "epoch": 0.6440406199893105,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.77617544668874e-05,
+      "loss": 1.2324,
+      "step": 9640
+    },
+    {
+      "epoch": 0.6443746659540353,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.765139627891099e-05,
+      "loss": 1.1822,
+      "step": 9645
+    },
+    {
+      "epoch": 0.64470871191876,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.754108207502069e-05,
+      "loss": 1.223,
+      "step": 9650
+    },
+    {
+      "epoch": 0.6450427578834848,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 6.743081200520962e-05,
+      "loss": 1.1462,
+      "step": 9655
+    },
+    {
+      "epoch": 0.6453768038482095,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 6.732058621941092e-05,
+      "loss": 1.2425,
+      "step": 9660
+    },
+    {
+      "epoch": 0.6457108498129343,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.721040486749756e-05,
+      "loss": 1.245,
+      "step": 9665
+    },
+    {
+      "epoch": 0.646044895777659,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.710026809928206e-05,
+      "loss": 1.1829,
+      "step": 9670
+    },
+    {
+      "epoch": 0.6463789417423838,
+      "grad_norm": 0.1875,
+      "learning_rate": 6.69901760645163e-05,
+      "loss": 1.1876,
+      "step": 9675
+    },
+    {
+      "epoch": 0.6467129877071085,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.68801289128914e-05,
+      "loss": 1.3096,
+      "step": 9680
+    },
+    {
+      "epoch": 0.6470470336718332,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.677012679403743e-05,
+      "loss": 1.2622,
+      "step": 9685
+    },
+    {
+      "epoch": 0.647381079636558,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.666016985752316e-05,
+      "loss": 1.2291,
+      "step": 9690
+    },
+    {
+      "epoch": 0.6477151256012827,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.655025825285601e-05,
+      "loss": 1.2274,
+      "step": 9695
+    },
+    {
+      "epoch": 0.6480491715660075,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.644039212948177e-05,
+      "loss": 1.2035,
+      "step": 9700
+    },
+    {
+      "epoch": 0.6483832175307322,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 6.633057163678426e-05,
+      "loss": 1.229,
+      "step": 9705
+    },
+    {
+      "epoch": 0.648717263495457,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.622079692408545e-05,
+      "loss": 1.319,
+      "step": 9710
+    },
+    {
+      "epoch": 0.6490513094601817,
+      "grad_norm": 0.166015625,
+      "learning_rate": 6.611106814064492e-05,
+      "loss": 1.1854,
+      "step": 9715
+    },
+    {
+      "epoch": 0.6493853554249065,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.600138543565986e-05,
+      "loss": 1.1671,
+      "step": 9720
+    },
+    {
+      "epoch": 0.6497194013896312,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.589174895826475e-05,
+      "loss": 1.2083,
+      "step": 9725
+    },
+    {
+      "epoch": 0.650053447354356,
+      "grad_norm": 0.162109375,
+      "learning_rate": 6.57821588575313e-05,
+      "loss": 1.147,
+      "step": 9730
+    },
+    {
+      "epoch": 0.6503874933190807,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.567261528246806e-05,
+      "loss": 1.2656,
+      "step": 9735
+    },
+    {
+      "epoch": 0.6507215392838055,
+      "grad_norm": 0.228515625,
+      "learning_rate": 6.556311838202046e-05,
+      "loss": 1.2546,
+      "step": 9740
+    },
+    {
+      "epoch": 0.6510555852485302,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.545366830507034e-05,
+      "loss": 1.2111,
+      "step": 9745
+    },
+    {
+      "epoch": 0.651389631213255,
+      "grad_norm": 0.16796875,
+      "learning_rate": 6.534426520043594e-05,
+      "loss": 1.2212,
+      "step": 9750
+    },
+    {
+      "epoch": 0.6517236771779797,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.523490921687157e-05,
+      "loss": 1.128,
+      "step": 9755
+    },
+    {
+      "epoch": 0.6520577231427044,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.512560050306756e-05,
+      "loss": 1.197,
+      "step": 9760
+    },
+    {
+      "epoch": 0.6523917691074291,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 6.501633920764989e-05,
+      "loss": 1.2123,
+      "step": 9765
+    },
+    {
+      "epoch": 0.6527258150721539,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.490712547918006e-05,
+      "loss": 1.2868,
+      "step": 9770
+    },
+    {
+      "epoch": 0.6530598610368786,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.479795946615501e-05,
+      "loss": 1.2247,
+      "step": 9775
+    },
+    {
+      "epoch": 0.6533939070016034,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.468884131700668e-05,
+      "loss": 1.2421,
+      "step": 9780
+    },
+    {
+      "epoch": 0.6537279529663281,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.457977118010196e-05,
+      "loss": 1.2825,
+      "step": 9785
+    },
+    {
+      "epoch": 0.6540619989310529,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.44707492037425e-05,
+      "loss": 1.1924,
+      "step": 9790
+    },
+    {
+      "epoch": 0.6543960448957776,
+      "grad_norm": 0.1953125,
+      "learning_rate": 6.436177553616438e-05,
+      "loss": 1.2154,
+      "step": 9795
+    },
+    {
+      "epoch": 0.6547300908605024,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.42528503255381e-05,
+      "loss": 1.1975,
+      "step": 9800
+    },
+    {
+      "epoch": 0.6550641368252271,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 6.414397371996821e-05,
+      "loss": 1.0709,
+      "step": 9805
+    },
+    {
+      "epoch": 0.6553981827899519,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.403514586749318e-05,
+      "loss": 1.2159,
+      "step": 9810
+    },
+    {
+      "epoch": 0.6557322287546766,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.392636691608521e-05,
+      "loss": 1.2559,
+      "step": 9815
+    },
+    {
+      "epoch": 0.6560662747194014,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.381763701365e-05,
+      "loss": 1.1734,
+      "step": 9820
+    },
+    {
+      "epoch": 0.6564003206841261,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 6.370895630802652e-05,
+      "loss": 1.1594,
+      "step": 9825
+    },
+    {
+      "epoch": 0.6567343666488509,
+      "grad_norm": 0.154296875,
+      "learning_rate": 6.36003249469869e-05,
+      "loss": 1.1063,
+      "step": 9830
+    },
+    {
+      "epoch": 0.6570684126135756,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.349174307823616e-05,
+      "loss": 1.1498,
+      "step": 9835
+    },
+    {
+      "epoch": 0.6574024585783004,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 6.338321084941205e-05,
+      "loss": 1.2358,
+      "step": 9840
+    },
+    {
+      "epoch": 0.6577365045430251,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.327472840808478e-05,
+      "loss": 1.2042,
+      "step": 9845
+    },
+    {
+      "epoch": 0.6580705505077499,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.316629590175688e-05,
+      "loss": 1.1814,
+      "step": 9850
+    },
+    {
+      "epoch": 0.6584045964724746,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.305791347786299e-05,
+      "loss": 1.2021,
+      "step": 9855
+    },
+    {
+      "epoch": 0.6587386424371994,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.294958128376962e-05,
+      "loss": 1.1821,
+      "step": 9860
+    },
+    {
+      "epoch": 0.6590726884019241,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.284129946677508e-05,
+      "loss": 1.1893,
+      "step": 9865
+    },
+    {
+      "epoch": 0.6594067343666489,
+      "grad_norm": 0.166015625,
+      "learning_rate": 6.27330681741091e-05,
+      "loss": 1.1647,
+      "step": 9870
+    },
+    {
+      "epoch": 0.6597407803313736,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.26248875529327e-05,
+      "loss": 1.2649,
+      "step": 9875
+    },
+    {
+      "epoch": 0.6600748262960984,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.251675775033804e-05,
+      "loss": 1.2161,
+      "step": 9880
+    },
+    {
+      "epoch": 0.6604088722608231,
+      "grad_norm": 0.1640625,
+      "learning_rate": 6.24086789133482e-05,
+      "loss": 1.1798,
+      "step": 9885
+    },
+    {
+      "epoch": 0.6607429182255479,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.23006511889169e-05,
+      "loss": 1.1734,
+      "step": 9890
+    },
+    {
+      "epoch": 0.6610769641902726,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.219267472392843e-05,
+      "loss": 1.215,
+      "step": 9895
+    },
+    {
+      "epoch": 0.6614110101549974,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.208474966519735e-05,
+      "loss": 1.1571,
+      "step": 9900
+    },
+    {
+      "epoch": 0.6617450561197221,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.197687615946832e-05,
+      "loss": 1.2906,
+      "step": 9905
+    },
+    {
+      "epoch": 0.6620791020844469,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 6.186905435341592e-05,
+      "loss": 1.1861,
+      "step": 9910
+    },
+    {
+      "epoch": 0.6624131480491716,
+      "grad_norm": 0.166015625,
+      "learning_rate": 6.17612843936444e-05,
+      "loss": 1.1958,
+      "step": 9915
+    },
+    {
+      "epoch": 0.6627471940138963,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.165356642668754e-05,
+      "loss": 1.2155,
+      "step": 9920
+    },
+    {
+      "epoch": 0.663081239978621,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 6.154590059900849e-05,
+      "loss": 1.1376,
+      "step": 9925
+    },
+    {
+      "epoch": 0.6634152859433458,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.143828705699936e-05,
+      "loss": 1.2243,
+      "step": 9930
+    },
+    {
+      "epoch": 0.6637493319080705,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.13307259469813e-05,
+      "loss": 1.2159,
+      "step": 9935
+    },
+    {
+      "epoch": 0.6640833778727953,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.122321741520412e-05,
+      "loss": 1.2191,
+      "step": 9940
+    },
+    {
+      "epoch": 0.66441742383752,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.111576160784611e-05,
+      "loss": 1.1936,
+      "step": 9945
+    },
+    {
+      "epoch": 0.6647514698022448,
+      "grad_norm": 0.205078125,
+      "learning_rate": 6.1008358671013885e-05,
+      "loss": 1.2535,
+      "step": 9950
+    },
+    {
+      "epoch": 0.6650855157669695,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.090100875074225e-05,
+      "loss": 1.1455,
+      "step": 9955
+    },
+    {
+      "epoch": 0.6654195617316943,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.079371199299384e-05,
+      "loss": 1.1946,
+      "step": 9960
+    },
+    {
+      "epoch": 0.665753607696419,
+      "grad_norm": 0.16796875,
+      "learning_rate": 6.0686468543659005e-05,
+      "loss": 1.2062,
+      "step": 9965
+    },
+    {
+      "epoch": 0.6660876536611438,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.057927854855565e-05,
+      "loss": 1.2187,
+      "step": 9970
+    },
+    {
+      "epoch": 0.6664216996258685,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.0472142153428954e-05,
+      "loss": 1.2224,
+      "step": 9975
+    },
+    {
+      "epoch": 0.6667557455905933,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.036505950395126e-05,
+      "loss": 1.1661,
+      "step": 9980
+    },
+    {
+      "epoch": 0.667089791555318,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.025803074572185e-05,
+      "loss": 1.2129,
+      "step": 9985
+    },
+    {
+      "epoch": 0.6674238375200428,
+      "grad_norm": 0.19140625,
+      "learning_rate": 6.0151056024266695e-05,
+      "loss": 1.2235,
+      "step": 9990
+    },
+    {
+      "epoch": 0.6677578834847675,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.0044135485038265e-05,
+      "loss": 1.1978,
+      "step": 9995
+    },
+    {
+      "epoch": 0.6680919294494923,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.9937269273415386e-05,
+      "loss": 1.2076,
+      "step": 10000
+    },
+    {
+      "epoch": 0.668425975414217,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.983045753470308e-05,
+      "loss": 1.2702,
+      "step": 10005
+    },
+    {
+      "epoch": 0.6687600213789417,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 5.972370041413218e-05,
+      "loss": 1.241,
+      "step": 10010
+    },
+    {
+      "epoch": 0.6690940673436665,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 5.961699805685932e-05,
+      "loss": 1.2256,
+      "step": 10015
+    },
+    {
+      "epoch": 0.6694281133083912,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 5.95103506079667e-05,
+      "loss": 1.1762,
+      "step": 10020
+    },
+    {
+      "epoch": 0.669762159273116,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.940375821246186e-05,
+      "loss": 1.1551,
+      "step": 10025
+    },
+    {
+      "epoch": 0.6700962052378407,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 5.9297221015277394e-05,
+      "loss": 1.1554,
+      "step": 10030
+    },
+    {
+      "epoch": 0.6704302512025655,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.9190739161270956e-05,
+      "loss": 1.1697,
+      "step": 10035
+    },
+    {
+      "epoch": 0.6707642971672902,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 5.9084312795224874e-05,
+      "loss": 1.1717,
+      "step": 10040
+    },
+    {
+      "epoch": 0.671098343132015,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.8977942061846034e-05,
+      "loss": 1.2296,
+      "step": 10045
+    },
+    {
+      "epoch": 0.6714323890967397,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.8871627105765746e-05,
+      "loss": 1.2098,
+      "step": 10050
+    },
+    {
+      "epoch": 0.6717664350614645,
+      "grad_norm": 0.166015625,
+      "learning_rate": 5.876536807153943e-05,
+      "loss": 1.1903,
+      "step": 10055
+    },
+    {
+      "epoch": 0.6721004810261892,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.865916510364648e-05,
+      "loss": 1.1912,
+      "step": 10060
+    },
+    {
+      "epoch": 0.672434526990914,
+      "grad_norm": 0.16015625,
+      "learning_rate": 5.855301834649003e-05,
+      "loss": 1.1724,
+      "step": 10065
+    },
+    {
+      "epoch": 0.6727685729556387,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.8446927944396815e-05,
+      "loss": 1.2208,
+      "step": 10070
+    },
+    {
+      "epoch": 0.6731026189203635,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 5.834089404161689e-05,
+      "loss": 1.2138,
+      "step": 10075
+    },
+    {
+      "epoch": 0.6734366648850882,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.8234916782323646e-05,
+      "loss": 1.2382,
+      "step": 10080
+    },
+    {
+      "epoch": 0.6737707108498129,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.812899631061327e-05,
+      "loss": 1.1906,
+      "step": 10085
+    },
+    {
+      "epoch": 0.6741047568145376,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 5.80231327705048e-05,
+      "loss": 1.2238,
+      "step": 10090
+    },
+    {
+      "epoch": 0.6744388027792624,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.791732630593991e-05,
+      "loss": 1.1796,
+      "step": 10095
+    },
+    {
+      "epoch": 0.6747728487439871,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 5.781157706078264e-05,
+      "loss": 1.247,
+      "step": 10100
+    },
+    {
+      "epoch": 0.6751068947087119,
+      "grad_norm": 0.18359375,
+      "learning_rate": 5.770588517881918e-05,
+      "loss": 1.2604,
+      "step": 10105
+    },
+    {
+      "epoch": 0.6754409406734366,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.760025080375777e-05,
+      "loss": 1.2595,
+      "step": 10110
+    },
+    {
+      "epoch": 0.6757749866381614,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.749467407922853e-05,
+      "loss": 1.1983,
+      "step": 10115
+    },
+    {
+      "epoch": 0.6761090326028861,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.738915514878307e-05,
+      "loss": 1.2557,
+      "step": 10120
+    },
+    {
+      "epoch": 0.6764430785676109,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 5.728369415589443e-05,
+      "loss": 1.1756,
+      "step": 10125
+    },
+    {
+      "epoch": 0.6767771245323356,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.717829124395699e-05,
+      "loss": 1.2396,
+      "step": 10130
+    },
+    {
+      "epoch": 0.6771111704970604,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 5.707294655628599e-05,
+      "loss": 1.1818,
+      "step": 10135
+    },
+    {
+      "epoch": 0.6774452164617851,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.696766023611768e-05,
+      "loss": 1.1824,
+      "step": 10140
+    },
+    {
+      "epoch": 0.6777792624265099,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 5.686243242660876e-05,
+      "loss": 1.1367,
+      "step": 10145
+    },
+    {
+      "epoch": 0.6781133083912346,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 5.6757263270836594e-05,
+      "loss": 1.2218,
+      "step": 10150
+    },
+    {
+      "epoch": 0.6784473543559594,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.6652152911798585e-05,
+      "loss": 1.2477,
+      "step": 10155
+    },
+    {
+      "epoch": 0.6787814003206841,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 5.6547101492412265e-05,
+      "loss": 1.2255,
+      "step": 10160
+    },
+    {
+      "epoch": 0.6791154462854089,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 5.644210915551509e-05,
+      "loss": 1.2308,
+      "step": 10165
+    },
+    {
+      "epoch": 0.6794494922501336,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 5.633717604386415e-05,
+      "loss": 1.1479,
+      "step": 10170
+    },
+    {
+      "epoch": 0.6797835382148584,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.6232302300135966e-05,
+      "loss": 1.171,
+      "step": 10175
+    },
+    {
+      "epoch": 0.6801175841795831,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 5.612748806692632e-05,
+      "loss": 1.2255,
+      "step": 10180
+    },
+    {
+      "epoch": 0.6804516301443079,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.602273348675019e-05,
+      "loss": 1.1911,
+      "step": 10185
+    },
+    {
+      "epoch": 0.6807856761090326,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 5.5918038702041343e-05,
+      "loss": 1.1586,
+      "step": 10190
+    },
+    {
+      "epoch": 0.6811197220737574,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.5813403855152224e-05,
+      "loss": 1.2018,
+      "step": 10195
+    },
+    {
+      "epoch": 0.6814537680384821,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.5708829088353875e-05,
+      "loss": 1.1946,
+      "step": 10200
+    },
+    {
+      "epoch": 0.6817878140032069,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.560431454383566e-05,
+      "loss": 1.2084,
+      "step": 10205
+    },
+    {
+      "epoch": 0.6821218599679316,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 5.549986036370491e-05,
+      "loss": 1.2056,
+      "step": 10210
+    },
+    {
+      "epoch": 0.6824559059326564,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.5395466689987044e-05,
+      "loss": 1.1836,
+      "step": 10215
+    },
+    {
+      "epoch": 0.6827899518973811,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.529113366462504e-05,
+      "loss": 1.1598,
+      "step": 10220
+    },
+    {
+      "epoch": 0.6831239978621059,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.518686142947962e-05,
+      "loss": 1.1612,
+      "step": 10225
+    },
+    {
+      "epoch": 0.6834580438268306,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.508265012632865e-05,
+      "loss": 1.2547,
+      "step": 10230
+    },
+    {
+      "epoch": 0.6837920897915554,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 5.497849989686732e-05,
+      "loss": 1.2068,
+      "step": 10235
+    },
+    {
+      "epoch": 0.6841261357562801,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.4874410882707635e-05,
+      "loss": 1.2069,
+      "step": 10240
+    },
+    {
+      "epoch": 0.6844601817210048,
+      "grad_norm": 0.1640625,
+      "learning_rate": 5.4770383225378486e-05,
+      "loss": 1.1858,
+      "step": 10245
+    },
+    {
+      "epoch": 0.6847942276857295,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.466641706632525e-05,
+      "loss": 1.1981,
+      "step": 10250
+    },
+    {
+      "epoch": 0.6851282736504543,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.456251254690967e-05,
+      "loss": 1.2036,
+      "step": 10255
+    },
+    {
+      "epoch": 0.685462319615179,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 5.4458669808409766e-05,
+      "loss": 1.2206,
+      "step": 10260
+    },
+    {
+      "epoch": 0.6857963655799038,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 5.435488899201957e-05,
+      "loss": 1.1534,
+      "step": 10265
+    },
+    {
+      "epoch": 0.6861304115446285,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.42511702388488e-05,
+      "loss": 1.2095,
+      "step": 10270
+    },
+    {
+      "epoch": 0.6864644575093533,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.4147513689922815e-05,
+      "loss": 1.1465,
+      "step": 10275
+    },
+    {
+      "epoch": 0.686798503474078,
+      "grad_norm": 0.162109375,
+      "learning_rate": 5.404391948618252e-05,
+      "loss": 1.2494,
+      "step": 10280
+    },
+    {
+      "epoch": 0.6871325494388028,
+      "grad_norm": 0.166015625,
+      "learning_rate": 5.39403877684839e-05,
+      "loss": 1.2441,
+      "step": 10285
+    },
+    {
+      "epoch": 0.6874665954035275,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 5.3836918677598015e-05,
+      "loss": 1.1477,
+      "step": 10290
+    },
+    {
+      "epoch": 0.6878006413682523,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 5.373351235421084e-05,
+      "loss": 1.244,
+      "step": 10295
+    },
+    {
+      "epoch": 0.688134687332977,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 5.3630168938922984e-05,
+      "loss": 1.252,
+      "step": 10300
+    },
+    {
+      "epoch": 0.6884687332977018,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 5.352688857224945e-05,
+      "loss": 1.1898,
+      "step": 10305
+    },
+    {
+      "epoch": 0.6888027792624265,
+      "grad_norm": 0.16015625,
+      "learning_rate": 5.3423671394619554e-05,
+      "loss": 1.1629,
+      "step": 10310
+    },
+    {
+      "epoch": 0.6891368252271513,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 5.332051754637676e-05,
+      "loss": 1.2108,
+      "step": 10315
+    },
+    {
+      "epoch": 0.689470871191876,
+      "grad_norm": 0.15625,
+      "learning_rate": 5.321742716777829e-05,
+      "loss": 1.1881,
+      "step": 10320
+    },
+    {
+      "epoch": 0.6898049171566007,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 5.311440039899521e-05,
+      "loss": 1.1336,
+      "step": 10325
+    },
+    {
+      "epoch": 0.6901389631213255,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.301143738011197e-05,
+      "loss": 1.1887,
+      "step": 10330
+    },
+    {
+      "epoch": 0.6904730090860502,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 5.290853825112647e-05,
+      "loss": 1.1783,
+      "step": 10335
+    },
+    {
+      "epoch": 0.690807055050775,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.2805703151949616e-05,
+      "loss": 1.2454,
+      "step": 10340
+    },
+    {
+      "epoch": 0.6911411010154997,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 5.2702932222405286e-05,
+      "loss": 1.2103,
+      "step": 10345
+    },
+    {
+      "epoch": 0.6914751469802245,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.2600225602230166e-05,
+      "loss": 1.2498,
+      "step": 10350
+    },
+    {
+      "epoch": 0.6918091929449492,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 5.249758343107348e-05,
+      "loss": 1.1575,
+      "step": 10355
+    },
+    {
+      "epoch": 0.692143238909674,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 5.239500584849678e-05,
+      "loss": 1.2488,
+      "step": 10360
+    },
+    {
+      "epoch": 0.6924772848743987,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.229249299397378e-05,
+      "loss": 1.2236,
+      "step": 10365
+    },
+    {
+      "epoch": 0.6928113308391235,
+      "grad_norm": 0.181640625,
+      "learning_rate": 5.219004500689031e-05,
+      "loss": 1.1939,
+      "step": 10370
+    },
+    {
+      "epoch": 0.6931453768038482,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.2087662026543846e-05,
+      "loss": 1.2497,
+      "step": 10375
+    },
+    {
+      "epoch": 0.693479422768573,
+      "grad_norm": 0.193359375,
+      "learning_rate": 5.1985344192143534e-05,
+      "loss": 1.1932,
+      "step": 10380
+    },
+    {
+      "epoch": 0.6938134687332977,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.188309164281e-05,
+      "loss": 1.1782,
+      "step": 10385
+    },
+    {
+      "epoch": 0.6941475146980225,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.1780904517575046e-05,
+      "loss": 1.2641,
+      "step": 10390
+    },
+    {
+      "epoch": 0.6944815606627472,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.1678782955381534e-05,
+      "loss": 1.1262,
+      "step": 10395
+    },
+    {
+      "epoch": 0.694815606627472,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 5.157672709508312e-05,
+      "loss": 1.2646,
+      "step": 10400
+    },
+    {
+      "epoch": 0.6951496525921966,
+      "grad_norm": 0.166015625,
+      "learning_rate": 5.147473707544425e-05,
+      "loss": 1.1851,
+      "step": 10405
+    },
+    {
+      "epoch": 0.6954836985569214,
+      "grad_norm": 0.1875,
+      "learning_rate": 5.137281303513969e-05,
+      "loss": 1.1714,
+      "step": 10410
+    },
+    {
+      "epoch": 0.6958177445216461,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.127095511275466e-05,
+      "loss": 1.2067,
+      "step": 10415
+    },
+    {
+      "epoch": 0.6961517904863709,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 5.116916344678435e-05,
+      "loss": 1.2317,
+      "step": 10420
+    },
+    {
+      "epoch": 0.6964858364510956,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 5.106743817563395e-05,
+      "loss": 1.2716,
+      "step": 10425
+    },
+    {
+      "epoch": 0.6968198824158204,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.0965779437618314e-05,
+      "loss": 1.1254,
+      "step": 10430
+    },
+    {
+      "epoch": 0.6971539283805451,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 5.086418737096186e-05,
+      "loss": 1.211,
+      "step": 10435
+    },
+    {
+      "epoch": 0.6974879743452699,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.076266211379826e-05,
+      "loss": 1.2339,
+      "step": 10440
+    },
+    {
+      "epoch": 0.6978220203099946,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.066120380417056e-05,
+      "loss": 1.2717,
+      "step": 10445
+    },
+    {
+      "epoch": 0.6981560662747194,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.055981258003064e-05,
+      "loss": 1.2531,
+      "step": 10450
+    },
+    {
+      "epoch": 0.6984901122394441,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 5.0458488579239116e-05,
+      "loss": 1.1889,
+      "step": 10455
+    },
+    {
+      "epoch": 0.6988241582041689,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.035723193956523e-05,
+      "loss": 1.1753,
+      "step": 10460
+    },
+    {
+      "epoch": 0.6991582041688936,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 5.025604279868676e-05,
+      "loss": 1.303,
+      "step": 10465
+    },
+    {
+      "epoch": 0.6994922501336184,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.015492129418957e-05,
+      "loss": 1.2091,
+      "step": 10470
+    },
+    {
+      "epoch": 0.6998262960983431,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.005386756356754e-05,
+      "loss": 1.2594,
+      "step": 10475
+    },
+    {
+      "epoch": 0.7001603420630679,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.995288174422251e-05,
+      "loss": 1.1741,
+      "step": 10480
+    },
+    {
+      "epoch": 0.7004943880277926,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.985196397346397e-05,
+      "loss": 1.2114,
+      "step": 10485
+    },
+    {
+      "epoch": 0.7008284339925174,
+      "grad_norm": 0.181640625,
+      "learning_rate": 4.97511143885088e-05,
+      "loss": 1.1966,
+      "step": 10490
+    },
+    {
+      "epoch": 0.7011624799572421,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.9650333126481174e-05,
+      "loss": 1.1899,
+      "step": 10495
+    },
+    {
+      "epoch": 0.7014965259219669,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.954962032441249e-05,
+      "loss": 1.1736,
+      "step": 10500
+    },
+    {
+      "epoch": 0.7018305718866916,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 4.9448976119240895e-05,
+      "loss": 1.1972,
+      "step": 10505
+    },
+    {
+      "epoch": 0.7021646178514164,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 4.934840064781143e-05,
+      "loss": 1.2075,
+      "step": 10510
+    },
+    {
+      "epoch": 0.7024986638161411,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.924789404687552e-05,
+      "loss": 1.2075,
+      "step": 10515
+    },
+    {
+      "epoch": 0.7028327097808659,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.914745645309111e-05,
+      "loss": 1.198,
+      "step": 10520
+    },
+    {
+      "epoch": 0.7031667557455906,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.904708800302218e-05,
+      "loss": 1.2011,
+      "step": 10525
+    },
+    {
+      "epoch": 0.7035008017103154,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.8946788833138724e-05,
+      "loss": 1.1074,
+      "step": 10530
+    },
+    {
+      "epoch": 0.7038348476750401,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.884655907981659e-05,
+      "loss": 1.1243,
+      "step": 10535
+    },
+    {
+      "epoch": 0.7041688936397649,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.874639887933725e-05,
+      "loss": 1.1672,
+      "step": 10540
+    },
+    {
+      "epoch": 0.7045029396044896,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 4.864630836788753e-05,
+      "loss": 1.1574,
+      "step": 10545
+    },
+    {
+      "epoch": 0.7048369855692144,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.854628768155951e-05,
+      "loss": 1.1272,
+      "step": 10550
+    },
+    {
+      "epoch": 0.7051710315339391,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.844633695635041e-05,
+      "loss": 1.2193,
+      "step": 10555
+    },
+    {
+      "epoch": 0.7055050774986639,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.834645632816227e-05,
+      "loss": 1.2342,
+      "step": 10560
+    },
+    {
+      "epoch": 0.7058391234633885,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.824664593280175e-05,
+      "loss": 1.2653,
+      "step": 10565
+    },
+    {
+      "epoch": 0.7061731694281133,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.8146905905980144e-05,
+      "loss": 1.1794,
+      "step": 10570
+    },
+    {
+      "epoch": 0.706507215392838,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.804723638331303e-05,
+      "loss": 1.2321,
+      "step": 10575
+    },
+    {
+      "epoch": 0.7068412613575628,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 4.7947637500320084e-05,
+      "loss": 1.1332,
+      "step": 10580
+    },
+    {
+      "epoch": 0.7071753073222875,
+      "grad_norm": 0.177734375,
+      "learning_rate": 4.7848109392424914e-05,
+      "loss": 1.2459,
+      "step": 10585
+    },
+    {
+      "epoch": 0.7075093532870123,
+      "grad_norm": 0.18359375,
+      "learning_rate": 4.774865219495498e-05,
+      "loss": 1.254,
+      "step": 10590
+    },
+    {
+      "epoch": 0.707843399251737,
+      "grad_norm": 0.1953125,
+      "learning_rate": 4.7649266043141236e-05,
+      "loss": 1.2396,
+      "step": 10595
+    },
+    {
+      "epoch": 0.7081774452164618,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.7549951072118137e-05,
+      "loss": 1.125,
+      "step": 10600
+    },
+    {
+      "epoch": 0.7085114911811865,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 4.7450707416923225e-05,
+      "loss": 1.2224,
+      "step": 10605
+    },
+    {
+      "epoch": 0.7088455371459113,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.7351535212497213e-05,
+      "loss": 1.2234,
+      "step": 10610
+    },
+    {
+      "epoch": 0.709179583110636,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.725243459368358e-05,
+      "loss": 1.1677,
+      "step": 10615
+    },
+    {
+      "epoch": 0.7095136290753608,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.7153405695228457e-05,
+      "loss": 1.2209,
+      "step": 10620
+    },
+    {
+      "epoch": 0.7098476750400855,
+      "grad_norm": 0.18359375,
+      "learning_rate": 4.7054448651780434e-05,
+      "loss": 1.1557,
+      "step": 10625
+    },
+    {
+      "epoch": 0.7101817210048103,
+      "grad_norm": 0.181640625,
+      "learning_rate": 4.695556359789061e-05,
+      "loss": 1.2074,
+      "step": 10630
+    },
+    {
+      "epoch": 0.710515766969535,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 4.685675066801194e-05,
+      "loss": 1.0386,
+      "step": 10635
+    },
+    {
+      "epoch": 0.7108498129342598,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.675800999649943e-05,
+      "loss": 1.2289,
+      "step": 10640
+    },
+    {
+      "epoch": 0.7111838588989845,
+      "grad_norm": 0.1640625,
+      "learning_rate": 4.66593417176098e-05,
+      "loss": 1.2262,
+      "step": 10645
+    },
+    {
+      "epoch": 0.7115179048637092,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.656074596550142e-05,
+      "loss": 1.1148,
+      "step": 10650
+    },
+    {
+      "epoch": 0.711851950828434,
+      "grad_norm": 0.158203125,
+      "learning_rate": 4.646222287423391e-05,
+      "loss": 1.2033,
+      "step": 10655
+    },
+    {
+      "epoch": 0.7121859967931587,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 4.636377257776826e-05,
+      "loss": 1.1739,
+      "step": 10660
+    },
+    {
+      "epoch": 0.7125200427578835,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.626539520996632e-05,
+      "loss": 1.1661,
+      "step": 10665
+    },
+    {
+      "epoch": 0.7128540887226082,
+      "grad_norm": 0.16015625,
+      "learning_rate": 4.6167090904590916e-05,
+      "loss": 1.2023,
+      "step": 10670
+    },
+    {
+      "epoch": 0.713188134687333,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 4.606885979530544e-05,
+      "loss": 1.2215,
+      "step": 10675
+    },
+    {
+      "epoch": 0.7135221806520577,
+      "grad_norm": 0.181640625,
+      "learning_rate": 4.597070201567374e-05,
+      "loss": 1.2264,
+      "step": 10680
+    },
+    {
+      "epoch": 0.7138562266167825,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.5872617699160095e-05,
+      "loss": 1.1838,
+      "step": 10685
+    },
+    {
+      "epoch": 0.7141902725815072,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 4.577460697912873e-05,
+      "loss": 1.2051,
+      "step": 10690
+    },
+    {
+      "epoch": 0.714524318546232,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.567666998884395e-05,
+      "loss": 1.1821,
+      "step": 10695
+    },
+    {
+      "epoch": 0.7148583645109567,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.557880686146968e-05,
+      "loss": 1.2537,
+      "step": 10700
+    },
+    {
+      "epoch": 0.7151924104756815,
+      "grad_norm": 0.189453125,
+      "learning_rate": 4.548101773006953e-05,
+      "loss": 1.1674,
+      "step": 10705
+    },
+    {
+      "epoch": 0.7155264564404062,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.5383302727606426e-05,
+      "loss": 1.1552,
+      "step": 10710
+    },
+    {
+      "epoch": 0.715860502405131,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.528566198694246e-05,
+      "loss": 1.2517,
+      "step": 10715
+    },
+    {
+      "epoch": 0.7161945483698557,
+      "grad_norm": 0.193359375,
+      "learning_rate": 4.5188095640838865e-05,
+      "loss": 1.176,
+      "step": 10720
+    },
+    {
+      "epoch": 0.7165285943345805,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 4.50906038219557e-05,
+      "loss": 1.1779,
+      "step": 10725
+    },
+    {
+      "epoch": 0.7168626402993051,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 4.499318666285162e-05,
+      "loss": 1.1945,
+      "step": 10730
+    },
+    {
+      "epoch": 0.7171966862640299,
+      "grad_norm": 0.177734375,
+      "learning_rate": 4.489584429598375e-05,
+      "loss": 1.2406,
+      "step": 10735
+    },
+    {
+      "epoch": 0.7175307322287546,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.4798576853707664e-05,
+      "loss": 1.207,
+      "step": 10740
+    },
+    {
+      "epoch": 0.7178647781934794,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 4.470138446827692e-05,
+      "loss": 1.2622,
+      "step": 10745
+    },
+    {
+      "epoch": 0.7181988241582041,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 4.4604267271843046e-05,
+      "loss": 1.1559,
+      "step": 10750
+    },
+    {
+      "epoch": 0.7185328701229289,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.4507225396455385e-05,
+      "loss": 1.2233,
+      "step": 10755
+    },
+    {
+      "epoch": 0.7188669160876536,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.44102589740609e-05,
+      "loss": 1.2279,
+      "step": 10760
+    },
+    {
+      "epoch": 0.7192009620523784,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.431336813650385e-05,
+      "loss": 1.2143,
+      "step": 10765
+    },
+    {
+      "epoch": 0.7195350080171031,
+      "grad_norm": 0.21484375,
+      "learning_rate": 4.421655301552575e-05,
+      "loss": 1.2,
+      "step": 10770
+    },
+    {
+      "epoch": 0.7198690539818279,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.411981374276527e-05,
+      "loss": 1.2192,
+      "step": 10775
+    },
+    {
+      "epoch": 0.7202030999465526,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.402315044975778e-05,
+      "loss": 1.1967,
+      "step": 10780
+    },
+    {
+      "epoch": 0.7205371459112774,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 4.3926563267935514e-05,
+      "loss": 1.2115,
+      "step": 10785
+    },
+    {
+      "epoch": 0.7208711918760021,
+      "grad_norm": 0.1796875,
+      "learning_rate": 4.383005232862707e-05,
+      "loss": 1.2096,
+      "step": 10790
+    },
+    {
+      "epoch": 0.7212052378407269,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 4.37336177630575e-05,
+      "loss": 1.1461,
+      "step": 10795
+    },
+    {
+      "epoch": 0.7215392838054516,
+      "grad_norm": 0.177734375,
+      "learning_rate": 4.363725970234794e-05,
+      "loss": 1.2384,
+      "step": 10800
+    },
+    {
+      "epoch": 0.7218733297701764,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.354097827751552e-05,
+      "loss": 1.2205,
+      "step": 10805
+    },
+    {
+      "epoch": 0.7222073757349011,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 4.344477361947309e-05,
+      "loss": 1.1431,
+      "step": 10810
+    },
+    {
+      "epoch": 0.7225414216996259,
+      "grad_norm": 0.177734375,
+      "learning_rate": 4.334864585902935e-05,
+      "loss": 1.1884,
+      "step": 10815
+    },
+    {
+      "epoch": 0.7228754676643506,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 4.3252595126888205e-05,
+      "loss": 1.1578,
+      "step": 10820
+    },
+    {
+      "epoch": 0.7232095136290754,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.31566215536489e-05,
+      "loss": 1.2039,
+      "step": 10825
+    },
+    {
+      "epoch": 0.7235435595938001,
+      "grad_norm": 0.181640625,
+      "learning_rate": 4.3060725269805846e-05,
+      "loss": 1.1527,
+      "step": 10830
+    },
+    {
+      "epoch": 0.7238776055585249,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.296490640574826e-05,
+      "loss": 1.174,
+      "step": 10835
+    },
+    {
+      "epoch": 0.7242116515232496,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.2869165091760086e-05,
+      "loss": 1.2951,
+      "step": 10840
+    },
+    {
+      "epoch": 0.7245456974879744,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 4.2773501458019936e-05,
+      "loss": 1.2161,
+      "step": 10845
+    },
+    {
+      "epoch": 0.7248797434526991,
+      "grad_norm": 0.162109375,
+      "learning_rate": 4.267791563460074e-05,
+      "loss": 1.1598,
+      "step": 10850
+    },
+    {
+      "epoch": 0.7252137894174239,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 4.258240775146961e-05,
+      "loss": 1.2698,
+      "step": 10855
+    },
+    {
+      "epoch": 0.7255478353821486,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.248697793848768e-05,
+      "loss": 1.2178,
+      "step": 10860
+    },
+    {
+      "epoch": 0.7258818813468734,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 4.239162632540994e-05,
+      "loss": 1.1651,
+      "step": 10865
+    },
+    {
+      "epoch": 0.7262159273115981,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.229635304188507e-05,
+      "loss": 1.2835,
+      "step": 10870
+    },
+    {
+      "epoch": 0.7265499732763229,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.2201158217455296e-05,
+      "loss": 1.1627,
+      "step": 10875
+    },
+    {
+      "epoch": 0.7268840192410476,
+      "grad_norm": 0.15234375,
+      "learning_rate": 4.210604198155607e-05,
+      "loss": 1.2025,
+      "step": 10880
+    },
+    {
+      "epoch": 0.7272180652057724,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.201100446351597e-05,
+      "loss": 1.2049,
+      "step": 10885
+    },
+    {
+      "epoch": 0.727552111170497,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.1916045792556694e-05,
+      "loss": 1.2161,
+      "step": 10890
+    },
+    {
+      "epoch": 0.7278861571352218,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 4.182116609779259e-05,
+      "loss": 1.1577,
+      "step": 10895
+    },
+    {
+      "epoch": 0.7282202030999465,
+      "grad_norm": 0.189453125,
+      "learning_rate": 4.1726365508230616e-05,
+      "loss": 1.2406,
+      "step": 10900
+    },
+    {
+      "epoch": 0.7285542490646713,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 4.163164415277029e-05,
+      "loss": 1.173,
+      "step": 10905
+    },
+    {
+      "epoch": 0.728888295029396,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 4.1537002160203344e-05,
+      "loss": 1.2019,
+      "step": 10910
+    },
+    {
+      "epoch": 0.7292223409941208,
+      "grad_norm": 0.166015625,
+      "learning_rate": 4.1442439659213564e-05,
+      "loss": 1.2311,
+      "step": 10915
+    },
+    {
+      "epoch": 0.7295563869588455,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.134795677837663e-05,
+      "loss": 1.1683,
+      "step": 10920
+    },
+    {
+      "epoch": 0.7298904329235703,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 4.125355364616009e-05,
+      "loss": 1.1664,
+      "step": 10925
+    },
+    {
+      "epoch": 0.730224478888295,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.115923039092293e-05,
+      "loss": 1.1981,
+      "step": 10930
+    },
+    {
+      "epoch": 0.7305585248530198,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.1064987140915544e-05,
+      "loss": 1.1545,
+      "step": 10935
+    },
+    {
+      "epoch": 0.7308925708177445,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.097082402427962e-05,
+      "loss": 1.1627,
+      "step": 10940
+    },
+    {
+      "epoch": 0.7312266167824693,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.087674116904786e-05,
+      "loss": 1.2434,
+      "step": 10945
+    },
+    {
+      "epoch": 0.731560662747194,
+      "grad_norm": 0.18359375,
+      "learning_rate": 4.07827387031438e-05,
+      "loss": 1.1772,
+      "step": 10950
+    },
+    {
+      "epoch": 0.7318947087119188,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.068881675438165e-05,
+      "loss": 1.1721,
+      "step": 10955
+    },
+    {
+      "epoch": 0.7322287546766435,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 4.0594975450466255e-05,
+      "loss": 1.141,
+      "step": 10960
+    },
+    {
+      "epoch": 0.7325628006413683,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.050121491899266e-05,
+      "loss": 1.2159,
+      "step": 10965
+    },
+    {
+      "epoch": 0.732896846606093,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 4.040753528744623e-05,
+      "loss": 1.219,
+      "step": 10970
+    },
+    {
+      "epoch": 0.7332308925708177,
+      "grad_norm": 0.166015625,
+      "learning_rate": 4.0313936683202205e-05,
+      "loss": 1.1963,
+      "step": 10975
+    },
+    {
+      "epoch": 0.7335649385355425,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.0220419233525754e-05,
+      "loss": 1.1704,
+      "step": 10980
+    },
+    {
+      "epoch": 0.7338989845002672,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 4.0126983065571643e-05,
+      "loss": 1.1832,
+      "step": 10985
+    },
+    {
+      "epoch": 0.734233030464992,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.003362830638409e-05,
+      "loss": 1.2385,
+      "step": 10990
+    },
+    {
+      "epoch": 0.7345670764297167,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.9940355082896694e-05,
+      "loss": 1.1976,
+      "step": 10995
+    },
+    {
+      "epoch": 0.7349011223944415,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.984716352193222e-05,
+      "loss": 1.1941,
+      "step": 11000
+    },
+    {
+      "epoch": 0.7352351683591662,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.975405375020228e-05,
+      "loss": 1.1729,
+      "step": 11005
+    },
+    {
+      "epoch": 0.735569214323891,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.96610258943073e-05,
+      "loss": 1.2462,
+      "step": 11010
+    },
+    {
+      "epoch": 0.7359032602886157,
+      "grad_norm": 0.1875,
+      "learning_rate": 3.956808008073646e-05,
+      "loss": 1.2269,
+      "step": 11015
+    },
+    {
+      "epoch": 0.7362373062533405,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 3.9475216435867225e-05,
+      "loss": 1.1627,
+      "step": 11020
+    },
+    {
+      "epoch": 0.7365713522180652,
+      "grad_norm": 0.181640625,
+      "learning_rate": 3.938243508596539e-05,
+      "loss": 1.1939,
+      "step": 11025
+    },
+    {
+      "epoch": 0.73690539818279,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 3.9289736157184876e-05,
+      "loss": 1.2428,
+      "step": 11030
+    },
+    {
+      "epoch": 0.7372394441475147,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.9197119775567595e-05,
+      "loss": 1.1643,
+      "step": 11035
+    },
+    {
+      "epoch": 0.7375734901122395,
+      "grad_norm": 0.181640625,
+      "learning_rate": 3.910458606704309e-05,
+      "loss": 1.1663,
+      "step": 11040
+    },
+    {
+      "epoch": 0.7379075360769642,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.901213515742856e-05,
+      "loss": 1.2113,
+      "step": 11045
+    },
+    {
+      "epoch": 0.7382415820416889,
+      "grad_norm": 0.181640625,
+      "learning_rate": 3.891976717242861e-05,
+      "loss": 1.0892,
+      "step": 11050
+    },
+    {
+      "epoch": 0.7385756280064136,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.8827482237635105e-05,
+      "loss": 1.2396,
+      "step": 11055
+    },
+    {
+      "epoch": 0.7389096739711384,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.8735280478527035e-05,
+      "loss": 1.1426,
+      "step": 11060
+    },
+    {
+      "epoch": 0.7392437199358631,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.8643162020470224e-05,
+      "loss": 1.1233,
+      "step": 11065
+    },
+    {
+      "epoch": 0.7395777659005879,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.85511269887172e-05,
+      "loss": 1.2431,
+      "step": 11070
+    },
+    {
+      "epoch": 0.7399118118653126,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.8459175508407184e-05,
+      "loss": 1.283,
+      "step": 11075
+    },
+    {
+      "epoch": 0.7402458578300374,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.8367307704565706e-05,
+      "loss": 1.1919,
+      "step": 11080
+    },
+    {
+      "epoch": 0.7405799037947621,
+      "grad_norm": 0.18359375,
+      "learning_rate": 3.827552370210448e-05,
+      "loss": 1.2189,
+      "step": 11085
+    },
+    {
+      "epoch": 0.7409139497594869,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.818382362582137e-05,
+      "loss": 1.2248,
+      "step": 11090
+    },
+    {
+      "epoch": 0.7412479957242116,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.809220760040014e-05,
+      "loss": 1.1446,
+      "step": 11095
+    },
+    {
+      "epoch": 0.7415820416889364,
+      "grad_norm": 0.1640625,
+      "learning_rate": 3.8000675750410186e-05,
+      "loss": 1.1236,
+      "step": 11100
+    },
+    {
+      "epoch": 0.7419160876536611,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 3.7909228200306436e-05,
+      "loss": 1.1346,
+      "step": 11105
+    },
+    {
+      "epoch": 0.7422501336183859,
+      "grad_norm": 0.1640625,
+      "learning_rate": 3.7817865074429314e-05,
+      "loss": 1.1425,
+      "step": 11110
+    },
+    {
+      "epoch": 0.7425841795831106,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.7726586497004334e-05,
+      "loss": 1.1919,
+      "step": 11115
+    },
+    {
+      "epoch": 0.7429182255478354,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.7635392592142174e-05,
+      "loss": 1.1797,
+      "step": 11120
+    },
+    {
+      "epoch": 0.7432522715125601,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.7544283483838215e-05,
+      "loss": 1.2874,
+      "step": 11125
+    },
+    {
+      "epoch": 0.7435863174772849,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.745325929597272e-05,
+      "loss": 1.1107,
+      "step": 11130
+    },
+    {
+      "epoch": 0.7439203634420096,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.736232015231038e-05,
+      "loss": 1.1931,
+      "step": 11135
+    },
+    {
+      "epoch": 0.7442544094067344,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.7271466176500224e-05,
+      "loss": 1.1346,
+      "step": 11140
+    },
+    {
+      "epoch": 0.7445884553714591,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.718069749207559e-05,
+      "loss": 1.2507,
+      "step": 11145
+    },
+    {
+      "epoch": 0.7449225013361839,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.7090014222453794e-05,
+      "loss": 1.2236,
+      "step": 11150
+    },
+    {
+      "epoch": 0.7452565473009086,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.699941649093599e-05,
+      "loss": 1.2067,
+      "step": 11155
+    },
+    {
+      "epoch": 0.7455905932656334,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.6908904420707e-05,
+      "loss": 1.2222,
+      "step": 11160
+    },
+    {
+      "epoch": 0.7459246392303581,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 3.6818478134835285e-05,
+      "loss": 1.1647,
+      "step": 11165
+    },
+    {
+      "epoch": 0.7462586851950829,
+      "grad_norm": 0.177734375,
+      "learning_rate": 3.672813775627259e-05,
+      "loss": 1.2522,
+      "step": 11170
+    },
+    {
+      "epoch": 0.7465927311598076,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 3.663788340785379e-05,
+      "loss": 1.2151,
+      "step": 11175
+    },
+    {
+      "epoch": 0.7469267771245324,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.6547715212296906e-05,
+      "loss": 1.1831,
+      "step": 11180
+    },
+    {
+      "epoch": 0.7472608230892571,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.645763329220281e-05,
+      "loss": 1.2159,
+      "step": 11185
+    },
+    {
+      "epoch": 0.7475948690539819,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.636763777005499e-05,
+      "loss": 1.2279,
+      "step": 11190
+    },
+    {
+      "epoch": 0.7479289150187066,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.627772876821944e-05,
+      "loss": 1.1907,
+      "step": 11195
+    },
+    {
+      "epoch": 0.7482629609834314,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.618790640894465e-05,
+      "loss": 1.1535,
+      "step": 11200
+    },
+    {
+      "epoch": 0.7485970069481561,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 3.609817081436119e-05,
+      "loss": 1.1594,
+      "step": 11205
+    },
+    {
+      "epoch": 0.7489310529128808,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.600852210648164e-05,
+      "loss": 1.2243,
+      "step": 11210
+    },
+    {
+      "epoch": 0.7492650988776055,
+      "grad_norm": 0.205078125,
+      "learning_rate": 3.591896040720054e-05,
+      "loss": 1.2602,
+      "step": 11215
+    },
+    {
+      "epoch": 0.7495991448423303,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.5829485838294093e-05,
+      "loss": 1.2086,
+      "step": 11220
+    },
+    {
+      "epoch": 0.749933190807055,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.5740098521419985e-05,
+      "loss": 1.2441,
+      "step": 11225
+    },
+    {
+      "epoch": 0.7502672367717798,
+      "grad_norm": 0.18359375,
+      "learning_rate": 3.565079857811728e-05,
+      "loss": 1.1788,
+      "step": 11230
+    },
+    {
+      "epoch": 0.7506012827365045,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 3.556158612980624e-05,
+      "loss": 1.1962,
+      "step": 11235
+    },
+    {
+      "epoch": 0.7509353287012293,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 3.5472461297788185e-05,
+      "loss": 1.2109,
+      "step": 11240
+    },
+    {
+      "epoch": 0.751269374665954,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.538342420324534e-05,
+      "loss": 1.2655,
+      "step": 11245
+    },
+    {
+      "epoch": 0.7516034206306788,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.529447496724053e-05,
+      "loss": 1.1728,
+      "step": 11250
+    },
+    {
+      "epoch": 0.7519374665954035,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 3.5205613710717234e-05,
+      "loss": 1.121,
+      "step": 11255
+    },
+    {
+      "epoch": 0.7522715125601283,
+      "grad_norm": 0.18359375,
+      "learning_rate": 3.511684055449922e-05,
+      "loss": 1.1685,
+      "step": 11260
+    },
+    {
+      "epoch": 0.752605558524853,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 3.5028155619290495e-05,
+      "loss": 1.162,
+      "step": 11265
+    },
+    {
+      "epoch": 0.7529396044895778,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 3.493955902567505e-05,
+      "loss": 1.2662,
+      "step": 11270
+    },
+    {
+      "epoch": 0.7532736504543025,
+      "grad_norm": 0.16796875,
+      "learning_rate": 3.4851050894116946e-05,
+      "loss": 1.1836,
+      "step": 11275
+    },
+    {
+      "epoch": 0.7536076964190273,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.476263134495978e-05,
+      "loss": 1.18,
+      "step": 11280
+    },
+    {
+      "epoch": 0.753941742383752,
+      "grad_norm": 0.162109375,
+      "learning_rate": 3.467430049842678e-05,
+      "loss": 1.1368,
+      "step": 11285
+    },
+    {
+      "epoch": 0.7542757883484768,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.4586058474620495e-05,
+      "loss": 1.2558,
+      "step": 11290
+    },
+    {
+      "epoch": 0.7546098343132015,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.4497905393522835e-05,
+      "loss": 1.1602,
+      "step": 11295
+    },
+    {
+      "epoch": 0.7549438802779262,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.4409841374994634e-05,
+      "loss": 1.264,
+      "step": 11300
+    },
+    {
+      "epoch": 0.755277926242651,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 3.432186653877575e-05,
+      "loss": 1.2065,
+      "step": 11305
+    },
+    {
+      "epoch": 0.7556119722073757,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 3.423398100448466e-05,
+      "loss": 1.2051,
+      "step": 11310
+    },
+    {
+      "epoch": 0.7559460181721005,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.414618489161856e-05,
+      "loss": 1.112,
+      "step": 11315
+    },
+    {
+      "epoch": 0.7562800641368252,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.4058478319552936e-05,
+      "loss": 1.2844,
+      "step": 11320
+    },
+    {
+      "epoch": 0.75661411010155,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.397086140754153e-05,
+      "loss": 1.2591,
+      "step": 11325
+    },
+    {
+      "epoch": 0.7569481560662747,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.388333427471627e-05,
+      "loss": 1.1908,
+      "step": 11330
+    },
+    {
+      "epoch": 0.7572822020309995,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.3795897040087e-05,
+      "loss": 1.21,
+      "step": 11335
+    },
+    {
+      "epoch": 0.7576162479957242,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.3708549822541225e-05,
+      "loss": 1.1525,
+      "step": 11340
+    },
+    {
+      "epoch": 0.757950293960449,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.36212927408441e-05,
+      "loss": 1.2159,
+      "step": 11345
+    },
+    {
+      "epoch": 0.7582843399251737,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.3534125913638316e-05,
+      "loss": 1.2143,
+      "step": 11350
+    },
+    {
+      "epoch": 0.7586183858898985,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 3.344704945944372e-05,
+      "loss": 1.2032,
+      "step": 11355
+    },
+    {
+      "epoch": 0.7589524318546232,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.336006349665731e-05,
+      "loss": 1.1575,
+      "step": 11360
+    },
+    {
+      "epoch": 0.759286477819348,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.32731681435531e-05,
+      "loss": 1.1845,
+      "step": 11365
+    },
+    {
+      "epoch": 0.7596205237840727,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.3186363518281907e-05,
+      "loss": 1.1324,
+      "step": 11370
+    },
+    {
+      "epoch": 0.7599545697487974,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.30996497388711e-05,
+      "loss": 1.2672,
+      "step": 11375
+    },
+    {
+      "epoch": 0.7602886157135221,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 3.301302692322453e-05,
+      "loss": 1.2024,
+      "step": 11380
+    },
+    {
+      "epoch": 0.7606226616782469,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 3.292649518912251e-05,
+      "loss": 1.2042,
+      "step": 11385
+    },
+    {
+      "epoch": 0.7609567076429716,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.284005465422134e-05,
+      "loss": 1.1876,
+      "step": 11390
+    },
+    {
+      "epoch": 0.7612907536076964,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 3.275370543605337e-05,
+      "loss": 1.1774,
+      "step": 11395
+    },
+    {
+      "epoch": 0.7616247995724211,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.266744765202684e-05,
+      "loss": 1.1879,
+      "step": 11400
+    },
+    {
+      "epoch": 0.7619588455371459,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.2581281419425644e-05,
+      "loss": 1.1583,
+      "step": 11405
+    },
+    {
+      "epoch": 0.7622928915018706,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.2495206855409165e-05,
+      "loss": 1.1837,
+      "step": 11410
+    },
+    {
+      "epoch": 0.7626269374665954,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.2409224077012134e-05,
+      "loss": 1.1778,
+      "step": 11415
+    },
+    {
+      "epoch": 0.7629609834313201,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.232333320114457e-05,
+      "loss": 1.1836,
+      "step": 11420
+    },
+    {
+      "epoch": 0.7632950293960449,
+      "grad_norm": 0.173828125,
+      "learning_rate": 3.223753434459139e-05,
+      "loss": 1.194,
+      "step": 11425
+    },
+    {
+      "epoch": 0.7636290753607696,
+      "grad_norm": 0.1640625,
+      "learning_rate": 3.2151827624012574e-05,
+      "loss": 1.2349,
+      "step": 11430
+    },
+    {
+      "epoch": 0.7639631213254944,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.206621315594264e-05,
+      "loss": 1.2218,
+      "step": 11435
+    },
+    {
+      "epoch": 0.7642971672902191,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.1980691056790814e-05,
+      "loss": 1.2077,
+      "step": 11440
+    },
+    {
+      "epoch": 0.7646312132549439,
+      "grad_norm": 0.1640625,
+      "learning_rate": 3.189526144284066e-05,
+      "loss": 1.1235,
+      "step": 11445
+    },
+    {
+      "epoch": 0.7649652592196686,
+      "grad_norm": 0.18359375,
+      "learning_rate": 3.180992443025001e-05,
+      "loss": 1.2072,
+      "step": 11450
+    },
+    {
+      "epoch": 0.7652993051843934,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.17246801350507e-05,
+      "loss": 1.2245,
+      "step": 11455
+    },
+    {
+      "epoch": 0.7656333511491181,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.163952867314871e-05,
+      "loss": 1.2003,
+      "step": 11460
+    },
+    {
+      "epoch": 0.7659673971138429,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 3.155447016032361e-05,
+      "loss": 1.2126,
+      "step": 11465
+    },
+    {
+      "epoch": 0.7663014430785676,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.146950471222865e-05,
+      "loss": 1.2016,
+      "step": 11470
+    },
+    {
+      "epoch": 0.7666354890432924,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.138463244439048e-05,
+      "loss": 1.2201,
+      "step": 11475
+    },
+    {
+      "epoch": 0.7669695350080171,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.1299853472209186e-05,
+      "loss": 1.2002,
+      "step": 11480
+    },
+    {
+      "epoch": 0.7673035809727419,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.121516791095787e-05,
+      "loss": 1.2091,
+      "step": 11485
+    },
+    {
+      "epoch": 0.7676376269374666,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.113057587578271e-05,
+      "loss": 1.2976,
+      "step": 11490
+    },
+    {
+      "epoch": 0.7679716729021914,
+      "grad_norm": 0.16796875,
+      "learning_rate": 3.1046077481702654e-05,
+      "loss": 1.1261,
+      "step": 11495
+    },
+    {
+      "epoch": 0.7683057188669161,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 3.096167284360939e-05,
+      "loss": 1.2017,
+      "step": 11500
+    },
+    {
+      "epoch": 0.7686397648316409,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.087736207626709e-05,
+      "loss": 1.1789,
+      "step": 11505
+    },
+    {
+      "epoch": 0.7689738107963656,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.0793145294312255e-05,
+      "loss": 1.2321,
+      "step": 11510
+    },
+    {
+      "epoch": 0.7693078567610904,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.0709022612253656e-05,
+      "loss": 1.2757,
+      "step": 11515
+    },
+    {
+      "epoch": 0.7696419027258151,
+      "grad_norm": 0.181640625,
+      "learning_rate": 3.062499414447215e-05,
+      "loss": 1.2314,
+      "step": 11520
+    },
+    {
+      "epoch": 0.7699759486905399,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.054106000522039e-05,
+      "loss": 1.1666,
+      "step": 11525
+    },
+    {
+      "epoch": 0.7703099946552646,
+      "grad_norm": 0.16796875,
+      "learning_rate": 3.0457220308622782e-05,
+      "loss": 1.1709,
+      "step": 11530
+    },
+    {
+      "epoch": 0.7706440406199893,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.0373475168675435e-05,
+      "loss": 1.2248,
+      "step": 11535
+    },
+    {
+      "epoch": 0.770978086584714,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.0289824699245784e-05,
+      "loss": 1.2131,
+      "step": 11540
+    },
+    {
+      "epoch": 0.7713121325494388,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 3.0206269014072518e-05,
+      "loss": 1.2146,
+      "step": 11545
+    },
+    {
+      "epoch": 0.7716461785141635,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.0122808226765554e-05,
+      "loss": 1.1952,
+      "step": 11550
+    },
+    {
+      "epoch": 0.7719802244788883,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.003944245080573e-05,
+      "loss": 1.1605,
+      "step": 11555
+    },
+    {
+      "epoch": 0.772314270443613,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.9956171799544686e-05,
+      "loss": 1.2247,
+      "step": 11560
+    },
+    {
+      "epoch": 0.7726483164083378,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.9872996386204678e-05,
+      "loss": 1.1881,
+      "step": 11565
+    },
+    {
+      "epoch": 0.7729823623730625,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.9789916323878597e-05,
+      "loss": 1.2223,
+      "step": 11570
+    },
+    {
+      "epoch": 0.7733164083377873,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.970693172552953e-05,
+      "loss": 1.2298,
+      "step": 11575
+    },
+    {
+      "epoch": 0.773650454302512,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.9624042703990896e-05,
+      "loss": 1.2414,
+      "step": 11580
+    },
+    {
+      "epoch": 0.7739845002672368,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.9541249371966064e-05,
+      "loss": 1.1649,
+      "step": 11585
+    },
+    {
+      "epoch": 0.7743185462319615,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.945855184202837e-05,
+      "loss": 1.1757,
+      "step": 11590
+    },
+    {
+      "epoch": 0.7746525921966863,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.937595022662083e-05,
+      "loss": 1.1983,
+      "step": 11595
+    },
+    {
+      "epoch": 0.774986638161411,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.9293444638056045e-05,
+      "loss": 1.27,
+      "step": 11600
+    },
+    {
+      "epoch": 0.7753206841261358,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.921103518851609e-05,
+      "loss": 1.2107,
+      "step": 11605
+    },
+    {
+      "epoch": 0.7756547300908605,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.9128721990052345e-05,
+      "loss": 1.167,
+      "step": 11610
+    },
+    {
+      "epoch": 0.7759887760555852,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.9046505154585235e-05,
+      "loss": 1.2401,
+      "step": 11615
+    },
+    {
+      "epoch": 0.77632282202031,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.8964384793904188e-05,
+      "loss": 1.1416,
+      "step": 11620
+    },
+    {
+      "epoch": 0.7766568679850347,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.8882361019667502e-05,
+      "loss": 1.2621,
+      "step": 11625
+    },
+    {
+      "epoch": 0.7769909139497595,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.8800433943402115e-05,
+      "loss": 1.2146,
+      "step": 11630
+    },
+    {
+      "epoch": 0.7773249599144842,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.8718603676503475e-05,
+      "loss": 1.2552,
+      "step": 11635
+    },
+    {
+      "epoch": 0.777659005879209,
+      "grad_norm": 0.203125,
+      "learning_rate": 2.8636870330235356e-05,
+      "loss": 1.2535,
+      "step": 11640
+    },
+    {
+      "epoch": 0.7779930518439337,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.8555234015729904e-05,
+      "loss": 1.1534,
+      "step": 11645
+    },
+    {
+      "epoch": 0.7783270978086585,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.8473694843987198e-05,
+      "loss": 1.2312,
+      "step": 11650
+    },
+    {
+      "epoch": 0.7786611437733832,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.839225292587525e-05,
+      "loss": 1.1423,
+      "step": 11655
+    },
+    {
+      "epoch": 0.778995189738108,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.831090837212984e-05,
+      "loss": 1.2034,
+      "step": 11660
+    },
+    {
+      "epoch": 0.7793292357028327,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 2.8229661293354427e-05,
+      "loss": 1.1359,
+      "step": 11665
+    },
+    {
+      "epoch": 0.7796632816675575,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.8148511800019827e-05,
+      "loss": 1.2567,
+      "step": 11670
+    },
+    {
+      "epoch": 0.7799973276322822,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.8067460002464252e-05,
+      "loss": 1.2415,
+      "step": 11675
+    },
+    {
+      "epoch": 0.780331373597007,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.7986506010893088e-05,
+      "loss": 1.1646,
+      "step": 11680
+    },
+    {
+      "epoch": 0.7806654195617317,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 2.7905649935378673e-05,
+      "loss": 1.1926,
+      "step": 11685
+    },
+    {
+      "epoch": 0.7809994655264565,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.7824891885860227e-05,
+      "loss": 1.2113,
+      "step": 11690
+    },
+    {
+      "epoch": 0.7813335114911811,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.7744231972143687e-05,
+      "loss": 1.1887,
+      "step": 11695
+    },
+    {
+      "epoch": 0.7816675574559059,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.7663670303901566e-05,
+      "loss": 1.2085,
+      "step": 11700
+    },
+    {
+      "epoch": 0.7820016034206306,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.758320699067284e-05,
+      "loss": 1.2066,
+      "step": 11705
+    },
+    {
+      "epoch": 0.7823356493853554,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.7502842141862672e-05,
+      "loss": 1.2673,
+      "step": 11710
+    },
+    {
+      "epoch": 0.7826696953500801,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.742257586674233e-05,
+      "loss": 1.2183,
+      "step": 11715
+    },
+    {
+      "epoch": 0.7830037413148049,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.7342408274449184e-05,
+      "loss": 1.166,
+      "step": 11720
+    },
+    {
+      "epoch": 0.7833377872795296,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 2.7262339473986286e-05,
+      "loss": 1.1372,
+      "step": 11725
+    },
+    {
+      "epoch": 0.7836718332442544,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.71823695742224e-05,
+      "loss": 1.1694,
+      "step": 11730
+    },
+    {
+      "epoch": 0.7840058792089791,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.710249868389185e-05,
+      "loss": 1.219,
+      "step": 11735
+    },
+    {
+      "epoch": 0.7843399251737039,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.7022726911594363e-05,
+      "loss": 1.2113,
+      "step": 11740
+    },
+    {
+      "epoch": 0.7846739711384286,
+      "grad_norm": 0.158203125,
+      "learning_rate": 2.6943054365794818e-05,
+      "loss": 1.1246,
+      "step": 11745
+    },
+    {
+      "epoch": 0.7850080171031534,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.6863481154823168e-05,
+      "loss": 1.1696,
+      "step": 11750
+    },
+    {
+      "epoch": 0.7853420630678781,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.678400738687442e-05,
+      "loss": 1.1715,
+      "step": 11755
+    },
+    {
+      "epoch": 0.7856761090326029,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.6704633170008232e-05,
+      "loss": 1.1638,
+      "step": 11760
+    },
+    {
+      "epoch": 0.7860101549973276,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.662535861214902e-05,
+      "loss": 1.2395,
+      "step": 11765
+    },
+    {
+      "epoch": 0.7863442009620524,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.654618382108558e-05,
+      "loss": 1.2303,
+      "step": 11770
+    },
+    {
+      "epoch": 0.7866782469267771,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.6467108904471184e-05,
+      "loss": 1.2723,
+      "step": 11775
+    },
+    {
+      "epoch": 0.7870122928915019,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.6388133969823193e-05,
+      "loss": 1.1856,
+      "step": 11780
+    },
+    {
+      "epoch": 0.7873463388562266,
+      "grad_norm": 0.1640625,
+      "learning_rate": 2.6309259124523046e-05,
+      "loss": 1.2234,
+      "step": 11785
+    },
+    {
+      "epoch": 0.7876803848209514,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.6230484475816132e-05,
+      "loss": 1.1786,
+      "step": 11790
+    },
+    {
+      "epoch": 0.7880144307856761,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.6151810130811638e-05,
+      "loss": 1.2053,
+      "step": 11795
+    },
+    {
+      "epoch": 0.7883484767504009,
+      "grad_norm": 0.1640625,
+      "learning_rate": 2.6073236196482263e-05,
+      "loss": 1.1418,
+      "step": 11800
+    },
+    {
+      "epoch": 0.7886825227151256,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.599476277966423e-05,
+      "loss": 1.1912,
+      "step": 11805
+    },
+    {
+      "epoch": 0.7890165686798504,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 2.591638998705711e-05,
+      "loss": 1.1552,
+      "step": 11810
+    },
+    {
+      "epoch": 0.7893506146445751,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 2.583811792522365e-05,
+      "loss": 1.148,
+      "step": 11815
+    },
+    {
+      "epoch": 0.7896846606092999,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 2.5759946700589556e-05,
+      "loss": 1.2014,
+      "step": 11820
+    },
+    {
+      "epoch": 0.7900187065740246,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.568187641944354e-05,
+      "loss": 1.1741,
+      "step": 11825
+    },
+    {
+      "epoch": 0.7903527525387494,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.5603907187937038e-05,
+      "loss": 1.1774,
+      "step": 11830
+    },
+    {
+      "epoch": 0.7906867985034741,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.5526039112084044e-05,
+      "loss": 1.2101,
+      "step": 11835
+    },
+    {
+      "epoch": 0.7910208444681989,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.5448272297761e-05,
+      "loss": 1.1802,
+      "step": 11840
+    },
+    {
+      "epoch": 0.7913548904329236,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.5370606850706757e-05,
+      "loss": 1.194,
+      "step": 11845
+    },
+    {
+      "epoch": 0.7916889363976484,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.5293042876522245e-05,
+      "loss": 1.2415,
+      "step": 11850
+    },
+    {
+      "epoch": 0.7920229823623731,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.521558048067042e-05,
+      "loss": 1.2278,
+      "step": 11855
+    },
+    {
+      "epoch": 0.7923570283270978,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 2.5138219768476203e-05,
+      "loss": 1.2145,
+      "step": 11860
+    },
+    {
+      "epoch": 0.7926910742918225,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.5060960845126235e-05,
+      "loss": 1.179,
+      "step": 11865
+    },
+    {
+      "epoch": 0.7930251202565473,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 2.4983803815668694e-05,
+      "loss": 1.1905,
+      "step": 11870
+    },
+    {
+      "epoch": 0.793359166221272,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.4906748785013267e-05,
+      "loss": 1.1982,
+      "step": 11875
+    },
+    {
+      "epoch": 0.7936932121859968,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.4829795857930904e-05,
+      "loss": 1.2422,
+      "step": 11880
+    },
+    {
+      "epoch": 0.7940272581507215,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.4752945139053785e-05,
+      "loss": 1.2127,
+      "step": 11885
+    },
+    {
+      "epoch": 0.7943613041154463,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.4676196732875144e-05,
+      "loss": 1.1725,
+      "step": 11890
+    },
+    {
+      "epoch": 0.794695350080171,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.4599550743749e-05,
+      "loss": 1.2145,
+      "step": 11895
+    },
+    {
+      "epoch": 0.7950293960448958,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.4523007275890152e-05,
+      "loss": 1.1748,
+      "step": 11900
+    },
+    {
+      "epoch": 0.7953634420096205,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.4446566433374065e-05,
+      "loss": 1.2528,
+      "step": 11905
+    },
+    {
+      "epoch": 0.7956974879743453,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.4370228320136613e-05,
+      "loss": 1.2329,
+      "step": 11910
+    },
+    {
+      "epoch": 0.79603153393907,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.429399303997394e-05,
+      "loss": 1.2073,
+      "step": 11915
+    },
+    {
+      "epoch": 0.7963655799037948,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.4217860696542482e-05,
+      "loss": 1.1722,
+      "step": 11920
+    },
+    {
+      "epoch": 0.7966996258685195,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.414183139335866e-05,
+      "loss": 1.24,
+      "step": 11925
+    },
+    {
+      "epoch": 0.7970336718332443,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.406590523379877e-05,
+      "loss": 1.1457,
+      "step": 11930
+    },
+    {
+      "epoch": 0.797367717797969,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.399008232109885e-05,
+      "loss": 1.2234,
+      "step": 11935
+    },
+    {
+      "epoch": 0.7977017637626937,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.3914362758354658e-05,
+      "loss": 1.2298,
+      "step": 11940
+    },
+    {
+      "epoch": 0.7980358097274185,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.383874664852127e-05,
+      "loss": 1.1725,
+      "step": 11945
+    },
+    {
+      "epoch": 0.7983698556921432,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.3763234094413277e-05,
+      "loss": 1.171,
+      "step": 11950
+    },
+    {
+      "epoch": 0.798703901656868,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.3687825198704296e-05,
+      "loss": 1.1286,
+      "step": 11955
+    },
+    {
+      "epoch": 0.7990379476215927,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.3612520063927145e-05,
+      "loss": 1.22,
+      "step": 11960
+    },
+    {
+      "epoch": 0.7993719935863175,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.353731879247345e-05,
+      "loss": 1.1227,
+      "step": 11965
+    },
+    {
+      "epoch": 0.7997060395510422,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.346222148659365e-05,
+      "loss": 1.1955,
+      "step": 11970
+    },
+    {
+      "epoch": 0.800040085515767,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 2.3387228248396842e-05,
+      "loss": 1.1408,
+      "step": 11975
+    },
+    {
+      "epoch": 0.8003741314804917,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.3312339179850652e-05,
+      "loss": 1.1877,
+      "step": 11980
+    },
+    {
+      "epoch": 0.8007081774452165,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.3237554382781002e-05,
+      "loss": 1.1573,
+      "step": 11985
+    },
+    {
+      "epoch": 0.8010422234099412,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.316287395887202e-05,
+      "loss": 1.1767,
+      "step": 11990
+    },
+    {
+      "epoch": 0.801376269374666,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.3088298009666033e-05,
+      "loss": 1.3085,
+      "step": 11995
+    },
+    {
+      "epoch": 0.8017103153393907,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.3013826636563198e-05,
+      "loss": 1.1452,
+      "step": 12000
+    },
+    {
+      "epoch": 0.8020443613041155,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.2939459940821518e-05,
+      "loss": 1.2353,
+      "step": 12005
+    },
+    {
+      "epoch": 0.8023784072688402,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.2865198023556698e-05,
+      "loss": 1.2847,
+      "step": 12010
+    },
+    {
+      "epoch": 0.802712453233565,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.2791040985741974e-05,
+      "loss": 1.2256,
+      "step": 12015
+    },
+    {
+      "epoch": 0.8030464991982896,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.271698892820794e-05,
+      "loss": 1.1917,
+      "step": 12020
+    },
+    {
+      "epoch": 0.8033805451630144,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 2.264304195164243e-05,
+      "loss": 1.2397,
+      "step": 12025
+    },
+    {
+      "epoch": 0.8037145911277391,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.2569200156590507e-05,
+      "loss": 1.1771,
+      "step": 12030
+    },
+    {
+      "epoch": 0.8040486370924639,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.2495463643454085e-05,
+      "loss": 1.1452,
+      "step": 12035
+    },
+    {
+      "epoch": 0.8043826830571886,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 2.2421832512492057e-05,
+      "loss": 1.2384,
+      "step": 12040
+    },
+    {
+      "epoch": 0.8047167290219134,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.23483068638199e-05,
+      "loss": 1.1587,
+      "step": 12045
+    },
+    {
+      "epoch": 0.8050507749866381,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.22748867974098e-05,
+      "loss": 1.2088,
+      "step": 12050
+    },
+    {
+      "epoch": 0.8053848209513629,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.220157241309028e-05,
+      "loss": 1.2517,
+      "step": 12055
+    },
+    {
+      "epoch": 0.8057188669160876,
+      "grad_norm": 0.162109375,
+      "learning_rate": 2.2128363810546205e-05,
+      "loss": 1.0946,
+      "step": 12060
+    },
+    {
+      "epoch": 0.8060529128808124,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.205526108931857e-05,
+      "loss": 1.1852,
+      "step": 12065
+    },
+    {
+      "epoch": 0.8063869588455371,
+      "grad_norm": 0.1640625,
+      "learning_rate": 2.1982264348804525e-05,
+      "loss": 1.1549,
+      "step": 12070
+    },
+    {
+      "epoch": 0.8067210048102619,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.1909373688257008e-05,
+      "loss": 1.2467,
+      "step": 12075
+    },
+    {
+      "epoch": 0.8070550507749866,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.183658920678474e-05,
+      "loss": 1.1084,
+      "step": 12080
+    },
+    {
+      "epoch": 0.8073890967397114,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.1763911003352055e-05,
+      "loss": 1.2701,
+      "step": 12085
+    },
+    {
+      "epoch": 0.8077231427044361,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.1691339176778856e-05,
+      "loss": 1.1389,
+      "step": 12090
+    },
+    {
+      "epoch": 0.8080571886691609,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 2.161887382574035e-05,
+      "loss": 1.2096,
+      "step": 12095
+    },
+    {
+      "epoch": 0.8083912346338856,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.1546515048766914e-05,
+      "loss": 1.2264,
+      "step": 12100
+    },
+    {
+      "epoch": 0.8087252805986104,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.1474262944244196e-05,
+      "loss": 1.1749,
+      "step": 12105
+    },
+    {
+      "epoch": 0.8090593265633351,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.140211761041262e-05,
+      "loss": 1.2378,
+      "step": 12110
+    },
+    {
+      "epoch": 0.8093933725280599,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.133007914536753e-05,
+      "loss": 1.1184,
+      "step": 12115
+    },
+    {
+      "epoch": 0.8097274184927846,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.125814764705889e-05,
+      "loss": 1.1524,
+      "step": 12120
+    },
+    {
+      "epoch": 0.8100614644575094,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.1186323213291316e-05,
+      "loss": 1.2009,
+      "step": 12125
+    },
+    {
+      "epoch": 0.8103955104222341,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.1114605941723777e-05,
+      "loss": 1.1628,
+      "step": 12130
+    },
+    {
+      "epoch": 0.8107295563869589,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 2.104299592986958e-05,
+      "loss": 1.2373,
+      "step": 12135
+    },
+    {
+      "epoch": 0.8110636023516836,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.0971493275096133e-05,
+      "loss": 1.2578,
+      "step": 12140
+    },
+    {
+      "epoch": 0.8113976483164084,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.0900098074624952e-05,
+      "loss": 1.1665,
+      "step": 12145
+    },
+    {
+      "epoch": 0.8117316942811331,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.08288104255314e-05,
+      "loss": 1.1408,
+      "step": 12150
+    },
+    {
+      "epoch": 0.8120657402458579,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 2.0757630424744568e-05,
+      "loss": 1.1506,
+      "step": 12155
+    },
+    {
+      "epoch": 0.8123997862105826,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.0686558169047256e-05,
+      "loss": 1.2123,
+      "step": 12160
+    },
+    {
+      "epoch": 0.8127338321753074,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.0615593755075734e-05,
+      "loss": 1.1628,
+      "step": 12165
+    },
+    {
+      "epoch": 0.8130678781400321,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.0544737279319636e-05,
+      "loss": 1.2352,
+      "step": 12170
+    },
+    {
+      "epoch": 0.8134019241047569,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 2.0473988838121783e-05,
+      "loss": 1.1936,
+      "step": 12175
+    },
+    {
+      "epoch": 0.8137359700694815,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.0403348527678222e-05,
+      "loss": 1.1521,
+      "step": 12180
+    },
+    {
+      "epoch": 0.8140700160342063,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.0332816444037873e-05,
+      "loss": 1.2135,
+      "step": 12185
+    },
+    {
+      "epoch": 0.814404061998931,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.0262392683102493e-05,
+      "loss": 1.2112,
+      "step": 12190
+    },
+    {
+      "epoch": 0.8147381079636558,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.0192077340626636e-05,
+      "loss": 1.1762,
+      "step": 12195
+    },
+    {
+      "epoch": 0.8150721539283805,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.012187051221742e-05,
+      "loss": 1.1734,
+      "step": 12200
+    },
+    {
+      "epoch": 0.8154061998931053,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.005177229333437e-05,
+      "loss": 1.1807,
+      "step": 12205
+    },
+    {
+      "epoch": 0.81574024585783,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 1.998178277928934e-05,
+      "loss": 1.1999,
+      "step": 12210
+    },
+    {
+      "epoch": 0.8160742918225548,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 1.9911902065246447e-05,
+      "loss": 1.1723,
+      "step": 12215
+    },
+    {
+      "epoch": 0.8164083377872795,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.9842130246221768e-05,
+      "loss": 1.1958,
+      "step": 12220
+    },
+    {
+      "epoch": 0.8167423837520043,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.977246741708344e-05,
+      "loss": 1.2349,
+      "step": 12225
+    },
+    {
+      "epoch": 0.817076429716729,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.9702913672551292e-05,
+      "loss": 1.1717,
+      "step": 12230
+    },
+    {
+      "epoch": 0.8174104756814538,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.9633469107196932e-05,
+      "loss": 1.2384,
+      "step": 12235
+    },
+    {
+      "epoch": 0.8177445216461785,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.956413381544344e-05,
+      "loss": 1.1851,
+      "step": 12240
+    },
+    {
+      "epoch": 0.8180785676109033,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.9494907891565316e-05,
+      "loss": 1.2535,
+      "step": 12245
+    },
+    {
+      "epoch": 0.818412613575628,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.942579142968842e-05,
+      "loss": 1.1402,
+      "step": 12250
+    },
+    {
+      "epoch": 0.8187466595403528,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.9356784523789772e-05,
+      "loss": 1.2167,
+      "step": 12255
+    },
+    {
+      "epoch": 0.8190807055050775,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.928788726769737e-05,
+      "loss": 1.1381,
+      "step": 12260
+    },
+    {
+      "epoch": 0.8194147514698022,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.9219099755090107e-05,
+      "loss": 1.1904,
+      "step": 12265
+    },
+    {
+      "epoch": 0.819748797434527,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.915042207949779e-05,
+      "loss": 1.2316,
+      "step": 12270
+    },
+    {
+      "epoch": 0.8200828433992517,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 1.908185433430074e-05,
+      "loss": 1.1822,
+      "step": 12275
+    },
+    {
+      "epoch": 0.8204168893639765,
+      "grad_norm": 0.19921875,
+      "learning_rate": 1.901339661272985e-05,
+      "loss": 1.2357,
+      "step": 12280
+    },
+    {
+      "epoch": 0.8207509353287012,
+      "grad_norm": 0.19140625,
+      "learning_rate": 1.8945049007866446e-05,
+      "loss": 1.1806,
+      "step": 12285
+    },
+    {
+      "epoch": 0.821084981293426,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.887681161264214e-05,
+      "loss": 1.1728,
+      "step": 12290
+    },
+    {
+      "epoch": 0.8214190272581507,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.880868451983865e-05,
+      "loss": 1.2062,
+      "step": 12295
+    },
+    {
+      "epoch": 0.8217530732228755,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.874066782208771e-05,
+      "loss": 1.1982,
+      "step": 12300
+    },
+    {
+      "epoch": 0.8220871191876002,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.8672761611870958e-05,
+      "loss": 1.2017,
+      "step": 12305
+    },
+    {
+      "epoch": 0.822421165152325,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.8604965981519827e-05,
+      "loss": 1.1544,
+      "step": 12310
+    },
+    {
+      "epoch": 0.8227552111170497,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.8537281023215436e-05,
+      "loss": 1.0842,
+      "step": 12315
+    },
+    {
+      "epoch": 0.8230892570817745,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 1.846970682898833e-05,
+      "loss": 1.2518,
+      "step": 12320
+    },
+    {
+      "epoch": 0.8234233030464992,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.8402243490718474e-05,
+      "loss": 1.2632,
+      "step": 12325
+    },
+    {
+      "epoch": 0.823757349011224,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.8334891100135166e-05,
+      "loss": 1.2021,
+      "step": 12330
+    },
+    {
+      "epoch": 0.8240913949759487,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.8267649748816772e-05,
+      "loss": 1.1112,
+      "step": 12335
+    },
+    {
+      "epoch": 0.8244254409406734,
+      "grad_norm": 0.16015625,
+      "learning_rate": 1.820051952819072e-05,
+      "loss": 1.2062,
+      "step": 12340
+    },
+    {
+      "epoch": 0.8247594869053981,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.8133500529533308e-05,
+      "loss": 1.2504,
+      "step": 12345
+    },
+    {
+      "epoch": 0.8250935328701229,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.806659284396969e-05,
+      "loss": 1.2482,
+      "step": 12350
+    },
+    {
+      "epoch": 0.8254275788348476,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.799979656247355e-05,
+      "loss": 1.1268,
+      "step": 12355
+    },
+    {
+      "epoch": 0.8257616247995724,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 1.793311177586714e-05,
+      "loss": 1.1863,
+      "step": 12360
+    },
+    {
+      "epoch": 0.8260956707642971,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 1.786653857482118e-05,
+      "loss": 1.2579,
+      "step": 12365
+    },
+    {
+      "epoch": 0.8264297167290219,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.780007704985457e-05,
+      "loss": 1.1932,
+      "step": 12370
+    },
+    {
+      "epoch": 0.8267637626937466,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.77337272913344e-05,
+      "loss": 1.2517,
+      "step": 12375
+    },
+    {
+      "epoch": 0.8270978086584714,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.766748938947581e-05,
+      "loss": 1.2773,
+      "step": 12380
+    },
+    {
+      "epoch": 0.8274318546231961,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.760136343434188e-05,
+      "loss": 1.1525,
+      "step": 12385
+    },
+    {
+      "epoch": 0.8277659005879209,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.7535349515843392e-05,
+      "loss": 1.1876,
+      "step": 12390
+    },
+    {
+      "epoch": 0.8280999465526456,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.746944772373883e-05,
+      "loss": 1.1088,
+      "step": 12395
+    },
+    {
+      "epoch": 0.8284339925173704,
+      "grad_norm": 0.19140625,
+      "learning_rate": 1.740365814763427e-05,
+      "loss": 1.2222,
+      "step": 12400
+    },
+    {
+      "epoch": 0.8287680384820951,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.733798087698313e-05,
+      "loss": 1.2101,
+      "step": 12405
+    },
+    {
+      "epoch": 0.8291020844468199,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.727241600108619e-05,
+      "loss": 1.2127,
+      "step": 12410
+    },
+    {
+      "epoch": 0.8294361304115446,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.7206963609091352e-05,
+      "loss": 1.1727,
+      "step": 12415
+    },
+    {
+      "epoch": 0.8297701763762694,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.7141623789993655e-05,
+      "loss": 1.1811,
+      "step": 12420
+    },
+    {
+      "epoch": 0.8301042223409941,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.7076396632634994e-05,
+      "loss": 1.179,
+      "step": 12425
+    },
+    {
+      "epoch": 0.8304382683057189,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.7011282225704074e-05,
+      "loss": 1.1736,
+      "step": 12430
+    },
+    {
+      "epoch": 0.8307723142704436,
+      "grad_norm": 0.16015625,
+      "learning_rate": 1.694628065773638e-05,
+      "loss": 1.1563,
+      "step": 12435
+    },
+    {
+      "epoch": 0.8311063602351684,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.6881392017113917e-05,
+      "loss": 1.2117,
+      "step": 12440
+    },
+    {
+      "epoch": 0.8314404061998931,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.6816616392065142e-05,
+      "loss": 1.1943,
+      "step": 12445
+    },
+    {
+      "epoch": 0.8317744521646179,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.6751953870664817e-05,
+      "loss": 1.2046,
+      "step": 12450
+    },
+    {
+      "epoch": 0.8321084981293426,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.6687404540833996e-05,
+      "loss": 1.1596,
+      "step": 12455
+    },
+    {
+      "epoch": 0.8324425440940674,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.6622968490339773e-05,
+      "loss": 1.265,
+      "step": 12460
+    },
+    {
+      "epoch": 0.8327765900587921,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.6558645806795193e-05,
+      "loss": 1.2546,
+      "step": 12465
+    },
+    {
+      "epoch": 0.8331106360235169,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.6494436577659222e-05,
+      "loss": 1.1452,
+      "step": 12470
+    },
+    {
+      "epoch": 0.8334446819882416,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.643034089023655e-05,
+      "loss": 1.2115,
+      "step": 12475
+    },
+    {
+      "epoch": 0.8337787279529664,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.6366358831677454e-05,
+      "loss": 1.156,
+      "step": 12480
+    },
+    {
+      "epoch": 0.8341127739176911,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.6302490488977705e-05,
+      "loss": 1.1572,
+      "step": 12485
+    },
+    {
+      "epoch": 0.8344468198824159,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.623873594897848e-05,
+      "loss": 1.1823,
+      "step": 12490
+    },
+    {
+      "epoch": 0.8347808658471406,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.6175095298366217e-05,
+      "loss": 1.1672,
+      "step": 12495
+    },
+    {
+      "epoch": 0.8351149118118654,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.6111568623672533e-05,
+      "loss": 1.1418,
+      "step": 12500
+    },
+    {
+      "epoch": 0.83544895777659,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.6048156011274018e-05,
+      "loss": 1.2192,
+      "step": 12505
+    },
+    {
+      "epoch": 0.8357830037413148,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.598485754739215e-05,
+      "loss": 1.2165,
+      "step": 12510
+    },
+    {
+      "epoch": 0.8361170497060395,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.59216733180933e-05,
+      "loss": 1.2849,
+      "step": 12515
+    },
+    {
+      "epoch": 0.8364510956707643,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.585860340928844e-05,
+      "loss": 1.2269,
+      "step": 12520
+    },
+    {
+      "epoch": 0.836785141635489,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.579564790673308e-05,
+      "loss": 1.2391,
+      "step": 12525
+    },
+    {
+      "epoch": 0.8371191876002138,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.5732806896027287e-05,
+      "loss": 1.1784,
+      "step": 12530
+    },
+    {
+      "epoch": 0.8374532335649385,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.5670080462615345e-05,
+      "loss": 1.1549,
+      "step": 12535
+    },
+    {
+      "epoch": 0.8377872795296633,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.5607468691785776e-05,
+      "loss": 1.1978,
+      "step": 12540
+    },
+    {
+      "epoch": 0.838121325494388,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.554497166867118e-05,
+      "loss": 1.2579,
+      "step": 12545
+    },
+    {
+      "epoch": 0.8384553714591128,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.5482589478248222e-05,
+      "loss": 1.2783,
+      "step": 12550
+    },
+    {
+      "epoch": 0.8387894174238375,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.5420322205337333e-05,
+      "loss": 1.2231,
+      "step": 12555
+    },
+    {
+      "epoch": 0.8391234633885623,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.5358169934602706e-05,
+      "loss": 1.1347,
+      "step": 12560
+    },
+    {
+      "epoch": 0.839457509353287,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 1.5296132750552207e-05,
+      "loss": 1.1165,
+      "step": 12565
+    },
+    {
+      "epoch": 0.8397915553180118,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 1.5234210737537225e-05,
+      "loss": 1.2383,
+      "step": 12570
+    },
+    {
+      "epoch": 0.8401256012827365,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.5172403979752492e-05,
+      "loss": 1.1735,
+      "step": 12575
+    },
+    {
+      "epoch": 0.8404596472474613,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.5110712561236062e-05,
+      "loss": 1.2072,
+      "step": 12580
+    },
+    {
+      "epoch": 0.840793693212186,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.5049136565869205e-05,
+      "loss": 1.1671,
+      "step": 12585
+    },
+    {
+      "epoch": 0.8411277391769107,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 1.4987676077376156e-05,
+      "loss": 1.2047,
+      "step": 12590
+    },
+    {
+      "epoch": 0.8414617851416355,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.4926331179324205e-05,
+      "loss": 1.1631,
+      "step": 12595
+    },
+    {
+      "epoch": 0.8417958311063602,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.4865101955123362e-05,
+      "loss": 1.1725,
+      "step": 12600
+    },
+    {
+      "epoch": 0.842129877071085,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.4803988488026487e-05,
+      "loss": 1.2299,
+      "step": 12605
+    },
+    {
+      "epoch": 0.8424639230358097,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.4742990861128924e-05,
+      "loss": 1.1771,
+      "step": 12610
+    },
+    {
+      "epoch": 0.8427979690005345,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.4682109157368561e-05,
+      "loss": 1.2413,
+      "step": 12615
+    },
+    {
+      "epoch": 0.8431320149652592,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.4621343459525671e-05,
+      "loss": 1.1604,
+      "step": 12620
+    },
+    {
+      "epoch": 0.843466060929984,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.4560693850222828e-05,
+      "loss": 1.1893,
+      "step": 12625
+    },
+    {
+      "epoch": 0.8438001068947087,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.450016041192469e-05,
+      "loss": 1.1732,
+      "step": 12630
+    },
+    {
+      "epoch": 0.8441341528594335,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.4439743226937975e-05,
+      "loss": 1.1983,
+      "step": 12635
+    },
+    {
+      "epoch": 0.8444681988241582,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.437944237741139e-05,
+      "loss": 1.2534,
+      "step": 12640
+    },
+    {
+      "epoch": 0.844802244788883,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.4319257945335408e-05,
+      "loss": 1.2628,
+      "step": 12645
+    },
+    {
+      "epoch": 0.8451362907536077,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.425919001254219e-05,
+      "loss": 1.1942,
+      "step": 12650
+    },
+    {
+      "epoch": 0.8454703367183325,
+      "grad_norm": 0.158203125,
+      "learning_rate": 1.419923866070556e-05,
+      "loss": 1.1969,
+      "step": 12655
+    },
+    {
+      "epoch": 0.8458043826830572,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.4139403971340815e-05,
+      "loss": 1.2407,
+      "step": 12660
+    },
+    {
+      "epoch": 0.8461384286477819,
+      "grad_norm": 0.18359375,
+      "learning_rate": 1.4079686025804584e-05,
+      "loss": 1.2333,
+      "step": 12665
+    },
+    {
+      "epoch": 0.8464724746125066,
+      "grad_norm": 0.185546875,
+      "learning_rate": 1.4020084905294761e-05,
+      "loss": 1.2168,
+      "step": 12670
+    },
+    {
+      "epoch": 0.8468065205772314,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.3960600690850466e-05,
+      "loss": 1.2205,
+      "step": 12675
+    },
+    {
+      "epoch": 0.8471405665419561,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.3901233463351771e-05,
+      "loss": 1.1949,
+      "step": 12680
+    },
+    {
+      "epoch": 0.8474746125066809,
+      "grad_norm": 0.16015625,
+      "learning_rate": 1.3841983303519756e-05,
+      "loss": 1.22,
+      "step": 12685
+    },
+    {
+      "epoch": 0.8478086584714056,
+      "grad_norm": 0.18359375,
+      "learning_rate": 1.3782850291916271e-05,
+      "loss": 1.1553,
+      "step": 12690
+    },
+    {
+      "epoch": 0.8481427044361304,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.3723834508943945e-05,
+      "loss": 1.2315,
+      "step": 12695
+    },
+    {
+      "epoch": 0.8484767504008551,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.3664936034845933e-05,
+      "loss": 1.1529,
+      "step": 12700
+    },
+    {
+      "epoch": 0.8488107963655799,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.360615494970594e-05,
+      "loss": 1.1605,
+      "step": 12705
+    },
+    {
+      "epoch": 0.8491448423303046,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.3547491333448003e-05,
+      "loss": 1.1639,
+      "step": 12710
+    },
+    {
+      "epoch": 0.8494788882950294,
+      "grad_norm": 0.205078125,
+      "learning_rate": 1.348894526583655e-05,
+      "loss": 1.189,
+      "step": 12715
+    },
+    {
+      "epoch": 0.8498129342597541,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.343051682647607e-05,
+      "loss": 1.2305,
+      "step": 12720
+    },
+    {
+      "epoch": 0.8501469802244789,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.3372206094811158e-05,
+      "loss": 1.1985,
+      "step": 12725
+    },
+    {
+      "epoch": 0.8504810261892036,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.3314013150126336e-05,
+      "loss": 1.1517,
+      "step": 12730
+    },
+    {
+      "epoch": 0.8508150721539284,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.3255938071546026e-05,
+      "loss": 1.1212,
+      "step": 12735
+    },
+    {
+      "epoch": 0.8511491181186531,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.3197980938034305e-05,
+      "loss": 1.1748,
+      "step": 12740
+    },
+    {
+      "epoch": 0.8514831640833779,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.3140141828394992e-05,
+      "loss": 1.2564,
+      "step": 12745
+    },
+    {
+      "epoch": 0.8518172100481026,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.3082420821271324e-05,
+      "loss": 1.2606,
+      "step": 12750
+    },
+    {
+      "epoch": 0.8521512560128274,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.3024817995146032e-05,
+      "loss": 1.2375,
+      "step": 12755
+    },
+    {
+      "epoch": 0.8524853019775521,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.2967333428341121e-05,
+      "loss": 1.1506,
+      "step": 12760
+    },
+    {
+      "epoch": 0.8528193479422769,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.290996719901777e-05,
+      "loss": 1.2125,
+      "step": 12765
+    },
+    {
+      "epoch": 0.8531533939070016,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.2852719385176303e-05,
+      "loss": 1.231,
+      "step": 12770
+    },
+    {
+      "epoch": 0.8534874398717264,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.279559006465607e-05,
+      "loss": 1.2045,
+      "step": 12775
+    },
+    {
+      "epoch": 0.8538214858364511,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.2738579315135224e-05,
+      "loss": 1.1227,
+      "step": 12780
+    },
+    {
+      "epoch": 0.8541555318011759,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.268168721413071e-05,
+      "loss": 1.1242,
+      "step": 12785
+    },
+    {
+      "epoch": 0.8544895777659006,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.262491383899823e-05,
+      "loss": 1.1677,
+      "step": 12790
+    },
+    {
+      "epoch": 0.8548236237306254,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.2568259266931958e-05,
+      "loss": 1.2305,
+      "step": 12795
+    },
+    {
+      "epoch": 0.8551576696953501,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.251172357496455e-05,
+      "loss": 1.1436,
+      "step": 12800
+    },
+    {
+      "epoch": 0.8554917156600749,
+      "grad_norm": 0.19140625,
+      "learning_rate": 1.2455306839967085e-05,
+      "loss": 1.178,
+      "step": 12805
+    },
+    {
+      "epoch": 0.8558257616247996,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.2399009138648854e-05,
+      "loss": 1.1356,
+      "step": 12810
+    },
+    {
+      "epoch": 0.8561598075895244,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.2342830547557305e-05,
+      "loss": 1.2149,
+      "step": 12815
+    },
+    {
+      "epoch": 0.8564938535542491,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.228677114307788e-05,
+      "loss": 1.185,
+      "step": 12820
+    },
+    {
+      "epoch": 0.8568278995189738,
+      "grad_norm": 0.185546875,
+      "learning_rate": 1.2230831001434084e-05,
+      "loss": 1.2487,
+      "step": 12825
+    },
+    {
+      "epoch": 0.8571619454836985,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.2175010198687143e-05,
+      "loss": 1.1808,
+      "step": 12830
+    },
+    {
+      "epoch": 0.8574959914484233,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.2119308810736064e-05,
+      "loss": 1.234,
+      "step": 12835
+    },
+    {
+      "epoch": 0.857830037413148,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.2063726913317508e-05,
+      "loss": 1.1794,
+      "step": 12840
+    },
+    {
+      "epoch": 0.8581640833778728,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.2008264582005657e-05,
+      "loss": 1.2039,
+      "step": 12845
+    },
+    {
+      "epoch": 0.8584981293425975,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.195292189221211e-05,
+      "loss": 1.1237,
+      "step": 12850
+    },
+    {
+      "epoch": 0.8588321753073223,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.189769891918575e-05,
+      "loss": 1.2079,
+      "step": 12855
+    },
+    {
+      "epoch": 0.859166221272047,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 1.1842595738012774e-05,
+      "loss": 1.1594,
+      "step": 12860
+    },
+    {
+      "epoch": 0.8595002672367718,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.1787612423616412e-05,
+      "loss": 1.2106,
+      "step": 12865
+    },
+    {
+      "epoch": 0.8598343132014965,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.1732749050756987e-05,
+      "loss": 1.2124,
+      "step": 12870
+    },
+    {
+      "epoch": 0.8601683591662213,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.1678005694031657e-05,
+      "loss": 1.2354,
+      "step": 12875
+    },
+    {
+      "epoch": 0.860502405130946,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.1623382427874474e-05,
+      "loss": 1.1976,
+      "step": 12880
+    },
+    {
+      "epoch": 0.8608364510956708,
+      "grad_norm": 0.185546875,
+      "learning_rate": 1.156887932655616e-05,
+      "loss": 1.1636,
+      "step": 12885
+    },
+    {
+      "epoch": 0.8611704970603955,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.1514496464184056e-05,
+      "loss": 1.2191,
+      "step": 12890
+    },
+    {
+      "epoch": 0.8615045430251203,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.1460233914701968e-05,
+      "loss": 1.2312,
+      "step": 12895
+    },
+    {
+      "epoch": 0.861838588989845,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.1406091751890257e-05,
+      "loss": 1.2598,
+      "step": 12900
+    },
+    {
+      "epoch": 0.8621726349545697,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.135207004936546e-05,
+      "loss": 1.175,
+      "step": 12905
+    },
+    {
+      "epoch": 0.8625066809192945,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.1298168880580362e-05,
+      "loss": 1.2295,
+      "step": 12910
+    },
+    {
+      "epoch": 0.8628407268840192,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.1244388318823851e-05,
+      "loss": 1.173,
+      "step": 12915
+    },
+    {
+      "epoch": 0.863174772848744,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.1190728437220877e-05,
+      "loss": 1.2834,
+      "step": 12920
+    },
+    {
+      "epoch": 0.8635088188134687,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.113718930873222e-05,
+      "loss": 1.2072,
+      "step": 12925
+    },
+    {
+      "epoch": 0.8638428647781935,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.1083771006154553e-05,
+      "loss": 1.2174,
+      "step": 12930
+    },
+    {
+      "epoch": 0.8641769107429182,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.103047360212024e-05,
+      "loss": 1.1909,
+      "step": 12935
+    },
+    {
+      "epoch": 0.864510956707643,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.0977297169097234e-05,
+      "loss": 1.1942,
+      "step": 12940
+    },
+    {
+      "epoch": 0.8648450026723677,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.0924241779389011e-05,
+      "loss": 1.1526,
+      "step": 12945
+    },
+    {
+      "epoch": 0.8651790486370925,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.0871307505134476e-05,
+      "loss": 1.2266,
+      "step": 12950
+    },
+    {
+      "epoch": 0.8655130946018172,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.0818494418307845e-05,
+      "loss": 1.2367,
+      "step": 12955
+    },
+    {
+      "epoch": 0.865847140566542,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.076580259071861e-05,
+      "loss": 1.2017,
+      "step": 12960
+    },
+    {
+      "epoch": 0.8661811865312667,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.0713232094011316e-05,
+      "loss": 1.1665,
+      "step": 12965
+    },
+    {
+      "epoch": 0.8665152324959915,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.0660782999665542e-05,
+      "loss": 1.2247,
+      "step": 12970
+    },
+    {
+      "epoch": 0.8668492784607162,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.0608455378995851e-05,
+      "loss": 1.1821,
+      "step": 12975
+    },
+    {
+      "epoch": 0.867183324425441,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.0556249303151599e-05,
+      "loss": 1.1619,
+      "step": 12980
+    },
+    {
+      "epoch": 0.8675173703901656,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.050416484311686e-05,
+      "loss": 1.1964,
+      "step": 12985
+    },
+    {
+      "epoch": 0.8678514163548904,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.0452202069710393e-05,
+      "loss": 1.1573,
+      "step": 12990
+    },
+    {
+      "epoch": 0.8681854623196151,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.0400361053585506e-05,
+      "loss": 1.1311,
+      "step": 12995
+    },
+    {
+      "epoch": 0.8685195082843399,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.0348641865229914e-05,
+      "loss": 1.1306,
+      "step": 13000
+    },
+    {
+      "epoch": 0.8688535542490646,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.0297044574965675e-05,
+      "loss": 1.2016,
+      "step": 13005
+    },
+    {
+      "epoch": 0.8691876002137894,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.024556925294916e-05,
+      "loss": 1.2006,
+      "step": 13010
+    },
+    {
+      "epoch": 0.8695216461785141,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.0194215969170872e-05,
+      "loss": 1.1202,
+      "step": 13015
+    },
+    {
+      "epoch": 0.8698556921432389,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.0142984793455346e-05,
+      "loss": 1.2478,
+      "step": 13020
+    },
+    {
+      "epoch": 0.8701897381079636,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.0091875795461147e-05,
+      "loss": 1.1605,
+      "step": 13025
+    },
+    {
+      "epoch": 0.8705237840726884,
+      "grad_norm": 0.19140625,
+      "learning_rate": 1.00408890446807e-05,
+      "loss": 1.2451,
+      "step": 13030
+    },
+    {
+      "epoch": 0.8708578300374131,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.990024610440185e-06,
+      "loss": 1.2272,
+      "step": 13035
+    },
+    {
+      "epoch": 0.8711918760021379,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.939282561899466e-06,
+      "loss": 1.112,
+      "step": 13040
+    },
+    {
+      "epoch": 0.8715259219668626,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.888662968052053e-06,
+      "loss": 1.2417,
+      "step": 13045
+    },
+    {
+      "epoch": 0.8718599679315874,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.838165897724894e-06,
+      "loss": 1.2227,
+      "step": 13050
+    },
+    {
+      "epoch": 0.8721940138963121,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.787791419578407e-06,
+      "loss": 1.2224,
+      "step": 13055
+    },
+    {
+      "epoch": 0.8725280598610369,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 9.737539602106238e-06,
+      "loss": 1.1799,
+      "step": 13060
+    },
+    {
+      "epoch": 0.8728621058257616,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 9.687410513635354e-06,
+      "loss": 1.2489,
+      "step": 13065
+    },
+    {
+      "epoch": 0.8731961517904864,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.637404222325763e-06,
+      "loss": 1.1243,
+      "step": 13070
+    },
+    {
+      "epoch": 0.8735301977552111,
+      "grad_norm": 0.162109375,
+      "learning_rate": 9.587520796170524e-06,
+      "loss": 1.1682,
+      "step": 13075
+    },
+    {
+      "epoch": 0.8738642437199359,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 9.53776030299568e-06,
+      "loss": 1.1071,
+      "step": 13080
+    },
+    {
+      "epoch": 0.8741982896846606,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 9.488122810460097e-06,
+      "loss": 1.2338,
+      "step": 13085
+    },
+    {
+      "epoch": 0.8745323356493854,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.438608386055403e-06,
+      "loss": 1.23,
+      "step": 13090
+    },
+    {
+      "epoch": 0.8748663816141101,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 9.38921709710585e-06,
+      "loss": 1.1198,
+      "step": 13095
+    },
+    {
+      "epoch": 0.8752004275788349,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.339949010768346e-06,
+      "loss": 1.2296,
+      "step": 13100
+    },
+    {
+      "epoch": 0.8755344735435596,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 9.290804194032199e-06,
+      "loss": 1.1535,
+      "step": 13105
+    },
+    {
+      "epoch": 0.8758685195082844,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.24178271371915e-06,
+      "loss": 1.2012,
+      "step": 13110
+    },
+    {
+      "epoch": 0.8762025654730091,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 9.192884636483246e-06,
+      "loss": 1.2506,
+      "step": 13115
+    },
+    {
+      "epoch": 0.8765366114377339,
+      "grad_norm": 0.1875,
+      "learning_rate": 9.144110028810737e-06,
+      "loss": 1.2568,
+      "step": 13120
+    },
+    {
+      "epoch": 0.8768706574024586,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.095458957019987e-06,
+      "loss": 1.1589,
+      "step": 13125
+    },
+    {
+      "epoch": 0.8772047033671834,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.04693148726139e-06,
+      "loss": 1.2607,
+      "step": 13130
+    },
+    {
+      "epoch": 0.8775387493319081,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.998527685517255e-06,
+      "loss": 1.2026,
+      "step": 13135
+    },
+    {
+      "epoch": 0.8778727952966329,
+      "grad_norm": 0.16796875,
+      "learning_rate": 8.950247617601793e-06,
+      "loss": 1.1373,
+      "step": 13140
+    },
+    {
+      "epoch": 0.8782068412613576,
+      "grad_norm": 0.158203125,
+      "learning_rate": 8.902091349160968e-06,
+      "loss": 1.1837,
+      "step": 13145
+    },
+    {
+      "epoch": 0.8785408872260823,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.854058945672372e-06,
+      "loss": 1.1745,
+      "step": 13150
+    },
+    {
+      "epoch": 0.878874933190807,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 8.806150472445185e-06,
+      "loss": 1.2169,
+      "step": 13155
+    },
+    {
+      "epoch": 0.8792089791555318,
+      "grad_norm": 0.181640625,
+      "learning_rate": 8.758365994620132e-06,
+      "loss": 1.1807,
+      "step": 13160
+    },
+    {
+      "epoch": 0.8795430251202565,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 8.710705577169299e-06,
+      "loss": 1.1301,
+      "step": 13165
+    },
+    {
+      "epoch": 0.8798770710849813,
+      "grad_norm": 0.17578125,
+      "learning_rate": 8.663169284896078e-06,
+      "loss": 1.2744,
+      "step": 13170
+    },
+    {
+      "epoch": 0.880211117049706,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 8.615757182435124e-06,
+      "loss": 1.1536,
+      "step": 13175
+    },
+    {
+      "epoch": 0.8805451630144308,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.568469334252238e-06,
+      "loss": 1.2433,
+      "step": 13180
+    },
+    {
+      "epoch": 0.8808792089791555,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.52130580464423e-06,
+      "loss": 1.2025,
+      "step": 13185
+    },
+    {
+      "epoch": 0.8812132549438803,
+      "grad_norm": 0.1640625,
+      "learning_rate": 8.474266657738895e-06,
+      "loss": 1.069,
+      "step": 13190
+    },
+    {
+      "epoch": 0.881547300908605,
+      "grad_norm": 0.171875,
+      "learning_rate": 8.427351957494921e-06,
+      "loss": 1.1772,
+      "step": 13195
+    },
+    {
+      "epoch": 0.8818813468733298,
+      "grad_norm": 0.1796875,
+      "learning_rate": 8.380561767701778e-06,
+      "loss": 1.2106,
+      "step": 13200
+    },
+    {
+      "epoch": 0.8822153928380545,
+      "grad_norm": 0.208984375,
+      "learning_rate": 8.333896151979636e-06,
+      "loss": 1.2206,
+      "step": 13205
+    },
+    {
+      "epoch": 0.8825494388027793,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 8.287355173779265e-06,
+      "loss": 1.1584,
+      "step": 13210
+    },
+    {
+      "epoch": 0.882883484767504,
+      "grad_norm": 0.181640625,
+      "learning_rate": 8.240938896382022e-06,
+      "loss": 1.1826,
+      "step": 13215
+    },
+    {
+      "epoch": 0.8832175307322288,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 8.194647382899656e-06,
+      "loss": 1.2083,
+      "step": 13220
+    },
+    {
+      "epoch": 0.8835515766969535,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 8.148480696274275e-06,
+      "loss": 1.2252,
+      "step": 13225
+    },
+    {
+      "epoch": 0.8838856226616782,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.102438899278298e-06,
+      "loss": 1.1475,
+      "step": 13230
+    },
+    {
+      "epoch": 0.884219668626403,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.056522054514337e-06,
+      "loss": 1.2059,
+      "step": 13235
+    },
+    {
+      "epoch": 0.8845537145911277,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.010730224415064e-06,
+      "loss": 1.2262,
+      "step": 13240
+    },
+    {
+      "epoch": 0.8848877605558525,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.965063471243161e-06,
+      "loss": 1.1197,
+      "step": 13245
+    },
+    {
+      "epoch": 0.8852218065205772,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.919521857091328e-06,
+      "loss": 1.1822,
+      "step": 13250
+    },
+    {
+      "epoch": 0.885555852485302,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 7.874105443882041e-06,
+      "loss": 1.1623,
+      "step": 13255
+    },
+    {
+      "epoch": 0.8858898984500267,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.828814293367525e-06,
+      "loss": 1.2486,
+      "step": 13260
+    },
+    {
+      "epoch": 0.8862239444147515,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 7.783648467129767e-06,
+      "loss": 1.1448,
+      "step": 13265
+    },
+    {
+      "epoch": 0.8865579903794762,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 7.738608026580296e-06,
+      "loss": 1.1805,
+      "step": 13270
+    },
+    {
+      "epoch": 0.886892036344201,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.693693032960181e-06,
+      "loss": 1.2759,
+      "step": 13275
+    },
+    {
+      "epoch": 0.8872260823089257,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.648903547339858e-06,
+      "loss": 1.1487,
+      "step": 13280
+    },
+    {
+      "epoch": 0.8875601282736505,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.604239630619225e-06,
+      "loss": 1.1906,
+      "step": 13285
+    },
+    {
+      "epoch": 0.8878941742383752,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.559701343527348e-06,
+      "loss": 1.1496,
+      "step": 13290
+    },
+    {
+      "epoch": 0.8882282202031,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.515288746622495e-06,
+      "loss": 1.1543,
+      "step": 13295
+    },
+    {
+      "epoch": 0.8885622661678247,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.471001900292063e-06,
+      "loss": 1.1974,
+      "step": 13300
+    },
+    {
+      "epoch": 0.8888963121325495,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.42684086475246e-06,
+      "loss": 1.2546,
+      "step": 13305
+    },
+    {
+      "epoch": 0.8892303580972741,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.382805700049023e-06,
+      "loss": 1.2219,
+      "step": 13310
+    },
+    {
+      "epoch": 0.8895644040619989,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.338896466055934e-06,
+      "loss": 1.1687,
+      "step": 13315
+    },
+    {
+      "epoch": 0.8898984500267236,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.295113222476147e-06,
+      "loss": 1.1647,
+      "step": 13320
+    },
+    {
+      "epoch": 0.8902324959914484,
+      "grad_norm": 0.16015625,
+      "learning_rate": 7.251456028841319e-06,
+      "loss": 1.171,
+      "step": 13325
+    },
+    {
+      "epoch": 0.8905665419561731,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.207924944511757e-06,
+      "loss": 1.1786,
+      "step": 13330
+    },
+    {
+      "epoch": 0.8909005879208979,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.1645200286762245e-06,
+      "loss": 1.2024,
+      "step": 13335
+    },
+    {
+      "epoch": 0.8912346338856226,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.121241340351947e-06,
+      "loss": 1.2028,
+      "step": 13340
+    },
+    {
+      "epoch": 0.8915686798503474,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 7.078088938384597e-06,
+      "loss": 1.2511,
+      "step": 13345
+    },
+    {
+      "epoch": 0.8919027258150721,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.035062881448051e-06,
+      "loss": 1.1716,
+      "step": 13350
+    },
+    {
+      "epoch": 0.8922367717797969,
+      "grad_norm": 0.158203125,
+      "learning_rate": 6.99216322804439e-06,
+      "loss": 1.1883,
+      "step": 13355
+    },
+    {
+      "epoch": 0.8925708177445216,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.9493900365039335e-06,
+      "loss": 1.2068,
+      "step": 13360
+    },
+    {
+      "epoch": 0.8929048637092464,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.9067433649849465e-06,
+      "loss": 1.2287,
+      "step": 13365
+    },
+    {
+      "epoch": 0.8932389096739711,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.864223271473702e-06,
+      "loss": 1.2195,
+      "step": 13370
+    },
+    {
+      "epoch": 0.8935729556386959,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 6.821829813784331e-06,
+      "loss": 1.2318,
+      "step": 13375
+    },
+    {
+      "epoch": 0.8939070016034206,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.779563049558857e-06,
+      "loss": 1.1857,
+      "step": 13380
+    },
+    {
+      "epoch": 0.8942410475681454,
+      "grad_norm": 0.162109375,
+      "learning_rate": 6.7374230362669655e-06,
+      "loss": 1.1986,
+      "step": 13385
+    },
+    {
+      "epoch": 0.8945750935328701,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.695409831206045e-06,
+      "loss": 1.1192,
+      "step": 13390
+    },
+    {
+      "epoch": 0.8949091394975949,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.653523491501035e-06,
+      "loss": 1.1814,
+      "step": 13395
+    },
+    {
+      "epoch": 0.8952431854623196,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.611764074104409e-06,
+      "loss": 1.2354,
+      "step": 13400
+    },
+    {
+      "epoch": 0.8955772314270444,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.5701316357960285e-06,
+      "loss": 1.1647,
+      "step": 13405
+    },
+    {
+      "epoch": 0.8959112773917691,
+      "grad_norm": 0.18359375,
+      "learning_rate": 6.52862623318311e-06,
+      "loss": 1.1939,
+      "step": 13410
+    },
+    {
+      "epoch": 0.8962453233564939,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.487247922700157e-06,
+      "loss": 1.1808,
+      "step": 13415
+    },
+    {
+      "epoch": 0.8965793693212186,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.445996760608896e-06,
+      "loss": 1.2396,
+      "step": 13420
+    },
+    {
+      "epoch": 0.8969134152859434,
+      "grad_norm": 0.16796875,
+      "learning_rate": 6.404872802998108e-06,
+      "loss": 1.182,
+      "step": 13425
+    },
+    {
+      "epoch": 0.8972474612506681,
+      "grad_norm": 0.16015625,
+      "learning_rate": 6.3638761057836285e-06,
+      "loss": 1.1529,
+      "step": 13430
+    },
+    {
+      "epoch": 0.8975815072153929,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.323006724708302e-06,
+      "loss": 1.2241,
+      "step": 13435
+    },
+    {
+      "epoch": 0.8979155531801176,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.2822647153418235e-06,
+      "loss": 1.2137,
+      "step": 13440
+    },
+    {
+      "epoch": 0.8982495991448424,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 6.24165013308069e-06,
+      "loss": 1.1781,
+      "step": 13445
+    },
+    {
+      "epoch": 0.8985836451095671,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 6.2011630331481826e-06,
+      "loss": 1.2033,
+      "step": 13450
+    },
+    {
+      "epoch": 0.8989176910742919,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 6.160803470594234e-06,
+      "loss": 1.3137,
+      "step": 13455
+    },
+    {
+      "epoch": 0.8992517370390166,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.120571500295347e-06,
+      "loss": 1.2285,
+      "step": 13460
+    },
+    {
+      "epoch": 0.8995857830037414,
+      "grad_norm": 0.16796875,
+      "learning_rate": 6.080467176954519e-06,
+      "loss": 1.1531,
+      "step": 13465
+    },
+    {
+      "epoch": 0.899919828968466,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.040490555101241e-06,
+      "loss": 1.1854,
+      "step": 13470
+    },
+    {
+      "epoch": 0.9002538749331908,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.000641689091324e-06,
+      "loss": 1.1357,
+      "step": 13475
+    },
+    {
+      "epoch": 0.9005879208979155,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 5.960920633106926e-06,
+      "loss": 1.2068,
+      "step": 13480
+    },
+    {
+      "epoch": 0.9009219668626403,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.921327441156321e-06,
+      "loss": 1.2706,
+      "step": 13485
+    },
+    {
+      "epoch": 0.901256012827365,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.8818621670740595e-06,
+      "loss": 1.2044,
+      "step": 13490
+    },
+    {
+      "epoch": 0.9015900587920898,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.84252486452066e-06,
+      "loss": 1.1731,
+      "step": 13495
+    },
+    {
+      "epoch": 0.9019241047568145,
+      "grad_norm": 0.1796875,
+      "learning_rate": 5.803315586982671e-06,
+      "loss": 1.1598,
+      "step": 13500
+    },
+    {
+      "epoch": 0.9022581507215393,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.764234387772593e-06,
+      "loss": 1.1737,
+      "step": 13505
+    },
+    {
+      "epoch": 0.902592196686264,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 5.7252813200287306e-06,
+      "loss": 1.1653,
+      "step": 13510
+    },
+    {
+      "epoch": 0.9029262426509888,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 5.686456436715227e-06,
+      "loss": 1.1973,
+      "step": 13515
+    },
+    {
+      "epoch": 0.9032602886157135,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.647759790621876e-06,
+      "loss": 1.2162,
+      "step": 13520
+    },
+    {
+      "epoch": 0.9035943345804383,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 5.609191434364159e-06,
+      "loss": 1.2118,
+      "step": 13525
+    },
+    {
+      "epoch": 0.903928380545163,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 5.5707514203830915e-06,
+      "loss": 1.2077,
+      "step": 13530
+    },
+    {
+      "epoch": 0.9042624265098878,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.532439800945188e-06,
+      "loss": 1.2367,
+      "step": 13535
+    },
+    {
+      "epoch": 0.9045964724746125,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.494256628142358e-06,
+      "loss": 1.2476,
+      "step": 13540
+    },
+    {
+      "epoch": 0.9049305184393373,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.45620195389196e-06,
+      "loss": 1.1336,
+      "step": 13545
+    },
+    {
+      "epoch": 0.905264564404062,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 5.418275829936537e-06,
+      "loss": 1.2059,
+      "step": 13550
+    },
+    {
+      "epoch": 0.9055986103687867,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 5.380478307843872e-06,
+      "loss": 1.181,
+      "step": 13555
+    },
+    {
+      "epoch": 0.9059326563335115,
+      "grad_norm": 0.162109375,
+      "learning_rate": 5.3428094390068906e-06,
+      "loss": 1.1992,
+      "step": 13560
+    },
+    {
+      "epoch": 0.9062667022982362,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.3052692746436095e-06,
+      "loss": 1.1258,
+      "step": 13565
+    },
+    {
+      "epoch": 0.906600748262961,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 5.2678578657970105e-06,
+      "loss": 1.1891,
+      "step": 13570
+    },
+    {
+      "epoch": 0.9069347942276857,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.2305752633350355e-06,
+      "loss": 1.1366,
+      "step": 13575
+    },
+    {
+      "epoch": 0.9072688401924105,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 5.193421517950481e-06,
+      "loss": 1.1117,
+      "step": 13580
+    },
+    {
+      "epoch": 0.9076028861571352,
+      "grad_norm": 0.1796875,
+      "learning_rate": 5.1563966801609465e-06,
+      "loss": 1.228,
+      "step": 13585
+    },
+    {
+      "epoch": 0.90793693212186,
+      "grad_norm": 0.1640625,
+      "learning_rate": 5.119500800308741e-06,
+      "loss": 1.2213,
+      "step": 13590
+    },
+    {
+      "epoch": 0.9082709780865847,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.082733928560835e-06,
+      "loss": 1.2232,
+      "step": 13595
+    },
+    {
+      "epoch": 0.9086050240513095,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.0460961149087824e-06,
+      "loss": 1.1233,
+      "step": 13600
+    },
+    {
+      "epoch": 0.9089390700160342,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.009587409168703e-06,
+      "loss": 1.2048,
+      "step": 13605
+    },
+    {
+      "epoch": 0.909273115980759,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.9732078609811e-06,
+      "loss": 1.1446,
+      "step": 13610
+    },
+    {
+      "epoch": 0.9096071619454837,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 4.936957519810892e-06,
+      "loss": 1.2648,
+      "step": 13615
+    },
+    {
+      "epoch": 0.9099412079102085,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.900836434947353e-06,
+      "loss": 1.2087,
+      "step": 13620
+    },
+    {
+      "epoch": 0.9102752538749332,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.864844655503953e-06,
+      "loss": 1.2085,
+      "step": 13625
+    },
+    {
+      "epoch": 0.9106092998396579,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 4.8289822304183665e-06,
+      "loss": 1.2029,
+      "step": 13630
+    },
+    {
+      "epoch": 0.9109433458043826,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.793249208452388e-06,
+      "loss": 1.1732,
+      "step": 13635
+    },
+    {
+      "epoch": 0.9112773917691074,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 4.757645638191876e-06,
+      "loss": 1.2723,
+      "step": 13640
+    },
+    {
+      "epoch": 0.9116114377338321,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.722171568046674e-06,
+      "loss": 1.2018,
+      "step": 13645
+    },
+    {
+      "epoch": 0.9119454836985569,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.686827046250497e-06,
+      "loss": 1.1453,
+      "step": 13650
+    },
+    {
+      "epoch": 0.9122795296632816,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.651612120860993e-06,
+      "loss": 1.2069,
+      "step": 13655
+    },
+    {
+      "epoch": 0.9126135756280064,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.616526839759516e-06,
+      "loss": 1.2072,
+      "step": 13660
+    },
+    {
+      "epoch": 0.9129476215927311,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.58157125065124e-06,
+      "loss": 1.2006,
+      "step": 13665
+    },
+    {
+      "epoch": 0.9132816675574559,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.546745401064889e-06,
+      "loss": 1.237,
+      "step": 13670
+    },
+    {
+      "epoch": 0.9136157135221806,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.512049338352875e-06,
+      "loss": 1.1559,
+      "step": 13675
+    },
+    {
+      "epoch": 0.9139497594869054,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.477483109691083e-06,
+      "loss": 1.2091,
+      "step": 13680
+    },
+    {
+      "epoch": 0.9142838054516301,
+      "grad_norm": 0.1640625,
+      "learning_rate": 4.443046762078884e-06,
+      "loss": 1.1398,
+      "step": 13685
+    },
+    {
+      "epoch": 0.9146178514163549,
+      "grad_norm": 0.1796875,
+      "learning_rate": 4.408740342339046e-06,
+      "loss": 1.1943,
+      "step": 13690
+    },
+    {
+      "epoch": 0.9149518973810796,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 4.3745638971177005e-06,
+      "loss": 1.1941,
+      "step": 13695
+    },
+    {
+      "epoch": 0.9152859433458044,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 4.34051747288422e-06,
+      "loss": 1.2382,
+      "step": 13700
+    },
+    {
+      "epoch": 0.9156199893105291,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.306601115931175e-06,
+      "loss": 1.1271,
+      "step": 13705
+    },
+    {
+      "epoch": 0.9159540352752539,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.2728148723743536e-06,
+      "loss": 1.2612,
+      "step": 13710
+    },
+    {
+      "epoch": 0.9162880812399786,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 4.239158788152564e-06,
+      "loss": 1.1464,
+      "step": 13715
+    },
+    {
+      "epoch": 0.9166221272047034,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.205632909027668e-06,
+      "loss": 1.1814,
+      "step": 13720
+    },
+    {
+      "epoch": 0.9169561731694281,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 4.172237280584445e-06,
+      "loss": 1.2345,
+      "step": 13725
+    },
+    {
+      "epoch": 0.9172902191341529,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.138971948230674e-06,
+      "loss": 1.1822,
+      "step": 13730
+    },
+    {
+      "epoch": 0.9176242650988776,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.105836957196873e-06,
+      "loss": 1.1968,
+      "step": 13735
+    },
+    {
+      "epoch": 0.9179583110636024,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 4.072832352536382e-06,
+      "loss": 1.1399,
+      "step": 13740
+    },
+    {
+      "epoch": 0.9182923570283271,
+      "grad_norm": 0.16015625,
+      "learning_rate": 4.039958179125225e-06,
+      "loss": 1.1361,
+      "step": 13745
+    },
+    {
+      "epoch": 0.9186264029930519,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.007214481662125e-06,
+      "loss": 1.1929,
+      "step": 13750
+    },
+    {
+      "epoch": 0.9189604489577766,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.974601304668346e-06,
+      "loss": 1.1354,
+      "step": 13755
+    },
+    {
+      "epoch": 0.9192944949225014,
+      "grad_norm": 0.162109375,
+      "learning_rate": 3.94211869248774e-06,
+      "loss": 1.1908,
+      "step": 13760
+    },
+    {
+      "epoch": 0.9196285408872261,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.909766689286576e-06,
+      "loss": 1.1672,
+      "step": 13765
+    },
+    {
+      "epoch": 0.9199625868519509,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.877545339053601e-06,
+      "loss": 1.1175,
+      "step": 13770
+    },
+    {
+      "epoch": 0.9202966328166756,
+      "grad_norm": 0.177734375,
+      "learning_rate": 3.845454685599847e-06,
+      "loss": 1.1686,
+      "step": 13775
+    },
+    {
+      "epoch": 0.9206306787814004,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.813494772558657e-06,
+      "loss": 1.1441,
+      "step": 13780
+    },
+    {
+      "epoch": 0.9209647247461251,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 3.7816656433856348e-06,
+      "loss": 1.1806,
+      "step": 13785
+    },
+    {
+      "epoch": 0.9212987707108499,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 3.7499673413585513e-06,
+      "loss": 1.1616,
+      "step": 13790
+    },
+    {
+      "epoch": 0.9216328166755745,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.718399909577275e-06,
+      "loss": 1.1416,
+      "step": 13795
+    },
+    {
+      "epoch": 0.9219668626402993,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.686963390963727e-06,
+      "loss": 1.1897,
+      "step": 13800
+    },
+    {
+      "epoch": 0.922300908605024,
+      "grad_norm": 0.18359375,
+      "learning_rate": 3.65565782826186e-06,
+      "loss": 1.1332,
+      "step": 13805
+    },
+    {
+      "epoch": 0.9226349545697488,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 3.624483264037537e-06,
+      "loss": 1.1628,
+      "step": 13810
+    },
+    {
+      "epoch": 0.9229690005344735,
+      "grad_norm": 0.162109375,
+      "learning_rate": 3.593439740678506e-06,
+      "loss": 1.2139,
+      "step": 13815
+    },
+    {
+      "epoch": 0.9233030464991983,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.5625273003943583e-06,
+      "loss": 1.176,
+      "step": 13820
+    },
+    {
+      "epoch": 0.923637092463923,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.5317459852164502e-06,
+      "loss": 1.2045,
+      "step": 13825
+    },
+    {
+      "epoch": 0.9239711384286478,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.501095836997825e-06,
+      "loss": 1.2195,
+      "step": 13830
+    },
+    {
+      "epoch": 0.9243051843933725,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 3.4705768974132024e-06,
+      "loss": 1.1469,
+      "step": 13835
+    },
+    {
+      "epoch": 0.9246392303580973,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 3.4401892079588993e-06,
+      "loss": 1.1442,
+      "step": 13840
+    },
+    {
+      "epoch": 0.924973276322822,
+      "grad_norm": 0.18359375,
+      "learning_rate": 3.4099328099527427e-06,
+      "loss": 1.2137,
+      "step": 13845
+    },
+    {
+      "epoch": 0.9253073222875468,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.3798077445340915e-06,
+      "loss": 1.2369,
+      "step": 13850
+    },
+    {
+      "epoch": 0.9256413682522715,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 3.34981405266368e-06,
+      "loss": 1.1292,
+      "step": 13855
+    },
+    {
+      "epoch": 0.9259754142169963,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.3199517751236753e-06,
+      "loss": 1.1946,
+      "step": 13860
+    },
+    {
+      "epoch": 0.926309460181721,
+      "grad_norm": 0.16796875,
+      "learning_rate": 3.29022095251752e-06,
+      "loss": 1.156,
+      "step": 13865
+    },
+    {
+      "epoch": 0.9266435061464458,
+      "grad_norm": 0.162109375,
+      "learning_rate": 3.2606216252699106e-06,
+      "loss": 1.2873,
+      "step": 13870
+    },
+    {
+      "epoch": 0.9269775521111705,
+      "grad_norm": 0.173828125,
+      "learning_rate": 3.2311538336267766e-06,
+      "loss": 1.2188,
+      "step": 13875
+    },
+    {
+      "epoch": 0.9273115980758952,
+      "grad_norm": 0.1640625,
+      "learning_rate": 3.2018176176552116e-06,
+      "loss": 1.1445,
+      "step": 13880
+    },
+    {
+      "epoch": 0.92764564404062,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.1726130172433755e-06,
+      "loss": 1.1774,
+      "step": 13885
+    },
+    {
+      "epoch": 0.9279796900053447,
+      "grad_norm": 0.177734375,
+      "learning_rate": 3.143540072100459e-06,
+      "loss": 1.2333,
+      "step": 13890
+    },
+    {
+      "epoch": 0.9283137359700695,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.114598821756698e-06,
+      "loss": 1.2043,
+      "step": 13895
+    },
+    {
+      "epoch": 0.9286477819347942,
+      "grad_norm": 0.177734375,
+      "learning_rate": 3.0857893055632246e-06,
+      "loss": 1.2579,
+      "step": 13900
+    },
+    {
+      "epoch": 0.928981827899519,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.0571115626920498e-06,
+      "loss": 1.1417,
+      "step": 13905
+    },
+    {
+      "epoch": 0.9293158738642437,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.0285656321360267e-06,
+      "loss": 1.1325,
+      "step": 13910
+    },
+    {
+      "epoch": 0.9296499198289685,
+      "grad_norm": 0.1640625,
+      "learning_rate": 3.0001515527088074e-06,
+      "loss": 1.1791,
+      "step": 13915
+    },
+    {
+      "epoch": 0.9299839657936932,
+      "grad_norm": 0.1640625,
+      "learning_rate": 2.9718693630447214e-06,
+      "loss": 1.1859,
+      "step": 13920
+    },
+    {
+      "epoch": 0.930318011758418,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.9437191015987854e-06,
+      "loss": 1.1939,
+      "step": 13925
+    },
+    {
+      "epoch": 0.9306520577231427,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.9157008066466707e-06,
+      "loss": 1.2171,
+      "step": 13930
+    },
+    {
+      "epoch": 0.9309861036878675,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.8878145162845373e-06,
+      "loss": 1.2151,
+      "step": 13935
+    },
+    {
+      "epoch": 0.9313201496525922,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.8600602684291546e-06,
+      "loss": 1.2412,
+      "step": 13940
+    },
+    {
+      "epoch": 0.931654195617317,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.832438100817658e-06,
+      "loss": 1.1997,
+      "step": 13945
+    },
+    {
+      "epoch": 0.9319882415820417,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.8049480510076944e-06,
+      "loss": 1.2391,
+      "step": 13950
+    },
+    {
+      "epoch": 0.9323222875467664,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 2.7775901563771967e-06,
+      "loss": 1.267,
+      "step": 13955
+    },
+    {
+      "epoch": 0.9326563335114911,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.750364454124432e-06,
+      "loss": 1.1957,
+      "step": 13960
+    },
+    {
+      "epoch": 0.9329903794762159,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.7232709812679114e-06,
+      "loss": 1.1725,
+      "step": 13965
+    },
+    {
+      "epoch": 0.9333244254409406,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.6963097746463882e-06,
+      "loss": 1.085,
+      "step": 13970
+    },
+    {
+      "epoch": 0.9336584714056654,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.6694808709187613e-06,
+      "loss": 1.1619,
+      "step": 13975
+    },
+    {
+      "epoch": 0.9339925173703901,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.6427843065640167e-06,
+      "loss": 1.2746,
+      "step": 13980
+    },
+    {
+      "epoch": 0.9343265633351149,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.6162201178812294e-06,
+      "loss": 1.2403,
+      "step": 13985
+    },
+    {
+      "epoch": 0.9346606092998396,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.589788340989474e-06,
+      "loss": 1.2107,
+      "step": 13990
+    },
+    {
+      "epoch": 0.9349946552645644,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.5634890118277686e-06,
+      "loss": 1.2215,
+      "step": 13995
+    },
+    {
+      "epoch": 0.9353287012292891,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.537322166155065e-06,
+      "loss": 1.1664,
+      "step": 14000
+    },
+    {
+      "epoch": 0.9356627471940139,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.5112878395501805e-06,
+      "loss": 1.209,
+      "step": 14005
+    },
+    {
+      "epoch": 0.9359967931587386,
+      "grad_norm": 0.203125,
+      "learning_rate": 2.4853860674117325e-06,
+      "loss": 1.1431,
+      "step": 14010
+    },
+    {
+      "epoch": 0.9363308391234634,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.4596168849581158e-06,
+      "loss": 1.2497,
+      "step": 14015
+    },
+    {
+      "epoch": 0.9366648850881881,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.4339803272274366e-06,
+      "loss": 1.1754,
+      "step": 14020
+    },
+    {
+      "epoch": 0.9369989310529129,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.4084764290774775e-06,
+      "loss": 1.1658,
+      "step": 14025
+    },
+    {
+      "epoch": 0.9373329770176376,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.3831052251856445e-06,
+      "loss": 1.2201,
+      "step": 14030
+    },
+    {
+      "epoch": 0.9376670229823624,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.3578667500489206e-06,
+      "loss": 1.2038,
+      "step": 14035
+    },
+    {
+      "epoch": 0.9380010689470871,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 2.3327610379838105e-06,
+      "loss": 1.1809,
+      "step": 14040
+    },
+    {
+      "epoch": 0.9383351149118119,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.307788123126331e-06,
+      "loss": 1.1757,
+      "step": 14045
+    },
+    {
+      "epoch": 0.9386691608765366,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.282948039431898e-06,
+      "loss": 1.209,
+      "step": 14050
+    },
+    {
+      "epoch": 0.9390032068412614,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.258240820675317e-06,
+      "loss": 1.1629,
+      "step": 14055
+    },
+    {
+      "epoch": 0.9393372528059861,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.233666500450793e-06,
+      "loss": 1.1975,
+      "step": 14060
+    },
+    {
+      "epoch": 0.9396712987707109,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.2092251121717757e-06,
+      "loss": 1.1843,
+      "step": 14065
+    },
+    {
+      "epoch": 0.9400053447354356,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.1849166890709927e-06,
+      "loss": 1.1829,
+      "step": 14070
+    },
+    {
+      "epoch": 0.9403393907001604,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.160741264200361e-06,
+      "loss": 1.1872,
+      "step": 14075
+    },
+    {
+      "epoch": 0.9406734366648851,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.1366988704309976e-06,
+      "loss": 1.2531,
+      "step": 14080
+    },
+    {
+      "epoch": 0.9410074826296099,
+      "grad_norm": 0.158203125,
+      "learning_rate": 2.112789540453086e-06,
+      "loss": 1.1454,
+      "step": 14085
+    },
+    {
+      "epoch": 0.9413415285943346,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.089013306775922e-06,
+      "loss": 1.1905,
+      "step": 14090
+    },
+    {
+      "epoch": 0.9416755745590594,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.065370201727823e-06,
+      "loss": 1.1781,
+      "step": 14095
+    },
+    {
+      "epoch": 0.9420096205237841,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.0418602574561074e-06,
+      "loss": 1.158,
+      "step": 14100
+    },
+    {
+      "epoch": 0.9423436664885089,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.0184835059270047e-06,
+      "loss": 1.2032,
+      "step": 14105
+    },
+    {
+      "epoch": 0.9426777124532336,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.9952399789256558e-06,
+      "loss": 1.135,
+      "step": 14110
+    },
+    {
+      "epoch": 0.9430117584179583,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.9721297080560807e-06,
+      "loss": 1.225,
+      "step": 14115
+    },
+    {
+      "epoch": 0.943345804382683,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.9491527247410657e-06,
+      "loss": 1.2238,
+      "step": 14120
+    },
+    {
+      "epoch": 0.9436798503474078,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.92630906022222e-06,
+      "loss": 1.0923,
+      "step": 14125
+    },
+    {
+      "epoch": 0.9440138963121325,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.9035987455598425e-06,
+      "loss": 1.2037,
+      "step": 14130
+    },
+    {
+      "epoch": 0.9443479422768573,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.8810218116329326e-06,
+      "loss": 1.2147,
+      "step": 14135
+    },
+    {
+      "epoch": 0.944681988241582,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.8585782891391345e-06,
+      "loss": 1.1872,
+      "step": 14140
+    },
+    {
+      "epoch": 0.9450160342063068,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 1.8362682085946825e-06,
+      "loss": 1.2518,
+      "step": 14145
+    },
+    {
+      "epoch": 0.9453500801710315,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.8140916003343778e-06,
+      "loss": 1.2239,
+      "step": 14150
+    },
+    {
+      "epoch": 0.9456841261357563,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.7920484945115557e-06,
+      "loss": 1.1243,
+      "step": 14155
+    },
+    {
+      "epoch": 0.946018172100481,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 1.770138921098008e-06,
+      "loss": 1.1931,
+      "step": 14160
+    },
+    {
+      "epoch": 0.9463522180652058,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.7483629098839605e-06,
+      "loss": 1.2633,
+      "step": 14165
+    },
+    {
+      "epoch": 0.9466862640299305,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.7267204904780621e-06,
+      "loss": 1.2289,
+      "step": 14170
+    },
+    {
+      "epoch": 0.9470203099946553,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.7052116923072958e-06,
+      "loss": 1.2129,
+      "step": 14175
+    },
+    {
+      "epoch": 0.94735435595938,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.683836544616979e-06,
+      "loss": 1.2649,
+      "step": 14180
+    },
+    {
+      "epoch": 0.9476884019241048,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.6625950764706743e-06,
+      "loss": 1.1021,
+      "step": 14185
+    },
+    {
+      "epoch": 0.9480224478888295,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.641487316750212e-06,
+      "loss": 1.2487,
+      "step": 14190
+    },
+    {
+      "epoch": 0.9483564938535542,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.6205132941556122e-06,
+      "loss": 1.2798,
+      "step": 14195
+    },
+    {
+      "epoch": 0.948690539818279,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.5996730372050627e-06,
+      "loss": 1.1935,
+      "step": 14200
+    },
+    {
+      "epoch": 0.9490245857830037,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.5789665742348415e-06,
+      "loss": 1.1511,
+      "step": 14205
+    },
+    {
+      "epoch": 0.9493586317477285,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.558393933399338e-06,
+      "loss": 1.2232,
+      "step": 14210
+    },
+    {
+      "epoch": 0.9496926777124532,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.537955142670966e-06,
+      "loss": 1.2133,
+      "step": 14215
+    },
+    {
+      "epoch": 0.950026723677178,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.517650229840173e-06,
+      "loss": 1.1668,
+      "step": 14220
+    },
+    {
+      "epoch": 0.9503607696419027,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.4974792225153188e-06,
+      "loss": 1.2862,
+      "step": 14225
+    },
+    {
+      "epoch": 0.9506948156066275,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.4774421481227762e-06,
+      "loss": 1.2089,
+      "step": 14230
+    },
+    {
+      "epoch": 0.9510288615713522,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.4575390339067296e-06,
+      "loss": 1.1126,
+      "step": 14235
+    },
+    {
+      "epoch": 0.951362907536077,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.4377699069292427e-06,
+      "loss": 1.1774,
+      "step": 14240
+    },
+    {
+      "epoch": 0.9516969535008017,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.418134794070236e-06,
+      "loss": 1.1765,
+      "step": 14245
+    },
+    {
+      "epoch": 0.9520309994655265,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.3986337220273759e-06,
+      "loss": 1.1659,
+      "step": 14250
+    },
+    {
+      "epoch": 0.9523650454302512,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.3792667173160855e-06,
+      "loss": 1.2359,
+      "step": 14255
+    },
+    {
+      "epoch": 0.952699091394976,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.3600338062694784e-06,
+      "loss": 1.1513,
+      "step": 14260
+    },
+    {
+      "epoch": 0.9530331373597007,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.3409350150383803e-06,
+      "loss": 1.1959,
+      "step": 14265
+    },
+    {
+      "epoch": 0.9533671833244255,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.32197036959123e-06,
+      "loss": 1.1792,
+      "step": 14270
+    },
+    {
+      "epoch": 0.9537012292891501,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.3031398957140562e-06,
+      "loss": 1.2564,
+      "step": 14275
+    },
+    {
+      "epoch": 0.9540352752538749,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.284443619010478e-06,
+      "loss": 1.2568,
+      "step": 14280
+    },
+    {
+      "epoch": 0.9543693212185996,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 1.265881564901672e-06,
+      "loss": 1.2136,
+      "step": 14285
+    },
+    {
+      "epoch": 0.9547033671833244,
+      "grad_norm": 0.201171875,
+      "learning_rate": 1.2474537586262603e-06,
+      "loss": 1.2579,
+      "step": 14290
+    },
+    {
+      "epoch": 0.9550374131480491,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.2291602252403444e-06,
+      "loss": 1.2214,
+      "step": 14295
+    },
+    {
+      "epoch": 0.9553714591127739,
+      "grad_norm": 0.189453125,
+      "learning_rate": 1.2110009896174944e-06,
+      "loss": 1.1595,
+      "step": 14300
+    },
+    {
+      "epoch": 0.9557055050774986,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.1929760764486264e-06,
+      "loss": 1.1858,
+      "step": 14305
+    },
+    {
+      "epoch": 0.9560395510422234,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.1750855102420578e-06,
+      "loss": 1.1783,
+      "step": 14310
+    },
+    {
+      "epoch": 0.9563735970069481,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.1573293153233966e-06,
+      "loss": 1.257,
+      "step": 14315
+    },
+    {
+      "epoch": 0.9567076429716729,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.1397075158355975e-06,
+      "loss": 1.197,
+      "step": 14320
+    },
+    {
+      "epoch": 0.9570416889363976,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.1222201357388496e-06,
+      "loss": 1.2113,
+      "step": 14325
+    },
+    {
+      "epoch": 0.9573757349011224,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.1048671988105553e-06,
+      "loss": 1.1578,
+      "step": 14330
+    },
+    {
+      "epoch": 0.9577097808658471,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.0876487286453408e-06,
+      "loss": 1.1104,
+      "step": 14335
+    },
+    {
+      "epoch": 0.9580438268305719,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.0705647486550118e-06,
+      "loss": 1.23,
+      "step": 14340
+    },
+    {
+      "epoch": 0.9583778727952966,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.0536152820684874e-06,
+      "loss": 1.2529,
+      "step": 14345
+    },
+    {
+      "epoch": 0.9587119187600214,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.036800351931777e-06,
+      "loss": 1.2079,
+      "step": 14350
+    },
+    {
+      "epoch": 0.9590459647247461,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.0201199811080032e-06,
+      "loss": 1.2061,
+      "step": 14355
+    },
+    {
+      "epoch": 0.9593800106894709,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.0035741922772902e-06,
+      "loss": 1.1448,
+      "step": 14360
+    },
+    {
+      "epoch": 0.9597140566541956,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.871630079367977e-07,
+      "loss": 1.2618,
+      "step": 14365
+    },
+    {
+      "epoch": 0.9600481026189204,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.708864504006433e-07,
+      "loss": 1.1832,
+      "step": 14370
+    },
+    {
+      "epoch": 0.9603821485836451,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 9.547445417999123e-07,
+      "loss": 1.1876,
+      "step": 14375
+    },
+    {
+      "epoch": 0.9607161945483699,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.387373040826153e-07,
+      "loss": 1.1768,
+      "step": 14380
+    },
+    {
+      "epoch": 0.9610502405130946,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 9.228647590136308e-07,
+      "loss": 1.1675,
+      "step": 14385
+    },
+    {
+      "epoch": 0.9613842864778194,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 9.071269281746842e-07,
+      "loss": 1.1261,
+      "step": 14390
+    },
+    {
+      "epoch": 0.9617183324425441,
+      "grad_norm": 0.357421875,
+      "learning_rate": 8.915238329643805e-07,
+      "loss": 1.198,
+      "step": 14395
+    },
+    {
+      "epoch": 0.9620523784072689,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.760554945981048e-07,
+      "loss": 1.2774,
+      "step": 14400
+    },
+    {
+      "epoch": 0.9623864243719936,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.607219341079887e-07,
+      "loss": 1.2643,
+      "step": 14405
+    },
+    {
+      "epoch": 0.9627204703367184,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.455231723429324e-07,
+      "loss": 1.1497,
+      "step": 14410
+    },
+    {
+      "epoch": 0.9630545163014431,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.304592299685387e-07,
+      "loss": 1.2001,
+      "step": 14415
+    },
+    {
+      "epoch": 0.9633885622661679,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 8.155301274671234e-07,
+      "loss": 1.2162,
+      "step": 14420
+    },
+    {
+      "epoch": 0.9637226082308926,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 8.007358851376378e-07,
+      "loss": 1.2046,
+      "step": 14425
+    },
+    {
+      "epoch": 0.9640566541956174,
+      "grad_norm": 0.162109375,
+      "learning_rate": 7.860765230956579e-07,
+      "loss": 1.1698,
+      "step": 14430
+    },
+    {
+      "epoch": 0.9643907001603421,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.71552061273395e-07,
+      "loss": 1.2303,
+      "step": 14435
+    },
+    {
+      "epoch": 0.9647247461250668,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 7.571625194196074e-07,
+      "loss": 1.1499,
+      "step": 14440
+    },
+    {
+      "epoch": 0.9650587920897915,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.429079170996222e-07,
+      "loss": 1.1865,
+      "step": 14445
+    },
+    {
+      "epoch": 0.9653928380545163,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.287882736952912e-07,
+      "loss": 1.2239,
+      "step": 14450
+    },
+    {
+      "epoch": 0.965726884019241,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.14803608404957e-07,
+      "loss": 1.134,
+      "step": 14455
+    },
+    {
+      "epoch": 0.9660609299839658,
+      "grad_norm": 0.189453125,
+      "learning_rate": 7.009539402434207e-07,
+      "loss": 1.2218,
+      "step": 14460
+    },
+    {
+      "epoch": 0.9663949759486905,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.87239288041952e-07,
+      "loss": 1.2575,
+      "step": 14465
+    },
+    {
+      "epoch": 0.9667290219134153,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.736596704482456e-07,
+      "loss": 1.1718,
+      "step": 14470
+    },
+    {
+      "epoch": 0.96706306787814,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.60215105926365e-07,
+      "loss": 1.1994,
+      "step": 14475
+    },
+    {
+      "epoch": 0.9673971138428648,
+      "grad_norm": 0.16796875,
+      "learning_rate": 6.46905612756743e-07,
+      "loss": 1.1799,
+      "step": 14480
+    },
+    {
+      "epoch": 0.9677311598075895,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 6.337312090361924e-07,
+      "loss": 1.2205,
+      "step": 14485
+    },
+    {
+      "epoch": 0.9680652057723143,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.206919126777955e-07,
+      "loss": 1.2612,
+      "step": 14490
+    },
+    {
+      "epoch": 0.968399251737039,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 6.077877414109923e-07,
+      "loss": 1.1688,
+      "step": 14495
+    },
+    {
+      "epoch": 0.9687332977017638,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.950187127814477e-07,
+      "loss": 1.2077,
+      "step": 14500
+    },
+    {
+      "epoch": 0.9690673436664885,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.823848441510737e-07,
+      "loss": 1.1738,
+      "step": 14505
+    },
+    {
+      "epoch": 0.9694013896312133,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.698861526980404e-07,
+      "loss": 1.1966,
+      "step": 14510
+    },
+    {
+      "epoch": 0.969735435595938,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 5.575226554166757e-07,
+      "loss": 1.145,
+      "step": 14515
+    },
+    {
+      "epoch": 0.9700694815606627,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.452943691175216e-07,
+      "loss": 1.1979,
+      "step": 14520
+    },
+    {
+      "epoch": 0.9704035275253875,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 5.332013104272782e-07,
+      "loss": 1.1926,
+      "step": 14525
+    },
+    {
+      "epoch": 0.9707375734901122,
+      "grad_norm": 0.1796875,
+      "learning_rate": 5.212434957887369e-07,
+      "loss": 1.2438,
+      "step": 14530
+    },
+    {
+      "epoch": 0.971071619454837,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.094209414608253e-07,
+      "loss": 1.2502,
+      "step": 14535
+    },
+    {
+      "epoch": 0.9714056654195617,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.977336635185736e-07,
+      "loss": 1.1783,
+      "step": 14540
+    },
+    {
+      "epoch": 0.9717397113842865,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.861816778530371e-07,
+      "loss": 1.2503,
+      "step": 14545
+    },
+    {
+      "epoch": 0.9720737573490112,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 4.7476500017136215e-07,
+      "loss": 1.2635,
+      "step": 14550
+    },
+    {
+      "epoch": 0.972407803313736,
+      "grad_norm": 0.166015625,
+      "learning_rate": 4.6348364599668737e-07,
+      "loss": 1.1629,
+      "step": 14555
+    },
+    {
+      "epoch": 0.9727418492784607,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.523376306681537e-07,
+      "loss": 1.1369,
+      "step": 14560
+    },
+    {
+      "epoch": 0.9730758952431855,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.413269693409161e-07,
+      "loss": 1.2048,
+      "step": 14565
+    },
+    {
+      "epoch": 0.9734099412079102,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.3045167698603226e-07,
+      "loss": 1.2306,
+      "step": 14570
+    },
+    {
+      "epoch": 0.973743987172635,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.197117683905627e-07,
+      "loss": 1.1437,
+      "step": 14575
+    },
+    {
+      "epoch": 0.9740780331373597,
+      "grad_norm": 0.166015625,
+      "learning_rate": 4.0910725815742626e-07,
+      "loss": 1.0792,
+      "step": 14580
+    },
+    {
+      "epoch": 0.9744120791020845,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.986381607055112e-07,
+      "loss": 1.2339,
+      "step": 14585
+    },
+    {
+      "epoch": 0.9747461250668092,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.883044902695199e-07,
+      "loss": 1.1773,
+      "step": 14590
+    },
+    {
+      "epoch": 0.975080171031534,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.781062609000463e-07,
+      "loss": 1.1915,
+      "step": 14595
+    },
+    {
+      "epoch": 0.9754142169962586,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.680434864635429e-07,
+      "loss": 1.1747,
+      "step": 14600
+    },
+    {
+      "epoch": 0.9757482629609834,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 3.5811618064226505e-07,
+      "loss": 1.1539,
+      "step": 14605
+    },
+    {
+      "epoch": 0.9760823089257081,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.483243569342487e-07,
+      "loss": 1.2025,
+      "step": 14610
+    },
+    {
+      "epoch": 0.9764163548904329,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.3866802865337723e-07,
+      "loss": 1.1391,
+      "step": 14615
+    },
+    {
+      "epoch": 0.9767504008551576,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 3.29147208929248e-07,
+      "loss": 1.1749,
+      "step": 14620
+    },
+    {
+      "epoch": 0.9770844468198824,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.1976191070722803e-07,
+      "loss": 1.1628,
+      "step": 14625
+    },
+    {
+      "epoch": 0.9774184927846071,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.1051214674843174e-07,
+      "loss": 1.1674,
+      "step": 14630
+    },
+    {
+      "epoch": 0.9777525387493319,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.013979296296543e-07,
+      "loss": 1.2079,
+      "step": 14635
+    },
+    {
+      "epoch": 0.9780865847140566,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.92419271743416e-07,
+      "loss": 1.2202,
+      "step": 14640
+    },
+    {
+      "epoch": 0.9784206306787814,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.8357618529792905e-07,
+      "loss": 1.2067,
+      "step": 14645
+    },
+    {
+      "epoch": 0.9787546766435061,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.74868682317031e-07,
+      "loss": 1.191,
+      "step": 14650
+    },
+    {
+      "epoch": 0.9790887226082309,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.6629677464024006e-07,
+      "loss": 1.1576,
+      "step": 14655
+    },
+    {
+      "epoch": 0.9794227685729556,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.5786047392268864e-07,
+      "loss": 1.1496,
+      "step": 14660
+    },
+    {
+      "epoch": 0.9797568145376804,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.4955979163514554e-07,
+      "loss": 1.2359,
+      "step": 14665
+    },
+    {
+      "epoch": 0.9800908605024051,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.413947390639715e-07,
+      "loss": 1.2322,
+      "step": 14670
+    },
+    {
+      "epoch": 0.9804249064671299,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.3336532731108584e-07,
+      "loss": 1.2106,
+      "step": 14675
+    },
+    {
+      "epoch": 0.9807589524318546,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.2547156729403329e-07,
+      "loss": 1.1849,
+      "step": 14680
+    },
+    {
+      "epoch": 0.9810929983965794,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.177134697458616e-07,
+      "loss": 1.225,
+      "step": 14685
+    },
+    {
+      "epoch": 0.9814270443613041,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.100910452151883e-07,
+      "loss": 1.1283,
+      "step": 14690
+    },
+    {
+      "epoch": 0.9817610903260289,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.0260430406615626e-07,
+      "loss": 1.1439,
+      "step": 14695
+    },
+    {
+      "epoch": 0.9820951362907536,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.9525325647841152e-07,
+      "loss": 1.1861,
+      "step": 14700
+    },
+    {
+      "epoch": 0.9824291822554784,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.8803791244710322e-07,
+      "loss": 1.2404,
+      "step": 14705
+    },
+    {
+      "epoch": 0.9827632282202031,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.8095828178286145e-07,
+      "loss": 1.2448,
+      "step": 14710
+    },
+    {
+      "epoch": 0.9830972741849279,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.7401437411179722e-07,
+      "loss": 1.2503,
+      "step": 14715
+    },
+    {
+      "epoch": 0.9834313201496526,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.6720619887548028e-07,
+      "loss": 1.2129,
+      "step": 14720
+    },
+    {
+      "epoch": 0.9837653661143774,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.605337653309058e-07,
+      "loss": 1.1759,
+      "step": 14725
+    },
+    {
+      "epoch": 0.9840994120791021,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.539970825505277e-07,
+      "loss": 1.1244,
+      "step": 14730
+    },
+    {
+      "epoch": 0.9844334580438269,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.4759615942220306e-07,
+      "loss": 1.1535,
+      "step": 14735
+    },
+    {
+      "epoch": 0.9847675040085516,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.4133100464922555e-07,
+      "loss": 1.1768,
+      "step": 14740
+    },
+    {
+      "epoch": 0.9851015499732764,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 1.352016267502365e-07,
+      "loss": 1.2574,
+      "step": 14745
+    },
+    {
+      "epoch": 0.9854355959380011,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 1.292080340593249e-07,
+      "loss": 1.1745,
+      "step": 14750
+    },
+    {
+      "epoch": 0.9857696419027259,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.2335023472589414e-07,
+      "loss": 1.2315,
+      "step": 14755
+    },
+    {
+      "epoch": 0.9861036878674505,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.1762823671475076e-07,
+      "loss": 1.2231,
+      "step": 14760
+    },
+    {
+      "epoch": 0.9864377338321753,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.12042047806038e-07,
+      "loss": 1.2253,
+      "step": 14765
+    },
+    {
+      "epoch": 0.9867717797969,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.0659167559523564e-07,
+      "loss": 1.1431,
+      "step": 14770
+    },
+    {
+      "epoch": 0.9871058257616248,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.012771274931823e-07,
+      "loss": 1.1837,
+      "step": 14775
+    },
+    {
+      "epoch": 0.9874398717263495,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.609841072599768e-08,
+      "loss": 1.2645,
+      "step": 14780
+    },
+    {
+      "epoch": 0.9877739176910743,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.105553233513809e-08,
+      "loss": 1.2164,
+      "step": 14785
+    },
+    {
+      "epoch": 0.988107963655799,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.614849917737423e-08,
+      "loss": 1.2218,
+      "step": 14790
+    },
+    {
+      "epoch": 0.9884420096205238,
+      "grad_norm": 0.16796875,
+      "learning_rate": 8.137731792472458e-08,
+      "loss": 1.1711,
+      "step": 14795
+    },
+    {
+      "epoch": 0.9887760555852485,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.674199506453317e-08,
+      "loss": 1.2162,
+      "step": 14800
+    },
+    {
+      "epoch": 0.9891101015499733,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.224253689940286e-08,
+      "loss": 1.2323,
+      "step": 14805
+    },
+    {
+      "epoch": 0.989444147514698,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.787894954720653e-08,
+      "loss": 1.1196,
+      "step": 14810
+    },
+    {
+      "epoch": 0.9897781934794228,
+      "grad_norm": 0.158203125,
+      "learning_rate": 6.365123894107594e-08,
+      "loss": 1.2014,
+      "step": 14815
+    },
+    {
+      "epoch": 0.9901122394441475,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.955941082936845e-08,
+      "loss": 1.1848,
+      "step": 14820
+    },
+    {
+      "epoch": 0.9904462854088723,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 5.560347077572248e-08,
+      "loss": 1.1779,
+      "step": 14825
+    },
+    {
+      "epoch": 0.990780331373597,
+      "grad_norm": 0.162109375,
+      "learning_rate": 5.1783424158990954e-08,
+      "loss": 1.208,
+      "step": 14830
+    },
+    {
+      "epoch": 0.9911143773383218,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.809927617324128e-08,
+      "loss": 1.2122,
+      "step": 14835
+    },
+    {
+      "epoch": 0.9914484233030465,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.4551031827788636e-08,
+      "loss": 1.1781,
+      "step": 14840
+    },
+    {
+      "epoch": 0.9917824692677712,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.113869594712938e-08,
+      "loss": 1.1899,
+      "step": 14845
+    },
+    {
+      "epoch": 0.992116515232496,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 3.7862273171007657e-08,
+      "loss": 1.1715,
+      "step": 14850
+    },
+    {
+      "epoch": 0.9924505611972207,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 3.472176795432658e-08,
+      "loss": 1.1051,
+      "step": 14855
+    },
+    {
+      "epoch": 0.9927846071619455,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.171718456720374e-08,
+      "loss": 1.1639,
+      "step": 14860
+    },
+    {
+      "epoch": 0.9931186531266702,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 2.8848527094949006e-08,
+      "loss": 1.1834,
+      "step": 14865
+    },
+    {
+      "epoch": 0.993452699091395,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.611579943804232e-08,
+      "loss": 1.274,
+      "step": 14870
+    },
+    {
+      "epoch": 0.9937867450561197,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 2.3519005312144792e-08,
+      "loss": 1.1908,
+      "step": 14875
+    },
+    {
+      "epoch": 0.9941207910208445,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.105814824810981e-08,
+      "loss": 1.2198,
+      "step": 14880
+    },
+    {
+      "epoch": 0.9944548369855692,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.873323159191642e-08,
+      "loss": 1.1147,
+      "step": 14885
+    },
+    {
+      "epoch": 0.994788882950294,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.6544258504747058e-08,
+      "loss": 1.1683,
+      "step": 14890
+    },
+    {
+      "epoch": 0.9951229289150187,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.4491231962920904e-08,
+      "loss": 1.2365,
+      "step": 14895
+    },
+    {
+      "epoch": 0.9954569748797435,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.2574154757916123e-08,
+      "loss": 1.1823,
+      "step": 14900
+    },
+    {
+      "epoch": 0.9957910208444682,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.0793029496369844e-08,
+      "loss": 1.1889,
+      "step": 14905
+    },
+    {
+      "epoch": 0.996125066809193,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.147858600067061e-09,
+      "loss": 1.2029,
+      "step": 14910
+    },
+    {
+      "epoch": 0.9964591127739177,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.63864430591843e-09,
+      "loss": 1.199,
+      "step": 14915
+    },
+    {
+      "epoch": 0.9967931587386424,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.265388666004679e-09,
+      "loss": 1.1928,
+      "step": 14920
+    },
+    {
+      "epoch": 0.9971272047033671,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 5.028093547498891e-09,
+      "loss": 1.2173,
+      "step": 14925
+    },
+    {
+      "epoch": 0.9974612506680919,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.926760632777526e-09,
+      "loss": 1.1781,
+      "step": 14930
+    },
+    {
+      "epoch": 0.9977952966328166,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.9613914192760937e-09,
+      "loss": 1.1657,
+      "step": 14935
+    },
+    {
+      "epoch": 0.9981293425975414,
+      "grad_norm": 0.1953125,
+      "learning_rate": 2.131987219622378e-09,
+      "loss": 1.1723,
+      "step": 14940
+    },
+    {
+      "epoch": 0.9984633885622661,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.4385491615365177e-09,
+      "loss": 1.1883,
+      "step": 14945
+    },
+    {
+      "epoch": 0.9987974345269909,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.810781878865193e-10,
+      "loss": 1.23,
+      "step": 14950
+    },
+    {
+      "epoch": 0.9991314804917156,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.595750566660506e-10,
+      "loss": 1.1041,
+      "step": 14955
+    },
+    {
+      "epoch": 0.9994655264564404,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.7404034097223688e-10,
+      "loss": 1.2243,
+      "step": 14960
+    },
+    {
+      "epoch": 0.9997995724211651,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.4474429050069803e-11,
+      "loss": 1.2883,
+      "step": 14965
+    },
+    {
+      "epoch": 1.0,
+      "step": 14968,
+      "total_flos": 1.1104476526052114e+19,
+      "train_loss": 0.0,
+      "train_runtime": 10.0721,
+      "train_samples_per_second": 11888.555,
+      "train_steps_per_second": 1486.082
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 14968,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.1104476526052114e+19,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}