diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,37311 @@
+{
+  "best_metric": 0.29454880952835083,
+  "best_model_checkpoint": "./cifar100_outputs/checkpoint-47817",
+  "epoch": 10.0,
+  "eval_steps": 500,
+  "global_step": 53130,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1900806427001953,
+      "learning_rate": 1.9996235648409565e-05,
+      "loss": 4.6367,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6411144733428955,
+      "learning_rate": 1.9992471296819125e-05,
+      "loss": 4.5998,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.594529151916504,
+      "learning_rate": 1.9988706945228688e-05,
+      "loss": 4.5914,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5973572731018066,
+      "learning_rate": 1.9984942593638247e-05,
+      "loss": 4.6124,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.887483596801758,
+      "learning_rate": 1.998117824204781e-05,
+      "loss": 4.6123,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.584811210632324,
+      "learning_rate": 1.997741389045737e-05,
+      "loss": 4.6138,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9465954303741455,
+      "learning_rate": 1.9973649538866934e-05,
+      "loss": 4.5905,
+      "step": 70
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4353935718536377,
+      "learning_rate": 1.9969885187276493e-05,
+      "loss": 4.5707,
+      "step": 80
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8148584365844727,
+      "learning_rate": 1.9966120835686053e-05,
+      "loss": 4.5688,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.962184429168701,
+      "learning_rate": 1.9962356484095616e-05,
+      "loss": 4.5784,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.464444637298584,
+      "learning_rate": 1.9958592132505176e-05,
+      "loss": 4.5557,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1651523113250732,
+      "learning_rate": 1.995482778091474e-05,
+      "loss": 4.5485,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.603081226348877,
+      "learning_rate": 1.99510634293243e-05,
+      "loss": 4.5642,
+      "step": 130
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5633766651153564,
+      "learning_rate": 1.9947299077733862e-05,
+      "loss": 4.5619,
+      "step": 140
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.072338342666626,
+      "learning_rate": 1.9943534726143422e-05,
+      "loss": 4.5694,
+      "step": 150
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9631261825561523,
+      "learning_rate": 1.9939770374552985e-05,
+      "loss": 4.554,
+      "step": 160
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.998347043991089,
+      "learning_rate": 1.9936006022962545e-05,
+      "loss": 4.5413,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.716461420059204,
+      "learning_rate": 1.9932241671372108e-05,
+      "loss": 4.5095,
+      "step": 180
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.613055944442749,
+      "learning_rate": 1.992847731978167e-05,
+      "loss": 4.5277,
+      "step": 190
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.56919002532959,
+      "learning_rate": 1.992471296819123e-05,
+      "loss": 4.5149,
+      "step": 200
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0990872383117676,
+      "learning_rate": 1.9920948616600794e-05,
+      "loss": 4.5029,
+      "step": 210
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.658292770385742,
+      "learning_rate": 1.9917184265010354e-05,
+      "loss": 4.4927,
+      "step": 220
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.613175868988037,
+      "learning_rate": 1.9913419913419917e-05,
+      "loss": 4.46,
+      "step": 230
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9422857761383057,
+      "learning_rate": 1.9909655561829477e-05,
+      "loss": 4.4698,
+      "step": 240
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.384212017059326,
+      "learning_rate": 1.990589121023904e-05,
+      "loss": 4.4693,
+      "step": 250
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8845982551574707,
+      "learning_rate": 1.99021268586486e-05,
+      "loss": 4.4814,
+      "step": 260
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9230496883392334,
+      "learning_rate": 1.989836250705816e-05,
+      "loss": 4.4504,
+      "step": 270
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3872103691101074,
+      "learning_rate": 1.9894598155467723e-05,
+      "loss": 4.4218,
+      "step": 280
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.033067226409912,
+      "learning_rate": 1.9890833803877282e-05,
+      "loss": 4.4298,
+      "step": 290
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.819579601287842,
+      "learning_rate": 1.9887069452286845e-05,
+      "loss": 4.4118,
+      "step": 300
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8507134914398193,
+      "learning_rate": 1.9883305100696405e-05,
+      "loss": 4.4365,
+      "step": 310
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.762634038925171,
+      "learning_rate": 1.987954074910597e-05,
+      "loss": 4.4232,
+      "step": 320
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0568439960479736,
+      "learning_rate": 1.9875776397515528e-05,
+      "loss": 4.3622,
+      "step": 330
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.941974639892578,
+      "learning_rate": 1.987201204592509e-05,
+      "loss": 4.4076,
+      "step": 340
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0038816928863525,
+      "learning_rate": 1.986824769433465e-05,
+      "loss": 4.3652,
+      "step": 350
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.681337356567383,
+      "learning_rate": 1.9864483342744214e-05,
+      "loss": 4.3819,
+      "step": 360
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8940606117248535,
+      "learning_rate": 1.9860718991153774e-05,
+      "loss": 4.3638,
+      "step": 370
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.30637526512146,
+      "learning_rate": 1.9856954639563337e-05,
+      "loss": 4.385,
+      "step": 380
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1481170654296875,
+      "learning_rate": 1.98531902879729e-05,
+      "loss": 4.3957,
+      "step": 390
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.795309543609619,
+      "learning_rate": 1.984942593638246e-05,
+      "loss": 4.3601,
+      "step": 400
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1323978900909424,
+      "learning_rate": 1.9845661584792023e-05,
+      "loss": 4.3536,
+      "step": 410
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0077526569366455,
+      "learning_rate": 1.9841897233201583e-05,
+      "loss": 4.3306,
+      "step": 420
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.095212936401367,
+      "learning_rate": 1.9838132881611146e-05,
+      "loss": 4.3473,
+      "step": 430
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.200291633605957,
+      "learning_rate": 1.9834368530020706e-05,
+      "loss": 4.2868,
+      "step": 440
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.210118532180786,
+      "learning_rate": 1.9830604178430266e-05,
+      "loss": 4.3114,
+      "step": 450
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.286947011947632,
+      "learning_rate": 1.982683982683983e-05,
+      "loss": 4.3222,
+      "step": 460
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8678014278411865,
+      "learning_rate": 1.982307547524939e-05,
+      "loss": 4.2983,
+      "step": 470
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.726501703262329,
+      "learning_rate": 1.9819311123658952e-05,
+      "loss": 4.2952,
+      "step": 480
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8604493141174316,
+      "learning_rate": 1.981554677206851e-05,
+      "loss": 4.2516,
+      "step": 490
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8486504554748535,
+      "learning_rate": 1.9811782420478075e-05,
+      "loss": 4.2516,
+      "step": 500
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9079558849334717,
+      "learning_rate": 1.9808018068887634e-05,
+      "loss": 4.2404,
+      "step": 510
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.027830123901367,
+      "learning_rate": 1.9804253717297198e-05,
+      "loss": 4.2424,
+      "step": 520
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.154351234436035,
+      "learning_rate": 1.9800489365706757e-05,
+      "loss": 4.21,
+      "step": 530
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0818393230438232,
+      "learning_rate": 1.979672501411632e-05,
+      "loss": 4.242,
+      "step": 540
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.877248525619507,
+      "learning_rate": 1.979296066252588e-05,
+      "loss": 4.2019,
+      "step": 550
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.109707832336426,
+      "learning_rate": 1.9789196310935443e-05,
+      "loss": 4.1811,
+      "step": 560
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.032790184020996,
+      "learning_rate": 1.9785431959345007e-05,
+      "loss": 4.1958,
+      "step": 570
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.237541913986206,
+      "learning_rate": 1.9781667607754566e-05,
+      "loss": 4.1735,
+      "step": 580
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.851644992828369,
+      "learning_rate": 1.977790325616413e-05,
+      "loss": 4.1478,
+      "step": 590
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9260008335113525,
+      "learning_rate": 1.977413890457369e-05,
+      "loss": 4.1436,
+      "step": 600
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1182501316070557,
+      "learning_rate": 1.977037455298325e-05,
+      "loss": 4.1768,
+      "step": 610
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.236636161804199,
+      "learning_rate": 1.9766610201392812e-05,
+      "loss": 4.1488,
+      "step": 620
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9359562397003174,
+      "learning_rate": 1.9762845849802372e-05,
+      "loss": 4.1619,
+      "step": 630
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9437806606292725,
+      "learning_rate": 1.9759081498211935e-05,
+      "loss": 4.092,
+      "step": 640
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.566126585006714,
+      "learning_rate": 1.9755317146621495e-05,
+      "loss": 4.1258,
+      "step": 650
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.742216110229492,
+      "learning_rate": 1.9751552795031058e-05,
+      "loss": 4.0587,
+      "step": 660
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9798004627227783,
+      "learning_rate": 1.9747788443440618e-05,
+      "loss": 4.0615,
+      "step": 670
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.074636697769165,
+      "learning_rate": 1.974402409185018e-05,
+      "loss": 4.0414,
+      "step": 680
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1997222900390625,
+      "learning_rate": 1.974025974025974e-05,
+      "loss": 4.0918,
+      "step": 690
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.321722984313965,
+      "learning_rate": 1.9736495388669304e-05,
+      "loss": 4.0527,
+      "step": 700
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0019032955169678,
+      "learning_rate": 1.9732731037078864e-05,
+      "loss": 3.9662,
+      "step": 710
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8338370323181152,
+      "learning_rate": 1.9728966685488427e-05,
+      "loss": 4.0178,
+      "step": 720
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2571487426757812,
+      "learning_rate": 1.9725202333897987e-05,
+      "loss": 4.0006,
+      "step": 730
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9352447986602783,
+      "learning_rate": 1.972143798230755e-05,
+      "loss": 4.0078,
+      "step": 740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.793808460235596,
+      "learning_rate": 1.9717673630717113e-05,
+      "loss": 4.0501,
+      "step": 750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.661139488220215,
+      "learning_rate": 1.9713909279126673e-05,
+      "loss": 3.9689,
+      "step": 760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5628387928009033,
+      "learning_rate": 1.9710144927536236e-05,
+      "loss": 3.9937,
+      "step": 770
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0853545665740967,
+      "learning_rate": 1.9706380575945796e-05,
+      "loss": 3.9623,
+      "step": 780
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3460278511047363,
+      "learning_rate": 1.9702616224355355e-05,
+      "loss": 4.0009,
+      "step": 790
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.079176425933838,
+      "learning_rate": 1.969885187276492e-05,
+      "loss": 3.9128,
+      "step": 800
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0661251544952393,
+      "learning_rate": 1.9695087521174478e-05,
+      "loss": 3.9185,
+      "step": 810
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.881580352783203,
+      "learning_rate": 1.969132316958404e-05,
+      "loss": 3.8848,
+      "step": 820
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.219837665557861,
+      "learning_rate": 1.96875588179936e-05,
+      "loss": 3.9397,
+      "step": 830
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.501627206802368,
+      "learning_rate": 1.9683794466403164e-05,
+      "loss": 3.9385,
+      "step": 840
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5525853633880615,
+      "learning_rate": 1.9680030114812724e-05,
+      "loss": 3.9145,
+      "step": 850
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5347719192504883,
+      "learning_rate": 1.9676265763222287e-05,
+      "loss": 3.9678,
+      "step": 860
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.558157920837402,
+      "learning_rate": 1.9672501411631847e-05,
+      "loss": 3.8212,
+      "step": 870
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4307472705841064,
+      "learning_rate": 1.966873706004141e-05,
+      "loss": 3.8341,
+      "step": 880
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.318007469177246,
+      "learning_rate": 1.966497270845097e-05,
+      "loss": 3.8516,
+      "step": 890
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.074690341949463,
+      "learning_rate": 1.9661208356860533e-05,
+      "loss": 3.8625,
+      "step": 900
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.376126289367676,
+      "learning_rate": 1.9657444005270093e-05,
+      "loss": 3.8391,
+      "step": 910
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.645514011383057,
+      "learning_rate": 1.9653679653679656e-05,
+      "loss": 3.8403,
+      "step": 920
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.457108497619629,
+      "learning_rate": 1.964991530208922e-05,
+      "loss": 3.8908,
+      "step": 930
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9738640785217285,
+      "learning_rate": 1.964615095049878e-05,
+      "loss": 3.7746,
+      "step": 940
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.399070739746094,
+      "learning_rate": 1.9642386598908342e-05,
+      "loss": 3.7559,
+      "step": 950
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.6740546226501465,
+      "learning_rate": 1.96386222473179e-05,
+      "loss": 3.7348,
+      "step": 960
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3767216205596924,
+      "learning_rate": 1.963485789572746e-05,
+      "loss": 3.6984,
+      "step": 970
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3283941745758057,
+      "learning_rate": 1.9631093544137025e-05,
+      "loss": 3.756,
+      "step": 980
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.651248931884766,
+      "learning_rate": 1.9627329192546585e-05,
+      "loss": 3.7782,
+      "step": 990
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.265860080718994,
+      "learning_rate": 1.9623564840956148e-05,
+      "loss": 3.787,
+      "step": 1000
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8226921558380127,
+      "learning_rate": 1.9619800489365707e-05,
+      "loss": 3.8101,
+      "step": 1010
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2111263275146484,
+      "learning_rate": 1.961603613777527e-05,
+      "loss": 3.737,
+      "step": 1020
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.9517107009887695,
+      "learning_rate": 1.961227178618483e-05,
+      "loss": 3.7126,
+      "step": 1030
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.050525665283203,
+      "learning_rate": 1.9608507434594394e-05,
+      "loss": 3.7004,
+      "step": 1040
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8831429481506348,
+      "learning_rate": 1.9604743083003953e-05,
+      "loss": 3.6664,
+      "step": 1050
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.123471260070801,
+      "learning_rate": 1.9600978731413516e-05,
+      "loss": 3.7305,
+      "step": 1060
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.836909770965576,
+      "learning_rate": 1.9597214379823076e-05,
+      "loss": 3.7419,
+      "step": 1070
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.690175771713257,
+      "learning_rate": 1.959345002823264e-05,
+      "loss": 3.6338,
+      "step": 1080
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.598996639251709,
+      "learning_rate": 1.95896856766422e-05,
+      "loss": 3.7188,
+      "step": 1090
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.010788917541504,
+      "learning_rate": 1.9585921325051762e-05,
+      "loss": 3.7787,
+      "step": 1100
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.156946182250977,
+      "learning_rate": 1.9582156973461322e-05,
+      "loss": 3.6394,
+      "step": 1110
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7809722423553467,
+      "learning_rate": 1.9578392621870885e-05,
+      "loss": 3.6945,
+      "step": 1120
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.646225929260254,
+      "learning_rate": 1.957462827028045e-05,
+      "loss": 3.5862,
+      "step": 1130
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2124147415161133,
+      "learning_rate": 1.9570863918690005e-05,
+      "loss": 3.6443,
+      "step": 1140
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.297073364257812,
+      "learning_rate": 1.9567099567099568e-05,
+      "loss": 3.7029,
+      "step": 1150
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.9652862548828125,
+      "learning_rate": 1.956333521550913e-05,
+      "loss": 3.6561,
+      "step": 1160
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.712035179138184,
+      "learning_rate": 1.955957086391869e-05,
+      "loss": 3.5105,
+      "step": 1170
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.600637912750244,
+      "learning_rate": 1.9555806512328254e-05,
+      "loss": 3.5531,
+      "step": 1180
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.265107154846191,
+      "learning_rate": 1.9552042160737814e-05,
+      "loss": 3.6188,
+      "step": 1190
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.627386093139648,
+      "learning_rate": 1.9548277809147377e-05,
+      "loss": 3.5911,
+      "step": 1200
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.785702228546143,
+      "learning_rate": 1.9544513457556937e-05,
+      "loss": 3.5837,
+      "step": 1210
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3369293212890625,
+      "learning_rate": 1.95407491059665e-05,
+      "loss": 3.6222,
+      "step": 1220
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.7859039306640625,
+      "learning_rate": 1.953698475437606e-05,
+      "loss": 3.5098,
+      "step": 1230
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.6498477458953857,
+      "learning_rate": 1.9533220402785623e-05,
+      "loss": 3.4831,
+      "step": 1240
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.522767543792725,
+      "learning_rate": 1.9529456051195183e-05,
+      "loss": 3.6801,
+      "step": 1250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2972137928009033,
+      "learning_rate": 1.9525691699604746e-05,
+      "loss": 3.3404,
+      "step": 1260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.638528823852539,
+      "learning_rate": 1.9521927348014305e-05,
+      "loss": 3.5866,
+      "step": 1270
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.175476551055908,
+      "learning_rate": 1.951816299642387e-05,
+      "loss": 3.509,
+      "step": 1280
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4966530799865723,
+      "learning_rate": 1.951439864483343e-05,
+      "loss": 3.4481,
+      "step": 1290
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.191817283630371,
+      "learning_rate": 1.951063429324299e-05,
+      "loss": 3.622,
+      "step": 1300
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.43389368057251,
+      "learning_rate": 1.950686994165255e-05,
+      "loss": 3.5212,
+      "step": 1310
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.5795392990112305,
+      "learning_rate": 1.950310559006211e-05,
+      "loss": 3.6054,
+      "step": 1320
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8179938793182373,
+      "learning_rate": 1.9499341238471674e-05,
+      "loss": 3.3792,
+      "step": 1330
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.200837135314941,
+      "learning_rate": 1.9495576886881234e-05,
+      "loss": 3.5091,
+      "step": 1340
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6297414302825928,
+      "learning_rate": 1.9491812535290797e-05,
+      "loss": 3.4068,
+      "step": 1350
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.7957377433776855,
+      "learning_rate": 1.948804818370036e-05,
+      "loss": 3.4862,
+      "step": 1360
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.713080644607544,
+      "learning_rate": 1.948428383210992e-05,
+      "loss": 3.4861,
+      "step": 1370
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.54100227355957,
+      "learning_rate": 1.9480519480519483e-05,
+      "loss": 3.4529,
+      "step": 1380
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8583154678344727,
+      "learning_rate": 1.9476755128929043e-05,
+      "loss": 3.4531,
+      "step": 1390
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.6709675788879395,
+      "learning_rate": 1.9472990777338606e-05,
+      "loss": 3.4798,
+      "step": 1400
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.9218077659606934,
+      "learning_rate": 1.9469226425748166e-05,
+      "loss": 3.4589,
+      "step": 1410
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0315423011779785,
+      "learning_rate": 1.946546207415773e-05,
+      "loss": 3.3474,
+      "step": 1420
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.207186222076416,
+      "learning_rate": 1.946169772256729e-05,
+      "loss": 3.3297,
+      "step": 1430
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.950998544692993,
+      "learning_rate": 1.9457933370976852e-05,
+      "loss": 3.4374,
+      "step": 1440
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.596767902374268,
+      "learning_rate": 1.9454169019386412e-05,
+      "loss": 3.417,
+      "step": 1450
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.453456878662109,
+      "learning_rate": 1.9450404667795975e-05,
+      "loss": 3.2764,
+      "step": 1460
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.382648944854736,
+      "learning_rate": 1.9446640316205535e-05,
+      "loss": 3.4189,
+      "step": 1470
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.689038276672363,
+      "learning_rate": 1.9442875964615098e-05,
+      "loss": 3.4015,
+      "step": 1480
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.036463737487793,
+      "learning_rate": 1.9439111613024658e-05,
+      "loss": 3.3156,
+      "step": 1490
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.8814964294433594,
+      "learning_rate": 1.9435347261434217e-05,
+      "loss": 3.1903,
+      "step": 1500
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.564530849456787,
+      "learning_rate": 1.943158290984378e-05,
+      "loss": 3.3632,
+      "step": 1510
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.282833576202393,
+      "learning_rate": 1.942781855825334e-05,
+      "loss": 3.4336,
+      "step": 1520
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.326153755187988,
+      "learning_rate": 1.9424054206662903e-05,
+      "loss": 3.4133,
+      "step": 1530
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.867426872253418,
+      "learning_rate": 1.9420289855072467e-05,
+      "loss": 3.3124,
+      "step": 1540
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.446542739868164,
+      "learning_rate": 1.9416525503482026e-05,
+      "loss": 3.2558,
+      "step": 1550
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.327835083007812,
+      "learning_rate": 1.941276115189159e-05,
+      "loss": 3.3257,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.354964256286621,
+      "learning_rate": 1.940899680030115e-05,
+      "loss": 3.2928,
+      "step": 1570
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.222498893737793,
+      "learning_rate": 1.9405232448710712e-05,
+      "loss": 3.3359,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6612141132354736,
+      "learning_rate": 1.9401468097120272e-05,
+      "loss": 3.2173,
+      "step": 1590
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.745842695236206,
+      "learning_rate": 1.9397703745529835e-05,
+      "loss": 3.1372,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.753840446472168,
+      "learning_rate": 1.9393939393939395e-05,
+      "loss": 3.321,
+      "step": 1610
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.043251037597656,
+      "learning_rate": 1.9390175042348958e-05,
+      "loss": 3.3141,
+      "step": 1620
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.991616249084473,
+      "learning_rate": 1.9386410690758518e-05,
+      "loss": 3.2403,
+      "step": 1630
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.049126148223877,
+      "learning_rate": 1.938264633916808e-05,
+      "loss": 3.2457,
+      "step": 1640
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2904117107391357,
+      "learning_rate": 1.937888198757764e-05,
+      "loss": 3.3252,
+      "step": 1650
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.309184551239014,
+      "learning_rate": 1.93751176359872e-05,
+      "loss": 3.1543,
+      "step": 1660
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.324444532394409,
+      "learning_rate": 1.9371353284396764e-05,
+      "loss": 3.3012,
+      "step": 1670
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.101102352142334,
+      "learning_rate": 1.9367588932806324e-05,
+      "loss": 3.2415,
+      "step": 1680
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.955221652984619,
+      "learning_rate": 1.9363824581215887e-05,
+      "loss": 3.1479,
+      "step": 1690
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.092071056365967,
+      "learning_rate": 1.9360060229625447e-05,
+      "loss": 3.2071,
+      "step": 1700
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.205142021179199,
+      "learning_rate": 1.935629587803501e-05,
+      "loss": 3.1963,
+      "step": 1710
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.192953109741211,
+      "learning_rate": 1.9352531526444573e-05,
+      "loss": 3.0651,
+      "step": 1720
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.651326656341553,
+      "learning_rate": 1.9348767174854133e-05,
+      "loss": 3.3063,
+      "step": 1730
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.780113697052002,
+      "learning_rate": 1.9345002823263696e-05,
+      "loss": 3.1667,
+      "step": 1740
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.112650394439697,
+      "learning_rate": 1.9341238471673256e-05,
+      "loss": 3.2213,
+      "step": 1750
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.035951137542725,
+      "learning_rate": 1.933747412008282e-05,
+      "loss": 3.2067,
+      "step": 1760
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.030355930328369,
+      "learning_rate": 1.933370976849238e-05,
+      "loss": 3.1413,
+      "step": 1770
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.994670391082764,
+      "learning_rate": 1.932994541690194e-05,
+      "loss": 3.1571,
+      "step": 1780
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.542640686035156,
+      "learning_rate": 1.93261810653115e-05,
+      "loss": 3.0721,
+      "step": 1790
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.707923173904419,
+      "learning_rate": 1.9322416713721065e-05,
+      "loss": 3.0077,
+      "step": 1800
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.954961776733398,
+      "learning_rate": 1.9318652362130624e-05,
+      "loss": 3.1853,
+      "step": 1810
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.040740013122559,
+      "learning_rate": 1.9314888010540187e-05,
+      "loss": 3.1621,
+      "step": 1820
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.269940137863159,
+      "learning_rate": 1.9311123658949747e-05,
+      "loss": 3.0776,
+      "step": 1830
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.951628684997559,
+      "learning_rate": 1.9307359307359307e-05,
+      "loss": 3.0865,
+      "step": 1840
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.079604625701904,
+      "learning_rate": 1.930359495576887e-05,
+      "loss": 3.1071,
+      "step": 1850
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.9071431159973145,
+      "learning_rate": 1.929983060417843e-05,
+      "loss": 3.1679,
+      "step": 1860
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.998239994049072,
+      "learning_rate": 1.9296066252587993e-05,
+      "loss": 2.9328,
+      "step": 1870
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.7196455001831055,
+      "learning_rate": 1.9292301900997553e-05,
+      "loss": 3.1693,
+      "step": 1880
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.726539611816406,
+      "learning_rate": 1.9288537549407116e-05,
+      "loss": 3.119,
+      "step": 1890
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.929136276245117,
+      "learning_rate": 1.9284773197816676e-05,
+      "loss": 3.0396,
+      "step": 1900
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6181488037109375,
+      "learning_rate": 1.928100884622624e-05,
+      "loss": 2.9703,
+      "step": 1910
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.3397040367126465,
+      "learning_rate": 1.9277244494635802e-05,
+      "loss": 3.1033,
+      "step": 1920
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.7239484786987305,
+      "learning_rate": 1.9273480143045362e-05,
+      "loss": 3.001,
+      "step": 1930
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.317354440689087,
+      "learning_rate": 1.9269715791454925e-05,
+      "loss": 2.9419,
+      "step": 1940
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.748729705810547,
+      "learning_rate": 1.9265951439864485e-05,
+      "loss": 3.1594,
+      "step": 1950
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.6273064613342285,
+      "learning_rate": 1.9262187088274048e-05,
+      "loss": 2.9821,
+      "step": 1960
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.60408878326416,
+      "learning_rate": 1.9258422736683608e-05,
+      "loss": 3.1565,
+      "step": 1970
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.767194747924805,
+      "learning_rate": 1.925465838509317e-05,
+      "loss": 2.9479,
+      "step": 1980
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 13.294413566589355,
+      "learning_rate": 1.925089403350273e-05,
+      "loss": 3.0082,
+      "step": 1990
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.2884979248046875,
+      "learning_rate": 1.9247129681912294e-05,
+      "loss": 2.9004,
+      "step": 2000
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.24321174621582,
+      "learning_rate": 1.9243365330321854e-05,
+      "loss": 2.8892,
+      "step": 2010
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.944282054901123,
+      "learning_rate": 1.9239600978731413e-05,
+      "loss": 3.0362,
+      "step": 2020
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.824367046356201,
+      "learning_rate": 1.9235836627140976e-05,
+      "loss": 3.0755,
+      "step": 2030
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.201323509216309,
+      "learning_rate": 1.9232072275550536e-05,
+      "loss": 2.9256,
+      "step": 2040
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.465519428253174,
+      "learning_rate": 1.92283079239601e-05,
+      "loss": 2.8423,
+      "step": 2050
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.563433647155762,
+      "learning_rate": 1.922454357236966e-05,
+      "loss": 3.063,
+      "step": 2060
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.595127582550049,
+      "learning_rate": 1.9220779220779222e-05,
+      "loss": 2.89,
+      "step": 2070
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.29541540145874,
+      "learning_rate": 1.9217014869188782e-05,
+      "loss": 2.9488,
+      "step": 2080
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.494131088256836,
+      "learning_rate": 1.9213250517598345e-05,
+      "loss": 2.9473,
+      "step": 2090
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.640602111816406,
+      "learning_rate": 1.920948616600791e-05,
+      "loss": 2.9225,
+      "step": 2100
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.552557945251465,
+      "learning_rate": 1.9205721814417468e-05,
+      "loss": 2.9173,
+      "step": 2110
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.50587272644043,
+      "learning_rate": 1.920195746282703e-05,
+      "loss": 2.8535,
+      "step": 2120
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.200937271118164,
+      "learning_rate": 1.919819311123659e-05,
+      "loss": 2.9673,
+      "step": 2130
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.5765485763549805,
+      "learning_rate": 1.9194428759646154e-05,
+      "loss": 2.8064,
+      "step": 2140
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.502497673034668,
+      "learning_rate": 1.9190664408055714e-05,
+      "loss": 2.8947,
+      "step": 2150
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 19.691614151000977,
+      "learning_rate": 1.9186900056465277e-05,
+      "loss": 3.028,
+      "step": 2160
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.6259846687316895,
+      "learning_rate": 1.9183135704874837e-05,
+      "loss": 2.8456,
+      "step": 2170
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.246203422546387,
+      "learning_rate": 1.9179371353284397e-05,
+      "loss": 2.8903,
+      "step": 2180
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.460795879364014,
+      "learning_rate": 1.917560700169396e-05,
+      "loss": 2.9129,
+      "step": 2190
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.957250118255615,
+      "learning_rate": 1.917184265010352e-05,
+      "loss": 2.8096,
+      "step": 2200
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.38149642944336,
+      "learning_rate": 1.9168078298513083e-05,
+      "loss": 2.9384,
+      "step": 2210
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.8041586875915527,
+      "learning_rate": 1.9164313946922643e-05,
+      "loss": 2.7674,
+      "step": 2220
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.8668904304504395,
+      "learning_rate": 1.9160549595332206e-05,
+      "loss": 2.7463,
+      "step": 2230
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.3397698402404785,
+      "learning_rate": 1.9156785243741765e-05,
+      "loss": 2.8838,
+      "step": 2240
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.79181957244873,
+      "learning_rate": 1.915302089215133e-05,
+      "loss": 2.8304,
+      "step": 2250
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.324551582336426,
+      "learning_rate": 1.914925654056089e-05,
+      "loss": 2.8287,
+      "step": 2260
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.550736427307129,
+      "learning_rate": 1.914549218897045e-05,
+      "loss": 2.7671,
+      "step": 2270
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.587412357330322,
+      "learning_rate": 1.9141727837380015e-05,
+      "loss": 3.0649,
+      "step": 2280
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.18621301651001,
+      "learning_rate": 1.9137963485789574e-05,
+      "loss": 2.863,
+      "step": 2290
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.899294376373291,
+      "learning_rate": 1.9134199134199138e-05,
+      "loss": 2.9058,
+      "step": 2300
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.400335788726807,
+      "learning_rate": 1.9130434782608697e-05,
+      "loss": 2.7415,
+      "step": 2310
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 18.84127426147461,
+      "learning_rate": 1.912667043101826e-05,
+      "loss": 2.8094,
+      "step": 2320
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.325419425964355,
+      "learning_rate": 1.912290607942782e-05,
+      "loss": 2.7769,
+      "step": 2330
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.26507568359375,
+      "learning_rate": 1.9119141727837383e-05,
+      "loss": 2.8841,
+      "step": 2340
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.831374168395996,
+      "learning_rate": 1.9115377376246943e-05,
+      "loss": 2.832,
+      "step": 2350
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.735857963562012,
+      "learning_rate": 1.9111613024656503e-05,
+      "loss": 2.7335,
+      "step": 2360
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.749835968017578,
+      "learning_rate": 1.9107848673066066e-05,
+      "loss": 2.7282,
+      "step": 2370
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.772495269775391,
+      "learning_rate": 1.9104084321475626e-05,
+      "loss": 2.6789,
+      "step": 2380
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.5235466957092285,
+      "learning_rate": 1.910031996988519e-05,
+      "loss": 2.8611,
+      "step": 2390
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.360028028488159,
+      "learning_rate": 1.909655561829475e-05,
+      "loss": 2.6936,
+      "step": 2400
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.748390197753906,
+      "learning_rate": 1.9092791266704312e-05,
+      "loss": 2.6904,
+      "step": 2410
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.634457588195801,
+      "learning_rate": 1.9089026915113872e-05,
+      "loss": 2.9342,
+      "step": 2420
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.747628211975098,
+      "learning_rate": 1.9085262563523435e-05,
+      "loss": 2.7683,
+      "step": 2430
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.515625476837158,
+      "learning_rate": 1.9081498211932995e-05,
+      "loss": 2.7442,
+      "step": 2440
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4130609035491943,
+      "learning_rate": 1.9077733860342558e-05,
+      "loss": 2.6693,
+      "step": 2450
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.92776346206665,
+      "learning_rate": 1.907396950875212e-05,
+      "loss": 2.733,
+      "step": 2460
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.010396957397461,
+      "learning_rate": 1.907020515716168e-05,
+      "loss": 2.6891,
+      "step": 2470
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.8040144443511963,
+      "learning_rate": 1.9066440805571244e-05,
+      "loss": 2.5664,
+      "step": 2480
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.726337909698486,
+      "learning_rate": 1.9062676453980804e-05,
+      "loss": 2.7416,
+      "step": 2490
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.524297714233398,
+      "learning_rate": 1.9058912102390367e-05,
+      "loss": 2.6905,
+      "step": 2500
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.6792001724243164,
+      "learning_rate": 1.9055147750799927e-05,
+      "loss": 2.5483,
+      "step": 2510
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.578388690948486,
+      "learning_rate": 1.905138339920949e-05,
+      "loss": 2.7227,
+      "step": 2520
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9775335788726807,
+      "learning_rate": 1.904761904761905e-05,
+      "loss": 2.5024,
+      "step": 2530
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.870952606201172,
+      "learning_rate": 1.904385469602861e-05,
+      "loss": 2.622,
+      "step": 2540
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.206267356872559,
+      "learning_rate": 1.9040090344438172e-05,
+      "loss": 2.8171,
+      "step": 2550
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.320977210998535,
+      "learning_rate": 1.9036325992847732e-05,
+      "loss": 2.6359,
+      "step": 2560
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.017043113708496,
+      "learning_rate": 1.9032561641257295e-05,
+      "loss": 2.6382,
+      "step": 2570
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.2981061935424805,
+      "learning_rate": 1.9028797289666855e-05,
+      "loss": 2.5545,
+      "step": 2580
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.075060844421387,
+      "learning_rate": 1.9025032938076418e-05,
+      "loss": 2.6773,
+      "step": 2590
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.952000379562378,
+      "learning_rate": 1.9021268586485978e-05,
+      "loss": 2.4791,
+      "step": 2600
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.052979469299316,
+      "learning_rate": 1.901750423489554e-05,
+      "loss": 2.7196,
+      "step": 2610
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.638850212097168,
+      "learning_rate": 1.90137398833051e-05,
+      "loss": 2.7139,
+      "step": 2620
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.109199523925781,
+      "learning_rate": 1.9009975531714664e-05,
+      "loss": 2.5936,
+      "step": 2630
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.101054668426514,
+      "learning_rate": 1.9006211180124224e-05,
+      "loss": 2.5606,
+      "step": 2640
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.660862445831299,
+      "learning_rate": 1.9002446828533787e-05,
+      "loss": 2.5895,
+      "step": 2650
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.114608287811279,
+      "learning_rate": 1.899868247694335e-05,
+      "loss": 2.4963,
+      "step": 2660
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.300898551940918,
+      "learning_rate": 1.899491812535291e-05,
+      "loss": 2.7821,
+      "step": 2670
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.744706630706787,
+      "learning_rate": 1.8991153773762473e-05,
+      "loss": 2.5695,
+      "step": 2680
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.060439586639404,
+      "learning_rate": 1.8987389422172033e-05,
+      "loss": 2.4313,
+      "step": 2690
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.207651138305664,
+      "learning_rate": 1.8983625070581596e-05,
+      "loss": 2.4718,
+      "step": 2700
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.556578636169434,
+      "learning_rate": 1.8979860718991156e-05,
+      "loss": 2.5789,
+      "step": 2710
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.473084449768066,
+      "learning_rate": 1.8976096367400716e-05,
+      "loss": 2.6073,
+      "step": 2720
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.552806854248047,
+      "learning_rate": 1.897233201581028e-05,
+      "loss": 2.568,
+      "step": 2730
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.12427043914795,
+      "learning_rate": 1.896856766421984e-05,
+      "loss": 2.4706,
+      "step": 2740
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.1217474937438965,
+      "learning_rate": 1.89648033126294e-05,
+      "loss": 2.4005,
+      "step": 2750
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.9127049446105957,
+      "learning_rate": 1.896103896103896e-05,
+      "loss": 2.517,
+      "step": 2760
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 15.194258689880371,
+      "learning_rate": 1.8957274609448525e-05,
+      "loss": 2.5601,
+      "step": 2770
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 12.40319538116455,
+      "learning_rate": 1.8953510257858084e-05,
+      "loss": 2.3259,
+      "step": 2780
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.481034278869629,
+      "learning_rate": 1.8949745906267647e-05,
+      "loss": 2.4099,
+      "step": 2790
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.578127861022949,
+      "learning_rate": 1.8945981554677207e-05,
+      "loss": 2.416,
+      "step": 2800
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 15.104098320007324,
+      "learning_rate": 1.894221720308677e-05,
+      "loss": 2.4252,
+      "step": 2810
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.515500545501709,
+      "learning_rate": 1.893845285149633e-05,
+      "loss": 2.4741,
+      "step": 2820
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.786322593688965,
+      "learning_rate": 1.8934688499905893e-05,
+      "loss": 2.5193,
+      "step": 2830
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.358415365219116,
+      "learning_rate": 1.8930924148315456e-05,
+      "loss": 2.603,
+      "step": 2840
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.374852180480957,
+      "learning_rate": 1.8927159796725016e-05,
+      "loss": 2.5356,
+      "step": 2850
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.016167640686035,
+      "learning_rate": 1.892339544513458e-05,
+      "loss": 2.4606,
+      "step": 2860
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.327517509460449,
+      "learning_rate": 1.891963109354414e-05,
+      "loss": 2.3619,
+      "step": 2870
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.286544799804688,
+      "learning_rate": 1.89158667419537e-05,
+      "loss": 2.4749,
+      "step": 2880
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.052702903747559,
+      "learning_rate": 1.8912102390363262e-05,
+      "loss": 2.5101,
+      "step": 2890
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 24.803468704223633,
+      "learning_rate": 1.8908338038772822e-05,
+      "loss": 2.4248,
+      "step": 2900
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5042014122009277,
+      "learning_rate": 1.8904573687182385e-05,
+      "loss": 2.3821,
+      "step": 2910
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.960639476776123,
+      "learning_rate": 1.8900809335591945e-05,
+      "loss": 2.3184,
+      "step": 2920
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.575072288513184,
+      "learning_rate": 1.8897044984001508e-05,
+      "loss": 2.5079,
+      "step": 2930
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.817319869995117,
+      "learning_rate": 1.8893280632411068e-05,
+      "loss": 2.3782,
+      "step": 2940
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.542531967163086,
+      "learning_rate": 1.888951628082063e-05,
+      "loss": 2.3612,
+      "step": 2950
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.432719230651855,
+      "learning_rate": 1.888575192923019e-05,
+      "loss": 2.6094,
+      "step": 2960
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.157316207885742,
+      "learning_rate": 1.8881987577639754e-05,
+      "loss": 2.3173,
+      "step": 2970
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.01808500289917,
+      "learning_rate": 1.8878223226049314e-05,
+      "loss": 2.4611,
+      "step": 2980
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.0088324546813965,
+      "learning_rate": 1.8874458874458877e-05,
+      "loss": 2.3906,
+      "step": 2990
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.18332052230835,
+      "learning_rate": 1.8870694522868436e-05,
+      "loss": 2.699,
+      "step": 3000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.170907974243164,
+      "learning_rate": 1.8866930171278e-05,
+      "loss": 2.1945,
+      "step": 3010
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.540265083312988,
+      "learning_rate": 1.8863165819687563e-05,
+      "loss": 2.3214,
+      "step": 3020
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.900252342224121,
+      "learning_rate": 1.8859401468097123e-05,
+      "loss": 2.4009,
+      "step": 3030
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.43867826461792,
+      "learning_rate": 1.8855637116506686e-05,
+      "loss": 2.5143,
+      "step": 3040
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.265262126922607,
+      "learning_rate": 1.8851872764916242e-05,
+      "loss": 2.3603,
+      "step": 3050
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.436113357543945,
+      "learning_rate": 1.8848108413325805e-05,
+      "loss": 2.1487,
+      "step": 3060
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.59679651260376,
+      "learning_rate": 1.884434406173537e-05,
+      "loss": 2.3253,
+      "step": 3070
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.955831527709961,
+      "learning_rate": 1.8840579710144928e-05,
+      "loss": 2.2582,
+      "step": 3080
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.666378021240234,
+      "learning_rate": 1.883681535855449e-05,
+      "loss": 2.2552,
+      "step": 3090
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.003193855285645,
+      "learning_rate": 1.883305100696405e-05,
+      "loss": 2.2303,
+      "step": 3100
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.098687648773193,
+      "learning_rate": 1.8829286655373614e-05,
+      "loss": 2.4797,
+      "step": 3110
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.299673557281494,
+      "learning_rate": 1.8825522303783174e-05,
+      "loss": 2.3256,
+      "step": 3120
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.408268451690674,
+      "learning_rate": 1.8821757952192737e-05,
+      "loss": 2.2441,
+      "step": 3130
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.5181303024292,
+      "learning_rate": 1.8817993600602297e-05,
+      "loss": 2.2284,
+      "step": 3140
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.531221389770508,
+      "learning_rate": 1.881422924901186e-05,
+      "loss": 2.3657,
+      "step": 3150
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.5880889892578125,
+      "learning_rate": 1.881046489742142e-05,
+      "loss": 2.3858,
+      "step": 3160
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.181473731994629,
+      "learning_rate": 1.8806700545830983e-05,
+      "loss": 2.3284,
+      "step": 3170
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.959000587463379,
+      "learning_rate": 1.8802936194240543e-05,
+      "loss": 2.265,
+      "step": 3180
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.405294418334961,
+      "learning_rate": 1.8799171842650106e-05,
+      "loss": 2.4588,
+      "step": 3190
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.867427825927734,
+      "learning_rate": 1.8795407491059666e-05,
+      "loss": 2.2892,
+      "step": 3200
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.539100646972656,
+      "learning_rate": 1.879164313946923e-05,
+      "loss": 2.033,
+      "step": 3210
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.679889678955078,
+      "learning_rate": 1.8787878787878792e-05,
+      "loss": 2.344,
+      "step": 3220
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.749494552612305,
+      "learning_rate": 1.878411443628835e-05,
+      "loss": 2.1338,
+      "step": 3230
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.74128246307373,
+      "learning_rate": 1.878035008469791e-05,
+      "loss": 2.4541,
+      "step": 3240
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 18.931474685668945,
+      "learning_rate": 1.8776585733107475e-05,
+      "loss": 2.2487,
+      "step": 3250
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.824002265930176,
+      "learning_rate": 1.8772821381517034e-05,
+      "loss": 2.407,
+      "step": 3260
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.731551170349121,
+      "learning_rate": 1.8769057029926598e-05,
+      "loss": 2.2347,
+      "step": 3270
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.22593879699707,
+      "learning_rate": 1.8765292678336157e-05,
+      "loss": 2.0125,
+      "step": 3280
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.457214832305908,
+      "learning_rate": 1.876152832674572e-05,
+      "loss": 2.1522,
+      "step": 3290
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.144113540649414,
+      "learning_rate": 1.875776397515528e-05,
+      "loss": 2.4606,
+      "step": 3300
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9879465103149414,
+      "learning_rate": 1.8753999623564843e-05,
+      "loss": 2.0853,
+      "step": 3310
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.082098960876465,
+      "learning_rate": 1.8750235271974403e-05,
+      "loss": 2.1247,
+      "step": 3320
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.393769264221191,
+      "learning_rate": 1.8746470920383966e-05,
+      "loss": 2.1507,
+      "step": 3330
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.922086477279663,
+      "learning_rate": 1.8742706568793526e-05,
+      "loss": 2.2208,
+      "step": 3340
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.737308025360107,
+      "learning_rate": 1.873894221720309e-05,
+      "loss": 2.2848,
+      "step": 3350
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 18.515275955200195,
+      "learning_rate": 1.873517786561265e-05,
+      "loss": 2.0712,
+      "step": 3360
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.918493270874023,
+      "learning_rate": 1.8731413514022212e-05,
+      "loss": 2.1302,
+      "step": 3370
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.196310043334961,
+      "learning_rate": 1.8727649162431772e-05,
+      "loss": 2.1687,
+      "step": 3380
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.755160808563232,
+      "learning_rate": 1.8723884810841335e-05,
+      "loss": 2.3199,
+      "step": 3390
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.229613780975342,
+      "learning_rate": 1.8720120459250895e-05,
+      "loss": 2.0273,
+      "step": 3400
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.063007831573486,
+      "learning_rate": 1.8716356107660455e-05,
+      "loss": 2.2429,
+      "step": 3410
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.2730302810668945,
+      "learning_rate": 1.8712591756070018e-05,
+      "loss": 2.2487,
+      "step": 3420
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.997701644897461,
+      "learning_rate": 1.8708827404479578e-05,
+      "loss": 2.288,
+      "step": 3430
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 13.065136909484863,
+      "learning_rate": 1.870506305288914e-05,
+      "loss": 2.2238,
+      "step": 3440
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.368946075439453,
+      "learning_rate": 1.8701298701298704e-05,
+      "loss": 2.2514,
+      "step": 3450
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.36475944519043,
+      "learning_rate": 1.8697534349708264e-05,
+      "loss": 2.1219,
+      "step": 3460
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.188838958740234,
+      "learning_rate": 1.8693769998117827e-05,
+      "loss": 2.2102,
+      "step": 3470
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 12.989961624145508,
+      "learning_rate": 1.8690005646527387e-05,
+      "loss": 2.2069,
+      "step": 3480
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.382188320159912,
+      "learning_rate": 1.868624129493695e-05,
+      "loss": 2.0466,
+      "step": 3490
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.724734306335449,
+      "learning_rate": 1.868247694334651e-05,
+      "loss": 2.0416,
+      "step": 3500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.628301620483398,
+      "learning_rate": 1.8678712591756073e-05,
+      "loss": 2.3185,
+      "step": 3510
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1391384601593018,
+      "learning_rate": 1.8674948240165632e-05,
+      "loss": 1.9034,
+      "step": 3520
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.9121732711792,
+      "learning_rate": 1.8671183888575196e-05,
+      "loss": 2.1419,
+      "step": 3530
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.588180541992188,
+      "learning_rate": 1.8667419536984755e-05,
+      "loss": 2.1772,
+      "step": 3540
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.538999557495117,
+      "learning_rate": 1.866365518539432e-05,
+      "loss": 2.1558,
+      "step": 3550
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.513119697570801,
+      "learning_rate": 1.8659890833803878e-05,
+      "loss": 2.2185,
+      "step": 3560
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.248276710510254,
+      "learning_rate": 1.865612648221344e-05,
+      "loss": 2.0644,
+      "step": 3570
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.285011291503906,
+      "learning_rate": 1.8652362130623e-05,
+      "loss": 2.1054,
+      "step": 3580
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.439683437347412,
+      "learning_rate": 1.864859777903256e-05,
+      "loss": 2.123,
+      "step": 3590
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.896665573120117,
+      "learning_rate": 1.8644833427442124e-05,
+      "loss": 2.1271,
+      "step": 3600
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.368461608886719,
+      "learning_rate": 1.8641069075851684e-05,
+      "loss": 1.907,
+      "step": 3610
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7022368907928467,
+      "learning_rate": 1.8637304724261247e-05,
+      "loss": 1.9676,
+      "step": 3620
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.112764358520508,
+      "learning_rate": 1.863354037267081e-05,
+      "loss": 1.9633,
+      "step": 3630
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.671046257019043,
+      "learning_rate": 1.862977602108037e-05,
+      "loss": 2.1857,
+      "step": 3640
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.5004682540893555,
+      "learning_rate": 1.8626011669489933e-05,
+      "loss": 1.9923,
+      "step": 3650
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 27.06603240966797,
+      "learning_rate": 1.8622247317899493e-05,
+      "loss": 2.2683,
+      "step": 3660
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.998562812805176,
+      "learning_rate": 1.8618482966309056e-05,
+      "loss": 2.1817,
+      "step": 3670
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.846683979034424,
+      "learning_rate": 1.8614718614718616e-05,
+      "loss": 2.0747,
+      "step": 3680
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.303011894226074,
+      "learning_rate": 1.861095426312818e-05,
+      "loss": 2.0505,
+      "step": 3690
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 14.131027221679688,
+      "learning_rate": 1.860718991153774e-05,
+      "loss": 2.143,
+      "step": 3700
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.333893775939941,
+      "learning_rate": 1.8603425559947302e-05,
+      "loss": 2.0173,
+      "step": 3710
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 13.341752052307129,
+      "learning_rate": 1.859966120835686e-05,
+      "loss": 2.0513,
+      "step": 3720
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.092610359191895,
+      "learning_rate": 1.8595896856766425e-05,
+      "loss": 1.966,
+      "step": 3730
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.7696356773376465,
+      "learning_rate": 1.8592132505175985e-05,
+      "loss": 2.0219,
+      "step": 3740
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.491209030151367,
+      "learning_rate": 1.8588368153585544e-05,
+      "loss": 2.062,
+      "step": 3750
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.02304458618164,
+      "learning_rate": 1.8584603801995107e-05,
+      "loss": 2.1111,
+      "step": 3760
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.15708065032959,
+      "learning_rate": 1.8580839450404667e-05,
+      "loss": 1.8681,
+      "step": 3770
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 14.145309448242188,
+      "learning_rate": 1.857707509881423e-05,
+      "loss": 1.9796,
+      "step": 3780
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.182771682739258,
+      "learning_rate": 1.857331074722379e-05,
+      "loss": 2.2717,
+      "step": 3790
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.4951090812683105,
+      "learning_rate": 1.8569546395633353e-05,
+      "loss": 2.2107,
+      "step": 3800
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 15.642072677612305,
+      "learning_rate": 1.8565782044042916e-05,
+      "loss": 2.1561,
+      "step": 3810
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.078226327896118,
+      "learning_rate": 1.8562017692452476e-05,
+      "loss": 1.9989,
+      "step": 3820
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.49596643447876,
+      "learning_rate": 1.855825334086204e-05,
+      "loss": 1.9954,
+      "step": 3830
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.259866714477539,
+      "learning_rate": 1.85544889892716e-05,
+      "loss": 1.8364,
+      "step": 3840
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.790380477905273,
+      "learning_rate": 1.8550724637681162e-05,
+      "loss": 2.0872,
+      "step": 3850
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.723867416381836,
+      "learning_rate": 1.8546960286090722e-05,
+      "loss": 1.9063,
+      "step": 3860
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.424127101898193,
+      "learning_rate": 1.8543195934500285e-05,
+      "loss": 1.9285,
+      "step": 3870
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.509077548980713,
+      "learning_rate": 1.8539431582909845e-05,
+      "loss": 1.9326,
+      "step": 3880
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.685775756835938,
+      "learning_rate": 1.8535667231319408e-05,
+      "loss": 2.2902,
+      "step": 3890
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.874910354614258,
+      "learning_rate": 1.8531902879728968e-05,
+      "loss": 1.8581,
+      "step": 3900
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.72626256942749,
+      "learning_rate": 1.852813852813853e-05,
+      "loss": 2.0148,
+      "step": 3910
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 21.550472259521484,
+      "learning_rate": 1.852437417654809e-05,
+      "loss": 1.9228,
+      "step": 3920
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.180908203125,
+      "learning_rate": 1.852060982495765e-05,
+      "loss": 1.94,
+      "step": 3930
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4823896884918213,
+      "learning_rate": 1.8516845473367214e-05,
+      "loss": 1.8551,
+      "step": 3940
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 37.56202697753906,
+      "learning_rate": 1.8513081121776774e-05,
+      "loss": 2.1671,
+      "step": 3950
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 13.235817909240723,
+      "learning_rate": 1.8509316770186337e-05,
+      "loss": 1.9974,
+      "step": 3960
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 13.139102935791016,
+      "learning_rate": 1.8505552418595896e-05,
+      "loss": 2.0052,
+      "step": 3970
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.566539287567139,
+      "learning_rate": 1.850178806700546e-05,
+      "loss": 1.8644,
+      "step": 3980
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.045722007751465,
+      "learning_rate": 1.849802371541502e-05,
+      "loss": 1.9597,
+      "step": 3990
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.117792129516602,
+      "learning_rate": 1.8494259363824583e-05,
+      "loss": 1.9613,
+      "step": 4000
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 30.70831298828125,
+      "learning_rate": 1.8490495012234146e-05,
+      "loss": 1.9947,
+      "step": 4010
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.068875312805176,
+      "learning_rate": 1.8486730660643705e-05,
+      "loss": 2.0513,
+      "step": 4020
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.033820629119873,
+      "learning_rate": 1.848296630905327e-05,
+      "loss": 2.1018,
+      "step": 4030
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.325448036193848,
+      "learning_rate": 1.847920195746283e-05,
+      "loss": 2.0123,
+      "step": 4040
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.722583770751953,
+      "learning_rate": 1.847543760587239e-05,
+      "loss": 2.086,
+      "step": 4050
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.869627952575684,
+      "learning_rate": 1.847167325428195e-05,
+      "loss": 1.8988,
+      "step": 4060
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.616646766662598,
+      "learning_rate": 1.8467908902691514e-05,
+      "loss": 1.9823,
+      "step": 4070
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.294058799743652,
+      "learning_rate": 1.8464144551101074e-05,
+      "loss": 1.97,
+      "step": 4080
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1566827297210693,
+      "learning_rate": 1.8460380199510637e-05,
+      "loss": 1.9226,
+      "step": 4090
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.810595512390137,
+      "learning_rate": 1.8456615847920197e-05,
+      "loss": 2.0752,
+      "step": 4100
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.686856269836426,
+      "learning_rate": 1.8452851496329757e-05,
+      "loss": 2.0722,
+      "step": 4110
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.495172500610352,
+      "learning_rate": 1.844908714473932e-05,
+      "loss": 1.7657,
+      "step": 4120
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.293805122375488,
+      "learning_rate": 1.844532279314888e-05,
+      "loss": 1.6995,
+      "step": 4130
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.600557327270508,
+      "learning_rate": 1.8441558441558443e-05,
+      "loss": 1.7201,
+      "step": 4140
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.685958862304688,
+      "learning_rate": 1.8437794089968003e-05,
+      "loss": 1.7472,
+      "step": 4150
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 20.721675872802734,
+      "learning_rate": 1.8434029738377566e-05,
+      "loss": 1.8939,
+      "step": 4160
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.649245262145996,
+      "learning_rate": 1.8430265386787126e-05,
+      "loss": 1.8974,
+      "step": 4170
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 17.670217514038086,
+      "learning_rate": 1.842650103519669e-05,
+      "loss": 1.7713,
+      "step": 4180
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 13.489444732666016,
+      "learning_rate": 1.8422736683606252e-05,
+      "loss": 2.0021,
+      "step": 4190
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 15.630731582641602,
+      "learning_rate": 1.8418972332015812e-05,
+      "loss": 1.9024,
+      "step": 4200
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.856411933898926,
+      "learning_rate": 1.8415207980425375e-05,
+      "loss": 2.1115,
+      "step": 4210
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.349586486816406,
+      "learning_rate": 1.8411443628834935e-05,
+      "loss": 1.7986,
+      "step": 4220
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.628458023071289,
+      "learning_rate": 1.8407679277244498e-05,
+      "loss": 1.7519,
+      "step": 4230
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.205367565155029,
+      "learning_rate": 1.8403914925654058e-05,
+      "loss": 1.8736,
+      "step": 4240
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.882181167602539,
+      "learning_rate": 1.840015057406362e-05,
+      "loss": 1.9139,
+      "step": 4250
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.525898933410645,
+      "learning_rate": 1.839638622247318e-05,
+      "loss": 2.0817,
+      "step": 4260
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.283459663391113,
+      "learning_rate": 1.839262187088274e-05,
+      "loss": 1.8557,
+      "step": 4270
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.506168365478516,
+      "learning_rate": 1.8388857519292303e-05,
+      "loss": 1.9177,
+      "step": 4280
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.862067222595215,
+      "learning_rate": 1.8385093167701863e-05,
+      "loss": 2.0223,
+      "step": 4290
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.25103759765625,
+      "learning_rate": 1.8381328816111426e-05,
+      "loss": 2.045,
+      "step": 4300
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.032151699066162,
+      "learning_rate": 1.8377564464520986e-05,
+      "loss": 1.8646,
+      "step": 4310
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.486144065856934,
+      "learning_rate": 1.837380011293055e-05,
+      "loss": 1.6828,
+      "step": 4320
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.718717575073242,
+      "learning_rate": 1.837003576134011e-05,
+      "loss": 2.1125,
+      "step": 4330
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.391996383666992,
+      "learning_rate": 1.8366271409749672e-05,
+      "loss": 1.7862,
+      "step": 4340
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.340251922607422,
+      "learning_rate": 1.8362507058159232e-05,
+      "loss": 1.8666,
+      "step": 4350
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 13.326081275939941,
+      "learning_rate": 1.8358742706568795e-05,
+      "loss": 1.7263,
+      "step": 4360
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.439601898193359,
+      "learning_rate": 1.8354978354978358e-05,
+      "loss": 1.9045,
+      "step": 4370
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.711633682250977,
+      "learning_rate": 1.8351214003387918e-05,
+      "loss": 1.6444,
+      "step": 4380
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 27.964046478271484,
+      "learning_rate": 1.834744965179748e-05,
+      "loss": 1.9252,
+      "step": 4390
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 23.466896057128906,
+      "learning_rate": 1.834368530020704e-05,
+      "loss": 1.9257,
+      "step": 4400
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.017694473266602,
+      "learning_rate": 1.8339920948616604e-05,
+      "loss": 1.8272,
+      "step": 4410
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.13419246673584,
+      "learning_rate": 1.8336156597026164e-05,
+      "loss": 1.8617,
+      "step": 4420
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.761314392089844,
+      "learning_rate": 1.8332392245435727e-05,
+      "loss": 1.6775,
+      "step": 4430
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 19.422325134277344,
+      "learning_rate": 1.8328627893845287e-05,
+      "loss": 2.015,
+      "step": 4440
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.556143760681152,
+      "learning_rate": 1.8324863542254847e-05,
+      "loss": 1.7302,
+      "step": 4450
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.221558570861816,
+      "learning_rate": 1.832109919066441e-05,
+      "loss": 1.6569,
+      "step": 4460
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.56251335144043,
+      "learning_rate": 1.831733483907397e-05,
+      "loss": 1.7976,
+      "step": 4470
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 15.487319946289062,
+      "learning_rate": 1.8313570487483533e-05,
+      "loss": 2.0049,
+      "step": 4480
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.269038200378418,
+      "learning_rate": 1.8309806135893092e-05,
+      "loss": 1.6322,
+      "step": 4490
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.545343399047852,
+      "learning_rate": 1.8306041784302656e-05,
+      "loss": 1.7523,
+      "step": 4500
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.680045127868652,
+      "learning_rate": 1.8302277432712215e-05,
+      "loss": 1.7489,
+      "step": 4510
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 13.965753555297852,
+      "learning_rate": 1.829851308112178e-05,
+      "loss": 1.7063,
+      "step": 4520
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 14.397643089294434,
+      "learning_rate": 1.8294748729531338e-05,
+      "loss": 1.8438,
+      "step": 4530
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 15.942564010620117,
+      "learning_rate": 1.82909843779409e-05,
+      "loss": 1.8897,
+      "step": 4540
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.38701868057251,
+      "learning_rate": 1.8287220026350465e-05,
+      "loss": 1.8604,
+      "step": 4550
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3334455490112305,
+      "learning_rate": 1.8283455674760024e-05,
+      "loss": 1.6651,
+      "step": 4560
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.184940338134766,
+      "learning_rate": 1.8279691323169587e-05,
+      "loss": 2.0425,
+      "step": 4570
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 29.6932430267334,
+      "learning_rate": 1.8275926971579147e-05,
+      "loss": 1.7136,
+      "step": 4580
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.31380844116211,
+      "learning_rate": 1.827216261998871e-05,
+      "loss": 1.6088,
+      "step": 4590
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.644775390625,
+      "learning_rate": 1.826839826839827e-05,
+      "loss": 1.7825,
+      "step": 4600
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 14.836319923400879,
+      "learning_rate": 1.8264633916807833e-05,
+      "loss": 1.7212,
+      "step": 4610
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.639488220214844,
+      "learning_rate": 1.8260869565217393e-05,
+      "loss": 1.6093,
+      "step": 4620
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.97711181640625,
+      "learning_rate": 1.8257105213626953e-05,
+      "loss": 1.6979,
+      "step": 4630
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 18.360660552978516,
+      "learning_rate": 1.8253340862036516e-05,
+      "loss": 1.8738,
+      "step": 4640
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.36877727508545,
+      "learning_rate": 1.8249576510446076e-05,
+      "loss": 1.6484,
+      "step": 4650
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 13.112272262573242,
+      "learning_rate": 1.824581215885564e-05,
+      "loss": 1.6616,
+      "step": 4660
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.932171821594238,
+      "learning_rate": 1.82420478072652e-05,
+      "loss": 1.6586,
+      "step": 4670
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.852755546569824,
+      "learning_rate": 1.8238283455674762e-05,
+      "loss": 1.8107,
+      "step": 4680
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.760582447052002,
+      "learning_rate": 1.823451910408432e-05,
+      "loss": 1.7808,
+      "step": 4690
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.635812759399414,
+      "learning_rate": 1.8230754752493885e-05,
+      "loss": 1.6713,
+      "step": 4700
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 16.030855178833008,
+      "learning_rate": 1.8226990400903445e-05,
+      "loss": 1.6924,
+      "step": 4710
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 16.42034149169922,
+      "learning_rate": 1.8223226049313008e-05,
+      "loss": 1.8233,
+      "step": 4720
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.589072227478027,
+      "learning_rate": 1.8219461697722567e-05,
+      "loss": 1.4428,
+      "step": 4730
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 14.824429512023926,
+      "learning_rate": 1.821569734613213e-05,
+      "loss": 1.7412,
+      "step": 4740
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6960065364837646,
+      "learning_rate": 1.8211932994541694e-05,
+      "loss": 1.7846,
+      "step": 4750
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.252869606018066,
+      "learning_rate": 1.8208168642951254e-05,
+      "loss": 1.6731,
+      "step": 4760
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.624210357666016,
+      "learning_rate": 1.8204404291360817e-05,
+      "loss": 1.7592,
+      "step": 4770
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.478265762329102,
+      "learning_rate": 1.8200639939770376e-05,
+      "loss": 1.6657,
+      "step": 4780
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.12108039855957,
+      "learning_rate": 1.819687558817994e-05,
+      "loss": 1.6616,
+      "step": 4790
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.339282035827637,
+      "learning_rate": 1.81931112365895e-05,
+      "loss": 1.5382,
+      "step": 4800
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 30.758148193359375,
+      "learning_rate": 1.818934688499906e-05,
+      "loss": 1.7256,
+      "step": 4810
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.9476451873779297,
+      "learning_rate": 1.8185582533408622e-05,
+      "loss": 1.5978,
+      "step": 4820
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4273171424865723,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 1.4934,
+      "step": 4830
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.683277130126953,
+      "learning_rate": 1.8178053830227745e-05,
+      "loss": 1.8227,
+      "step": 4840
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.911778450012207,
+      "learning_rate": 1.8174289478637305e-05,
+      "loss": 1.5653,
+      "step": 4850
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 30.177553176879883,
+      "learning_rate": 1.8170525127046868e-05,
+      "loss": 1.6419,
+      "step": 4860
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 22.67075538635254,
+      "learning_rate": 1.8166760775456428e-05,
+      "loss": 1.8511,
+      "step": 4870
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 16.48545265197754,
+      "learning_rate": 1.816299642386599e-05,
+      "loss": 1.2912,
+      "step": 4880
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.775650978088379,
+      "learning_rate": 1.815923207227555e-05,
+      "loss": 1.7698,
+      "step": 4890
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.288990020751953,
+      "learning_rate": 1.8155467720685114e-05,
+      "loss": 1.5706,
+      "step": 4900
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 17.974271774291992,
+      "learning_rate": 1.8151703369094674e-05,
+      "loss": 1.6064,
+      "step": 4910
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 28.941722869873047,
+      "learning_rate": 1.8147939017504237e-05,
+      "loss": 1.6713,
+      "step": 4920
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.917637825012207,
+      "learning_rate": 1.81441746659138e-05,
+      "loss": 1.6204,
+      "step": 4930
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 24.066226959228516,
+      "learning_rate": 1.814041031432336e-05,
+      "loss": 1.7583,
+      "step": 4940
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.762685775756836,
+      "learning_rate": 1.8136645962732923e-05,
+      "loss": 1.4719,
+      "step": 4950
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6561665534973145,
+      "learning_rate": 1.8132881611142483e-05,
+      "loss": 1.5726,
+      "step": 4960
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.348560333251953,
+      "learning_rate": 1.8129117259552043e-05,
+      "loss": 1.6183,
+      "step": 4970
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.388601303100586,
+      "learning_rate": 1.8125352907961606e-05,
+      "loss": 1.614,
+      "step": 4980
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.767553806304932,
+      "learning_rate": 1.8121588556371165e-05,
+      "loss": 1.5365,
+      "step": 4990
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.167011260986328,
+      "learning_rate": 1.811782420478073e-05,
+      "loss": 1.6598,
+      "step": 5000
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 33.257511138916016,
+      "learning_rate": 1.811405985319029e-05,
+      "loss": 1.9184,
+      "step": 5010
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 21.02366065979004,
+      "learning_rate": 1.811029550159985e-05,
+      "loss": 1.7108,
+      "step": 5020
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.491157531738281,
+      "learning_rate": 1.810653115000941e-05,
+      "loss": 1.5522,
+      "step": 5030
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 13.046422004699707,
+      "learning_rate": 1.8102766798418974e-05,
+      "loss": 1.7517,
+      "step": 5040
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.477202415466309,
+      "learning_rate": 1.8099002446828534e-05,
+      "loss": 1.6329,
+      "step": 5050
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.039795875549316,
+      "learning_rate": 1.8095238095238097e-05,
+      "loss": 1.7396,
+      "step": 5060
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.725845336914062,
+      "learning_rate": 1.8091473743647657e-05,
+      "loss": 1.5248,
+      "step": 5070
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.213160514831543,
+      "learning_rate": 1.808770939205722e-05,
+      "loss": 1.5598,
+      "step": 5080
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.273130893707275,
+      "learning_rate": 1.808394504046678e-05,
+      "loss": 1.5013,
+      "step": 5090
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.758931636810303,
+      "learning_rate": 1.8080180688876343e-05,
+      "loss": 1.6013,
+      "step": 5100
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 12.065877914428711,
+      "learning_rate": 1.8076416337285906e-05,
+      "loss": 1.5301,
+      "step": 5110
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.5146260261535645,
+      "learning_rate": 1.8072651985695466e-05,
+      "loss": 1.5861,
+      "step": 5120
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.56678295135498,
+      "learning_rate": 1.806888763410503e-05,
+      "loss": 1.5615,
+      "step": 5130
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.599604606628418,
+      "learning_rate": 1.8065123282514586e-05,
+      "loss": 1.5666,
+      "step": 5140
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.8899827003479,
+      "learning_rate": 1.806135893092415e-05,
+      "loss": 1.544,
+      "step": 5150
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.435726165771484,
+      "learning_rate": 1.8057594579333712e-05,
+      "loss": 1.4651,
+      "step": 5160
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 17.321407318115234,
+      "learning_rate": 1.8053830227743272e-05,
+      "loss": 1.7957,
+      "step": 5170
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.785209655761719,
+      "learning_rate": 1.8050065876152835e-05,
+      "loss": 1.5213,
+      "step": 5180
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.476792812347412,
+      "learning_rate": 1.8046301524562395e-05,
+      "loss": 1.4325,
+      "step": 5190
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.630148887634277,
+      "learning_rate": 1.8042537172971958e-05,
+      "loss": 1.4048,
+      "step": 5200
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 18.134206771850586,
+      "learning_rate": 1.8038772821381518e-05,
+      "loss": 1.6637,
+      "step": 5210
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 41.112239837646484,
+      "learning_rate": 1.803500846979108e-05,
+      "loss": 1.4036,
+      "step": 5220
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.7763662338256836,
+      "learning_rate": 1.803124411820064e-05,
+      "loss": 1.7532,
+      "step": 5230
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 24.516996383666992,
+      "learning_rate": 1.8027479766610204e-05,
+      "loss": 1.4504,
+      "step": 5240
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 18.604209899902344,
+      "learning_rate": 1.8023715415019763e-05,
+      "loss": 1.509,
+      "step": 5250
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.184398651123047,
+      "learning_rate": 1.8019951063429327e-05,
+      "loss": 1.4567,
+      "step": 5260
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.315374374389648,
+      "learning_rate": 1.8016186711838886e-05,
+      "loss": 1.5755,
+      "step": 5270
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.645417213439941,
+      "learning_rate": 1.801242236024845e-05,
+      "loss": 1.6434,
+      "step": 5280
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.885428428649902,
+      "learning_rate": 1.800865800865801e-05,
+      "loss": 1.4227,
+      "step": 5290
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.746971607208252,
+      "learning_rate": 1.8004893657067572e-05,
+      "loss": 1.2959,
+      "step": 5300
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.56287145614624,
+      "learning_rate": 1.8001129305477136e-05,
+      "loss": 1.3866,
+      "step": 5310
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.8746666666666667,
+      "eval_loss": 1.0967950820922852,
+      "eval_runtime": 33.6773,
+      "eval_samples_per_second": 222.702,
+      "eval_steps_per_second": 27.853,
+      "step": 5313
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.499802112579346,
+      "learning_rate": 1.7997364953886692e-05,
+      "loss": 1.676,
+      "step": 5320
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 16.520263671875,
+      "learning_rate": 1.7993600602296255e-05,
+      "loss": 1.5193,
+      "step": 5330
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.004297256469727,
+      "learning_rate": 1.7989836250705818e-05,
+      "loss": 1.5083,
+      "step": 5340
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 25.948406219482422,
+      "learning_rate": 1.7986071899115378e-05,
+      "loss": 1.2522,
+      "step": 5350
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 38.36083221435547,
+      "learning_rate": 1.798230754752494e-05,
+      "loss": 1.3645,
+      "step": 5360
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 7.533431529998779,
+      "learning_rate": 1.79785431959345e-05,
+      "loss": 1.3528,
+      "step": 5370
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.777375221252441,
+      "learning_rate": 1.7974778844344064e-05,
+      "loss": 1.5204,
+      "step": 5380
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.859736442565918,
+      "learning_rate": 1.7971014492753624e-05,
+      "loss": 1.3621,
+      "step": 5390
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 14.00033187866211,
+      "learning_rate": 1.7967250141163187e-05,
+      "loss": 1.6082,
+      "step": 5400
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 33.97581481933594,
+      "learning_rate": 1.7963485789572747e-05,
+      "loss": 1.6728,
+      "step": 5410
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 15.276714324951172,
+      "learning_rate": 1.795972143798231e-05,
+      "loss": 1.3271,
+      "step": 5420
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.051756858825684,
+      "learning_rate": 1.795595708639187e-05,
+      "loss": 1.6152,
+      "step": 5430
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 14.027779579162598,
+      "learning_rate": 1.7952192734801433e-05,
+      "loss": 1.3804,
+      "step": 5440
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 20.677108764648438,
+      "learning_rate": 1.7948428383210993e-05,
+      "loss": 1.4668,
+      "step": 5450
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 7.414674758911133,
+      "learning_rate": 1.7944664031620556e-05,
+      "loss": 1.3469,
+      "step": 5460
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 8.591064453125,
+      "learning_rate": 1.7940899680030116e-05,
+      "loss": 1.0537,
+      "step": 5470
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 14.56933307647705,
+      "learning_rate": 1.793713532843968e-05,
+      "loss": 1.2944,
+      "step": 5480
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 9.722732543945312,
+      "learning_rate": 1.793337097684924e-05,
+      "loss": 1.7081,
+      "step": 5490
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 22.593751907348633,
+      "learning_rate": 1.7929606625258798e-05,
+      "loss": 1.4611,
+      "step": 5500
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.620069980621338,
+      "learning_rate": 1.792584227366836e-05,
+      "loss": 1.3241,
+      "step": 5510
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 18.490928649902344,
+      "learning_rate": 1.792207792207792e-05,
+      "loss": 1.2878,
+      "step": 5520
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.842093467712402,
+      "learning_rate": 1.7918313570487484e-05,
+      "loss": 1.3626,
+      "step": 5530
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.8665945529937744,
+      "learning_rate": 1.7914549218897047e-05,
+      "loss": 1.3749,
+      "step": 5540
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 8.311807632446289,
+      "learning_rate": 1.7910784867306607e-05,
+      "loss": 1.0602,
+      "step": 5550
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 11.766958236694336,
+      "learning_rate": 1.790702051571617e-05,
+      "loss": 1.2955,
+      "step": 5560
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.863016605377197,
+      "learning_rate": 1.790325616412573e-05,
+      "loss": 1.7003,
+      "step": 5570
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 18.07254409790039,
+      "learning_rate": 1.7899491812535293e-05,
+      "loss": 1.3364,
+      "step": 5580
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.024524211883545,
+      "learning_rate": 1.7895727460944853e-05,
+      "loss": 1.3305,
+      "step": 5590
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 15.396920204162598,
+      "learning_rate": 1.7891963109354416e-05,
+      "loss": 1.309,
+      "step": 5600
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.364221572875977,
+      "learning_rate": 1.7888198757763976e-05,
+      "loss": 1.4572,
+      "step": 5610
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 21.09225082397461,
+      "learning_rate": 1.788443440617354e-05,
+      "loss": 1.4568,
+      "step": 5620
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 37.32011795043945,
+      "learning_rate": 1.78806700545831e-05,
+      "loss": 1.5728,
+      "step": 5630
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.8537468910217285,
+      "learning_rate": 1.7876905702992662e-05,
+      "loss": 1.4453,
+      "step": 5640
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 10.12431812286377,
+      "learning_rate": 1.7873141351402222e-05,
+      "loss": 1.4191,
+      "step": 5650
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 6.774322032928467,
+      "learning_rate": 1.7869376999811785e-05,
+      "loss": 1.3566,
+      "step": 5660
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 9.487374305725098,
+      "learning_rate": 1.7865612648221345e-05,
+      "loss": 1.2845,
+      "step": 5670
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.616371154785156,
+      "learning_rate": 1.7861848296630905e-05,
+      "loss": 1.2827,
+      "step": 5680
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 18.26116943359375,
+      "learning_rate": 1.7858083945040468e-05,
+      "loss": 1.279,
+      "step": 5690
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 6.658090114593506,
+      "learning_rate": 1.7854319593450027e-05,
+      "loss": 1.1976,
+      "step": 5700
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.6320950984954834,
+      "learning_rate": 1.785055524185959e-05,
+      "loss": 1.3875,
+      "step": 5710
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 13.106246948242188,
+      "learning_rate": 1.7846790890269154e-05,
+      "loss": 1.42,
+      "step": 5720
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 7.456079006195068,
+      "learning_rate": 1.7843026538678714e-05,
+      "loss": 1.106,
+      "step": 5730
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 13.17140007019043,
+      "learning_rate": 1.7839262187088277e-05,
+      "loss": 1.153,
+      "step": 5740
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 28.36896324157715,
+      "learning_rate": 1.7835497835497836e-05,
+      "loss": 1.6023,
+      "step": 5750
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 13.630576133728027,
+      "learning_rate": 1.78317334839074e-05,
+      "loss": 1.5223,
+      "step": 5760
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.064805030822754,
+      "learning_rate": 1.782796913231696e-05,
+      "loss": 1.3474,
+      "step": 5770
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 16.060394287109375,
+      "learning_rate": 1.7824204780726523e-05,
+      "loss": 1.3724,
+      "step": 5780
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.598823547363281,
+      "learning_rate": 1.7820440429136082e-05,
+      "loss": 1.2749,
+      "step": 5790
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 15.959213256835938,
+      "learning_rate": 1.7816676077545645e-05,
+      "loss": 1.5515,
+      "step": 5800
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 11.059527397155762,
+      "learning_rate": 1.7812911725955205e-05,
+      "loss": 1.3217,
+      "step": 5810
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 14.561382293701172,
+      "learning_rate": 1.780914737436477e-05,
+      "loss": 1.4639,
+      "step": 5820
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 7.75567626953125,
+      "learning_rate": 1.7805383022774328e-05,
+      "loss": 1.4613,
+      "step": 5830
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 5.322535037994385,
+      "learning_rate": 1.7801618671183888e-05,
+      "loss": 1.3931,
+      "step": 5840
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.112421035766602,
+      "learning_rate": 1.779785431959345e-05,
+      "loss": 1.3793,
+      "step": 5850
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 10.282398223876953,
+      "learning_rate": 1.779408996800301e-05,
+      "loss": 1.4861,
+      "step": 5860
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 16.15000343322754,
+      "learning_rate": 1.7790325616412574e-05,
+      "loss": 1.4151,
+      "step": 5870
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 16.440019607543945,
+      "learning_rate": 1.7786561264822134e-05,
+      "loss": 1.4436,
+      "step": 5880
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.281665802001953,
+      "learning_rate": 1.7782796913231697e-05,
+      "loss": 1.416,
+      "step": 5890
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 6.84842586517334,
+      "learning_rate": 1.777903256164126e-05,
+      "loss": 1.1374,
+      "step": 5900
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 6.96786642074585,
+      "learning_rate": 1.777526821005082e-05,
+      "loss": 1.0473,
+      "step": 5910
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 17.65937042236328,
+      "learning_rate": 1.7771503858460383e-05,
+      "loss": 1.1968,
+      "step": 5920
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.487143516540527,
+      "learning_rate": 1.7767739506869943e-05,
+      "loss": 1.4706,
+      "step": 5930
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 16.491477966308594,
+      "learning_rate": 1.7763975155279506e-05,
+      "loss": 1.3157,
+      "step": 5940
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 6.16838264465332,
+      "learning_rate": 1.7760210803689066e-05,
+      "loss": 1.1594,
+      "step": 5950
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 27.0855712890625,
+      "learning_rate": 1.775644645209863e-05,
+      "loss": 1.3721,
+      "step": 5960
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.84747314453125,
+      "learning_rate": 1.775268210050819e-05,
+      "loss": 1.3545,
+      "step": 5970
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 12.929715156555176,
+      "learning_rate": 1.7748917748917752e-05,
+      "loss": 1.2785,
+      "step": 5980
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.559564590454102,
+      "learning_rate": 1.774515339732731e-05,
+      "loss": 1.242,
+      "step": 5990
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 8.087242126464844,
+      "learning_rate": 1.7741389045736875e-05,
+      "loss": 1.1266,
+      "step": 6000
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.025975227355957,
+      "learning_rate": 1.7737624694146434e-05,
+      "loss": 1.4599,
+      "step": 6010
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 8.053915023803711,
+      "learning_rate": 1.7733860342555994e-05,
+      "loss": 0.8851,
+      "step": 6020
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 11.262309074401855,
+      "learning_rate": 1.7730095990965557e-05,
+      "loss": 1.1816,
+      "step": 6030
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.548508644104004,
+      "learning_rate": 1.7726331639375117e-05,
+      "loss": 1.2975,
+      "step": 6040
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 6.463540554046631,
+      "learning_rate": 1.772256728778468e-05,
+      "loss": 1.1834,
+      "step": 6050
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 8.140536308288574,
+      "learning_rate": 1.771880293619424e-05,
+      "loss": 1.0621,
+      "step": 6060
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.9653189182281494,
+      "learning_rate": 1.7715038584603803e-05,
+      "loss": 1.1374,
+      "step": 6070
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 16.860267639160156,
+      "learning_rate": 1.7711274233013366e-05,
+      "loss": 1.3068,
+      "step": 6080
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.63691520690918,
+      "learning_rate": 1.7707509881422926e-05,
+      "loss": 1.1172,
+      "step": 6090
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.925235748291016,
+      "learning_rate": 1.770374552983249e-05,
+      "loss": 0.7959,
+      "step": 6100
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 27.463420867919922,
+      "learning_rate": 1.769998117824205e-05,
+      "loss": 1.3255,
+      "step": 6110
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.057517051696777,
+      "learning_rate": 1.7696216826651612e-05,
+      "loss": 1.4139,
+      "step": 6120
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.9874982833862305,
+      "learning_rate": 1.7692452475061172e-05,
+      "loss": 1.3465,
+      "step": 6130
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 16.482505798339844,
+      "learning_rate": 1.7688688123470735e-05,
+      "loss": 0.9673,
+      "step": 6140
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.9901885986328125,
+      "learning_rate": 1.7684923771880295e-05,
+      "loss": 1.3453,
+      "step": 6150
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 18.844236373901367,
+      "learning_rate": 1.7681159420289858e-05,
+      "loss": 1.1033,
+      "step": 6160
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 18.83871841430664,
+      "learning_rate": 1.7677395068699418e-05,
+      "loss": 1.4439,
+      "step": 6170
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.778878211975098,
+      "learning_rate": 1.767363071710898e-05,
+      "loss": 1.2751,
+      "step": 6180
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 26.04317283630371,
+      "learning_rate": 1.766986636551854e-05,
+      "loss": 1.1188,
+      "step": 6190
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 14.1301851272583,
+      "learning_rate": 1.76661020139281e-05,
+      "loss": 1.376,
+      "step": 6200
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.929794311523438,
+      "learning_rate": 1.7662337662337664e-05,
+      "loss": 1.2898,
+      "step": 6210
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 24.299001693725586,
+      "learning_rate": 1.7658573310747223e-05,
+      "loss": 1.1609,
+      "step": 6220
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 15.158361434936523,
+      "learning_rate": 1.7654808959156787e-05,
+      "loss": 1.2065,
+      "step": 6230
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 7.862684726715088,
+      "learning_rate": 1.7651044607566346e-05,
+      "loss": 1.302,
+      "step": 6240
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.036644458770752,
+      "learning_rate": 1.764728025597591e-05,
+      "loss": 0.9381,
+      "step": 6250
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 17.739803314208984,
+      "learning_rate": 1.764351590438547e-05,
+      "loss": 0.9986,
+      "step": 6260
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 23.925811767578125,
+      "learning_rate": 1.7639751552795032e-05,
+      "loss": 1.4922,
+      "step": 6270
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.70884895324707,
+      "learning_rate": 1.7635987201204596e-05,
+      "loss": 1.2282,
+      "step": 6280
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.17802619934082,
+      "learning_rate": 1.7632222849614155e-05,
+      "loss": 1.3497,
+      "step": 6290
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 7.948269844055176,
+      "learning_rate": 1.762845849802372e-05,
+      "loss": 1.0691,
+      "step": 6300
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 19.185148239135742,
+      "learning_rate": 1.7624694146433278e-05,
+      "loss": 1.206,
+      "step": 6310
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 19.54599952697754,
+      "learning_rate": 1.762092979484284e-05,
+      "loss": 1.2538,
+      "step": 6320
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.534640789031982,
+      "learning_rate": 1.76171654432524e-05,
+      "loss": 0.9516,
+      "step": 6330
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.249228477478027,
+      "learning_rate": 1.7613401091661964e-05,
+      "loss": 1.1624,
+      "step": 6340
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 17.44942283630371,
+      "learning_rate": 1.7609636740071524e-05,
+      "loss": 1.4045,
+      "step": 6350
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.629911422729492,
+      "learning_rate": 1.7605872388481084e-05,
+      "loss": 1.3624,
+      "step": 6360
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 14.105700492858887,
+      "learning_rate": 1.7602108036890647e-05,
+      "loss": 1.2062,
+      "step": 6370
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.206585884094238,
+      "learning_rate": 1.7598343685300207e-05,
+      "loss": 1.4023,
+      "step": 6380
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 34.754150390625,
+      "learning_rate": 1.759457933370977e-05,
+      "loss": 1.1324,
+      "step": 6390
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 21.429977416992188,
+      "learning_rate": 1.759081498211933e-05,
+      "loss": 1.3611,
+      "step": 6400
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 13.360370635986328,
+      "learning_rate": 1.7587050630528893e-05,
+      "loss": 1.2232,
+      "step": 6410
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.976438522338867,
+      "learning_rate": 1.7583286278938453e-05,
+      "loss": 1.2726,
+      "step": 6420
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.3813695907592773,
+      "learning_rate": 1.7579521927348016e-05,
+      "loss": 1.3038,
+      "step": 6430
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.168612480163574,
+      "learning_rate": 1.7575757575757576e-05,
+      "loss": 1.0494,
+      "step": 6440
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.296903133392334,
+      "learning_rate": 1.757199322416714e-05,
+      "loss": 1.3312,
+      "step": 6450
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.388733863830566,
+      "learning_rate": 1.7568228872576702e-05,
+      "loss": 1.0723,
+      "step": 6460
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.61970043182373,
+      "learning_rate": 1.756446452098626e-05,
+      "loss": 1.1115,
+      "step": 6470
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.065197944641113,
+      "learning_rate": 1.7560700169395825e-05,
+      "loss": 1.173,
+      "step": 6480
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 20.004718780517578,
+      "learning_rate": 1.7556935817805385e-05,
+      "loss": 1.0418,
+      "step": 6490
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 5.888309001922607,
+      "learning_rate": 1.7553171466214948e-05,
+      "loss": 1.2062,
+      "step": 6500
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 31.428281784057617,
+      "learning_rate": 1.7549407114624507e-05,
+      "loss": 1.3629,
+      "step": 6510
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 37.77729034423828,
+      "learning_rate": 1.754564276303407e-05,
+      "loss": 1.1825,
+      "step": 6520
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.3252882957458496,
+      "learning_rate": 1.754187841144363e-05,
+      "loss": 1.3147,
+      "step": 6530
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 14.839897155761719,
+      "learning_rate": 1.753811405985319e-05,
+      "loss": 1.2901,
+      "step": 6540
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 13.03376293182373,
+      "learning_rate": 1.7534349708262753e-05,
+      "loss": 0.8962,
+      "step": 6550
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 15.040424346923828,
+      "learning_rate": 1.7530585356672313e-05,
+      "loss": 1.3194,
+      "step": 6560
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 12.113794326782227,
+      "learning_rate": 1.7526821005081876e-05,
+      "loss": 1.1243,
+      "step": 6570
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 13.116803169250488,
+      "learning_rate": 1.7523056653491436e-05,
+      "loss": 0.9943,
+      "step": 6580
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 6.835216522216797,
+      "learning_rate": 1.7519292301901e-05,
+      "loss": 1.2691,
+      "step": 6590
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.555135726928711,
+      "learning_rate": 1.751552795031056e-05,
+      "loss": 0.9653,
+      "step": 6600
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 20.608491897583008,
+      "learning_rate": 1.7511763598720122e-05,
+      "loss": 1.2499,
+      "step": 6610
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.590780258178711,
+      "learning_rate": 1.7507999247129682e-05,
+      "loss": 1.1381,
+      "step": 6620
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9735676050186157,
+      "learning_rate": 1.7504234895539245e-05,
+      "loss": 0.9156,
+      "step": 6630
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.672266006469727,
+      "learning_rate": 1.7500470543948808e-05,
+      "loss": 1.3334,
+      "step": 6640
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 16.784568786621094,
+      "learning_rate": 1.7496706192358368e-05,
+      "loss": 1.4105,
+      "step": 6650
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 14.953975677490234,
+      "learning_rate": 1.749294184076793e-05,
+      "loss": 1.0408,
+      "step": 6660
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 14.848883628845215,
+      "learning_rate": 1.748917748917749e-05,
+      "loss": 1.1514,
+      "step": 6670
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 13.122929573059082,
+      "learning_rate": 1.7485413137587054e-05,
+      "loss": 1.1443,
+      "step": 6680
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 20.22705078125,
+      "learning_rate": 1.7481648785996614e-05,
+      "loss": 1.0515,
+      "step": 6690
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 11.579082489013672,
+      "learning_rate": 1.7477884434406177e-05,
+      "loss": 1.3199,
+      "step": 6700
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 20.08841323852539,
+      "learning_rate": 1.7474120082815737e-05,
+      "loss": 1.0948,
+      "step": 6710
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 23.67656135559082,
+      "learning_rate": 1.7470355731225296e-05,
+      "loss": 1.158,
+      "step": 6720
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 11.358781814575195,
+      "learning_rate": 1.746659137963486e-05,
+      "loss": 1.1823,
+      "step": 6730
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 11.688865661621094,
+      "learning_rate": 1.746282702804442e-05,
+      "loss": 1.3804,
+      "step": 6740
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.149205207824707,
+      "learning_rate": 1.7459062676453983e-05,
+      "loss": 0.9094,
+      "step": 6750
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.609593391418457,
+      "learning_rate": 1.7455298324863542e-05,
+      "loss": 1.2674,
+      "step": 6760
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 15.846113204956055,
+      "learning_rate": 1.7451533973273105e-05,
+      "loss": 1.2045,
+      "step": 6770
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 14.391237258911133,
+      "learning_rate": 1.7447769621682665e-05,
+      "loss": 1.0173,
+      "step": 6780
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 18.767101287841797,
+      "learning_rate": 1.744400527009223e-05,
+      "loss": 1.1205,
+      "step": 6790
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.494179725646973,
+      "learning_rate": 1.7440240918501788e-05,
+      "loss": 1.0519,
+      "step": 6800
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 7.717184543609619,
+      "learning_rate": 1.743647656691135e-05,
+      "loss": 1.0126,
+      "step": 6810
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 19.109561920166016,
+      "learning_rate": 1.743271221532091e-05,
+      "loss": 0.9986,
+      "step": 6820
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.908644199371338,
+      "learning_rate": 1.7428947863730474e-05,
+      "loss": 0.9898,
+      "step": 6830
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.440939426422119,
+      "learning_rate": 1.7425183512140037e-05,
+      "loss": 1.1074,
+      "step": 6840
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 6.752496719360352,
+      "learning_rate": 1.7421419160549597e-05,
+      "loss": 1.1364,
+      "step": 6850
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 16.471498489379883,
+      "learning_rate": 1.741765480895916e-05,
+      "loss": 1.385,
+      "step": 6860
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 5.562114715576172,
+      "learning_rate": 1.741389045736872e-05,
+      "loss": 1.0686,
+      "step": 6870
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.111942768096924,
+      "learning_rate": 1.7410126105778283e-05,
+      "loss": 1.0202,
+      "step": 6880
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 26.472091674804688,
+      "learning_rate": 1.7406361754187843e-05,
+      "loss": 1.2902,
+      "step": 6890
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.381438255310059,
+      "learning_rate": 1.7402597402597403e-05,
+      "loss": 1.2116,
+      "step": 6900
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 61.49063491821289,
+      "learning_rate": 1.7398833051006966e-05,
+      "loss": 1.4989,
+      "step": 6910
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 20.3765869140625,
+      "learning_rate": 1.7395068699416526e-05,
+      "loss": 0.8747,
+      "step": 6920
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.292556762695312,
+      "learning_rate": 1.739130434782609e-05,
+      "loss": 1.1964,
+      "step": 6930
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 6.413318634033203,
+      "learning_rate": 1.738753999623565e-05,
+      "loss": 0.9384,
+      "step": 6940
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.494325637817383,
+      "learning_rate": 1.7383775644645212e-05,
+      "loss": 1.4534,
+      "step": 6950
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 16.83256721496582,
+      "learning_rate": 1.738001129305477e-05,
+      "loss": 1.1801,
+      "step": 6960
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.633999824523926,
+      "learning_rate": 1.7376246941464335e-05,
+      "loss": 1.3302,
+      "step": 6970
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 23.226150512695312,
+      "learning_rate": 1.7372482589873894e-05,
+      "loss": 1.1845,
+      "step": 6980
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 14.6954345703125,
+      "learning_rate": 1.7368718238283458e-05,
+      "loss": 1.2775,
+      "step": 6990
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.616198539733887,
+      "learning_rate": 1.7364953886693017e-05,
+      "loss": 0.9493,
+      "step": 7000
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.6046929359436035,
+      "learning_rate": 1.736118953510258e-05,
+      "loss": 1.2158,
+      "step": 7010
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 7.577897071838379,
+      "learning_rate": 1.7357425183512144e-05,
+      "loss": 1.092,
+      "step": 7020
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 14.711727142333984,
+      "learning_rate": 1.7353660831921703e-05,
+      "loss": 1.3447,
+      "step": 7030
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 22.08068084716797,
+      "learning_rate": 1.7349896480331267e-05,
+      "loss": 0.9474,
+      "step": 7040
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 5.154607772827148,
+      "learning_rate": 1.7346132128740826e-05,
+      "loss": 1.0093,
+      "step": 7050
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.215937614440918,
+      "learning_rate": 1.7342367777150386e-05,
+      "loss": 1.178,
+      "step": 7060
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 3.9813942909240723,
+      "learning_rate": 1.733860342555995e-05,
+      "loss": 1.1342,
+      "step": 7070
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.915732383728027,
+      "learning_rate": 1.733483907396951e-05,
+      "loss": 1.1109,
+      "step": 7080
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.5986876487731934,
+      "learning_rate": 1.7331074722379072e-05,
+      "loss": 1.1772,
+      "step": 7090
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 5.484806537628174,
+      "learning_rate": 1.7327310370788632e-05,
+      "loss": 1.1488,
+      "step": 7100
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 15.327614784240723,
+      "learning_rate": 1.7323546019198195e-05,
+      "loss": 1.1301,
+      "step": 7110
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.505776405334473,
+      "learning_rate": 1.7319781667607755e-05,
+      "loss": 1.1446,
+      "step": 7120
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 11.313884735107422,
+      "learning_rate": 1.7316017316017318e-05,
+      "loss": 1.0336,
+      "step": 7130
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.289346694946289,
+      "learning_rate": 1.7312252964426878e-05,
+      "loss": 0.9494,
+      "step": 7140
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 14.154651641845703,
+      "learning_rate": 1.730848861283644e-05,
+      "loss": 0.8998,
+      "step": 7150
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.944982528686523,
+      "learning_rate": 1.7304724261246e-05,
+      "loss": 0.9689,
+      "step": 7160
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.77503490447998,
+      "learning_rate": 1.7300959909655564e-05,
+      "loss": 1.0465,
+      "step": 7170
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 28.841171264648438,
+      "learning_rate": 1.7297195558065124e-05,
+      "loss": 1.2161,
+      "step": 7180
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.071671485900879,
+      "learning_rate": 1.7293431206474687e-05,
+      "loss": 1.1167,
+      "step": 7190
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.291550636291504,
+      "learning_rate": 1.728966685488425e-05,
+      "loss": 1.2706,
+      "step": 7200
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 13.803507804870605,
+      "learning_rate": 1.728590250329381e-05,
+      "loss": 1.1404,
+      "step": 7210
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.958276748657227,
+      "learning_rate": 1.7282138151703373e-05,
+      "loss": 1.2336,
+      "step": 7220
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.445544242858887,
+      "learning_rate": 1.727837380011293e-05,
+      "loss": 1.16,
+      "step": 7230
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 20.41545867919922,
+      "learning_rate": 1.7274609448522492e-05,
+      "loss": 1.3856,
+      "step": 7240
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.721564292907715,
+      "learning_rate": 1.7270845096932056e-05,
+      "loss": 1.0896,
+      "step": 7250
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 18.577259063720703,
+      "learning_rate": 1.7267080745341615e-05,
+      "loss": 1.181,
+      "step": 7260
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 24.051130294799805,
+      "learning_rate": 1.726331639375118e-05,
+      "loss": 1.0202,
+      "step": 7270
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.581955909729004,
+      "learning_rate": 1.7259552042160738e-05,
+      "loss": 1.0709,
+      "step": 7280
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 6.8670854568481445,
+      "learning_rate": 1.72557876905703e-05,
+      "loss": 0.9722,
+      "step": 7290
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.2810640335083,
+      "learning_rate": 1.725202333897986e-05,
+      "loss": 1.1988,
+      "step": 7300
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.035595893859863,
+      "learning_rate": 1.7248258987389424e-05,
+      "loss": 1.059,
+      "step": 7310
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 5.461163520812988,
+      "learning_rate": 1.7244494635798984e-05,
+      "loss": 1.1715,
+      "step": 7320
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 6.015925407409668,
+      "learning_rate": 1.7240730284208547e-05,
+      "loss": 0.8844,
+      "step": 7330
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 17.79153060913086,
+      "learning_rate": 1.7236965932618107e-05,
+      "loss": 1.1316,
+      "step": 7340
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 65.43628692626953,
+      "learning_rate": 1.723320158102767e-05,
+      "loss": 0.9798,
+      "step": 7350
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 12.030508041381836,
+      "learning_rate": 1.722943722943723e-05,
+      "loss": 1.093,
+      "step": 7360
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 5.5272297859191895,
+      "learning_rate": 1.7225672877846793e-05,
+      "loss": 0.882,
+      "step": 7370
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 6.576135635375977,
+      "learning_rate": 1.7221908526256356e-05,
+      "loss": 1.0853,
+      "step": 7380
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 13.047865867614746,
+      "learning_rate": 1.7218144174665916e-05,
+      "loss": 1.0521,
+      "step": 7390
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.292452812194824,
+      "learning_rate": 1.721437982307548e-05,
+      "loss": 1.3856,
+      "step": 7400
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 4.0016326904296875,
+      "learning_rate": 1.7210615471485036e-05,
+      "loss": 0.9876,
+      "step": 7410
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.368799209594727,
+      "learning_rate": 1.72068511198946e-05,
+      "loss": 0.7977,
+      "step": 7420
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.0331449508667,
+      "learning_rate": 1.7203086768304162e-05,
+      "loss": 1.0317,
+      "step": 7430
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 4.535419940948486,
+      "learning_rate": 1.719932241671372e-05,
+      "loss": 1.2162,
+      "step": 7440
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 4.63490104675293,
+      "learning_rate": 1.7195558065123285e-05,
+      "loss": 0.9855,
+      "step": 7450
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.621866226196289,
+      "learning_rate": 1.7191793713532845e-05,
+      "loss": 1.1798,
+      "step": 7460
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 6.641757488250732,
+      "learning_rate": 1.7188029361942408e-05,
+      "loss": 1.2643,
+      "step": 7470
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 4.4592604637146,
+      "learning_rate": 1.7184265010351967e-05,
+      "loss": 0.8314,
+      "step": 7480
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 22.9270076751709,
+      "learning_rate": 1.718050065876153e-05,
+      "loss": 1.2297,
+      "step": 7490
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 13.613977432250977,
+      "learning_rate": 1.717673630717109e-05,
+      "loss": 0.957,
+      "step": 7500
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 18.736854553222656,
+      "learning_rate": 1.7172971955580654e-05,
+      "loss": 1.0854,
+      "step": 7510
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 15.172490119934082,
+      "learning_rate": 1.7169207603990213e-05,
+      "loss": 0.9997,
+      "step": 7520
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 14.035980224609375,
+      "learning_rate": 1.7165443252399776e-05,
+      "loss": 0.9522,
+      "step": 7530
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 6.1426310539245605,
+      "learning_rate": 1.7161678900809336e-05,
+      "loss": 0.8932,
+      "step": 7540
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 4.456810474395752,
+      "learning_rate": 1.71579145492189e-05,
+      "loss": 1.0575,
+      "step": 7550
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 18.907272338867188,
+      "learning_rate": 1.715415019762846e-05,
+      "loss": 0.9667,
+      "step": 7560
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 24.286773681640625,
+      "learning_rate": 1.7150385846038022e-05,
+      "loss": 1.1699,
+      "step": 7570
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 6.49676513671875,
+      "learning_rate": 1.7146621494447582e-05,
+      "loss": 1.1024,
+      "step": 7580
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 13.635810852050781,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 0.9841,
+      "step": 7590
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 13.655981063842773,
+      "learning_rate": 1.7139092791266705e-05,
+      "loss": 1.0624,
+      "step": 7600
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.580477714538574,
+      "learning_rate": 1.7135328439676268e-05,
+      "loss": 0.943,
+      "step": 7610
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 19.59247589111328,
+      "learning_rate": 1.7131564088085828e-05,
+      "loss": 0.9837,
+      "step": 7620
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.9449715614318848,
+      "learning_rate": 1.712779973649539e-05,
+      "loss": 1.0507,
+      "step": 7630
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.523272514343262,
+      "learning_rate": 1.712403538490495e-05,
+      "loss": 1.0263,
+      "step": 7640
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 14.362699508666992,
+      "learning_rate": 1.7120271033314514e-05,
+      "loss": 1.1896,
+      "step": 7650
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.5842766761779785,
+      "learning_rate": 1.7116506681724074e-05,
+      "loss": 1.0641,
+      "step": 7660
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 17.754718780517578,
+      "learning_rate": 1.7112742330133637e-05,
+      "loss": 0.9287,
+      "step": 7670
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.024535179138184,
+      "learning_rate": 1.7108977978543197e-05,
+      "loss": 1.2929,
+      "step": 7680
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 14.328852653503418,
+      "learning_rate": 1.710521362695276e-05,
+      "loss": 1.2506,
+      "step": 7690
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.0855348110198975,
+      "learning_rate": 1.710144927536232e-05,
+      "loss": 0.9465,
+      "step": 7700
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.1114163398742676,
+      "learning_rate": 1.7097684923771883e-05,
+      "loss": 1.1016,
+      "step": 7710
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.601757526397705,
+      "learning_rate": 1.7093920572181443e-05,
+      "loss": 1.1521,
+      "step": 7720
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.17392349243164,
+      "learning_rate": 1.7090156220591006e-05,
+      "loss": 1.1438,
+      "step": 7730
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 14.822709083557129,
+      "learning_rate": 1.7086391869000565e-05,
+      "loss": 1.0884,
+      "step": 7740
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 12.926187515258789,
+      "learning_rate": 1.708262751741013e-05,
+      "loss": 1.2607,
+      "step": 7750
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 13.08968734741211,
+      "learning_rate": 1.707886316581969e-05,
+      "loss": 1.1789,
+      "step": 7760
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 13.06160831451416,
+      "learning_rate": 1.7075098814229248e-05,
+      "loss": 1.2074,
+      "step": 7770
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.818836212158203,
+      "learning_rate": 1.707133446263881e-05,
+      "loss": 1.2368,
+      "step": 7780
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 15.557538032531738,
+      "learning_rate": 1.706757011104837e-05,
+      "loss": 1.1518,
+      "step": 7790
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 12.945401191711426,
+      "learning_rate": 1.7063805759457934e-05,
+      "loss": 0.9489,
+      "step": 7800
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 8.303050994873047,
+      "learning_rate": 1.7060041407867497e-05,
+      "loss": 1.0736,
+      "step": 7810
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 19.958223342895508,
+      "learning_rate": 1.7056277056277057e-05,
+      "loss": 1.3085,
+      "step": 7820
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 13.087350845336914,
+      "learning_rate": 1.705251270468662e-05,
+      "loss": 1.2354,
+      "step": 7830
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 5.501480579376221,
+      "learning_rate": 1.704874835309618e-05,
+      "loss": 1.1402,
+      "step": 7840
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 16.800434112548828,
+      "learning_rate": 1.7044984001505743e-05,
+      "loss": 1.176,
+      "step": 7850
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 15.918781280517578,
+      "learning_rate": 1.7041219649915303e-05,
+      "loss": 1.0657,
+      "step": 7860
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.953459739685059,
+      "learning_rate": 1.7037455298324866e-05,
+      "loss": 0.9552,
+      "step": 7870
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.856039047241211,
+      "learning_rate": 1.7033690946734426e-05,
+      "loss": 0.9727,
+      "step": 7880
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 6.3316755294799805,
+      "learning_rate": 1.702992659514399e-05,
+      "loss": 0.9975,
+      "step": 7890
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 54.56752014160156,
+      "learning_rate": 1.702616224355355e-05,
+      "loss": 1.0682,
+      "step": 7900
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 16.042255401611328,
+      "learning_rate": 1.7022397891963112e-05,
+      "loss": 1.2348,
+      "step": 7910
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 14.975419044494629,
+      "learning_rate": 1.7018633540372672e-05,
+      "loss": 1.0173,
+      "step": 7920
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.27600383758545,
+      "learning_rate": 1.701486918878223e-05,
+      "loss": 1.1986,
+      "step": 7930
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 25.487232208251953,
+      "learning_rate": 1.7011104837191795e-05,
+      "loss": 1.5951,
+      "step": 7940
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 19.516244888305664,
+      "learning_rate": 1.7007340485601354e-05,
+      "loss": 1.0892,
+      "step": 7950
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.6688463687896729,
+      "learning_rate": 1.7003576134010918e-05,
+      "loss": 0.8421,
+      "step": 7960
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.081192493438721,
+      "learning_rate": 1.6999811782420477e-05,
+      "loss": 1.0691,
+      "step": 7970
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.31060791015625,
+      "learning_rate": 1.699604743083004e-05,
+      "loss": 0.8777,
+      "step": 7980
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.9369659423828125,
+      "learning_rate": 1.6992283079239604e-05,
+      "loss": 0.9824,
+      "step": 7990
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 13.844595909118652,
+      "learning_rate": 1.6988518727649163e-05,
+      "loss": 0.9572,
+      "step": 8000
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 32.09144592285156,
+      "learning_rate": 1.6984754376058727e-05,
+      "loss": 1.0308,
+      "step": 8010
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.795945167541504,
+      "learning_rate": 1.6980990024468286e-05,
+      "loss": 1.1595,
+      "step": 8020
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 22.064987182617188,
+      "learning_rate": 1.697722567287785e-05,
+      "loss": 0.7747,
+      "step": 8030
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 8.580666542053223,
+      "learning_rate": 1.697346132128741e-05,
+      "loss": 0.9154,
+      "step": 8040
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.91849136352539,
+      "learning_rate": 1.6969696969696972e-05,
+      "loss": 1.2361,
+      "step": 8050
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 13.927072525024414,
+      "learning_rate": 1.6965932618106532e-05,
+      "loss": 0.8912,
+      "step": 8060
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 7.275015830993652,
+      "learning_rate": 1.6962168266516095e-05,
+      "loss": 1.1334,
+      "step": 8070
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 24.040142059326172,
+      "learning_rate": 1.6958403914925655e-05,
+      "loss": 0.9843,
+      "step": 8080
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 22.510576248168945,
+      "learning_rate": 1.6954639563335218e-05,
+      "loss": 0.9083,
+      "step": 8090
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 7.329137325286865,
+      "learning_rate": 1.6950875211744778e-05,
+      "loss": 1.0092,
+      "step": 8100
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 5.5893330574035645,
+      "learning_rate": 1.6947110860154338e-05,
+      "loss": 0.9781,
+      "step": 8110
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 22.0462703704834,
+      "learning_rate": 1.69433465085639e-05,
+      "loss": 0.9981,
+      "step": 8120
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.5774564743042,
+      "learning_rate": 1.693958215697346e-05,
+      "loss": 1.0343,
+      "step": 8130
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 15.137490272521973,
+      "learning_rate": 1.6935817805383024e-05,
+      "loss": 1.0412,
+      "step": 8140
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 16.09028434753418,
+      "learning_rate": 1.6932053453792584e-05,
+      "loss": 0.8931,
+      "step": 8150
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 3.9936368465423584,
+      "learning_rate": 1.6928289102202147e-05,
+      "loss": 0.787,
+      "step": 8160
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 20.518054962158203,
+      "learning_rate": 1.692452475061171e-05,
+      "loss": 1.2866,
+      "step": 8170
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 6.0525689125061035,
+      "learning_rate": 1.692076039902127e-05,
+      "loss": 1.1158,
+      "step": 8180
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.403233528137207,
+      "learning_rate": 1.6916996047430833e-05,
+      "loss": 1.0815,
+      "step": 8190
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 22.706296920776367,
+      "learning_rate": 1.6913231695840393e-05,
+      "loss": 1.0685,
+      "step": 8200
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.937347412109375,
+      "learning_rate": 1.6909467344249956e-05,
+      "loss": 1.1039,
+      "step": 8210
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 20.09309959411621,
+      "learning_rate": 1.6905702992659516e-05,
+      "loss": 0.9035,
+      "step": 8220
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.693171501159668,
+      "learning_rate": 1.690193864106908e-05,
+      "loss": 1.0463,
+      "step": 8230
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 15.067790031433105,
+      "learning_rate": 1.689817428947864e-05,
+      "loss": 1.1981,
+      "step": 8240
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 5.470821857452393,
+      "learning_rate": 1.68944099378882e-05,
+      "loss": 1.1446,
+      "step": 8250
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 7.007253646850586,
+      "learning_rate": 1.689064558629776e-05,
+      "loss": 1.101,
+      "step": 8260
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.471108913421631,
+      "learning_rate": 1.6886881234707325e-05,
+      "loss": 0.7692,
+      "step": 8270
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.229903221130371,
+      "learning_rate": 1.6883116883116884e-05,
+      "loss": 0.9428,
+      "step": 8280
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 13.042211532592773,
+      "learning_rate": 1.6879352531526444e-05,
+      "loss": 1.2651,
+      "step": 8290
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.159402847290039,
+      "learning_rate": 1.6875588179936007e-05,
+      "loss": 0.7305,
+      "step": 8300
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.143734931945801,
+      "learning_rate": 1.6871823828345567e-05,
+      "loss": 1.3042,
+      "step": 8310
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 5.597439765930176,
+      "learning_rate": 1.686805947675513e-05,
+      "loss": 0.7694,
+      "step": 8320
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 24.404653549194336,
+      "learning_rate": 1.686429512516469e-05,
+      "loss": 0.7511,
+      "step": 8330
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.326371908187866,
+      "learning_rate": 1.6860530773574253e-05,
+      "loss": 0.8211,
+      "step": 8340
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 16.690340042114258,
+      "learning_rate": 1.6856766421983813e-05,
+      "loss": 1.0918,
+      "step": 8350
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.873743772506714,
+      "learning_rate": 1.6853002070393376e-05,
+      "loss": 0.8984,
+      "step": 8360
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 38.66802978515625,
+      "learning_rate": 1.684923771880294e-05,
+      "loss": 0.6514,
+      "step": 8370
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 23.641817092895508,
+      "learning_rate": 1.68454733672125e-05,
+      "loss": 0.9783,
+      "step": 8380
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 14.112013816833496,
+      "learning_rate": 1.6841709015622062e-05,
+      "loss": 0.8311,
+      "step": 8390
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 5.782837867736816,
+      "learning_rate": 1.6837944664031622e-05,
+      "loss": 0.9418,
+      "step": 8400
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 5.840491771697998,
+      "learning_rate": 1.6834180312441185e-05,
+      "loss": 0.9358,
+      "step": 8410
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 28.026111602783203,
+      "learning_rate": 1.6830415960850745e-05,
+      "loss": 1.5066,
+      "step": 8420
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 13.618431091308594,
+      "learning_rate": 1.6826651609260308e-05,
+      "loss": 1.0053,
+      "step": 8430
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.274946212768555,
+      "learning_rate": 1.6822887257669868e-05,
+      "loss": 1.0545,
+      "step": 8440
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 19.648679733276367,
+      "learning_rate": 1.6819122906079427e-05,
+      "loss": 0.8776,
+      "step": 8450
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 15.531411170959473,
+      "learning_rate": 1.681535855448899e-05,
+      "loss": 1.0128,
+      "step": 8460
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.063401222229004,
+      "learning_rate": 1.681159420289855e-05,
+      "loss": 1.0426,
+      "step": 8470
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.043712615966797,
+      "learning_rate": 1.6807829851308114e-05,
+      "loss": 0.7918,
+      "step": 8480
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.745186805725098,
+      "learning_rate": 1.6804065499717673e-05,
+      "loss": 0.8653,
+      "step": 8490
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 13.314781188964844,
+      "learning_rate": 1.6800301148127236e-05,
+      "loss": 0.904,
+      "step": 8500
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 4.7882208824157715,
+      "learning_rate": 1.6796536796536796e-05,
+      "loss": 0.8867,
+      "step": 8510
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.435511589050293,
+      "learning_rate": 1.679277244494636e-05,
+      "loss": 1.0631,
+      "step": 8520
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 30.225162506103516,
+      "learning_rate": 1.678900809335592e-05,
+      "loss": 1.235,
+      "step": 8530
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 17.80291175842285,
+      "learning_rate": 1.6785243741765482e-05,
+      "loss": 1.2388,
+      "step": 8540
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 19.908802032470703,
+      "learning_rate": 1.6781479390175045e-05,
+      "loss": 1.1334,
+      "step": 8550
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 4.993206977844238,
+      "learning_rate": 1.6777715038584605e-05,
+      "loss": 0.9747,
+      "step": 8560
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 27.299789428710938,
+      "learning_rate": 1.677395068699417e-05,
+      "loss": 1.1113,
+      "step": 8570
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 12.851706504821777,
+      "learning_rate": 1.6770186335403728e-05,
+      "loss": 0.907,
+      "step": 8580
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 24.122529983520508,
+      "learning_rate": 1.676642198381329e-05,
+      "loss": 0.9915,
+      "step": 8590
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 22.42790985107422,
+      "learning_rate": 1.676265763222285e-05,
+      "loss": 0.9087,
+      "step": 8600
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.046234130859375,
+      "learning_rate": 1.6758893280632414e-05,
+      "loss": 0.9194,
+      "step": 8610
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 11.75545883178711,
+      "learning_rate": 1.6755128929041974e-05,
+      "loss": 1.2203,
+      "step": 8620
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.967751502990723,
+      "learning_rate": 1.6751364577451534e-05,
+      "loss": 1.1011,
+      "step": 8630
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 6.418219566345215,
+      "learning_rate": 1.6747600225861097e-05,
+      "loss": 1.0657,
+      "step": 8640
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 39.12499237060547,
+      "learning_rate": 1.6743835874270657e-05,
+      "loss": 1.2006,
+      "step": 8650
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.25285267829895,
+      "learning_rate": 1.674007152268022e-05,
+      "loss": 0.8309,
+      "step": 8660
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.486536026000977,
+      "learning_rate": 1.673630717108978e-05,
+      "loss": 0.9349,
+      "step": 8670
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 3.4122467041015625,
+      "learning_rate": 1.6732542819499343e-05,
+      "loss": 1.1027,
+      "step": 8680
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.690104484558105,
+      "learning_rate": 1.6728778467908903e-05,
+      "loss": 0.9033,
+      "step": 8690
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 16.22601318359375,
+      "learning_rate": 1.6725014116318466e-05,
+      "loss": 1.0882,
+      "step": 8700
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 20.739831924438477,
+      "learning_rate": 1.6721249764728025e-05,
+      "loss": 0.9375,
+      "step": 8710
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 15.394549369812012,
+      "learning_rate": 1.671748541313759e-05,
+      "loss": 0.7345,
+      "step": 8720
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 26.133543014526367,
+      "learning_rate": 1.6713721061547152e-05,
+      "loss": 0.64,
+      "step": 8730
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 13.604958534240723,
+      "learning_rate": 1.670995670995671e-05,
+      "loss": 0.9136,
+      "step": 8740
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 42.65677261352539,
+      "learning_rate": 1.6706192358366275e-05,
+      "loss": 1.0659,
+      "step": 8750
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.75528621673584,
+      "learning_rate": 1.6702428006775834e-05,
+      "loss": 0.8149,
+      "step": 8760
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 5.066274642944336,
+      "learning_rate": 1.6698663655185398e-05,
+      "loss": 1.0136,
+      "step": 8770
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 22.15322494506836,
+      "learning_rate": 1.6694899303594957e-05,
+      "loss": 1.0216,
+      "step": 8780
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 14.151029586791992,
+      "learning_rate": 1.669113495200452e-05,
+      "loss": 1.1095,
+      "step": 8790
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 6.58624267578125,
+      "learning_rate": 1.668737060041408e-05,
+      "loss": 0.8408,
+      "step": 8800
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 25.148658752441406,
+      "learning_rate": 1.668360624882364e-05,
+      "loss": 1.0391,
+      "step": 8810
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 6.866160869598389,
+      "learning_rate": 1.6679841897233203e-05,
+      "loss": 0.9131,
+      "step": 8820
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 32.80194854736328,
+      "learning_rate": 1.6676077545642763e-05,
+      "loss": 0.8875,
+      "step": 8830
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.0241827964782715,
+      "learning_rate": 1.6672313194052326e-05,
+      "loss": 0.7572,
+      "step": 8840
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 14.243484497070312,
+      "learning_rate": 1.6668548842461886e-05,
+      "loss": 0.9565,
+      "step": 8850
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 18.31822395324707,
+      "learning_rate": 1.666478449087145e-05,
+      "loss": 1.0248,
+      "step": 8860
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 18.321674346923828,
+      "learning_rate": 1.666102013928101e-05,
+      "loss": 0.9848,
+      "step": 8870
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 14.554122924804688,
+      "learning_rate": 1.6657255787690572e-05,
+      "loss": 0.9658,
+      "step": 8880
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.484946250915527,
+      "learning_rate": 1.6653491436100132e-05,
+      "loss": 0.8559,
+      "step": 8890
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 5.653412342071533,
+      "learning_rate": 1.6649727084509695e-05,
+      "loss": 0.9627,
+      "step": 8900
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 7.155862331390381,
+      "learning_rate": 1.6645962732919258e-05,
+      "loss": 0.8445,
+      "step": 8910
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.255889892578125,
+      "learning_rate": 1.6642198381328818e-05,
+      "loss": 0.7791,
+      "step": 8920
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.364027976989746,
+      "learning_rate": 1.663843402973838e-05,
+      "loss": 1.0761,
+      "step": 8930
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 19.489355087280273,
+      "learning_rate": 1.663466967814794e-05,
+      "loss": 0.8686,
+      "step": 8940
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.537185668945312,
+      "learning_rate": 1.6630905326557504e-05,
+      "loss": 0.9145,
+      "step": 8950
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 26.021825790405273,
+      "learning_rate": 1.6627140974967064e-05,
+      "loss": 0.8122,
+      "step": 8960
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 6.992552757263184,
+      "learning_rate": 1.6623376623376627e-05,
+      "loss": 0.7718,
+      "step": 8970
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.073301315307617,
+      "learning_rate": 1.6619612271786187e-05,
+      "loss": 0.9453,
+      "step": 8980
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 25.186731338500977,
+      "learning_rate": 1.6615847920195746e-05,
+      "loss": 0.7758,
+      "step": 8990
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 15.207765579223633,
+      "learning_rate": 1.661208356860531e-05,
+      "loss": 1.0578,
+      "step": 9000
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 14.172220230102539,
+      "learning_rate": 1.660831921701487e-05,
+      "loss": 1.0341,
+      "step": 9010
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.55916976928711,
+      "learning_rate": 1.6604554865424432e-05,
+      "loss": 0.9354,
+      "step": 9020
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 13.917641639709473,
+      "learning_rate": 1.6600790513833992e-05,
+      "loss": 1.136,
+      "step": 9030
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.619650840759277,
+      "learning_rate": 1.6597026162243555e-05,
+      "loss": 0.7212,
+      "step": 9040
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 17.172563552856445,
+      "learning_rate": 1.6593261810653115e-05,
+      "loss": 0.7703,
+      "step": 9050
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 18.772756576538086,
+      "learning_rate": 1.6589497459062678e-05,
+      "loss": 0.8157,
+      "step": 9060
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 18.383899688720703,
+      "learning_rate": 1.6585733107472238e-05,
+      "loss": 1.1417,
+      "step": 9070
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 15.073466300964355,
+      "learning_rate": 1.65819687558818e-05,
+      "loss": 0.9564,
+      "step": 9080
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 23.901710510253906,
+      "learning_rate": 1.657820440429136e-05,
+      "loss": 1.1246,
+      "step": 9090
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 7.932938575744629,
+      "learning_rate": 1.6574440052700924e-05,
+      "loss": 0.949,
+      "step": 9100
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 4.0341410636901855,
+      "learning_rate": 1.6570675701110487e-05,
+      "loss": 1.0209,
+      "step": 9110
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.543979167938232,
+      "learning_rate": 1.6566911349520047e-05,
+      "loss": 1.0694,
+      "step": 9120
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 18.913318634033203,
+      "learning_rate": 1.656314699792961e-05,
+      "loss": 0.9295,
+      "step": 9130
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.830410003662109,
+      "learning_rate": 1.655938264633917e-05,
+      "loss": 0.7556,
+      "step": 9140
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.6431379318237305,
+      "learning_rate": 1.655561829474873e-05,
+      "loss": 0.8671,
+      "step": 9150
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 27.64036750793457,
+      "learning_rate": 1.6551853943158293e-05,
+      "loss": 0.6891,
+      "step": 9160
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 15.285219192504883,
+      "learning_rate": 1.6548089591567853e-05,
+      "loss": 1.1812,
+      "step": 9170
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 11.814650535583496,
+      "learning_rate": 1.6544325239977416e-05,
+      "loss": 0.8133,
+      "step": 9180
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 3.3252789974212646,
+      "learning_rate": 1.6540560888386976e-05,
+      "loss": 1.0719,
+      "step": 9190
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.2466318607330322,
+      "learning_rate": 1.653679653679654e-05,
+      "loss": 0.8765,
+      "step": 9200
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 34.163726806640625,
+      "learning_rate": 1.65330321852061e-05,
+      "loss": 1.2877,
+      "step": 9210
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 27.794078826904297,
+      "learning_rate": 1.652926783361566e-05,
+      "loss": 0.8177,
+      "step": 9220
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 16.112585067749023,
+      "learning_rate": 1.652550348202522e-05,
+      "loss": 1.1625,
+      "step": 9230
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 19.92578125,
+      "learning_rate": 1.6521739130434785e-05,
+      "loss": 1.1242,
+      "step": 9240
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 17.408260345458984,
+      "learning_rate": 1.6517974778844344e-05,
+      "loss": 1.1734,
+      "step": 9250
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 23.571901321411133,
+      "learning_rate": 1.6514210427253907e-05,
+      "loss": 0.9096,
+      "step": 9260
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.09569263458252,
+      "learning_rate": 1.6510446075663467e-05,
+      "loss": 0.8421,
+      "step": 9270
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 22.68370246887207,
+      "learning_rate": 1.650668172407303e-05,
+      "loss": 0.8,
+      "step": 9280
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 26.997875213623047,
+      "learning_rate": 1.6502917372482594e-05,
+      "loss": 0.9072,
+      "step": 9290
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 26.56907081604004,
+      "learning_rate": 1.6499153020892153e-05,
+      "loss": 1.1413,
+      "step": 9300
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.025582313537598,
+      "learning_rate": 1.6495388669301716e-05,
+      "loss": 0.712,
+      "step": 9310
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.8579206466674805,
+      "learning_rate": 1.6491624317711273e-05,
+      "loss": 1.0311,
+      "step": 9320
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.947895526885986,
+      "learning_rate": 1.6487859966120836e-05,
+      "loss": 0.8033,
+      "step": 9330
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 6.863089561462402,
+      "learning_rate": 1.64840956145304e-05,
+      "loss": 1.0714,
+      "step": 9340
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.841645240783691,
+      "learning_rate": 1.648033126293996e-05,
+      "loss": 0.6842,
+      "step": 9350
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 27.438404083251953,
+      "learning_rate": 1.6476566911349522e-05,
+      "loss": 0.851,
+      "step": 9360
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.9211061000823975,
+      "learning_rate": 1.6472802559759082e-05,
+      "loss": 0.8752,
+      "step": 9370
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.311325073242188,
+      "learning_rate": 1.6469038208168645e-05,
+      "loss": 1.0659,
+      "step": 9380
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.285094261169434,
+      "learning_rate": 1.6465273856578205e-05,
+      "loss": 0.8645,
+      "step": 9390
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 12.683711051940918,
+      "learning_rate": 1.6461509504987768e-05,
+      "loss": 0.8845,
+      "step": 9400
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 16.096887588500977,
+      "learning_rate": 1.6457745153397328e-05,
+      "loss": 0.9758,
+      "step": 9410
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 10.897958755493164,
+      "learning_rate": 1.645398080180689e-05,
+      "loss": 1.2507,
+      "step": 9420
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 22.48199462890625,
+      "learning_rate": 1.645021645021645e-05,
+      "loss": 0.7385,
+      "step": 9430
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 16.660545349121094,
+      "learning_rate": 1.6446452098626014e-05,
+      "loss": 0.8564,
+      "step": 9440
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.656999588012695,
+      "learning_rate": 1.6442687747035574e-05,
+      "loss": 0.7063,
+      "step": 9450
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.140769004821777,
+      "learning_rate": 1.6438923395445137e-05,
+      "loss": 1.0147,
+      "step": 9460
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 4.824342727661133,
+      "learning_rate": 1.64351590438547e-05,
+      "loss": 0.7609,
+      "step": 9470
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.159531593322754,
+      "learning_rate": 1.643139469226426e-05,
+      "loss": 1.0369,
+      "step": 9480
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 14.373122215270996,
+      "learning_rate": 1.6427630340673823e-05,
+      "loss": 1.1801,
+      "step": 9490
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 5.078024864196777,
+      "learning_rate": 1.642386598908338e-05,
+      "loss": 0.9217,
+      "step": 9500
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.723122596740723,
+      "learning_rate": 1.6420101637492942e-05,
+      "loss": 0.9727,
+      "step": 9510
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.917556762695312,
+      "learning_rate": 1.6416337285902505e-05,
+      "loss": 1.2133,
+      "step": 9520
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.582098007202148,
+      "learning_rate": 1.6412572934312065e-05,
+      "loss": 0.9397,
+      "step": 9530
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.489495277404785,
+      "learning_rate": 1.640880858272163e-05,
+      "loss": 1.093,
+      "step": 9540
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 21.351308822631836,
+      "learning_rate": 1.6405044231131188e-05,
+      "loss": 0.9417,
+      "step": 9550
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 14.546730041503906,
+      "learning_rate": 1.640127987954075e-05,
+      "loss": 1.0752,
+      "step": 9560
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 14.62718391418457,
+      "learning_rate": 1.639751552795031e-05,
+      "loss": 0.85,
+      "step": 9570
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.844496726989746,
+      "learning_rate": 1.6393751176359874e-05,
+      "loss": 1.3426,
+      "step": 9580
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.517404317855835,
+      "learning_rate": 1.6389986824769434e-05,
+      "loss": 0.8968,
+      "step": 9590
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 39.88374328613281,
+      "learning_rate": 1.6386222473178997e-05,
+      "loss": 0.6198,
+      "step": 9600
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 23.119050979614258,
+      "learning_rate": 1.6382458121588557e-05,
+      "loss": 0.7901,
+      "step": 9610
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.583620071411133,
+      "learning_rate": 1.637869376999812e-05,
+      "loss": 1.0529,
+      "step": 9620
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 29.128244400024414,
+      "learning_rate": 1.637492941840768e-05,
+      "loss": 1.237,
+      "step": 9630
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.6622776985168457,
+      "learning_rate": 1.6371165066817243e-05,
+      "loss": 0.9682,
+      "step": 9640
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 28.509607315063477,
+      "learning_rate": 1.6367400715226803e-05,
+      "loss": 1.0373,
+      "step": 9650
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.624483346939087,
+      "learning_rate": 1.6363636363636366e-05,
+      "loss": 0.9508,
+      "step": 9660
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 14.195404052734375,
+      "learning_rate": 1.6359872012045926e-05,
+      "loss": 1.0253,
+      "step": 9670
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.364255905151367,
+      "learning_rate": 1.6356107660455485e-05,
+      "loss": 1.1393,
+      "step": 9680
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 17.5921688079834,
+      "learning_rate": 1.635234330886505e-05,
+      "loss": 1.2439,
+      "step": 9690
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 20.8980712890625,
+      "learning_rate": 1.6348578957274612e-05,
+      "loss": 1.1815,
+      "step": 9700
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 5.478924751281738,
+      "learning_rate": 1.634481460568417e-05,
+      "loss": 1.0223,
+      "step": 9710
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 20.507476806640625,
+      "learning_rate": 1.6341050254093735e-05,
+      "loss": 0.6629,
+      "step": 9720
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 28.950838088989258,
+      "learning_rate": 1.6337285902503294e-05,
+      "loss": 0.858,
+      "step": 9730
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 14.215471267700195,
+      "learning_rate": 1.6333521550912858e-05,
+      "loss": 1.0337,
+      "step": 9740
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 9.36776065826416,
+      "learning_rate": 1.6329757199322417e-05,
+      "loss": 1.4074,
+      "step": 9750
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.266942501068115,
+      "learning_rate": 1.632599284773198e-05,
+      "loss": 0.8383,
+      "step": 9760
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.291203498840332,
+      "learning_rate": 1.632222849614154e-05,
+      "loss": 0.8997,
+      "step": 9770
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 28.071760177612305,
+      "learning_rate": 1.6318464144551103e-05,
+      "loss": 0.8101,
+      "step": 9780
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.149245262145996,
+      "learning_rate": 1.6314699792960663e-05,
+      "loss": 0.7052,
+      "step": 9790
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 34.350502014160156,
+      "learning_rate": 1.6310935441370226e-05,
+      "loss": 0.9759,
+      "step": 9800
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 4.613625526428223,
+      "learning_rate": 1.6307171089779786e-05,
+      "loss": 0.9999,
+      "step": 9810
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.235222816467285,
+      "learning_rate": 1.630340673818935e-05,
+      "loss": 0.8468,
+      "step": 9820
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.5585298538208,
+      "learning_rate": 1.629964238659891e-05,
+      "loss": 1.0039,
+      "step": 9830
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 23.900362014770508,
+      "learning_rate": 1.6295878035008472e-05,
+      "loss": 1.15,
+      "step": 9840
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.563354969024658,
+      "learning_rate": 1.6292113683418032e-05,
+      "loss": 0.723,
+      "step": 9850
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 38.29470443725586,
+      "learning_rate": 1.6288349331827592e-05,
+      "loss": 0.9061,
+      "step": 9860
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 17.7316837310791,
+      "learning_rate": 1.6284584980237155e-05,
+      "loss": 1.1306,
+      "step": 9870
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 22.314489364624023,
+      "learning_rate": 1.6280820628646715e-05,
+      "loss": 0.8234,
+      "step": 9880
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 10.5621337890625,
+      "learning_rate": 1.6277056277056278e-05,
+      "loss": 0.7809,
+      "step": 9890
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 13.799981117248535,
+      "learning_rate": 1.627329192546584e-05,
+      "loss": 0.9947,
+      "step": 9900
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 15.118837356567383,
+      "learning_rate": 1.62695275738754e-05,
+      "loss": 0.9937,
+      "step": 9910
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.086974143981934,
+      "learning_rate": 1.6265763222284964e-05,
+      "loss": 0.8989,
+      "step": 9920
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 24.501052856445312,
+      "learning_rate": 1.6261998870694524e-05,
+      "loss": 0.9246,
+      "step": 9930
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 17.838918685913086,
+      "learning_rate": 1.6258234519104087e-05,
+      "loss": 0.6852,
+      "step": 9940
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 14.205941200256348,
+      "learning_rate": 1.6254470167513647e-05,
+      "loss": 0.9937,
+      "step": 9950
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 37.4649658203125,
+      "learning_rate": 1.625070581592321e-05,
+      "loss": 1.1566,
+      "step": 9960
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 3.642307758331299,
+      "learning_rate": 1.624694146433277e-05,
+      "loss": 0.6511,
+      "step": 9970
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.570235252380371,
+      "learning_rate": 1.6243177112742333e-05,
+      "loss": 1.0627,
+      "step": 9980
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 49.44059753417969,
+      "learning_rate": 1.6239412761151892e-05,
+      "loss": 0.87,
+      "step": 9990
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 4.772658824920654,
+      "learning_rate": 1.6235648409561456e-05,
+      "loss": 0.9856,
+      "step": 10000
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.388487815856934,
+      "learning_rate": 1.6231884057971015e-05,
+      "loss": 0.8331,
+      "step": 10010
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 28.38582992553711,
+      "learning_rate": 1.6228119706380575e-05,
+      "loss": 0.7234,
+      "step": 10020
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 35.28871154785156,
+      "learning_rate": 1.6224355354790138e-05,
+      "loss": 1.2131,
+      "step": 10030
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.386420249938965,
+      "learning_rate": 1.6220591003199698e-05,
+      "loss": 0.5296,
+      "step": 10040
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.352224349975586,
+      "learning_rate": 1.621682665160926e-05,
+      "loss": 0.8543,
+      "step": 10050
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 45.16379928588867,
+      "learning_rate": 1.621306230001882e-05,
+      "loss": 1.0114,
+      "step": 10060
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 18.55881118774414,
+      "learning_rate": 1.6209297948428384e-05,
+      "loss": 0.8992,
+      "step": 10070
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 35.752506256103516,
+      "learning_rate": 1.6205533596837947e-05,
+      "loss": 1.1698,
+      "step": 10080
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.165136337280273,
+      "learning_rate": 1.6201769245247507e-05,
+      "loss": 0.809,
+      "step": 10090
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.1471781730651855,
+      "learning_rate": 1.619800489365707e-05,
+      "loss": 0.9719,
+      "step": 10100
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.279614448547363,
+      "learning_rate": 1.619424054206663e-05,
+      "loss": 0.8327,
+      "step": 10110
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 4.469071865081787,
+      "learning_rate": 1.6190476190476193e-05,
+      "loss": 0.7403,
+      "step": 10120
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 28.841690063476562,
+      "learning_rate": 1.6186711838885753e-05,
+      "loss": 0.9098,
+      "step": 10130
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 13.314472198486328,
+      "learning_rate": 1.6182947487295316e-05,
+      "loss": 0.8678,
+      "step": 10140
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 12.73503589630127,
+      "learning_rate": 1.6179183135704876e-05,
+      "loss": 0.9671,
+      "step": 10150
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.612974643707275,
+      "learning_rate": 1.617541878411444e-05,
+      "loss": 0.8328,
+      "step": 10160
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 16.328927993774414,
+      "learning_rate": 1.6171654432524e-05,
+      "loss": 0.9885,
+      "step": 10170
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 22.157751083374023,
+      "learning_rate": 1.6167890080933562e-05,
+      "loss": 0.9805,
+      "step": 10180
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 41.27034378051758,
+      "learning_rate": 1.616412572934312e-05,
+      "loss": 1.0366,
+      "step": 10190
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.670302391052246,
+      "learning_rate": 1.616036137775268e-05,
+      "loss": 1.0358,
+      "step": 10200
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 23.558927536010742,
+      "learning_rate": 1.6156597026162245e-05,
+      "loss": 1.0069,
+      "step": 10210
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.577465057373047,
+      "learning_rate": 1.6152832674571804e-05,
+      "loss": 0.9131,
+      "step": 10220
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 7.152612209320068,
+      "learning_rate": 1.6149068322981367e-05,
+      "loss": 0.9243,
+      "step": 10230
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 14.56041431427002,
+      "learning_rate": 1.6145303971390927e-05,
+      "loss": 1.2452,
+      "step": 10240
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 4.7469611167907715,
+      "learning_rate": 1.614153961980049e-05,
+      "loss": 0.8841,
+      "step": 10250
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6863610148429871,
+      "learning_rate": 1.6137775268210054e-05,
+      "loss": 0.6211,
+      "step": 10260
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 18.57179832458496,
+      "learning_rate": 1.6134010916619613e-05,
+      "loss": 1.0319,
+      "step": 10270
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.881142616271973,
+      "learning_rate": 1.6130246565029176e-05,
+      "loss": 1.0036,
+      "step": 10280
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 12.16629695892334,
+      "learning_rate": 1.6126482213438736e-05,
+      "loss": 0.7445,
+      "step": 10290
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 26.047819137573242,
+      "learning_rate": 1.61227178618483e-05,
+      "loss": 1.1173,
+      "step": 10300
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 4.4914116859436035,
+      "learning_rate": 1.611895351025786e-05,
+      "loss": 0.9002,
+      "step": 10310
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 12.42726993560791,
+      "learning_rate": 1.6115189158667422e-05,
+      "loss": 0.9133,
+      "step": 10320
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 12.923616409301758,
+      "learning_rate": 1.6111424807076982e-05,
+      "loss": 0.4911,
+      "step": 10330
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 22.348865509033203,
+      "learning_rate": 1.6107660455486545e-05,
+      "loss": 0.8791,
+      "step": 10340
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 4.061618804931641,
+      "learning_rate": 1.6103896103896105e-05,
+      "loss": 0.9309,
+      "step": 10350
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 25.17974090576172,
+      "learning_rate": 1.6100131752305668e-05,
+      "loss": 0.995,
+      "step": 10360
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 25.17315673828125,
+      "learning_rate": 1.6096367400715228e-05,
+      "loss": 0.6279,
+      "step": 10370
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 33.93449020385742,
+      "learning_rate": 1.6092603049124788e-05,
+      "loss": 0.9279,
+      "step": 10380
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.33711051940918,
+      "learning_rate": 1.608883869753435e-05,
+      "loss": 1.0123,
+      "step": 10390
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 4.49125862121582,
+      "learning_rate": 1.608507434594391e-05,
+      "loss": 0.7202,
+      "step": 10400
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 21.362960815429688,
+      "learning_rate": 1.6081309994353474e-05,
+      "loss": 1.0845,
+      "step": 10410
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 22.759014129638672,
+      "learning_rate": 1.6077545642763034e-05,
+      "loss": 0.6752,
+      "step": 10420
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 20.3575382232666,
+      "learning_rate": 1.6073781291172597e-05,
+      "loss": 0.9548,
+      "step": 10430
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.1825947761535645,
+      "learning_rate": 1.6070016939582156e-05,
+      "loss": 0.7794,
+      "step": 10440
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 13.626837730407715,
+      "learning_rate": 1.606625258799172e-05,
+      "loss": 0.8118,
+      "step": 10450
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 20.532129287719727,
+      "learning_rate": 1.6062488236401283e-05,
+      "loss": 1.0943,
+      "step": 10460
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 23.677160263061523,
+      "learning_rate": 1.6058723884810843e-05,
+      "loss": 1.4279,
+      "step": 10470
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 30.627092361450195,
+      "learning_rate": 1.6054959533220406e-05,
+      "loss": 0.798,
+      "step": 10480
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 11.574738502502441,
+      "learning_rate": 1.6051195181629965e-05,
+      "loss": 0.7298,
+      "step": 10490
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 20.059284210205078,
+      "learning_rate": 1.604743083003953e-05,
+      "loss": 1.1131,
+      "step": 10500
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 14.673868179321289,
+      "learning_rate": 1.604366647844909e-05,
+      "loss": 0.8992,
+      "step": 10510
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.70554256439209,
+      "learning_rate": 1.603990212685865e-05,
+      "loss": 0.9923,
+      "step": 10520
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 22.268571853637695,
+      "learning_rate": 1.603613777526821e-05,
+      "loss": 0.9369,
+      "step": 10530
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 7.241332530975342,
+      "learning_rate": 1.603237342367777e-05,
+      "loss": 0.918,
+      "step": 10540
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.223257064819336,
+      "learning_rate": 1.6028609072087334e-05,
+      "loss": 0.7295,
+      "step": 10550
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 14.106549263000488,
+      "learning_rate": 1.6024844720496894e-05,
+      "loss": 0.9356,
+      "step": 10560
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.024641513824463,
+      "learning_rate": 1.6021080368906457e-05,
+      "loss": 1.0827,
+      "step": 10570
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.588906288146973,
+      "learning_rate": 1.6017316017316017e-05,
+      "loss": 0.9986,
+      "step": 10580
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 13.386589050292969,
+      "learning_rate": 1.601355166572558e-05,
+      "loss": 0.801,
+      "step": 10590
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 14.955891609191895,
+      "learning_rate": 1.600978731413514e-05,
+      "loss": 0.611,
+      "step": 10600
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 26.955785751342773,
+      "learning_rate": 1.6006022962544703e-05,
+      "loss": 0.8322,
+      "step": 10610
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.895423889160156,
+      "learning_rate": 1.6002258610954263e-05,
+      "loss": 0.6479,
+      "step": 10620
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.9004,
+      "eval_loss": 0.43774810433387756,
+      "eval_runtime": 33.2538,
+      "eval_samples_per_second": 225.538,
+      "eval_steps_per_second": 28.207,
+      "step": 10626
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.223067283630371,
+      "learning_rate": 1.5998494259363826e-05,
+      "loss": 0.8513,
+      "step": 10630
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.11199951171875,
+      "learning_rate": 1.599472990777339e-05,
+      "loss": 0.8083,
+      "step": 10640
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.889083385467529,
+      "learning_rate": 1.599096555618295e-05,
+      "loss": 0.6345,
+      "step": 10650
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.47759780287742615,
+      "learning_rate": 1.5987201204592512e-05,
+      "loss": 0.7421,
+      "step": 10660
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 7.3583879470825195,
+      "learning_rate": 1.5983436853002072e-05,
+      "loss": 0.791,
+      "step": 10670
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 14.362602233886719,
+      "learning_rate": 1.5979672501411635e-05,
+      "loss": 0.7772,
+      "step": 10680
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 14.467147827148438,
+      "learning_rate": 1.5975908149821195e-05,
+      "loss": 0.8084,
+      "step": 10690
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 27.702394485473633,
+      "learning_rate": 1.5972143798230758e-05,
+      "loss": 1.2076,
+      "step": 10700
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 29.06777572631836,
+      "learning_rate": 1.5968379446640318e-05,
+      "loss": 0.9421,
+      "step": 10710
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 17.13526725769043,
+      "learning_rate": 1.5964615095049877e-05,
+      "loss": 0.7547,
+      "step": 10720
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.489052772521973,
+      "learning_rate": 1.596085074345944e-05,
+      "loss": 0.6932,
+      "step": 10730
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 18.08340072631836,
+      "learning_rate": 1.5957086391869e-05,
+      "loss": 0.9143,
+      "step": 10740
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 7.827996730804443,
+      "learning_rate": 1.5953322040278563e-05,
+      "loss": 0.7774,
+      "step": 10750
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 29.571741104125977,
+      "learning_rate": 1.5949557688688123e-05,
+      "loss": 0.7913,
+      "step": 10760
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 12.85148811340332,
+      "learning_rate": 1.5945793337097686e-05,
+      "loss": 0.5545,
+      "step": 10770
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.041319847106934,
+      "learning_rate": 1.5942028985507246e-05,
+      "loss": 0.7221,
+      "step": 10780
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 13.464383125305176,
+      "learning_rate": 1.593826463391681e-05,
+      "loss": 1.0252,
+      "step": 10790
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 267.48870849609375,
+      "learning_rate": 1.593450028232637e-05,
+      "loss": 1.032,
+      "step": 10800
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.915716648101807,
+      "learning_rate": 1.5930735930735932e-05,
+      "loss": 0.6537,
+      "step": 10810
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.0029585361480713,
+      "learning_rate": 1.5926971579145495e-05,
+      "loss": 0.8683,
+      "step": 10820
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 51.99099349975586,
+      "learning_rate": 1.5923207227555055e-05,
+      "loss": 0.714,
+      "step": 10830
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 18.225906372070312,
+      "learning_rate": 1.5919442875964618e-05,
+      "loss": 0.6733,
+      "step": 10840
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 14.735111236572266,
+      "learning_rate": 1.5915678524374178e-05,
+      "loss": 0.7899,
+      "step": 10850
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.9460431337356567,
+      "learning_rate": 1.591191417278374e-05,
+      "loss": 0.697,
+      "step": 10860
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.448774814605713,
+      "learning_rate": 1.59081498211933e-05,
+      "loss": 0.699,
+      "step": 10870
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 12.799127578735352,
+      "learning_rate": 1.5904385469602864e-05,
+      "loss": 1.0552,
+      "step": 10880
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 44.27261734008789,
+      "learning_rate": 1.5900621118012424e-05,
+      "loss": 0.6706,
+      "step": 10890
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 23.952608108520508,
+      "learning_rate": 1.5896856766421984e-05,
+      "loss": 1.0494,
+      "step": 10900
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 29.93094825744629,
+      "learning_rate": 1.5893092414831547e-05,
+      "loss": 0.7551,
+      "step": 10910
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 12.698896408081055,
+      "learning_rate": 1.5889328063241107e-05,
+      "loss": 0.8114,
+      "step": 10920
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 6.4680867195129395,
+      "learning_rate": 1.588556371165067e-05,
+      "loss": 0.6495,
+      "step": 10930
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 20.92882537841797,
+      "learning_rate": 1.588179936006023e-05,
+      "loss": 0.5763,
+      "step": 10940
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 16.703866958618164,
+      "learning_rate": 1.5878035008469793e-05,
+      "loss": 0.7678,
+      "step": 10950
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 26.071792602539062,
+      "learning_rate": 1.5874270656879352e-05,
+      "loss": 0.7934,
+      "step": 10960
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 8.398687362670898,
+      "learning_rate": 1.5870506305288916e-05,
+      "loss": 0.6339,
+      "step": 10970
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 7.431175231933594,
+      "learning_rate": 1.5866741953698475e-05,
+      "loss": 0.4395,
+      "step": 10980
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 14.827509880065918,
+      "learning_rate": 1.586297760210804e-05,
+      "loss": 0.6793,
+      "step": 10990
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 3.574143409729004,
+      "learning_rate": 1.58592132505176e-05,
+      "loss": 1.0553,
+      "step": 11000
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 11.981563568115234,
+      "learning_rate": 1.585544889892716e-05,
+      "loss": 1.0152,
+      "step": 11010
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 33.9033088684082,
+      "learning_rate": 1.5851684547336725e-05,
+      "loss": 0.8337,
+      "step": 11020
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.125753879547119,
+      "learning_rate": 1.5847920195746284e-05,
+      "loss": 0.6738,
+      "step": 11030
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 16.212665557861328,
+      "learning_rate": 1.5844155844155847e-05,
+      "loss": 0.7257,
+      "step": 11040
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 8.9789457321167,
+      "learning_rate": 1.5840391492565407e-05,
+      "loss": 0.8129,
+      "step": 11050
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 27.197084426879883,
+      "learning_rate": 1.583662714097497e-05,
+      "loss": 0.8489,
+      "step": 11060
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 9.654715538024902,
+      "learning_rate": 1.583286278938453e-05,
+      "loss": 1.1031,
+      "step": 11070
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 7.9157514572143555,
+      "learning_rate": 1.582909843779409e-05,
+      "loss": 0.649,
+      "step": 11080
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 26.433883666992188,
+      "learning_rate": 1.5825334086203653e-05,
+      "loss": 1.0084,
+      "step": 11090
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.050842046737671,
+      "learning_rate": 1.5821569734613213e-05,
+      "loss": 0.7625,
+      "step": 11100
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 22.36996078491211,
+      "learning_rate": 1.5817805383022776e-05,
+      "loss": 0.877,
+      "step": 11110
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.3118711709976196,
+      "learning_rate": 1.5814041031432336e-05,
+      "loss": 0.5775,
+      "step": 11120
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 17.231550216674805,
+      "learning_rate": 1.58102766798419e-05,
+      "loss": 1.0835,
+      "step": 11130
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 8.116622924804688,
+      "learning_rate": 1.580651232825146e-05,
+      "loss": 0.7954,
+      "step": 11140
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 10.011574745178223,
+      "learning_rate": 1.5802747976661022e-05,
+      "loss": 0.817,
+      "step": 11150
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 17.308115005493164,
+      "learning_rate": 1.579898362507058e-05,
+      "loss": 0.4949,
+      "step": 11160
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 15.126243591308594,
+      "learning_rate": 1.5795219273480145e-05,
+      "loss": 0.5666,
+      "step": 11170
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 8.26230525970459,
+      "learning_rate": 1.5791454921889705e-05,
+      "loss": 0.5484,
+      "step": 11180
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.2589051723480225,
+      "learning_rate": 1.5787690570299268e-05,
+      "loss": 0.8625,
+      "step": 11190
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 18.615032196044922,
+      "learning_rate": 1.578392621870883e-05,
+      "loss": 1.1054,
+      "step": 11200
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 14.837967872619629,
+      "learning_rate": 1.578016186711839e-05,
+      "loss": 0.9737,
+      "step": 11210
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 11.36581802368164,
+      "learning_rate": 1.5776397515527954e-05,
+      "loss": 0.77,
+      "step": 11220
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 7.043651103973389,
+      "learning_rate": 1.5772633163937514e-05,
+      "loss": 0.7536,
+      "step": 11230
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 18.54761505126953,
+      "learning_rate": 1.5768868812347073e-05,
+      "loss": 1.0175,
+      "step": 11240
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 7.881554126739502,
+      "learning_rate": 1.5765104460756636e-05,
+      "loss": 0.6701,
+      "step": 11250
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 18.58554458618164,
+      "learning_rate": 1.5761340109166196e-05,
+      "loss": 1.0165,
+      "step": 11260
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 6.169960021972656,
+      "learning_rate": 1.575757575757576e-05,
+      "loss": 0.7932,
+      "step": 11270
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 13.55213737487793,
+      "learning_rate": 1.575381140598532e-05,
+      "loss": 0.6286,
+      "step": 11280
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 26.40308952331543,
+      "learning_rate": 1.5750047054394882e-05,
+      "loss": 1.0074,
+      "step": 11290
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 14.90792465209961,
+      "learning_rate": 1.5746282702804442e-05,
+      "loss": 0.7594,
+      "step": 11300
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 25.512937545776367,
+      "learning_rate": 1.5742518351214005e-05,
+      "loss": 0.8933,
+      "step": 11310
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 20.684497833251953,
+      "learning_rate": 1.5738753999623565e-05,
+      "loss": 0.8769,
+      "step": 11320
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 26.62982940673828,
+      "learning_rate": 1.5734989648033128e-05,
+      "loss": 1.0551,
+      "step": 11330
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 12.02119255065918,
+      "learning_rate": 1.5731225296442688e-05,
+      "loss": 0.8013,
+      "step": 11340
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.007152080535889,
+      "learning_rate": 1.572746094485225e-05,
+      "loss": 0.7307,
+      "step": 11350
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 11.850004196166992,
+      "learning_rate": 1.572369659326181e-05,
+      "loss": 0.7143,
+      "step": 11360
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 11.220576286315918,
+      "learning_rate": 1.5719932241671374e-05,
+      "loss": 0.7637,
+      "step": 11370
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.0567941665649414,
+      "learning_rate": 1.5716167890080937e-05,
+      "loss": 0.9187,
+      "step": 11380
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 24.867557525634766,
+      "learning_rate": 1.5712403538490497e-05,
+      "loss": 0.86,
+      "step": 11390
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 13.429787635803223,
+      "learning_rate": 1.570863918690006e-05,
+      "loss": 0.919,
+      "step": 11400
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.4714537858963013,
+      "learning_rate": 1.5704874835309616e-05,
+      "loss": 0.8753,
+      "step": 11410
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 26.924331665039062,
+      "learning_rate": 1.570111048371918e-05,
+      "loss": 0.61,
+      "step": 11420
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 19.854873657226562,
+      "learning_rate": 1.5697346132128743e-05,
+      "loss": 0.765,
+      "step": 11430
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 10.933598518371582,
+      "learning_rate": 1.5693581780538303e-05,
+      "loss": 0.9729,
+      "step": 11440
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 12.186927795410156,
+      "learning_rate": 1.5689817428947866e-05,
+      "loss": 1.1194,
+      "step": 11450
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 15.089810371398926,
+      "learning_rate": 1.5686053077357425e-05,
+      "loss": 0.979,
+      "step": 11460
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 34.98917007446289,
+      "learning_rate": 1.568228872576699e-05,
+      "loss": 0.7631,
+      "step": 11470
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 7.224375247955322,
+      "learning_rate": 1.567852437417655e-05,
+      "loss": 0.7461,
+      "step": 11480
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.452284812927246,
+      "learning_rate": 1.567476002258611e-05,
+      "loss": 0.6214,
+      "step": 11490
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 12.092110633850098,
+      "learning_rate": 1.567099567099567e-05,
+      "loss": 0.6558,
+      "step": 11500
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 10.905387878417969,
+      "learning_rate": 1.5667231319405234e-05,
+      "loss": 0.6047,
+      "step": 11510
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 10.892648696899414,
+      "learning_rate": 1.5663466967814794e-05,
+      "loss": 0.8369,
+      "step": 11520
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.8297064304351807,
+      "learning_rate": 1.5659702616224357e-05,
+      "loss": 0.6804,
+      "step": 11530
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 6.064371585845947,
+      "learning_rate": 1.5655938264633917e-05,
+      "loss": 0.6657,
+      "step": 11540
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 8.448015213012695,
+      "learning_rate": 1.565217391304348e-05,
+      "loss": 0.6027,
+      "step": 11550
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 39.2706413269043,
+      "learning_rate": 1.5648409561453043e-05,
+      "loss": 0.7446,
+      "step": 11560
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 17.710491180419922,
+      "learning_rate": 1.5644645209862603e-05,
+      "loss": 0.9606,
+      "step": 11570
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 8.026297569274902,
+      "learning_rate": 1.5640880858272166e-05,
+      "loss": 0.6021,
+      "step": 11580
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.3070855140686035,
+      "learning_rate": 1.5637116506681723e-05,
+      "loss": 0.5271,
+      "step": 11590
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 10.307164192199707,
+      "learning_rate": 1.5633352155091286e-05,
+      "loss": 0.6096,
+      "step": 11600
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.217769145965576,
+      "learning_rate": 1.562958780350085e-05,
+      "loss": 0.4626,
+      "step": 11610
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 33.302310943603516,
+      "learning_rate": 1.562582345191041e-05,
+      "loss": 0.5173,
+      "step": 11620
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.040510654449463,
+      "learning_rate": 1.5622059100319972e-05,
+      "loss": 0.9346,
+      "step": 11630
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 30.57442855834961,
+      "learning_rate": 1.5618294748729532e-05,
+      "loss": 0.8653,
+      "step": 11640
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.821202754974365,
+      "learning_rate": 1.5614530397139095e-05,
+      "loss": 0.8027,
+      "step": 11650
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.6210570335388184,
+      "learning_rate": 1.5610766045548655e-05,
+      "loss": 0.7766,
+      "step": 11660
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.999725818634033,
+      "learning_rate": 1.5607001693958218e-05,
+      "loss": 0.9771,
+      "step": 11670
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 9.624933242797852,
+      "learning_rate": 1.5603237342367778e-05,
+      "loss": 0.5936,
+      "step": 11680
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 10.115823745727539,
+      "learning_rate": 1.559947299077734e-05,
+      "loss": 1.1893,
+      "step": 11690
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 7.536423683166504,
+      "learning_rate": 1.55957086391869e-05,
+      "loss": 0.8263,
+      "step": 11700
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 15.722282409667969,
+      "learning_rate": 1.5591944287596464e-05,
+      "loss": 0.5882,
+      "step": 11710
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 22.392576217651367,
+      "learning_rate": 1.5588179936006023e-05,
+      "loss": 0.6934,
+      "step": 11720
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 28.885343551635742,
+      "learning_rate": 1.5584415584415587e-05,
+      "loss": 1.0459,
+      "step": 11730
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 20.01675796508789,
+      "learning_rate": 1.5580651232825146e-05,
+      "loss": 0.7343,
+      "step": 11740
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.547024726867676,
+      "learning_rate": 1.557688688123471e-05,
+      "loss": 0.8088,
+      "step": 11750
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 7.1423726081848145,
+      "learning_rate": 1.557312252964427e-05,
+      "loss": 0.5475,
+      "step": 11760
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 29.471176147460938,
+      "learning_rate": 1.556935817805383e-05,
+      "loss": 0.8382,
+      "step": 11770
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 16.62720489501953,
+      "learning_rate": 1.5565593826463392e-05,
+      "loss": 0.7917,
+      "step": 11780
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 12.257611274719238,
+      "learning_rate": 1.5561829474872955e-05,
+      "loss": 0.7627,
+      "step": 11790
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 16.232980728149414,
+      "learning_rate": 1.5558065123282515e-05,
+      "loss": 0.5723,
+      "step": 11800
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 15.463409423828125,
+      "learning_rate": 1.5554300771692078e-05,
+      "loss": 0.6709,
+      "step": 11810
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 11.793865203857422,
+      "learning_rate": 1.5550536420101638e-05,
+      "loss": 0.6177,
+      "step": 11820
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 23.931665420532227,
+      "learning_rate": 1.55467720685112e-05,
+      "loss": 0.8577,
+      "step": 11830
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 12.713443756103516,
+      "learning_rate": 1.554300771692076e-05,
+      "loss": 0.7765,
+      "step": 11840
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 19.967254638671875,
+      "learning_rate": 1.5539243365330324e-05,
+      "loss": 1.0167,
+      "step": 11850
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 15.038551330566406,
+      "learning_rate": 1.5535479013739884e-05,
+      "loss": 1.0065,
+      "step": 11860
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 11.947111129760742,
+      "learning_rate": 1.5531714662149447e-05,
+      "loss": 0.7426,
+      "step": 11870
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 23.100839614868164,
+      "learning_rate": 1.5527950310559007e-05,
+      "loss": 1.2186,
+      "step": 11880
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 18.139780044555664,
+      "learning_rate": 1.552418595896857e-05,
+      "loss": 0.5974,
+      "step": 11890
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 12.937055587768555,
+      "learning_rate": 1.552042160737813e-05,
+      "loss": 1.0542,
+      "step": 11900
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 25.850690841674805,
+      "learning_rate": 1.5516657255787693e-05,
+      "loss": 0.6753,
+      "step": 11910
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 11.147032737731934,
+      "learning_rate": 1.5512892904197253e-05,
+      "loss": 0.8321,
+      "step": 11920
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 16.01250648498535,
+      "learning_rate": 1.5509128552606816e-05,
+      "loss": 0.6835,
+      "step": 11930
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 14.47998332977295,
+      "learning_rate": 1.5505364201016376e-05,
+      "loss": 0.4995,
+      "step": 11940
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 29.964704513549805,
+      "learning_rate": 1.5501599849425935e-05,
+      "loss": 0.6967,
+      "step": 11950
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 17.843643188476562,
+      "learning_rate": 1.54978354978355e-05,
+      "loss": 0.8046,
+      "step": 11960
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 10.963820457458496,
+      "learning_rate": 1.5494071146245058e-05,
+      "loss": 0.8361,
+      "step": 11970
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 15.843602180480957,
+      "learning_rate": 1.549030679465462e-05,
+      "loss": 0.6498,
+      "step": 11980
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7258365154266357,
+      "learning_rate": 1.5486542443064185e-05,
+      "loss": 0.8831,
+      "step": 11990
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.571528434753418,
+      "learning_rate": 1.5482778091473744e-05,
+      "loss": 0.8198,
+      "step": 12000
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 8.469624519348145,
+      "learning_rate": 1.5479013739883307e-05,
+      "loss": 0.7245,
+      "step": 12010
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.154909133911133,
+      "learning_rate": 1.5475249388292867e-05,
+      "loss": 0.7674,
+      "step": 12020
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 21.369901657104492,
+      "learning_rate": 1.547148503670243e-05,
+      "loss": 0.9809,
+      "step": 12030
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 31.945472717285156,
+      "learning_rate": 1.546772068511199e-05,
+      "loss": 0.7597,
+      "step": 12040
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 25.05544662475586,
+      "learning_rate": 1.5463956333521553e-05,
+      "loss": 0.5625,
+      "step": 12050
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 21.71693992614746,
+      "learning_rate": 1.5460191981931113e-05,
+      "loss": 0.5271,
+      "step": 12060
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 24.646568298339844,
+      "learning_rate": 1.5456427630340676e-05,
+      "loss": 0.7604,
+      "step": 12070
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.2769088745117188,
+      "learning_rate": 1.5452663278750236e-05,
+      "loss": 0.7491,
+      "step": 12080
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 15.733813285827637,
+      "learning_rate": 1.54488989271598e-05,
+      "loss": 0.9759,
+      "step": 12090
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 17.16128158569336,
+      "learning_rate": 1.544513457556936e-05,
+      "loss": 0.8524,
+      "step": 12100
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 11.384461402893066,
+      "learning_rate": 1.544137022397892e-05,
+      "loss": 0.7966,
+      "step": 12110
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 13.934250831604004,
+      "learning_rate": 1.5437605872388482e-05,
+      "loss": 0.8525,
+      "step": 12120
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 11.015745162963867,
+      "learning_rate": 1.543384152079804e-05,
+      "loss": 0.617,
+      "step": 12130
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 31.421207427978516,
+      "learning_rate": 1.5430077169207605e-05,
+      "loss": 0.4724,
+      "step": 12140
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 27.373411178588867,
+      "learning_rate": 1.5426312817617165e-05,
+      "loss": 0.6447,
+      "step": 12150
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.831554889678955,
+      "learning_rate": 1.5422548466026728e-05,
+      "loss": 0.7196,
+      "step": 12160
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 17.245033264160156,
+      "learning_rate": 1.541878411443629e-05,
+      "loss": 0.9213,
+      "step": 12170
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 12.337189674377441,
+      "learning_rate": 1.541501976284585e-05,
+      "loss": 0.8848,
+      "step": 12180
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 14.200904846191406,
+      "learning_rate": 1.5411255411255414e-05,
+      "loss": 0.9418,
+      "step": 12190
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 25.789546966552734,
+      "learning_rate": 1.5407491059664974e-05,
+      "loss": 0.9909,
+      "step": 12200
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 67.17901611328125,
+      "learning_rate": 1.5403726708074537e-05,
+      "loss": 0.7414,
+      "step": 12210
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 11.221137046813965,
+      "learning_rate": 1.5399962356484096e-05,
+      "loss": 1.1104,
+      "step": 12220
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 13.688097953796387,
+      "learning_rate": 1.539619800489366e-05,
+      "loss": 0.6626,
+      "step": 12230
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 16.81260871887207,
+      "learning_rate": 1.539243365330322e-05,
+      "loss": 0.6778,
+      "step": 12240
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 7.933690071105957,
+      "learning_rate": 1.5388669301712783e-05,
+      "loss": 0.913,
+      "step": 12250
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 30.486112594604492,
+      "learning_rate": 1.5384904950122342e-05,
+      "loss": 1.057,
+      "step": 12260
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 19.049509048461914,
+      "learning_rate": 1.5381140598531905e-05,
+      "loss": 0.7562,
+      "step": 12270
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 21.68576431274414,
+      "learning_rate": 1.5377376246941465e-05,
+      "loss": 0.696,
+      "step": 12280
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 21.92389678955078,
+      "learning_rate": 1.5373611895351025e-05,
+      "loss": 0.845,
+      "step": 12290
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 11.414402961730957,
+      "learning_rate": 1.5369847543760588e-05,
+      "loss": 0.7288,
+      "step": 12300
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 12.205060005187988,
+      "learning_rate": 1.5366083192170148e-05,
+      "loss": 0.7567,
+      "step": 12310
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 21.142671585083008,
+      "learning_rate": 1.536231884057971e-05,
+      "loss": 0.6949,
+      "step": 12320
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.8780596256256104,
+      "learning_rate": 1.535855448898927e-05,
+      "loss": 0.5205,
+      "step": 12330
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.610788106918335,
+      "learning_rate": 1.5354790137398834e-05,
+      "loss": 0.4817,
+      "step": 12340
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.773406505584717,
+      "learning_rate": 1.5351025785808397e-05,
+      "loss": 0.9445,
+      "step": 12350
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.580637454986572,
+      "learning_rate": 1.5347261434217957e-05,
+      "loss": 0.7737,
+      "step": 12360
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 20.938676834106445,
+      "learning_rate": 1.534349708262752e-05,
+      "loss": 0.8886,
+      "step": 12370
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 16.42961311340332,
+      "learning_rate": 1.533973273103708e-05,
+      "loss": 0.48,
+      "step": 12380
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.3253493309021,
+      "learning_rate": 1.5335968379446643e-05,
+      "loss": 0.7892,
+      "step": 12390
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.634984970092773,
+      "learning_rate": 1.5332204027856203e-05,
+      "loss": 0.805,
+      "step": 12400
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 19.899145126342773,
+      "learning_rate": 1.5328439676265766e-05,
+      "loss": 0.7257,
+      "step": 12410
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 11.208027839660645,
+      "learning_rate": 1.5324675324675326e-05,
+      "loss": 0.732,
+      "step": 12420
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 33.87583541870117,
+      "learning_rate": 1.532091097308489e-05,
+      "loss": 0.6432,
+      "step": 12430
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 8.736035346984863,
+      "learning_rate": 1.531714662149445e-05,
+      "loss": 0.7492,
+      "step": 12440
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 13.03171443939209,
+      "learning_rate": 1.5313382269904012e-05,
+      "loss": 0.9966,
+      "step": 12450
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 9.763032913208008,
+      "learning_rate": 1.530961791831357e-05,
+      "loss": 0.8488,
+      "step": 12460
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.931615948677063,
+      "learning_rate": 1.530585356672313e-05,
+      "loss": 0.6715,
+      "step": 12470
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 12.652198791503906,
+      "learning_rate": 1.5302089215132694e-05,
+      "loss": 1.1259,
+      "step": 12480
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 6.345569133758545,
+      "learning_rate": 1.5298324863542254e-05,
+      "loss": 0.7731,
+      "step": 12490
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 22.567358016967773,
+      "learning_rate": 1.5294560511951817e-05,
+      "loss": 0.6017,
+      "step": 12500
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.477397918701172,
+      "learning_rate": 1.5290796160361377e-05,
+      "loss": 1.0937,
+      "step": 12510
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 24.150218963623047,
+      "learning_rate": 1.528703180877094e-05,
+      "loss": 0.8269,
+      "step": 12520
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 33.526763916015625,
+      "learning_rate": 1.5283267457180503e-05,
+      "loss": 0.7596,
+      "step": 12530
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 11.938372611999512,
+      "learning_rate": 1.5279503105590063e-05,
+      "loss": 0.7197,
+      "step": 12540
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7802831530570984,
+      "learning_rate": 1.5275738753999626e-05,
+      "loss": 0.5031,
+      "step": 12550
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 33.820281982421875,
+      "learning_rate": 1.5271974402409186e-05,
+      "loss": 0.7919,
+      "step": 12560
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 16.24278450012207,
+      "learning_rate": 1.526821005081875e-05,
+      "loss": 0.5727,
+      "step": 12570
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 8.651189804077148,
+      "learning_rate": 1.526444569922831e-05,
+      "loss": 0.6429,
+      "step": 12580
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 42.32835388183594,
+      "learning_rate": 1.5260681347637872e-05,
+      "loss": 0.9006,
+      "step": 12590
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 15.94575309753418,
+      "learning_rate": 1.5256916996047434e-05,
+      "loss": 0.7776,
+      "step": 12600
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 34.81913375854492,
+      "learning_rate": 1.5253152644456995e-05,
+      "loss": 1.2596,
+      "step": 12610
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.684762954711914,
+      "learning_rate": 1.5249388292866557e-05,
+      "loss": 0.6622,
+      "step": 12620
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 14.016319274902344,
+      "learning_rate": 1.5245623941276115e-05,
+      "loss": 0.7527,
+      "step": 12630
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 24.80596351623535,
+      "learning_rate": 1.5241859589685676e-05,
+      "loss": 1.1414,
+      "step": 12640
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 11.010673522949219,
+      "learning_rate": 1.523809523809524e-05,
+      "loss": 0.6522,
+      "step": 12650
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 11.180171966552734,
+      "learning_rate": 1.52343308865048e-05,
+      "loss": 0.4195,
+      "step": 12660
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 18.080059051513672,
+      "learning_rate": 1.5230566534914362e-05,
+      "loss": 0.532,
+      "step": 12670
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 16.197202682495117,
+      "learning_rate": 1.5226802183323924e-05,
+      "loss": 1.0089,
+      "step": 12680
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 13.255231857299805,
+      "learning_rate": 1.5223037831733485e-05,
+      "loss": 0.9259,
+      "step": 12690
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 6.038920879364014,
+      "learning_rate": 1.5219273480143047e-05,
+      "loss": 0.7511,
+      "step": 12700
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 6.505653381347656,
+      "learning_rate": 1.5215509128552608e-05,
+      "loss": 0.8873,
+      "step": 12710
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 10.185829162597656,
+      "learning_rate": 1.521174477696217e-05,
+      "loss": 0.9914,
+      "step": 12720
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 22.103660583496094,
+      "learning_rate": 1.5207980425371731e-05,
+      "loss": 1.0945,
+      "step": 12730
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.609415531158447,
+      "learning_rate": 1.5204216073781292e-05,
+      "loss": 0.7768,
+      "step": 12740
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 21.126070022583008,
+      "learning_rate": 1.5200451722190854e-05,
+      "loss": 0.8569,
+      "step": 12750
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.481779098510742,
+      "learning_rate": 1.5196687370600415e-05,
+      "loss": 0.6686,
+      "step": 12760
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.01765775680542,
+      "learning_rate": 1.5192923019009977e-05,
+      "loss": 0.6434,
+      "step": 12770
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 8.629416465759277,
+      "learning_rate": 1.5189158667419538e-05,
+      "loss": 0.6469,
+      "step": 12780
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 14.828179359436035,
+      "learning_rate": 1.5185394315829101e-05,
+      "loss": 0.7293,
+      "step": 12790
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 10.851454734802246,
+      "learning_rate": 1.5181629964238663e-05,
+      "loss": 0.7607,
+      "step": 12800
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 22.300085067749023,
+      "learning_rate": 1.5177865612648221e-05,
+      "loss": 0.8268,
+      "step": 12810
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 8.352980613708496,
+      "learning_rate": 1.5174101261057782e-05,
+      "loss": 0.6191,
+      "step": 12820
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 10.153834342956543,
+      "learning_rate": 1.5170336909467346e-05,
+      "loss": 0.8384,
+      "step": 12830
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 9.206123352050781,
+      "learning_rate": 1.5166572557876907e-05,
+      "loss": 0.7043,
+      "step": 12840
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.315232276916504,
+      "learning_rate": 1.5162808206286468e-05,
+      "loss": 0.6646,
+      "step": 12850
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 34.999691009521484,
+      "learning_rate": 1.515904385469603e-05,
+      "loss": 0.874,
+      "step": 12860
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 9.736954689025879,
+      "learning_rate": 1.5155279503105591e-05,
+      "loss": 1.0045,
+      "step": 12870
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 13.932580947875977,
+      "learning_rate": 1.5151515151515153e-05,
+      "loss": 0.8253,
+      "step": 12880
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 34.856239318847656,
+      "learning_rate": 1.5147750799924714e-05,
+      "loss": 0.9148,
+      "step": 12890
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 8.915196418762207,
+      "learning_rate": 1.5143986448334276e-05,
+      "loss": 0.799,
+      "step": 12900
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 23.154916763305664,
+      "learning_rate": 1.5140222096743837e-05,
+      "loss": 0.7314,
+      "step": 12910
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 16.326953887939453,
+      "learning_rate": 1.5136457745153399e-05,
+      "loss": 0.7348,
+      "step": 12920
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 25.118789672851562,
+      "learning_rate": 1.513269339356296e-05,
+      "loss": 0.6603,
+      "step": 12930
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 24.18308448791504,
+      "learning_rate": 1.5128929041972522e-05,
+      "loss": 0.7526,
+      "step": 12940
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.591280937194824,
+      "learning_rate": 1.5125164690382083e-05,
+      "loss": 0.8457,
+      "step": 12950
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 13.854071617126465,
+      "learning_rate": 1.5121400338791645e-05,
+      "loss": 0.7127,
+      "step": 12960
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.194824457168579,
+      "learning_rate": 1.5117635987201206e-05,
+      "loss": 0.9448,
+      "step": 12970
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 10.191553115844727,
+      "learning_rate": 1.5113871635610766e-05,
+      "loss": 0.4985,
+      "step": 12980
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 21.121976852416992,
+      "learning_rate": 1.5110107284020327e-05,
+      "loss": 0.5263,
+      "step": 12990
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 15.498823165893555,
+      "learning_rate": 1.5106342932429889e-05,
+      "loss": 0.9089,
+      "step": 13000
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 10.255945205688477,
+      "learning_rate": 1.510257858083945e-05,
+      "loss": 0.7219,
+      "step": 13010
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.906744003295898,
+      "learning_rate": 1.5098814229249013e-05,
+      "loss": 0.8106,
+      "step": 13020
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.751705169677734,
+      "learning_rate": 1.5095049877658575e-05,
+      "loss": 0.742,
+      "step": 13030
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 10.848084449768066,
+      "learning_rate": 1.5091285526068136e-05,
+      "loss": 0.6807,
+      "step": 13040
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 31.987140655517578,
+      "learning_rate": 1.5087521174477698e-05,
+      "loss": 0.4382,
+      "step": 13050
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 23.37934112548828,
+      "learning_rate": 1.5083756822887259e-05,
+      "loss": 0.7205,
+      "step": 13060
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.923102855682373,
+      "learning_rate": 1.507999247129682e-05,
+      "loss": 0.9366,
+      "step": 13070
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.3516485691070557,
+      "learning_rate": 1.5076228119706382e-05,
+      "loss": 0.6969,
+      "step": 13080
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 16.754817962646484,
+      "learning_rate": 1.5072463768115944e-05,
+      "loss": 0.6468,
+      "step": 13090
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.9715334177017212,
+      "learning_rate": 1.5068699416525505e-05,
+      "loss": 0.9202,
+      "step": 13100
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 16.75336456298828,
+      "learning_rate": 1.5064935064935066e-05,
+      "loss": 0.5497,
+      "step": 13110
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 11.597676277160645,
+      "learning_rate": 1.5061170713344628e-05,
+      "loss": 1.031,
+      "step": 13120
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 19.306570053100586,
+      "learning_rate": 1.505740636175419e-05,
+      "loss": 0.5285,
+      "step": 13130
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 11.259316444396973,
+      "learning_rate": 1.505364201016375e-05,
+      "loss": 0.6111,
+      "step": 13140
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.920182704925537,
+      "learning_rate": 1.5049877658573312e-05,
+      "loss": 0.9721,
+      "step": 13150
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.26685905456543,
+      "learning_rate": 1.5046113306982872e-05,
+      "loss": 0.5111,
+      "step": 13160
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.7324987649917603,
+      "learning_rate": 1.5042348955392434e-05,
+      "loss": 0.3585,
+      "step": 13170
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 12.388806343078613,
+      "learning_rate": 1.5038584603801995e-05,
+      "loss": 0.6431,
+      "step": 13180
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 26.238208770751953,
+      "learning_rate": 1.5034820252211556e-05,
+      "loss": 0.881,
+      "step": 13190
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 17.09404945373535,
+      "learning_rate": 1.5031055900621118e-05,
+      "loss": 0.7259,
+      "step": 13200
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 24.81346321105957,
+      "learning_rate": 1.5027291549030681e-05,
+      "loss": 0.8976,
+      "step": 13210
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 18.42681121826172,
+      "learning_rate": 1.5023527197440243e-05,
+      "loss": 0.6566,
+      "step": 13220
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 14.247604370117188,
+      "learning_rate": 1.5019762845849804e-05,
+      "loss": 0.5223,
+      "step": 13230
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 10.424066543579102,
+      "learning_rate": 1.5015998494259365e-05,
+      "loss": 0.7308,
+      "step": 13240
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.2744300067424774,
+      "learning_rate": 1.5012234142668927e-05,
+      "loss": 0.5581,
+      "step": 13250
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 13.965598106384277,
+      "learning_rate": 1.5008469791078488e-05,
+      "loss": 0.6971,
+      "step": 13260
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 15.680059432983398,
+      "learning_rate": 1.500470543948805e-05,
+      "loss": 0.823,
+      "step": 13270
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 64.32691955566406,
+      "learning_rate": 1.5000941087897611e-05,
+      "loss": 0.7521,
+      "step": 13280
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 17.173147201538086,
+      "learning_rate": 1.4997176736307173e-05,
+      "loss": 0.7143,
+      "step": 13290
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 12.266512870788574,
+      "learning_rate": 1.4993412384716734e-05,
+      "loss": 0.5146,
+      "step": 13300
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 27.415990829467773,
+      "learning_rate": 1.4989648033126296e-05,
+      "loss": 0.6267,
+      "step": 13310
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 38.314414978027344,
+      "learning_rate": 1.4985883681535857e-05,
+      "loss": 0.8287,
+      "step": 13320
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 14.848710060119629,
+      "learning_rate": 1.4982119329945417e-05,
+      "loss": 0.9447,
+      "step": 13330
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 16.18121910095215,
+      "learning_rate": 1.4978354978354978e-05,
+      "loss": 0.8354,
+      "step": 13340
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 18.633420944213867,
+      "learning_rate": 1.497459062676454e-05,
+      "loss": 0.5659,
+      "step": 13350
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 8.951250076293945,
+      "learning_rate": 1.4970826275174101e-05,
+      "loss": 0.7976,
+      "step": 13360
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 8.146002769470215,
+      "learning_rate": 1.4967061923583663e-05,
+      "loss": 0.5475,
+      "step": 13370
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 10.315155982971191,
+      "learning_rate": 1.4963297571993224e-05,
+      "loss": 0.7532,
+      "step": 13380
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 6.590341567993164,
+      "learning_rate": 1.4959533220402787e-05,
+      "loss": 0.7397,
+      "step": 13390
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 7.777267932891846,
+      "learning_rate": 1.4955768868812349e-05,
+      "loss": 0.7557,
+      "step": 13400
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 13.434835433959961,
+      "learning_rate": 1.495200451722191e-05,
+      "loss": 0.6464,
+      "step": 13410
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 15.336316108703613,
+      "learning_rate": 1.4948240165631472e-05,
+      "loss": 1.0214,
+      "step": 13420
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 9.77111530303955,
+      "learning_rate": 1.4944475814041033e-05,
+      "loss": 0.4507,
+      "step": 13430
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 10.342262268066406,
+      "learning_rate": 1.4940711462450595e-05,
+      "loss": 0.993,
+      "step": 13440
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 22.338380813598633,
+      "learning_rate": 1.4936947110860156e-05,
+      "loss": 1.1721,
+      "step": 13450
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 12.006577491760254,
+      "learning_rate": 1.4933182759269718e-05,
+      "loss": 0.7968,
+      "step": 13460
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 27.795074462890625,
+      "learning_rate": 1.4929418407679279e-05,
+      "loss": 0.9014,
+      "step": 13470
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 38.55152893066406,
+      "learning_rate": 1.492565405608884e-05,
+      "loss": 0.8012,
+      "step": 13480
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.785322666168213,
+      "learning_rate": 1.4921889704498402e-05,
+      "loss": 0.7915,
+      "step": 13490
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 11.960060119628906,
+      "learning_rate": 1.4918125352907963e-05,
+      "loss": 0.949,
+      "step": 13500
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 11.774707794189453,
+      "learning_rate": 1.4914361001317523e-05,
+      "loss": 0.5137,
+      "step": 13510
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 10.362529754638672,
+      "learning_rate": 1.4910596649727085e-05,
+      "loss": 0.7564,
+      "step": 13520
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 14.903637886047363,
+      "learning_rate": 1.4906832298136646e-05,
+      "loss": 1.0717,
+      "step": 13530
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 24.547834396362305,
+      "learning_rate": 1.4903067946546208e-05,
+      "loss": 0.675,
+      "step": 13540
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 15.174468040466309,
+      "learning_rate": 1.4899303594955769e-05,
+      "loss": 0.7592,
+      "step": 13550
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.759738922119141,
+      "learning_rate": 1.489553924336533e-05,
+      "loss": 0.8173,
+      "step": 13560
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 30.83299446105957,
+      "learning_rate": 1.4891774891774892e-05,
+      "loss": 0.6815,
+      "step": 13570
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 18.987077713012695,
+      "learning_rate": 1.4888010540184455e-05,
+      "loss": 0.7051,
+      "step": 13580
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 25.075693130493164,
+      "learning_rate": 1.4884246188594017e-05,
+      "loss": 0.9777,
+      "step": 13590
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 12.43553352355957,
+      "learning_rate": 1.4880481837003578e-05,
+      "loss": 0.6671,
+      "step": 13600
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 7.164165496826172,
+      "learning_rate": 1.487671748541314e-05,
+      "loss": 0.7385,
+      "step": 13610
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 12.529688835144043,
+      "learning_rate": 1.4872953133822701e-05,
+      "loss": 0.9938,
+      "step": 13620
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 11.181074142456055,
+      "learning_rate": 1.4869188782232262e-05,
+      "loss": 1.0485,
+      "step": 13630
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 15.510049819946289,
+      "learning_rate": 1.4865424430641824e-05,
+      "loss": 0.6744,
+      "step": 13640
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 14.156461715698242,
+      "learning_rate": 1.4861660079051385e-05,
+      "loss": 0.8552,
+      "step": 13650
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 17.33963394165039,
+      "learning_rate": 1.4857895727460947e-05,
+      "loss": 0.9491,
+      "step": 13660
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.6199512481689453,
+      "learning_rate": 1.4854131375870508e-05,
+      "loss": 0.7211,
+      "step": 13670
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 6.893851280212402,
+      "learning_rate": 1.4850367024280068e-05,
+      "loss": 0.717,
+      "step": 13680
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 11.729772567749023,
+      "learning_rate": 1.484660267268963e-05,
+      "loss": 0.4342,
+      "step": 13690
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.8101961612701416,
+      "learning_rate": 1.4842838321099191e-05,
+      "loss": 0.8445,
+      "step": 13700
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 28.97067642211914,
+      "learning_rate": 1.4839073969508752e-05,
+      "loss": 0.8899,
+      "step": 13710
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 50.73062515258789,
+      "learning_rate": 1.4835309617918314e-05,
+      "loss": 0.7837,
+      "step": 13720
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 15.3909330368042,
+      "learning_rate": 1.4831545266327875e-05,
+      "loss": 0.9568,
+      "step": 13730
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 30.514009475708008,
+      "learning_rate": 1.4827780914737437e-05,
+      "loss": 0.9959,
+      "step": 13740
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 8.172021865844727,
+      "learning_rate": 1.4824016563146998e-05,
+      "loss": 0.7495,
+      "step": 13750
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 20.11758041381836,
+      "learning_rate": 1.4820252211556561e-05,
+      "loss": 0.549,
+      "step": 13760
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.79815274477005,
+      "learning_rate": 1.4816487859966123e-05,
+      "loss": 0.5851,
+      "step": 13770
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 60.54981231689453,
+      "learning_rate": 1.4812723508375684e-05,
+      "loss": 0.7835,
+      "step": 13780
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 10.375343322753906,
+      "learning_rate": 1.4808959156785246e-05,
+      "loss": 0.9386,
+      "step": 13790
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 11.545114517211914,
+      "learning_rate": 1.4805194805194807e-05,
+      "loss": 0.585,
+      "step": 13800
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 18.797042846679688,
+      "learning_rate": 1.4801430453604369e-05,
+      "loss": 0.7882,
+      "step": 13810
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 18.879865646362305,
+      "learning_rate": 1.479766610201393e-05,
+      "loss": 0.6514,
+      "step": 13820
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 21.993356704711914,
+      "learning_rate": 1.4793901750423492e-05,
+      "loss": 0.7005,
+      "step": 13830
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 27.9875431060791,
+      "learning_rate": 1.4790137398833053e-05,
+      "loss": 0.9488,
+      "step": 13840
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 11.482915878295898,
+      "learning_rate": 1.4786373047242613e-05,
+      "loss": 0.7769,
+      "step": 13850
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 7.644073486328125,
+      "learning_rate": 1.4782608695652174e-05,
+      "loss": 0.3767,
+      "step": 13860
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 15.943230628967285,
+      "learning_rate": 1.4778844344061736e-05,
+      "loss": 0.7765,
+      "step": 13870
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 23.832786560058594,
+      "learning_rate": 1.4775079992471297e-05,
+      "loss": 0.7166,
+      "step": 13880
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 9.894569396972656,
+      "learning_rate": 1.4771315640880859e-05,
+      "loss": 0.7431,
+      "step": 13890
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 18.379024505615234,
+      "learning_rate": 1.476755128929042e-05,
+      "loss": 0.8358,
+      "step": 13900
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.7742643356323242,
+      "learning_rate": 1.4763786937699982e-05,
+      "loss": 0.5125,
+      "step": 13910
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 10.236210823059082,
+      "learning_rate": 1.4760022586109543e-05,
+      "loss": 0.6742,
+      "step": 13920
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 21.775062561035156,
+      "learning_rate": 1.4756258234519105e-05,
+      "loss": 0.5442,
+      "step": 13930
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 12.80639362335205,
+      "learning_rate": 1.4752493882928666e-05,
+      "loss": 0.8378,
+      "step": 13940
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 12.556268692016602,
+      "learning_rate": 1.4748729531338229e-05,
+      "loss": 0.7747,
+      "step": 13950
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 9.039406776428223,
+      "learning_rate": 1.474496517974779e-05,
+      "loss": 0.7077,
+      "step": 13960
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 15.263288497924805,
+      "learning_rate": 1.4741200828157352e-05,
+      "loss": 0.6179,
+      "step": 13970
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 18.868026733398438,
+      "learning_rate": 1.4737436476566914e-05,
+      "loss": 0.8249,
+      "step": 13980
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 35.89374923706055,
+      "learning_rate": 1.4733672124976475e-05,
+      "loss": 0.9226,
+      "step": 13990
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 15.616324424743652,
+      "learning_rate": 1.4729907773386036e-05,
+      "loss": 1.0765,
+      "step": 14000
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.1768288612365723,
+      "learning_rate": 1.4726143421795598e-05,
+      "loss": 0.7274,
+      "step": 14010
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 24.081043243408203,
+      "learning_rate": 1.472237907020516e-05,
+      "loss": 0.8053,
+      "step": 14020
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.48431396484375,
+      "learning_rate": 1.4718614718614719e-05,
+      "loss": 0.5182,
+      "step": 14030
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 15.65145492553711,
+      "learning_rate": 1.471485036702428e-05,
+      "loss": 0.4912,
+      "step": 14040
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 7.0879316329956055,
+      "learning_rate": 1.4711086015433842e-05,
+      "loss": 0.6764,
+      "step": 14050
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 12.771845817565918,
+      "learning_rate": 1.4707321663843404e-05,
+      "loss": 0.9551,
+      "step": 14060
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 14.33671760559082,
+      "learning_rate": 1.4703557312252965e-05,
+      "loss": 0.8178,
+      "step": 14070
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 7.756180286407471,
+      "learning_rate": 1.4699792960662526e-05,
+      "loss": 0.8128,
+      "step": 14080
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.8817169666290283,
+      "learning_rate": 1.4696028609072088e-05,
+      "loss": 0.6818,
+      "step": 14090
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 25.685022354125977,
+      "learning_rate": 1.469226425748165e-05,
+      "loss": 1.0637,
+      "step": 14100
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 7.975883483886719,
+      "learning_rate": 1.468849990589121e-05,
+      "loss": 0.6875,
+      "step": 14110
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 11.329687118530273,
+      "learning_rate": 1.4684735554300772e-05,
+      "loss": 0.7874,
+      "step": 14120
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 28.03009033203125,
+      "learning_rate": 1.4680971202710335e-05,
+      "loss": 0.3694,
+      "step": 14130
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 15.965397834777832,
+      "learning_rate": 1.4677206851119897e-05,
+      "loss": 0.8764,
+      "step": 14140
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 25.158184051513672,
+      "learning_rate": 1.4673442499529458e-05,
+      "loss": 0.7737,
+      "step": 14150
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 6.6749725341796875,
+      "learning_rate": 1.466967814793902e-05,
+      "loss": 0.587,
+      "step": 14160
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 11.725072860717773,
+      "learning_rate": 1.4665913796348581e-05,
+      "loss": 0.7353,
+      "step": 14170
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 17.462539672851562,
+      "learning_rate": 1.4662149444758143e-05,
+      "loss": 0.9037,
+      "step": 14180
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 7.446224212646484,
+      "learning_rate": 1.4658385093167704e-05,
+      "loss": 0.7478,
+      "step": 14190
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 26.2742919921875,
+      "learning_rate": 1.4654620741577264e-05,
+      "loss": 0.8547,
+      "step": 14200
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 26.45722770690918,
+      "learning_rate": 1.4650856389986825e-05,
+      "loss": 1.2598,
+      "step": 14210
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 30.59162139892578,
+      "learning_rate": 1.4647092038396387e-05,
+      "loss": 0.9227,
+      "step": 14220
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 2.223261594772339,
+      "learning_rate": 1.4643327686805948e-05,
+      "loss": 0.6894,
+      "step": 14230
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 12.304634094238281,
+      "learning_rate": 1.463956333521551e-05,
+      "loss": 0.938,
+      "step": 14240
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 18.860149383544922,
+      "learning_rate": 1.4635798983625071e-05,
+      "loss": 0.7099,
+      "step": 14250
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 26.4725399017334,
+      "learning_rate": 1.4632034632034633e-05,
+      "loss": 0.6649,
+      "step": 14260
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 16.648344039916992,
+      "learning_rate": 1.4628270280444194e-05,
+      "loss": 0.7711,
+      "step": 14270
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 10.517045974731445,
+      "learning_rate": 1.4624505928853756e-05,
+      "loss": 0.9333,
+      "step": 14280
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 11.410792350769043,
+      "learning_rate": 1.4620741577263317e-05,
+      "loss": 0.5923,
+      "step": 14290
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 16.357027053833008,
+      "learning_rate": 1.4616977225672879e-05,
+      "loss": 0.9859,
+      "step": 14300
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 16.898576736450195,
+      "learning_rate": 1.461321287408244e-05,
+      "loss": 0.5869,
+      "step": 14310
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 24.384197235107422,
+      "learning_rate": 1.4609448522492003e-05,
+      "loss": 1.1913,
+      "step": 14320
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 19.52752685546875,
+      "learning_rate": 1.4605684170901565e-05,
+      "loss": 0.4424,
+      "step": 14330
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 24.645038604736328,
+      "learning_rate": 1.4601919819311126e-05,
+      "loss": 0.8484,
+      "step": 14340
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 6.663630485534668,
+      "learning_rate": 1.4598155467720688e-05,
+      "loss": 0.8926,
+      "step": 14350
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 36.7334098815918,
+      "learning_rate": 1.4594391116130249e-05,
+      "loss": 0.8592,
+      "step": 14360
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 8.041152954101562,
+      "learning_rate": 1.459062676453981e-05,
+      "loss": 0.4931,
+      "step": 14370
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.520535469055176,
+      "learning_rate": 1.458686241294937e-05,
+      "loss": 0.6731,
+      "step": 14380
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 18.924875259399414,
+      "learning_rate": 1.4583098061358932e-05,
+      "loss": 0.6853,
+      "step": 14390
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 8.557247161865234,
+      "learning_rate": 1.4579333709768493e-05,
+      "loss": 0.7192,
+      "step": 14400
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 14.18856143951416,
+      "learning_rate": 1.4575569358178055e-05,
+      "loss": 0.6096,
+      "step": 14410
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 8.3154296875,
+      "learning_rate": 1.4571805006587616e-05,
+      "loss": 0.6547,
+      "step": 14420
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 22.346403121948242,
+      "learning_rate": 1.4568040654997178e-05,
+      "loss": 0.6901,
+      "step": 14430
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 26.471521377563477,
+      "learning_rate": 1.4564276303406739e-05,
+      "loss": 0.8754,
+      "step": 14440
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 18.310527801513672,
+      "learning_rate": 1.45605119518163e-05,
+      "loss": 0.6152,
+      "step": 14450
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 9.209930419921875,
+      "learning_rate": 1.4556747600225862e-05,
+      "loss": 0.8826,
+      "step": 14460
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 7.652730464935303,
+      "learning_rate": 1.4552983248635423e-05,
+      "loss": 0.4942,
+      "step": 14470
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7166558504104614,
+      "learning_rate": 1.4549218897044985e-05,
+      "loss": 0.6347,
+      "step": 14480
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 23.243921279907227,
+      "learning_rate": 1.4545454545454546e-05,
+      "loss": 0.9109,
+      "step": 14490
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 24.10265350341797,
+      "learning_rate": 1.4541690193864108e-05,
+      "loss": 0.5445,
+      "step": 14500
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 7.200760364532471,
+      "learning_rate": 1.4537925842273671e-05,
+      "loss": 0.8351,
+      "step": 14510
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 23.413759231567383,
+      "learning_rate": 1.4534161490683232e-05,
+      "loss": 0.9889,
+      "step": 14520
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 11.395479202270508,
+      "learning_rate": 1.4530397139092794e-05,
+      "loss": 0.548,
+      "step": 14530
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 18.534324645996094,
+      "learning_rate": 1.4526632787502355e-05,
+      "loss": 0.5068,
+      "step": 14540
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 22.6841983795166,
+      "learning_rate": 1.4522868435911915e-05,
+      "loss": 0.933,
+      "step": 14550
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 26.522497177124023,
+      "learning_rate": 1.4519104084321477e-05,
+      "loss": 0.9139,
+      "step": 14560
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.6627997756004333,
+      "learning_rate": 1.4515339732731038e-05,
+      "loss": 0.7262,
+      "step": 14570
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 8.96130084991455,
+      "learning_rate": 1.45115753811406e-05,
+      "loss": 0.8326,
+      "step": 14580
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 8.547167778015137,
+      "learning_rate": 1.4507811029550161e-05,
+      "loss": 0.8872,
+      "step": 14590
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 8.127227783203125,
+      "learning_rate": 1.4504046677959722e-05,
+      "loss": 0.7324,
+      "step": 14600
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 33.58380126953125,
+      "learning_rate": 1.4500282326369284e-05,
+      "loss": 0.8574,
+      "step": 14610
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.6204919815063477,
+      "learning_rate": 1.4496517974778845e-05,
+      "loss": 0.6359,
+      "step": 14620
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 32.16131591796875,
+      "learning_rate": 1.4492753623188407e-05,
+      "loss": 0.9616,
+      "step": 14630
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 12.819900512695312,
+      "learning_rate": 1.4488989271597968e-05,
+      "loss": 0.6547,
+      "step": 14640
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 22.329343795776367,
+      "learning_rate": 1.448522492000753e-05,
+      "loss": 0.5535,
+      "step": 14650
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 24.37099266052246,
+      "learning_rate": 1.4481460568417091e-05,
+      "loss": 0.935,
+      "step": 14660
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 24.315109252929688,
+      "learning_rate": 1.4477696216826653e-05,
+      "loss": 0.8781,
+      "step": 14670
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 2.761948823928833,
+      "learning_rate": 1.4473931865236214e-05,
+      "loss": 0.7095,
+      "step": 14680
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 20.340417861938477,
+      "learning_rate": 1.4470167513645777e-05,
+      "loss": 0.9713,
+      "step": 14690
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 14.535849571228027,
+      "learning_rate": 1.4466403162055339e-05,
+      "loss": 0.5641,
+      "step": 14700
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 18.312532424926758,
+      "learning_rate": 1.44626388104649e-05,
+      "loss": 0.8118,
+      "step": 14710
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 15.691104888916016,
+      "learning_rate": 1.4458874458874458e-05,
+      "loss": 0.6316,
+      "step": 14720
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 10.579209327697754,
+      "learning_rate": 1.445511010728402e-05,
+      "loss": 0.4896,
+      "step": 14730
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 13.167963981628418,
+      "learning_rate": 1.4451345755693583e-05,
+      "loss": 0.9028,
+      "step": 14740
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 18.209197998046875,
+      "learning_rate": 1.4447581404103144e-05,
+      "loss": 0.7617,
+      "step": 14750
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 7.098175525665283,
+      "learning_rate": 1.4443817052512706e-05,
+      "loss": 0.9601,
+      "step": 14760
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 29.88236427307129,
+      "learning_rate": 1.4440052700922267e-05,
+      "loss": 0.9673,
+      "step": 14770
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 14.183192253112793,
+      "learning_rate": 1.4436288349331829e-05,
+      "loss": 0.9759,
+      "step": 14780
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.7883967161178589,
+      "learning_rate": 1.443252399774139e-05,
+      "loss": 0.4751,
+      "step": 14790
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 25.07923698425293,
+      "learning_rate": 1.4428759646150952e-05,
+      "loss": 0.8102,
+      "step": 14800
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.43187662959098816,
+      "learning_rate": 1.4424995294560513e-05,
+      "loss": 0.6616,
+      "step": 14810
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 3.4543023109436035,
+      "learning_rate": 1.4421230942970075e-05,
+      "loss": 0.8503,
+      "step": 14820
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 9.902512550354004,
+      "learning_rate": 1.4417466591379636e-05,
+      "loss": 0.7454,
+      "step": 14830
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 14.120229721069336,
+      "learning_rate": 1.4413702239789197e-05,
+      "loss": 0.7271,
+      "step": 14840
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 42.34109878540039,
+      "learning_rate": 1.4409937888198759e-05,
+      "loss": 0.6191,
+      "step": 14850
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 8.808544158935547,
+      "learning_rate": 1.440617353660832e-05,
+      "loss": 0.7787,
+      "step": 14860
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 18.57027244567871,
+      "learning_rate": 1.4402409185017882e-05,
+      "loss": 0.6487,
+      "step": 14870
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 12.555221557617188,
+      "learning_rate": 1.4398644833427445e-05,
+      "loss": 0.5236,
+      "step": 14880
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.785314083099365,
+      "learning_rate": 1.4394880481837006e-05,
+      "loss": 0.6572,
+      "step": 14890
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.444519996643066,
+      "learning_rate": 1.4391116130246565e-05,
+      "loss": 0.4675,
+      "step": 14900
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 16.588285446166992,
+      "learning_rate": 1.4387351778656126e-05,
+      "loss": 0.9234,
+      "step": 14910
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 6.433644771575928,
+      "learning_rate": 1.4383587427065689e-05,
+      "loss": 0.7184,
+      "step": 14920
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 27.862564086914062,
+      "learning_rate": 1.437982307547525e-05,
+      "loss": 0.8789,
+      "step": 14930
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 8.804747581481934,
+      "learning_rate": 1.4376058723884812e-05,
+      "loss": 0.8987,
+      "step": 14940
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 15.815366744995117,
+      "learning_rate": 1.4372294372294374e-05,
+      "loss": 0.601,
+      "step": 14950
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 16.371736526489258,
+      "learning_rate": 1.4368530020703935e-05,
+      "loss": 0.9209,
+      "step": 14960
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 20.936168670654297,
+      "learning_rate": 1.4364765669113496e-05,
+      "loss": 0.6826,
+      "step": 14970
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.5109462141990662,
+      "learning_rate": 1.4361001317523058e-05,
+      "loss": 0.5467,
+      "step": 14980
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 15.911187171936035,
+      "learning_rate": 1.435723696593262e-05,
+      "loss": 0.7498,
+      "step": 14990
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 10.85964584350586,
+      "learning_rate": 1.435347261434218e-05,
+      "loss": 0.7946,
+      "step": 15000
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 12.648409843444824,
+      "learning_rate": 1.4349708262751742e-05,
+      "loss": 0.8198,
+      "step": 15010
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 25.404829025268555,
+      "learning_rate": 1.4345943911161304e-05,
+      "loss": 0.6896,
+      "step": 15020
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 21.891569137573242,
+      "learning_rate": 1.4342179559570865e-05,
+      "loss": 0.5017,
+      "step": 15030
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 8.692461967468262,
+      "learning_rate": 1.4338415207980427e-05,
+      "loss": 0.6032,
+      "step": 15040
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 6.623054027557373,
+      "learning_rate": 1.4334650856389988e-05,
+      "loss": 0.5454,
+      "step": 15050
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 14.266157150268555,
+      "learning_rate": 1.4330886504799551e-05,
+      "loss": 0.8083,
+      "step": 15060
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 21.779327392578125,
+      "learning_rate": 1.432712215320911e-05,
+      "loss": 0.7511,
+      "step": 15070
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 14.891936302185059,
+      "learning_rate": 1.432335780161867e-05,
+      "loss": 0.7927,
+      "step": 15080
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 3.073430061340332,
+      "learning_rate": 1.4319593450028232e-05,
+      "loss": 0.6556,
+      "step": 15090
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 13.47904109954834,
+      "learning_rate": 1.4315829098437794e-05,
+      "loss": 0.8525,
+      "step": 15100
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 22.618297576904297,
+      "learning_rate": 1.4312064746847357e-05,
+      "loss": 0.7522,
+      "step": 15110
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 21.820899963378906,
+      "learning_rate": 1.4308300395256918e-05,
+      "loss": 1.0761,
+      "step": 15120
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.3761250078678131,
+      "learning_rate": 1.430453604366648e-05,
+      "loss": 0.5225,
+      "step": 15130
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.551104545593262,
+      "learning_rate": 1.4300771692076041e-05,
+      "loss": 0.7417,
+      "step": 15140
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 17.213136672973633,
+      "learning_rate": 1.4297007340485603e-05,
+      "loss": 0.7196,
+      "step": 15150
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 29.140161514282227,
+      "learning_rate": 1.4293242988895164e-05,
+      "loss": 0.6101,
+      "step": 15160
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 23.715343475341797,
+      "learning_rate": 1.4289478637304726e-05,
+      "loss": 0.9656,
+      "step": 15170
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 16.146928787231445,
+      "learning_rate": 1.4285714285714287e-05,
+      "loss": 0.5416,
+      "step": 15180
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 23.387487411499023,
+      "learning_rate": 1.4281949934123849e-05,
+      "loss": 1.0218,
+      "step": 15190
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 19.16897201538086,
+      "learning_rate": 1.427818558253341e-05,
+      "loss": 0.7901,
+      "step": 15200
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 20.92551040649414,
+      "learning_rate": 1.4274421230942972e-05,
+      "loss": 1.0405,
+      "step": 15210
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.3972930908203125,
+      "learning_rate": 1.4270656879352533e-05,
+      "loss": 1.0935,
+      "step": 15220
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 9.615336418151855,
+      "learning_rate": 1.4266892527762094e-05,
+      "loss": 0.7947,
+      "step": 15230
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 12.70883560180664,
+      "learning_rate": 1.4263128176171656e-05,
+      "loss": 0.7954,
+      "step": 15240
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 19.39516830444336,
+      "learning_rate": 1.4259363824581216e-05,
+      "loss": 0.8201,
+      "step": 15250
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 16.59307098388672,
+      "learning_rate": 1.4255599472990777e-05,
+      "loss": 0.766,
+      "step": 15260
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 7.716126441955566,
+      "learning_rate": 1.4251835121400339e-05,
+      "loss": 0.7935,
+      "step": 15270
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 3.0978472232818604,
+      "learning_rate": 1.42480707698099e-05,
+      "loss": 0.7058,
+      "step": 15280
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 2.391120195388794,
+      "learning_rate": 1.4244306418219463e-05,
+      "loss": 0.88,
+      "step": 15290
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 16.929407119750977,
+      "learning_rate": 1.4240542066629025e-05,
+      "loss": 0.6826,
+      "step": 15300
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 15.712681770324707,
+      "learning_rate": 1.4236777715038586e-05,
+      "loss": 0.9275,
+      "step": 15310
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 2.698317527770996,
+      "learning_rate": 1.4233013363448148e-05,
+      "loss": 0.7053,
+      "step": 15320
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.682209014892578,
+      "learning_rate": 1.4229249011857709e-05,
+      "loss": 0.3361,
+      "step": 15330
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 18.427043914794922,
+      "learning_rate": 1.422548466026727e-05,
+      "loss": 0.5728,
+      "step": 15340
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.598134994506836,
+      "learning_rate": 1.4221720308676832e-05,
+      "loss": 0.6496,
+      "step": 15350
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 13.333558082580566,
+      "learning_rate": 1.4217955957086393e-05,
+      "loss": 0.7265,
+      "step": 15360
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 15.676292419433594,
+      "learning_rate": 1.4214191605495955e-05,
+      "loss": 0.6903,
+      "step": 15370
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 15.788107872009277,
+      "learning_rate": 1.4210427253905516e-05,
+      "loss": 0.7214,
+      "step": 15380
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 18.944698333740234,
+      "learning_rate": 1.4206662902315078e-05,
+      "loss": 0.6256,
+      "step": 15390
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 17.18710708618164,
+      "learning_rate": 1.420289855072464e-05,
+      "loss": 0.9801,
+      "step": 15400
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 24.525888442993164,
+      "learning_rate": 1.41991341991342e-05,
+      "loss": 0.8311,
+      "step": 15410
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 15.199677467346191,
+      "learning_rate": 1.419536984754376e-05,
+      "loss": 0.6035,
+      "step": 15420
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.7926441431045532,
+      "learning_rate": 1.4191605495953322e-05,
+      "loss": 0.7877,
+      "step": 15430
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 10.499088287353516,
+      "learning_rate": 1.4187841144362883e-05,
+      "loss": 0.5538,
+      "step": 15440
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 3.2541844844818115,
+      "learning_rate": 1.4184076792772445e-05,
+      "loss": 0.5928,
+      "step": 15450
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.6842057108879089,
+      "learning_rate": 1.4180312441182006e-05,
+      "loss": 0.8708,
+      "step": 15460
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 11.153387069702148,
+      "learning_rate": 1.4176548089591568e-05,
+      "loss": 0.55,
+      "step": 15470
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 14.324100494384766,
+      "learning_rate": 1.4172783738001131e-05,
+      "loss": 0.9598,
+      "step": 15480
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 9.769564628601074,
+      "learning_rate": 1.4169019386410692e-05,
+      "loss": 0.6915,
+      "step": 15490
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 5.7416863441467285,
+      "learning_rate": 1.4165255034820254e-05,
+      "loss": 1.0917,
+      "step": 15500
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 8.69757080078125,
+      "learning_rate": 1.4161490683229815e-05,
+      "loss": 0.5941,
+      "step": 15510
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 5.992401123046875,
+      "learning_rate": 1.4157726331639377e-05,
+      "loss": 0.5866,
+      "step": 15520
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 18.785144805908203,
+      "learning_rate": 1.4153961980048938e-05,
+      "loss": 0.721,
+      "step": 15530
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 18.443172454833984,
+      "learning_rate": 1.41501976284585e-05,
+      "loss": 0.513,
+      "step": 15540
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 8.198022842407227,
+      "learning_rate": 1.4146433276868061e-05,
+      "loss": 0.8079,
+      "step": 15550
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 8.908957481384277,
+      "learning_rate": 1.4142668925277623e-05,
+      "loss": 0.656,
+      "step": 15560
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 16.126298904418945,
+      "learning_rate": 1.4138904573687184e-05,
+      "loss": 0.9063,
+      "step": 15570
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 20.169885635375977,
+      "learning_rate": 1.4135140222096746e-05,
+      "loss": 0.741,
+      "step": 15580
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.3554701507091522,
+      "learning_rate": 1.4131375870506307e-05,
+      "loss": 0.614,
+      "step": 15590
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 9.904111862182617,
+      "learning_rate": 1.4127611518915867e-05,
+      "loss": 0.8123,
+      "step": 15600
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 24.91057586669922,
+      "learning_rate": 1.4123847167325428e-05,
+      "loss": 0.7551,
+      "step": 15610
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 5.963438510894775,
+      "learning_rate": 1.412008281573499e-05,
+      "loss": 0.6714,
+      "step": 15620
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 15.89360523223877,
+      "learning_rate": 1.4116318464144551e-05,
+      "loss": 0.8974,
+      "step": 15630
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 12.206485748291016,
+      "learning_rate": 1.4112554112554113e-05,
+      "loss": 1.0773,
+      "step": 15640
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 2.8578970432281494,
+      "learning_rate": 1.4108789760963674e-05,
+      "loss": 0.5777,
+      "step": 15650
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 3.7192718982696533,
+      "learning_rate": 1.4105025409373236e-05,
+      "loss": 0.6443,
+      "step": 15660
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 15.652264595031738,
+      "learning_rate": 1.4101261057782799e-05,
+      "loss": 0.7183,
+      "step": 15670
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 4.038358688354492,
+      "learning_rate": 1.409749670619236e-05,
+      "loss": 0.3951,
+      "step": 15680
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 23.8483829498291,
+      "learning_rate": 1.4093732354601922e-05,
+      "loss": 0.7791,
+      "step": 15690
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 3.3910269737243652,
+      "learning_rate": 1.4089968003011483e-05,
+      "loss": 0.5433,
+      "step": 15700
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 3.8397445678710938,
+      "learning_rate": 1.4086203651421045e-05,
+      "loss": 0.7548,
+      "step": 15710
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 11.651528358459473,
+      "learning_rate": 1.4082439299830606e-05,
+      "loss": 0.4432,
+      "step": 15720
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 11.280261039733887,
+      "learning_rate": 1.4078674948240167e-05,
+      "loss": 0.5902,
+      "step": 15730
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 22.70057487487793,
+      "learning_rate": 1.4074910596649729e-05,
+      "loss": 0.5237,
+      "step": 15740
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 12.416793823242188,
+      "learning_rate": 1.407114624505929e-05,
+      "loss": 0.5404,
+      "step": 15750
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 23.81106948852539,
+      "learning_rate": 1.4067381893468852e-05,
+      "loss": 0.7493,
+      "step": 15760
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 21.926897048950195,
+      "learning_rate": 1.4063617541878412e-05,
+      "loss": 0.8389,
+      "step": 15770
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.98796272277832,
+      "learning_rate": 1.4059853190287973e-05,
+      "loss": 0.7776,
+      "step": 15780
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 3.577707052230835,
+      "learning_rate": 1.4056088838697535e-05,
+      "loss": 0.3931,
+      "step": 15790
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 6.863688945770264,
+      "learning_rate": 1.4052324487107096e-05,
+      "loss": 0.8432,
+      "step": 15800
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.5216230154037476,
+      "learning_rate": 1.4048560135516657e-05,
+      "loss": 0.8054,
+      "step": 15810
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 30.641324996948242,
+      "learning_rate": 1.4044795783926219e-05,
+      "loss": 0.7257,
+      "step": 15820
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 10.028465270996094,
+      "learning_rate": 1.404103143233578e-05,
+      "loss": 0.9489,
+      "step": 15830
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 27.563535690307617,
+      "learning_rate": 1.4037267080745342e-05,
+      "loss": 0.8339,
+      "step": 15840
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 8.942625045776367,
+      "learning_rate": 1.4033502729154905e-05,
+      "loss": 0.7258,
+      "step": 15850
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 16.081558227539062,
+      "learning_rate": 1.4029738377564466e-05,
+      "loss": 0.6341,
+      "step": 15860
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 6.025300025939941,
+      "learning_rate": 1.4025974025974028e-05,
+      "loss": 0.6362,
+      "step": 15870
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 15.980204582214355,
+      "learning_rate": 1.402220967438359e-05,
+      "loss": 0.8119,
+      "step": 15880
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 34.32817840576172,
+      "learning_rate": 1.401844532279315e-05,
+      "loss": 0.8354,
+      "step": 15890
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 9.960022926330566,
+      "learning_rate": 1.4014680971202712e-05,
+      "loss": 0.727,
+      "step": 15900
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 17.459871292114258,
+      "learning_rate": 1.4010916619612274e-05,
+      "loss": 0.4824,
+      "step": 15910
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 7.697139263153076,
+      "learning_rate": 1.4007152268021835e-05,
+      "loss": 0.4821,
+      "step": 15920
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 21.81731605529785,
+      "learning_rate": 1.4003387916431397e-05,
+      "loss": 0.6092,
+      "step": 15930
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.9081333333333333,
+      "eval_loss": 0.3438812494277954,
+      "eval_runtime": 51.1646,
+      "eval_samples_per_second": 146.586,
+      "eval_steps_per_second": 18.333,
+      "step": 15939
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 30.34982681274414,
+      "learning_rate": 1.3999623564840956e-05,
+      "loss": 1.0665,
+      "step": 15940
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 11.997414588928223,
+      "learning_rate": 1.3995859213250518e-05,
+      "loss": 0.8418,
+      "step": 15950
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 12.47726058959961,
+      "learning_rate": 1.399209486166008e-05,
+      "loss": 0.6524,
+      "step": 15960
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 5.827723026275635,
+      "learning_rate": 1.398833051006964e-05,
+      "loss": 0.6862,
+      "step": 15970
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 11.972338676452637,
+      "learning_rate": 1.3984566158479202e-05,
+      "loss": 0.6641,
+      "step": 15980
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 16.457120895385742,
+      "learning_rate": 1.3980801806888764e-05,
+      "loss": 0.9417,
+      "step": 15990
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 11.473155975341797,
+      "learning_rate": 1.3977037455298325e-05,
+      "loss": 0.6272,
+      "step": 16000
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 26.96430778503418,
+      "learning_rate": 1.3973273103707887e-05,
+      "loss": 0.3731,
+      "step": 16010
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 12.995346069335938,
+      "learning_rate": 1.3969508752117448e-05,
+      "loss": 0.7276,
+      "step": 16020
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 9.100690841674805,
+      "learning_rate": 1.396574440052701e-05,
+      "loss": 0.4041,
+      "step": 16030
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 29.41590690612793,
+      "learning_rate": 1.3961980048936573e-05,
+      "loss": 0.5657,
+      "step": 16040
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 1.4971169233322144,
+      "learning_rate": 1.3958215697346134e-05,
+      "loss": 0.8166,
+      "step": 16050
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 27.66693687438965,
+      "learning_rate": 1.3954451345755696e-05,
+      "loss": 0.5622,
+      "step": 16060
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 19.55706214904785,
+      "learning_rate": 1.3950686994165257e-05,
+      "loss": 0.4681,
+      "step": 16070
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 8.878369331359863,
+      "learning_rate": 1.3946922642574819e-05,
+      "loss": 0.4866,
+      "step": 16080
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 1.7334586381912231,
+      "learning_rate": 1.394315829098438e-05,
+      "loss": 0.52,
+      "step": 16090
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 11.960335731506348,
+      "learning_rate": 1.3939393939393942e-05,
+      "loss": 1.061,
+      "step": 16100
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 9.888663291931152,
+      "learning_rate": 1.3935629587803503e-05,
+      "loss": 0.6879,
+      "step": 16110
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 5.383222579956055,
+      "learning_rate": 1.3931865236213063e-05,
+      "loss": 0.8313,
+      "step": 16120
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.5276015996932983,
+      "learning_rate": 1.3928100884622624e-05,
+      "loss": 0.6315,
+      "step": 16130
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 12.056757926940918,
+      "learning_rate": 1.3924336533032186e-05,
+      "loss": 0.5598,
+      "step": 16140
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 6.8872971534729,
+      "learning_rate": 1.3920572181441747e-05,
+      "loss": 0.3837,
+      "step": 16150
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 3.6625053882598877,
+      "learning_rate": 1.3916807829851309e-05,
+      "loss": 0.6365,
+      "step": 16160
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.40750113129615784,
+      "learning_rate": 1.391304347826087e-05,
+      "loss": 0.545,
+      "step": 16170
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 5.1755170822143555,
+      "learning_rate": 1.3909279126670432e-05,
+      "loss": 0.5662,
+      "step": 16180
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 27.666248321533203,
+      "learning_rate": 1.3905514775079993e-05,
+      "loss": 0.7077,
+      "step": 16190
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 6.176600456237793,
+      "learning_rate": 1.3901750423489554e-05,
+      "loss": 0.5436,
+      "step": 16200
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 40.43038558959961,
+      "learning_rate": 1.3897986071899116e-05,
+      "loss": 0.6344,
+      "step": 16210
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 13.843911170959473,
+      "learning_rate": 1.3894221720308679e-05,
+      "loss": 0.7547,
+      "step": 16220
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 17.818954467773438,
+      "learning_rate": 1.389045736871824e-05,
+      "loss": 0.8285,
+      "step": 16230
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 23.41729164123535,
+      "learning_rate": 1.3886693017127802e-05,
+      "loss": 0.5006,
+      "step": 16240
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 24.877290725708008,
+      "learning_rate": 1.3882928665537363e-05,
+      "loss": 1.136,
+      "step": 16250
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 12.05526065826416,
+      "learning_rate": 1.3879164313946925e-05,
+      "loss": 0.703,
+      "step": 16260
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 12.669425964355469,
+      "learning_rate": 1.3875399962356486e-05,
+      "loss": 0.7638,
+      "step": 16270
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 4.3127899169921875,
+      "learning_rate": 1.3871635610766048e-05,
+      "loss": 0.8595,
+      "step": 16280
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 27.62041473388672,
+      "learning_rate": 1.3867871259175608e-05,
+      "loss": 0.8843,
+      "step": 16290
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 19.740711212158203,
+      "learning_rate": 1.3864106907585169e-05,
+      "loss": 0.6221,
+      "step": 16300
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 4.2288312911987305,
+      "learning_rate": 1.386034255599473e-05,
+      "loss": 0.6524,
+      "step": 16310
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 15.323369026184082,
+      "learning_rate": 1.3856578204404292e-05,
+      "loss": 0.9442,
+      "step": 16320
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 7.15553617477417,
+      "learning_rate": 1.3852813852813853e-05,
+      "loss": 0.6764,
+      "step": 16330
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 22.282184600830078,
+      "learning_rate": 1.3849049501223415e-05,
+      "loss": 0.6296,
+      "step": 16340
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 20.944698333740234,
+      "learning_rate": 1.3845285149632976e-05,
+      "loss": 0.6328,
+      "step": 16350
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 11.877099990844727,
+      "learning_rate": 1.3841520798042538e-05,
+      "loss": 0.668,
+      "step": 16360
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 7.8392863273620605,
+      "learning_rate": 1.38377564464521e-05,
+      "loss": 0.4543,
+      "step": 16370
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 19.019250869750977,
+      "learning_rate": 1.383399209486166e-05,
+      "loss": 0.8038,
+      "step": 16380
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.9338988065719604,
+      "learning_rate": 1.3830227743271222e-05,
+      "loss": 0.4974,
+      "step": 16390
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.640573263168335,
+      "learning_rate": 1.3826463391680784e-05,
+      "loss": 0.6984,
+      "step": 16400
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 15.033812522888184,
+      "learning_rate": 1.3822699040090347e-05,
+      "loss": 0.9574,
+      "step": 16410
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 12.06335163116455,
+      "learning_rate": 1.3818934688499908e-05,
+      "loss": 0.6254,
+      "step": 16420
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 4.687410354614258,
+      "learning_rate": 1.381517033690947e-05,
+      "loss": 0.6764,
+      "step": 16430
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 27.31790542602539,
+      "learning_rate": 1.3811405985319031e-05,
+      "loss": 0.5824,
+      "step": 16440
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 9.797171592712402,
+      "learning_rate": 1.3807641633728593e-05,
+      "loss": 0.4788,
+      "step": 16450
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 17.67564582824707,
+      "learning_rate": 1.3803877282138154e-05,
+      "loss": 1.209,
+      "step": 16460
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.271230548620224,
+      "learning_rate": 1.3800112930547714e-05,
+      "loss": 0.4775,
+      "step": 16470
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 33.21747589111328,
+      "learning_rate": 1.3796348578957275e-05,
+      "loss": 1.0736,
+      "step": 16480
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 88.40038299560547,
+      "learning_rate": 1.3792584227366837e-05,
+      "loss": 0.8484,
+      "step": 16490
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 17.692691802978516,
+      "learning_rate": 1.3788819875776398e-05,
+      "loss": 0.7804,
+      "step": 16500
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 9.918412208557129,
+      "learning_rate": 1.378505552418596e-05,
+      "loss": 0.7316,
+      "step": 16510
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 5.417355537414551,
+      "learning_rate": 1.3781291172595521e-05,
+      "loss": 0.6182,
+      "step": 16520
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 11.936298370361328,
+      "learning_rate": 1.3777526821005083e-05,
+      "loss": 0.4077,
+      "step": 16530
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 13.960946083068848,
+      "learning_rate": 1.3773762469414644e-05,
+      "loss": 0.526,
+      "step": 16540
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 22.762529373168945,
+      "learning_rate": 1.3769998117824206e-05,
+      "loss": 0.6749,
+      "step": 16550
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 14.109274864196777,
+      "learning_rate": 1.3766233766233767e-05,
+      "loss": 0.5186,
+      "step": 16560
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 35.54779815673828,
+      "learning_rate": 1.3762469414643328e-05,
+      "loss": 0.7046,
+      "step": 16570
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 12.966293334960938,
+      "learning_rate": 1.375870506305289e-05,
+      "loss": 0.7402,
+      "step": 16580
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 8.81118392944336,
+      "learning_rate": 1.3754940711462453e-05,
+      "loss": 0.6679,
+      "step": 16590
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.5349300503730774,
+      "learning_rate": 1.3751176359872015e-05,
+      "loss": 0.6342,
+      "step": 16600
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 14.064860343933105,
+      "learning_rate": 1.3747412008281576e-05,
+      "loss": 0.8408,
+      "step": 16610
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 13.386101722717285,
+      "learning_rate": 1.3743647656691137e-05,
+      "loss": 0.6574,
+      "step": 16620
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.2948550879955292,
+      "learning_rate": 1.3739883305100699e-05,
+      "loss": 0.8882,
+      "step": 16630
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 8.500995635986328,
+      "learning_rate": 1.3736118953510259e-05,
+      "loss": 0.6037,
+      "step": 16640
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 18.741573333740234,
+      "learning_rate": 1.373235460191982e-05,
+      "loss": 0.629,
+      "step": 16650
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.3394733667373657,
+      "learning_rate": 1.3728590250329382e-05,
+      "loss": 0.7729,
+      "step": 16660
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 11.697025299072266,
+      "learning_rate": 1.3724825898738943e-05,
+      "loss": 0.512,
+      "step": 16670
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 9.116700172424316,
+      "learning_rate": 1.3721061547148505e-05,
+      "loss": 0.6958,
+      "step": 16680
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 59.24357223510742,
+      "learning_rate": 1.3717297195558066e-05,
+      "loss": 0.9625,
+      "step": 16690
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 20.930126190185547,
+      "learning_rate": 1.3713532843967627e-05,
+      "loss": 0.7139,
+      "step": 16700
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 46.4968147277832,
+      "learning_rate": 1.3709768492377189e-05,
+      "loss": 0.8418,
+      "step": 16710
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.5635331273078918,
+      "learning_rate": 1.370600414078675e-05,
+      "loss": 0.7676,
+      "step": 16720
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.2128181755542755,
+      "learning_rate": 1.3702239789196312e-05,
+      "loss": 0.9451,
+      "step": 16730
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 21.21966552734375,
+      "learning_rate": 1.3698475437605873e-05,
+      "loss": 0.5834,
+      "step": 16740
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 37.33884811401367,
+      "learning_rate": 1.3694711086015435e-05,
+      "loss": 0.8985,
+      "step": 16750
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 7.059642314910889,
+      "learning_rate": 1.3690946734424996e-05,
+      "loss": 0.7002,
+      "step": 16760
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 14.130393981933594,
+      "learning_rate": 1.3687182382834558e-05,
+      "loss": 0.8018,
+      "step": 16770
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 15.686850547790527,
+      "learning_rate": 1.368341803124412e-05,
+      "loss": 0.7179,
+      "step": 16780
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 7.120452404022217,
+      "learning_rate": 1.3679653679653682e-05,
+      "loss": 0.6912,
+      "step": 16790
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 15.659395217895508,
+      "learning_rate": 1.3675889328063244e-05,
+      "loss": 0.8925,
+      "step": 16800
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 27.15504264831543,
+      "learning_rate": 1.3672124976472802e-05,
+      "loss": 0.4575,
+      "step": 16810
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 7.112409591674805,
+      "learning_rate": 1.3668360624882365e-05,
+      "loss": 0.7047,
+      "step": 16820
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 21.725502014160156,
+      "learning_rate": 1.3664596273291926e-05,
+      "loss": 0.5739,
+      "step": 16830
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 17.80892562866211,
+      "learning_rate": 1.3660831921701488e-05,
+      "loss": 0.8959,
+      "step": 16840
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 1.4555948972702026,
+      "learning_rate": 1.365706757011105e-05,
+      "loss": 0.4714,
+      "step": 16850
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 2.813807725906372,
+      "learning_rate": 1.365330321852061e-05,
+      "loss": 0.5298,
+      "step": 16860
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 15.255325317382812,
+      "learning_rate": 1.3649538866930172e-05,
+      "loss": 0.7289,
+      "step": 16870
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 10.276988983154297,
+      "learning_rate": 1.3645774515339734e-05,
+      "loss": 0.3811,
+      "step": 16880
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 4.02056884765625,
+      "learning_rate": 1.3642010163749295e-05,
+      "loss": 0.7805,
+      "step": 16890
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 20.98756217956543,
+      "learning_rate": 1.3638245812158857e-05,
+      "loss": 0.7219,
+      "step": 16900
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 1.9777402877807617,
+      "learning_rate": 1.3634481460568418e-05,
+      "loss": 0.4574,
+      "step": 16910
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 17.15428924560547,
+      "learning_rate": 1.363071710897798e-05,
+      "loss": 0.6216,
+      "step": 16920
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 26.12958335876465,
+      "learning_rate": 1.3626952757387541e-05,
+      "loss": 0.5892,
+      "step": 16930
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 2.257452964782715,
+      "learning_rate": 1.3623188405797103e-05,
+      "loss": 0.6007,
+      "step": 16940
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 21.12044334411621,
+      "learning_rate": 1.3619424054206664e-05,
+      "loss": 0.6436,
+      "step": 16950
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 27.5797061920166,
+      "learning_rate": 1.3615659702616225e-05,
+      "loss": 0.6577,
+      "step": 16960
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.14470890164375305,
+      "learning_rate": 1.3611895351025789e-05,
+      "loss": 0.7087,
+      "step": 16970
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 25.093280792236328,
+      "learning_rate": 1.360813099943535e-05,
+      "loss": 0.8523,
+      "step": 16980
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 9.746846199035645,
+      "learning_rate": 1.3604366647844908e-05,
+      "loss": 0.6063,
+      "step": 16990
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 26.40338897705078,
+      "learning_rate": 1.360060229625447e-05,
+      "loss": 0.7969,
+      "step": 17000
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 16.754243850708008,
+      "learning_rate": 1.3596837944664033e-05,
+      "loss": 0.7463,
+      "step": 17010
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 30.17478370666504,
+      "learning_rate": 1.3593073593073594e-05,
+      "loss": 0.8098,
+      "step": 17020
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 10.640960693359375,
+      "learning_rate": 1.3589309241483156e-05,
+      "loss": 0.5837,
+      "step": 17030
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 21.33949089050293,
+      "learning_rate": 1.3585544889892717e-05,
+      "loss": 0.627,
+      "step": 17040
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 11.06455135345459,
+      "learning_rate": 1.3581780538302279e-05,
+      "loss": 0.6147,
+      "step": 17050
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 25.937108993530273,
+      "learning_rate": 1.357801618671184e-05,
+      "loss": 1.1128,
+      "step": 17060
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 6.419862270355225,
+      "learning_rate": 1.3574251835121402e-05,
+      "loss": 0.8132,
+      "step": 17070
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 26.92257308959961,
+      "learning_rate": 1.3570487483530963e-05,
+      "loss": 0.4862,
+      "step": 17080
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 10.812932968139648,
+      "learning_rate": 1.3566723131940524e-05,
+      "loss": 1.0484,
+      "step": 17090
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 23.167081832885742,
+      "learning_rate": 1.3562958780350086e-05,
+      "loss": 0.815,
+      "step": 17100
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 19.610149383544922,
+      "learning_rate": 1.3559194428759647e-05,
+      "loss": 0.788,
+      "step": 17110
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 13.85695743560791,
+      "learning_rate": 1.3555430077169209e-05,
+      "loss": 0.65,
+      "step": 17120
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 4.279960632324219,
+      "learning_rate": 1.355166572557877e-05,
+      "loss": 0.6214,
+      "step": 17130
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 21.195491790771484,
+      "learning_rate": 1.3547901373988332e-05,
+      "loss": 0.5917,
+      "step": 17140
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 43.27781295776367,
+      "learning_rate": 1.3544137022397895e-05,
+      "loss": 0.5481,
+      "step": 17150
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 32.89248275756836,
+      "learning_rate": 1.3540372670807453e-05,
+      "loss": 0.7128,
+      "step": 17160
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 1.7565685510635376,
+      "learning_rate": 1.3536608319217014e-05,
+      "loss": 0.5305,
+      "step": 17170
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.9114441871643066,
+      "learning_rate": 1.3532843967626576e-05,
+      "loss": 0.4674,
+      "step": 17180
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 14.061701774597168,
+      "learning_rate": 1.3529079616036137e-05,
+      "loss": 0.947,
+      "step": 17190
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 14.889267921447754,
+      "learning_rate": 1.35253152644457e-05,
+      "loss": 0.8785,
+      "step": 17200
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 10.14665412902832,
+      "learning_rate": 1.3521550912855262e-05,
+      "loss": 0.728,
+      "step": 17210
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 16.317060470581055,
+      "learning_rate": 1.3517786561264823e-05,
+      "loss": 0.615,
+      "step": 17220
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 14.949483871459961,
+      "learning_rate": 1.3514022209674385e-05,
+      "loss": 0.4112,
+      "step": 17230
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 42.36807632446289,
+      "learning_rate": 1.3510257858083946e-05,
+      "loss": 0.5082,
+      "step": 17240
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 10.184826850891113,
+      "learning_rate": 1.3506493506493508e-05,
+      "loss": 0.63,
+      "step": 17250
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 7.961637496948242,
+      "learning_rate": 1.350272915490307e-05,
+      "loss": 0.373,
+      "step": 17260
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 15.121786117553711,
+      "learning_rate": 1.349896480331263e-05,
+      "loss": 1.0031,
+      "step": 17270
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 17.208942413330078,
+      "learning_rate": 1.3495200451722192e-05,
+      "loss": 0.7613,
+      "step": 17280
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 3.7287890911102295,
+      "learning_rate": 1.3491436100131754e-05,
+      "loss": 0.4566,
+      "step": 17290
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 34.4078254699707,
+      "learning_rate": 1.3487671748541315e-05,
+      "loss": 0.8477,
+      "step": 17300
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 6.368019104003906,
+      "learning_rate": 1.3483907396950877e-05,
+      "loss": 0.6715,
+      "step": 17310
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 6.000946044921875,
+      "learning_rate": 1.3480143045360438e-05,
+      "loss": 0.3316,
+      "step": 17320
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 8.809701919555664,
+      "learning_rate": 1.347637869377e-05,
+      "loss": 0.9712,
+      "step": 17330
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 2.8006772994995117,
+      "learning_rate": 1.347261434217956e-05,
+      "loss": 0.5571,
+      "step": 17340
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 3.8963782787323,
+      "learning_rate": 1.346884999058912e-05,
+      "loss": 0.8123,
+      "step": 17350
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 21.654504776000977,
+      "learning_rate": 1.3465085638998682e-05,
+      "loss": 0.46,
+      "step": 17360
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 21.95523452758789,
+      "learning_rate": 1.3461321287408244e-05,
+      "loss": 0.8436,
+      "step": 17370
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 21.384435653686523,
+      "learning_rate": 1.3457556935817807e-05,
+      "loss": 0.6596,
+      "step": 17380
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 2.9826455116271973,
+      "learning_rate": 1.3453792584227368e-05,
+      "loss": 0.7622,
+      "step": 17390
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 1.4264588356018066,
+      "learning_rate": 1.345002823263693e-05,
+      "loss": 0.739,
+      "step": 17400
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 7.517381191253662,
+      "learning_rate": 1.3446263881046491e-05,
+      "loss": 0.6091,
+      "step": 17410
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 32.621761322021484,
+      "learning_rate": 1.3442499529456053e-05,
+      "loss": 0.6992,
+      "step": 17420
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 4.760457992553711,
+      "learning_rate": 1.3438735177865614e-05,
+      "loss": 1.0,
+      "step": 17430
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.4995521008968353,
+      "learning_rate": 1.3434970826275176e-05,
+      "loss": 0.523,
+      "step": 17440
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 8.695629119873047,
+      "learning_rate": 1.3431206474684737e-05,
+      "loss": 0.646,
+      "step": 17450
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.28502389788627625,
+      "learning_rate": 1.3427442123094298e-05,
+      "loss": 0.5188,
+      "step": 17460
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.6979949474334717,
+      "learning_rate": 1.342367777150386e-05,
+      "loss": 0.5536,
+      "step": 17470
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 24.936214447021484,
+      "learning_rate": 1.3419913419913421e-05,
+      "loss": 0.647,
+      "step": 17480
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 8.776481628417969,
+      "learning_rate": 1.3416149068322983e-05,
+      "loss": 0.8732,
+      "step": 17490
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 19.038969039916992,
+      "learning_rate": 1.3412384716732544e-05,
+      "loss": 0.77,
+      "step": 17500
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 4.335228443145752,
+      "learning_rate": 1.3408620365142104e-05,
+      "loss": 0.6023,
+      "step": 17510
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 7.746741771697998,
+      "learning_rate": 1.3404856013551666e-05,
+      "loss": 0.6851,
+      "step": 17520
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 24.021316528320312,
+      "learning_rate": 1.3401091661961227e-05,
+      "loss": 0.8683,
+      "step": 17530
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 40.04768371582031,
+      "learning_rate": 1.3397327310370788e-05,
+      "loss": 0.7139,
+      "step": 17540
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 28.369369506835938,
+      "learning_rate": 1.339356295878035e-05,
+      "loss": 0.7984,
+      "step": 17550
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 12.895668983459473,
+      "learning_rate": 1.3389798607189911e-05,
+      "loss": 0.8768,
+      "step": 17560
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 13.10739517211914,
+      "learning_rate": 1.3386034255599475e-05,
+      "loss": 0.9014,
+      "step": 17570
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 6.7018537521362305,
+      "learning_rate": 1.3382269904009036e-05,
+      "loss": 0.6438,
+      "step": 17580
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 20.47855567932129,
+      "learning_rate": 1.3378505552418597e-05,
+      "loss": 0.5956,
+      "step": 17590
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 34.99126052856445,
+      "learning_rate": 1.3374741200828159e-05,
+      "loss": 0.6245,
+      "step": 17600
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 31.813350677490234,
+      "learning_rate": 1.337097684923772e-05,
+      "loss": 0.9122,
+      "step": 17610
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 15.438579559326172,
+      "learning_rate": 1.3367212497647282e-05,
+      "loss": 0.7581,
+      "step": 17620
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 10.1848726272583,
+      "learning_rate": 1.3363448146056843e-05,
+      "loss": 0.3745,
+      "step": 17630
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.6416832804679871,
+      "learning_rate": 1.3359683794466405e-05,
+      "loss": 0.5186,
+      "step": 17640
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 17.22964859008789,
+      "learning_rate": 1.3355919442875966e-05,
+      "loss": 0.517,
+      "step": 17650
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 23.54407501220703,
+      "learning_rate": 1.3352155091285528e-05,
+      "loss": 0.6123,
+      "step": 17660
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 24.60756492614746,
+      "learning_rate": 1.3348390739695089e-05,
+      "loss": 0.6709,
+      "step": 17670
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 18.41277313232422,
+      "learning_rate": 1.334462638810465e-05,
+      "loss": 0.724,
+      "step": 17680
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 13.058307647705078,
+      "learning_rate": 1.334086203651421e-05,
+      "loss": 0.6962,
+      "step": 17690
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 13.01867389678955,
+      "learning_rate": 1.3337097684923772e-05,
+      "loss": 0.5346,
+      "step": 17700
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 13.48055362701416,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.5472,
+      "step": 17710
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 18.029436111450195,
+      "learning_rate": 1.3329568981742895e-05,
+      "loss": 0.4834,
+      "step": 17720
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 2.22222900390625,
+      "learning_rate": 1.3325804630152456e-05,
+      "loss": 0.6628,
+      "step": 17730
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 7.793396472930908,
+      "learning_rate": 1.3322040278562018e-05,
+      "loss": 0.866,
+      "step": 17740
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 42.029598236083984,
+      "learning_rate": 1.331827592697158e-05,
+      "loss": 0.5733,
+      "step": 17750
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 8.039556503295898,
+      "learning_rate": 1.3314511575381142e-05,
+      "loss": 0.6282,
+      "step": 17760
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 13.458028793334961,
+      "learning_rate": 1.3310747223790704e-05,
+      "loss": 0.911,
+      "step": 17770
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 14.425552368164062,
+      "learning_rate": 1.3306982872200265e-05,
+      "loss": 1.0583,
+      "step": 17780
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 3.239820718765259,
+      "learning_rate": 1.3303218520609827e-05,
+      "loss": 0.5288,
+      "step": 17790
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 49.50566482543945,
+      "learning_rate": 1.3299454169019388e-05,
+      "loss": 0.9413,
+      "step": 17800
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 15.337508201599121,
+      "learning_rate": 1.329568981742895e-05,
+      "loss": 0.6259,
+      "step": 17810
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 22.253278732299805,
+      "learning_rate": 1.3291925465838511e-05,
+      "loss": 0.8045,
+      "step": 17820
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 8.542914390563965,
+      "learning_rate": 1.3288161114248073e-05,
+      "loss": 0.5947,
+      "step": 17830
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 9.067444801330566,
+      "learning_rate": 1.3284396762657634e-05,
+      "loss": 0.5303,
+      "step": 17840
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 15.519441604614258,
+      "learning_rate": 1.3280632411067195e-05,
+      "loss": 1.0939,
+      "step": 17850
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 8.470025062561035,
+      "learning_rate": 1.3276868059476755e-05,
+      "loss": 0.7933,
+      "step": 17860
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 17.665695190429688,
+      "learning_rate": 1.3273103707886317e-05,
+      "loss": 0.7463,
+      "step": 17870
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 17.912128448486328,
+      "learning_rate": 1.3269339356295878e-05,
+      "loss": 0.6041,
+      "step": 17880
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 26.95820426940918,
+      "learning_rate": 1.326557500470544e-05,
+      "loss": 0.7265,
+      "step": 17890
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 13.88226318359375,
+      "learning_rate": 1.3261810653115001e-05,
+      "loss": 0.7233,
+      "step": 17900
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 9.494647979736328,
+      "learning_rate": 1.3258046301524563e-05,
+      "loss": 0.5738,
+      "step": 17910
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 23.288188934326172,
+      "learning_rate": 1.3254281949934124e-05,
+      "loss": 0.9191,
+      "step": 17920
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 12.423943519592285,
+      "learning_rate": 1.3250517598343685e-05,
+      "loss": 0.5722,
+      "step": 17930
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 20.487674713134766,
+      "learning_rate": 1.3246753246753249e-05,
+      "loss": 0.5988,
+      "step": 17940
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 8.707905769348145,
+      "learning_rate": 1.324298889516281e-05,
+      "loss": 0.3331,
+      "step": 17950
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.4977474808692932,
+      "learning_rate": 1.3239224543572372e-05,
+      "loss": 0.5852,
+      "step": 17960
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 6.157559871673584,
+      "learning_rate": 1.3235460191981933e-05,
+      "loss": 0.6341,
+      "step": 17970
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 23.23993682861328,
+      "learning_rate": 1.3231695840391494e-05,
+      "loss": 0.9592,
+      "step": 17980
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 10.312979698181152,
+      "learning_rate": 1.3227931488801056e-05,
+      "loss": 0.7303,
+      "step": 17990
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 24.67930030822754,
+      "learning_rate": 1.3224167137210617e-05,
+      "loss": 0.5402,
+      "step": 18000
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 22.200210571289062,
+      "learning_rate": 1.3220402785620179e-05,
+      "loss": 0.8094,
+      "step": 18010
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 5.099223613739014,
+      "learning_rate": 1.321663843402974e-05,
+      "loss": 0.8314,
+      "step": 18020
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 19.834396362304688,
+      "learning_rate": 1.32128740824393e-05,
+      "loss": 1.009,
+      "step": 18030
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 11.00228214263916,
+      "learning_rate": 1.3209109730848861e-05,
+      "loss": 0.8001,
+      "step": 18040
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 19.782005310058594,
+      "learning_rate": 1.3205345379258423e-05,
+      "loss": 0.5065,
+      "step": 18050
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 9.906347274780273,
+      "learning_rate": 1.3201581027667984e-05,
+      "loss": 0.4765,
+      "step": 18060
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 28.395952224731445,
+      "learning_rate": 1.3197816676077546e-05,
+      "loss": 0.9283,
+      "step": 18070
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 4.160763740539551,
+      "learning_rate": 1.3194052324487107e-05,
+      "loss": 0.6412,
+      "step": 18080
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 16.857194900512695,
+      "learning_rate": 1.3190287972896669e-05,
+      "loss": 0.888,
+      "step": 18090
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 3.8556270599365234,
+      "learning_rate": 1.318652362130623e-05,
+      "loss": 0.8731,
+      "step": 18100
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 29.509401321411133,
+      "learning_rate": 1.3182759269715792e-05,
+      "loss": 0.6451,
+      "step": 18110
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 8.988554954528809,
+      "learning_rate": 1.3178994918125355e-05,
+      "loss": 0.9483,
+      "step": 18120
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 19.900251388549805,
+      "learning_rate": 1.3175230566534916e-05,
+      "loss": 1.0578,
+      "step": 18130
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 11.77758502960205,
+      "learning_rate": 1.3171466214944478e-05,
+      "loss": 0.6442,
+      "step": 18140
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 6.527004718780518,
+      "learning_rate": 1.316770186335404e-05,
+      "loss": 0.7697,
+      "step": 18150
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 5.472777843475342,
+      "learning_rate": 1.31639375117636e-05,
+      "loss": 0.9177,
+      "step": 18160
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 11.783276557922363,
+      "learning_rate": 1.3160173160173162e-05,
+      "loss": 0.6803,
+      "step": 18170
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 9.348353385925293,
+      "learning_rate": 1.3156408808582724e-05,
+      "loss": 0.9163,
+      "step": 18180
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 8.337559700012207,
+      "learning_rate": 1.3152644456992285e-05,
+      "loss": 0.5074,
+      "step": 18190
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 18.181156158447266,
+      "learning_rate": 1.3148880105401847e-05,
+      "loss": 0.6039,
+      "step": 18200
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 24.175315856933594,
+      "learning_rate": 1.3145115753811406e-05,
+      "loss": 0.6791,
+      "step": 18210
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 11.990982055664062,
+      "learning_rate": 1.3141351402220968e-05,
+      "loss": 0.5122,
+      "step": 18220
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 12.18565845489502,
+      "learning_rate": 1.313758705063053e-05,
+      "loss": 0.8138,
+      "step": 18230
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 20.374452590942383,
+      "learning_rate": 1.313382269904009e-05,
+      "loss": 0.4286,
+      "step": 18240
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 11.977636337280273,
+      "learning_rate": 1.3130058347449652e-05,
+      "loss": 0.6633,
+      "step": 18250
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 5.555273056030273,
+      "learning_rate": 1.3126293995859214e-05,
+      "loss": 0.3768,
+      "step": 18260
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 12.093500137329102,
+      "learning_rate": 1.3122529644268775e-05,
+      "loss": 0.5081,
+      "step": 18270
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 12.523153305053711,
+      "learning_rate": 1.3118765292678337e-05,
+      "loss": 0.7281,
+      "step": 18280
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 34.439266204833984,
+      "learning_rate": 1.3115000941087898e-05,
+      "loss": 0.7471,
+      "step": 18290
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 28.79136848449707,
+      "learning_rate": 1.311123658949746e-05,
+      "loss": 0.6478,
+      "step": 18300
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 5.4620561599731445,
+      "learning_rate": 1.3107472237907023e-05,
+      "loss": 0.8856,
+      "step": 18310
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 14.058579444885254,
+      "learning_rate": 1.3103707886316584e-05,
+      "loss": 0.4122,
+      "step": 18320
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 17.870243072509766,
+      "learning_rate": 1.3099943534726146e-05,
+      "loss": 0.4202,
+      "step": 18330
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 32.40336990356445,
+      "learning_rate": 1.3096179183135707e-05,
+      "loss": 0.9052,
+      "step": 18340
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 10.193222999572754,
+      "learning_rate": 1.3092414831545268e-05,
+      "loss": 0.5433,
+      "step": 18350
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 7.227964401245117,
+      "learning_rate": 1.308865047995483e-05,
+      "loss": 0.6235,
+      "step": 18360
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 17.942949295043945,
+      "learning_rate": 1.3084886128364391e-05,
+      "loss": 0.9279,
+      "step": 18370
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 2.4786839485168457,
+      "learning_rate": 1.3081121776773951e-05,
+      "loss": 0.5688,
+      "step": 18380
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 13.7864408493042,
+      "learning_rate": 1.3077357425183513e-05,
+      "loss": 0.6254,
+      "step": 18390
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 11.290862083435059,
+      "learning_rate": 1.3073593073593074e-05,
+      "loss": 0.5499,
+      "step": 18400
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 27.211687088012695,
+      "learning_rate": 1.3069828722002636e-05,
+      "loss": 0.7081,
+      "step": 18410
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 31.036930084228516,
+      "learning_rate": 1.3066064370412197e-05,
+      "loss": 0.692,
+      "step": 18420
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 9.828187942504883,
+      "learning_rate": 1.3062300018821758e-05,
+      "loss": 0.848,
+      "step": 18430
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 22.125829696655273,
+      "learning_rate": 1.305853566723132e-05,
+      "loss": 0.7898,
+      "step": 18440
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 15.01488971710205,
+      "learning_rate": 1.3054771315640881e-05,
+      "loss": 0.4647,
+      "step": 18450
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 14.60596752166748,
+      "learning_rate": 1.3051006964050443e-05,
+      "loss": 0.7542,
+      "step": 18460
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 23.092918395996094,
+      "learning_rate": 1.3047242612460004e-05,
+      "loss": 0.8983,
+      "step": 18470
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 12.916792869567871,
+      "learning_rate": 1.3043478260869566e-05,
+      "loss": 0.6377,
+      "step": 18480
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 30.594213485717773,
+      "learning_rate": 1.3039713909279127e-05,
+      "loss": 0.9676,
+      "step": 18490
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 18.934587478637695,
+      "learning_rate": 1.303594955768869e-05,
+      "loss": 0.8782,
+      "step": 18500
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 2.794344663619995,
+      "learning_rate": 1.3032185206098252e-05,
+      "loss": 0.3121,
+      "step": 18510
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 2.3938238620758057,
+      "learning_rate": 1.3028420854507813e-05,
+      "loss": 0.9087,
+      "step": 18520
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 32.515846252441406,
+      "learning_rate": 1.3024656502917375e-05,
+      "loss": 0.6834,
+      "step": 18530
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 4.718508720397949,
+      "learning_rate": 1.3020892151326936e-05,
+      "loss": 0.6034,
+      "step": 18540
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 16.135196685791016,
+      "learning_rate": 1.3017127799736498e-05,
+      "loss": 0.7444,
+      "step": 18550
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 15.95462417602539,
+      "learning_rate": 1.3013363448146057e-05,
+      "loss": 0.482,
+      "step": 18560
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 15.838091850280762,
+      "learning_rate": 1.3009599096555619e-05,
+      "loss": 0.7012,
+      "step": 18570
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 9.506869316101074,
+      "learning_rate": 1.300583474496518e-05,
+      "loss": 0.5384,
+      "step": 18580
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 11.965920448303223,
+      "learning_rate": 1.3002070393374742e-05,
+      "loss": 0.5091,
+      "step": 18590
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 21.009666442871094,
+      "learning_rate": 1.2998306041784303e-05,
+      "loss": 0.9806,
+      "step": 18600
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 19.624370574951172,
+      "learning_rate": 1.2994541690193865e-05,
+      "loss": 0.7226,
+      "step": 18610
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 18.199827194213867,
+      "learning_rate": 1.2990777338603426e-05,
+      "loss": 0.7008,
+      "step": 18620
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 16.78691864013672,
+      "learning_rate": 1.2987012987012988e-05,
+      "loss": 0.8391,
+      "step": 18630
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 11.214224815368652,
+      "learning_rate": 1.2983248635422549e-05,
+      "loss": 0.7674,
+      "step": 18640
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 12.816788673400879,
+      "learning_rate": 1.297948428383211e-05,
+      "loss": 0.6571,
+      "step": 18650
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 26.343637466430664,
+      "learning_rate": 1.2975719932241672e-05,
+      "loss": 0.7407,
+      "step": 18660
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 10.422934532165527,
+      "learning_rate": 1.2971955580651234e-05,
+      "loss": 0.571,
+      "step": 18670
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 4.7122015953063965,
+      "learning_rate": 1.2968191229060797e-05,
+      "loss": 0.4944,
+      "step": 18680
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 10.654007911682129,
+      "learning_rate": 1.2964426877470358e-05,
+      "loss": 0.4871,
+      "step": 18690
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 16.5805606842041,
+      "learning_rate": 1.296066252587992e-05,
+      "loss": 0.7417,
+      "step": 18700
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 12.002159118652344,
+      "learning_rate": 1.2956898174289481e-05,
+      "loss": 0.686,
+      "step": 18710
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 22.57682991027832,
+      "learning_rate": 1.2953133822699043e-05,
+      "loss": 0.9398,
+      "step": 18720
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 8.608962059020996,
+      "learning_rate": 1.2949369471108602e-05,
+      "loss": 0.5677,
+      "step": 18730
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 14.006396293640137,
+      "learning_rate": 1.2945605119518164e-05,
+      "loss": 0.635,
+      "step": 18740
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 10.348053932189941,
+      "learning_rate": 1.2941840767927725e-05,
+      "loss": 0.7357,
+      "step": 18750
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 18.81407928466797,
+      "learning_rate": 1.2938076416337287e-05,
+      "loss": 0.9314,
+      "step": 18760
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 11.178011894226074,
+      "learning_rate": 1.2934312064746848e-05,
+      "loss": 0.454,
+      "step": 18770
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 23.306005477905273,
+      "learning_rate": 1.293054771315641e-05,
+      "loss": 0.5233,
+      "step": 18780
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 5.833727836608887,
+      "learning_rate": 1.2926783361565971e-05,
+      "loss": 0.5184,
+      "step": 18790
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 11.050354957580566,
+      "learning_rate": 1.2923019009975533e-05,
+      "loss": 0.6627,
+      "step": 18800
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 17.933313369750977,
+      "learning_rate": 1.2919254658385094e-05,
+      "loss": 0.6955,
+      "step": 18810
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 27.6488037109375,
+      "learning_rate": 1.2915490306794655e-05,
+      "loss": 0.7085,
+      "step": 18820
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 19.928560256958008,
+      "learning_rate": 1.2911725955204217e-05,
+      "loss": 0.4012,
+      "step": 18830
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 12.084012985229492,
+      "learning_rate": 1.2907961603613778e-05,
+      "loss": 0.679,
+      "step": 18840
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 36.998104095458984,
+      "learning_rate": 1.290419725202334e-05,
+      "loss": 0.8037,
+      "step": 18850
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 15.250351905822754,
+      "learning_rate": 1.2900432900432901e-05,
+      "loss": 0.5382,
+      "step": 18860
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 30.874340057373047,
+      "learning_rate": 1.2896668548842464e-05,
+      "loss": 0.4945,
+      "step": 18870
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 22.66957664489746,
+      "learning_rate": 1.2892904197252026e-05,
+      "loss": 0.5362,
+      "step": 18880
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 29.47343635559082,
+      "learning_rate": 1.2889139845661587e-05,
+      "loss": 0.6591,
+      "step": 18890
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 11.1913423538208,
+      "learning_rate": 1.2885375494071149e-05,
+      "loss": 0.6665,
+      "step": 18900
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 7.629205703735352,
+      "learning_rate": 1.2881611142480709e-05,
+      "loss": 0.879,
+      "step": 18910
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 9.654862403869629,
+      "learning_rate": 1.287784679089027e-05,
+      "loss": 0.5893,
+      "step": 18920
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.7786961793899536,
+      "learning_rate": 1.2874082439299832e-05,
+      "loss": 0.7301,
+      "step": 18930
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 18.76505470275879,
+      "learning_rate": 1.2870318087709393e-05,
+      "loss": 0.8441,
+      "step": 18940
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 34.90388870239258,
+      "learning_rate": 1.2866553736118954e-05,
+      "loss": 0.7965,
+      "step": 18950
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 14.218279838562012,
+      "learning_rate": 1.2862789384528516e-05,
+      "loss": 0.7871,
+      "step": 18960
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 12.61451244354248,
+      "learning_rate": 1.2859025032938077e-05,
+      "loss": 0.3458,
+      "step": 18970
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 7.381354331970215,
+      "learning_rate": 1.2855260681347639e-05,
+      "loss": 0.3608,
+      "step": 18980
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 5.548383712768555,
+      "learning_rate": 1.28514963297572e-05,
+      "loss": 0.3053,
+      "step": 18990
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 24.012208938598633,
+      "learning_rate": 1.2847731978166762e-05,
+      "loss": 0.73,
+      "step": 19000
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 8.774457931518555,
+      "learning_rate": 1.2843967626576323e-05,
+      "loss": 0.6298,
+      "step": 19010
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 30.910158157348633,
+      "learning_rate": 1.2840203274985885e-05,
+      "loss": 0.7751,
+      "step": 19020
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 34.78523635864258,
+      "learning_rate": 1.2836438923395446e-05,
+      "loss": 0.7953,
+      "step": 19030
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 19.161802291870117,
+      "learning_rate": 1.2832674571805008e-05,
+      "loss": 0.6131,
+      "step": 19040
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.4744579792022705,
+      "learning_rate": 1.282891022021457e-05,
+      "loss": 0.5262,
+      "step": 19050
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 12.062616348266602,
+      "learning_rate": 1.2825145868624132e-05,
+      "loss": 0.8151,
+      "step": 19060
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 11.607182502746582,
+      "learning_rate": 1.2821381517033694e-05,
+      "loss": 0.5112,
+      "step": 19070
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 1.1499428749084473,
+      "learning_rate": 1.2817617165443252e-05,
+      "loss": 0.9765,
+      "step": 19080
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 4.771254062652588,
+      "learning_rate": 1.2813852813852813e-05,
+      "loss": 0.5044,
+      "step": 19090
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 11.462512016296387,
+      "learning_rate": 1.2810088462262376e-05,
+      "loss": 0.7876,
+      "step": 19100
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 29.608722686767578,
+      "learning_rate": 1.2806324110671938e-05,
+      "loss": 0.9717,
+      "step": 19110
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 18.42778968811035,
+      "learning_rate": 1.28025597590815e-05,
+      "loss": 0.4388,
+      "step": 19120
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 8.988313674926758,
+      "learning_rate": 1.279879540749106e-05,
+      "loss": 0.5285,
+      "step": 19130
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 17.52735710144043,
+      "learning_rate": 1.2795031055900622e-05,
+      "loss": 0.5965,
+      "step": 19140
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 24.257164001464844,
+      "learning_rate": 1.2791266704310184e-05,
+      "loss": 0.8865,
+      "step": 19150
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 5.7428460121154785,
+      "learning_rate": 1.2787502352719745e-05,
+      "loss": 0.5267,
+      "step": 19160
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 8.993688583374023,
+      "learning_rate": 1.2783738001129307e-05,
+      "loss": 0.8054,
+      "step": 19170
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 6.0230631828308105,
+      "learning_rate": 1.2779973649538868e-05,
+      "loss": 0.4008,
+      "step": 19180
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 7.1093220710754395,
+      "learning_rate": 1.277620929794843e-05,
+      "loss": 0.5139,
+      "step": 19190
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 6.246382713317871,
+      "learning_rate": 1.2772444946357991e-05,
+      "loss": 0.7167,
+      "step": 19200
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 9.685704231262207,
+      "learning_rate": 1.2768680594767552e-05,
+      "loss": 0.6331,
+      "step": 19210
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 3.0230801105499268,
+      "learning_rate": 1.2764916243177114e-05,
+      "loss": 0.6397,
+      "step": 19220
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 17.838150024414062,
+      "learning_rate": 1.2761151891586675e-05,
+      "loss": 0.5742,
+      "step": 19230
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 21.397781372070312,
+      "learning_rate": 1.2757387539996238e-05,
+      "loss": 0.8132,
+      "step": 19240
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 8.271471977233887,
+      "learning_rate": 1.2753623188405797e-05,
+      "loss": 0.6171,
+      "step": 19250
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 2.5334246158599854,
+      "learning_rate": 1.2749858836815358e-05,
+      "loss": 0.8327,
+      "step": 19260
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 4.004415988922119,
+      "learning_rate": 1.274609448522492e-05,
+      "loss": 0.5323,
+      "step": 19270
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 22.41325569152832,
+      "learning_rate": 1.2742330133634483e-05,
+      "loss": 0.5696,
+      "step": 19280
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 10.288199424743652,
+      "learning_rate": 1.2738565782044044e-05,
+      "loss": 0.6255,
+      "step": 19290
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 7.55498743057251,
+      "learning_rate": 1.2734801430453606e-05,
+      "loss": 0.6258,
+      "step": 19300
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 5.852587699890137,
+      "learning_rate": 1.2731037078863167e-05,
+      "loss": 0.6389,
+      "step": 19310
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 27.851564407348633,
+      "learning_rate": 1.2727272727272728e-05,
+      "loss": 0.8596,
+      "step": 19320
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 9.760847091674805,
+      "learning_rate": 1.272350837568229e-05,
+      "loss": 0.5892,
+      "step": 19330
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 14.357726097106934,
+      "learning_rate": 1.2719744024091851e-05,
+      "loss": 0.4372,
+      "step": 19340
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 2.770073175430298,
+      "learning_rate": 1.2715979672501413e-05,
+      "loss": 0.6819,
+      "step": 19350
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 32.54544448852539,
+      "learning_rate": 1.2712215320910974e-05,
+      "loss": 0.8689,
+      "step": 19360
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 28.392282485961914,
+      "learning_rate": 1.2708450969320536e-05,
+      "loss": 0.6886,
+      "step": 19370
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 29.35589599609375,
+      "learning_rate": 1.2704686617730097e-05,
+      "loss": 0.5073,
+      "step": 19380
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 4.598623752593994,
+      "learning_rate": 1.2700922266139659e-05,
+      "loss": 0.3873,
+      "step": 19390
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 31.010652542114258,
+      "learning_rate": 1.269715791454922e-05,
+      "loss": 0.6336,
+      "step": 19400
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 17.593690872192383,
+      "learning_rate": 1.2693393562958782e-05,
+      "loss": 0.7767,
+      "step": 19410
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 6.044388771057129,
+      "learning_rate": 1.2689629211368343e-05,
+      "loss": 0.683,
+      "step": 19420
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 9.18665599822998,
+      "learning_rate": 1.2685864859777903e-05,
+      "loss": 0.9063,
+      "step": 19430
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 12.083257675170898,
+      "learning_rate": 1.2682100508187464e-05,
+      "loss": 1.0481,
+      "step": 19440
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 18.31049919128418,
+      "learning_rate": 1.2678336156597026e-05,
+      "loss": 0.8629,
+      "step": 19450
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 5.21190881729126,
+      "learning_rate": 1.2674571805006587e-05,
+      "loss": 0.5297,
+      "step": 19460
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 14.538198471069336,
+      "learning_rate": 1.267080745341615e-05,
+      "loss": 0.6776,
+      "step": 19470
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 3.3181121349334717,
+      "learning_rate": 1.2667043101825712e-05,
+      "loss": 0.6773,
+      "step": 19480
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 1.1416178941726685,
+      "learning_rate": 1.2663278750235273e-05,
+      "loss": 0.4913,
+      "step": 19490
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 4.539183139801025,
+      "learning_rate": 1.2659514398644835e-05,
+      "loss": 0.8529,
+      "step": 19500
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 11.008049011230469,
+      "learning_rate": 1.2655750047054396e-05,
+      "loss": 0.5791,
+      "step": 19510
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 26.218887329101562,
+      "learning_rate": 1.2651985695463958e-05,
+      "loss": 0.6222,
+      "step": 19520
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 9.582514762878418,
+      "learning_rate": 1.2648221343873519e-05,
+      "loss": 0.5567,
+      "step": 19530
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 33.36174011230469,
+      "learning_rate": 1.264445699228308e-05,
+      "loss": 0.5457,
+      "step": 19540
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 5.197388648986816,
+      "learning_rate": 1.2640692640692642e-05,
+      "loss": 1.0242,
+      "step": 19550
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 29.88850212097168,
+      "learning_rate": 1.2636928289102204e-05,
+      "loss": 0.6045,
+      "step": 19560
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 14.333205223083496,
+      "learning_rate": 1.2633163937511765e-05,
+      "loss": 0.8825,
+      "step": 19570
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 16.74696159362793,
+      "learning_rate": 1.2629399585921326e-05,
+      "loss": 0.5199,
+      "step": 19580
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 10.93972110748291,
+      "learning_rate": 1.2625635234330888e-05,
+      "loss": 0.5204,
+      "step": 19590
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 26.37276268005371,
+      "learning_rate": 1.2621870882740448e-05,
+      "loss": 0.6918,
+      "step": 19600
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 30.152820587158203,
+      "learning_rate": 1.2618106531150009e-05,
+      "loss": 0.6616,
+      "step": 19610
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 4.047630310058594,
+      "learning_rate": 1.261434217955957e-05,
+      "loss": 0.3501,
+      "step": 19620
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 23.427507400512695,
+      "learning_rate": 1.2610577827969132e-05,
+      "loss": 0.7386,
+      "step": 19630
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 17.632530212402344,
+      "learning_rate": 1.2606813476378694e-05,
+      "loss": 0.3864,
+      "step": 19640
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 13.861688613891602,
+      "learning_rate": 1.2603049124788255e-05,
+      "loss": 0.4736,
+      "step": 19650
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.12637591361999512,
+      "learning_rate": 1.2599284773197818e-05,
+      "loss": 0.459,
+      "step": 19660
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 12.556920051574707,
+      "learning_rate": 1.259552042160738e-05,
+      "loss": 0.5324,
+      "step": 19670
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 15.041160583496094,
+      "learning_rate": 1.2591756070016941e-05,
+      "loss": 0.7857,
+      "step": 19680
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 8.056427955627441,
+      "learning_rate": 1.2587991718426503e-05,
+      "loss": 0.5911,
+      "step": 19690
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 6.120177268981934,
+      "learning_rate": 1.2584227366836064e-05,
+      "loss": 0.4515,
+      "step": 19700
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 5.58326530456543,
+      "learning_rate": 1.2580463015245625e-05,
+      "loss": 0.8182,
+      "step": 19710
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 13.465167045593262,
+      "learning_rate": 1.2576698663655187e-05,
+      "loss": 0.48,
+      "step": 19720
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 2.778507709503174,
+      "learning_rate": 1.2572934312064748e-05,
+      "loss": 0.5068,
+      "step": 19730
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 7.899831771850586,
+      "learning_rate": 1.256916996047431e-05,
+      "loss": 0.5613,
+      "step": 19740
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 8.313310623168945,
+      "learning_rate": 1.2565405608883871e-05,
+      "loss": 0.583,
+      "step": 19750
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 9.49411678314209,
+      "learning_rate": 1.2561641257293433e-05,
+      "loss": 1.0025,
+      "step": 19760
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 24.08152198791504,
+      "learning_rate": 1.2557876905702994e-05,
+      "loss": 0.7991,
+      "step": 19770
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 6.793327808380127,
+      "learning_rate": 1.2554112554112554e-05,
+      "loss": 0.8257,
+      "step": 19780
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 17.370943069458008,
+      "learning_rate": 1.2550348202522115e-05,
+      "loss": 0.8301,
+      "step": 19790
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 9.847330093383789,
+      "learning_rate": 1.2546583850931677e-05,
+      "loss": 0.6801,
+      "step": 19800
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 7.884547710418701,
+      "learning_rate": 1.2542819499341238e-05,
+      "loss": 0.4344,
+      "step": 19810
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.6122704744338989,
+      "learning_rate": 1.25390551477508e-05,
+      "loss": 0.649,
+      "step": 19820
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 12.32701301574707,
+      "learning_rate": 1.2535290796160361e-05,
+      "loss": 0.4236,
+      "step": 19830
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 11.68048095703125,
+      "learning_rate": 1.2531526444569924e-05,
+      "loss": 0.6277,
+      "step": 19840
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 27.56481170654297,
+      "learning_rate": 1.2527762092979486e-05,
+      "loss": 0.6161,
+      "step": 19850
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 10.342185020446777,
+      "learning_rate": 1.2523997741389047e-05,
+      "loss": 0.7714,
+      "step": 19860
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 15.999653816223145,
+      "learning_rate": 1.2520233389798609e-05,
+      "loss": 0.4018,
+      "step": 19870
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 14.883124351501465,
+      "learning_rate": 1.251646903820817e-05,
+      "loss": 0.9517,
+      "step": 19880
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 4.749094009399414,
+      "learning_rate": 1.2512704686617732e-05,
+      "loss": 0.4089,
+      "step": 19890
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 23.341344833374023,
+      "learning_rate": 1.2508940335027293e-05,
+      "loss": 0.5633,
+      "step": 19900
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 15.799015045166016,
+      "learning_rate": 1.2505175983436855e-05,
+      "loss": 0.662,
+      "step": 19910
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 25.26270294189453,
+      "learning_rate": 1.2501411631846416e-05,
+      "loss": 0.6803,
+      "step": 19920
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 19.431108474731445,
+      "learning_rate": 1.2497647280255978e-05,
+      "loss": 0.6431,
+      "step": 19930
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 6.340080261230469,
+      "learning_rate": 1.2493882928665539e-05,
+      "loss": 0.6813,
+      "step": 19940
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 10.563408851623535,
+      "learning_rate": 1.2490118577075099e-05,
+      "loss": 0.5349,
+      "step": 19950
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 18.89433479309082,
+      "learning_rate": 1.248635422548466e-05,
+      "loss": 0.4407,
+      "step": 19960
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 5.784712791442871,
+      "learning_rate": 1.2482589873894222e-05,
+      "loss": 0.7272,
+      "step": 19970
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 24.42642593383789,
+      "learning_rate": 1.2478825522303783e-05,
+      "loss": 0.6009,
+      "step": 19980
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 21.03069496154785,
+      "learning_rate": 1.2475061170713345e-05,
+      "loss": 0.652,
+      "step": 19990
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.11219408363103867,
+      "learning_rate": 1.2471296819122906e-05,
+      "loss": 0.5116,
+      "step": 20000
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 10.447319030761719,
+      "learning_rate": 1.2467532467532468e-05,
+      "loss": 0.3867,
+      "step": 20010
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 8.39681625366211,
+      "learning_rate": 1.2463768115942029e-05,
+      "loss": 0.7164,
+      "step": 20020
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 13.545207977294922,
+      "learning_rate": 1.2460003764351592e-05,
+      "loss": 0.7469,
+      "step": 20030
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 8.322916030883789,
+      "learning_rate": 1.2456239412761154e-05,
+      "loss": 0.6016,
+      "step": 20040
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 10.224037170410156,
+      "learning_rate": 1.2452475061170715e-05,
+      "loss": 0.6915,
+      "step": 20050
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 10.355185508728027,
+      "learning_rate": 1.2448710709580277e-05,
+      "loss": 0.5082,
+      "step": 20060
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 11.303367614746094,
+      "learning_rate": 1.2444946357989838e-05,
+      "loss": 0.574,
+      "step": 20070
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 15.811086654663086,
+      "learning_rate": 1.24411820063994e-05,
+      "loss": 0.8375,
+      "step": 20080
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 19.942611694335938,
+      "learning_rate": 1.2437417654808961e-05,
+      "loss": 0.6268,
+      "step": 20090
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 8.685022354125977,
+      "learning_rate": 1.2433653303218522e-05,
+      "loss": 0.5048,
+      "step": 20100
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 30.918336868286133,
+      "learning_rate": 1.2429888951628084e-05,
+      "loss": 0.5821,
+      "step": 20110
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 22.617633819580078,
+      "learning_rate": 1.2426124600037644e-05,
+      "loss": 0.6512,
+      "step": 20120
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 7.3372483253479,
+      "learning_rate": 1.2422360248447205e-05,
+      "loss": 0.958,
+      "step": 20130
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 9.846896171569824,
+      "learning_rate": 1.2418595896856767e-05,
+      "loss": 0.61,
+      "step": 20140
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 27.524892807006836,
+      "learning_rate": 1.2414831545266328e-05,
+      "loss": 0.885,
+      "step": 20150
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 5.012969017028809,
+      "learning_rate": 1.241106719367589e-05,
+      "loss": 0.5705,
+      "step": 20160
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 14.11467456817627,
+      "learning_rate": 1.2407302842085451e-05,
+      "loss": 0.7462,
+      "step": 20170
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 23.304222106933594,
+      "learning_rate": 1.2403538490495012e-05,
+      "loss": 0.8118,
+      "step": 20180
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 5.021602630615234,
+      "learning_rate": 1.2399774138904574e-05,
+      "loss": 0.6483,
+      "step": 20190
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 4.954014301300049,
+      "learning_rate": 1.2396009787314135e-05,
+      "loss": 0.5144,
+      "step": 20200
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 15.003413200378418,
+      "learning_rate": 1.2392245435723698e-05,
+      "loss": 0.6549,
+      "step": 20210
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.8703759908676147,
+      "learning_rate": 1.238848108413326e-05,
+      "loss": 0.477,
+      "step": 20220
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 19.85989761352539,
+      "learning_rate": 1.2384716732542821e-05,
+      "loss": 0.6608,
+      "step": 20230
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 10.10875129699707,
+      "learning_rate": 1.2380952380952383e-05,
+      "loss": 0.7041,
+      "step": 20240
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 6.1762471199035645,
+      "learning_rate": 1.2377188029361944e-05,
+      "loss": 0.9546,
+      "step": 20250
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 44.439823150634766,
+      "learning_rate": 1.2373423677771506e-05,
+      "loss": 1.0269,
+      "step": 20260
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 17.293577194213867,
+      "learning_rate": 1.2369659326181067e-05,
+      "loss": 0.6694,
+      "step": 20270
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 5.511439800262451,
+      "learning_rate": 1.2365894974590629e-05,
+      "loss": 0.502,
+      "step": 20280
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 18.506452560424805,
+      "learning_rate": 1.236213062300019e-05,
+      "loss": 1.0257,
+      "step": 20290
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 10.748970031738281,
+      "learning_rate": 1.235836627140975e-05,
+      "loss": 0.7194,
+      "step": 20300
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 39.78062057495117,
+      "learning_rate": 1.2354601919819311e-05,
+      "loss": 0.5527,
+      "step": 20310
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 16.632009506225586,
+      "learning_rate": 1.2350837568228873e-05,
+      "loss": 0.7186,
+      "step": 20320
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 3.820341110229492,
+      "learning_rate": 1.2347073216638434e-05,
+      "loss": 0.6947,
+      "step": 20330
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 1.6036685705184937,
+      "learning_rate": 1.2343308865047996e-05,
+      "loss": 0.4921,
+      "step": 20340
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 1.2304280996322632,
+      "learning_rate": 1.2339544513457557e-05,
+      "loss": 0.6795,
+      "step": 20350
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 18.970306396484375,
+      "learning_rate": 1.2335780161867119e-05,
+      "loss": 0.6016,
+      "step": 20360
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 8.561296463012695,
+      "learning_rate": 1.233201581027668e-05,
+      "loss": 0.9288,
+      "step": 20370
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 11.71761417388916,
+      "learning_rate": 1.2328251458686242e-05,
+      "loss": 0.7948,
+      "step": 20380
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 38.39409637451172,
+      "learning_rate": 1.2324487107095803e-05,
+      "loss": 0.6868,
+      "step": 20390
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 19.950315475463867,
+      "learning_rate": 1.2320722755505366e-05,
+      "loss": 0.4955,
+      "step": 20400
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 6.911773204803467,
+      "learning_rate": 1.2316958403914928e-05,
+      "loss": 0.3525,
+      "step": 20410
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 13.918087005615234,
+      "learning_rate": 1.2313194052324489e-05,
+      "loss": 0.3229,
+      "step": 20420
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 14.926342964172363,
+      "learning_rate": 1.230942970073405e-05,
+      "loss": 0.6098,
+      "step": 20430
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 16.889162063598633,
+      "learning_rate": 1.2305665349143612e-05,
+      "loss": 0.6025,
+      "step": 20440
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 17.502153396606445,
+      "learning_rate": 1.2301900997553174e-05,
+      "loss": 0.4296,
+      "step": 20450
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 17.864023208618164,
+      "learning_rate": 1.2298136645962735e-05,
+      "loss": 0.9841,
+      "step": 20460
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 8.081547737121582,
+      "learning_rate": 1.2294372294372295e-05,
+      "loss": 0.5448,
+      "step": 20470
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 10.248896598815918,
+      "learning_rate": 1.2290607942781856e-05,
+      "loss": 0.7463,
+      "step": 20480
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 43.104915618896484,
+      "learning_rate": 1.2286843591191418e-05,
+      "loss": 0.868,
+      "step": 20490
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 23.549795150756836,
+      "learning_rate": 1.2283079239600979e-05,
+      "loss": 0.551,
+      "step": 20500
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 18.552968978881836,
+      "learning_rate": 1.227931488801054e-05,
+      "loss": 0.5959,
+      "step": 20510
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 15.869778633117676,
+      "learning_rate": 1.2275550536420102e-05,
+      "loss": 0.7179,
+      "step": 20520
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 14.710973739624023,
+      "learning_rate": 1.2271786184829664e-05,
+      "loss": 0.6882,
+      "step": 20530
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 8.942631721496582,
+      "learning_rate": 1.2268021833239225e-05,
+      "loss": 0.8396,
+      "step": 20540
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 6.654508113861084,
+      "learning_rate": 1.2264257481648786e-05,
+      "loss": 0.3994,
+      "step": 20550
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 47.291954040527344,
+      "learning_rate": 1.2260493130058348e-05,
+      "loss": 0.7838,
+      "step": 20560
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 13.667283058166504,
+      "learning_rate": 1.225672877846791e-05,
+      "loss": 0.4754,
+      "step": 20570
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 41.05614471435547,
+      "learning_rate": 1.2252964426877473e-05,
+      "loss": 0.643,
+      "step": 20580
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 16.537818908691406,
+      "learning_rate": 1.2249200075287034e-05,
+      "loss": 0.7026,
+      "step": 20590
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 16.900470733642578,
+      "learning_rate": 1.2245435723696595e-05,
+      "loss": 0.4249,
+      "step": 20600
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 11.43319034576416,
+      "learning_rate": 1.2241671372106157e-05,
+      "loss": 0.8047,
+      "step": 20610
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 1.8628902435302734,
+      "learning_rate": 1.2237907020515718e-05,
+      "loss": 0.5949,
+      "step": 20620
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 2.925947427749634,
+      "learning_rate": 1.223414266892528e-05,
+      "loss": 0.6901,
+      "step": 20630
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 10.759583473205566,
+      "learning_rate": 1.2230378317334841e-05,
+      "loss": 0.4635,
+      "step": 20640
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 13.309684753417969,
+      "learning_rate": 1.2226613965744401e-05,
+      "loss": 0.5326,
+      "step": 20650
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 23.810779571533203,
+      "learning_rate": 1.2222849614153963e-05,
+      "loss": 0.603,
+      "step": 20660
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 14.398448944091797,
+      "learning_rate": 1.2219085262563524e-05,
+      "loss": 0.9449,
+      "step": 20670
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 10.191229820251465,
+      "learning_rate": 1.2215320910973085e-05,
+      "loss": 0.4746,
+      "step": 20680
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 18.595684051513672,
+      "learning_rate": 1.2211556559382647e-05,
+      "loss": 1.0004,
+      "step": 20690
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 32.1865119934082,
+      "learning_rate": 1.2207792207792208e-05,
+      "loss": 0.4889,
+      "step": 20700
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.28670892119407654,
+      "learning_rate": 1.220402785620177e-05,
+      "loss": 0.6854,
+      "step": 20710
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 5.30502986907959,
+      "learning_rate": 1.2200263504611331e-05,
+      "loss": 0.5467,
+      "step": 20720
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 12.79511547088623,
+      "learning_rate": 1.2196499153020893e-05,
+      "loss": 0.4574,
+      "step": 20730
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 17.650203704833984,
+      "learning_rate": 1.2192734801430454e-05,
+      "loss": 0.5639,
+      "step": 20740
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.7680536508560181,
+      "learning_rate": 1.2188970449840016e-05,
+      "loss": 0.6494,
+      "step": 20750
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.08082833886146545,
+      "learning_rate": 1.2185206098249577e-05,
+      "loss": 0.4382,
+      "step": 20760
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 14.408023834228516,
+      "learning_rate": 1.218144174665914e-05,
+      "loss": 0.8581,
+      "step": 20770
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 10.851282119750977,
+      "learning_rate": 1.2177677395068702e-05,
+      "loss": 0.7481,
+      "step": 20780
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 29.99077796936035,
+      "learning_rate": 1.2173913043478263e-05,
+      "loss": 1.0,
+      "step": 20790
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.19973182678222656,
+      "learning_rate": 1.2170148691887825e-05,
+      "loss": 0.6387,
+      "step": 20800
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 7.756618499755859,
+      "learning_rate": 1.2166384340297386e-05,
+      "loss": 0.5351,
+      "step": 20810
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 20.255647659301758,
+      "learning_rate": 1.2162619988706946e-05,
+      "loss": 0.7969,
+      "step": 20820
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.8674776554107666,
+      "learning_rate": 1.2158855637116507e-05,
+      "loss": 0.7177,
+      "step": 20830
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 8.670698165893555,
+      "learning_rate": 1.2155091285526069e-05,
+      "loss": 0.624,
+      "step": 20840
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.19249022006988525,
+      "learning_rate": 1.215132693393563e-05,
+      "loss": 0.4953,
+      "step": 20850
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 26.02955436706543,
+      "learning_rate": 1.2147562582345192e-05,
+      "loss": 0.5161,
+      "step": 20860
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 29.887535095214844,
+      "learning_rate": 1.2143798230754753e-05,
+      "loss": 0.5798,
+      "step": 20870
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 17.09107208251953,
+      "learning_rate": 1.2140033879164315e-05,
+      "loss": 0.5887,
+      "step": 20880
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 21.360456466674805,
+      "learning_rate": 1.2136269527573876e-05,
+      "loss": 0.76,
+      "step": 20890
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 8.325711250305176,
+      "learning_rate": 1.2132505175983438e-05,
+      "loss": 0.6259,
+      "step": 20900
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 8.927617073059082,
+      "learning_rate": 1.2128740824392999e-05,
+      "loss": 0.5466,
+      "step": 20910
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 9.248173713684082,
+      "learning_rate": 1.212497647280256e-05,
+      "loss": 0.4414,
+      "step": 20920
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 9.122405052185059,
+      "learning_rate": 1.2121212121212122e-05,
+      "loss": 0.5872,
+      "step": 20930
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 8.157663345336914,
+      "learning_rate": 1.2117447769621683e-05,
+      "loss": 0.577,
+      "step": 20940
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 19.186445236206055,
+      "learning_rate": 1.2113683418031245e-05,
+      "loss": 0.7963,
+      "step": 20950
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 8.585948944091797,
+      "learning_rate": 1.2109919066440808e-05,
+      "loss": 0.6799,
+      "step": 20960
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 11.789979934692383,
+      "learning_rate": 1.210615471485037e-05,
+      "loss": 0.5179,
+      "step": 20970
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 5.875610828399658,
+      "learning_rate": 1.2102390363259931e-05,
+      "loss": 0.437,
+      "step": 20980
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 18.89061164855957,
+      "learning_rate": 1.2098626011669492e-05,
+      "loss": 0.577,
+      "step": 20990
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 4.297598838806152,
+      "learning_rate": 1.2094861660079052e-05,
+      "loss": 0.7225,
+      "step": 21000
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.36552512645721436,
+      "learning_rate": 1.2091097308488614e-05,
+      "loss": 0.2754,
+      "step": 21010
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 13.013652801513672,
+      "learning_rate": 1.2087332956898175e-05,
+      "loss": 0.669,
+      "step": 21020
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 6.945061683654785,
+      "learning_rate": 1.2083568605307737e-05,
+      "loss": 0.5892,
+      "step": 21030
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 4.39546012878418,
+      "learning_rate": 1.2079804253717298e-05,
+      "loss": 0.6006,
+      "step": 21040
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 24.449954986572266,
+      "learning_rate": 1.207603990212686e-05,
+      "loss": 0.503,
+      "step": 21050
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 9.902664184570312,
+      "learning_rate": 1.2072275550536421e-05,
+      "loss": 0.6385,
+      "step": 21060
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 44.686317443847656,
+      "learning_rate": 1.2068511198945982e-05,
+      "loss": 0.4779,
+      "step": 21070
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 13.766404151916504,
+      "learning_rate": 1.2064746847355544e-05,
+      "loss": 0.6803,
+      "step": 21080
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 41.97696304321289,
+      "learning_rate": 1.2060982495765105e-05,
+      "loss": 1.1873,
+      "step": 21090
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 15.863570213317871,
+      "learning_rate": 1.2057218144174667e-05,
+      "loss": 0.5433,
+      "step": 21100
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 10.411752700805664,
+      "learning_rate": 1.2053453792584228e-05,
+      "loss": 0.5845,
+      "step": 21110
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 4.70853853225708,
+      "learning_rate": 1.204968944099379e-05,
+      "loss": 0.9748,
+      "step": 21120
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 41.68354034423828,
+      "learning_rate": 1.2045925089403351e-05,
+      "loss": 0.7451,
+      "step": 21130
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 12.279452323913574,
+      "learning_rate": 1.2042160737812914e-05,
+      "loss": 0.6239,
+      "step": 21140
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 1.5319571495056152,
+      "learning_rate": 1.2038396386222476e-05,
+      "loss": 0.6755,
+      "step": 21150
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 19.07404899597168,
+      "learning_rate": 1.2034632034632037e-05,
+      "loss": 0.9424,
+      "step": 21160
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 23.491230010986328,
+      "learning_rate": 1.2030867683041595e-05,
+      "loss": 0.7338,
+      "step": 21170
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 30.06789207458496,
+      "learning_rate": 1.2027103331451157e-05,
+      "loss": 0.705,
+      "step": 21180
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 7.1863789558410645,
+      "learning_rate": 1.202333897986072e-05,
+      "loss": 0.964,
+      "step": 21190
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 19.22167205810547,
+      "learning_rate": 1.2019574628270281e-05,
+      "loss": 0.5426,
+      "step": 21200
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 18.454984664916992,
+      "learning_rate": 1.2015810276679843e-05,
+      "loss": 0.5774,
+      "step": 21210
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 12.360136985778809,
+      "learning_rate": 1.2012045925089404e-05,
+      "loss": 0.9418,
+      "step": 21220
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 11.939126968383789,
+      "learning_rate": 1.2008281573498966e-05,
+      "loss": 0.7091,
+      "step": 21230
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 32.17779541015625,
+      "learning_rate": 1.2004517221908527e-05,
+      "loss": 0.7587,
+      "step": 21240
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 7.594665050506592,
+      "learning_rate": 1.2000752870318089e-05,
+      "loss": 0.4173,
+      "step": 21250
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.9169333333333334,
+      "eval_loss": 0.3204882740974426,
+      "eval_runtime": 51.2733,
+      "eval_samples_per_second": 146.275,
+      "eval_steps_per_second": 18.294,
+      "step": 21252
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 20.44147491455078,
+      "learning_rate": 1.199698851872765e-05,
+      "loss": 0.7655,
+      "step": 21260
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 5.429800987243652,
+      "learning_rate": 1.1993224167137212e-05,
+      "loss": 0.507,
+      "step": 21270
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 2.7339699268341064,
+      "learning_rate": 1.1989459815546773e-05,
+      "loss": 0.5326,
+      "step": 21280
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 10.058930397033691,
+      "learning_rate": 1.1985695463956335e-05,
+      "loss": 0.923,
+      "step": 21290
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 17.565692901611328,
+      "learning_rate": 1.1981931112365896e-05,
+      "loss": 0.5113,
+      "step": 21300
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 6.418165683746338,
+      "learning_rate": 1.1978166760775457e-05,
+      "loss": 0.7591,
+      "step": 21310
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 8.731622695922852,
+      "learning_rate": 1.1974402409185019e-05,
+      "loss": 0.4488,
+      "step": 21320
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 6.82095193862915,
+      "learning_rate": 1.1970638057594582e-05,
+      "loss": 0.5737,
+      "step": 21330
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 19.75332260131836,
+      "learning_rate": 1.196687370600414e-05,
+      "loss": 0.5355,
+      "step": 21340
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 8.986883163452148,
+      "learning_rate": 1.1963109354413702e-05,
+      "loss": 0.749,
+      "step": 21350
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 5.045528411865234,
+      "learning_rate": 1.1959345002823263e-05,
+      "loss": 0.7225,
+      "step": 21360
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 24.6765079498291,
+      "learning_rate": 1.1955580651232826e-05,
+      "loss": 0.5938,
+      "step": 21370
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 14.547659873962402,
+      "learning_rate": 1.1951816299642388e-05,
+      "loss": 0.554,
+      "step": 21380
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 21.172176361083984,
+      "learning_rate": 1.1948051948051949e-05,
+      "loss": 0.3325,
+      "step": 21390
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 35.6104736328125,
+      "learning_rate": 1.194428759646151e-05,
+      "loss": 0.8367,
+      "step": 21400
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 4.693765163421631,
+      "learning_rate": 1.1940523244871072e-05,
+      "loss": 0.6465,
+      "step": 21410
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 12.738019943237305,
+      "learning_rate": 1.1936758893280634e-05,
+      "loss": 0.5504,
+      "step": 21420
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 26.901885986328125,
+      "learning_rate": 1.1932994541690195e-05,
+      "loss": 0.4753,
+      "step": 21430
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 21.4942569732666,
+      "learning_rate": 1.1929230190099756e-05,
+      "loss": 0.8649,
+      "step": 21440
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 14.137928009033203,
+      "learning_rate": 1.1925465838509318e-05,
+      "loss": 0.5552,
+      "step": 21450
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 18.820674896240234,
+      "learning_rate": 1.192170148691888e-05,
+      "loss": 0.577,
+      "step": 21460
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 3.413235664367676,
+      "learning_rate": 1.191793713532844e-05,
+      "loss": 0.5026,
+      "step": 21470
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 1.3790395259857178,
+      "learning_rate": 1.1914172783738002e-05,
+      "loss": 0.4477,
+      "step": 21480
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 16.56000328063965,
+      "learning_rate": 1.1910408432147564e-05,
+      "loss": 0.6547,
+      "step": 21490
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 6.222777843475342,
+      "learning_rate": 1.1906644080557125e-05,
+      "loss": 0.6857,
+      "step": 21500
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 9.765985488891602,
+      "learning_rate": 1.1902879728966688e-05,
+      "loss": 0.6043,
+      "step": 21510
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 21.691728591918945,
+      "learning_rate": 1.1899115377376246e-05,
+      "loss": 0.3465,
+      "step": 21520
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 0.24102507531642914,
+      "learning_rate": 1.1895351025785808e-05,
+      "loss": 0.4715,
+      "step": 21530
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 6.748645305633545,
+      "learning_rate": 1.189158667419537e-05,
+      "loss": 0.4595,
+      "step": 21540
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 6.551207065582275,
+      "learning_rate": 1.188782232260493e-05,
+      "loss": 0.8763,
+      "step": 21550
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 9.870513916015625,
+      "learning_rate": 1.1884057971014494e-05,
+      "loss": 0.4955,
+      "step": 21560
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 15.267050743103027,
+      "learning_rate": 1.1880293619424055e-05,
+      "loss": 0.5343,
+      "step": 21570
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 16.510562896728516,
+      "learning_rate": 1.1876529267833617e-05,
+      "loss": 0.7292,
+      "step": 21580
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 5.880744457244873,
+      "learning_rate": 1.1872764916243178e-05,
+      "loss": 0.8015,
+      "step": 21590
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 12.049949645996094,
+      "learning_rate": 1.186900056465274e-05,
+      "loss": 0.5739,
+      "step": 21600
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 13.833176612854004,
+      "learning_rate": 1.1865236213062301e-05,
+      "loss": 0.9261,
+      "step": 21610
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 21.508180618286133,
+      "learning_rate": 1.1861471861471863e-05,
+      "loss": 0.6697,
+      "step": 21620
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 0.2058361917734146,
+      "learning_rate": 1.1857707509881424e-05,
+      "loss": 0.7873,
+      "step": 21630
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 31.943452835083008,
+      "learning_rate": 1.1853943158290986e-05,
+      "loss": 0.5298,
+      "step": 21640
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 11.333353042602539,
+      "learning_rate": 1.1850178806700547e-05,
+      "loss": 0.3821,
+      "step": 21650
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 8.284579277038574,
+      "learning_rate": 1.1846414455110109e-05,
+      "loss": 0.4522,
+      "step": 21660
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 11.12885570526123,
+      "learning_rate": 1.184265010351967e-05,
+      "loss": 0.9707,
+      "step": 21670
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 5.9140191078186035,
+      "learning_rate": 1.1838885751929231e-05,
+      "loss": 0.5,
+      "step": 21680
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 6.177058696746826,
+      "learning_rate": 1.1835121400338791e-05,
+      "loss": 0.6024,
+      "step": 21690
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 2.904001235961914,
+      "learning_rate": 1.1831357048748353e-05,
+      "loss": 0.8046,
+      "step": 21700
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 16.1057186126709,
+      "learning_rate": 1.1827592697157914e-05,
+      "loss": 0.5458,
+      "step": 21710
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 19.81020736694336,
+      "learning_rate": 1.1823828345567476e-05,
+      "loss": 0.3633,
+      "step": 21720
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 16.992023468017578,
+      "learning_rate": 1.1820063993977037e-05,
+      "loss": 0.9944,
+      "step": 21730
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 24.561140060424805,
+      "learning_rate": 1.18162996423866e-05,
+      "loss": 0.4752,
+      "step": 21740
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 0.726672887802124,
+      "learning_rate": 1.1812535290796162e-05,
+      "loss": 0.4632,
+      "step": 21750
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 0.690126359462738,
+      "learning_rate": 1.1808770939205723e-05,
+      "loss": 0.5868,
+      "step": 21760
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 36.9278564453125,
+      "learning_rate": 1.1805006587615285e-05,
+      "loss": 0.772,
+      "step": 21770
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 16.300630569458008,
+      "learning_rate": 1.1801242236024846e-05,
+      "loss": 0.4303,
+      "step": 21780
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 0.3904343843460083,
+      "learning_rate": 1.1797477884434408e-05,
+      "loss": 0.5117,
+      "step": 21790
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 26.473928451538086,
+      "learning_rate": 1.1793713532843969e-05,
+      "loss": 0.4777,
+      "step": 21800
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 12.805113792419434,
+      "learning_rate": 1.178994918125353e-05,
+      "loss": 0.7928,
+      "step": 21810
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 26.79267692565918,
+      "learning_rate": 1.1786184829663092e-05,
+      "loss": 0.6495,
+      "step": 21820
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 18.316246032714844,
+      "learning_rate": 1.1782420478072653e-05,
+      "loss": 0.7989,
+      "step": 21830
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 15.82396411895752,
+      "learning_rate": 1.1778656126482215e-05,
+      "loss": 0.6794,
+      "step": 21840
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 15.773408889770508,
+      "learning_rate": 1.1774891774891776e-05,
+      "loss": 0.7149,
+      "step": 21850
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 9.06588077545166,
+      "learning_rate": 1.1771127423301338e-05,
+      "loss": 0.7622,
+      "step": 21860
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 0.6174959540367126,
+      "learning_rate": 1.1767363071710898e-05,
+      "loss": 0.7434,
+      "step": 21870
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 4.278200149536133,
+      "learning_rate": 1.1763598720120459e-05,
+      "loss": 0.6226,
+      "step": 21880
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 16.598983764648438,
+      "learning_rate": 1.175983436853002e-05,
+      "loss": 0.5363,
+      "step": 21890
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 25.885684967041016,
+      "learning_rate": 1.1756070016939582e-05,
+      "loss": 0.5998,
+      "step": 21900
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 7.993864059448242,
+      "learning_rate": 1.1752305665349143e-05,
+      "loss": 0.3801,
+      "step": 21910
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 0.9169643521308899,
+      "learning_rate": 1.1748541313758705e-05,
+      "loss": 0.6745,
+      "step": 21920
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 48.847991943359375,
+      "learning_rate": 1.1744776962168268e-05,
+      "loss": 0.5772,
+      "step": 21930
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 34.63701629638672,
+      "learning_rate": 1.174101261057783e-05,
+      "loss": 0.6175,
+      "step": 21940
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 55.32223129272461,
+      "learning_rate": 1.1737248258987391e-05,
+      "loss": 1.0425,
+      "step": 21950
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 8.96568489074707,
+      "learning_rate": 1.1733483907396952e-05,
+      "loss": 0.5169,
+      "step": 21960
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 13.289989471435547,
+      "learning_rate": 1.1729719555806514e-05,
+      "loss": 0.6216,
+      "step": 21970
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 0.6537827253341675,
+      "learning_rate": 1.1725955204216075e-05,
+      "loss": 0.7422,
+      "step": 21980
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 3.4758856296539307,
+      "learning_rate": 1.1722190852625637e-05,
+      "loss": 0.6289,
+      "step": 21990
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 3.0647480487823486,
+      "learning_rate": 1.1718426501035198e-05,
+      "loss": 0.598,
+      "step": 22000
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 9.152379989624023,
+      "learning_rate": 1.171466214944476e-05,
+      "loss": 0.4765,
+      "step": 22010
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 8.172050476074219,
+      "learning_rate": 1.1710897797854321e-05,
+      "loss": 0.6093,
+      "step": 22020
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 9.752901077270508,
+      "learning_rate": 1.1707133446263883e-05,
+      "loss": 0.6086,
+      "step": 22030
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 12.649826049804688,
+      "learning_rate": 1.1703369094673442e-05,
+      "loss": 0.7863,
+      "step": 22040
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 11.14371109008789,
+      "learning_rate": 1.1699604743083004e-05,
+      "loss": 0.6756,
+      "step": 22050
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 0.7610763311386108,
+      "learning_rate": 1.1695840391492565e-05,
+      "loss": 0.6398,
+      "step": 22060
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 1.436449408531189,
+      "learning_rate": 1.1692076039902127e-05,
+      "loss": 0.5272,
+      "step": 22070
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 22.43182373046875,
+      "learning_rate": 1.1688311688311688e-05,
+      "loss": 0.5553,
+      "step": 22080
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.680665910243988,
+      "learning_rate": 1.168454733672125e-05,
+      "loss": 0.3977,
+      "step": 22090
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 12.118298530578613,
+      "learning_rate": 1.1680782985130811e-05,
+      "loss": 0.482,
+      "step": 22100
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 19.042518615722656,
+      "learning_rate": 1.1677018633540373e-05,
+      "loss": 0.5045,
+      "step": 22110
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 9.090872764587402,
+      "learning_rate": 1.1673254281949936e-05,
+      "loss": 0.4631,
+      "step": 22120
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 2.18812894821167,
+      "learning_rate": 1.1669489930359497e-05,
+      "loss": 0.6592,
+      "step": 22130
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 36.4178466796875,
+      "learning_rate": 1.1665725578769059e-05,
+      "loss": 0.6483,
+      "step": 22140
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 28.83828353881836,
+      "learning_rate": 1.166196122717862e-05,
+      "loss": 0.4226,
+      "step": 22150
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 31.52997398376465,
+      "learning_rate": 1.1658196875588182e-05,
+      "loss": 0.3637,
+      "step": 22160
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 8.760178565979004,
+      "learning_rate": 1.1654432523997743e-05,
+      "loss": 0.7955,
+      "step": 22170
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 0.6389203071594238,
+      "learning_rate": 1.1650668172407305e-05,
+      "loss": 0.4444,
+      "step": 22180
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 29.18195915222168,
+      "learning_rate": 1.1646903820816866e-05,
+      "loss": 0.7053,
+      "step": 22190
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 22.56359100341797,
+      "learning_rate": 1.1643139469226427e-05,
+      "loss": 0.4031,
+      "step": 22200
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 23.801780700683594,
+      "learning_rate": 1.1639375117635989e-05,
+      "loss": 0.624,
+      "step": 22210
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 13.610224723815918,
+      "learning_rate": 1.1635610766045549e-05,
+      "loss": 0.9822,
+      "step": 22220
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 18.63134765625,
+      "learning_rate": 1.163184641445511e-05,
+      "loss": 0.6779,
+      "step": 22230
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 16.268205642700195,
+      "learning_rate": 1.1628082062864672e-05,
+      "loss": 0.5979,
+      "step": 22240
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 8.651780128479004,
+      "learning_rate": 1.1624317711274233e-05,
+      "loss": 0.7667,
+      "step": 22250
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 42.6129035949707,
+      "learning_rate": 1.1620553359683795e-05,
+      "loss": 0.6944,
+      "step": 22260
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 3.1881678104400635,
+      "learning_rate": 1.1616789008093356e-05,
+      "loss": 0.471,
+      "step": 22270
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 0.7028786540031433,
+      "learning_rate": 1.1613024656502917e-05,
+      "loss": 0.5197,
+      "step": 22280
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 47.69763946533203,
+      "learning_rate": 1.1609260304912479e-05,
+      "loss": 0.6065,
+      "step": 22290
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 6.314722061157227,
+      "learning_rate": 1.1605495953322042e-05,
+      "loss": 0.5556,
+      "step": 22300
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 8.165750503540039,
+      "learning_rate": 1.1601731601731604e-05,
+      "loss": 0.7189,
+      "step": 22310
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 7.349857807159424,
+      "learning_rate": 1.1597967250141165e-05,
+      "loss": 0.5367,
+      "step": 22320
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 10.950774192810059,
+      "learning_rate": 1.1594202898550726e-05,
+      "loss": 0.773,
+      "step": 22330
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 16.20650291442871,
+      "learning_rate": 1.1590438546960288e-05,
+      "loss": 0.6576,
+      "step": 22340
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 20.38640594482422,
+      "learning_rate": 1.158667419536985e-05,
+      "loss": 0.5126,
+      "step": 22350
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 8.257250785827637,
+      "learning_rate": 1.158290984377941e-05,
+      "loss": 0.591,
+      "step": 22360
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 5.747903823852539,
+      "learning_rate": 1.1579145492188972e-05,
+      "loss": 0.6314,
+      "step": 22370
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 19.206281661987305,
+      "learning_rate": 1.1575381140598534e-05,
+      "loss": 0.6481,
+      "step": 22380
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 22.839046478271484,
+      "learning_rate": 1.1571616789008094e-05,
+      "loss": 0.5929,
+      "step": 22390
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 22.537134170532227,
+      "learning_rate": 1.1567852437417655e-05,
+      "loss": 0.4775,
+      "step": 22400
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 0.13030314445495605,
+      "learning_rate": 1.1564088085827216e-05,
+      "loss": 0.4585,
+      "step": 22410
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 11.866349220275879,
+      "learning_rate": 1.1560323734236778e-05,
+      "loss": 0.6309,
+      "step": 22420
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 15.4808349609375,
+      "learning_rate": 1.155655938264634e-05,
+      "loss": 0.7321,
+      "step": 22430
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 14.688508033752441,
+      "learning_rate": 1.15527950310559e-05,
+      "loss": 0.8127,
+      "step": 22440
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 8.376343727111816,
+      "learning_rate": 1.1549030679465462e-05,
+      "loss": 0.6156,
+      "step": 22450
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 22.739774703979492,
+      "learning_rate": 1.1545266327875024e-05,
+      "loss": 0.6053,
+      "step": 22460
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 1.1516119241714478,
+      "learning_rate": 1.1541501976284585e-05,
+      "loss": 0.5797,
+      "step": 22470
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 38.92045974731445,
+      "learning_rate": 1.1537737624694147e-05,
+      "loss": 0.7583,
+      "step": 22480
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 11.236977577209473,
+      "learning_rate": 1.153397327310371e-05,
+      "loss": 0.7248,
+      "step": 22490
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 1.5737559795379639,
+      "learning_rate": 1.1530208921513271e-05,
+      "loss": 0.7463,
+      "step": 22500
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 1.3805655241012573,
+      "learning_rate": 1.1526444569922833e-05,
+      "loss": 0.4989,
+      "step": 22510
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 115.92575073242188,
+      "learning_rate": 1.1522680218332394e-05,
+      "loss": 0.4902,
+      "step": 22520
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 1.0640676021575928,
+      "learning_rate": 1.1518915866741956e-05,
+      "loss": 0.5065,
+      "step": 22530
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 9.37736988067627,
+      "learning_rate": 1.1515151515151517e-05,
+      "loss": 0.493,
+      "step": 22540
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 11.277339935302734,
+      "learning_rate": 1.1511387163561079e-05,
+      "loss": 0.5904,
+      "step": 22550
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 33.85553741455078,
+      "learning_rate": 1.1507622811970638e-05,
+      "loss": 0.5691,
+      "step": 22560
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 16.795196533203125,
+      "learning_rate": 1.15038584603802e-05,
+      "loss": 0.6498,
+      "step": 22570
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 6.320994853973389,
+      "learning_rate": 1.1500094108789761e-05,
+      "loss": 0.5735,
+      "step": 22580
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 0.7785674929618835,
+      "learning_rate": 1.1496329757199323e-05,
+      "loss": 0.5169,
+      "step": 22590
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 1.8020360469818115,
+      "learning_rate": 1.1492565405608884e-05,
+      "loss": 0.3877,
+      "step": 22600
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 28.67214584350586,
+      "learning_rate": 1.1488801054018446e-05,
+      "loss": 0.6102,
+      "step": 22610
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 33.07846450805664,
+      "learning_rate": 1.1485036702428007e-05,
+      "loss": 0.6309,
+      "step": 22620
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 41.755577087402344,
+      "learning_rate": 1.1481272350837569e-05,
+      "loss": 0.9355,
+      "step": 22630
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 7.226475238800049,
+      "learning_rate": 1.147750799924713e-05,
+      "loss": 0.6734,
+      "step": 22640
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 20.88840675354004,
+      "learning_rate": 1.1473743647656691e-05,
+      "loss": 0.4802,
+      "step": 22650
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 18.926301956176758,
+      "learning_rate": 1.1469979296066253e-05,
+      "loss": 0.6868,
+      "step": 22660
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 9.090205192565918,
+      "learning_rate": 1.1466214944475816e-05,
+      "loss": 0.7658,
+      "step": 22670
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 55.62081527709961,
+      "learning_rate": 1.1462450592885378e-05,
+      "loss": 0.6087,
+      "step": 22680
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 20.29065704345703,
+      "learning_rate": 1.1458686241294939e-05,
+      "loss": 0.4497,
+      "step": 22690
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 6.561438083648682,
+      "learning_rate": 1.14549218897045e-05,
+      "loss": 0.5936,
+      "step": 22700
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 12.46176815032959,
+      "learning_rate": 1.1451157538114062e-05,
+      "loss": 0.8636,
+      "step": 22710
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 31.131755828857422,
+      "learning_rate": 1.1447393186523623e-05,
+      "loss": 0.7126,
+      "step": 22720
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 20.14628028869629,
+      "learning_rate": 1.1443628834933185e-05,
+      "loss": 0.4201,
+      "step": 22730
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 8.134221076965332,
+      "learning_rate": 1.1439864483342745e-05,
+      "loss": 0.595,
+      "step": 22740
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 23.64768409729004,
+      "learning_rate": 1.1436100131752306e-05,
+      "loss": 0.8233,
+      "step": 22750
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 3.0115549564361572,
+      "learning_rate": 1.1432335780161868e-05,
+      "loss": 0.3933,
+      "step": 22760
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 0.10152573138475418,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 0.4964,
+      "step": 22770
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 12.97737979888916,
+      "learning_rate": 1.142480707698099e-05,
+      "loss": 0.281,
+      "step": 22780
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 10.339591979980469,
+      "learning_rate": 1.1421042725390552e-05,
+      "loss": 0.7338,
+      "step": 22790
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 21.024656295776367,
+      "learning_rate": 1.1417278373800113e-05,
+      "loss": 0.4871,
+      "step": 22800
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 0.6640139818191528,
+      "learning_rate": 1.1413514022209675e-05,
+      "loss": 0.5691,
+      "step": 22810
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 19.123470306396484,
+      "learning_rate": 1.1409749670619236e-05,
+      "loss": 0.8221,
+      "step": 22820
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 16.33519172668457,
+      "learning_rate": 1.1405985319028798e-05,
+      "loss": 0.6879,
+      "step": 22830
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 19.19475746154785,
+      "learning_rate": 1.140222096743836e-05,
+      "loss": 0.5056,
+      "step": 22840
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 4.6960883140563965,
+      "learning_rate": 1.139845661584792e-05,
+      "loss": 0.6242,
+      "step": 22850
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 0.4895837604999542,
+      "learning_rate": 1.1394692264257484e-05,
+      "loss": 0.6206,
+      "step": 22860
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 18.621294021606445,
+      "learning_rate": 1.1390927912667045e-05,
+      "loss": 0.6418,
+      "step": 22870
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 12.699724197387695,
+      "learning_rate": 1.1387163561076607e-05,
+      "loss": 0.4555,
+      "step": 22880
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 8.417510986328125,
+      "learning_rate": 1.1383399209486168e-05,
+      "loss": 0.5236,
+      "step": 22890
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 9.543535232543945,
+      "learning_rate": 1.137963485789573e-05,
+      "loss": 0.5137,
+      "step": 22900
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 21.985715866088867,
+      "learning_rate": 1.137587050630529e-05,
+      "loss": 0.4621,
+      "step": 22910
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 2.261923313140869,
+      "learning_rate": 1.1372106154714851e-05,
+      "loss": 0.6878,
+      "step": 22920
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 12.996893882751465,
+      "learning_rate": 1.1368341803124412e-05,
+      "loss": 0.7822,
+      "step": 22930
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 1.7661455869674683,
+      "learning_rate": 1.1364577451533974e-05,
+      "loss": 0.627,
+      "step": 22940
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 16.084827423095703,
+      "learning_rate": 1.1360813099943535e-05,
+      "loss": 0.4727,
+      "step": 22950
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 10.136256217956543,
+      "learning_rate": 1.1357048748353097e-05,
+      "loss": 0.5857,
+      "step": 22960
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 13.618847846984863,
+      "learning_rate": 1.1353284396762658e-05,
+      "loss": 0.5905,
+      "step": 22970
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 1.0733072757720947,
+      "learning_rate": 1.134952004517222e-05,
+      "loss": 0.6132,
+      "step": 22980
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 4.069360733032227,
+      "learning_rate": 1.1345755693581781e-05,
+      "loss": 0.3998,
+      "step": 22990
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 27.945581436157227,
+      "learning_rate": 1.1341991341991343e-05,
+      "loss": 0.5658,
+      "step": 23000
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 0.5599634647369385,
+      "learning_rate": 1.1338226990400904e-05,
+      "loss": 0.3289,
+      "step": 23010
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 21.056047439575195,
+      "learning_rate": 1.1334462638810466e-05,
+      "loss": 0.4886,
+      "step": 23020
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 19.576793670654297,
+      "learning_rate": 1.1330698287220027e-05,
+      "loss": 0.584,
+      "step": 23030
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 8.226250648498535,
+      "learning_rate": 1.132693393562959e-05,
+      "loss": 0.7993,
+      "step": 23040
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 7.203392028808594,
+      "learning_rate": 1.1323169584039152e-05,
+      "loss": 0.5483,
+      "step": 23050
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 27.118764877319336,
+      "learning_rate": 1.1319405232448713e-05,
+      "loss": 0.6039,
+      "step": 23060
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 8.184642791748047,
+      "learning_rate": 1.1315640880858275e-05,
+      "loss": 0.5206,
+      "step": 23070
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 1.8194717168807983,
+      "learning_rate": 1.1311876529267836e-05,
+      "loss": 0.7129,
+      "step": 23080
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 7.509031772613525,
+      "learning_rate": 1.1308112177677396e-05,
+      "loss": 0.7681,
+      "step": 23090
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 19.41481590270996,
+      "learning_rate": 1.1304347826086957e-05,
+      "loss": 0.5878,
+      "step": 23100
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 9.031009674072266,
+      "learning_rate": 1.1300583474496519e-05,
+      "loss": 0.4608,
+      "step": 23110
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 3.5254902839660645,
+      "learning_rate": 1.129681912290608e-05,
+      "loss": 0.6231,
+      "step": 23120
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 14.69057559967041,
+      "learning_rate": 1.1293054771315642e-05,
+      "loss": 0.7666,
+      "step": 23130
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 26.575754165649414,
+      "learning_rate": 1.1289290419725203e-05,
+      "loss": 0.5632,
+      "step": 23140
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 12.611248016357422,
+      "learning_rate": 1.1285526068134765e-05,
+      "loss": 0.6975,
+      "step": 23150
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 19.788602828979492,
+      "learning_rate": 1.1281761716544326e-05,
+      "loss": 0.6446,
+      "step": 23160
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 20.97464370727539,
+      "learning_rate": 1.1277997364953887e-05,
+      "loss": 0.7142,
+      "step": 23170
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 11.11438274383545,
+      "learning_rate": 1.1274233013363449e-05,
+      "loss": 0.4329,
+      "step": 23180
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 6.7174296379089355,
+      "learning_rate": 1.127046866177301e-05,
+      "loss": 0.642,
+      "step": 23190
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 12.077559471130371,
+      "learning_rate": 1.1266704310182572e-05,
+      "loss": 0.6088,
+      "step": 23200
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 0.6835032105445862,
+      "learning_rate": 1.1262939958592133e-05,
+      "loss": 0.7096,
+      "step": 23210
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 13.644143104553223,
+      "learning_rate": 1.1259175607001695e-05,
+      "loss": 0.5963,
+      "step": 23220
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 17.367053985595703,
+      "learning_rate": 1.1255411255411258e-05,
+      "loss": 0.5982,
+      "step": 23230
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 19.097530364990234,
+      "learning_rate": 1.125164690382082e-05,
+      "loss": 0.7949,
+      "step": 23240
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 11.06638240814209,
+      "learning_rate": 1.124788255223038e-05,
+      "loss": 0.5742,
+      "step": 23250
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 6.701301097869873,
+      "learning_rate": 1.1244118200639939e-05,
+      "loss": 0.3653,
+      "step": 23260
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 7.139585018157959,
+      "learning_rate": 1.1240353849049502e-05,
+      "loss": 0.2213,
+      "step": 23270
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 15.305197715759277,
+      "learning_rate": 1.1236589497459064e-05,
+      "loss": 0.4505,
+      "step": 23280
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 9.9895601272583,
+      "learning_rate": 1.1232825145868625e-05,
+      "loss": 0.7347,
+      "step": 23290
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 30.742244720458984,
+      "learning_rate": 1.1229060794278186e-05,
+      "loss": 0.7636,
+      "step": 23300
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 24.35445785522461,
+      "learning_rate": 1.1225296442687748e-05,
+      "loss": 0.7456,
+      "step": 23310
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 24.246612548828125,
+      "learning_rate": 1.122153209109731e-05,
+      "loss": 0.6049,
+      "step": 23320
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 33.06283950805664,
+      "learning_rate": 1.121776773950687e-05,
+      "loss": 0.5307,
+      "step": 23330
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 25.850997924804688,
+      "learning_rate": 1.1214003387916432e-05,
+      "loss": 0.7956,
+      "step": 23340
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 11.392745018005371,
+      "learning_rate": 1.1210239036325994e-05,
+      "loss": 0.6775,
+      "step": 23350
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 10.8051118850708,
+      "learning_rate": 1.1206474684735555e-05,
+      "loss": 0.4612,
+      "step": 23360
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 32.45104217529297,
+      "learning_rate": 1.1202710333145117e-05,
+      "loss": 0.6303,
+      "step": 23370
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 5.765473365783691,
+      "learning_rate": 1.1198945981554678e-05,
+      "loss": 0.7854,
+      "step": 23380
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 0.4003254771232605,
+      "learning_rate": 1.119518162996424e-05,
+      "loss": 0.5774,
+      "step": 23390
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 8.617371559143066,
+      "learning_rate": 1.1191417278373801e-05,
+      "loss": 1.0461,
+      "step": 23400
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 17.002267837524414,
+      "learning_rate": 1.1187652926783363e-05,
+      "loss": 0.7655,
+      "step": 23410
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 20.015209197998047,
+      "learning_rate": 1.1183888575192926e-05,
+      "loss": 0.6856,
+      "step": 23420
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 5.477625846862793,
+      "learning_rate": 1.1180124223602484e-05,
+      "loss": 0.7369,
+      "step": 23430
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 9.53620719909668,
+      "learning_rate": 1.1176359872012045e-05,
+      "loss": 0.4399,
+      "step": 23440
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 3.9531502723693848,
+      "learning_rate": 1.1172595520421607e-05,
+      "loss": 0.9006,
+      "step": 23450
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 5.668755054473877,
+      "learning_rate": 1.116883116883117e-05,
+      "loss": 0.432,
+      "step": 23460
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 36.14069366455078,
+      "learning_rate": 1.1165066817240731e-05,
+      "loss": 0.7544,
+      "step": 23470
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 9.506393432617188,
+      "learning_rate": 1.1161302465650293e-05,
+      "loss": 0.7204,
+      "step": 23480
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 22.080291748046875,
+      "learning_rate": 1.1157538114059854e-05,
+      "loss": 0.5264,
+      "step": 23490
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 6.617203712463379,
+      "learning_rate": 1.1153773762469416e-05,
+      "loss": 0.5839,
+      "step": 23500
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 13.228835105895996,
+      "learning_rate": 1.1150009410878977e-05,
+      "loss": 0.7058,
+      "step": 23510
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 0.6176514625549316,
+      "learning_rate": 1.1146245059288539e-05,
+      "loss": 0.4557,
+      "step": 23520
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 0.5672368407249451,
+      "learning_rate": 1.11424807076981e-05,
+      "loss": 0.4541,
+      "step": 23530
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 2.1428656578063965,
+      "learning_rate": 1.1138716356107661e-05,
+      "loss": 0.7391,
+      "step": 23540
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 10.963018417358398,
+      "learning_rate": 1.1134952004517223e-05,
+      "loss": 0.5223,
+      "step": 23550
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 9.472241401672363,
+      "learning_rate": 1.1131187652926784e-05,
+      "loss": 0.5409,
+      "step": 23560
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 18.68202781677246,
+      "learning_rate": 1.1127423301336346e-05,
+      "loss": 0.6596,
+      "step": 23570
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 4.82119083404541,
+      "learning_rate": 1.1123658949745907e-05,
+      "loss": 0.6226,
+      "step": 23580
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 5.042635917663574,
+      "learning_rate": 1.1119894598155469e-05,
+      "loss": 0.6738,
+      "step": 23590
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 12.62334156036377,
+      "learning_rate": 1.1116130246565032e-05,
+      "loss": 0.7358,
+      "step": 23600
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 15.830984115600586,
+      "learning_rate": 1.111236589497459e-05,
+      "loss": 0.7408,
+      "step": 23610
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 9.996474266052246,
+      "learning_rate": 1.1108601543384151e-05,
+      "loss": 0.3432,
+      "step": 23620
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 42.45836639404297,
+      "learning_rate": 1.1104837191793713e-05,
+      "loss": 0.7615,
+      "step": 23630
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 32.250850677490234,
+      "learning_rate": 1.1101072840203274e-05,
+      "loss": 0.4661,
+      "step": 23640
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 24.006662368774414,
+      "learning_rate": 1.1097308488612838e-05,
+      "loss": 0.7126,
+      "step": 23650
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 7.966846942901611,
+      "learning_rate": 1.1093544137022399e-05,
+      "loss": 0.7873,
+      "step": 23660
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 16.041297912597656,
+      "learning_rate": 1.108977978543196e-05,
+      "loss": 0.5328,
+      "step": 23670
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 12.257124900817871,
+      "learning_rate": 1.1086015433841522e-05,
+      "loss": 0.549,
+      "step": 23680
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 9.333507537841797,
+      "learning_rate": 1.1082251082251083e-05,
+      "loss": 0.4787,
+      "step": 23690
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 15.94845199584961,
+      "learning_rate": 1.1078486730660645e-05,
+      "loss": 0.5384,
+      "step": 23700
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 15.848910331726074,
+      "learning_rate": 1.1074722379070206e-05,
+      "loss": 0.6707,
+      "step": 23710
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 12.358846664428711,
+      "learning_rate": 1.1070958027479768e-05,
+      "loss": 0.6222,
+      "step": 23720
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 16.612165451049805,
+      "learning_rate": 1.106719367588933e-05,
+      "loss": 0.4296,
+      "step": 23730
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 7.621057987213135,
+      "learning_rate": 1.106342932429889e-05,
+      "loss": 0.4142,
+      "step": 23740
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 26.88300895690918,
+      "learning_rate": 1.1059664972708452e-05,
+      "loss": 0.3773,
+      "step": 23750
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 14.327881813049316,
+      "learning_rate": 1.1055900621118014e-05,
+      "loss": 0.5598,
+      "step": 23760
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 7.642137050628662,
+      "learning_rate": 1.1052136269527575e-05,
+      "loss": 0.5747,
+      "step": 23770
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 11.110989570617676,
+      "learning_rate": 1.1048371917937135e-05,
+      "loss": 0.4744,
+      "step": 23780
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 11.560811042785645,
+      "learning_rate": 1.1044607566346696e-05,
+      "loss": 0.7039,
+      "step": 23790
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 17.159273147583008,
+      "learning_rate": 1.1040843214756258e-05,
+      "loss": 0.7643,
+      "step": 23800
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 15.362770080566406,
+      "learning_rate": 1.103707886316582e-05,
+      "loss": 0.6629,
+      "step": 23810
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.8261994123458862,
+      "learning_rate": 1.103331451157538e-05,
+      "loss": 0.5071,
+      "step": 23820
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 6.509442329406738,
+      "learning_rate": 1.1029550159984944e-05,
+      "loss": 0.4378,
+      "step": 23830
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 15.329565048217773,
+      "learning_rate": 1.1025785808394505e-05,
+      "loss": 0.5858,
+      "step": 23840
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 5.65818977355957,
+      "learning_rate": 1.1022021456804067e-05,
+      "loss": 0.4484,
+      "step": 23850
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 4.112398624420166,
+      "learning_rate": 1.1018257105213628e-05,
+      "loss": 0.4658,
+      "step": 23860
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 12.523710250854492,
+      "learning_rate": 1.101449275362319e-05,
+      "loss": 0.5546,
+      "step": 23870
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 0.8319380879402161,
+      "learning_rate": 1.1010728402032751e-05,
+      "loss": 0.4997,
+      "step": 23880
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 17.31648063659668,
+      "learning_rate": 1.1006964050442313e-05,
+      "loss": 0.4323,
+      "step": 23890
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 34.52608871459961,
+      "learning_rate": 1.1003199698851874e-05,
+      "loss": 0.544,
+      "step": 23900
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 9.72608757019043,
+      "learning_rate": 1.0999435347261436e-05,
+      "loss": 0.6388,
+      "step": 23910
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 17.97516441345215,
+      "learning_rate": 1.0995670995670997e-05,
+      "loss": 0.8557,
+      "step": 23920
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 0.5164489150047302,
+      "learning_rate": 1.0991906644080558e-05,
+      "loss": 0.7708,
+      "step": 23930
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 4.0153117179870605,
+      "learning_rate": 1.098814229249012e-05,
+      "loss": 0.6138,
+      "step": 23940
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 1.1666315793991089,
+      "learning_rate": 1.0984377940899681e-05,
+      "loss": 0.7278,
+      "step": 23950
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 9.93372917175293,
+      "learning_rate": 1.0980613589309241e-05,
+      "loss": 0.4048,
+      "step": 23960
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 21.520671844482422,
+      "learning_rate": 1.0976849237718803e-05,
+      "loss": 0.6345,
+      "step": 23970
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 3.4470133781433105,
+      "learning_rate": 1.0973084886128364e-05,
+      "loss": 1.0712,
+      "step": 23980
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 29.81308364868164,
+      "learning_rate": 1.0969320534537926e-05,
+      "loss": 0.4262,
+      "step": 23990
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 4.830698490142822,
+      "learning_rate": 1.0965556182947487e-05,
+      "loss": 0.7396,
+      "step": 24000
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 9.33254623413086,
+      "learning_rate": 1.0961791831357048e-05,
+      "loss": 0.8639,
+      "step": 24010
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 14.065550804138184,
+      "learning_rate": 1.0958027479766612e-05,
+      "loss": 0.623,
+      "step": 24020
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 11.378796577453613,
+      "learning_rate": 1.0954263128176173e-05,
+      "loss": 0.581,
+      "step": 24030
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 13.996289253234863,
+      "learning_rate": 1.0950498776585735e-05,
+      "loss": 0.5295,
+      "step": 24040
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 8.590402603149414,
+      "learning_rate": 1.0946734424995296e-05,
+      "loss": 0.7987,
+      "step": 24050
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 11.498170852661133,
+      "learning_rate": 1.0942970073404857e-05,
+      "loss": 0.6092,
+      "step": 24060
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 10.157093048095703,
+      "learning_rate": 1.0939205721814419e-05,
+      "loss": 0.3375,
+      "step": 24070
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 10.8042573928833,
+      "learning_rate": 1.093544137022398e-05,
+      "loss": 0.7559,
+      "step": 24080
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 0.13831891119480133,
+      "learning_rate": 1.0931677018633542e-05,
+      "loss": 0.655,
+      "step": 24090
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 21.07738494873047,
+      "learning_rate": 1.0927912667043103e-05,
+      "loss": 0.8755,
+      "step": 24100
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 25.695125579833984,
+      "learning_rate": 1.0924148315452665e-05,
+      "loss": 0.4189,
+      "step": 24110
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 23.678091049194336,
+      "learning_rate": 1.0920383963862226e-05,
+      "loss": 0.6966,
+      "step": 24120
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 17.928577423095703,
+      "learning_rate": 1.0916619612271786e-05,
+      "loss": 0.4332,
+      "step": 24130
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 40.44157409667969,
+      "learning_rate": 1.0912855260681347e-05,
+      "loss": 0.4427,
+      "step": 24140
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 4.393035411834717,
+      "learning_rate": 1.0909090909090909e-05,
+      "loss": 0.5062,
+      "step": 24150
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 46.50747299194336,
+      "learning_rate": 1.090532655750047e-05,
+      "loss": 0.7385,
+      "step": 24160
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 0.6289504170417786,
+      "learning_rate": 1.0901562205910032e-05,
+      "loss": 0.3068,
+      "step": 24170
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 26.921764373779297,
+      "learning_rate": 1.0897797854319593e-05,
+      "loss": 0.6155,
+      "step": 24180
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 38.250587463378906,
+      "learning_rate": 1.0894033502729155e-05,
+      "loss": 0.576,
+      "step": 24190
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 2.182755470275879,
+      "learning_rate": 1.0890269151138718e-05,
+      "loss": 0.5502,
+      "step": 24200
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 8.932013511657715,
+      "learning_rate": 1.088650479954828e-05,
+      "loss": 0.4656,
+      "step": 24210
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 8.519400596618652,
+      "learning_rate": 1.088274044795784e-05,
+      "loss": 0.3739,
+      "step": 24220
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 21.846633911132812,
+      "learning_rate": 1.0878976096367402e-05,
+      "loss": 0.5867,
+      "step": 24230
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 19.62921142578125,
+      "learning_rate": 1.0875211744776964e-05,
+      "loss": 0.5858,
+      "step": 24240
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 60.666866302490234,
+      "learning_rate": 1.0871447393186525e-05,
+      "loss": 0.6691,
+      "step": 24250
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 12.690692901611328,
+      "learning_rate": 1.0867683041596087e-05,
+      "loss": 0.5031,
+      "step": 24260
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 22.332693099975586,
+      "learning_rate": 1.0863918690005648e-05,
+      "loss": 0.6551,
+      "step": 24270
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 8.846254348754883,
+      "learning_rate": 1.086015433841521e-05,
+      "loss": 0.5339,
+      "step": 24280
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 30.030580520629883,
+      "learning_rate": 1.0856389986824771e-05,
+      "loss": 0.8483,
+      "step": 24290
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 30.104228973388672,
+      "learning_rate": 1.0852625635234333e-05,
+      "loss": 0.4223,
+      "step": 24300
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 26.719762802124023,
+      "learning_rate": 1.0848861283643892e-05,
+      "loss": 0.5728,
+      "step": 24310
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 8.01773452758789,
+      "learning_rate": 1.0845096932053454e-05,
+      "loss": 0.7319,
+      "step": 24320
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 17.984601974487305,
+      "learning_rate": 1.0841332580463015e-05,
+      "loss": 0.6263,
+      "step": 24330
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 1.179288625717163,
+      "learning_rate": 1.0837568228872577e-05,
+      "loss": 0.2777,
+      "step": 24340
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 17.420080184936523,
+      "learning_rate": 1.0833803877282138e-05,
+      "loss": 0.3534,
+      "step": 24350
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 20.7567138671875,
+      "learning_rate": 1.08300395256917e-05,
+      "loss": 1.0827,
+      "step": 24360
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 28.857418060302734,
+      "learning_rate": 1.0826275174101261e-05,
+      "loss": 0.6296,
+      "step": 24370
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 20.736568450927734,
+      "learning_rate": 1.0822510822510823e-05,
+      "loss": 0.9884,
+      "step": 24380
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 22.843528747558594,
+      "learning_rate": 1.0818746470920386e-05,
+      "loss": 0.4609,
+      "step": 24390
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 14.99477767944336,
+      "learning_rate": 1.0814982119329947e-05,
+      "loss": 0.7755,
+      "step": 24400
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 13.172017097473145,
+      "learning_rate": 1.0811217767739509e-05,
+      "loss": 0.4842,
+      "step": 24410
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 15.066924095153809,
+      "learning_rate": 1.080745341614907e-05,
+      "loss": 0.4145,
+      "step": 24420
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 0.11462697386741638,
+      "learning_rate": 1.0803689064558631e-05,
+      "loss": 0.5677,
+      "step": 24430
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 0.14985047280788422,
+      "learning_rate": 1.0799924712968193e-05,
+      "loss": 0.4298,
+      "step": 24440
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 13.569608688354492,
+      "learning_rate": 1.0796160361377754e-05,
+      "loss": 0.5217,
+      "step": 24450
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 9.481203079223633,
+      "learning_rate": 1.0792396009787316e-05,
+      "loss": 0.8526,
+      "step": 24460
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 4.702198028564453,
+      "learning_rate": 1.0788631658196877e-05,
+      "loss": 0.5461,
+      "step": 24470
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 10.113847732543945,
+      "learning_rate": 1.0784867306606437e-05,
+      "loss": 0.7229,
+      "step": 24480
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 9.467721939086914,
+      "learning_rate": 1.0781102955015999e-05,
+      "loss": 0.6535,
+      "step": 24490
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 18.867504119873047,
+      "learning_rate": 1.077733860342556e-05,
+      "loss": 0.7685,
+      "step": 24500
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 8.213113784790039,
+      "learning_rate": 1.0773574251835121e-05,
+      "loss": 0.7004,
+      "step": 24510
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 2.475691795349121,
+      "learning_rate": 1.0769809900244683e-05,
+      "loss": 0.6569,
+      "step": 24520
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 17.931053161621094,
+      "learning_rate": 1.0766045548654244e-05,
+      "loss": 0.6393,
+      "step": 24530
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 8.332486152648926,
+      "learning_rate": 1.0762281197063806e-05,
+      "loss": 0.6192,
+      "step": 24540
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 0.999433159828186,
+      "learning_rate": 1.0758516845473367e-05,
+      "loss": 0.5856,
+      "step": 24550
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 5.174107074737549,
+      "learning_rate": 1.0754752493882929e-05,
+      "loss": 0.6075,
+      "step": 24560
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 9.829113006591797,
+      "learning_rate": 1.0750988142292492e-05,
+      "loss": 0.5431,
+      "step": 24570
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 41.149696350097656,
+      "learning_rate": 1.0747223790702053e-05,
+      "loss": 0.4506,
+      "step": 24580
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 15.11819839477539,
+      "learning_rate": 1.0743459439111615e-05,
+      "loss": 0.7719,
+      "step": 24590
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 7.7300124168396,
+      "learning_rate": 1.0739695087521176e-05,
+      "loss": 0.5971,
+      "step": 24600
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 20.490882873535156,
+      "learning_rate": 1.0735930735930738e-05,
+      "loss": 0.4271,
+      "step": 24610
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 7.432632923126221,
+      "learning_rate": 1.07321663843403e-05,
+      "loss": 0.7363,
+      "step": 24620
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 15.051981925964355,
+      "learning_rate": 1.072840203274986e-05,
+      "loss": 0.5787,
+      "step": 24630
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 7.251710414886475,
+      "learning_rate": 1.0724637681159422e-05,
+      "loss": 0.7642,
+      "step": 24640
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 17.7536678314209,
+      "learning_rate": 1.0720873329568982e-05,
+      "loss": 0.6854,
+      "step": 24650
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 16.003955841064453,
+      "learning_rate": 1.0717108977978543e-05,
+      "loss": 0.409,
+      "step": 24660
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 4.870722770690918,
+      "learning_rate": 1.0713344626388105e-05,
+      "loss": 0.7751,
+      "step": 24670
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 17.760967254638672,
+      "learning_rate": 1.0709580274797666e-05,
+      "loss": 0.3988,
+      "step": 24680
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 1.264788031578064,
+      "learning_rate": 1.0705815923207228e-05,
+      "loss": 0.5068,
+      "step": 24690
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 18.01626205444336,
+      "learning_rate": 1.070205157161679e-05,
+      "loss": 0.4463,
+      "step": 24700
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 6.05813455581665,
+      "learning_rate": 1.069828722002635e-05,
+      "loss": 0.6254,
+      "step": 24710
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 25.01399803161621,
+      "learning_rate": 1.0694522868435912e-05,
+      "loss": 0.5808,
+      "step": 24720
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 5.349250316619873,
+      "learning_rate": 1.0690758516845474e-05,
+      "loss": 0.5773,
+      "step": 24730
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 0.2522965967655182,
+      "learning_rate": 1.0686994165255035e-05,
+      "loss": 0.3343,
+      "step": 24740
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 18.957963943481445,
+      "learning_rate": 1.0683229813664597e-05,
+      "loss": 0.6543,
+      "step": 24750
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 40.8657112121582,
+      "learning_rate": 1.067946546207416e-05,
+      "loss": 0.4904,
+      "step": 24760
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 20.800670623779297,
+      "learning_rate": 1.0675701110483721e-05,
+      "loss": 0.4968,
+      "step": 24770
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 7.221461296081543,
+      "learning_rate": 1.0671936758893283e-05,
+      "loss": 0.4546,
+      "step": 24780
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 2.7894575595855713,
+      "learning_rate": 1.0668172407302844e-05,
+      "loss": 0.4324,
+      "step": 24790
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 2.2078590393066406,
+      "learning_rate": 1.0664408055712406e-05,
+      "loss": 0.3639,
+      "step": 24800
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 13.342238426208496,
+      "learning_rate": 1.0660643704121967e-05,
+      "loss": 0.3975,
+      "step": 24810
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 15.394166946411133,
+      "learning_rate": 1.0656879352531528e-05,
+      "loss": 0.696,
+      "step": 24820
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 13.436577796936035,
+      "learning_rate": 1.0653115000941088e-05,
+      "loss": 0.8646,
+      "step": 24830
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 10.768505096435547,
+      "learning_rate": 1.064935064935065e-05,
+      "loss": 0.6416,
+      "step": 24840
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 21.5819034576416,
+      "learning_rate": 1.0645586297760211e-05,
+      "loss": 0.4017,
+      "step": 24850
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 4.942729473114014,
+      "learning_rate": 1.0641821946169773e-05,
+      "loss": 0.5493,
+      "step": 24860
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 28.3702392578125,
+      "learning_rate": 1.0638057594579334e-05,
+      "loss": 0.4549,
+      "step": 24870
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 18.81791877746582,
+      "learning_rate": 1.0634293242988896e-05,
+      "loss": 0.4436,
+      "step": 24880
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 14.989437103271484,
+      "learning_rate": 1.0630528891398457e-05,
+      "loss": 0.5157,
+      "step": 24890
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 10.192904472351074,
+      "learning_rate": 1.0626764539808018e-05,
+      "loss": 0.6498,
+      "step": 24900
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 11.987055778503418,
+      "learning_rate": 1.062300018821758e-05,
+      "loss": 0.5607,
+      "step": 24910
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 8.632203102111816,
+      "learning_rate": 1.0619235836627141e-05,
+      "loss": 0.7684,
+      "step": 24920
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 18.148733139038086,
+      "learning_rate": 1.0615471485036703e-05,
+      "loss": 0.4516,
+      "step": 24930
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 13.001415252685547,
+      "learning_rate": 1.0611707133446264e-05,
+      "loss": 0.6813,
+      "step": 24940
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 9.432195663452148,
+      "learning_rate": 1.0607942781855827e-05,
+      "loss": 0.7401,
+      "step": 24950
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 14.515077590942383,
+      "learning_rate": 1.0604178430265389e-05,
+      "loss": 0.411,
+      "step": 24960
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 7.635622024536133,
+      "learning_rate": 1.060041407867495e-05,
+      "loss": 0.55,
+      "step": 24970
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 21.15432357788086,
+      "learning_rate": 1.0596649727084512e-05,
+      "loss": 0.9953,
+      "step": 24980
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 12.600613594055176,
+      "learning_rate": 1.0592885375494073e-05,
+      "loss": 0.5475,
+      "step": 24990
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 15.378485679626465,
+      "learning_rate": 1.0589121023903633e-05,
+      "loss": 0.6201,
+      "step": 25000
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 16.676759719848633,
+      "learning_rate": 1.0585356672313195e-05,
+      "loss": 0.6947,
+      "step": 25010
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 9.91386604309082,
+      "learning_rate": 1.0581592320722756e-05,
+      "loss": 0.7228,
+      "step": 25020
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 21.984731674194336,
+      "learning_rate": 1.0577827969132317e-05,
+      "loss": 0.6805,
+      "step": 25030
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 0.7576162219047546,
+      "learning_rate": 1.0574063617541879e-05,
+      "loss": 0.4934,
+      "step": 25040
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 4.221212387084961,
+      "learning_rate": 1.057029926595144e-05,
+      "loss": 0.5943,
+      "step": 25050
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 11.483154296875,
+      "learning_rate": 1.0566534914361002e-05,
+      "loss": 0.3277,
+      "step": 25060
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 6.1331024169921875,
+      "learning_rate": 1.0562770562770563e-05,
+      "loss": 0.9332,
+      "step": 25070
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 1.2644256353378296,
+      "learning_rate": 1.0559006211180125e-05,
+      "loss": 0.894,
+      "step": 25080
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 4.269558906555176,
+      "learning_rate": 1.0555241859589686e-05,
+      "loss": 0.6306,
+      "step": 25090
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 17.586618423461914,
+      "learning_rate": 1.0551477507999248e-05,
+      "loss": 0.5255,
+      "step": 25100
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 0.7638131976127625,
+      "learning_rate": 1.0547713156408809e-05,
+      "loss": 0.6211,
+      "step": 25110
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 6.860489368438721,
+      "learning_rate": 1.054394880481837e-05,
+      "loss": 0.5254,
+      "step": 25120
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 3.1576449871063232,
+      "learning_rate": 1.0540184453227934e-05,
+      "loss": 0.7581,
+      "step": 25130
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 23.722333908081055,
+      "learning_rate": 1.0536420101637495e-05,
+      "loss": 0.3437,
+      "step": 25140
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 2.296964168548584,
+      "learning_rate": 1.0532655750047057e-05,
+      "loss": 0.4786,
+      "step": 25150
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 18.260684967041016,
+      "learning_rate": 1.0528891398456618e-05,
+      "loss": 0.8271,
+      "step": 25160
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 11.189855575561523,
+      "learning_rate": 1.052512704686618e-05,
+      "loss": 0.4455,
+      "step": 25170
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 25.472238540649414,
+      "learning_rate": 1.052136269527574e-05,
+      "loss": 0.5042,
+      "step": 25180
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 3.388648271560669,
+      "learning_rate": 1.05175983436853e-05,
+      "loss": 0.6337,
+      "step": 25190
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 6.780930042266846,
+      "learning_rate": 1.0513833992094862e-05,
+      "loss": 0.7147,
+      "step": 25200
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 28.922266006469727,
+      "learning_rate": 1.0510069640504424e-05,
+      "loss": 0.4112,
+      "step": 25210
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 14.36142349243164,
+      "learning_rate": 1.0506305288913985e-05,
+      "loss": 0.7538,
+      "step": 25220
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 16.995763778686523,
+      "learning_rate": 1.0502540937323547e-05,
+      "loss": 0.7889,
+      "step": 25230
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 0.24860745668411255,
+      "learning_rate": 1.0498776585733108e-05,
+      "loss": 0.4844,
+      "step": 25240
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 7.232578754425049,
+      "learning_rate": 1.049501223414267e-05,
+      "loss": 0.9642,
+      "step": 25250
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 6.162290573120117,
+      "learning_rate": 1.0491247882552231e-05,
+      "loss": 0.4363,
+      "step": 25260
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 4.756533145904541,
+      "learning_rate": 1.0487483530961793e-05,
+      "loss": 0.7687,
+      "step": 25270
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 5.880683422088623,
+      "learning_rate": 1.0483719179371354e-05,
+      "loss": 0.3593,
+      "step": 25280
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 12.597949981689453,
+      "learning_rate": 1.0479954827780915e-05,
+      "loss": 0.5685,
+      "step": 25290
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 4.098395824432373,
+      "learning_rate": 1.0476190476190477e-05,
+      "loss": 0.4729,
+      "step": 25300
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 30.62727928161621,
+      "learning_rate": 1.0472426124600038e-05,
+      "loss": 0.8119,
+      "step": 25310
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 18.5235595703125,
+      "learning_rate": 1.0468661773009602e-05,
+      "loss": 0.4133,
+      "step": 25320
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 26.45856285095215,
+      "learning_rate": 1.0464897421419163e-05,
+      "loss": 0.5182,
+      "step": 25330
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 16.52037811279297,
+      "learning_rate": 1.0461133069828724e-05,
+      "loss": 0.5143,
+      "step": 25340
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 0.0581025630235672,
+      "learning_rate": 1.0457368718238282e-05,
+      "loss": 0.6238,
+      "step": 25350
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 2.4997506141662598,
+      "learning_rate": 1.0453604366647846e-05,
+      "loss": 0.4463,
+      "step": 25360
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 11.980353355407715,
+      "learning_rate": 1.0449840015057407e-05,
+      "loss": 0.4376,
+      "step": 25370
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 14.3879976272583,
+      "learning_rate": 1.0446075663466969e-05,
+      "loss": 1.0135,
+      "step": 25380
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 20.742895126342773,
+      "learning_rate": 1.044231131187653e-05,
+      "loss": 0.5509,
+      "step": 25390
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 7.918363094329834,
+      "learning_rate": 1.0438546960286091e-05,
+      "loss": 0.6053,
+      "step": 25400
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 15.34104061126709,
+      "learning_rate": 1.0434782608695653e-05,
+      "loss": 0.7682,
+      "step": 25410
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 10.88797378540039,
+      "learning_rate": 1.0431018257105214e-05,
+      "loss": 0.6793,
+      "step": 25420
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 21.449735641479492,
+      "learning_rate": 1.0427253905514776e-05,
+      "loss": 0.7585,
+      "step": 25430
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 13.024362564086914,
+      "learning_rate": 1.0423489553924337e-05,
+      "loss": 0.6198,
+      "step": 25440
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 1.5722072124481201,
+      "learning_rate": 1.0419725202333899e-05,
+      "loss": 0.466,
+      "step": 25450
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 4.080865383148193,
+      "learning_rate": 1.041596085074346e-05,
+      "loss": 0.5095,
+      "step": 25460
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 23.28003692626953,
+      "learning_rate": 1.0412196499153022e-05,
+      "loss": 0.8175,
+      "step": 25470
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 8.450470924377441,
+      "learning_rate": 1.0408432147562583e-05,
+      "loss": 0.5053,
+      "step": 25480
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 18.8877010345459,
+      "learning_rate": 1.0404667795972145e-05,
+      "loss": 0.6964,
+      "step": 25490
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 4.335949420928955,
+      "learning_rate": 1.0400903444381708e-05,
+      "loss": 0.4675,
+      "step": 25500
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 6.426523685455322,
+      "learning_rate": 1.039713909279127e-05,
+      "loss": 0.6782,
+      "step": 25510
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 2.119450330734253,
+      "learning_rate": 1.0393374741200827e-05,
+      "loss": 0.4604,
+      "step": 25520
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 59.50654983520508,
+      "learning_rate": 1.0389610389610389e-05,
+      "loss": 0.6908,
+      "step": 25530
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 1.951970100402832,
+      "learning_rate": 1.038584603801995e-05,
+      "loss": 0.6409,
+      "step": 25540
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 1.0901389122009277,
+      "learning_rate": 1.0382081686429513e-05,
+      "loss": 0.5373,
+      "step": 25550
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 7.754991054534912,
+      "learning_rate": 1.0378317334839075e-05,
+      "loss": 0.4142,
+      "step": 25560
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 12.59786605834961,
+      "learning_rate": 1.0374552983248636e-05,
+      "loss": 0.544,
+      "step": 25570
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 32.1705436706543,
+      "learning_rate": 1.0370788631658198e-05,
+      "loss": 0.4405,
+      "step": 25580
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 12.26833438873291,
+      "learning_rate": 1.036702428006776e-05,
+      "loss": 0.9665,
+      "step": 25590
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 7.950512886047363,
+      "learning_rate": 1.036325992847732e-05,
+      "loss": 0.7382,
+      "step": 25600
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 14.049162864685059,
+      "learning_rate": 1.0359495576886882e-05,
+      "loss": 0.4283,
+      "step": 25610
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 19.716205596923828,
+      "learning_rate": 1.0355731225296444e-05,
+      "loss": 0.3962,
+      "step": 25620
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 10.237818717956543,
+      "learning_rate": 1.0351966873706005e-05,
+      "loss": 0.5062,
+      "step": 25630
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 27.731966018676758,
+      "learning_rate": 1.0348202522115567e-05,
+      "loss": 0.4489,
+      "step": 25640
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 2.626387596130371,
+      "learning_rate": 1.0344438170525128e-05,
+      "loss": 0.3363,
+      "step": 25650
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 23.135208129882812,
+      "learning_rate": 1.034067381893469e-05,
+      "loss": 0.7282,
+      "step": 25660
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 17.033761978149414,
+      "learning_rate": 1.0336909467344251e-05,
+      "loss": 0.5404,
+      "step": 25670
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 5.097890377044678,
+      "learning_rate": 1.0333145115753812e-05,
+      "loss": 1.1042,
+      "step": 25680
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 10.225035667419434,
+      "learning_rate": 1.0329380764163376e-05,
+      "loss": 0.4477,
+      "step": 25690
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 7.208095550537109,
+      "learning_rate": 1.0325616412572934e-05,
+      "loss": 0.699,
+      "step": 25700
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 9.896465301513672,
+      "learning_rate": 1.0321852060982495e-05,
+      "loss": 0.4655,
+      "step": 25710
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 12.159602165222168,
+      "learning_rate": 1.0318087709392057e-05,
+      "loss": 0.7253,
+      "step": 25720
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 0.5531998872756958,
+      "learning_rate": 1.031432335780162e-05,
+      "loss": 0.3742,
+      "step": 25730
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 8.790497779846191,
+      "learning_rate": 1.0310559006211181e-05,
+      "loss": 0.5695,
+      "step": 25740
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 3.4259729385375977,
+      "learning_rate": 1.0306794654620743e-05,
+      "loss": 0.7616,
+      "step": 25750
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 21.889638900756836,
+      "learning_rate": 1.0303030303030304e-05,
+      "loss": 0.6106,
+      "step": 25760
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 13.903435707092285,
+      "learning_rate": 1.0299265951439866e-05,
+      "loss": 0.6131,
+      "step": 25770
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 19.042236328125,
+      "learning_rate": 1.0295501599849427e-05,
+      "loss": 0.6028,
+      "step": 25780
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 25.542591094970703,
+      "learning_rate": 1.0291737248258988e-05,
+      "loss": 0.5859,
+      "step": 25790
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 12.444345474243164,
+      "learning_rate": 1.028797289666855e-05,
+      "loss": 0.6576,
+      "step": 25800
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 0.6395838260650635,
+      "learning_rate": 1.0284208545078111e-05,
+      "loss": 0.7174,
+      "step": 25810
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 13.468929290771484,
+      "learning_rate": 1.0280444193487673e-05,
+      "loss": 0.6934,
+      "step": 25820
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 16.97991180419922,
+      "learning_rate": 1.0276679841897234e-05,
+      "loss": 0.9544,
+      "step": 25830
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 22.151426315307617,
+      "learning_rate": 1.0272915490306796e-05,
+      "loss": 0.5222,
+      "step": 25840
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 9.480589866638184,
+      "learning_rate": 1.0269151138716357e-05,
+      "loss": 0.6157,
+      "step": 25850
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 15.794319152832031,
+      "learning_rate": 1.0265386787125919e-05,
+      "loss": 0.8519,
+      "step": 25860
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 9.91077709197998,
+      "learning_rate": 1.0261622435535478e-05,
+      "loss": 0.3128,
+      "step": 25870
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 10.460318565368652,
+      "learning_rate": 1.025785808394504e-05,
+      "loss": 0.7739,
+      "step": 25880
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 29.492027282714844,
+      "learning_rate": 1.0254093732354601e-05,
+      "loss": 0.6284,
+      "step": 25890
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 7.003443241119385,
+      "learning_rate": 1.0250329380764163e-05,
+      "loss": 0.4957,
+      "step": 25900
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 36.19068145751953,
+      "learning_rate": 1.0246565029173724e-05,
+      "loss": 0.5009,
+      "step": 25910
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 17.596288681030273,
+      "learning_rate": 1.0242800677583287e-05,
+      "loss": 0.5891,
+      "step": 25920
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 34.26020812988281,
+      "learning_rate": 1.0239036325992849e-05,
+      "loss": 0.767,
+      "step": 25930
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 0.572287380695343,
+      "learning_rate": 1.023527197440241e-05,
+      "loss": 0.4655,
+      "step": 25940
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 11.611014366149902,
+      "learning_rate": 1.0231507622811972e-05,
+      "loss": 0.6033,
+      "step": 25950
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 5.741128444671631,
+      "learning_rate": 1.0227743271221533e-05,
+      "loss": 0.2666,
+      "step": 25960
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 27.423297882080078,
+      "learning_rate": 1.0223978919631095e-05,
+      "loss": 0.5137,
+      "step": 25970
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 7.721019744873047,
+      "learning_rate": 1.0220214568040656e-05,
+      "loss": 0.8591,
+      "step": 25980
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 5.386199474334717,
+      "learning_rate": 1.0216450216450218e-05,
+      "loss": 0.5424,
+      "step": 25990
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 16.02313804626465,
+      "learning_rate": 1.0212685864859779e-05,
+      "loss": 0.9359,
+      "step": 26000
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 28.932086944580078,
+      "learning_rate": 1.020892151326934e-05,
+      "loss": 0.4697,
+      "step": 26010
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 7.247927665710449,
+      "learning_rate": 1.0205157161678902e-05,
+      "loss": 0.5748,
+      "step": 26020
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 0.293780118227005,
+      "learning_rate": 1.0201392810088464e-05,
+      "loss": 0.4207,
+      "step": 26030
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 9.770951271057129,
+      "learning_rate": 1.0197628458498025e-05,
+      "loss": 0.6643,
+      "step": 26040
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 17.128610610961914,
+      "learning_rate": 1.0193864106907585e-05,
+      "loss": 0.4047,
+      "step": 26050
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 20.82026481628418,
+      "learning_rate": 1.0190099755317146e-05,
+      "loss": 0.7626,
+      "step": 26060
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 9.654380798339844,
+      "learning_rate": 1.0186335403726708e-05,
+      "loss": 0.4447,
+      "step": 26070
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 19.486652374267578,
+      "learning_rate": 1.0182571052136269e-05,
+      "loss": 0.4849,
+      "step": 26080
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 15.955955505371094,
+      "learning_rate": 1.017880670054583e-05,
+      "loss": 0.7018,
+      "step": 26090
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 6.0235915184021,
+      "learning_rate": 1.0175042348955392e-05,
+      "loss": 0.5864,
+      "step": 26100
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 41.60502624511719,
+      "learning_rate": 1.0171277997364955e-05,
+      "loss": 0.9113,
+      "step": 26110
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 10.696873664855957,
+      "learning_rate": 1.0167513645774517e-05,
+      "loss": 0.4178,
+      "step": 26120
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 27.66557502746582,
+      "learning_rate": 1.0163749294184078e-05,
+      "loss": 0.5074,
+      "step": 26130
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 4.632658958435059,
+      "learning_rate": 1.015998494259364e-05,
+      "loss": 0.3351,
+      "step": 26140
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 26.457387924194336,
+      "learning_rate": 1.0156220591003201e-05,
+      "loss": 1.3971,
+      "step": 26150
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 3.868096351623535,
+      "learning_rate": 1.0152456239412763e-05,
+      "loss": 0.5699,
+      "step": 26160
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 1.3630057573318481,
+      "learning_rate": 1.0148691887822324e-05,
+      "loss": 0.4219,
+      "step": 26170
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 34.00837326049805,
+      "learning_rate": 1.0144927536231885e-05,
+      "loss": 0.8625,
+      "step": 26180
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 2.9621551036834717,
+      "learning_rate": 1.0141163184641447e-05,
+      "loss": 0.6725,
+      "step": 26190
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 0.44564133882522583,
+      "learning_rate": 1.0137398833051008e-05,
+      "loss": 0.6897,
+      "step": 26200
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 10.090829849243164,
+      "learning_rate": 1.013363448146057e-05,
+      "loss": 0.7626,
+      "step": 26210
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 10.721175193786621,
+      "learning_rate": 1.012987012987013e-05,
+      "loss": 0.7807,
+      "step": 26220
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 8.438685417175293,
+      "learning_rate": 1.0126105778279691e-05,
+      "loss": 0.6646,
+      "step": 26230
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 20.982872009277344,
+      "learning_rate": 1.0122341426689252e-05,
+      "loss": 0.7103,
+      "step": 26240
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 24.98394012451172,
+      "learning_rate": 1.0118577075098814e-05,
+      "loss": 0.658,
+      "step": 26250
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 10.647464752197266,
+      "learning_rate": 1.0114812723508375e-05,
+      "loss": 0.5797,
+      "step": 26260
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 5.082032203674316,
+      "learning_rate": 1.0111048371917937e-05,
+      "loss": 0.3246,
+      "step": 26270
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 34.345130920410156,
+      "learning_rate": 1.0107284020327498e-05,
+      "loss": 0.7128,
+      "step": 26280
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 1.867757797241211,
+      "learning_rate": 1.0103519668737061e-05,
+      "loss": 0.4022,
+      "step": 26290
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 17.364198684692383,
+      "learning_rate": 1.0099755317146623e-05,
+      "loss": 0.4057,
+      "step": 26300
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 0.3039032518863678,
+      "learning_rate": 1.0095990965556184e-05,
+      "loss": 0.6772,
+      "step": 26310
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 7.0659050941467285,
+      "learning_rate": 1.0092226613965746e-05,
+      "loss": 0.4634,
+      "step": 26320
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 6.5546135902404785,
+      "learning_rate": 1.0088462262375307e-05,
+      "loss": 0.6775,
+      "step": 26330
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 19.550182342529297,
+      "learning_rate": 1.0084697910784869e-05,
+      "loss": 0.3757,
+      "step": 26340
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 10.655951499938965,
+      "learning_rate": 1.008093355919443e-05,
+      "loss": 0.5657,
+      "step": 26350
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 7.180771350860596,
+      "learning_rate": 1.0077169207603992e-05,
+      "loss": 0.7091,
+      "step": 26360
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 10.868316650390625,
+      "learning_rate": 1.0073404856013553e-05,
+      "loss": 0.6567,
+      "step": 26370
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 43.5999641418457,
+      "learning_rate": 1.0069640504423115e-05,
+      "loss": 0.4698,
+      "step": 26380
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 12.528432846069336,
+      "learning_rate": 1.0065876152832676e-05,
+      "loss": 0.7588,
+      "step": 26390
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 13.703747749328613,
+      "learning_rate": 1.0062111801242236e-05,
+      "loss": 0.7794,
+      "step": 26400
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 0.5737854838371277,
+      "learning_rate": 1.0058347449651797e-05,
+      "loss": 0.6142,
+      "step": 26410
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 15.537364959716797,
+      "learning_rate": 1.0054583098061359e-05,
+      "loss": 0.585,
+      "step": 26420
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 14.01577377319336,
+      "learning_rate": 1.005081874647092e-05,
+      "loss": 0.7088,
+      "step": 26430
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 5.803639888763428,
+      "learning_rate": 1.0047054394880482e-05,
+      "loss": 0.7963,
+      "step": 26440
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 16.50397300720215,
+      "learning_rate": 1.0043290043290043e-05,
+      "loss": 0.5466,
+      "step": 26450
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 10.755285263061523,
+      "learning_rate": 1.0039525691699605e-05,
+      "loss": 0.4802,
+      "step": 26460
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 4.0640869140625,
+      "learning_rate": 1.0035761340109166e-05,
+      "loss": 0.8407,
+      "step": 26470
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 10.815974235534668,
+      "learning_rate": 1.003199698851873e-05,
+      "loss": 0.5498,
+      "step": 26480
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 21.308177947998047,
+      "learning_rate": 1.002823263692829e-05,
+      "loss": 0.6421,
+      "step": 26490
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 20.949392318725586,
+      "learning_rate": 1.0024468285337852e-05,
+      "loss": 0.4232,
+      "step": 26500
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 6.295169830322266,
+      "learning_rate": 1.0020703933747414e-05,
+      "loss": 0.3888,
+      "step": 26510
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 22.155059814453125,
+      "learning_rate": 1.0016939582156975e-05,
+      "loss": 0.4725,
+      "step": 26520
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 11.023420333862305,
+      "learning_rate": 1.0013175230566537e-05,
+      "loss": 0.9657,
+      "step": 26530
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 7.431711673736572,
+      "learning_rate": 1.0009410878976098e-05,
+      "loss": 0.8442,
+      "step": 26540
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 7.200937271118164,
+      "learning_rate": 1.000564652738566e-05,
+      "loss": 0.4408,
+      "step": 26550
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 16.981725692749023,
+      "learning_rate": 1.0001882175795221e-05,
+      "loss": 0.4665,
+      "step": 26560
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.9174666666666667,
+      "eval_loss": 0.3038625717163086,
+      "eval_runtime": 51.1218,
+      "eval_samples_per_second": 146.708,
+      "eval_steps_per_second": 18.348,
+      "step": 26565
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 11.93405532836914,
+      "learning_rate": 9.998117824204782e-06,
+      "loss": 0.5697,
+      "step": 26570
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 5.752691745758057,
+      "learning_rate": 9.994353472614344e-06,
+      "loss": 0.407,
+      "step": 26580
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.38019922375679016,
+      "learning_rate": 9.990589121023905e-06,
+      "loss": 0.5588,
+      "step": 26590
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 23.82176399230957,
+      "learning_rate": 9.986824769433467e-06,
+      "loss": 0.7845,
+      "step": 26600
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 18.493980407714844,
+      "learning_rate": 9.983060417843027e-06,
+      "loss": 1.0395,
+      "step": 26610
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 0.250558078289032,
+      "learning_rate": 9.979296066252588e-06,
+      "loss": 0.2666,
+      "step": 26620
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 12.60075569152832,
+      "learning_rate": 9.97553171466215e-06,
+      "loss": 0.4569,
+      "step": 26630
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 12.024946212768555,
+      "learning_rate": 9.971767363071711e-06,
+      "loss": 0.4128,
+      "step": 26640
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 0.44587206840515137,
+      "learning_rate": 9.968003011481272e-06,
+      "loss": 0.5314,
+      "step": 26650
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 3.9360930919647217,
+      "learning_rate": 9.964238659890836e-06,
+      "loss": 0.5811,
+      "step": 26660
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 5.4166035652160645,
+      "learning_rate": 9.960474308300397e-06,
+      "loss": 0.5869,
+      "step": 26670
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 13.620014190673828,
+      "learning_rate": 9.956709956709958e-06,
+      "loss": 0.5772,
+      "step": 26680
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 18.73905372619629,
+      "learning_rate": 9.95294560511952e-06,
+      "loss": 0.4458,
+      "step": 26690
+    },
+    {
+      "epoch": 5.03,
+      "grad_norm": 9.41979694366455,
+      "learning_rate": 9.94918125352908e-06,
+      "loss": 0.6342,
+      "step": 26700
+    },
+    {
+      "epoch": 5.03,
+      "grad_norm": 6.560514450073242,
+      "learning_rate": 9.945416901938641e-06,
+      "loss": 0.8343,
+      "step": 26710
+    },
+    {
+      "epoch": 5.03,
+      "grad_norm": 24.214908599853516,
+      "learning_rate": 9.941652550348203e-06,
+      "loss": 0.3422,
+      "step": 26720
+    },
+    {
+      "epoch": 5.03,
+      "grad_norm": 13.99776554107666,
+      "learning_rate": 9.937888198757764e-06,
+      "loss": 0.6253,
+      "step": 26730
+    },
+    {
+      "epoch": 5.03,
+      "grad_norm": 13.242351531982422,
+      "learning_rate": 9.934123847167326e-06,
+      "loss": 0.543,
+      "step": 26740
+    },
+    {
+      "epoch": 5.03,
+      "grad_norm": 33.52753448486328,
+      "learning_rate": 9.930359495576887e-06,
+      "loss": 0.4902,
+      "step": 26750
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 23.321645736694336,
+      "learning_rate": 9.92659514398645e-06,
+      "loss": 0.656,
+      "step": 26760
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 16.174564361572266,
+      "learning_rate": 9.922830792396012e-06,
+      "loss": 0.5337,
+      "step": 26770
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 4.621311664581299,
+      "learning_rate": 9.919066440805573e-06,
+      "loss": 0.7531,
+      "step": 26780
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 29.9796085357666,
+      "learning_rate": 9.915302089215133e-06,
+      "loss": 0.6874,
+      "step": 26790
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 26.092121124267578,
+      "learning_rate": 9.911537737624694e-06,
+      "loss": 0.7789,
+      "step": 26800
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 7.301915645599365,
+      "learning_rate": 9.907773386034256e-06,
+      "loss": 0.385,
+      "step": 26810
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 19.605470657348633,
+      "learning_rate": 9.904009034443817e-06,
+      "loss": 0.702,
+      "step": 26820
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 13.46226692199707,
+      "learning_rate": 9.900244682853379e-06,
+      "loss": 0.24,
+      "step": 26830
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 5.802454948425293,
+      "learning_rate": 9.89648033126294e-06,
+      "loss": 0.2608,
+      "step": 26840
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 8.033814430236816,
+      "learning_rate": 9.892715979672503e-06,
+      "loss": 0.6069,
+      "step": 26850
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 0.2305469959974289,
+      "learning_rate": 9.888951628082065e-06,
+      "loss": 0.4146,
+      "step": 26860
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 7.512425422668457,
+      "learning_rate": 9.885187276491625e-06,
+      "loss": 0.6316,
+      "step": 26870
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 3.023247480392456,
+      "learning_rate": 9.881422924901186e-06,
+      "loss": 0.7018,
+      "step": 26880
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 20.460777282714844,
+      "learning_rate": 9.877658573310747e-06,
+      "loss": 0.5622,
+      "step": 26890
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 5.8153839111328125,
+      "learning_rate": 9.873894221720309e-06,
+      "loss": 0.5251,
+      "step": 26900
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 27.608034133911133,
+      "learning_rate": 9.87012987012987e-06,
+      "loss": 0.6028,
+      "step": 26910
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 33.49705123901367,
+      "learning_rate": 9.866365518539432e-06,
+      "loss": 0.5239,
+      "step": 26920
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 1.4094599485397339,
+      "learning_rate": 9.862601166948993e-06,
+      "loss": 0.7576,
+      "step": 26930
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 8.930601119995117,
+      "learning_rate": 9.858836815358556e-06,
+      "loss": 0.4088,
+      "step": 26940
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 0.6943548321723938,
+      "learning_rate": 9.855072463768118e-06,
+      "loss": 0.633,
+      "step": 26950
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 33.55091094970703,
+      "learning_rate": 9.851308112177678e-06,
+      "loss": 0.7098,
+      "step": 26960
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 18.824798583984375,
+      "learning_rate": 9.847543760587239e-06,
+      "loss": 0.5024,
+      "step": 26970
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 19.024484634399414,
+      "learning_rate": 9.8437794089968e-06,
+      "loss": 0.5722,
+      "step": 26980
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 4.507168292999268,
+      "learning_rate": 9.840015057406362e-06,
+      "loss": 0.2062,
+      "step": 26990
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 31.550439834594727,
+      "learning_rate": 9.836250705815924e-06,
+      "loss": 0.4145,
+      "step": 27000
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 11.541763305664062,
+      "learning_rate": 9.832486354225485e-06,
+      "loss": 0.522,
+      "step": 27010
+    },
+    {
+      "epoch": 5.09,
+      "grad_norm": 14.351038932800293,
+      "learning_rate": 9.828722002635046e-06,
+      "loss": 0.5481,
+      "step": 27020
+    },
+    {
+      "epoch": 5.09,
+      "grad_norm": 4.900013446807861,
+      "learning_rate": 9.82495765104461e-06,
+      "loss": 0.4472,
+      "step": 27030
+    },
+    {
+      "epoch": 5.09,
+      "grad_norm": 14.692139625549316,
+      "learning_rate": 9.821193299454171e-06,
+      "loss": 0.6838,
+      "step": 27040
+    },
+    {
+      "epoch": 5.09,
+      "grad_norm": 0.5007912516593933,
+      "learning_rate": 9.81742894786373e-06,
+      "loss": 0.7123,
+      "step": 27050
+    },
+    {
+      "epoch": 5.09,
+      "grad_norm": 29.906484603881836,
+      "learning_rate": 9.813664596273292e-06,
+      "loss": 0.3299,
+      "step": 27060
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 21.051406860351562,
+      "learning_rate": 9.809900244682854e-06,
+      "loss": 0.471,
+      "step": 27070
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 8.312321662902832,
+      "learning_rate": 9.806135893092415e-06,
+      "loss": 0.5687,
+      "step": 27080
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 0.8224940896034241,
+      "learning_rate": 9.802371541501977e-06,
+      "loss": 0.3544,
+      "step": 27090
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 18.394412994384766,
+      "learning_rate": 9.798607189911538e-06,
+      "loss": 0.4271,
+      "step": 27100
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 14.561628341674805,
+      "learning_rate": 9.7948428383211e-06,
+      "loss": 0.7428,
+      "step": 27110
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 17.597429275512695,
+      "learning_rate": 9.791078486730661e-06,
+      "loss": 0.322,
+      "step": 27120
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 0.4523058831691742,
+      "learning_rate": 9.787314135140224e-06,
+      "loss": 0.4471,
+      "step": 27130
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 57.31825637817383,
+      "learning_rate": 9.783549783549784e-06,
+      "loss": 1.022,
+      "step": 27140
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 15.83234977722168,
+      "learning_rate": 9.779785431959345e-06,
+      "loss": 0.5429,
+      "step": 27150
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 17.558889389038086,
+      "learning_rate": 9.776021080368907e-06,
+      "loss": 0.4771,
+      "step": 27160
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 6.830644607543945,
+      "learning_rate": 9.772256728778468e-06,
+      "loss": 0.3859,
+      "step": 27170
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 23.56614875793457,
+      "learning_rate": 9.76849237718803e-06,
+      "loss": 0.2237,
+      "step": 27180
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 17.846912384033203,
+      "learning_rate": 9.764728025597591e-06,
+      "loss": 0.5612,
+      "step": 27190
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 22.106182098388672,
+      "learning_rate": 9.760963674007153e-06,
+      "loss": 0.4336,
+      "step": 27200
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 34.99475860595703,
+      "learning_rate": 9.757199322416714e-06,
+      "loss": 0.5471,
+      "step": 27210
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 22.88408660888672,
+      "learning_rate": 9.753434970826276e-06,
+      "loss": 0.5702,
+      "step": 27220
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 0.2636334300041199,
+      "learning_rate": 9.749670619235837e-06,
+      "loss": 0.6926,
+      "step": 27230
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 0.4821383059024811,
+      "learning_rate": 9.745906267645399e-06,
+      "loss": 0.4507,
+      "step": 27240
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 0.9522204399108887,
+      "learning_rate": 9.74214191605496e-06,
+      "loss": 0.7467,
+      "step": 27250
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 0.07715526968240738,
+      "learning_rate": 9.738377564464521e-06,
+      "loss": 0.5735,
+      "step": 27260
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 2.0689947605133057,
+      "learning_rate": 9.734613212874083e-06,
+      "loss": 0.621,
+      "step": 27270
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 4.146627902984619,
+      "learning_rate": 9.730848861283644e-06,
+      "loss": 0.3251,
+      "step": 27280
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 25.737661361694336,
+      "learning_rate": 9.727084509693206e-06,
+      "loss": 0.34,
+      "step": 27290
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 10.852874755859375,
+      "learning_rate": 9.723320158102767e-06,
+      "loss": 0.6041,
+      "step": 27300
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 18.879383087158203,
+      "learning_rate": 9.719555806512329e-06,
+      "loss": 0.6012,
+      "step": 27310
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 0.9953970909118652,
+      "learning_rate": 9.71579145492189e-06,
+      "loss": 0.62,
+      "step": 27320
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 1.2048535346984863,
+      "learning_rate": 9.712027103331452e-06,
+      "loss": 0.5219,
+      "step": 27330
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 13.980968475341797,
+      "learning_rate": 9.708262751741013e-06,
+      "loss": 0.7282,
+      "step": 27340
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 9.173256874084473,
+      "learning_rate": 9.704498400150575e-06,
+      "loss": 0.432,
+      "step": 27350
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 11.977608680725098,
+      "learning_rate": 9.700734048560136e-06,
+      "loss": 0.5185,
+      "step": 27360
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 9.067453384399414,
+      "learning_rate": 9.696969696969698e-06,
+      "loss": 0.4714,
+      "step": 27370
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 4.337044715881348,
+      "learning_rate": 9.693205345379259e-06,
+      "loss": 0.3863,
+      "step": 27380
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 6.323875427246094,
+      "learning_rate": 9.68944099378882e-06,
+      "loss": 0.7175,
+      "step": 27390
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 2.4866652488708496,
+      "learning_rate": 9.685676642198382e-06,
+      "loss": 0.3135,
+      "step": 27400
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 2.433683395385742,
+      "learning_rate": 9.681912290607943e-06,
+      "loss": 0.576,
+      "step": 27410
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 10.03541374206543,
+      "learning_rate": 9.678147939017505e-06,
+      "loss": 0.4971,
+      "step": 27420
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 22.778186798095703,
+      "learning_rate": 9.674383587427066e-06,
+      "loss": 0.5237,
+      "step": 27430
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 4.677865028381348,
+      "learning_rate": 9.670619235836628e-06,
+      "loss": 0.7147,
+      "step": 27440
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 21.216205596923828,
+      "learning_rate": 9.66685488424619e-06,
+      "loss": 0.5606,
+      "step": 27450
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 10.490550994873047,
+      "learning_rate": 9.66309053265575e-06,
+      "loss": 0.7098,
+      "step": 27460
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 7.556525230407715,
+      "learning_rate": 9.659326181065312e-06,
+      "loss": 0.6452,
+      "step": 27470
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 0.8397094011306763,
+      "learning_rate": 9.655561829474874e-06,
+      "loss": 0.1972,
+      "step": 27480
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 9.92435073852539,
+      "learning_rate": 9.651797477884435e-06,
+      "loss": 0.4522,
+      "step": 27490
+    },
+    {
+      "epoch": 5.18,
+      "grad_norm": 1.3592374324798584,
+      "learning_rate": 9.648033126293997e-06,
+      "loss": 0.4904,
+      "step": 27500
+    },
+    {
+      "epoch": 5.18,
+      "grad_norm": 10.651955604553223,
+      "learning_rate": 9.644268774703558e-06,
+      "loss": 0.5157,
+      "step": 27510
+    },
+    {
+      "epoch": 5.18,
+      "grad_norm": 18.821256637573242,
+      "learning_rate": 9.64050442311312e-06,
+      "loss": 0.5196,
+      "step": 27520
+    },
+    {
+      "epoch": 5.18,
+      "grad_norm": 3.651150703430176,
+      "learning_rate": 9.636740071522681e-06,
+      "loss": 0.5902,
+      "step": 27530
+    },
+    {
+      "epoch": 5.18,
+      "grad_norm": 13.892976760864258,
+      "learning_rate": 9.632975719932242e-06,
+      "loss": 0.8091,
+      "step": 27540
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 0.7897984981536865,
+      "learning_rate": 9.629211368341804e-06,
+      "loss": 0.4273,
+      "step": 27550
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 16.25598907470703,
+      "learning_rate": 9.625447016751365e-06,
+      "loss": 0.3544,
+      "step": 27560
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 15.311908721923828,
+      "learning_rate": 9.621682665160927e-06,
+      "loss": 0.3885,
+      "step": 27570
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 13.913825988769531,
+      "learning_rate": 9.617918313570488e-06,
+      "loss": 0.3768,
+      "step": 27580
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 0.829640805721283,
+      "learning_rate": 9.61415396198005e-06,
+      "loss": 0.3235,
+      "step": 27590
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 11.236851692199707,
+      "learning_rate": 9.610389610389611e-06,
+      "loss": 0.73,
+      "step": 27600
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 9.920482635498047,
+      "learning_rate": 9.606625258799173e-06,
+      "loss": 0.379,
+      "step": 27610
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 5.420136451721191,
+      "learning_rate": 9.602860907208734e-06,
+      "loss": 0.7466,
+      "step": 27620
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 32.80744934082031,
+      "learning_rate": 9.599096555618296e-06,
+      "loss": 0.906,
+      "step": 27630
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 21.18132209777832,
+      "learning_rate": 9.595332204027857e-06,
+      "loss": 0.8318,
+      "step": 27640
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 7.528763771057129,
+      "learning_rate": 9.591567852437418e-06,
+      "loss": 0.6328,
+      "step": 27650
+    },
+    {
+      "epoch": 5.21,
+      "grad_norm": 24.74061393737793,
+      "learning_rate": 9.58780350084698e-06,
+      "loss": 0.5783,
+      "step": 27660
+    },
+    {
+      "epoch": 5.21,
+      "grad_norm": 15.430523872375488,
+      "learning_rate": 9.584039149256541e-06,
+      "loss": 0.554,
+      "step": 27670
+    },
+    {
+      "epoch": 5.21,
+      "grad_norm": 23.323516845703125,
+      "learning_rate": 9.580274797666103e-06,
+      "loss": 0.246,
+      "step": 27680
+    },
+    {
+      "epoch": 5.21,
+      "grad_norm": 29.96780776977539,
+      "learning_rate": 9.576510446075664e-06,
+      "loss": 0.7304,
+      "step": 27690
+    },
+    {
+      "epoch": 5.21,
+      "grad_norm": 19.847814559936523,
+      "learning_rate": 9.572746094485226e-06,
+      "loss": 0.4962,
+      "step": 27700
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 11.185002326965332,
+      "learning_rate": 9.568981742894787e-06,
+      "loss": 0.4532,
+      "step": 27710
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 12.05192756652832,
+      "learning_rate": 9.565217391304349e-06,
+      "loss": 0.6854,
+      "step": 27720
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 18.75062370300293,
+      "learning_rate": 9.56145303971391e-06,
+      "loss": 0.7951,
+      "step": 27730
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 5.429780960083008,
+      "learning_rate": 9.557688688123472e-06,
+      "loss": 0.5807,
+      "step": 27740
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 6.862247943878174,
+      "learning_rate": 9.553924336533033e-06,
+      "loss": 0.4013,
+      "step": 27750
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 2.499727487564087,
+      "learning_rate": 9.550159984942595e-06,
+      "loss": 0.5839,
+      "step": 27760
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 20.708763122558594,
+      "learning_rate": 9.546395633352156e-06,
+      "loss": 0.5312,
+      "step": 27770
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 35.32229995727539,
+      "learning_rate": 9.542631281761717e-06,
+      "loss": 0.5386,
+      "step": 27780
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 4.731060028076172,
+      "learning_rate": 9.538866930171279e-06,
+      "loss": 0.3741,
+      "step": 27790
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 15.941679000854492,
+      "learning_rate": 9.53510257858084e-06,
+      "loss": 0.5086,
+      "step": 27800
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 33.2602653503418,
+      "learning_rate": 9.531338226990402e-06,
+      "loss": 0.6115,
+      "step": 27810
+    },
+    {
+      "epoch": 5.24,
+      "grad_norm": 10.409682273864746,
+      "learning_rate": 9.527573875399963e-06,
+      "loss": 0.3425,
+      "step": 27820
+    },
+    {
+      "epoch": 5.24,
+      "grad_norm": 18.776081085205078,
+      "learning_rate": 9.523809523809525e-06,
+      "loss": 0.4186,
+      "step": 27830
+    },
+    {
+      "epoch": 5.24,
+      "grad_norm": 45.04935073852539,
+      "learning_rate": 9.520045172219086e-06,
+      "loss": 0.5253,
+      "step": 27840
+    },
+    {
+      "epoch": 5.24,
+      "grad_norm": 8.127985954284668,
+      "learning_rate": 9.516280820628648e-06,
+      "loss": 0.5575,
+      "step": 27850
+    },
+    {
+      "epoch": 5.24,
+      "grad_norm": 3.796123504638672,
+      "learning_rate": 9.512516469038209e-06,
+      "loss": 0.3674,
+      "step": 27860
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 3.369060754776001,
+      "learning_rate": 9.50875211744777e-06,
+      "loss": 0.539,
+      "step": 27870
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 1.8295410871505737,
+      "learning_rate": 9.504987765857332e-06,
+      "loss": 0.3522,
+      "step": 27880
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 21.568384170532227,
+      "learning_rate": 9.501223414266894e-06,
+      "loss": 0.4866,
+      "step": 27890
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 6.660335540771484,
+      "learning_rate": 9.497459062676455e-06,
+      "loss": 0.6279,
+      "step": 27900
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 0.142924964427948,
+      "learning_rate": 9.493694711086016e-06,
+      "loss": 0.7068,
+      "step": 27910
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 0.3119409680366516,
+      "learning_rate": 9.489930359495578e-06,
+      "loss": 0.5291,
+      "step": 27920
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 41.5748405456543,
+      "learning_rate": 9.48616600790514e-06,
+      "loss": 0.6962,
+      "step": 27930
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 21.246931076049805,
+      "learning_rate": 9.4824016563147e-06,
+      "loss": 0.5247,
+      "step": 27940
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 9.162127494812012,
+      "learning_rate": 9.478637304724262e-06,
+      "loss": 0.6259,
+      "step": 27950
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 14.135149955749512,
+      "learning_rate": 9.474872953133824e-06,
+      "loss": 0.5191,
+      "step": 27960
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 14.310225486755371,
+      "learning_rate": 9.471108601543385e-06,
+      "loss": 0.6222,
+      "step": 27970
+    },
+    {
+      "epoch": 5.27,
+      "grad_norm": 23.264753341674805,
+      "learning_rate": 9.467344249952947e-06,
+      "loss": 0.8251,
+      "step": 27980
+    },
+    {
+      "epoch": 5.27,
+      "grad_norm": 18.3626766204834,
+      "learning_rate": 9.463579898362508e-06,
+      "loss": 0.6605,
+      "step": 27990
+    },
+    {
+      "epoch": 5.27,
+      "grad_norm": 1.1121855974197388,
+      "learning_rate": 9.45981554677207e-06,
+      "loss": 0.396,
+      "step": 28000
+    },
+    {
+      "epoch": 5.27,
+      "grad_norm": 0.5820768475532532,
+      "learning_rate": 9.456051195181631e-06,
+      "loss": 0.3876,
+      "step": 28010
+    },
+    {
+      "epoch": 5.27,
+      "grad_norm": 19.84071159362793,
+      "learning_rate": 9.452286843591193e-06,
+      "loss": 0.2798,
+      "step": 28020
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 14.78951644897461,
+      "learning_rate": 9.448522492000754e-06,
+      "loss": 0.7496,
+      "step": 28030
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 1.7860188484191895,
+      "learning_rate": 9.444758140410315e-06,
+      "loss": 0.3007,
+      "step": 28040
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 23.34420394897461,
+      "learning_rate": 9.440993788819877e-06,
+      "loss": 0.6097,
+      "step": 28050
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 11.595169067382812,
+      "learning_rate": 9.437229437229438e-06,
+      "loss": 0.7458,
+      "step": 28060
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 13.248278617858887,
+      "learning_rate": 9.433465085639e-06,
+      "loss": 0.4254,
+      "step": 28070
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 9.730267524719238,
+      "learning_rate": 9.429700734048561e-06,
+      "loss": 0.2484,
+      "step": 28080
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 10.137773513793945,
+      "learning_rate": 9.425936382458121e-06,
+      "loss": 0.5563,
+      "step": 28090
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 9.176275253295898,
+      "learning_rate": 9.422172030867684e-06,
+      "loss": 0.7219,
+      "step": 28100
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 18.387922286987305,
+      "learning_rate": 9.418407679277246e-06,
+      "loss": 0.4597,
+      "step": 28110
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 5.12867546081543,
+      "learning_rate": 9.414643327686807e-06,
+      "loss": 0.6968,
+      "step": 28120
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 5.446651935577393,
+      "learning_rate": 9.410878976096369e-06,
+      "loss": 0.5488,
+      "step": 28130
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 11.691588401794434,
+      "learning_rate": 9.40711462450593e-06,
+      "loss": 0.3528,
+      "step": 28140
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 5.106471538543701,
+      "learning_rate": 9.403350272915491e-06,
+      "loss": 0.5703,
+      "step": 28150
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 22.144344329833984,
+      "learning_rate": 9.399585921325053e-06,
+      "loss": 0.5555,
+      "step": 28160
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 11.630510330200195,
+      "learning_rate": 9.395821569734614e-06,
+      "loss": 0.6737,
+      "step": 28170
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 0.5806583166122437,
+      "learning_rate": 9.392057218144174e-06,
+      "loss": 0.5265,
+      "step": 28180
+    },
+    {
+      "epoch": 5.31,
+      "grad_norm": 4.3490214347839355,
+      "learning_rate": 9.388292866553737e-06,
+      "loss": 0.7698,
+      "step": 28190
+    },
+    {
+      "epoch": 5.31,
+      "grad_norm": 19.73464584350586,
+      "learning_rate": 9.384528514963299e-06,
+      "loss": 0.3166,
+      "step": 28200
+    },
+    {
+      "epoch": 5.31,
+      "grad_norm": 8.744988441467285,
+      "learning_rate": 9.38076416337286e-06,
+      "loss": 0.5361,
+      "step": 28210
+    },
+    {
+      "epoch": 5.31,
+      "grad_norm": 13.170635223388672,
+      "learning_rate": 9.376999811782422e-06,
+      "loss": 0.5091,
+      "step": 28220
+    },
+    {
+      "epoch": 5.31,
+      "grad_norm": 3.8502869606018066,
+      "learning_rate": 9.373235460191983e-06,
+      "loss": 0.6684,
+      "step": 28230
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 4.719302177429199,
+      "learning_rate": 9.369471108601545e-06,
+      "loss": 0.3989,
+      "step": 28240
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 32.28798294067383,
+      "learning_rate": 9.365706757011106e-06,
+      "loss": 0.509,
+      "step": 28250
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 17.933258056640625,
+      "learning_rate": 9.361942405420668e-06,
+      "loss": 0.5806,
+      "step": 28260
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 4.121888637542725,
+      "learning_rate": 9.358178053830227e-06,
+      "loss": 0.4916,
+      "step": 28270
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 7.130378723144531,
+      "learning_rate": 9.354413702239789e-06,
+      "loss": 0.4107,
+      "step": 28280
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 14.70533561706543,
+      "learning_rate": 9.350649350649352e-06,
+      "loss": 0.581,
+      "step": 28290
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 11.548562049865723,
+      "learning_rate": 9.346884999058913e-06,
+      "loss": 0.3762,
+      "step": 28300
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 15.485182762145996,
+      "learning_rate": 9.343120647468475e-06,
+      "loss": 0.7006,
+      "step": 28310
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 26.412918090820312,
+      "learning_rate": 9.339356295878036e-06,
+      "loss": 0.2112,
+      "step": 28320
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 13.118706703186035,
+      "learning_rate": 9.335591944287598e-06,
+      "loss": 0.3918,
+      "step": 28330
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 6.297987937927246,
+      "learning_rate": 9.33182759269716e-06,
+      "loss": 0.4208,
+      "step": 28340
+    },
+    {
+      "epoch": 5.34,
+      "grad_norm": 5.268683433532715,
+      "learning_rate": 9.32806324110672e-06,
+      "loss": 0.5968,
+      "step": 28350
+    },
+    {
+      "epoch": 5.34,
+      "grad_norm": 6.9022955894470215,
+      "learning_rate": 9.32429888951628e-06,
+      "loss": 0.5602,
+      "step": 28360
+    },
+    {
+      "epoch": 5.34,
+      "grad_norm": 1.084697961807251,
+      "learning_rate": 9.320534537925842e-06,
+      "loss": 0.9877,
+      "step": 28370
+    },
+    {
+      "epoch": 5.34,
+      "grad_norm": 18.268627166748047,
+      "learning_rate": 9.316770186335405e-06,
+      "loss": 0.5616,
+      "step": 28380
+    },
+    {
+      "epoch": 5.34,
+      "grad_norm": 11.595245361328125,
+      "learning_rate": 9.313005834744967e-06,
+      "loss": 0.3982,
+      "step": 28390
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 15.181175231933594,
+      "learning_rate": 9.309241483154528e-06,
+      "loss": 0.4555,
+      "step": 28400
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 20.354490280151367,
+      "learning_rate": 9.30547713156409e-06,
+      "loss": 0.4182,
+      "step": 28410
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 2.795022487640381,
+      "learning_rate": 9.301712779973651e-06,
+      "loss": 0.629,
+      "step": 28420
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 3.4192261695861816,
+      "learning_rate": 9.297948428383212e-06,
+      "loss": 0.6771,
+      "step": 28430
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 18.82028579711914,
+      "learning_rate": 9.294184076792772e-06,
+      "loss": 0.6471,
+      "step": 28440
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 10.711726188659668,
+      "learning_rate": 9.290419725202334e-06,
+      "loss": 0.707,
+      "step": 28450
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 18.926769256591797,
+      "learning_rate": 9.286655373611895e-06,
+      "loss": 0.499,
+      "step": 28460
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 6.421844482421875,
+      "learning_rate": 9.282891022021458e-06,
+      "loss": 0.506,
+      "step": 28470
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 30.580154418945312,
+      "learning_rate": 9.27912667043102e-06,
+      "loss": 0.4126,
+      "step": 28480
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 1.4444148540496826,
+      "learning_rate": 9.275362318840581e-06,
+      "loss": 0.321,
+      "step": 28490
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 0.1973607838153839,
+      "learning_rate": 9.271597967250143e-06,
+      "loss": 0.6311,
+      "step": 28500
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 4.095947265625,
+      "learning_rate": 9.267833615659704e-06,
+      "loss": 0.3348,
+      "step": 28510
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 3.3935463428497314,
+      "learning_rate": 9.264069264069266e-06,
+      "loss": 0.7266,
+      "step": 28520
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 1.5833213329315186,
+      "learning_rate": 9.260304912478825e-06,
+      "loss": 0.2602,
+      "step": 28530
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 32.14379119873047,
+      "learning_rate": 9.256540560888387e-06,
+      "loss": 0.6631,
+      "step": 28540
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 6.494132041931152,
+      "learning_rate": 9.252776209297948e-06,
+      "loss": 0.686,
+      "step": 28550
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 7.422737121582031,
+      "learning_rate": 9.24901185770751e-06,
+      "loss": 0.7162,
+      "step": 28560
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 21.754234313964844,
+      "learning_rate": 9.245247506117073e-06,
+      "loss": 0.5544,
+      "step": 28570
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 11.023734092712402,
+      "learning_rate": 9.241483154526634e-06,
+      "loss": 0.7434,
+      "step": 28580
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 14.888821601867676,
+      "learning_rate": 9.237718802936196e-06,
+      "loss": 0.4337,
+      "step": 28590
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 6.61778450012207,
+      "learning_rate": 9.233954451345757e-06,
+      "loss": 0.7,
+      "step": 28600
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 10.373224258422852,
+      "learning_rate": 9.230190099755319e-06,
+      "loss": 0.5232,
+      "step": 28610
+    },
+    {
+      "epoch": 5.39,
+      "grad_norm": 6.890915870666504,
+      "learning_rate": 9.226425748164878e-06,
+      "loss": 0.4981,
+      "step": 28620
+    },
+    {
+      "epoch": 5.39,
+      "grad_norm": 7.358579158782959,
+      "learning_rate": 9.22266139657444e-06,
+      "loss": 0.7168,
+      "step": 28630
+    },
+    {
+      "epoch": 5.39,
+      "grad_norm": 26.474361419677734,
+      "learning_rate": 9.218897044984001e-06,
+      "loss": 0.4434,
+      "step": 28640
+    },
+    {
+      "epoch": 5.39,
+      "grad_norm": 0.49774327874183655,
+      "learning_rate": 9.215132693393563e-06,
+      "loss": 0.4762,
+      "step": 28650
+    },
+    {
+      "epoch": 5.39,
+      "grad_norm": 0.7938456535339355,
+      "learning_rate": 9.211368341803126e-06,
+      "loss": 0.4004,
+      "step": 28660
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 16.776710510253906,
+      "learning_rate": 9.207603990212687e-06,
+      "loss": 0.6284,
+      "step": 28670
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 28.070096969604492,
+      "learning_rate": 9.203839638622249e-06,
+      "loss": 0.6981,
+      "step": 28680
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 10.797050476074219,
+      "learning_rate": 9.20007528703181e-06,
+      "loss": 0.4466,
+      "step": 28690
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 25.648643493652344,
+      "learning_rate": 9.19631093544137e-06,
+      "loss": 0.7307,
+      "step": 28700
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 20.66429901123047,
+      "learning_rate": 9.192546583850932e-06,
+      "loss": 0.5789,
+      "step": 28710
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 21.309707641601562,
+      "learning_rate": 9.188782232260493e-06,
+      "loss": 0.3682,
+      "step": 28720
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 7.1550703048706055,
+      "learning_rate": 9.185017880670055e-06,
+      "loss": 0.3696,
+      "step": 28730
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 19.179853439331055,
+      "learning_rate": 9.181253529079616e-06,
+      "loss": 0.4966,
+      "step": 28740
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 12.53691291809082,
+      "learning_rate": 9.177489177489179e-06,
+      "loss": 0.4346,
+      "step": 28750
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 17.41402244567871,
+      "learning_rate": 9.17372482589874e-06,
+      "loss": 0.469,
+      "step": 28760
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 9.987686157226562,
+      "learning_rate": 9.169960474308302e-06,
+      "loss": 0.4145,
+      "step": 28770
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 9.735391616821289,
+      "learning_rate": 9.166196122717864e-06,
+      "loss": 0.4125,
+      "step": 28780
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 40.975223541259766,
+      "learning_rate": 9.162431771127423e-06,
+      "loss": 0.6056,
+      "step": 28790
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 22.280542373657227,
+      "learning_rate": 9.158667419536985e-06,
+      "loss": 0.6244,
+      "step": 28800
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 8.667016983032227,
+      "learning_rate": 9.154903067946546e-06,
+      "loss": 0.3564,
+      "step": 28810
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 0.26233333349227905,
+      "learning_rate": 9.151138716356108e-06,
+      "loss": 0.5011,
+      "step": 28820
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 8.251745223999023,
+      "learning_rate": 9.147374364765669e-06,
+      "loss": 0.5049,
+      "step": 28830
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 18.818889617919922,
+      "learning_rate": 9.143610013175232e-06,
+      "loss": 0.3172,
+      "step": 28840
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 26.497833251953125,
+      "learning_rate": 9.139845661584794e-06,
+      "loss": 0.5095,
+      "step": 28850
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 8.361761093139648,
+      "learning_rate": 9.136081309994355e-06,
+      "loss": 0.5354,
+      "step": 28860
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 17.610157012939453,
+      "learning_rate": 9.132316958403917e-06,
+      "loss": 0.6836,
+      "step": 28870
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 7.032655715942383,
+      "learning_rate": 9.128552606813476e-06,
+      "loss": 0.5749,
+      "step": 28880
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 18.669261932373047,
+      "learning_rate": 9.124788255223038e-06,
+      "loss": 0.5949,
+      "step": 28890
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 12.058159828186035,
+      "learning_rate": 9.1210239036326e-06,
+      "loss": 0.5494,
+      "step": 28900
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 0.7247989773750305,
+      "learning_rate": 9.11725955204216e-06,
+      "loss": 0.4002,
+      "step": 28910
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 6.094273090362549,
+      "learning_rate": 9.113495200451722e-06,
+      "loss": 0.5376,
+      "step": 28920
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 17.795515060424805,
+      "learning_rate": 9.109730848861284e-06,
+      "loss": 0.5065,
+      "step": 28930
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 13.030588150024414,
+      "learning_rate": 9.105966497270847e-06,
+      "loss": 0.5984,
+      "step": 28940
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 0.7247048020362854,
+      "learning_rate": 9.102202145680408e-06,
+      "loss": 0.6524,
+      "step": 28950
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 0.13299456238746643,
+      "learning_rate": 9.09843779408997e-06,
+      "loss": 0.616,
+      "step": 28960
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 5.874953746795654,
+      "learning_rate": 9.09467344249953e-06,
+      "loss": 0.6393,
+      "step": 28970
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 14.227991104125977,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.3868,
+      "step": 28980
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 8.069024085998535,
+      "learning_rate": 9.087144739318652e-06,
+      "loss": 0.6467,
+      "step": 28990
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 12.728198051452637,
+      "learning_rate": 9.083380387728214e-06,
+      "loss": 0.8101,
+      "step": 29000
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 31.68919563293457,
+      "learning_rate": 9.079616036137775e-06,
+      "loss": 0.4018,
+      "step": 29010
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 37.730751037597656,
+      "learning_rate": 9.075851684547337e-06,
+      "loss": 0.3918,
+      "step": 29020
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 32.44590759277344,
+      "learning_rate": 9.0720873329569e-06,
+      "loss": 0.5716,
+      "step": 29030
+    },
+    {
+      "epoch": 5.47,
+      "grad_norm": 3.4160537719726562,
+      "learning_rate": 9.068322981366461e-06,
+      "loss": 0.6425,
+      "step": 29040
+    },
+    {
+      "epoch": 5.47,
+      "grad_norm": 15.794535636901855,
+      "learning_rate": 9.064558629776021e-06,
+      "loss": 0.4578,
+      "step": 29050
+    },
+    {
+      "epoch": 5.47,
+      "grad_norm": 14.576927185058594,
+      "learning_rate": 9.060794278185583e-06,
+      "loss": 0.2114,
+      "step": 29060
+    },
+    {
+      "epoch": 5.47,
+      "grad_norm": 12.429288864135742,
+      "learning_rate": 9.057029926595144e-06,
+      "loss": 0.571,
+      "step": 29070
+    },
+    {
+      "epoch": 5.47,
+      "grad_norm": 12.511153221130371,
+      "learning_rate": 9.053265575004706e-06,
+      "loss": 0.5288,
+      "step": 29080
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 11.03250789642334,
+      "learning_rate": 9.049501223414267e-06,
+      "loss": 0.6037,
+      "step": 29090
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 14.617107391357422,
+      "learning_rate": 9.045736871823829e-06,
+      "loss": 0.7927,
+      "step": 29100
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 10.177559852600098,
+      "learning_rate": 9.04197252023339e-06,
+      "loss": 0.5179,
+      "step": 29110
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 1.5865446329116821,
+      "learning_rate": 9.038208168642953e-06,
+      "loss": 0.6091,
+      "step": 29120
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 18.313032150268555,
+      "learning_rate": 9.034443817052515e-06,
+      "loss": 0.4665,
+      "step": 29130
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 23.01014518737793,
+      "learning_rate": 9.030679465462074e-06,
+      "loss": 0.6638,
+      "step": 29140
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 18.016878128051758,
+      "learning_rate": 9.026915113871636e-06,
+      "loss": 0.437,
+      "step": 29150
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 19.193099975585938,
+      "learning_rate": 9.023150762281197e-06,
+      "loss": 0.864,
+      "step": 29160
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 22.776002883911133,
+      "learning_rate": 9.019386410690759e-06,
+      "loss": 0.3799,
+      "step": 29170
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 19.30347442626953,
+      "learning_rate": 9.01562205910032e-06,
+      "loss": 0.4335,
+      "step": 29180
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 8.978057861328125,
+      "learning_rate": 9.011857707509882e-06,
+      "loss": 0.6437,
+      "step": 29190
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 24.524974822998047,
+      "learning_rate": 9.008093355919443e-06,
+      "loss": 0.6654,
+      "step": 29200
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 29.421937942504883,
+      "learning_rate": 9.004329004329005e-06,
+      "loss": 1.0022,
+      "step": 29210
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 9.427165985107422,
+      "learning_rate": 9.000564652738568e-06,
+      "loss": 0.618,
+      "step": 29220
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 3.5719804763793945,
+      "learning_rate": 8.996800301148128e-06,
+      "loss": 0.4899,
+      "step": 29230
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 4.902514934539795,
+      "learning_rate": 8.993035949557689e-06,
+      "loss": 0.6078,
+      "step": 29240
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 7.195416450500488,
+      "learning_rate": 8.98927159796725e-06,
+      "loss": 0.5264,
+      "step": 29250
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 5.761614799499512,
+      "learning_rate": 8.985507246376812e-06,
+      "loss": 0.5512,
+      "step": 29260
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 20.03443145751953,
+      "learning_rate": 8.981742894786373e-06,
+      "loss": 0.3887,
+      "step": 29270
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 20.721759796142578,
+      "learning_rate": 8.977978543195935e-06,
+      "loss": 0.4945,
+      "step": 29280
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 3.0783581733703613,
+      "learning_rate": 8.974214191605496e-06,
+      "loss": 0.5722,
+      "step": 29290
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 5.727044105529785,
+      "learning_rate": 8.970449840015058e-06,
+      "loss": 0.9616,
+      "step": 29300
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 12.99439811706543,
+      "learning_rate": 8.96668548842462e-06,
+      "loss": 0.282,
+      "step": 29310
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 22.195281982421875,
+      "learning_rate": 8.96292113683418e-06,
+      "loss": 0.8487,
+      "step": 29320
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 16.877878189086914,
+      "learning_rate": 8.959156785243742e-06,
+      "loss": 0.4369,
+      "step": 29330
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 3.1448419094085693,
+      "learning_rate": 8.955392433653304e-06,
+      "loss": 0.4618,
+      "step": 29340
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 11.687005996704102,
+      "learning_rate": 8.951628082062865e-06,
+      "loss": 0.5836,
+      "step": 29350
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 52.1478385925293,
+      "learning_rate": 8.947863730472427e-06,
+      "loss": 0.8731,
+      "step": 29360
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 8.80933952331543,
+      "learning_rate": 8.944099378881988e-06,
+      "loss": 0.6717,
+      "step": 29370
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 18.5314884185791,
+      "learning_rate": 8.94033502729155e-06,
+      "loss": 0.4894,
+      "step": 29380
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 11.329646110534668,
+      "learning_rate": 8.936570675701111e-06,
+      "loss": 0.4815,
+      "step": 29390
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 21.230852127075195,
+      "learning_rate": 8.932806324110672e-06,
+      "loss": 0.4756,
+      "step": 29400
+    },
+    {
+      "epoch": 5.54,
+      "grad_norm": 8.987369537353516,
+      "learning_rate": 8.929041972520234e-06,
+      "loss": 0.553,
+      "step": 29410
+    },
+    {
+      "epoch": 5.54,
+      "grad_norm": 9.444402694702148,
+      "learning_rate": 8.925277620929795e-06,
+      "loss": 0.4838,
+      "step": 29420
+    },
+    {
+      "epoch": 5.54,
+      "grad_norm": 0.40830114483833313,
+      "learning_rate": 8.921513269339357e-06,
+      "loss": 0.7624,
+      "step": 29430
+    },
+    {
+      "epoch": 5.54,
+      "grad_norm": 0.5162389874458313,
+      "learning_rate": 8.917748917748918e-06,
+      "loss": 0.2015,
+      "step": 29440
+    },
+    {
+      "epoch": 5.54,
+      "grad_norm": 37.219818115234375,
+      "learning_rate": 8.91398456615848e-06,
+      "loss": 0.5263,
+      "step": 29450
+    },
+    {
+      "epoch": 5.54,
+      "grad_norm": 8.195032119750977,
+      "learning_rate": 8.910220214568041e-06,
+      "loss": 0.6158,
+      "step": 29460
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 5.147323131561279,
+      "learning_rate": 8.906455862977603e-06,
+      "loss": 0.5521,
+      "step": 29470
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 3.7888503074645996,
+      "learning_rate": 8.902691511387164e-06,
+      "loss": 0.357,
+      "step": 29480
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 6.756317138671875,
+      "learning_rate": 8.898927159796726e-06,
+      "loss": 0.4016,
+      "step": 29490
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 26.7462158203125,
+      "learning_rate": 8.895162808206287e-06,
+      "loss": 0.6225,
+      "step": 29500
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 3.3406825065612793,
+      "learning_rate": 8.891398456615848e-06,
+      "loss": 0.5504,
+      "step": 29510
+    },
+    {
+      "epoch": 5.56,
+      "grad_norm": 7.413740634918213,
+      "learning_rate": 8.88763410502541e-06,
+      "loss": 0.4144,
+      "step": 29520
+    },
+    {
+      "epoch": 5.56,
+      "grad_norm": 0.5586540699005127,
+      "learning_rate": 8.883869753434971e-06,
+      "loss": 0.394,
+      "step": 29530
+    },
+    {
+      "epoch": 5.56,
+      "grad_norm": 28.930585861206055,
+      "learning_rate": 8.880105401844533e-06,
+      "loss": 0.436,
+      "step": 29540
+    },
+    {
+      "epoch": 5.56,
+      "grad_norm": 4.1949849128723145,
+      "learning_rate": 8.876341050254094e-06,
+      "loss": 0.7621,
+      "step": 29550
+    },
+    {
+      "epoch": 5.56,
+      "grad_norm": 15.377513885498047,
+      "learning_rate": 8.872576698663656e-06,
+      "loss": 0.3309,
+      "step": 29560
+    },
+    {
+      "epoch": 5.57,
+      "grad_norm": 10.37800121307373,
+      "learning_rate": 8.868812347073217e-06,
+      "loss": 0.5609,
+      "step": 29570
+    },
+    {
+      "epoch": 5.57,
+      "grad_norm": 16.737653732299805,
+      "learning_rate": 8.865047995482779e-06,
+      "loss": 0.5195,
+      "step": 29580
+    },
+    {
+      "epoch": 5.57,
+      "grad_norm": 10.057021141052246,
+      "learning_rate": 8.86128364389234e-06,
+      "loss": 0.5856,
+      "step": 29590
+    },
+    {
+      "epoch": 5.57,
+      "grad_norm": 18.19561195373535,
+      "learning_rate": 8.857519292301902e-06,
+      "loss": 0.6032,
+      "step": 29600
+    },
+    {
+      "epoch": 5.57,
+      "grad_norm": 8.191587448120117,
+      "learning_rate": 8.853754940711463e-06,
+      "loss": 0.3631,
+      "step": 29610
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 14.722908973693848,
+      "learning_rate": 8.849990589121025e-06,
+      "loss": 0.9588,
+      "step": 29620
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 25.523221969604492,
+      "learning_rate": 8.846226237530586e-06,
+      "loss": 0.7066,
+      "step": 29630
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 9.714911460876465,
+      "learning_rate": 8.842461885940147e-06,
+      "loss": 0.4797,
+      "step": 29640
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 5.88856315612793,
+      "learning_rate": 8.838697534349709e-06,
+      "loss": 0.4359,
+      "step": 29650
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 12.061199188232422,
+      "learning_rate": 8.83493318275927e-06,
+      "loss": 0.5571,
+      "step": 29660
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 18.20134162902832,
+      "learning_rate": 8.831168831168832e-06,
+      "loss": 0.3449,
+      "step": 29670
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 12.034881591796875,
+      "learning_rate": 8.827404479578393e-06,
+      "loss": 0.7451,
+      "step": 29680
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 4.537769317626953,
+      "learning_rate": 8.823640127987955e-06,
+      "loss": 0.7454,
+      "step": 29690
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 39.31631088256836,
+      "learning_rate": 8.819875776397516e-06,
+      "loss": 0.3897,
+      "step": 29700
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 20.764333724975586,
+      "learning_rate": 8.816111424807078e-06,
+      "loss": 0.4239,
+      "step": 29710
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 4.3475823402404785,
+      "learning_rate": 8.812347073216639e-06,
+      "loss": 0.4626,
+      "step": 29720
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 26.93927001953125,
+      "learning_rate": 8.8085827216262e-06,
+      "loss": 0.3852,
+      "step": 29730
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 23.00481414794922,
+      "learning_rate": 8.804818370035762e-06,
+      "loss": 0.5371,
+      "step": 29740
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 1.6085634231567383,
+      "learning_rate": 8.801054018445324e-06,
+      "loss": 0.3698,
+      "step": 29750
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 13.103279113769531,
+      "learning_rate": 8.797289666854885e-06,
+      "loss": 0.6233,
+      "step": 29760
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 14.591599464416504,
+      "learning_rate": 8.793525315264446e-06,
+      "loss": 0.4826,
+      "step": 29770
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 23.947324752807617,
+      "learning_rate": 8.789760963674008e-06,
+      "loss": 0.5319,
+      "step": 29780
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 11.421167373657227,
+      "learning_rate": 8.78599661208357e-06,
+      "loss": 0.5641,
+      "step": 29790
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 4.573395252227783,
+      "learning_rate": 8.78223226049313e-06,
+      "loss": 0.6296,
+      "step": 29800
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 30.208024978637695,
+      "learning_rate": 8.778467908902692e-06,
+      "loss": 0.4905,
+      "step": 29810
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 33.18610763549805,
+      "learning_rate": 8.774703557312254e-06,
+      "loss": 0.7144,
+      "step": 29820
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 34.4425048828125,
+      "learning_rate": 8.770939205721815e-06,
+      "loss": 0.8056,
+      "step": 29830
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 9.860750198364258,
+      "learning_rate": 8.767174854131377e-06,
+      "loss": 0.4445,
+      "step": 29840
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 10.247167587280273,
+      "learning_rate": 8.763410502540938e-06,
+      "loss": 0.472,
+      "step": 29850
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 30.189830780029297,
+      "learning_rate": 8.7596461509505e-06,
+      "loss": 0.4546,
+      "step": 29860
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 4.302022933959961,
+      "learning_rate": 8.755881799360061e-06,
+      "loss": 0.6976,
+      "step": 29870
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 0.30874672532081604,
+      "learning_rate": 8.752117447769623e-06,
+      "loss": 0.711,
+      "step": 29880
+    },
+    {
+      "epoch": 5.63,
+      "grad_norm": 5.02794075012207,
+      "learning_rate": 8.748353096179184e-06,
+      "loss": 0.6327,
+      "step": 29890
+    },
+    {
+      "epoch": 5.63,
+      "grad_norm": 1.8799618482589722,
+      "learning_rate": 8.744588744588745e-06,
+      "loss": 0.3782,
+      "step": 29900
+    },
+    {
+      "epoch": 5.63,
+      "grad_norm": 12.888080596923828,
+      "learning_rate": 8.740824392998307e-06,
+      "loss": 0.4036,
+      "step": 29910
+    },
+    {
+      "epoch": 5.63,
+      "grad_norm": 24.87213897705078,
+      "learning_rate": 8.737060041407868e-06,
+      "loss": 0.51,
+      "step": 29920
+    },
+    {
+      "epoch": 5.63,
+      "grad_norm": 20.8959903717041,
+      "learning_rate": 8.73329568981743e-06,
+      "loss": 0.434,
+      "step": 29930
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 15.75831413269043,
+      "learning_rate": 8.729531338226991e-06,
+      "loss": 0.5039,
+      "step": 29940
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 6.654936790466309,
+      "learning_rate": 8.725766986636553e-06,
+      "loss": 0.6428,
+      "step": 29950
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 12.773100852966309,
+      "learning_rate": 8.722002635046114e-06,
+      "loss": 0.5853,
+      "step": 29960
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 19.850297927856445,
+      "learning_rate": 8.718238283455676e-06,
+      "loss": 0.7984,
+      "step": 29970
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 11.2599458694458,
+      "learning_rate": 8.714473931865237e-06,
+      "loss": 0.7433,
+      "step": 29980
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 14.740336418151855,
+      "learning_rate": 8.710709580274799e-06,
+      "loss": 0.448,
+      "step": 29990
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 11.278069496154785,
+      "learning_rate": 8.70694522868436e-06,
+      "loss": 0.5628,
+      "step": 30000
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 17.312114715576172,
+      "learning_rate": 8.703180877093921e-06,
+      "loss": 0.8434,
+      "step": 30010
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 10.707963943481445,
+      "learning_rate": 8.699416525503483e-06,
+      "loss": 0.5479,
+      "step": 30020
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 33.91600036621094,
+      "learning_rate": 8.695652173913044e-06,
+      "loss": 1.1438,
+      "step": 30030
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 26.10776138305664,
+      "learning_rate": 8.691887822322606e-06,
+      "loss": 0.6822,
+      "step": 30040
+    },
+    {
+      "epoch": 5.66,
+      "grad_norm": 4.60282564163208,
+      "learning_rate": 8.688123470732167e-06,
+      "loss": 0.5197,
+      "step": 30050
+    },
+    {
+      "epoch": 5.66,
+      "grad_norm": 1.051890254020691,
+      "learning_rate": 8.684359119141729e-06,
+      "loss": 0.6023,
+      "step": 30060
+    },
+    {
+      "epoch": 5.66,
+      "grad_norm": 20.802509307861328,
+      "learning_rate": 8.68059476755129e-06,
+      "loss": 0.3724,
+      "step": 30070
+    },
+    {
+      "epoch": 5.66,
+      "grad_norm": 11.538596153259277,
+      "learning_rate": 8.676830415960852e-06,
+      "loss": 0.5301,
+      "step": 30080
+    },
+    {
+      "epoch": 5.66,
+      "grad_norm": 31.12389373779297,
+      "learning_rate": 8.673066064370413e-06,
+      "loss": 0.6386,
+      "step": 30090
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 17.99003028869629,
+      "learning_rate": 8.669301712779975e-06,
+      "loss": 0.4503,
+      "step": 30100
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 25.43760108947754,
+      "learning_rate": 8.665537361189536e-06,
+      "loss": 0.6649,
+      "step": 30110
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 0.21140538156032562,
+      "learning_rate": 8.661773009599098e-06,
+      "loss": 0.3958,
+      "step": 30120
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 6.105637550354004,
+      "learning_rate": 8.658008658008659e-06,
+      "loss": 0.7797,
+      "step": 30130
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 3.348720073699951,
+      "learning_rate": 8.65424430641822e-06,
+      "loss": 0.3967,
+      "step": 30140
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 17.45372772216797,
+      "learning_rate": 8.650479954827782e-06,
+      "loss": 0.4381,
+      "step": 30150
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 5.134333610534668,
+      "learning_rate": 8.646715603237343e-06,
+      "loss": 0.6194,
+      "step": 30160
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 28.561365127563477,
+      "learning_rate": 8.642951251646905e-06,
+      "loss": 0.8043,
+      "step": 30170
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 9.468137741088867,
+      "learning_rate": 8.639186900056465e-06,
+      "loss": 0.5759,
+      "step": 30180
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 5.760190010070801,
+      "learning_rate": 8.635422548466028e-06,
+      "loss": 0.6533,
+      "step": 30190
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 22.225126266479492,
+      "learning_rate": 8.63165819687559e-06,
+      "loss": 0.7,
+      "step": 30200
+    },
+    {
+      "epoch": 5.69,
+      "grad_norm": 14.245889663696289,
+      "learning_rate": 8.62789384528515e-06,
+      "loss": 0.6829,
+      "step": 30210
+    },
+    {
+      "epoch": 5.69,
+      "grad_norm": 30.38525390625,
+      "learning_rate": 8.624129493694712e-06,
+      "loss": 0.5493,
+      "step": 30220
+    },
+    {
+      "epoch": 5.69,
+      "grad_norm": 16.88604164123535,
+      "learning_rate": 8.620365142104274e-06,
+      "loss": 0.6734,
+      "step": 30230
+    },
+    {
+      "epoch": 5.69,
+      "grad_norm": 23.054962158203125,
+      "learning_rate": 8.616600790513835e-06,
+      "loss": 0.6553,
+      "step": 30240
+    },
+    {
+      "epoch": 5.69,
+      "grad_norm": 16.953598022460938,
+      "learning_rate": 8.612836438923397e-06,
+      "loss": 0.6512,
+      "step": 30250
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 22.24566078186035,
+      "learning_rate": 8.609072087332958e-06,
+      "loss": 0.5785,
+      "step": 30260
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 0.3672504723072052,
+      "learning_rate": 8.605307735742518e-06,
+      "loss": 0.6382,
+      "step": 30270
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 0.8988174796104431,
+      "learning_rate": 8.601543384152081e-06,
+      "loss": 0.4421,
+      "step": 30280
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 15.05488395690918,
+      "learning_rate": 8.597779032561642e-06,
+      "loss": 0.4611,
+      "step": 30290
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 24.66227912902832,
+      "learning_rate": 8.594014680971204e-06,
+      "loss": 0.5693,
+      "step": 30300
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 24.09734344482422,
+      "learning_rate": 8.590250329380765e-06,
+      "loss": 0.4477,
+      "step": 30310
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 9.480891227722168,
+      "learning_rate": 8.586485977790327e-06,
+      "loss": 0.5086,
+      "step": 30320
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 41.71274185180664,
+      "learning_rate": 8.582721626199888e-06,
+      "loss": 0.8377,
+      "step": 30330
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 30.149093627929688,
+      "learning_rate": 8.57895727460945e-06,
+      "loss": 0.6889,
+      "step": 30340
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 14.241212844848633,
+      "learning_rate": 8.575192923019011e-06,
+      "loss": 0.6828,
+      "step": 30350
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 8.023870468139648,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 0.5508,
+      "step": 30360
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 1.6237149238586426,
+      "learning_rate": 8.567664219838134e-06,
+      "loss": 0.3401,
+      "step": 30370
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 7.518566608428955,
+      "learning_rate": 8.563899868247696e-06,
+      "loss": 0.4684,
+      "step": 30380
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 15.331060409545898,
+      "learning_rate": 8.560135516657257e-06,
+      "loss": 0.5097,
+      "step": 30390
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 6.36647891998291,
+      "learning_rate": 8.556371165066818e-06,
+      "loss": 0.5214,
+      "step": 30400
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 17.332448959350586,
+      "learning_rate": 8.55260681347638e-06,
+      "loss": 0.4498,
+      "step": 30410
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 63.40143966674805,
+      "learning_rate": 8.548842461885941e-06,
+      "loss": 0.6639,
+      "step": 30420
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 9.465449333190918,
+      "learning_rate": 8.545078110295503e-06,
+      "loss": 0.5251,
+      "step": 30430
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 11.202866554260254,
+      "learning_rate": 8.541313758705064e-06,
+      "loss": 0.4721,
+      "step": 30440
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 66.58721923828125,
+      "learning_rate": 8.537549407114624e-06,
+      "loss": 0.6034,
+      "step": 30450
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 36.35587692260742,
+      "learning_rate": 8.533785055524186e-06,
+      "loss": 0.5666,
+      "step": 30460
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 0.15961378812789917,
+      "learning_rate": 8.530020703933749e-06,
+      "loss": 0.7364,
+      "step": 30470
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 6.215275287628174,
+      "learning_rate": 8.52625635234331e-06,
+      "loss": 0.4973,
+      "step": 30480
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 29.853361129760742,
+      "learning_rate": 8.522492000752872e-06,
+      "loss": 0.6356,
+      "step": 30490
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 18.131460189819336,
+      "learning_rate": 8.518727649162433e-06,
+      "loss": 0.7122,
+      "step": 30500
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 17.9971866607666,
+      "learning_rate": 8.514963297571995e-06,
+      "loss": 0.784,
+      "step": 30510
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 21.481882095336914,
+      "learning_rate": 8.511198945981556e-06,
+      "loss": 0.6354,
+      "step": 30520
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 7.704098224639893,
+      "learning_rate": 8.507434594391116e-06,
+      "loss": 0.5242,
+      "step": 30530
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 0.10782869905233383,
+      "learning_rate": 8.503670242800677e-06,
+      "loss": 0.3119,
+      "step": 30540
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 9.13038158416748,
+      "learning_rate": 8.499905891210239e-06,
+      "loss": 0.5088,
+      "step": 30550
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 1.3244730234146118,
+      "learning_rate": 8.496141539619802e-06,
+      "loss": 0.3711,
+      "step": 30560
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 12.541553497314453,
+      "learning_rate": 8.492377188029363e-06,
+      "loss": 0.4415,
+      "step": 30570
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 6.784567832946777,
+      "learning_rate": 8.488612836438925e-06,
+      "loss": 0.3377,
+      "step": 30580
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 29.57703971862793,
+      "learning_rate": 8.484848484848486e-06,
+      "loss": 0.5329,
+      "step": 30590
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 16.579954147338867,
+      "learning_rate": 8.481084133258048e-06,
+      "loss": 0.4068,
+      "step": 30600
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 13.801152229309082,
+      "learning_rate": 8.477319781667609e-06,
+      "loss": 0.5752,
+      "step": 30610
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 4.930255889892578,
+      "learning_rate": 8.473555430077169e-06,
+      "loss": 0.6852,
+      "step": 30620
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 6.195820331573486,
+      "learning_rate": 8.46979107848673e-06,
+      "loss": 0.4997,
+      "step": 30630
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 14.934310913085938,
+      "learning_rate": 8.466026726896292e-06,
+      "loss": 0.4644,
+      "step": 30640
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 13.503046989440918,
+      "learning_rate": 8.462262375305855e-06,
+      "loss": 0.6239,
+      "step": 30650
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 7.415577411651611,
+      "learning_rate": 8.458498023715416e-06,
+      "loss": 0.4476,
+      "step": 30660
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 11.47219181060791,
+      "learning_rate": 8.454733672124978e-06,
+      "loss": 0.4326,
+      "step": 30670
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 31.040847778320312,
+      "learning_rate": 8.45096932053454e-06,
+      "loss": 0.5231,
+      "step": 30680
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 14.181180953979492,
+      "learning_rate": 8.4472049689441e-06,
+      "loss": 0.6028,
+      "step": 30690
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 14.366339683532715,
+      "learning_rate": 8.443440617353662e-06,
+      "loss": 0.5395,
+      "step": 30700
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 1.154944896697998,
+      "learning_rate": 8.439676265763222e-06,
+      "loss": 0.6325,
+      "step": 30710
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 5.166085243225098,
+      "learning_rate": 8.435911914172784e-06,
+      "loss": 0.5345,
+      "step": 30720
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 2.16483211517334,
+      "learning_rate": 8.432147562582345e-06,
+      "loss": 0.4621,
+      "step": 30730
+    },
+    {
+      "epoch": 5.79,
+      "grad_norm": 24.309677124023438,
+      "learning_rate": 8.428383210991906e-06,
+      "loss": 0.4354,
+      "step": 30740
+    },
+    {
+      "epoch": 5.79,
+      "grad_norm": 30.56849479675293,
+      "learning_rate": 8.42461885940147e-06,
+      "loss": 0.6885,
+      "step": 30750
+    },
+    {
+      "epoch": 5.79,
+      "grad_norm": 30.05738067626953,
+      "learning_rate": 8.420854507811031e-06,
+      "loss": 0.3176,
+      "step": 30760
+    },
+    {
+      "epoch": 5.79,
+      "grad_norm": 1.265407919883728,
+      "learning_rate": 8.417090156220593e-06,
+      "loss": 0.415,
+      "step": 30770
+    },
+    {
+      "epoch": 5.79,
+      "grad_norm": 1.8372256755828857,
+      "learning_rate": 8.413325804630154e-06,
+      "loss": 0.3897,
+      "step": 30780
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 5.470216274261475,
+      "learning_rate": 8.409561453039714e-06,
+      "loss": 0.7885,
+      "step": 30790
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 28.494230270385742,
+      "learning_rate": 8.405797101449275e-06,
+      "loss": 0.5577,
+      "step": 30800
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 3.192377805709839,
+      "learning_rate": 8.402032749858837e-06,
+      "loss": 0.6976,
+      "step": 30810
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 16.5463924407959,
+      "learning_rate": 8.398268398268398e-06,
+      "loss": 0.4397,
+      "step": 30820
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 28.271276473999023,
+      "learning_rate": 8.39450404667796e-06,
+      "loss": 0.8356,
+      "step": 30830
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 1.6185463666915894,
+      "learning_rate": 8.390739695087523e-06,
+      "loss": 0.5331,
+      "step": 30840
+    },
+    {
+      "epoch": 5.81,
+      "grad_norm": 27.62196159362793,
+      "learning_rate": 8.386975343497084e-06,
+      "loss": 0.4344,
+      "step": 30850
+    },
+    {
+      "epoch": 5.81,
+      "grad_norm": 47.22684097290039,
+      "learning_rate": 8.383210991906646e-06,
+      "loss": 0.7706,
+      "step": 30860
+    },
+    {
+      "epoch": 5.81,
+      "grad_norm": 9.242682456970215,
+      "learning_rate": 8.379446640316207e-06,
+      "loss": 0.6522,
+      "step": 30870
+    },
+    {
+      "epoch": 5.81,
+      "grad_norm": 9.148713111877441,
+      "learning_rate": 8.375682288725767e-06,
+      "loss": 0.4746,
+      "step": 30880
+    },
+    {
+      "epoch": 5.81,
+      "grad_norm": 24.679914474487305,
+      "learning_rate": 8.371917937135328e-06,
+      "loss": 0.6837,
+      "step": 30890
+    },
+    {
+      "epoch": 5.82,
+      "grad_norm": 14.974303245544434,
+      "learning_rate": 8.36815358554489e-06,
+      "loss": 0.6234,
+      "step": 30900
+    },
+    {
+      "epoch": 5.82,
+      "grad_norm": 17.962249755859375,
+      "learning_rate": 8.364389233954451e-06,
+      "loss": 0.507,
+      "step": 30910
+    },
+    {
+      "epoch": 5.82,
+      "grad_norm": 10.956331253051758,
+      "learning_rate": 8.360624882364013e-06,
+      "loss": 0.5623,
+      "step": 30920
+    },
+    {
+      "epoch": 5.82,
+      "grad_norm": 0.729061484336853,
+      "learning_rate": 8.356860530773576e-06,
+      "loss": 0.6084,
+      "step": 30930
+    },
+    {
+      "epoch": 5.82,
+      "grad_norm": 12.048168182373047,
+      "learning_rate": 8.353096179183137e-06,
+      "loss": 0.3989,
+      "step": 30940
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 7.992116928100586,
+      "learning_rate": 8.349331827592699e-06,
+      "loss": 0.4077,
+      "step": 30950
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 19.040918350219727,
+      "learning_rate": 8.34556747600226e-06,
+      "loss": 0.9477,
+      "step": 30960
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 12.26267147064209,
+      "learning_rate": 8.34180312441182e-06,
+      "loss": 0.5282,
+      "step": 30970
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 7.242198467254639,
+      "learning_rate": 8.338038772821381e-06,
+      "loss": 0.4611,
+      "step": 30980
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 27.804534912109375,
+      "learning_rate": 8.334274421230943e-06,
+      "loss": 0.5615,
+      "step": 30990
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 0.4633367955684662,
+      "learning_rate": 8.330510069640504e-06,
+      "loss": 0.533,
+      "step": 31000
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 7.059324741363525,
+      "learning_rate": 8.326745718050066e-06,
+      "loss": 0.1811,
+      "step": 31010
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 13.364731788635254,
+      "learning_rate": 8.322981366459629e-06,
+      "loss": 0.4192,
+      "step": 31020
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 10.441876411437988,
+      "learning_rate": 8.31921701486919e-06,
+      "loss": 0.4543,
+      "step": 31030
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 1.0615036487579346,
+      "learning_rate": 8.315452663278752e-06,
+      "loss": 0.4342,
+      "step": 31040
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 0.598838210105896,
+      "learning_rate": 8.311688311688313e-06,
+      "loss": 0.6021,
+      "step": 31050
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 32.29551315307617,
+      "learning_rate": 8.307923960097873e-06,
+      "loss": 0.5029,
+      "step": 31060
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 42.38307571411133,
+      "learning_rate": 8.304159608507435e-06,
+      "loss": 0.4835,
+      "step": 31070
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 0.41502639651298523,
+      "learning_rate": 8.300395256916996e-06,
+      "loss": 0.2942,
+      "step": 31080
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 1.5998517274856567,
+      "learning_rate": 8.296630905326558e-06,
+      "loss": 0.4603,
+      "step": 31090
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 18.18465232849121,
+      "learning_rate": 8.292866553736119e-06,
+      "loss": 0.5552,
+      "step": 31100
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 19.617231369018555,
+      "learning_rate": 8.28910220214568e-06,
+      "loss": 0.5419,
+      "step": 31110
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 33.43571472167969,
+      "learning_rate": 8.285337850555244e-06,
+      "loss": 0.3924,
+      "step": 31120
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 14.464301109313965,
+      "learning_rate": 8.281573498964805e-06,
+      "loss": 0.5375,
+      "step": 31130
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 15.446751594543457,
+      "learning_rate": 8.277809147374365e-06,
+      "loss": 0.2812,
+      "step": 31140
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 24.76715087890625,
+      "learning_rate": 8.274044795783926e-06,
+      "loss": 0.4512,
+      "step": 31150
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 19.724828720092773,
+      "learning_rate": 8.270280444193488e-06,
+      "loss": 0.5171,
+      "step": 31160
+    },
+    {
+      "epoch": 5.87,
+      "grad_norm": 6.281129837036133,
+      "learning_rate": 8.26651609260305e-06,
+      "loss": 0.3646,
+      "step": 31170
+    },
+    {
+      "epoch": 5.87,
+      "grad_norm": 0.5755620002746582,
+      "learning_rate": 8.26275174101261e-06,
+      "loss": 0.6023,
+      "step": 31180
+    },
+    {
+      "epoch": 5.87,
+      "grad_norm": 24.553237915039062,
+      "learning_rate": 8.258987389422172e-06,
+      "loss": 0.4937,
+      "step": 31190
+    },
+    {
+      "epoch": 5.87,
+      "grad_norm": 25.367462158203125,
+      "learning_rate": 8.255223037831734e-06,
+      "loss": 0.4216,
+      "step": 31200
+    },
+    {
+      "epoch": 5.87,
+      "grad_norm": 5.50634765625,
+      "learning_rate": 8.251458686241297e-06,
+      "loss": 0.3745,
+      "step": 31210
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": 31.35790252685547,
+      "learning_rate": 8.247694334650858e-06,
+      "loss": 0.6542,
+      "step": 31220
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": 18.49880027770996,
+      "learning_rate": 8.243929983060418e-06,
+      "loss": 1.0202,
+      "step": 31230
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": 0.294521301984787,
+      "learning_rate": 8.24016563146998e-06,
+      "loss": 0.2461,
+      "step": 31240
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": 2.930119276046753,
+      "learning_rate": 8.236401279879541e-06,
+      "loss": 0.4741,
+      "step": 31250
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": 14.062408447265625,
+      "learning_rate": 8.232636928289102e-06,
+      "loss": 0.5901,
+      "step": 31260
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 1.9135679006576538,
+      "learning_rate": 8.228872576698664e-06,
+      "loss": 0.3728,
+      "step": 31270
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 12.453121185302734,
+      "learning_rate": 8.225108225108225e-06,
+      "loss": 0.612,
+      "step": 31280
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 18.037458419799805,
+      "learning_rate": 8.221343873517787e-06,
+      "loss": 0.6464,
+      "step": 31290
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 14.36330795288086,
+      "learning_rate": 8.21757952192735e-06,
+      "loss": 0.6451,
+      "step": 31300
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 17.178071975708008,
+      "learning_rate": 8.213815170336911e-06,
+      "loss": 0.4489,
+      "step": 31310
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 3.543149948120117,
+      "learning_rate": 8.210050818746471e-06,
+      "loss": 0.8048,
+      "step": 31320
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 8.747088432312012,
+      "learning_rate": 8.206286467156033e-06,
+      "loss": 0.5323,
+      "step": 31330
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 10.16124439239502,
+      "learning_rate": 8.202522115565594e-06,
+      "loss": 0.2721,
+      "step": 31340
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 25.606826782226562,
+      "learning_rate": 8.198757763975156e-06,
+      "loss": 0.3302,
+      "step": 31350
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 10.258896827697754,
+      "learning_rate": 8.194993412384717e-06,
+      "loss": 0.4167,
+      "step": 31360
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 11.452646255493164,
+      "learning_rate": 8.191229060794278e-06,
+      "loss": 0.5729,
+      "step": 31370
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 20.775026321411133,
+      "learning_rate": 8.18746470920384e-06,
+      "loss": 0.6972,
+      "step": 31380
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 0.9261711239814758,
+      "learning_rate": 8.183700357613401e-06,
+      "loss": 0.4341,
+      "step": 31390
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 23.06140899658203,
+      "learning_rate": 8.179936006022963e-06,
+      "loss": 0.6072,
+      "step": 31400
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 0.12934058904647827,
+      "learning_rate": 8.176171654432524e-06,
+      "loss": 0.5984,
+      "step": 31410
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 11.607454299926758,
+      "learning_rate": 8.172407302842086e-06,
+      "loss": 0.4704,
+      "step": 31420
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 17.176578521728516,
+      "learning_rate": 8.168642951251647e-06,
+      "loss": 0.6911,
+      "step": 31430
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 21.022741317749023,
+      "learning_rate": 8.164878599661209e-06,
+      "loss": 0.5338,
+      "step": 31440
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 1.4555150270462036,
+      "learning_rate": 8.16111424807077e-06,
+      "loss": 0.5478,
+      "step": 31450
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 0.8523018956184387,
+      "learning_rate": 8.157349896480332e-06,
+      "loss": 0.3781,
+      "step": 31460
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 0.12009706348180771,
+      "learning_rate": 8.153585544889893e-06,
+      "loss": 0.5797,
+      "step": 31470
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 3.4244158267974854,
+      "learning_rate": 8.149821193299455e-06,
+      "loss": 0.2886,
+      "step": 31480
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 5.182470798492432,
+      "learning_rate": 8.146056841709016e-06,
+      "loss": 0.4928,
+      "step": 31490
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 37.03120040893555,
+      "learning_rate": 8.142292490118577e-06,
+      "loss": 0.5429,
+      "step": 31500
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 10.818685531616211,
+      "learning_rate": 8.138528138528139e-06,
+      "loss": 0.7055,
+      "step": 31510
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 19.673473358154297,
+      "learning_rate": 8.1347637869377e-06,
+      "loss": 0.7041,
+      "step": 31520
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 19.44447135925293,
+      "learning_rate": 8.130999435347262e-06,
+      "loss": 0.4307,
+      "step": 31530
+    },
+    {
+      "epoch": 5.94,
+      "grad_norm": 15.414891242980957,
+      "learning_rate": 8.127235083756823e-06,
+      "loss": 0.433,
+      "step": 31540
+    },
+    {
+      "epoch": 5.94,
+      "grad_norm": 1.3869212865829468,
+      "learning_rate": 8.123470732166385e-06,
+      "loss": 0.4351,
+      "step": 31550
+    },
+    {
+      "epoch": 5.94,
+      "grad_norm": 6.919072151184082,
+      "learning_rate": 8.119706380575946e-06,
+      "loss": 0.4183,
+      "step": 31560
+    },
+    {
+      "epoch": 5.94,
+      "grad_norm": 18.920053482055664,
+      "learning_rate": 8.115942028985508e-06,
+      "loss": 0.6036,
+      "step": 31570
+    },
+    {
+      "epoch": 5.94,
+      "grad_norm": 23.368011474609375,
+      "learning_rate": 8.112177677395069e-06,
+      "loss": 0.4688,
+      "step": 31580
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 43.60259246826172,
+      "learning_rate": 8.10841332580463e-06,
+      "loss": 0.2959,
+      "step": 31590
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 3.7617850303649902,
+      "learning_rate": 8.104648974214192e-06,
+      "loss": 0.3229,
+      "step": 31600
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 28.480266571044922,
+      "learning_rate": 8.100884622623754e-06,
+      "loss": 0.5828,
+      "step": 31610
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 6.059977054595947,
+      "learning_rate": 8.097120271033315e-06,
+      "loss": 0.6493,
+      "step": 31620
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 21.183731079101562,
+      "learning_rate": 8.093355919442876e-06,
+      "loss": 0.4919,
+      "step": 31630
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 0.6658665537834167,
+      "learning_rate": 8.089591567852438e-06,
+      "loss": 0.3223,
+      "step": 31640
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 6.641610622406006,
+      "learning_rate": 8.085827216262e-06,
+      "loss": 0.452,
+      "step": 31650
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 2.345966339111328,
+      "learning_rate": 8.08206286467156e-06,
+      "loss": 0.4293,
+      "step": 31660
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 21.061580657958984,
+      "learning_rate": 8.078298513081122e-06,
+      "loss": 0.5237,
+      "step": 31670
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 20.903779983520508,
+      "learning_rate": 8.074534161490684e-06,
+      "loss": 0.4802,
+      "step": 31680
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 13.212108612060547,
+      "learning_rate": 8.070769809900245e-06,
+      "loss": 0.674,
+      "step": 31690
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 12.111379623413086,
+      "learning_rate": 8.067005458309807e-06,
+      "loss": 0.519,
+      "step": 31700
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 21.36750030517578,
+      "learning_rate": 8.063241106719368e-06,
+      "loss": 0.6218,
+      "step": 31710
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 35.358192443847656,
+      "learning_rate": 8.05947675512893e-06,
+      "loss": 0.5728,
+      "step": 31720
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 16.899436950683594,
+      "learning_rate": 8.055712403538491e-06,
+      "loss": 0.3583,
+      "step": 31730
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 27.319355010986328,
+      "learning_rate": 8.051948051948052e-06,
+      "loss": 0.5901,
+      "step": 31740
+    },
+    {
+      "epoch": 5.98,
+      "grad_norm": 14.794363021850586,
+      "learning_rate": 8.048183700357614e-06,
+      "loss": 0.5036,
+      "step": 31750
+    },
+    {
+      "epoch": 5.98,
+      "grad_norm": 0.04856634885072708,
+      "learning_rate": 8.044419348767175e-06,
+      "loss": 0.4173,
+      "step": 31760
+    },
+    {
+      "epoch": 5.98,
+      "grad_norm": 16.557096481323242,
+      "learning_rate": 8.040654997176737e-06,
+      "loss": 0.6479,
+      "step": 31770
+    },
+    {
+      "epoch": 5.98,
+      "grad_norm": 12.90949821472168,
+      "learning_rate": 8.036890645586298e-06,
+      "loss": 0.5475,
+      "step": 31780
+    },
+    {
+      "epoch": 5.98,
+      "grad_norm": 23.995361328125,
+      "learning_rate": 8.03312629399586e-06,
+      "loss": 0.4879,
+      "step": 31790
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 15.855318069458008,
+      "learning_rate": 8.029361942405421e-06,
+      "loss": 0.5636,
+      "step": 31800
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 17.926624298095703,
+      "learning_rate": 8.025597590814983e-06,
+      "loss": 0.6181,
+      "step": 31810
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 25.674468994140625,
+      "learning_rate": 8.021833239224544e-06,
+      "loss": 0.5092,
+      "step": 31820
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 15.86432933807373,
+      "learning_rate": 8.018068887634106e-06,
+      "loss": 0.5039,
+      "step": 31830
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 16.778894424438477,
+      "learning_rate": 8.014304536043667e-06,
+      "loss": 0.53,
+      "step": 31840
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 20.491600036621094,
+      "learning_rate": 8.010540184453229e-06,
+      "loss": 0.5336,
+      "step": 31850
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 22.663936614990234,
+      "learning_rate": 8.00677583286279e-06,
+      "loss": 0.2657,
+      "step": 31860
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 30.6973876953125,
+      "learning_rate": 8.003011481272351e-06,
+      "loss": 0.3944,
+      "step": 31870
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.9201333333333334,
+      "eval_loss": 0.3082124590873718,
+      "eval_runtime": 51.3947,
+      "eval_samples_per_second": 145.929,
+      "eval_steps_per_second": 18.251,
+      "step": 31878
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 16.97609519958496,
+      "learning_rate": 7.999247129681913e-06,
+      "loss": 0.5414,
+      "step": 31880
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 34.15602111816406,
+      "learning_rate": 7.995482778091474e-06,
+      "loss": 0.7876,
+      "step": 31890
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 18.55365753173828,
+      "learning_rate": 7.991718426501036e-06,
+      "loss": 0.5426,
+      "step": 31900
+    },
+    {
+      "epoch": 6.01,
+      "grad_norm": 23.530202865600586,
+      "learning_rate": 7.987954074910597e-06,
+      "loss": 0.5966,
+      "step": 31910
+    },
+    {
+      "epoch": 6.01,
+      "grad_norm": 22.947160720825195,
+      "learning_rate": 7.984189723320159e-06,
+      "loss": 0.5899,
+      "step": 31920
+    },
+    {
+      "epoch": 6.01,
+      "grad_norm": 21.2280216217041,
+      "learning_rate": 7.98042537172972e-06,
+      "loss": 0.2858,
+      "step": 31930
+    },
+    {
+      "epoch": 6.01,
+      "grad_norm": 13.020065307617188,
+      "learning_rate": 7.976661020139282e-06,
+      "loss": 0.6148,
+      "step": 31940
+    },
+    {
+      "epoch": 6.01,
+      "grad_norm": 25.60552978515625,
+      "learning_rate": 7.972896668548843e-06,
+      "loss": 0.4152,
+      "step": 31950
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 0.053030744194984436,
+      "learning_rate": 7.969132316958405e-06,
+      "loss": 0.4397,
+      "step": 31960
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 18.835439682006836,
+      "learning_rate": 7.965367965367966e-06,
+      "loss": 0.574,
+      "step": 31970
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 13.809188842773438,
+      "learning_rate": 7.961603613777528e-06,
+      "loss": 0.3238,
+      "step": 31980
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 17.327247619628906,
+      "learning_rate": 7.957839262187089e-06,
+      "loss": 0.7307,
+      "step": 31990
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 9.818562507629395,
+      "learning_rate": 7.95407491059665e-06,
+      "loss": 0.4293,
+      "step": 32000
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 0.10741060227155685,
+      "learning_rate": 7.950310559006212e-06,
+      "loss": 0.3728,
+      "step": 32010
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 12.27418041229248,
+      "learning_rate": 7.946546207415773e-06,
+      "loss": 0.5624,
+      "step": 32020
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 24.412097930908203,
+      "learning_rate": 7.942781855825335e-06,
+      "loss": 0.3476,
+      "step": 32030
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 11.939266204833984,
+      "learning_rate": 7.939017504234896e-06,
+      "loss": 0.6026,
+      "step": 32040
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 1.6670972108840942,
+      "learning_rate": 7.935253152644458e-06,
+      "loss": 0.461,
+      "step": 32050
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 18.398563385009766,
+      "learning_rate": 7.93148880105402e-06,
+      "loss": 0.8113,
+      "step": 32060
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 7.880075454711914,
+      "learning_rate": 7.92772444946358e-06,
+      "loss": 0.3885,
+      "step": 32070
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 12.342472076416016,
+      "learning_rate": 7.923960097873142e-06,
+      "loss": 0.251,
+      "step": 32080
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 49.20840072631836,
+      "learning_rate": 7.920195746282704e-06,
+      "loss": 0.5007,
+      "step": 32090
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 17.772216796875,
+      "learning_rate": 7.916431394692265e-06,
+      "loss": 0.3225,
+      "step": 32100
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 53.0165901184082,
+      "learning_rate": 7.912667043101827e-06,
+      "loss": 0.597,
+      "step": 32110
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 16.379194259643555,
+      "learning_rate": 7.908902691511388e-06,
+      "loss": 0.4052,
+      "step": 32120
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 0.09049464017152786,
+      "learning_rate": 7.90513833992095e-06,
+      "loss": 0.777,
+      "step": 32130
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 8.671985626220703,
+      "learning_rate": 7.901373988330511e-06,
+      "loss": 0.8194,
+      "step": 32140
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 34.3663215637207,
+      "learning_rate": 7.897609636740072e-06,
+      "loss": 0.6592,
+      "step": 32150
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 1.3141288757324219,
+      "learning_rate": 7.893845285149634e-06,
+      "loss": 0.3544,
+      "step": 32160
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 17.053375244140625,
+      "learning_rate": 7.890080933559195e-06,
+      "loss": 0.7546,
+      "step": 32170
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 40.75886917114258,
+      "learning_rate": 7.886316581968757e-06,
+      "loss": 0.5418,
+      "step": 32180
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 9.297530174255371,
+      "learning_rate": 7.882552230378318e-06,
+      "loss": 0.6633,
+      "step": 32190
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 13.147290229797363,
+      "learning_rate": 7.87878787878788e-06,
+      "loss": 0.4542,
+      "step": 32200
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 31.284879684448242,
+      "learning_rate": 7.875023527197441e-06,
+      "loss": 0.4669,
+      "step": 32210
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 15.669341087341309,
+      "learning_rate": 7.871259175607003e-06,
+      "loss": 0.1938,
+      "step": 32220
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 44.40217590332031,
+      "learning_rate": 7.867494824016564e-06,
+      "loss": 0.3318,
+      "step": 32230
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 6.310189247131348,
+      "learning_rate": 7.863730472426126e-06,
+      "loss": 0.6378,
+      "step": 32240
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 13.212310791015625,
+      "learning_rate": 7.859966120835687e-06,
+      "loss": 0.6015,
+      "step": 32250
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 15.912529945373535,
+      "learning_rate": 7.856201769245248e-06,
+      "loss": 0.5288,
+      "step": 32260
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 1.8978145122528076,
+      "learning_rate": 7.852437417654808e-06,
+      "loss": 0.4826,
+      "step": 32270
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 39.089073181152344,
+      "learning_rate": 7.848673066064371e-06,
+      "loss": 0.5948,
+      "step": 32280
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 8.845064163208008,
+      "learning_rate": 7.844908714473933e-06,
+      "loss": 0.4591,
+      "step": 32290
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 20.97511100769043,
+      "learning_rate": 7.841144362883494e-06,
+      "loss": 0.6119,
+      "step": 32300
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 8.488319396972656,
+      "learning_rate": 7.837380011293056e-06,
+      "loss": 0.6892,
+      "step": 32310
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 38.96818542480469,
+      "learning_rate": 7.833615659702617e-06,
+      "loss": 0.3597,
+      "step": 32320
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 1.3680989742279053,
+      "learning_rate": 7.829851308112179e-06,
+      "loss": 0.5009,
+      "step": 32330
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 26.558027267456055,
+      "learning_rate": 7.82608695652174e-06,
+      "loss": 0.6395,
+      "step": 32340
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 17.90494155883789,
+      "learning_rate": 7.822322604931302e-06,
+      "loss": 0.6256,
+      "step": 32350
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 11.778217315673828,
+      "learning_rate": 7.818558253340861e-06,
+      "loss": 0.3306,
+      "step": 32360
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 18.40743637084961,
+      "learning_rate": 7.814793901750425e-06,
+      "loss": 0.4534,
+      "step": 32370
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 16.67788314819336,
+      "learning_rate": 7.811029550159986e-06,
+      "loss": 0.6644,
+      "step": 32380
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 0.07633399218320847,
+      "learning_rate": 7.807265198569547e-06,
+      "loss": 0.5102,
+      "step": 32390
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 15.977582931518555,
+      "learning_rate": 7.803500846979109e-06,
+      "loss": 0.5423,
+      "step": 32400
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 53.08889389038086,
+      "learning_rate": 7.79973649538867e-06,
+      "loss": 0.596,
+      "step": 32410
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 7.981726169586182,
+      "learning_rate": 7.795972143798232e-06,
+      "loss": 0.6868,
+      "step": 32420
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 22.594266891479492,
+      "learning_rate": 7.792207792207793e-06,
+      "loss": 0.6394,
+      "step": 32430
+    },
+    {
+      "epoch": 6.11,
+      "grad_norm": 18.69639778137207,
+      "learning_rate": 7.788443440617355e-06,
+      "loss": 0.4663,
+      "step": 32440
+    },
+    {
+      "epoch": 6.11,
+      "grad_norm": 11.259078025817871,
+      "learning_rate": 7.784679089026915e-06,
+      "loss": 0.4948,
+      "step": 32450
+    },
+    {
+      "epoch": 6.11,
+      "grad_norm": 6.536806583404541,
+      "learning_rate": 7.780914737436478e-06,
+      "loss": 0.5363,
+      "step": 32460
+    },
+    {
+      "epoch": 6.11,
+      "grad_norm": 24.416154861450195,
+      "learning_rate": 7.777150385846039e-06,
+      "loss": 0.6788,
+      "step": 32470
+    },
+    {
+      "epoch": 6.11,
+      "grad_norm": 7.047547340393066,
+      "learning_rate": 7.7733860342556e-06,
+      "loss": 0.6391,
+      "step": 32480
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 32.992122650146484,
+      "learning_rate": 7.769621682665162e-06,
+      "loss": 0.4505,
+      "step": 32490
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 21.74233055114746,
+      "learning_rate": 7.765857331074724e-06,
+      "loss": 0.441,
+      "step": 32500
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 2.9683542251586914,
+      "learning_rate": 7.762092979484285e-06,
+      "loss": 0.4137,
+      "step": 32510
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 6.940702438354492,
+      "learning_rate": 7.758328627893846e-06,
+      "loss": 0.3561,
+      "step": 32520
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 13.663592338562012,
+      "learning_rate": 7.754564276303408e-06,
+      "loss": 0.4385,
+      "step": 32530
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 4.1214752197265625,
+      "learning_rate": 7.750799924712968e-06,
+      "loss": 0.5538,
+      "step": 32540
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 41.12890625,
+      "learning_rate": 7.747035573122529e-06,
+      "loss": 0.4315,
+      "step": 32550
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 6.651817798614502,
+      "learning_rate": 7.743271221532092e-06,
+      "loss": 0.5182,
+      "step": 32560
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 22.12759780883789,
+      "learning_rate": 7.739506869941654e-06,
+      "loss": 0.4464,
+      "step": 32570
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 9.155705451965332,
+      "learning_rate": 7.735742518351215e-06,
+      "loss": 0.5285,
+      "step": 32580
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 22.846324920654297,
+      "learning_rate": 7.731978166760777e-06,
+      "loss": 0.5287,
+      "step": 32590
+    },
+    {
+      "epoch": 6.14,
+      "grad_norm": 2.597435474395752,
+      "learning_rate": 7.728213815170338e-06,
+      "loss": 0.064,
+      "step": 32600
+    },
+    {
+      "epoch": 6.14,
+      "grad_norm": 10.483798027038574,
+      "learning_rate": 7.7244494635799e-06,
+      "loss": 0.3093,
+      "step": 32610
+    },
+    {
+      "epoch": 6.14,
+      "grad_norm": 9.434378623962402,
+      "learning_rate": 7.72068511198946e-06,
+      "loss": 0.6346,
+      "step": 32620
+    },
+    {
+      "epoch": 6.14,
+      "grad_norm": 0.27151790261268616,
+      "learning_rate": 7.71692076039902e-06,
+      "loss": 0.3794,
+      "step": 32630
+    },
+    {
+      "epoch": 6.14,
+      "grad_norm": 7.006466388702393,
+      "learning_rate": 7.713156408808582e-06,
+      "loss": 0.5533,
+      "step": 32640
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 6.906270980834961,
+      "learning_rate": 7.709392057218145e-06,
+      "loss": 0.4613,
+      "step": 32650
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 1.0680428743362427,
+      "learning_rate": 7.705627705627707e-06,
+      "loss": 0.474,
+      "step": 32660
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 0.06817318499088287,
+      "learning_rate": 7.701863354037268e-06,
+      "loss": 0.4843,
+      "step": 32670
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 9.567086219787598,
+      "learning_rate": 7.69809900244683e-06,
+      "loss": 0.6909,
+      "step": 32680
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 4.2329301834106445,
+      "learning_rate": 7.694334650856391e-06,
+      "loss": 0.4636,
+      "step": 32690
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 0.9662070870399475,
+      "learning_rate": 7.690570299265953e-06,
+      "loss": 0.4107,
+      "step": 32700
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 10.737784385681152,
+      "learning_rate": 7.686805947675512e-06,
+      "loss": 0.3815,
+      "step": 32710
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 16.375762939453125,
+      "learning_rate": 7.683041596085074e-06,
+      "loss": 0.4073,
+      "step": 32720
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 3.9611103534698486,
+      "learning_rate": 7.679277244494635e-06,
+      "loss": 0.5408,
+      "step": 32730
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 3.900883674621582,
+      "learning_rate": 7.675512892904199e-06,
+      "loss": 0.5193,
+      "step": 32740
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 13.367103576660156,
+      "learning_rate": 7.67174854131376e-06,
+      "loss": 0.7197,
+      "step": 32750
+    },
+    {
+      "epoch": 6.17,
+      "grad_norm": 0.168531134724617,
+      "learning_rate": 7.667984189723321e-06,
+      "loss": 0.3934,
+      "step": 32760
+    },
+    {
+      "epoch": 6.17,
+      "grad_norm": 2.455349922180176,
+      "learning_rate": 7.664219838132883e-06,
+      "loss": 0.5187,
+      "step": 32770
+    },
+    {
+      "epoch": 6.17,
+      "grad_norm": 17.506908416748047,
+      "learning_rate": 7.660455486542444e-06,
+      "loss": 0.5227,
+      "step": 32780
+    },
+    {
+      "epoch": 6.17,
+      "grad_norm": 7.857856750488281,
+      "learning_rate": 7.656691134952006e-06,
+      "loss": 0.6049,
+      "step": 32790
+    },
+    {
+      "epoch": 6.17,
+      "grad_norm": 10.179645538330078,
+      "learning_rate": 7.652926783361566e-06,
+      "loss": 0.5264,
+      "step": 32800
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 15.510612487792969,
+      "learning_rate": 7.649162431771127e-06,
+      "loss": 0.3707,
+      "step": 32810
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 4.712235927581787,
+      "learning_rate": 7.645398080180689e-06,
+      "loss": 0.5458,
+      "step": 32820
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 7.241546630859375,
+      "learning_rate": 7.641633728590252e-06,
+      "loss": 0.5487,
+      "step": 32830
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 1.5072888135910034,
+      "learning_rate": 7.637869376999813e-06,
+      "loss": 0.6038,
+      "step": 32840
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 0.652369499206543,
+      "learning_rate": 7.634105025409375e-06,
+      "loss": 0.5724,
+      "step": 32850
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 4.111026287078857,
+      "learning_rate": 7.630340673818936e-06,
+      "loss": 0.2281,
+      "step": 32860
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 0.2701318860054016,
+      "learning_rate": 7.6265763222284976e-06,
+      "loss": 0.6504,
+      "step": 32870
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 13.367432594299316,
+      "learning_rate": 7.622811970638057e-06,
+      "loss": 0.2885,
+      "step": 32880
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 5.478610515594482,
+      "learning_rate": 7.61904761904762e-06,
+      "loss": 0.5181,
+      "step": 32890
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 8.030369758605957,
+      "learning_rate": 7.615283267457181e-06,
+      "loss": 0.5897,
+      "step": 32900
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 13.834616661071777,
+      "learning_rate": 7.6115189158667426e-06,
+      "loss": 0.5119,
+      "step": 32910
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 5.764062881469727,
+      "learning_rate": 7.607754564276304e-06,
+      "loss": 0.5504,
+      "step": 32920
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 51.650718688964844,
+      "learning_rate": 7.6039902126858655e-06,
+      "loss": 0.4928,
+      "step": 32930
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 43.39817428588867,
+      "learning_rate": 7.600225861095427e-06,
+      "loss": 0.3655,
+      "step": 32940
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 1.6382009983062744,
+      "learning_rate": 7.596461509504988e-06,
+      "loss": 0.4539,
+      "step": 32950
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 0.764748215675354,
+      "learning_rate": 7.592697157914551e-06,
+      "loss": 0.6964,
+      "step": 32960
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 46.74039077758789,
+      "learning_rate": 7.5889328063241105e-06,
+      "loss": 0.4344,
+      "step": 32970
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 24.714452743530273,
+      "learning_rate": 7.585168454733673e-06,
+      "loss": 0.7972,
+      "step": 32980
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 15.897273063659668,
+      "learning_rate": 7.581404103143234e-06,
+      "loss": 0.443,
+      "step": 32990
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 5.593212604522705,
+      "learning_rate": 7.577639751552796e-06,
+      "loss": 0.5032,
+      "step": 33000
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 10.625405311584473,
+      "learning_rate": 7.573875399962357e-06,
+      "loss": 0.4239,
+      "step": 33010
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 28.671268463134766,
+      "learning_rate": 7.570111048371919e-06,
+      "loss": 0.3406,
+      "step": 33020
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 1.3727935552597046,
+      "learning_rate": 7.56634669678148e-06,
+      "loss": 0.4927,
+      "step": 33030
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 8.277888298034668,
+      "learning_rate": 7.5625823451910415e-06,
+      "loss": 0.5651,
+      "step": 33040
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 23.557905197143555,
+      "learning_rate": 7.558817993600603e-06,
+      "loss": 0.4431,
+      "step": 33050
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 27.223262786865234,
+      "learning_rate": 7.555053642010164e-06,
+      "loss": 0.8412,
+      "step": 33060
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 14.346407890319824,
+      "learning_rate": 7.551289290419725e-06,
+      "loss": 0.5623,
+      "step": 33070
+    },
+    {
+      "epoch": 6.23,
+      "grad_norm": 0.6994918584823608,
+      "learning_rate": 7.547524938829287e-06,
+      "loss": 0.6004,
+      "step": 33080
+    },
+    {
+      "epoch": 6.23,
+      "grad_norm": 3.5773377418518066,
+      "learning_rate": 7.543760587238849e-06,
+      "loss": 0.4514,
+      "step": 33090
+    },
+    {
+      "epoch": 6.23,
+      "grad_norm": 15.664129257202148,
+      "learning_rate": 7.53999623564841e-06,
+      "loss": 0.437,
+      "step": 33100
+    },
+    {
+      "epoch": 6.23,
+      "grad_norm": 14.454998016357422,
+      "learning_rate": 7.536231884057972e-06,
+      "loss": 0.3146,
+      "step": 33110
+    },
+    {
+      "epoch": 6.23,
+      "grad_norm": 14.287149429321289,
+      "learning_rate": 7.532467532467533e-06,
+      "loss": 0.3651,
+      "step": 33120
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 35.363502502441406,
+      "learning_rate": 7.528703180877095e-06,
+      "loss": 0.3985,
+      "step": 33130
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 11.074187278747559,
+      "learning_rate": 7.524938829286656e-06,
+      "loss": 0.7611,
+      "step": 33140
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 2.34920597076416,
+      "learning_rate": 7.521174477696217e-06,
+      "loss": 0.6438,
+      "step": 33150
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 3.194157361984253,
+      "learning_rate": 7.517410126105778e-06,
+      "loss": 0.9574,
+      "step": 33160
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 15.871007919311523,
+      "learning_rate": 7.5136457745153405e-06,
+      "loss": 0.8127,
+      "step": 33170
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 0.2083834558725357,
+      "learning_rate": 7.509881422924902e-06,
+      "loss": 0.438,
+      "step": 33180
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 9.858641624450684,
+      "learning_rate": 7.5061170713344635e-06,
+      "loss": 0.4504,
+      "step": 33190
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 9.694896697998047,
+      "learning_rate": 7.502352719744025e-06,
+      "loss": 0.6191,
+      "step": 33200
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 18.69541358947754,
+      "learning_rate": 7.498588368153586e-06,
+      "loss": 0.9686,
+      "step": 33210
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 18.836008071899414,
+      "learning_rate": 7.494824016563148e-06,
+      "loss": 0.4195,
+      "step": 33220
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 8.28152084350586,
+      "learning_rate": 7.4910596649727084e-06,
+      "loss": 0.4661,
+      "step": 33230
+    },
+    {
+      "epoch": 6.26,
+      "grad_norm": 0.8841540813446045,
+      "learning_rate": 7.48729531338227e-06,
+      "loss": 1.1192,
+      "step": 33240
+    },
+    {
+      "epoch": 6.26,
+      "grad_norm": 21.22022247314453,
+      "learning_rate": 7.483530961791831e-06,
+      "loss": 0.5756,
+      "step": 33250
+    },
+    {
+      "epoch": 6.26,
+      "grad_norm": 5.069668769836426,
+      "learning_rate": 7.479766610201394e-06,
+      "loss": 0.5685,
+      "step": 33260
+    },
+    {
+      "epoch": 6.26,
+      "grad_norm": 1.734766960144043,
+      "learning_rate": 7.476002258610955e-06,
+      "loss": 0.2923,
+      "step": 33270
+    },
+    {
+      "epoch": 6.26,
+      "grad_norm": 15.850728988647461,
+      "learning_rate": 7.472237907020517e-06,
+      "loss": 0.6755,
+      "step": 33280
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 0.8242527842521667,
+      "learning_rate": 7.468473555430078e-06,
+      "loss": 0.5514,
+      "step": 33290
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 22.578418731689453,
+      "learning_rate": 7.4647092038396395e-06,
+      "loss": 0.5968,
+      "step": 33300
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 17.16265869140625,
+      "learning_rate": 7.460944852249201e-06,
+      "loss": 0.4647,
+      "step": 33310
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 32.857601165771484,
+      "learning_rate": 7.457180500658762e-06,
+      "loss": 0.511,
+      "step": 33320
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 10.078173637390137,
+      "learning_rate": 7.453416149068323e-06,
+      "loss": 0.3435,
+      "step": 33330
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 18.10398292541504,
+      "learning_rate": 7.4496517974778845e-06,
+      "loss": 0.348,
+      "step": 33340
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 9.601995468139648,
+      "learning_rate": 7.445887445887446e-06,
+      "loss": 0.4952,
+      "step": 33350
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 13.778106689453125,
+      "learning_rate": 7.442123094297008e-06,
+      "loss": 0.4809,
+      "step": 33360
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 19.068946838378906,
+      "learning_rate": 7.43835874270657e-06,
+      "loss": 0.3458,
+      "step": 33370
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 1.7670868635177612,
+      "learning_rate": 7.434594391116131e-06,
+      "loss": 0.6565,
+      "step": 33380
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 25.797090530395508,
+      "learning_rate": 7.430830039525693e-06,
+      "loss": 0.7225,
+      "step": 33390
+    },
+    {
+      "epoch": 6.29,
+      "grad_norm": 30.813005447387695,
+      "learning_rate": 7.427065687935254e-06,
+      "loss": 0.8055,
+      "step": 33400
+    },
+    {
+      "epoch": 6.29,
+      "grad_norm": 7.973330497741699,
+      "learning_rate": 7.423301336344815e-06,
+      "loss": 0.4666,
+      "step": 33410
+    },
+    {
+      "epoch": 6.29,
+      "grad_norm": 9.906311988830566,
+      "learning_rate": 7.419536984754376e-06,
+      "loss": 0.3954,
+      "step": 33420
+    },
+    {
+      "epoch": 6.29,
+      "grad_norm": 43.313724517822266,
+      "learning_rate": 7.415772633163938e-06,
+      "loss": 0.6149,
+      "step": 33430
+    },
+    {
+      "epoch": 6.29,
+      "grad_norm": 7.175143241882324,
+      "learning_rate": 7.412008281573499e-06,
+      "loss": 0.6234,
+      "step": 33440
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 28.30812644958496,
+      "learning_rate": 7.4082439299830614e-06,
+      "loss": 0.3256,
+      "step": 33450
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 12.158953666687012,
+      "learning_rate": 7.404479578392623e-06,
+      "loss": 0.3962,
+      "step": 33460
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 17.756189346313477,
+      "learning_rate": 7.400715226802184e-06,
+      "loss": 0.5152,
+      "step": 33470
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 18.10654067993164,
+      "learning_rate": 7.396950875211746e-06,
+      "loss": 0.6907,
+      "step": 33480
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 18.467599868774414,
+      "learning_rate": 7.3931865236213064e-06,
+      "loss": 0.4874,
+      "step": 33490
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 1.991639256477356,
+      "learning_rate": 7.389422172030868e-06,
+      "loss": 0.3471,
+      "step": 33500
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 20.875844955444336,
+      "learning_rate": 7.385657820440429e-06,
+      "loss": 0.8077,
+      "step": 33510
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 4.168043613433838,
+      "learning_rate": 7.381893468849991e-06,
+      "loss": 0.3689,
+      "step": 33520
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 0.10714370012283325,
+      "learning_rate": 7.378129117259552e-06,
+      "loss": 0.4633,
+      "step": 33530
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 14.68402099609375,
+      "learning_rate": 7.3743647656691146e-06,
+      "loss": 0.5519,
+      "step": 33540
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 1.1338722705841064,
+      "learning_rate": 7.370600414078676e-06,
+      "loss": 0.6535,
+      "step": 33550
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 20.80348014831543,
+      "learning_rate": 7.3668360624882375e-06,
+      "loss": 0.6073,
+      "step": 33560
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 17.18090057373047,
+      "learning_rate": 7.363071710897799e-06,
+      "loss": 0.5032,
+      "step": 33570
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 0.5403892397880554,
+      "learning_rate": 7.3593073593073596e-06,
+      "loss": 0.4669,
+      "step": 33580
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 10.763050079345703,
+      "learning_rate": 7.355543007716921e-06,
+      "loss": 0.5767,
+      "step": 33590
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 46.35898971557617,
+      "learning_rate": 7.3517786561264825e-06,
+      "loss": 0.6032,
+      "step": 33600
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 12.066878318786621,
+      "learning_rate": 7.348014304536044e-06,
+      "loss": 0.3331,
+      "step": 33610
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 7.480466365814209,
+      "learning_rate": 7.344249952945605e-06,
+      "loss": 0.6212,
+      "step": 33620
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 18.744464874267578,
+      "learning_rate": 7.340485601355168e-06,
+      "loss": 0.7442,
+      "step": 33630
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 7.432648658752441,
+      "learning_rate": 7.336721249764729e-06,
+      "loss": 0.358,
+      "step": 33640
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 8.940939903259277,
+      "learning_rate": 7.332956898174291e-06,
+      "loss": 0.7015,
+      "step": 33650
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 0.28902769088745117,
+      "learning_rate": 7.329192546583852e-06,
+      "loss": 0.5588,
+      "step": 33660
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 0.46211880445480347,
+      "learning_rate": 7.325428194993413e-06,
+      "loss": 0.6197,
+      "step": 33670
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 25.7103328704834,
+      "learning_rate": 7.321663843402974e-06,
+      "loss": 0.4847,
+      "step": 33680
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 8.652647018432617,
+      "learning_rate": 7.317899491812536e-06,
+      "loss": 0.3872,
+      "step": 33690
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 10.216687202453613,
+      "learning_rate": 7.314135140222097e-06,
+      "loss": 0.4059,
+      "step": 33700
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 2.1355504989624023,
+      "learning_rate": 7.3103707886316586e-06,
+      "loss": 0.5013,
+      "step": 33710
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 18.312965393066406,
+      "learning_rate": 7.30660643704122e-06,
+      "loss": 0.536,
+      "step": 33720
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 11.597451210021973,
+      "learning_rate": 7.302842085450782e-06,
+      "loss": 0.4923,
+      "step": 33730
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 21.869190216064453,
+      "learning_rate": 7.299077733860344e-06,
+      "loss": 0.6294,
+      "step": 33740
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 9.284852981567383,
+      "learning_rate": 7.295313382269905e-06,
+      "loss": 0.4809,
+      "step": 33750
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 26.03097152709961,
+      "learning_rate": 7.291549030679466e-06,
+      "loss": 0.6448,
+      "step": 33760
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 22.911418914794922,
+      "learning_rate": 7.287784679089027e-06,
+      "loss": 0.5017,
+      "step": 33770
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 50.75450897216797,
+      "learning_rate": 7.284020327498589e-06,
+      "loss": 0.6046,
+      "step": 33780
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 22.943689346313477,
+      "learning_rate": 7.28025597590815e-06,
+      "loss": 0.4832,
+      "step": 33790
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 7.2119460105896,
+      "learning_rate": 7.276491624317712e-06,
+      "loss": 0.852,
+      "step": 33800
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 9.327651023864746,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 0.5135,
+      "step": 33810
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 20.863204956054688,
+      "learning_rate": 7.2689629211368355e-06,
+      "loss": 0.5866,
+      "step": 33820
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 6.156462669372559,
+      "learning_rate": 7.265198569546397e-06,
+      "loss": 0.62,
+      "step": 33830
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 24.723724365234375,
+      "learning_rate": 7.2614342179559576e-06,
+      "loss": 0.6837,
+      "step": 33840
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 0.5178024172782898,
+      "learning_rate": 7.257669866365519e-06,
+      "loss": 0.2405,
+      "step": 33850
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 7.65985107421875,
+      "learning_rate": 7.2539055147750805e-06,
+      "loss": 0.3353,
+      "step": 33860
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 21.34691619873047,
+      "learning_rate": 7.250141163184642e-06,
+      "loss": 0.6229,
+      "step": 33870
+    },
+    {
+      "epoch": 6.38,
+      "grad_norm": 20.477985382080078,
+      "learning_rate": 7.246376811594203e-06,
+      "loss": 0.4679,
+      "step": 33880
+    },
+    {
+      "epoch": 6.38,
+      "grad_norm": 22.93824005126953,
+      "learning_rate": 7.242612460003765e-06,
+      "loss": 0.5659,
+      "step": 33890
+    },
+    {
+      "epoch": 6.38,
+      "grad_norm": 0.7325233221054077,
+      "learning_rate": 7.238848108413326e-06,
+      "loss": 0.5032,
+      "step": 33900
+    },
+    {
+      "epoch": 6.38,
+      "grad_norm": 14.224431037902832,
+      "learning_rate": 7.235083756822889e-06,
+      "loss": 0.674,
+      "step": 33910
+    },
+    {
+      "epoch": 6.38,
+      "grad_norm": 19.057111740112305,
+      "learning_rate": 7.23131940523245e-06,
+      "loss": 0.4854,
+      "step": 33920
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 16.500669479370117,
+      "learning_rate": 7.22755505364201e-06,
+      "loss": 0.4265,
+      "step": 33930
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 18.075031280517578,
+      "learning_rate": 7.223790702051572e-06,
+      "loss": 0.8634,
+      "step": 33940
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 1.375807523727417,
+      "learning_rate": 7.220026350461134e-06,
+      "loss": 0.5841,
+      "step": 33950
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 2.372751474380493,
+      "learning_rate": 7.216261998870695e-06,
+      "loss": 0.3938,
+      "step": 33960
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 0.7880887985229492,
+      "learning_rate": 7.2124976472802565e-06,
+      "loss": 0.5251,
+      "step": 33970
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 10.042675018310547,
+      "learning_rate": 7.208733295689818e-06,
+      "loss": 0.3012,
+      "step": 33980
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 41.02911376953125,
+      "learning_rate": 7.2049689440993795e-06,
+      "loss": 0.6843,
+      "step": 33990
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 0.09337029606103897,
+      "learning_rate": 7.201204592508941e-06,
+      "loss": 0.6831,
+      "step": 34000
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 0.04473143443465233,
+      "learning_rate": 7.197440240918503e-06,
+      "loss": 0.5183,
+      "step": 34010
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 30.65087127685547,
+      "learning_rate": 7.193675889328063e-06,
+      "loss": 0.5636,
+      "step": 34020
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 22.237075805664062,
+      "learning_rate": 7.189911537737625e-06,
+      "loss": 0.6722,
+      "step": 34030
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 23.21628761291504,
+      "learning_rate": 7.186147186147187e-06,
+      "loss": 0.4067,
+      "step": 34040
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 6.441122055053711,
+      "learning_rate": 7.182382834556748e-06,
+      "loss": 0.373,
+      "step": 34050
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 17.526185989379883,
+      "learning_rate": 7.17861848296631e-06,
+      "loss": 0.4955,
+      "step": 34060
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 5.90300178527832,
+      "learning_rate": 7.174854131375871e-06,
+      "loss": 0.6357,
+      "step": 34070
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 14.964542388916016,
+      "learning_rate": 7.171089779785433e-06,
+      "loss": 0.6288,
+      "step": 34080
+    },
+    {
+      "epoch": 6.42,
+      "grad_norm": 14.822508811950684,
+      "learning_rate": 7.167325428194994e-06,
+      "loss": 0.6348,
+      "step": 34090
+    },
+    {
+      "epoch": 6.42,
+      "grad_norm": 40.991546630859375,
+      "learning_rate": 7.163561076604555e-06,
+      "loss": 0.3295,
+      "step": 34100
+    },
+    {
+      "epoch": 6.42,
+      "grad_norm": 13.501119613647461,
+      "learning_rate": 7.159796725014116e-06,
+      "loss": 0.5754,
+      "step": 34110
+    },
+    {
+      "epoch": 6.42,
+      "grad_norm": 18.85485076904297,
+      "learning_rate": 7.1560323734236784e-06,
+      "loss": 0.5432,
+      "step": 34120
+    },
+    {
+      "epoch": 6.42,
+      "grad_norm": 8.995746612548828,
+      "learning_rate": 7.15226802183324e-06,
+      "loss": 0.9828,
+      "step": 34130
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 13.933249473571777,
+      "learning_rate": 7.148503670242801e-06,
+      "loss": 0.5424,
+      "step": 34140
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 13.510233879089355,
+      "learning_rate": 7.144739318652363e-06,
+      "loss": 0.6617,
+      "step": 34150
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 13.763504981994629,
+      "learning_rate": 7.140974967061924e-06,
+      "loss": 0.4808,
+      "step": 34160
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 18.854583740234375,
+      "learning_rate": 7.137210615471486e-06,
+      "loss": 0.3383,
+      "step": 34170
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 0.48108717799186707,
+      "learning_rate": 7.133446263881047e-06,
+      "loss": 0.4345,
+      "step": 34180
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 50.08021926879883,
+      "learning_rate": 7.129681912290608e-06,
+      "loss": 0.4781,
+      "step": 34190
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 18.643566131591797,
+      "learning_rate": 7.125917560700169e-06,
+      "loss": 0.4631,
+      "step": 34200
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 12.921839714050293,
+      "learning_rate": 7.122153209109732e-06,
+      "loss": 0.5039,
+      "step": 34210
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 27.30744171142578,
+      "learning_rate": 7.118388857519293e-06,
+      "loss": 0.5912,
+      "step": 34220
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 27.566017150878906,
+      "learning_rate": 7.1146245059288545e-06,
+      "loss": 0.4702,
+      "step": 34230
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 12.484590530395508,
+      "learning_rate": 7.110860154338416e-06,
+      "loss": 0.5907,
+      "step": 34240
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 19.984474182128906,
+      "learning_rate": 7.1070958027479774e-06,
+      "loss": 0.2223,
+      "step": 34250
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 6.380549430847168,
+      "learning_rate": 7.103331451157539e-06,
+      "loss": 0.3962,
+      "step": 34260
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 8.179642677307129,
+      "learning_rate": 7.0995670995671e-06,
+      "loss": 0.778,
+      "step": 34270
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 0.15643277764320374,
+      "learning_rate": 7.095802747976661e-06,
+      "loss": 0.316,
+      "step": 34280
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 17.682392120361328,
+      "learning_rate": 7.0920383963862224e-06,
+      "loss": 0.3378,
+      "step": 34290
+    },
+    {
+      "epoch": 6.46,
+      "grad_norm": 10.891948699951172,
+      "learning_rate": 7.088274044795784e-06,
+      "loss": 0.345,
+      "step": 34300
+    },
+    {
+      "epoch": 6.46,
+      "grad_norm": 37.315853118896484,
+      "learning_rate": 7.084509693205346e-06,
+      "loss": 0.5888,
+      "step": 34310
+    },
+    {
+      "epoch": 6.46,
+      "grad_norm": 0.2774854302406311,
+      "learning_rate": 7.080745341614908e-06,
+      "loss": 0.2816,
+      "step": 34320
+    },
+    {
+      "epoch": 6.46,
+      "grad_norm": 15.914167404174805,
+      "learning_rate": 7.076980990024469e-06,
+      "loss": 0.6068,
+      "step": 34330
+    },
+    {
+      "epoch": 6.46,
+      "grad_norm": 17.352615356445312,
+      "learning_rate": 7.073216638434031e-06,
+      "loss": 0.4122,
+      "step": 34340
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 21.448284149169922,
+      "learning_rate": 7.069452286843592e-06,
+      "loss": 0.2653,
+      "step": 34350
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 21.003995895385742,
+      "learning_rate": 7.0656879352531535e-06,
+      "loss": 0.3752,
+      "step": 34360
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 12.033404350280762,
+      "learning_rate": 7.061923583662714e-06,
+      "loss": 0.5186,
+      "step": 34370
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 9.405254364013672,
+      "learning_rate": 7.058159232072276e-06,
+      "loss": 0.7365,
+      "step": 34380
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 25.861106872558594,
+      "learning_rate": 7.054394880481837e-06,
+      "loss": 0.7919,
+      "step": 34390
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 6.728229522705078,
+      "learning_rate": 7.050630528891399e-06,
+      "loss": 0.4589,
+      "step": 34400
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 29.374160766601562,
+      "learning_rate": 7.046866177300961e-06,
+      "loss": 0.7846,
+      "step": 34410
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 31.664583206176758,
+      "learning_rate": 7.043101825710522e-06,
+      "loss": 0.9016,
+      "step": 34420
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 9.76760482788086,
+      "learning_rate": 7.039337474120084e-06,
+      "loss": 0.5766,
+      "step": 34430
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 19.36331558227539,
+      "learning_rate": 7.035573122529645e-06,
+      "loss": 0.4637,
+      "step": 34440
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 7.009848594665527,
+      "learning_rate": 7.031808770939206e-06,
+      "loss": 0.6186,
+      "step": 34450
+    },
+    {
+      "epoch": 6.49,
+      "grad_norm": 11.30031967163086,
+      "learning_rate": 7.028044419348767e-06,
+      "loss": 0.859,
+      "step": 34460
+    },
+    {
+      "epoch": 6.49,
+      "grad_norm": 23.84954833984375,
+      "learning_rate": 7.024280067758329e-06,
+      "loss": 0.4672,
+      "step": 34470
+    },
+    {
+      "epoch": 6.49,
+      "grad_norm": 1.282763957977295,
+      "learning_rate": 7.02051571616789e-06,
+      "loss": 0.3136,
+      "step": 34480
+    },
+    {
+      "epoch": 6.49,
+      "grad_norm": 0.5825543999671936,
+      "learning_rate": 7.0167513645774525e-06,
+      "loss": 0.4409,
+      "step": 34490
+    },
+    {
+      "epoch": 6.49,
+      "grad_norm": 23.45937728881836,
+      "learning_rate": 7.012987012987014e-06,
+      "loss": 0.5154,
+      "step": 34500
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 0.2989109456539154,
+      "learning_rate": 7.009222661396575e-06,
+      "loss": 0.5457,
+      "step": 34510
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 12.64752197265625,
+      "learning_rate": 7.005458309806137e-06,
+      "loss": 0.4991,
+      "step": 34520
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 56.25718688964844,
+      "learning_rate": 7.001693958215698e-06,
+      "loss": 0.6106,
+      "step": 34530
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 16.893884658813477,
+      "learning_rate": 6.997929606625259e-06,
+      "loss": 0.4511,
+      "step": 34540
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 8.532828330993652,
+      "learning_rate": 6.99416525503482e-06,
+      "loss": 0.388,
+      "step": 34550
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 12.58056926727295,
+      "learning_rate": 6.990400903444382e-06,
+      "loss": 0.8358,
+      "step": 34560
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 20.3173885345459,
+      "learning_rate": 6.986636551853943e-06,
+      "loss": 0.3383,
+      "step": 34570
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 7.896786689758301,
+      "learning_rate": 6.982872200263505e-06,
+      "loss": 0.5333,
+      "step": 34580
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 6.4286065101623535,
+      "learning_rate": 6.979107848673067e-06,
+      "loss": 0.4064,
+      "step": 34590
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 23.140363693237305,
+      "learning_rate": 6.9753434970826286e-06,
+      "loss": 0.623,
+      "step": 34600
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 12.2300443649292,
+      "learning_rate": 6.97157914549219e-06,
+      "loss": 0.4408,
+      "step": 34610
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 30.439172744750977,
+      "learning_rate": 6.9678147939017515e-06,
+      "loss": 0.3854,
+      "step": 34620
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 13.196409225463867,
+      "learning_rate": 6.964050442311312e-06,
+      "loss": 0.4278,
+      "step": 34630
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 8.042349815368652,
+      "learning_rate": 6.9602860907208736e-06,
+      "loss": 0.4624,
+      "step": 34640
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 7.406027793884277,
+      "learning_rate": 6.956521739130435e-06,
+      "loss": 0.4726,
+      "step": 34650
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 8.683379173278809,
+      "learning_rate": 6.9527573875399965e-06,
+      "loss": 0.6854,
+      "step": 34660
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 23.16596794128418,
+      "learning_rate": 6.948993035949558e-06,
+      "loss": 0.6081,
+      "step": 34670
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 0.07282250374555588,
+      "learning_rate": 6.94522868435912e-06,
+      "loss": 0.5896,
+      "step": 34680
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 39.00768280029297,
+      "learning_rate": 6.941464332768682e-06,
+      "loss": 0.5436,
+      "step": 34690
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 35.062660217285156,
+      "learning_rate": 6.937699981178243e-06,
+      "loss": 0.6628,
+      "step": 34700
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 9.127306938171387,
+      "learning_rate": 6.933935629587804e-06,
+      "loss": 0.7176,
+      "step": 34710
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 0.9268749952316284,
+      "learning_rate": 6.930171277997365e-06,
+      "loss": 0.438,
+      "step": 34720
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 2.197774648666382,
+      "learning_rate": 6.926406926406927e-06,
+      "loss": 0.4989,
+      "step": 34730
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 32.574371337890625,
+      "learning_rate": 6.922642574816488e-06,
+      "loss": 0.586,
+      "step": 34740
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 31.461170196533203,
+      "learning_rate": 6.91887822322605e-06,
+      "loss": 0.5052,
+      "step": 34750
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 12.898176193237305,
+      "learning_rate": 6.915113871635611e-06,
+      "loss": 0.5327,
+      "step": 34760
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 2.2167086601257324,
+      "learning_rate": 6.911349520045173e-06,
+      "loss": 0.5105,
+      "step": 34770
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 15.682035446166992,
+      "learning_rate": 6.907585168454735e-06,
+      "loss": 0.3067,
+      "step": 34780
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 41.14264678955078,
+      "learning_rate": 6.903820816864296e-06,
+      "loss": 0.4904,
+      "step": 34790
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 0.08841241896152496,
+      "learning_rate": 6.900056465273857e-06,
+      "loss": 0.4581,
+      "step": 34800
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 38.545257568359375,
+      "learning_rate": 6.896292113683418e-06,
+      "loss": 0.4896,
+      "step": 34810
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 35.339149475097656,
+      "learning_rate": 6.89252776209298e-06,
+      "loss": 0.5188,
+      "step": 34820
+    },
+    {
+      "epoch": 6.56,
+      "grad_norm": 14.379655838012695,
+      "learning_rate": 6.888763410502541e-06,
+      "loss": 0.3942,
+      "step": 34830
+    },
+    {
+      "epoch": 6.56,
+      "grad_norm": 21.1137752532959,
+      "learning_rate": 6.884999058912103e-06,
+      "loss": 0.5441,
+      "step": 34840
+    },
+    {
+      "epoch": 6.56,
+      "grad_norm": 20.75609588623047,
+      "learning_rate": 6.881234707321664e-06,
+      "loss": 0.6482,
+      "step": 34850
+    },
+    {
+      "epoch": 6.56,
+      "grad_norm": 19.010169982910156,
+      "learning_rate": 6.8774703557312265e-06,
+      "loss": 0.4973,
+      "step": 34860
+    },
+    {
+      "epoch": 6.56,
+      "grad_norm": 19.815196990966797,
+      "learning_rate": 6.873706004140788e-06,
+      "loss": 0.6047,
+      "step": 34870
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 8.65109634399414,
+      "learning_rate": 6.8699416525503495e-06,
+      "loss": 0.3104,
+      "step": 34880
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 1.1485213041305542,
+      "learning_rate": 6.86617730095991e-06,
+      "loss": 0.4565,
+      "step": 34890
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 1.1633775234222412,
+      "learning_rate": 6.8624129493694715e-06,
+      "loss": 0.5844,
+      "step": 34900
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 7.708075523376465,
+      "learning_rate": 6.858648597779033e-06,
+      "loss": 0.6197,
+      "step": 34910
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 4.411716938018799,
+      "learning_rate": 6.8548842461885945e-06,
+      "loss": 0.5093,
+      "step": 34920
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 18.344188690185547,
+      "learning_rate": 6.851119894598156e-06,
+      "loss": 0.693,
+      "step": 34930
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 40.74114990234375,
+      "learning_rate": 6.847355543007717e-06,
+      "loss": 0.6018,
+      "step": 34940
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 0.057777296751737595,
+      "learning_rate": 6.843591191417279e-06,
+      "loss": 0.5193,
+      "step": 34950
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 13.32892894744873,
+      "learning_rate": 6.839826839826841e-06,
+      "loss": 0.5995,
+      "step": 34960
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 28.153160095214844,
+      "learning_rate": 6.836062488236401e-06,
+      "loss": 0.5686,
+      "step": 34970
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 7.822689056396484,
+      "learning_rate": 6.832298136645963e-06,
+      "loss": 0.4162,
+      "step": 34980
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 25.546289443969727,
+      "learning_rate": 6.828533785055525e-06,
+      "loss": 0.3894,
+      "step": 34990
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 0.47211772203445435,
+      "learning_rate": 6.824769433465086e-06,
+      "loss": 0.2991,
+      "step": 35000
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 6.658637523651123,
+      "learning_rate": 6.821005081874648e-06,
+      "loss": 0.6118,
+      "step": 35010
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 4.372053146362305,
+      "learning_rate": 6.817240730284209e-06,
+      "loss": 0.239,
+      "step": 35020
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 25.087793350219727,
+      "learning_rate": 6.8134763786937705e-06,
+      "loss": 0.6427,
+      "step": 35030
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 18.80016326904297,
+      "learning_rate": 6.809712027103332e-06,
+      "loss": 0.3743,
+      "step": 35040
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 0.2986818552017212,
+      "learning_rate": 6.805947675512894e-06,
+      "loss": 0.3221,
+      "step": 35050
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 15.778828620910645,
+      "learning_rate": 6.802183323922454e-06,
+      "loss": 0.855,
+      "step": 35060
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 1.3159215450286865,
+      "learning_rate": 6.798418972332016e-06,
+      "loss": 0.5559,
+      "step": 35070
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 0.06491217762231827,
+      "learning_rate": 6.794654620741578e-06,
+      "loss": 0.2105,
+      "step": 35080
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 19.969703674316406,
+      "learning_rate": 6.790890269151139e-06,
+      "loss": 0.4636,
+      "step": 35090
+    },
+    {
+      "epoch": 6.61,
+      "grad_norm": 11.362780570983887,
+      "learning_rate": 6.787125917560701e-06,
+      "loss": 0.5571,
+      "step": 35100
+    },
+    {
+      "epoch": 6.61,
+      "grad_norm": 8.18557357788086,
+      "learning_rate": 6.783361565970262e-06,
+      "loss": 0.4913,
+      "step": 35110
+    },
+    {
+      "epoch": 6.61,
+      "grad_norm": 10.069347381591797,
+      "learning_rate": 6.779597214379824e-06,
+      "loss": 0.2839,
+      "step": 35120
+    },
+    {
+      "epoch": 6.61,
+      "grad_norm": 2.1053032875061035,
+      "learning_rate": 6.775832862789385e-06,
+      "loss": 0.6116,
+      "step": 35130
+    },
+    {
+      "epoch": 6.61,
+      "grad_norm": 12.72517204284668,
+      "learning_rate": 6.7720685111989474e-06,
+      "loss": 0.4671,
+      "step": 35140
+    },
+    {
+      "epoch": 6.62,
+      "grad_norm": 19.65533447265625,
+      "learning_rate": 6.768304159608507e-06,
+      "loss": 0.5756,
+      "step": 35150
+    },
+    {
+      "epoch": 6.62,
+      "grad_norm": 15.998537063598633,
+      "learning_rate": 6.764539808018069e-06,
+      "loss": 0.5428,
+      "step": 35160
+    },
+    {
+      "epoch": 6.62,
+      "grad_norm": 24.56063461303711,
+      "learning_rate": 6.760775456427631e-06,
+      "loss": 0.6397,
+      "step": 35170
+    },
+    {
+      "epoch": 6.62,
+      "grad_norm": 0.9871610403060913,
+      "learning_rate": 6.7570111048371924e-06,
+      "loss": 0.5336,
+      "step": 35180
+    },
+    {
+      "epoch": 6.62,
+      "grad_norm": 0.14848846197128296,
+      "learning_rate": 6.753246753246754e-06,
+      "loss": 0.4377,
+      "step": 35190
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 55.045509338378906,
+      "learning_rate": 6.749482401656315e-06,
+      "loss": 0.5341,
+      "step": 35200
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 13.329508781433105,
+      "learning_rate": 6.745718050065877e-06,
+      "loss": 0.5933,
+      "step": 35210
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 3.753176212310791,
+      "learning_rate": 6.741953698475438e-06,
+      "loss": 0.301,
+      "step": 35220
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 0.3469744622707367,
+      "learning_rate": 6.738189346885e-06,
+      "loss": 0.3636,
+      "step": 35230
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 8.834991455078125,
+      "learning_rate": 6.73442499529456e-06,
+      "loss": 0.7362,
+      "step": 35240
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 13.375589370727539,
+      "learning_rate": 6.730660643704122e-06,
+      "loss": 0.4198,
+      "step": 35250
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 0.9780624508857727,
+      "learning_rate": 6.726896292113684e-06,
+      "loss": 0.2504,
+      "step": 35260
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 12.835625648498535,
+      "learning_rate": 6.723131940523246e-06,
+      "loss": 0.7162,
+      "step": 35270
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 1.3344264030456543,
+      "learning_rate": 6.719367588932807e-06,
+      "loss": 0.3267,
+      "step": 35280
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 33.066993713378906,
+      "learning_rate": 6.7156032373423685e-06,
+      "loss": 0.367,
+      "step": 35290
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 5.983303546905518,
+      "learning_rate": 6.71183888575193e-06,
+      "loss": 0.8041,
+      "step": 35300
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 34.84541320800781,
+      "learning_rate": 6.7080745341614914e-06,
+      "loss": 0.4726,
+      "step": 35310
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 15.427559852600098,
+      "learning_rate": 6.704310182571052e-06,
+      "loss": 0.2845,
+      "step": 35320
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 8.32028865814209,
+      "learning_rate": 6.7005458309806135e-06,
+      "loss": 0.5386,
+      "step": 35330
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 49.66559982299805,
+      "learning_rate": 6.696781479390175e-06,
+      "loss": 0.4041,
+      "step": 35340
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 3.6853106021881104,
+      "learning_rate": 6.693017127799737e-06,
+      "loss": 0.6245,
+      "step": 35350
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 2.9729301929473877,
+      "learning_rate": 6.689252776209299e-06,
+      "loss": 0.5603,
+      "step": 35360
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 7.048079490661621,
+      "learning_rate": 6.68548842461886e-06,
+      "loss": 0.5224,
+      "step": 35370
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 4.4709296226501465,
+      "learning_rate": 6.681724073028422e-06,
+      "loss": 0.2556,
+      "step": 35380
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 29.859085083007812,
+      "learning_rate": 6.677959721437983e-06,
+      "loss": 0.5444,
+      "step": 35390
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 6.1535797119140625,
+      "learning_rate": 6.6741953698475446e-06,
+      "loss": 0.5421,
+      "step": 35400
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 5.314403533935547,
+      "learning_rate": 6.670431018257105e-06,
+      "loss": 0.5939,
+      "step": 35410
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 17.58977508544922,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.4557,
+      "step": 35420
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 1.5328501462936401,
+      "learning_rate": 6.662902315076228e-06,
+      "loss": 0.4192,
+      "step": 35430
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 16.774066925048828,
+      "learning_rate": 6.65913796348579e-06,
+      "loss": 0.644,
+      "step": 35440
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 56.260440826416016,
+      "learning_rate": 6.655373611895352e-06,
+      "loss": 0.7726,
+      "step": 35450
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 20.503341674804688,
+      "learning_rate": 6.651609260304913e-06,
+      "loss": 0.5263,
+      "step": 35460
+    },
+    {
+      "epoch": 6.68,
+      "grad_norm": 4.7927422523498535,
+      "learning_rate": 6.647844908714475e-06,
+      "loss": 0.5455,
+      "step": 35470
+    },
+    {
+      "epoch": 6.68,
+      "grad_norm": 0.0458342470228672,
+      "learning_rate": 6.644080557124036e-06,
+      "loss": 0.3213,
+      "step": 35480
+    },
+    {
+      "epoch": 6.68,
+      "grad_norm": 12.544048309326172,
+      "learning_rate": 6.640316205533598e-06,
+      "loss": 0.7539,
+      "step": 35490
+    },
+    {
+      "epoch": 6.68,
+      "grad_norm": 23.94100570678711,
+      "learning_rate": 6.636551853943158e-06,
+      "loss": 0.7003,
+      "step": 35500
+    },
+    {
+      "epoch": 6.68,
+      "grad_norm": 1.0551177263259888,
+      "learning_rate": 6.63278750235272e-06,
+      "loss": 0.3386,
+      "step": 35510
+    },
+    {
+      "epoch": 6.69,
+      "grad_norm": 18.300785064697266,
+      "learning_rate": 6.629023150762281e-06,
+      "loss": 0.5915,
+      "step": 35520
+    },
+    {
+      "epoch": 6.69,
+      "grad_norm": 2.241436243057251,
+      "learning_rate": 6.625258799171843e-06,
+      "loss": 0.3854,
+      "step": 35530
+    },
+    {
+      "epoch": 6.69,
+      "grad_norm": 2.476534366607666,
+      "learning_rate": 6.621494447581405e-06,
+      "loss": 0.5499,
+      "step": 35540
+    },
+    {
+      "epoch": 6.69,
+      "grad_norm": 14.764324188232422,
+      "learning_rate": 6.6177300959909665e-06,
+      "loss": 0.4017,
+      "step": 35550
+    },
+    {
+      "epoch": 6.69,
+      "grad_norm": 20.741302490234375,
+      "learning_rate": 6.613965744400528e-06,
+      "loss": 0.533,
+      "step": 35560
+    },
+    {
+      "epoch": 6.69,
+      "grad_norm": 20.3760929107666,
+      "learning_rate": 6.610201392810089e-06,
+      "loss": 0.7259,
+      "step": 35570
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 22.055103302001953,
+      "learning_rate": 6.60643704121965e-06,
+      "loss": 0.7339,
+      "step": 35580
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 0.13808929920196533,
+      "learning_rate": 6.6026726896292115e-06,
+      "loss": 0.2974,
+      "step": 35590
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 0.07140693813562393,
+      "learning_rate": 6.598908338038773e-06,
+      "loss": 0.3145,
+      "step": 35600
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 24.929866790771484,
+      "learning_rate": 6.595143986448334e-06,
+      "loss": 0.2889,
+      "step": 35610
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 16.1369571685791,
+      "learning_rate": 6.591379634857896e-06,
+      "loss": 1.012,
+      "step": 35620
+    },
+    {
+      "epoch": 6.71,
+      "grad_norm": 29.283899307250977,
+      "learning_rate": 6.587615283267458e-06,
+      "loss": 0.5044,
+      "step": 35630
+    },
+    {
+      "epoch": 6.71,
+      "grad_norm": 8.903387069702148,
+      "learning_rate": 6.58385093167702e-06,
+      "loss": 0.3735,
+      "step": 35640
+    },
+    {
+      "epoch": 6.71,
+      "grad_norm": 32.65473937988281,
+      "learning_rate": 6.580086580086581e-06,
+      "loss": 0.3131,
+      "step": 35650
+    },
+    {
+      "epoch": 6.71,
+      "grad_norm": 3.421217203140259,
+      "learning_rate": 6.5763222284961426e-06,
+      "loss": 0.4596,
+      "step": 35660
+    },
+    {
+      "epoch": 6.71,
+      "grad_norm": 22.2224178314209,
+      "learning_rate": 6.572557876905703e-06,
+      "loss": 0.6354,
+      "step": 35670
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 10.610343933105469,
+      "learning_rate": 6.568793525315265e-06,
+      "loss": 0.6314,
+      "step": 35680
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 18.17767906188965,
+      "learning_rate": 6.565029173724826e-06,
+      "loss": 0.4187,
+      "step": 35690
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 11.569199562072754,
+      "learning_rate": 6.5612648221343875e-06,
+      "loss": 0.3357,
+      "step": 35700
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 2.6818912029266357,
+      "learning_rate": 6.557500470543949e-06,
+      "loss": 0.3558,
+      "step": 35710
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 7.722105026245117,
+      "learning_rate": 6.553736118953511e-06,
+      "loss": 0.3008,
+      "step": 35720
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 47.493011474609375,
+      "learning_rate": 6.549971767363073e-06,
+      "loss": 0.6678,
+      "step": 35730
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 21.441015243530273,
+      "learning_rate": 6.546207415772634e-06,
+      "loss": 0.5969,
+      "step": 35740
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 0.7386395931243896,
+      "learning_rate": 6.542443064182196e-06,
+      "loss": 0.6374,
+      "step": 35750
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 12.457733154296875,
+      "learning_rate": 6.538678712591756e-06,
+      "loss": 0.5366,
+      "step": 35760
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 16.376632690429688,
+      "learning_rate": 6.534914361001318e-06,
+      "loss": 0.3374,
+      "step": 35770
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 16.77738380432129,
+      "learning_rate": 6.531150009410879e-06,
+      "loss": 0.3973,
+      "step": 35780
+    },
+    {
+      "epoch": 6.74,
+      "grad_norm": 3.2299344539642334,
+      "learning_rate": 6.527385657820441e-06,
+      "loss": 0.3064,
+      "step": 35790
+    },
+    {
+      "epoch": 6.74,
+      "grad_norm": 29.193862915039062,
+      "learning_rate": 6.523621306230002e-06,
+      "loss": 0.5058,
+      "step": 35800
+    },
+    {
+      "epoch": 6.74,
+      "grad_norm": 0.34984686970710754,
+      "learning_rate": 6.519856954639564e-06,
+      "loss": 0.4511,
+      "step": 35810
+    },
+    {
+      "epoch": 6.74,
+      "grad_norm": 14.813493728637695,
+      "learning_rate": 6.516092603049126e-06,
+      "loss": 0.4569,
+      "step": 35820
+    },
+    {
+      "epoch": 6.74,
+      "grad_norm": 15.381288528442383,
+      "learning_rate": 6.512328251458687e-06,
+      "loss": 0.5122,
+      "step": 35830
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 4.633648872375488,
+      "learning_rate": 6.508563899868249e-06,
+      "loss": 0.4984,
+      "step": 35840
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 9.600768089294434,
+      "learning_rate": 6.5047995482778095e-06,
+      "loss": 0.4427,
+      "step": 35850
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 10.763100624084473,
+      "learning_rate": 6.501035196687371e-06,
+      "loss": 0.4402,
+      "step": 35860
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 11.68506145477295,
+      "learning_rate": 6.497270845096932e-06,
+      "loss": 0.4209,
+      "step": 35870
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 14.833422660827637,
+      "learning_rate": 6.493506493506494e-06,
+      "loss": 0.6574,
+      "step": 35880
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 4.88935661315918,
+      "learning_rate": 6.489742141916055e-06,
+      "loss": 0.4479,
+      "step": 35890
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 10.467138290405273,
+      "learning_rate": 6.485977790325617e-06,
+      "loss": 0.5853,
+      "step": 35900
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 1.4499626159667969,
+      "learning_rate": 6.482213438735179e-06,
+      "loss": 0.4039,
+      "step": 35910
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 14.744542121887207,
+      "learning_rate": 6.4784490871447405e-06,
+      "loss": 0.333,
+      "step": 35920
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 0.04979345574975014,
+      "learning_rate": 6.474684735554301e-06,
+      "loss": 0.4814,
+      "step": 35930
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 15.655257225036621,
+      "learning_rate": 6.470920383963863e-06,
+      "loss": 0.4449,
+      "step": 35940
+    },
+    {
+      "epoch": 6.77,
+      "grad_norm": 0.3292839229106903,
+      "learning_rate": 6.467156032373424e-06,
+      "loss": 0.2991,
+      "step": 35950
+    },
+    {
+      "epoch": 6.77,
+      "grad_norm": 22.464662551879883,
+      "learning_rate": 6.4633916807829855e-06,
+      "loss": 0.9058,
+      "step": 35960
+    },
+    {
+      "epoch": 6.77,
+      "grad_norm": 26.670724868774414,
+      "learning_rate": 6.459627329192547e-06,
+      "loss": 0.5506,
+      "step": 35970
+    },
+    {
+      "epoch": 6.77,
+      "grad_norm": 20.745296478271484,
+      "learning_rate": 6.4558629776021084e-06,
+      "loss": 0.2994,
+      "step": 35980
+    },
+    {
+      "epoch": 6.77,
+      "grad_norm": 25.35480499267578,
+      "learning_rate": 6.45209862601167e-06,
+      "loss": 0.5347,
+      "step": 35990
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 0.18468712270259857,
+      "learning_rate": 6.448334274421232e-06,
+      "loss": 0.4672,
+      "step": 36000
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 27.896060943603516,
+      "learning_rate": 6.444569922830794e-06,
+      "loss": 0.7579,
+      "step": 36010
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 8.810677528381348,
+      "learning_rate": 6.440805571240354e-06,
+      "loss": 0.4414,
+      "step": 36020
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 32.96968078613281,
+      "learning_rate": 6.437041219649916e-06,
+      "loss": 0.336,
+      "step": 36030
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 35.82292556762695,
+      "learning_rate": 6.433276868059477e-06,
+      "loss": 0.5106,
+      "step": 36040
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 27.85030746459961,
+      "learning_rate": 6.429512516469039e-06,
+      "loss": 0.5962,
+      "step": 36050
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 27.592071533203125,
+      "learning_rate": 6.4257481648786e-06,
+      "loss": 0.7332,
+      "step": 36060
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 23.796886444091797,
+      "learning_rate": 6.421983813288162e-06,
+      "loss": 0.4692,
+      "step": 36070
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 14.783629417419434,
+      "learning_rate": 6.418219461697723e-06,
+      "loss": 0.4037,
+      "step": 36080
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 14.800186157226562,
+      "learning_rate": 6.414455110107285e-06,
+      "loss": 0.3501,
+      "step": 36090
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 31.679777145385742,
+      "learning_rate": 6.410690758516847e-06,
+      "loss": 0.3348,
+      "step": 36100
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 27.46072769165039,
+      "learning_rate": 6.406926406926407e-06,
+      "loss": 0.7376,
+      "step": 36110
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 5.877180099487305,
+      "learning_rate": 6.403162055335969e-06,
+      "loss": 0.6983,
+      "step": 36120
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 2.0558340549468994,
+      "learning_rate": 6.39939770374553e-06,
+      "loss": 0.499,
+      "step": 36130
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 7.006868839263916,
+      "learning_rate": 6.395633352155092e-06,
+      "loss": 0.544,
+      "step": 36140
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 15.996200561523438,
+      "learning_rate": 6.391869000564653e-06,
+      "loss": 0.3828,
+      "step": 36150
+    },
+    {
+      "epoch": 6.81,
+      "grad_norm": 11.61379337310791,
+      "learning_rate": 6.388104648974215e-06,
+      "loss": 0.4883,
+      "step": 36160
+    },
+    {
+      "epoch": 6.81,
+      "grad_norm": 7.964717388153076,
+      "learning_rate": 6.384340297383776e-06,
+      "loss": 0.3079,
+      "step": 36170
+    },
+    {
+      "epoch": 6.81,
+      "grad_norm": 0.344844788312912,
+      "learning_rate": 6.380575945793338e-06,
+      "loss": 0.5976,
+      "step": 36180
+    },
+    {
+      "epoch": 6.81,
+      "grad_norm": 5.613010883331299,
+      "learning_rate": 6.376811594202898e-06,
+      "loss": 0.5082,
+      "step": 36190
+    },
+    {
+      "epoch": 6.81,
+      "grad_norm": 33.136268615722656,
+      "learning_rate": 6.37304724261246e-06,
+      "loss": 0.4065,
+      "step": 36200
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 26.772239685058594,
+      "learning_rate": 6.369282891022022e-06,
+      "loss": 0.6971,
+      "step": 36210
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 21.456703186035156,
+      "learning_rate": 6.3655185394315835e-06,
+      "loss": 0.3556,
+      "step": 36220
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 26.07500648498535,
+      "learning_rate": 6.361754187841145e-06,
+      "loss": 0.5151,
+      "step": 36230
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 19.035114288330078,
+      "learning_rate": 6.3579898362507064e-06,
+      "loss": 0.897,
+      "step": 36240
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 11.198712348937988,
+      "learning_rate": 6.354225484660268e-06,
+      "loss": 0.6124,
+      "step": 36250
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 7.805197715759277,
+      "learning_rate": 6.350461133069829e-06,
+      "loss": 0.6911,
+      "step": 36260
+    },
+    {
+      "epoch": 6.83,
+      "grad_norm": 29.66564178466797,
+      "learning_rate": 6.346696781479391e-06,
+      "loss": 0.2914,
+      "step": 36270
+    },
+    {
+      "epoch": 6.83,
+      "grad_norm": 42.64144515991211,
+      "learning_rate": 6.342932429888951e-06,
+      "loss": 0.8512,
+      "step": 36280
+    },
+    {
+      "epoch": 6.83,
+      "grad_norm": 6.411401748657227,
+      "learning_rate": 6.339168078298513e-06,
+      "loss": 0.5668,
+      "step": 36290
+    },
+    {
+      "epoch": 6.83,
+      "grad_norm": 11.873297691345215,
+      "learning_rate": 6.335403726708075e-06,
+      "loss": 0.6048,
+      "step": 36300
+    },
+    {
+      "epoch": 6.83,
+      "grad_norm": 18.666488647460938,
+      "learning_rate": 6.331639375117637e-06,
+      "loss": 0.6428,
+      "step": 36310
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 13.76421070098877,
+      "learning_rate": 6.327875023527198e-06,
+      "loss": 0.4083,
+      "step": 36320
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 10.075658798217773,
+      "learning_rate": 6.3241106719367596e-06,
+      "loss": 0.3769,
+      "step": 36330
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 19.852584838867188,
+      "learning_rate": 6.320346320346321e-06,
+      "loss": 0.6485,
+      "step": 36340
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 15.061224937438965,
+      "learning_rate": 6.3165819687558825e-06,
+      "loss": 0.4621,
+      "step": 36350
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 15.020421981811523,
+      "learning_rate": 6.312817617165444e-06,
+      "loss": 0.5403,
+      "step": 36360
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 0.03311420977115631,
+      "learning_rate": 6.3090532655750046e-06,
+      "loss": 0.2076,
+      "step": 36370
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 40.68346405029297,
+      "learning_rate": 6.305288913984566e-06,
+      "loss": 0.529,
+      "step": 36380
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 6.1792521476745605,
+      "learning_rate": 6.3015245623941275e-06,
+      "loss": 0.5218,
+      "step": 36390
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 18.217483520507812,
+      "learning_rate": 6.29776021080369e-06,
+      "loss": 0.3166,
+      "step": 36400
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 0.7558429837226868,
+      "learning_rate": 6.293995859213251e-06,
+      "loss": 0.6442,
+      "step": 36410
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 2.1329829692840576,
+      "learning_rate": 6.290231507622813e-06,
+      "loss": 0.5574,
+      "step": 36420
+    },
+    {
+      "epoch": 6.86,
+      "grad_norm": 19.163063049316406,
+      "learning_rate": 6.286467156032374e-06,
+      "loss": 0.4113,
+      "step": 36430
+    },
+    {
+      "epoch": 6.86,
+      "grad_norm": 21.05096435546875,
+      "learning_rate": 6.282702804441936e-06,
+      "loss": 0.644,
+      "step": 36440
+    },
+    {
+      "epoch": 6.86,
+      "grad_norm": 15.979854583740234,
+      "learning_rate": 6.278938452851497e-06,
+      "loss": 0.4465,
+      "step": 36450
+    },
+    {
+      "epoch": 6.86,
+      "grad_norm": 0.3811950087547302,
+      "learning_rate": 6.275174101261058e-06,
+      "loss": 0.472,
+      "step": 36460
+    },
+    {
+      "epoch": 6.86,
+      "grad_norm": 2.4271154403686523,
+      "learning_rate": 6.271409749670619e-06,
+      "loss": 0.6266,
+      "step": 36470
+    },
+    {
+      "epoch": 6.87,
+      "grad_norm": 21.372413635253906,
+      "learning_rate": 6.267645398080181e-06,
+      "loss": 0.5725,
+      "step": 36480
+    },
+    {
+      "epoch": 6.87,
+      "grad_norm": 13.645919799804688,
+      "learning_rate": 6.263881046489743e-06,
+      "loss": 0.4645,
+      "step": 36490
+    },
+    {
+      "epoch": 6.87,
+      "grad_norm": 56.878578186035156,
+      "learning_rate": 6.260116694899304e-06,
+      "loss": 0.5547,
+      "step": 36500
+    },
+    {
+      "epoch": 6.87,
+      "grad_norm": 2.6566834449768066,
+      "learning_rate": 6.256352343308866e-06,
+      "loss": 0.3839,
+      "step": 36510
+    },
+    {
+      "epoch": 6.87,
+      "grad_norm": 18.267131805419922,
+      "learning_rate": 6.252587991718427e-06,
+      "loss": 0.3353,
+      "step": 36520
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 29.942182540893555,
+      "learning_rate": 6.248823640127989e-06,
+      "loss": 0.8058,
+      "step": 36530
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 44.51432418823242,
+      "learning_rate": 6.245059288537549e-06,
+      "loss": 0.7491,
+      "step": 36540
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 22.656213760375977,
+      "learning_rate": 6.241294936947111e-06,
+      "loss": 0.3901,
+      "step": 36550
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 4.600666046142578,
+      "learning_rate": 6.237530585356672e-06,
+      "loss": 0.4892,
+      "step": 36560
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 0.16996584832668304,
+      "learning_rate": 6.233766233766234e-06,
+      "loss": 0.3893,
+      "step": 36570
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 7.973287105560303,
+      "learning_rate": 6.230001882175796e-06,
+      "loss": 0.4334,
+      "step": 36580
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 6.781386375427246,
+      "learning_rate": 6.2262375305853575e-06,
+      "loss": 0.4581,
+      "step": 36590
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 17.54666519165039,
+      "learning_rate": 6.222473178994919e-06,
+      "loss": 0.351,
+      "step": 36600
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 1.0639405250549316,
+      "learning_rate": 6.2187088274044805e-06,
+      "loss": 0.514,
+      "step": 36610
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 0.03612279146909714,
+      "learning_rate": 6.214944475814042e-06,
+      "loss": 0.6789,
+      "step": 36620
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 13.270979881286621,
+      "learning_rate": 6.2111801242236025e-06,
+      "loss": 0.4822,
+      "step": 36630
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 1.6162692308425903,
+      "learning_rate": 6.207415772633164e-06,
+      "loss": 0.4827,
+      "step": 36640
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 28.315080642700195,
+      "learning_rate": 6.2036514210427255e-06,
+      "loss": 0.616,
+      "step": 36650
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 11.296304702758789,
+      "learning_rate": 6.199887069452287e-06,
+      "loss": 0.6246,
+      "step": 36660
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 10.235584259033203,
+      "learning_rate": 6.196122717861849e-06,
+      "loss": 0.433,
+      "step": 36670
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 26.53406524658203,
+      "learning_rate": 6.192358366271411e-06,
+      "loss": 0.3869,
+      "step": 36680
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 3.4729597568511963,
+      "learning_rate": 6.188594014680972e-06,
+      "loss": 0.3534,
+      "step": 36690
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 20.50278091430664,
+      "learning_rate": 6.184829663090534e-06,
+      "loss": 0.3664,
+      "step": 36700
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 7.9125566482543945,
+      "learning_rate": 6.181065311500095e-06,
+      "loss": 0.3867,
+      "step": 36710
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 17.5158634185791,
+      "learning_rate": 6.177300959909656e-06,
+      "loss": 0.315,
+      "step": 36720
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 30.33664321899414,
+      "learning_rate": 6.173536608319217e-06,
+      "loss": 0.4541,
+      "step": 36730
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 24.297866821289062,
+      "learning_rate": 6.169772256728779e-06,
+      "loss": 0.772,
+      "step": 36740
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 16.661861419677734,
+      "learning_rate": 6.16600790513834e-06,
+      "loss": 0.5398,
+      "step": 36750
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 0.9989810585975647,
+      "learning_rate": 6.1622435535479015e-06,
+      "loss": 0.4729,
+      "step": 36760
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 20.895919799804688,
+      "learning_rate": 6.158479201957464e-06,
+      "loss": 0.5922,
+      "step": 36770
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 13.476861953735352,
+      "learning_rate": 6.154714850367025e-06,
+      "loss": 0.6265,
+      "step": 36780
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 29.94508171081543,
+      "learning_rate": 6.150950498776587e-06,
+      "loss": 0.5422,
+      "step": 36790
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 6.413869380950928,
+      "learning_rate": 6.147186147186147e-06,
+      "loss": 0.4729,
+      "step": 36800
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 26.468042373657227,
+      "learning_rate": 6.143421795595709e-06,
+      "loss": 0.4204,
+      "step": 36810
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 5.048950672149658,
+      "learning_rate": 6.13965744400527e-06,
+      "loss": 0.5322,
+      "step": 36820
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 20.66354751586914,
+      "learning_rate": 6.135893092414832e-06,
+      "loss": 0.4366,
+      "step": 36830
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 14.539073944091797,
+      "learning_rate": 6.132128740824393e-06,
+      "loss": 0.5468,
+      "step": 36840
+    },
+    {
+      "epoch": 6.94,
+      "grad_norm": 19.072439193725586,
+      "learning_rate": 6.128364389233955e-06,
+      "loss": 0.2972,
+      "step": 36850
+    },
+    {
+      "epoch": 6.94,
+      "grad_norm": 0.2176668494939804,
+      "learning_rate": 6.124600037643517e-06,
+      "loss": 0.4319,
+      "step": 36860
+    },
+    {
+      "epoch": 6.94,
+      "grad_norm": 1.570940613746643,
+      "learning_rate": 6.1208356860530784e-06,
+      "loss": 0.2465,
+      "step": 36870
+    },
+    {
+      "epoch": 6.94,
+      "grad_norm": 0.4804990291595459,
+      "learning_rate": 6.11707133446264e-06,
+      "loss": 0.5241,
+      "step": 36880
+    },
+    {
+      "epoch": 6.94,
+      "grad_norm": 20.45847511291504,
+      "learning_rate": 6.1133069828722005e-06,
+      "loss": 0.5885,
+      "step": 36890
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 0.5880195498466492,
+      "learning_rate": 6.109542631281762e-06,
+      "loss": 0.6118,
+      "step": 36900
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 19.450361251831055,
+      "learning_rate": 6.1057782796913234e-06,
+      "loss": 0.6389,
+      "step": 36910
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 6.449407577514648,
+      "learning_rate": 6.102013928100885e-06,
+      "loss": 0.4828,
+      "step": 36920
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 19.77794075012207,
+      "learning_rate": 6.098249576510446e-06,
+      "loss": 0.5739,
+      "step": 36930
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 19.867918014526367,
+      "learning_rate": 6.094485224920008e-06,
+      "loss": 0.5181,
+      "step": 36940
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 11.12002944946289,
+      "learning_rate": 6.09072087332957e-06,
+      "loss": 0.526,
+      "step": 36950
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 2.037919521331787,
+      "learning_rate": 6.086956521739132e-06,
+      "loss": 0.4911,
+      "step": 36960
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 31.55478286743164,
+      "learning_rate": 6.083192170148693e-06,
+      "loss": 0.856,
+      "step": 36970
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 18.756759643554688,
+      "learning_rate": 6.079427818558254e-06,
+      "loss": 0.7955,
+      "step": 36980
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 18.737565994262695,
+      "learning_rate": 6.075663466967815e-06,
+      "loss": 0.5185,
+      "step": 36990
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 8.580978393554688,
+      "learning_rate": 6.071899115377377e-06,
+      "loss": 0.5442,
+      "step": 37000
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 16.809736251831055,
+      "learning_rate": 6.068134763786938e-06,
+      "loss": 0.7495,
+      "step": 37010
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 11.04421329498291,
+      "learning_rate": 6.0643704121964995e-06,
+      "loss": 0.2769,
+      "step": 37020
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 11.780550956726074,
+      "learning_rate": 6.060606060606061e-06,
+      "loss": 0.8881,
+      "step": 37030
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 1.5381724834442139,
+      "learning_rate": 6.0568417090156224e-06,
+      "loss": 0.4434,
+      "step": 37040
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 2.1745412349700928,
+      "learning_rate": 6.053077357425185e-06,
+      "loss": 0.3855,
+      "step": 37050
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 1.2057762145996094,
+      "learning_rate": 6.049313005834746e-06,
+      "loss": 0.5586,
+      "step": 37060
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 4.387953758239746,
+      "learning_rate": 6.045548654244307e-06,
+      "loss": 0.6215,
+      "step": 37070
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 15.834346771240234,
+      "learning_rate": 6.041784302653868e-06,
+      "loss": 0.6887,
+      "step": 37080
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 22.549379348754883,
+      "learning_rate": 6.03801995106343e-06,
+      "loss": 0.5226,
+      "step": 37090
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 21.140119552612305,
+      "learning_rate": 6.034255599472991e-06,
+      "loss": 0.8414,
+      "step": 37100
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 8.247200012207031,
+      "learning_rate": 6.030491247882553e-06,
+      "loss": 0.49,
+      "step": 37110
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 7.904429912567139,
+      "learning_rate": 6.026726896292114e-06,
+      "loss": 0.2786,
+      "step": 37120
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 26.10804557800293,
+      "learning_rate": 6.022962544701676e-06,
+      "loss": 0.3379,
+      "step": 37130
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 12.558807373046875,
+      "learning_rate": 6.019198193111238e-06,
+      "loss": 0.6979,
+      "step": 37140
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 4.539823055267334,
+      "learning_rate": 6.015433841520798e-06,
+      "loss": 0.4851,
+      "step": 37150
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 17.593862533569336,
+      "learning_rate": 6.01166948993036e-06,
+      "loss": 0.4093,
+      "step": 37160
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 6.471458435058594,
+      "learning_rate": 6.007905138339921e-06,
+      "loss": 0.4581,
+      "step": 37170
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 6.4814348220825195,
+      "learning_rate": 6.004140786749483e-06,
+      "loss": 0.4462,
+      "step": 37180
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.26597660779953003,
+      "learning_rate": 6.000376435159044e-06,
+      "loss": 0.303,
+      "step": 37190
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.9241333333333334,
+      "eval_loss": 0.30108216404914856,
+      "eval_runtime": 52.0717,
+      "eval_samples_per_second": 144.032,
+      "eval_steps_per_second": 18.014,
+      "step": 37191
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 4.4914164543151855,
+      "learning_rate": 5.996612083568606e-06,
+      "loss": 0.5521,
+      "step": 37200
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 20.4888973236084,
+      "learning_rate": 5.992847731978167e-06,
+      "loss": 0.3498,
+      "step": 37210
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 18.38087272644043,
+      "learning_rate": 5.989083380387729e-06,
+      "loss": 0.6529,
+      "step": 37220
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 9.481815338134766,
+      "learning_rate": 5.985319028797291e-06,
+      "loss": 0.5024,
+      "step": 37230
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 2.8433196544647217,
+      "learning_rate": 5.981554677206851e-06,
+      "loss": 0.3982,
+      "step": 37240
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 26.133316040039062,
+      "learning_rate": 5.977790325616413e-06,
+      "loss": 0.9014,
+      "step": 37250
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 17.805519104003906,
+      "learning_rate": 5.9740259740259746e-06,
+      "loss": 0.4383,
+      "step": 37260
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 8.60024642944336,
+      "learning_rate": 5.970261622435536e-06,
+      "loss": 0.7659,
+      "step": 37270
+    },
+    {
+      "epoch": 7.02,
+      "grad_norm": 1.4241914749145508,
+      "learning_rate": 5.9664972708450975e-06,
+      "loss": 0.5184,
+      "step": 37280
+    },
+    {
+      "epoch": 7.02,
+      "grad_norm": 3.293872356414795,
+      "learning_rate": 5.962732919254659e-06,
+      "loss": 0.4679,
+      "step": 37290
+    },
+    {
+      "epoch": 7.02,
+      "grad_norm": 9.334550857543945,
+      "learning_rate": 5.95896856766422e-06,
+      "loss": 0.3055,
+      "step": 37300
+    },
+    {
+      "epoch": 7.02,
+      "grad_norm": 12.182844161987305,
+      "learning_rate": 5.955204216073782e-06,
+      "loss": 0.4598,
+      "step": 37310
+    },
+    {
+      "epoch": 7.02,
+      "grad_norm": 21.008893966674805,
+      "learning_rate": 5.951439864483344e-06,
+      "loss": 0.5906,
+      "step": 37320
+    },
+    {
+      "epoch": 7.03,
+      "grad_norm": 21.775415420532227,
+      "learning_rate": 5.947675512892904e-06,
+      "loss": 0.4748,
+      "step": 37330
+    },
+    {
+      "epoch": 7.03,
+      "grad_norm": 33.95808792114258,
+      "learning_rate": 5.943911161302465e-06,
+      "loss": 0.6582,
+      "step": 37340
+    },
+    {
+      "epoch": 7.03,
+      "grad_norm": 28.52260971069336,
+      "learning_rate": 5.940146809712028e-06,
+      "loss": 0.4331,
+      "step": 37350
+    },
+    {
+      "epoch": 7.03,
+      "grad_norm": 0.2158019095659256,
+      "learning_rate": 5.936382458121589e-06,
+      "loss": 0.4771,
+      "step": 37360
+    },
+    {
+      "epoch": 7.03,
+      "grad_norm": 11.500349998474121,
+      "learning_rate": 5.932618106531151e-06,
+      "loss": 0.2201,
+      "step": 37370
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 13.823716163635254,
+      "learning_rate": 5.928853754940712e-06,
+      "loss": 0.429,
+      "step": 37380
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 20.627649307250977,
+      "learning_rate": 5.9250894033502736e-06,
+      "loss": 0.526,
+      "step": 37390
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 1.0602643489837646,
+      "learning_rate": 5.921325051759835e-06,
+      "loss": 0.3343,
+      "step": 37400
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 34.25511932373047,
+      "learning_rate": 5.917560700169396e-06,
+      "loss": 0.7096,
+      "step": 37410
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 58.332923889160156,
+      "learning_rate": 5.913796348578957e-06,
+      "loss": 0.5007,
+      "step": 37420
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 0.520494282245636,
+      "learning_rate": 5.9100319969885186e-06,
+      "loss": 0.6852,
+      "step": 37430
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 14.561441421508789,
+      "learning_rate": 5.906267645398081e-06,
+      "loss": 0.7221,
+      "step": 37440
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 7.289056777954102,
+      "learning_rate": 5.902503293807642e-06,
+      "loss": 0.4541,
+      "step": 37450
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 15.227532386779785,
+      "learning_rate": 5.898738942217204e-06,
+      "loss": 0.6697,
+      "step": 37460
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 33.787776947021484,
+      "learning_rate": 5.894974590626765e-06,
+      "loss": 0.5081,
+      "step": 37470
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 18.24729347229004,
+      "learning_rate": 5.891210239036327e-06,
+      "loss": 0.5493,
+      "step": 37480
+    },
+    {
+      "epoch": 7.06,
+      "grad_norm": 17.62848472595215,
+      "learning_rate": 5.887445887445888e-06,
+      "loss": 0.6633,
+      "step": 37490
+    },
+    {
+      "epoch": 7.06,
+      "grad_norm": 13.691800117492676,
+      "learning_rate": 5.883681535855449e-06,
+      "loss": 0.5323,
+      "step": 37500
+    },
+    {
+      "epoch": 7.06,
+      "grad_norm": 33.989376068115234,
+      "learning_rate": 5.87991718426501e-06,
+      "loss": 0.5766,
+      "step": 37510
+    },
+    {
+      "epoch": 7.06,
+      "grad_norm": 6.813748359680176,
+      "learning_rate": 5.876152832674572e-06,
+      "loss": 0.4941,
+      "step": 37520
+    },
+    {
+      "epoch": 7.06,
+      "grad_norm": 10.321836471557617,
+      "learning_rate": 5.872388481084134e-06,
+      "loss": 0.6905,
+      "step": 37530
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 0.11116129904985428,
+      "learning_rate": 5.8686241294936955e-06,
+      "loss": 0.2757,
+      "step": 37540
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 15.543323516845703,
+      "learning_rate": 5.864859777903257e-06,
+      "loss": 0.5171,
+      "step": 37550
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 10.380474090576172,
+      "learning_rate": 5.861095426312818e-06,
+      "loss": 0.6989,
+      "step": 37560
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 4.593369007110596,
+      "learning_rate": 5.85733107472238e-06,
+      "loss": 0.3715,
+      "step": 37570
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 0.048009395599365234,
+      "learning_rate": 5.853566723131941e-06,
+      "loss": 0.4439,
+      "step": 37580
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 0.29102352261543274,
+      "learning_rate": 5.849802371541502e-06,
+      "loss": 0.3896,
+      "step": 37590
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 31.79117774963379,
+      "learning_rate": 5.846038019951063e-06,
+      "loss": 0.4598,
+      "step": 37600
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 0.6519231796264648,
+      "learning_rate": 5.842273668360625e-06,
+      "loss": 0.4029,
+      "step": 37610
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 0.04763523116707802,
+      "learning_rate": 5.838509316770186e-06,
+      "loss": 0.4446,
+      "step": 37620
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 5.059743881225586,
+      "learning_rate": 5.834744965179749e-06,
+      "loss": 0.2375,
+      "step": 37630
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 18.59803581237793,
+      "learning_rate": 5.83098061358931e-06,
+      "loss": 0.2464,
+      "step": 37640
+    },
+    {
+      "epoch": 7.09,
+      "grad_norm": 10.456582069396973,
+      "learning_rate": 5.8272162619988715e-06,
+      "loss": 0.3523,
+      "step": 37650
+    },
+    {
+      "epoch": 7.09,
+      "grad_norm": 32.90705871582031,
+      "learning_rate": 5.823451910408433e-06,
+      "loss": 0.5078,
+      "step": 37660
+    },
+    {
+      "epoch": 7.09,
+      "grad_norm": 3.806859254837036,
+      "learning_rate": 5.8196875588179945e-06,
+      "loss": 0.3444,
+      "step": 37670
+    },
+    {
+      "epoch": 7.09,
+      "grad_norm": 0.30914196372032166,
+      "learning_rate": 5.815923207227555e-06,
+      "loss": 0.4412,
+      "step": 37680
+    },
+    {
+      "epoch": 7.09,
+      "grad_norm": 1.1022928953170776,
+      "learning_rate": 5.8121588556371165e-06,
+      "loss": 0.4646,
+      "step": 37690
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 2.6765952110290527,
+      "learning_rate": 5.808394504046678e-06,
+      "loss": 0.4558,
+      "step": 37700
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 18.882295608520508,
+      "learning_rate": 5.8046301524562395e-06,
+      "loss": 0.3892,
+      "step": 37710
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 23.711225509643555,
+      "learning_rate": 5.800865800865802e-06,
+      "loss": 0.4363,
+      "step": 37720
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 15.44466781616211,
+      "learning_rate": 5.797101449275363e-06,
+      "loss": 0.3836,
+      "step": 37730
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 42.25343322753906,
+      "learning_rate": 5.793337097684925e-06,
+      "loss": 0.3638,
+      "step": 37740
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 5.034309387207031,
+      "learning_rate": 5.789572746094486e-06,
+      "loss": 0.4246,
+      "step": 37750
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 8.63538646697998,
+      "learning_rate": 5.785808394504047e-06,
+      "loss": 0.3743,
+      "step": 37760
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 7.230229377746582,
+      "learning_rate": 5.782044042913608e-06,
+      "loss": 0.3815,
+      "step": 37770
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 45.37762451171875,
+      "learning_rate": 5.77827969132317e-06,
+      "loss": 0.5341,
+      "step": 37780
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 13.632198333740234,
+      "learning_rate": 5.774515339732731e-06,
+      "loss": 0.5235,
+      "step": 37790
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 0.051364749670028687,
+      "learning_rate": 5.770750988142293e-06,
+      "loss": 0.2958,
+      "step": 37800
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 31.266891479492188,
+      "learning_rate": 5.766986636551855e-06,
+      "loss": 0.7418,
+      "step": 37810
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 21.40471649169922,
+      "learning_rate": 5.763222284961416e-06,
+      "loss": 0.5075,
+      "step": 37820
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 0.524456262588501,
+      "learning_rate": 5.759457933370978e-06,
+      "loss": 0.3901,
+      "step": 37830
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 4.217536926269531,
+      "learning_rate": 5.755693581780539e-06,
+      "loss": 0.5058,
+      "step": 37840
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 14.552117347717285,
+      "learning_rate": 5.7519292301901e-06,
+      "loss": 0.7943,
+      "step": 37850
+    },
+    {
+      "epoch": 7.13,
+      "grad_norm": 8.193163871765137,
+      "learning_rate": 5.748164878599661e-06,
+      "loss": 0.4663,
+      "step": 37860
+    },
+    {
+      "epoch": 7.13,
+      "grad_norm": 29.58894920349121,
+      "learning_rate": 5.744400527009223e-06,
+      "loss": 0.4782,
+      "step": 37870
+    },
+    {
+      "epoch": 7.13,
+      "grad_norm": 1.9362828731536865,
+      "learning_rate": 5.740636175418784e-06,
+      "loss": 0.4437,
+      "step": 37880
+    },
+    {
+      "epoch": 7.13,
+      "grad_norm": 11.760843276977539,
+      "learning_rate": 5.736871823828346e-06,
+      "loss": 0.6013,
+      "step": 37890
+    },
+    {
+      "epoch": 7.13,
+      "grad_norm": 11.666337966918945,
+      "learning_rate": 5.733107472237908e-06,
+      "loss": 0.477,
+      "step": 37900
+    },
+    {
+      "epoch": 7.14,
+      "grad_norm": 28.812885284423828,
+      "learning_rate": 5.7293431206474695e-06,
+      "loss": 0.5293,
+      "step": 37910
+    },
+    {
+      "epoch": 7.14,
+      "grad_norm": 38.73482894897461,
+      "learning_rate": 5.725578769057031e-06,
+      "loss": 0.5247,
+      "step": 37920
+    },
+    {
+      "epoch": 7.14,
+      "grad_norm": 14.522818565368652,
+      "learning_rate": 5.7218144174665924e-06,
+      "loss": 0.6026,
+      "step": 37930
+    },
+    {
+      "epoch": 7.14,
+      "grad_norm": 5.249240875244141,
+      "learning_rate": 5.718050065876153e-06,
+      "loss": 0.3719,
+      "step": 37940
+    },
+    {
+      "epoch": 7.14,
+      "grad_norm": 20.325544357299805,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.4992,
+      "step": 37950
+    },
+    {
+      "epoch": 7.14,
+      "grad_norm": 10.740631103515625,
+      "learning_rate": 5.710521362695276e-06,
+      "loss": 0.5879,
+      "step": 37960
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 12.676395416259766,
+      "learning_rate": 5.7067570111048374e-06,
+      "loss": 0.46,
+      "step": 37970
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 18.58064079284668,
+      "learning_rate": 5.702992659514399e-06,
+      "loss": 0.3504,
+      "step": 37980
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 4.198108196258545,
+      "learning_rate": 5.69922830792396e-06,
+      "loss": 0.5009,
+      "step": 37990
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 41.74122619628906,
+      "learning_rate": 5.695463956333523e-06,
+      "loss": 0.4566,
+      "step": 38000
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 4.8473219871521,
+      "learning_rate": 5.691699604743084e-06,
+      "loss": 0.437,
+      "step": 38010
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 0.8085455298423767,
+      "learning_rate": 5.687935253152645e-06,
+      "loss": 0.4446,
+      "step": 38020
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 12.186211585998535,
+      "learning_rate": 5.684170901562206e-06,
+      "loss": 0.3344,
+      "step": 38030
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 8.344589233398438,
+      "learning_rate": 5.680406549971768e-06,
+      "loss": 0.6024,
+      "step": 38040
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 16.539567947387695,
+      "learning_rate": 5.676642198381329e-06,
+      "loss": 0.4564,
+      "step": 38050
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 13.2172269821167,
+      "learning_rate": 5.672877846790891e-06,
+      "loss": 0.336,
+      "step": 38060
+    },
+    {
+      "epoch": 7.17,
+      "grad_norm": 0.8715003728866577,
+      "learning_rate": 5.669113495200452e-06,
+      "loss": 0.5934,
+      "step": 38070
+    },
+    {
+      "epoch": 7.17,
+      "grad_norm": 0.03967274725437164,
+      "learning_rate": 5.6653491436100135e-06,
+      "loss": 0.4955,
+      "step": 38080
+    },
+    {
+      "epoch": 7.17,
+      "grad_norm": 42.34977722167969,
+      "learning_rate": 5.661584792019576e-06,
+      "loss": 0.6749,
+      "step": 38090
+    },
+    {
+      "epoch": 7.17,
+      "grad_norm": 13.210389137268066,
+      "learning_rate": 5.657820440429137e-06,
+      "loss": 0.4131,
+      "step": 38100
+    },
+    {
+      "epoch": 7.17,
+      "grad_norm": 6.826159477233887,
+      "learning_rate": 5.654056088838698e-06,
+      "loss": 0.3849,
+      "step": 38110
+    },
+    {
+      "epoch": 7.17,
+      "grad_norm": 4.8396406173706055,
+      "learning_rate": 5.650291737248259e-06,
+      "loss": 0.4347,
+      "step": 38120
+    },
+    {
+      "epoch": 7.18,
+      "grad_norm": 0.492663711309433,
+      "learning_rate": 5.646527385657821e-06,
+      "loss": 0.462,
+      "step": 38130
+    },
+    {
+      "epoch": 7.18,
+      "grad_norm": 11.567475318908691,
+      "learning_rate": 5.642763034067382e-06,
+      "loss": 0.4399,
+      "step": 38140
+    },
+    {
+      "epoch": 7.18,
+      "grad_norm": 15.863377571105957,
+      "learning_rate": 5.638998682476944e-06,
+      "loss": 0.8885,
+      "step": 38150
+    },
+    {
+      "epoch": 7.18,
+      "grad_norm": 14.544013977050781,
+      "learning_rate": 5.635234330886505e-06,
+      "loss": 0.5326,
+      "step": 38160
+    },
+    {
+      "epoch": 7.18,
+      "grad_norm": 10.913338661193848,
+      "learning_rate": 5.631469979296067e-06,
+      "loss": 0.4607,
+      "step": 38170
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 9.55777645111084,
+      "learning_rate": 5.627705627705629e-06,
+      "loss": 0.2982,
+      "step": 38180
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 32.77585983276367,
+      "learning_rate": 5.62394127611519e-06,
+      "loss": 0.5706,
+      "step": 38190
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 17.475671768188477,
+      "learning_rate": 5.620176924524751e-06,
+      "loss": 0.5987,
+      "step": 38200
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 20.316225051879883,
+      "learning_rate": 5.6164125729343125e-06,
+      "loss": 0.2997,
+      "step": 38210
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 24.66366958618164,
+      "learning_rate": 5.612648221343874e-06,
+      "loss": 0.5617,
+      "step": 38220
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 2.09023380279541,
+      "learning_rate": 5.608883869753435e-06,
+      "loss": 0.1539,
+      "step": 38230
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 15.220237731933594,
+      "learning_rate": 5.605119518162997e-06,
+      "loss": 0.426,
+      "step": 38240
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 2.988645076751709,
+      "learning_rate": 5.601355166572558e-06,
+      "loss": 0.4009,
+      "step": 38250
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 10.845462799072266,
+      "learning_rate": 5.59759081498212e-06,
+      "loss": 0.3647,
+      "step": 38260
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 24.533771514892578,
+      "learning_rate": 5.593826463391681e-06,
+      "loss": 0.6294,
+      "step": 38270
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 33.65419006347656,
+      "learning_rate": 5.590062111801242e-06,
+      "loss": 0.608,
+      "step": 38280
+    },
+    {
+      "epoch": 7.21,
+      "grad_norm": 22.451557159423828,
+      "learning_rate": 5.586297760210803e-06,
+      "loss": 0.421,
+      "step": 38290
+    },
+    {
+      "epoch": 7.21,
+      "grad_norm": 10.964974403381348,
+      "learning_rate": 5.582533408620366e-06,
+      "loss": 0.4638,
+      "step": 38300
+    },
+    {
+      "epoch": 7.21,
+      "grad_norm": 0.03477161377668381,
+      "learning_rate": 5.578769057029927e-06,
+      "loss": 0.4434,
+      "step": 38310
+    },
+    {
+      "epoch": 7.21,
+      "grad_norm": 0.03130391240119934,
+      "learning_rate": 5.5750047054394886e-06,
+      "loss": 0.65,
+      "step": 38320
+    },
+    {
+      "epoch": 7.21,
+      "grad_norm": 7.138416290283203,
+      "learning_rate": 5.57124035384905e-06,
+      "loss": 0.4412,
+      "step": 38330
+    },
+    {
+      "epoch": 7.22,
+      "grad_norm": 0.3959789574146271,
+      "learning_rate": 5.5674760022586115e-06,
+      "loss": 0.3357,
+      "step": 38340
+    },
+    {
+      "epoch": 7.22,
+      "grad_norm": 14.614726066589355,
+      "learning_rate": 5.563711650668173e-06,
+      "loss": 0.6037,
+      "step": 38350
+    },
+    {
+      "epoch": 7.22,
+      "grad_norm": 9.566722869873047,
+      "learning_rate": 5.559947299077734e-06,
+      "loss": 0.4153,
+      "step": 38360
+    },
+    {
+      "epoch": 7.22,
+      "grad_norm": 24.345317840576172,
+      "learning_rate": 5.556182947487295e-06,
+      "loss": 0.3658,
+      "step": 38370
+    },
+    {
+      "epoch": 7.22,
+      "grad_norm": 45.182655334472656,
+      "learning_rate": 5.5524185958968565e-06,
+      "loss": 0.3317,
+      "step": 38380
+    },
+    {
+      "epoch": 7.23,
+      "grad_norm": 20.688343048095703,
+      "learning_rate": 5.548654244306419e-06,
+      "loss": 0.6442,
+      "step": 38390
+    },
+    {
+      "epoch": 7.23,
+      "grad_norm": 0.04111315310001373,
+      "learning_rate": 5.54488989271598e-06,
+      "loss": 0.6456,
+      "step": 38400
+    },
+    {
+      "epoch": 7.23,
+      "grad_norm": 25.130800247192383,
+      "learning_rate": 5.541125541125542e-06,
+      "loss": 0.4863,
+      "step": 38410
+    },
+    {
+      "epoch": 7.23,
+      "grad_norm": 8.105908393859863,
+      "learning_rate": 5.537361189535103e-06,
+      "loss": 0.6085,
+      "step": 38420
+    },
+    {
+      "epoch": 7.23,
+      "grad_norm": 20.2083740234375,
+      "learning_rate": 5.533596837944665e-06,
+      "loss": 0.5963,
+      "step": 38430
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 0.32264116406440735,
+      "learning_rate": 5.529832486354226e-06,
+      "loss": 0.2069,
+      "step": 38440
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 15.73507022857666,
+      "learning_rate": 5.5260681347637875e-06,
+      "loss": 0.3864,
+      "step": 38450
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 13.899942398071289,
+      "learning_rate": 5.522303783173348e-06,
+      "loss": 0.4887,
+      "step": 38460
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 29.511226654052734,
+      "learning_rate": 5.51853943158291e-06,
+      "loss": 0.4645,
+      "step": 38470
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 24.286212921142578,
+      "learning_rate": 5.514775079992472e-06,
+      "loss": 0.9881,
+      "step": 38480
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 7.7546491622924805,
+      "learning_rate": 5.511010728402033e-06,
+      "loss": 0.1187,
+      "step": 38490
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 2.176553726196289,
+      "learning_rate": 5.507246376811595e-06,
+      "loss": 0.2799,
+      "step": 38500
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 20.29352378845215,
+      "learning_rate": 5.503482025221156e-06,
+      "loss": 0.5774,
+      "step": 38510
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 11.283035278320312,
+      "learning_rate": 5.499717673630718e-06,
+      "loss": 0.6969,
+      "step": 38520
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 2.3551344871520996,
+      "learning_rate": 5.495953322040279e-06,
+      "loss": 0.6055,
+      "step": 38530
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 17.497970581054688,
+      "learning_rate": 5.492188970449841e-06,
+      "loss": 0.5596,
+      "step": 38540
+    },
+    {
+      "epoch": 7.26,
+      "grad_norm": 0.6391315460205078,
+      "learning_rate": 5.488424618859401e-06,
+      "loss": 0.5974,
+      "step": 38550
+    },
+    {
+      "epoch": 7.26,
+      "grad_norm": 10.734766960144043,
+      "learning_rate": 5.484660267268963e-06,
+      "loss": 0.5015,
+      "step": 38560
+    },
+    {
+      "epoch": 7.26,
+      "grad_norm": 28.943450927734375,
+      "learning_rate": 5.480895915678524e-06,
+      "loss": 0.291,
+      "step": 38570
+    },
+    {
+      "epoch": 7.26,
+      "grad_norm": 30.02581024169922,
+      "learning_rate": 5.4771315640880865e-06,
+      "loss": 0.5427,
+      "step": 38580
+    },
+    {
+      "epoch": 7.26,
+      "grad_norm": 2.238271474838257,
+      "learning_rate": 5.473367212497648e-06,
+      "loss": 0.5335,
+      "step": 38590
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 23.164579391479492,
+      "learning_rate": 5.4696028609072095e-06,
+      "loss": 0.6102,
+      "step": 38600
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 15.776400566101074,
+      "learning_rate": 5.465838509316771e-06,
+      "loss": 0.5805,
+      "step": 38610
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 15.679583549499512,
+      "learning_rate": 5.462074157726332e-06,
+      "loss": 0.667,
+      "step": 38620
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 13.96922492980957,
+      "learning_rate": 5.458309806135893e-06,
+      "loss": 0.1757,
+      "step": 38630
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 20.9793643951416,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 0.4583,
+      "step": 38640
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 13.186413764953613,
+      "learning_rate": 5.450781102955016e-06,
+      "loss": 0.3714,
+      "step": 38650
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 0.07995478063821793,
+      "learning_rate": 5.447016751364577e-06,
+      "loss": 0.3397,
+      "step": 38660
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 0.7243136167526245,
+      "learning_rate": 5.44325239977414e-06,
+      "loss": 0.4177,
+      "step": 38670
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 19.715696334838867,
+      "learning_rate": 5.439488048183701e-06,
+      "loss": 0.5963,
+      "step": 38680
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 13.43181037902832,
+      "learning_rate": 5.435723696593263e-06,
+      "loss": 0.4663,
+      "step": 38690
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 21.230018615722656,
+      "learning_rate": 5.431959345002824e-06,
+      "loss": 0.3512,
+      "step": 38700
+    },
+    {
+      "epoch": 7.29,
+      "grad_norm": 20.25225067138672,
+      "learning_rate": 5.4281949934123855e-06,
+      "loss": 0.7707,
+      "step": 38710
+    },
+    {
+      "epoch": 7.29,
+      "grad_norm": 0.15251778066158295,
+      "learning_rate": 5.424430641821946e-06,
+      "loss": 0.3512,
+      "step": 38720
+    },
+    {
+      "epoch": 7.29,
+      "grad_norm": 25.118371963500977,
+      "learning_rate": 5.420666290231508e-06,
+      "loss": 0.5093,
+      "step": 38730
+    },
+    {
+      "epoch": 7.29,
+      "grad_norm": 5.278887748718262,
+      "learning_rate": 5.416901938641069e-06,
+      "loss": 0.5096,
+      "step": 38740
+    },
+    {
+      "epoch": 7.29,
+      "grad_norm": 2.8506057262420654,
+      "learning_rate": 5.4131375870506305e-06,
+      "loss": 0.4723,
+      "step": 38750
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 12.97459602355957,
+      "learning_rate": 5.409373235460193e-06,
+      "loss": 0.74,
+      "step": 38760
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 12.791807174682617,
+      "learning_rate": 5.405608883869754e-06,
+      "loss": 0.5725,
+      "step": 38770
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 0.048314113169908524,
+      "learning_rate": 5.401844532279316e-06,
+      "loss": 0.5425,
+      "step": 38780
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 2.543400287628174,
+      "learning_rate": 5.398080180688877e-06,
+      "loss": 0.5112,
+      "step": 38790
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 8.345157623291016,
+      "learning_rate": 5.394315829098439e-06,
+      "loss": 0.2803,
+      "step": 38800
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 37.12696838378906,
+      "learning_rate": 5.390551477507999e-06,
+      "loss": 0.609,
+      "step": 38810
+    },
+    {
+      "epoch": 7.31,
+      "grad_norm": 15.115370750427246,
+      "learning_rate": 5.386787125917561e-06,
+      "loss": 0.4371,
+      "step": 38820
+    },
+    {
+      "epoch": 7.31,
+      "grad_norm": 24.323984146118164,
+      "learning_rate": 5.383022774327122e-06,
+      "loss": 0.7161,
+      "step": 38830
+    },
+    {
+      "epoch": 7.31,
+      "grad_norm": 27.755170822143555,
+      "learning_rate": 5.379258422736684e-06,
+      "loss": 0.559,
+      "step": 38840
+    },
+    {
+      "epoch": 7.31,
+      "grad_norm": 14.916714668273926,
+      "learning_rate": 5.375494071146246e-06,
+      "loss": 0.3077,
+      "step": 38850
+    },
+    {
+      "epoch": 7.31,
+      "grad_norm": 7.22851037979126,
+      "learning_rate": 5.3717297195558074e-06,
+      "loss": 0.4101,
+      "step": 38860
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 17.815229415893555,
+      "learning_rate": 5.367965367965369e-06,
+      "loss": 0.6718,
+      "step": 38870
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 43.300785064697266,
+      "learning_rate": 5.36420101637493e-06,
+      "loss": 0.2887,
+      "step": 38880
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 44.66895294189453,
+      "learning_rate": 5.360436664784491e-06,
+      "loss": 0.4488,
+      "step": 38890
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 13.247306823730469,
+      "learning_rate": 5.3566723131940524e-06,
+      "loss": 0.4967,
+      "step": 38900
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 9.896644592285156,
+      "learning_rate": 5.352907961603614e-06,
+      "loss": 0.5352,
+      "step": 38910
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 15.37808609008789,
+      "learning_rate": 5.349143610013175e-06,
+      "loss": 0.4616,
+      "step": 38920
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 21.590145111083984,
+      "learning_rate": 5.345379258422737e-06,
+      "loss": 0.8465,
+      "step": 38930
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 3.483672618865967,
+      "learning_rate": 5.341614906832298e-06,
+      "loss": 0.3051,
+      "step": 38940
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 24.26028060913086,
+      "learning_rate": 5.337850555241861e-06,
+      "loss": 0.3835,
+      "step": 38950
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 7.518918991088867,
+      "learning_rate": 5.334086203651422e-06,
+      "loss": 0.3848,
+      "step": 38960
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 25.557096481323242,
+      "learning_rate": 5.3303218520609835e-06,
+      "loss": 0.511,
+      "step": 38970
+    },
+    {
+      "epoch": 7.34,
+      "grad_norm": 5.811315059661865,
+      "learning_rate": 5.326557500470544e-06,
+      "loss": 0.3622,
+      "step": 38980
+    },
+    {
+      "epoch": 7.34,
+      "grad_norm": 0.346421480178833,
+      "learning_rate": 5.322793148880106e-06,
+      "loss": 0.4936,
+      "step": 38990
+    },
+    {
+      "epoch": 7.34,
+      "grad_norm": 28.79204559326172,
+      "learning_rate": 5.319028797289667e-06,
+      "loss": 0.8451,
+      "step": 39000
+    },
+    {
+      "epoch": 7.34,
+      "grad_norm": 44.75595474243164,
+      "learning_rate": 5.3152644456992285e-06,
+      "loss": 0.4559,
+      "step": 39010
+    },
+    {
+      "epoch": 7.34,
+      "grad_norm": 10.665853500366211,
+      "learning_rate": 5.31150009410879e-06,
+      "loss": 0.5485,
+      "step": 39020
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 11.347060203552246,
+      "learning_rate": 5.307735742518351e-06,
+      "loss": 0.5688,
+      "step": 39030
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 0.026885811239480972,
+      "learning_rate": 5.303971390927914e-06,
+      "loss": 0.6052,
+      "step": 39040
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 17.71526336669922,
+      "learning_rate": 5.300207039337475e-06,
+      "loss": 0.2693,
+      "step": 39050
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 19.759750366210938,
+      "learning_rate": 5.296442687747037e-06,
+      "loss": 0.4032,
+      "step": 39060
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 7.999401569366455,
+      "learning_rate": 5.292678336156597e-06,
+      "loss": 0.4863,
+      "step": 39070
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 0.07535995543003082,
+      "learning_rate": 5.288913984566159e-06,
+      "loss": 0.3263,
+      "step": 39080
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 15.817834854125977,
+      "learning_rate": 5.28514963297572e-06,
+      "loss": 0.7172,
+      "step": 39090
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 7.389345169067383,
+      "learning_rate": 5.281385281385282e-06,
+      "loss": 0.737,
+      "step": 39100
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 1.000137209892273,
+      "learning_rate": 5.277620929794843e-06,
+      "loss": 0.5336,
+      "step": 39110
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 0.07644683867692947,
+      "learning_rate": 5.2738565782044046e-06,
+      "loss": 0.3762,
+      "step": 39120
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 34.44324493408203,
+      "learning_rate": 5.270092226613967e-06,
+      "loss": 0.434,
+      "step": 39130
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 15.976019859313965,
+      "learning_rate": 5.266327875023528e-06,
+      "loss": 0.3836,
+      "step": 39140
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 0.10559114813804626,
+      "learning_rate": 5.26256352343309e-06,
+      "loss": 0.3044,
+      "step": 39150
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 13.347066879272461,
+      "learning_rate": 5.25879917184265e-06,
+      "loss": 0.5778,
+      "step": 39160
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 49.38035583496094,
+      "learning_rate": 5.255034820252212e-06,
+      "loss": 0.3162,
+      "step": 39170
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 3.3886990547180176,
+      "learning_rate": 5.251270468661773e-06,
+      "loss": 0.3553,
+      "step": 39180
+    },
+    {
+      "epoch": 7.38,
+      "grad_norm": 9.105963706970215,
+      "learning_rate": 5.247506117071335e-06,
+      "loss": 0.7682,
+      "step": 39190
+    },
+    {
+      "epoch": 7.38,
+      "grad_norm": 14.80863094329834,
+      "learning_rate": 5.243741765480896e-06,
+      "loss": 0.9301,
+      "step": 39200
+    },
+    {
+      "epoch": 7.38,
+      "grad_norm": 3.077606201171875,
+      "learning_rate": 5.239977413890458e-06,
+      "loss": 0.5738,
+      "step": 39210
+    },
+    {
+      "epoch": 7.38,
+      "grad_norm": 9.167863845825195,
+      "learning_rate": 5.236213062300019e-06,
+      "loss": 0.6652,
+      "step": 39220
+    },
+    {
+      "epoch": 7.38,
+      "grad_norm": 39.81913375854492,
+      "learning_rate": 5.2324487107095815e-06,
+      "loss": 0.6311,
+      "step": 39230
+    },
+    {
+      "epoch": 7.39,
+      "grad_norm": 0.21679580211639404,
+      "learning_rate": 5.228684359119141e-06,
+      "loss": 0.4019,
+      "step": 39240
+    },
+    {
+      "epoch": 7.39,
+      "grad_norm": 42.18350601196289,
+      "learning_rate": 5.2249200075287036e-06,
+      "loss": 0.4451,
+      "step": 39250
+    },
+    {
+      "epoch": 7.39,
+      "grad_norm": 4.216889381408691,
+      "learning_rate": 5.221155655938265e-06,
+      "loss": 0.4791,
+      "step": 39260
+    },
+    {
+      "epoch": 7.39,
+      "grad_norm": 27.158920288085938,
+      "learning_rate": 5.2173913043478265e-06,
+      "loss": 0.632,
+      "step": 39270
+    },
+    {
+      "epoch": 7.39,
+      "grad_norm": 3.5005364418029785,
+      "learning_rate": 5.213626952757388e-06,
+      "loss": 0.3373,
+      "step": 39280
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 5.771264553070068,
+      "learning_rate": 5.209862601166949e-06,
+      "loss": 0.8069,
+      "step": 39290
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 21.01832389831543,
+      "learning_rate": 5.206098249576511e-06,
+      "loss": 0.6151,
+      "step": 39300
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 0.21687443554401398,
+      "learning_rate": 5.202333897986072e-06,
+      "loss": 0.4028,
+      "step": 39310
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 17.40138816833496,
+      "learning_rate": 5.198569546395635e-06,
+      "loss": 0.3488,
+      "step": 39320
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 18.54778289794922,
+      "learning_rate": 5.194805194805194e-06,
+      "loss": 0.4226,
+      "step": 39330
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 0.8926262259483337,
+      "learning_rate": 5.191040843214757e-06,
+      "loss": 0.4072,
+      "step": 39340
+    },
+    {
+      "epoch": 7.41,
+      "grad_norm": 0.043081410229206085,
+      "learning_rate": 5.187276491624318e-06,
+      "loss": 0.4525,
+      "step": 39350
+    },
+    {
+      "epoch": 7.41,
+      "grad_norm": 28.826021194458008,
+      "learning_rate": 5.18351214003388e-06,
+      "loss": 0.4743,
+      "step": 39360
+    },
+    {
+      "epoch": 7.41,
+      "grad_norm": 19.19860076904297,
+      "learning_rate": 5.179747788443441e-06,
+      "loss": 0.6083,
+      "step": 39370
+    },
+    {
+      "epoch": 7.41,
+      "grad_norm": 23.61128044128418,
+      "learning_rate": 5.1759834368530025e-06,
+      "loss": 0.5457,
+      "step": 39380
+    },
+    {
+      "epoch": 7.41,
+      "grad_norm": 15.979894638061523,
+      "learning_rate": 5.172219085262564e-06,
+      "loss": 0.4737,
+      "step": 39390
+    },
+    {
+      "epoch": 7.42,
+      "grad_norm": 16.330854415893555,
+      "learning_rate": 5.1684547336721255e-06,
+      "loss": 0.6512,
+      "step": 39400
+    },
+    {
+      "epoch": 7.42,
+      "grad_norm": 3.0890164375305176,
+      "learning_rate": 5.164690382081688e-06,
+      "loss": 0.5336,
+      "step": 39410
+    },
+    {
+      "epoch": 7.42,
+      "grad_norm": 24.518455505371094,
+      "learning_rate": 5.1609260304912475e-06,
+      "loss": 0.4603,
+      "step": 39420
+    },
+    {
+      "epoch": 7.42,
+      "grad_norm": 0.07051704078912735,
+      "learning_rate": 5.15716167890081e-06,
+      "loss": 0.2793,
+      "step": 39430
+    },
+    {
+      "epoch": 7.42,
+      "grad_norm": 9.986491203308105,
+      "learning_rate": 5.153397327310371e-06,
+      "loss": 0.5185,
+      "step": 39440
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 6.8465142250061035,
+      "learning_rate": 5.149632975719933e-06,
+      "loss": 0.6369,
+      "step": 39450
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 0.4283851683139801,
+      "learning_rate": 5.145868624129494e-06,
+      "loss": 0.5151,
+      "step": 39460
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 13.111185073852539,
+      "learning_rate": 5.142104272539056e-06,
+      "loss": 0.4777,
+      "step": 39470
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 18.829418182373047,
+      "learning_rate": 5.138339920948617e-06,
+      "loss": 0.5422,
+      "step": 39480
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 18.628862380981445,
+      "learning_rate": 5.134575569358179e-06,
+      "loss": 0.5727,
+      "step": 39490
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 8.623425483703613,
+      "learning_rate": 5.130811217767739e-06,
+      "loss": 0.5578,
+      "step": 39500
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 7.155189514160156,
+      "learning_rate": 5.127046866177301e-06,
+      "loss": 0.6192,
+      "step": 39510
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 2.426950216293335,
+      "learning_rate": 5.123282514586862e-06,
+      "loss": 0.5366,
+      "step": 39520
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 5.574533939361572,
+      "learning_rate": 5.1195181629964245e-06,
+      "loss": 0.4428,
+      "step": 39530
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 0.15088945627212524,
+      "learning_rate": 5.115753811405986e-06,
+      "loss": 0.4344,
+      "step": 39540
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 53.62294387817383,
+      "learning_rate": 5.111989459815547e-06,
+      "loss": 0.3419,
+      "step": 39550
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 19.655786514282227,
+      "learning_rate": 5.108225108225109e-06,
+      "loss": 0.4297,
+      "step": 39560
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 9.113370895385742,
+      "learning_rate": 5.10446075663467e-06,
+      "loss": 0.4997,
+      "step": 39570
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 8.7140474319458,
+      "learning_rate": 5.100696405044232e-06,
+      "loss": 0.5054,
+      "step": 39580
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 0.8119073510169983,
+      "learning_rate": 5.096932053453792e-06,
+      "loss": 0.4112,
+      "step": 39590
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 24.722349166870117,
+      "learning_rate": 5.093167701863354e-06,
+      "loss": 0.4957,
+      "step": 39600
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 23.276451110839844,
+      "learning_rate": 5.089403350272915e-06,
+      "loss": 0.0708,
+      "step": 39610
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 10.128479957580566,
+      "learning_rate": 5.085638998682478e-06,
+      "loss": 0.3646,
+      "step": 39620
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 79.91693878173828,
+      "learning_rate": 5.081874647092039e-06,
+      "loss": 0.567,
+      "step": 39630
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 33.41339111328125,
+      "learning_rate": 5.0781102955016005e-06,
+      "loss": 0.3338,
+      "step": 39640
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 21.802555084228516,
+      "learning_rate": 5.074345943911162e-06,
+      "loss": 0.3191,
+      "step": 39650
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 18.43767738342285,
+      "learning_rate": 5.0705815923207234e-06,
+      "loss": 0.5815,
+      "step": 39660
+    },
+    {
+      "epoch": 7.47,
+      "grad_norm": 14.148484230041504,
+      "learning_rate": 5.066817240730285e-06,
+      "loss": 0.2298,
+      "step": 39670
+    },
+    {
+      "epoch": 7.47,
+      "grad_norm": 20.37163734436035,
+      "learning_rate": 5.0630528891398455e-06,
+      "loss": 0.8356,
+      "step": 39680
+    },
+    {
+      "epoch": 7.47,
+      "grad_norm": 6.101649284362793,
+      "learning_rate": 5.059288537549407e-06,
+      "loss": 0.4301,
+      "step": 39690
+    },
+    {
+      "epoch": 7.47,
+      "grad_norm": 1.8522956371307373,
+      "learning_rate": 5.0555241859589684e-06,
+      "loss": 0.3962,
+      "step": 39700
+    },
+    {
+      "epoch": 7.47,
+      "grad_norm": 37.072349548339844,
+      "learning_rate": 5.051759834368531e-06,
+      "loss": 0.4138,
+      "step": 39710
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 28.832242965698242,
+      "learning_rate": 5.047995482778092e-06,
+      "loss": 0.5587,
+      "step": 39720
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 29.438364028930664,
+      "learning_rate": 5.044231131187654e-06,
+      "loss": 0.3898,
+      "step": 39730
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 0.1980200558900833,
+      "learning_rate": 5.040466779597215e-06,
+      "loss": 0.1686,
+      "step": 39740
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 2.973400592803955,
+      "learning_rate": 5.036702428006777e-06,
+      "loss": 0.2205,
+      "step": 39750
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 41.77248001098633,
+      "learning_rate": 5.032938076416338e-06,
+      "loss": 0.5068,
+      "step": 39760
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 23.240005493164062,
+      "learning_rate": 5.029173724825899e-06,
+      "loss": 0.2396,
+      "step": 39770
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 29.466114044189453,
+      "learning_rate": 5.02540937323546e-06,
+      "loss": 0.7611,
+      "step": 39780
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 18.149154663085938,
+      "learning_rate": 5.021645021645022e-06,
+      "loss": 0.4035,
+      "step": 39790
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 15.7437105178833,
+      "learning_rate": 5.017880670054583e-06,
+      "loss": 0.5208,
+      "step": 39800
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 23.674476623535156,
+      "learning_rate": 5.014116318464145e-06,
+      "loss": 0.3757,
+      "step": 39810
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 2.990457057952881,
+      "learning_rate": 5.010351966873707e-06,
+      "loss": 0.3227,
+      "step": 39820
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 5.852288246154785,
+      "learning_rate": 5.006587615283268e-06,
+      "loss": 0.398,
+      "step": 39830
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 21.586061477661133,
+      "learning_rate": 5.00282326369283e-06,
+      "loss": 0.4684,
+      "step": 39840
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 18.58787727355957,
+      "learning_rate": 4.999058912102391e-06,
+      "loss": 0.4361,
+      "step": 39850
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 6.530487060546875,
+      "learning_rate": 4.995294560511953e-06,
+      "loss": 0.419,
+      "step": 39860
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 18.475072860717773,
+      "learning_rate": 4.991530208921513e-06,
+      "loss": 0.5548,
+      "step": 39870
+    },
+    {
+      "epoch": 7.51,
+      "grad_norm": 14.466314315795898,
+      "learning_rate": 4.987765857331075e-06,
+      "loss": 0.5023,
+      "step": 39880
+    },
+    {
+      "epoch": 7.51,
+      "grad_norm": 15.371254920959473,
+      "learning_rate": 4.984001505740636e-06,
+      "loss": 0.4835,
+      "step": 39890
+    },
+    {
+      "epoch": 7.51,
+      "grad_norm": 3.507498264312744,
+      "learning_rate": 4.9802371541501985e-06,
+      "loss": 0.4585,
+      "step": 39900
+    },
+    {
+      "epoch": 7.51,
+      "grad_norm": 30.916147232055664,
+      "learning_rate": 4.97647280255976e-06,
+      "loss": 0.5154,
+      "step": 39910
+    },
+    {
+      "epoch": 7.51,
+      "grad_norm": 14.572548866271973,
+      "learning_rate": 4.9727084509693206e-06,
+      "loss": 0.2884,
+      "step": 39920
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 0.2854940891265869,
+      "learning_rate": 4.968944099378882e-06,
+      "loss": 0.4503,
+      "step": 39930
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 30.854312896728516,
+      "learning_rate": 4.9651797477884435e-06,
+      "loss": 0.5331,
+      "step": 39940
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 19.74198341369629,
+      "learning_rate": 4.961415396198006e-06,
+      "loss": 0.1224,
+      "step": 39950
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 25.249618530273438,
+      "learning_rate": 4.957651044607566e-06,
+      "loss": 0.4178,
+      "step": 39960
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 8.16231632232666,
+      "learning_rate": 4.953886693017128e-06,
+      "loss": 0.5486,
+      "step": 39970
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 3.6326353549957275,
+      "learning_rate": 4.950122341426689e-06,
+      "loss": 0.5058,
+      "step": 39980
+    },
+    {
+      "epoch": 7.53,
+      "grad_norm": 20.812101364135742,
+      "learning_rate": 4.946357989836252e-06,
+      "loss": 0.5612,
+      "step": 39990
+    },
+    {
+      "epoch": 7.53,
+      "grad_norm": 24.15587615966797,
+      "learning_rate": 4.942593638245812e-06,
+      "loss": 0.5528,
+      "step": 40000
+    },
+    {
+      "epoch": 7.53,
+      "grad_norm": 0.3987553119659424,
+      "learning_rate": 4.938829286655374e-06,
+      "loss": 0.3515,
+      "step": 40010
+    },
+    {
+      "epoch": 7.53,
+      "grad_norm": 0.9960697889328003,
+      "learning_rate": 4.935064935064935e-06,
+      "loss": 0.3475,
+      "step": 40020
+    },
+    {
+      "epoch": 7.53,
+      "grad_norm": 11.244698524475098,
+      "learning_rate": 4.931300583474497e-06,
+      "loss": 0.7113,
+      "step": 40030
+    },
+    {
+      "epoch": 7.54,
+      "grad_norm": 12.09398365020752,
+      "learning_rate": 4.927536231884059e-06,
+      "loss": 0.3169,
+      "step": 40040
+    },
+    {
+      "epoch": 7.54,
+      "grad_norm": 17.471298217773438,
+      "learning_rate": 4.9237718802936196e-06,
+      "loss": 0.3766,
+      "step": 40050
+    },
+    {
+      "epoch": 7.54,
+      "grad_norm": 20.0872802734375,
+      "learning_rate": 4.920007528703181e-06,
+      "loss": 0.4371,
+      "step": 40060
+    },
+    {
+      "epoch": 7.54,
+      "grad_norm": 0.6726735830307007,
+      "learning_rate": 4.9162431771127425e-06,
+      "loss": 0.5611,
+      "step": 40070
+    },
+    {
+      "epoch": 7.54,
+      "grad_norm": 0.362667977809906,
+      "learning_rate": 4.912478825522305e-06,
+      "loss": 0.4279,
+      "step": 40080
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 48.890018463134766,
+      "learning_rate": 4.908714473931865e-06,
+      "loss": 0.3399,
+      "step": 40090
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 2.0744881629943848,
+      "learning_rate": 4.904950122341427e-06,
+      "loss": 0.2226,
+      "step": 40100
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 4.881161212921143,
+      "learning_rate": 4.901185770750988e-06,
+      "loss": 0.6647,
+      "step": 40110
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 15.641790390014648,
+      "learning_rate": 4.89742141916055e-06,
+      "loss": 0.4884,
+      "step": 40120
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 27.704832077026367,
+      "learning_rate": 4.893657067570112e-06,
+      "loss": 0.6596,
+      "step": 40130
+    },
+    {
+      "epoch": 7.56,
+      "grad_norm": 11.091574668884277,
+      "learning_rate": 4.889892715979673e-06,
+      "loss": 0.4945,
+      "step": 40140
+    },
+    {
+      "epoch": 7.56,
+      "grad_norm": 23.857437133789062,
+      "learning_rate": 4.886128364389234e-06,
+      "loss": 0.4619,
+      "step": 40150
+    },
+    {
+      "epoch": 7.56,
+      "grad_norm": 0.4577654004096985,
+      "learning_rate": 4.882364012798796e-06,
+      "loss": 0.2583,
+      "step": 40160
+    },
+    {
+      "epoch": 7.56,
+      "grad_norm": 0.02999284863471985,
+      "learning_rate": 4.878599661208357e-06,
+      "loss": 0.4674,
+      "step": 40170
+    },
+    {
+      "epoch": 7.56,
+      "grad_norm": 8.35001277923584,
+      "learning_rate": 4.8748353096179186e-06,
+      "loss": 0.4048,
+      "step": 40180
+    },
+    {
+      "epoch": 7.56,
+      "grad_norm": 24.522512435913086,
+      "learning_rate": 4.87107095802748e-06,
+      "loss": 0.3845,
+      "step": 40190
+    },
+    {
+      "epoch": 7.57,
+      "grad_norm": 11.478434562683105,
+      "learning_rate": 4.8673066064370415e-06,
+      "loss": 0.4528,
+      "step": 40200
+    },
+    {
+      "epoch": 7.57,
+      "grad_norm": 34.55609130859375,
+      "learning_rate": 4.863542254846603e-06,
+      "loss": 0.3256,
+      "step": 40210
+    },
+    {
+      "epoch": 7.57,
+      "grad_norm": 30.954259872436523,
+      "learning_rate": 4.859777903256164e-06,
+      "loss": 0.6146,
+      "step": 40220
+    },
+    {
+      "epoch": 7.57,
+      "grad_norm": 17.295373916625977,
+      "learning_rate": 4.856013551665726e-06,
+      "loss": 0.3439,
+      "step": 40230
+    },
+    {
+      "epoch": 7.57,
+      "grad_norm": 4.239554405212402,
+      "learning_rate": 4.852249200075287e-06,
+      "loss": 0.4998,
+      "step": 40240
+    },
+    {
+      "epoch": 7.58,
+      "grad_norm": 9.876043319702148,
+      "learning_rate": 4.848484848484849e-06,
+      "loss": 0.4356,
+      "step": 40250
+    },
+    {
+      "epoch": 7.58,
+      "grad_norm": 14.549151420593262,
+      "learning_rate": 4.84472049689441e-06,
+      "loss": 0.3502,
+      "step": 40260
+    },
+    {
+      "epoch": 7.58,
+      "grad_norm": 51.860286712646484,
+      "learning_rate": 4.840956145303972e-06,
+      "loss": 0.6799,
+      "step": 40270
+    },
+    {
+      "epoch": 7.58,
+      "grad_norm": 14.050992965698242,
+      "learning_rate": 4.837191793713533e-06,
+      "loss": 0.2468,
+      "step": 40280
+    },
+    {
+      "epoch": 7.58,
+      "grad_norm": 43.28739547729492,
+      "learning_rate": 4.833427442123095e-06,
+      "loss": 0.4276,
+      "step": 40290
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 0.5742427706718445,
+      "learning_rate": 4.829663090532656e-06,
+      "loss": 0.3327,
+      "step": 40300
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 21.811113357543945,
+      "learning_rate": 4.8258987389422175e-06,
+      "loss": 0.4486,
+      "step": 40310
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 0.1784685254096985,
+      "learning_rate": 4.822134387351779e-06,
+      "loss": 0.2737,
+      "step": 40320
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 38.78385543823242,
+      "learning_rate": 4.8183700357613405e-06,
+      "loss": 0.4103,
+      "step": 40330
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 26.003643035888672,
+      "learning_rate": 4.814605684170902e-06,
+      "loss": 0.4167,
+      "step": 40340
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 0.23788726329803467,
+      "learning_rate": 4.810841332580463e-06,
+      "loss": 0.5304,
+      "step": 40350
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 18.81310272216797,
+      "learning_rate": 4.807076980990025e-06,
+      "loss": 0.5927,
+      "step": 40360
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 36.79855728149414,
+      "learning_rate": 4.803312629399586e-06,
+      "loss": 0.3944,
+      "step": 40370
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 11.548039436340332,
+      "learning_rate": 4.799548277809148e-06,
+      "loss": 0.7024,
+      "step": 40380
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 17.895483016967773,
+      "learning_rate": 4.795783926218709e-06,
+      "loss": 0.4029,
+      "step": 40390
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 25.039541244506836,
+      "learning_rate": 4.792019574628271e-06,
+      "loss": 0.4372,
+      "step": 40400
+    },
+    {
+      "epoch": 7.61,
+      "grad_norm": 5.053126811981201,
+      "learning_rate": 4.788255223037832e-06,
+      "loss": 0.1909,
+      "step": 40410
+    },
+    {
+      "epoch": 7.61,
+      "grad_norm": 19.565786361694336,
+      "learning_rate": 4.784490871447394e-06,
+      "loss": 0.6079,
+      "step": 40420
+    },
+    {
+      "epoch": 7.61,
+      "grad_norm": 21.104755401611328,
+      "learning_rate": 4.780726519856955e-06,
+      "loss": 0.5151,
+      "step": 40430
+    },
+    {
+      "epoch": 7.61,
+      "grad_norm": 20.68798828125,
+      "learning_rate": 4.7769621682665165e-06,
+      "loss": 0.4636,
+      "step": 40440
+    },
+    {
+      "epoch": 7.61,
+      "grad_norm": 0.7437390685081482,
+      "learning_rate": 4.773197816676078e-06,
+      "loss": 0.4972,
+      "step": 40450
+    },
+    {
+      "epoch": 7.62,
+      "grad_norm": 5.112244606018066,
+      "learning_rate": 4.7694334650856395e-06,
+      "loss": 0.4124,
+      "step": 40460
+    },
+    {
+      "epoch": 7.62,
+      "grad_norm": 14.092161178588867,
+      "learning_rate": 4.765669113495201e-06,
+      "loss": 0.5287,
+      "step": 40470
+    },
+    {
+      "epoch": 7.62,
+      "grad_norm": 30.17268180847168,
+      "learning_rate": 4.761904761904762e-06,
+      "loss": 0.5739,
+      "step": 40480
+    },
+    {
+      "epoch": 7.62,
+      "grad_norm": 6.536983013153076,
+      "learning_rate": 4.758140410314324e-06,
+      "loss": 0.2285,
+      "step": 40490
+    },
+    {
+      "epoch": 7.62,
+      "grad_norm": 0.2827499508857727,
+      "learning_rate": 4.754376058723885e-06,
+      "loss": 0.6735,
+      "step": 40500
+    },
+    {
+      "epoch": 7.62,
+      "grad_norm": 6.0071563720703125,
+      "learning_rate": 4.750611707133447e-06,
+      "loss": 0.5643,
+      "step": 40510
+    },
+    {
+      "epoch": 7.63,
+      "grad_norm": 8.479032516479492,
+      "learning_rate": 4.746847355543008e-06,
+      "loss": 0.3364,
+      "step": 40520
+    },
+    {
+      "epoch": 7.63,
+      "grad_norm": 18.724328994750977,
+      "learning_rate": 4.74308300395257e-06,
+      "loss": 0.3524,
+      "step": 40530
+    },
+    {
+      "epoch": 7.63,
+      "grad_norm": 9.712454795837402,
+      "learning_rate": 4.739318652362131e-06,
+      "loss": 0.4297,
+      "step": 40540
+    },
+    {
+      "epoch": 7.63,
+      "grad_norm": 22.55899429321289,
+      "learning_rate": 4.735554300771693e-06,
+      "loss": 0.3444,
+      "step": 40550
+    },
+    {
+      "epoch": 7.63,
+      "grad_norm": 1.9720405340194702,
+      "learning_rate": 4.731789949181254e-06,
+      "loss": 0.4414,
+      "step": 40560
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 0.06697694957256317,
+      "learning_rate": 4.7280255975908155e-06,
+      "loss": 0.4789,
+      "step": 40570
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 11.40739631652832,
+      "learning_rate": 4.724261246000377e-06,
+      "loss": 0.5276,
+      "step": 40580
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 17.688581466674805,
+      "learning_rate": 4.7204968944099384e-06,
+      "loss": 0.4745,
+      "step": 40590
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 21.510601043701172,
+      "learning_rate": 4.7167325428195e-06,
+      "loss": 0.3477,
+      "step": 40600
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 0.09768573194742203,
+      "learning_rate": 4.7129681912290605e-06,
+      "loss": 0.2761,
+      "step": 40610
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 19.56352424621582,
+      "learning_rate": 4.709203839638623e-06,
+      "loss": 0.4043,
+      "step": 40620
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 29.654693603515625,
+      "learning_rate": 4.705439488048184e-06,
+      "loss": 0.4509,
+      "step": 40630
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 3.091660737991333,
+      "learning_rate": 4.701675136457746e-06,
+      "loss": 0.6926,
+      "step": 40640
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 0.025056788697838783,
+      "learning_rate": 4.697910784867307e-06,
+      "loss": 0.4671,
+      "step": 40650
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 16.61036491394043,
+      "learning_rate": 4.694146433276869e-06,
+      "loss": 0.4399,
+      "step": 40660
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 14.356207847595215,
+      "learning_rate": 4.69038208168643e-06,
+      "loss": 0.4281,
+      "step": 40670
+    },
+    {
+      "epoch": 7.66,
+      "grad_norm": 26.823406219482422,
+      "learning_rate": 4.686617730095992e-06,
+      "loss": 0.5089,
+      "step": 40680
+    },
+    {
+      "epoch": 7.66,
+      "grad_norm": 2.588717460632324,
+      "learning_rate": 4.682853378505553e-06,
+      "loss": 0.4385,
+      "step": 40690
+    },
+    {
+      "epoch": 7.66,
+      "grad_norm": 8.470640182495117,
+      "learning_rate": 4.679089026915114e-06,
+      "loss": 0.534,
+      "step": 40700
+    },
+    {
+      "epoch": 7.66,
+      "grad_norm": 1.754682183265686,
+      "learning_rate": 4.675324675324676e-06,
+      "loss": 0.5254,
+      "step": 40710
+    },
+    {
+      "epoch": 7.66,
+      "grad_norm": 37.55604934692383,
+      "learning_rate": 4.6715603237342374e-06,
+      "loss": 0.4852,
+      "step": 40720
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 4.329613208770752,
+      "learning_rate": 4.667795972143799e-06,
+      "loss": 0.3389,
+      "step": 40730
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 5.099806785583496,
+      "learning_rate": 4.66403162055336e-06,
+      "loss": 0.6732,
+      "step": 40740
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 8.931364059448242,
+      "learning_rate": 4.660267268962921e-06,
+      "loss": 0.4488,
+      "step": 40750
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 9.485333442687988,
+      "learning_rate": 4.656502917372483e-06,
+      "loss": 0.3706,
+      "step": 40760
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 10.791669845581055,
+      "learning_rate": 4.652738565782045e-06,
+      "loss": 0.5172,
+      "step": 40770
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 11.53160285949707,
+      "learning_rate": 4.648974214191606e-06,
+      "loss": 0.325,
+      "step": 40780
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 0.11603966355323792,
+      "learning_rate": 4.645209862601167e-06,
+      "loss": 0.5713,
+      "step": 40790
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 4.958864688873291,
+      "learning_rate": 4.641445511010729e-06,
+      "loss": 0.4887,
+      "step": 40800
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 6.0979814529418945,
+      "learning_rate": 4.637681159420291e-06,
+      "loss": 0.4514,
+      "step": 40810
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 37.97774124145508,
+      "learning_rate": 4.633916807829852e-06,
+      "loss": 0.5397,
+      "step": 40820
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 6.384561061859131,
+      "learning_rate": 4.630152456239413e-06,
+      "loss": 0.2419,
+      "step": 40830
+    },
+    {
+      "epoch": 7.69,
+      "grad_norm": 32.47441101074219,
+      "learning_rate": 4.626388104648974e-06,
+      "loss": 0.7723,
+      "step": 40840
+    },
+    {
+      "epoch": 7.69,
+      "grad_norm": 58.51771545410156,
+      "learning_rate": 4.622623753058536e-06,
+      "loss": 0.4001,
+      "step": 40850
+    },
+    {
+      "epoch": 7.69,
+      "grad_norm": 1.2339733839035034,
+      "learning_rate": 4.618859401468098e-06,
+      "loss": 0.2751,
+      "step": 40860
+    },
+    {
+      "epoch": 7.69,
+      "grad_norm": 1.42243492603302,
+      "learning_rate": 4.615095049877659e-06,
+      "loss": 0.4036,
+      "step": 40870
+    },
+    {
+      "epoch": 7.69,
+      "grad_norm": 0.05874158814549446,
+      "learning_rate": 4.61133069828722e-06,
+      "loss": 0.2054,
+      "step": 40880
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 0.06559593230485916,
+      "learning_rate": 4.607566346696781e-06,
+      "loss": 0.3678,
+      "step": 40890
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 37.953216552734375,
+      "learning_rate": 4.603801995106344e-06,
+      "loss": 0.7121,
+      "step": 40900
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 28.273658752441406,
+      "learning_rate": 4.600037643515905e-06,
+      "loss": 0.7176,
+      "step": 40910
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 0.9303427934646606,
+      "learning_rate": 4.596273291925466e-06,
+      "loss": 0.4192,
+      "step": 40920
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 6.678520202636719,
+      "learning_rate": 4.592508940335027e-06,
+      "loss": 0.4673,
+      "step": 40930
+    },
+    {
+      "epoch": 7.71,
+      "grad_norm": 10.7068510055542,
+      "learning_rate": 4.5887445887445896e-06,
+      "loss": 0.5474,
+      "step": 40940
+    },
+    {
+      "epoch": 7.71,
+      "grad_norm": 24.268388748168945,
+      "learning_rate": 4.584980237154151e-06,
+      "loss": 0.4693,
+      "step": 40950
+    },
+    {
+      "epoch": 7.71,
+      "grad_norm": 0.8732045292854309,
+      "learning_rate": 4.581215885563712e-06,
+      "loss": 0.3434,
+      "step": 40960
+    },
+    {
+      "epoch": 7.71,
+      "grad_norm": 4.5128865242004395,
+      "learning_rate": 4.577451533973273e-06,
+      "loss": 0.5951,
+      "step": 40970
+    },
+    {
+      "epoch": 7.71,
+      "grad_norm": 29.08911895751953,
+      "learning_rate": 4.5736871823828346e-06,
+      "loss": 0.8518,
+      "step": 40980
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 34.50959014892578,
+      "learning_rate": 4.569922830792397e-06,
+      "loss": 0.3967,
+      "step": 40990
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 12.16829776763916,
+      "learning_rate": 4.566158479201958e-06,
+      "loss": 0.568,
+      "step": 41000
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 8.815430641174316,
+      "learning_rate": 4.562394127611519e-06,
+      "loss": 0.6214,
+      "step": 41010
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 0.7041884064674377,
+      "learning_rate": 4.55862977602108e-06,
+      "loss": 0.3887,
+      "step": 41020
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 5.827664852142334,
+      "learning_rate": 4.554865424430642e-06,
+      "loss": 0.3346,
+      "step": 41030
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 13.621718406677246,
+      "learning_rate": 4.551101072840204e-06,
+      "loss": 0.5485,
+      "step": 41040
+    },
+    {
+      "epoch": 7.73,
+      "grad_norm": 0.27952587604522705,
+      "learning_rate": 4.547336721249765e-06,
+      "loss": 0.4817,
+      "step": 41050
+    },
+    {
+      "epoch": 7.73,
+      "grad_norm": 7.287542343139648,
+      "learning_rate": 4.543572369659326e-06,
+      "loss": 0.674,
+      "step": 41060
+    },
+    {
+      "epoch": 7.73,
+      "grad_norm": 8.12995433807373,
+      "learning_rate": 4.539808018068888e-06,
+      "loss": 0.6428,
+      "step": 41070
+    },
+    {
+      "epoch": 7.73,
+      "grad_norm": 20.937484741210938,
+      "learning_rate": 4.53604366647845e-06,
+      "loss": 0.5712,
+      "step": 41080
+    },
+    {
+      "epoch": 7.73,
+      "grad_norm": 15.425496101379395,
+      "learning_rate": 4.532279314888011e-06,
+      "loss": 0.7567,
+      "step": 41090
+    },
+    {
+      "epoch": 7.74,
+      "grad_norm": 14.051750183105469,
+      "learning_rate": 4.528514963297572e-06,
+      "loss": 0.5802,
+      "step": 41100
+    },
+    {
+      "epoch": 7.74,
+      "grad_norm": 0.16733068227767944,
+      "learning_rate": 4.5247506117071336e-06,
+      "loss": 0.2608,
+      "step": 41110
+    },
+    {
+      "epoch": 7.74,
+      "grad_norm": 17.37752914428711,
+      "learning_rate": 4.520986260116695e-06,
+      "loss": 0.5201,
+      "step": 41120
+    },
+    {
+      "epoch": 7.74,
+      "grad_norm": 18.55728530883789,
+      "learning_rate": 4.517221908526257e-06,
+      "loss": 0.4982,
+      "step": 41130
+    },
+    {
+      "epoch": 7.74,
+      "grad_norm": 15.515111923217773,
+      "learning_rate": 4.513457556935818e-06,
+      "loss": 0.5779,
+      "step": 41140
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 2.7760565280914307,
+      "learning_rate": 4.509693205345379e-06,
+      "loss": 0.2145,
+      "step": 41150
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 7.785154342651367,
+      "learning_rate": 4.505928853754941e-06,
+      "loss": 0.2332,
+      "step": 41160
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 27.99343490600586,
+      "learning_rate": 4.502164502164502e-06,
+      "loss": 0.4307,
+      "step": 41170
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 19.354997634887695,
+      "learning_rate": 4.498400150574064e-06,
+      "loss": 0.8148,
+      "step": 41180
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 2.94992995262146,
+      "learning_rate": 4.494635798983625e-06,
+      "loss": 0.1812,
+      "step": 41190
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 10.769566535949707,
+      "learning_rate": 4.490871447393187e-06,
+      "loss": 0.4019,
+      "step": 41200
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 15.878639221191406,
+      "learning_rate": 4.487107095802748e-06,
+      "loss": 0.4287,
+      "step": 41210
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 0.04310622438788414,
+      "learning_rate": 4.48334274421231e-06,
+      "loss": 0.333,
+      "step": 41220
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 5.737595081329346,
+      "learning_rate": 4.479578392621871e-06,
+      "loss": 0.4404,
+      "step": 41230
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 14.339284896850586,
+      "learning_rate": 4.4758140410314325e-06,
+      "loss": 0.4577,
+      "step": 41240
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 10.48139762878418,
+      "learning_rate": 4.472049689440994e-06,
+      "loss": 0.3453,
+      "step": 41250
+    },
+    {
+      "epoch": 7.77,
+      "grad_norm": 4.904701232910156,
+      "learning_rate": 4.4682853378505555e-06,
+      "loss": 0.3275,
+      "step": 41260
+    },
+    {
+      "epoch": 7.77,
+      "grad_norm": 0.7866021990776062,
+      "learning_rate": 4.464520986260117e-06,
+      "loss": 0.4991,
+      "step": 41270
+    },
+    {
+      "epoch": 7.77,
+      "grad_norm": 13.277933120727539,
+      "learning_rate": 4.460756634669678e-06,
+      "loss": 0.4167,
+      "step": 41280
+    },
+    {
+      "epoch": 7.77,
+      "grad_norm": 31.82375717163086,
+      "learning_rate": 4.45699228307924e-06,
+      "loss": 0.4556,
+      "step": 41290
+    },
+    {
+      "epoch": 7.77,
+      "grad_norm": 36.45254898071289,
+      "learning_rate": 4.453227931488801e-06,
+      "loss": 0.4507,
+      "step": 41300
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 38.87356948852539,
+      "learning_rate": 4.449463579898363e-06,
+      "loss": 0.4436,
+      "step": 41310
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 0.09693789482116699,
+      "learning_rate": 4.445699228307924e-06,
+      "loss": 0.3846,
+      "step": 41320
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 17.44363784790039,
+      "learning_rate": 4.441934876717486e-06,
+      "loss": 0.3734,
+      "step": 41330
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 3.033371925354004,
+      "learning_rate": 4.438170525127047e-06,
+      "loss": 0.3748,
+      "step": 41340
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 14.98725414276123,
+      "learning_rate": 4.434406173536609e-06,
+      "loss": 0.8152,
+      "step": 41350
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 6.687844753265381,
+      "learning_rate": 4.43064182194617e-06,
+      "loss": 0.6698,
+      "step": 41360
+    },
+    {
+      "epoch": 7.79,
+      "grad_norm": 0.14900249242782593,
+      "learning_rate": 4.4268774703557315e-06,
+      "loss": 0.3195,
+      "step": 41370
+    },
+    {
+      "epoch": 7.79,
+      "grad_norm": 9.252775192260742,
+      "learning_rate": 4.423113118765293e-06,
+      "loss": 0.4316,
+      "step": 41380
+    },
+    {
+      "epoch": 7.79,
+      "grad_norm": 7.680717945098877,
+      "learning_rate": 4.4193487671748545e-06,
+      "loss": 0.164,
+      "step": 41390
+    },
+    {
+      "epoch": 7.79,
+      "grad_norm": 7.291797637939453,
+      "learning_rate": 4.415584415584416e-06,
+      "loss": 0.5361,
+      "step": 41400
+    },
+    {
+      "epoch": 7.79,
+      "grad_norm": 3.572070360183716,
+      "learning_rate": 4.411820063993977e-06,
+      "loss": 0.4278,
+      "step": 41410
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 0.2998538911342621,
+      "learning_rate": 4.408055712403539e-06,
+      "loss": 0.3525,
+      "step": 41420
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 0.47384247183799744,
+      "learning_rate": 4.4042913608131e-06,
+      "loss": 0.2823,
+      "step": 41430
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 0.05279775708913803,
+      "learning_rate": 4.400527009222662e-06,
+      "loss": 0.4339,
+      "step": 41440
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 26.150938034057617,
+      "learning_rate": 4.396762657632223e-06,
+      "loss": 0.537,
+      "step": 41450
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 29.126239776611328,
+      "learning_rate": 4.392998306041785e-06,
+      "loss": 0.4351,
+      "step": 41460
+    },
+    {
+      "epoch": 7.81,
+      "grad_norm": 13.385824203491211,
+      "learning_rate": 4.389233954451346e-06,
+      "loss": 0.2548,
+      "step": 41470
+    },
+    {
+      "epoch": 7.81,
+      "grad_norm": 110.32833099365234,
+      "learning_rate": 4.385469602860908e-06,
+      "loss": 0.6887,
+      "step": 41480
+    },
+    {
+      "epoch": 7.81,
+      "grad_norm": 2.3352975845336914,
+      "learning_rate": 4.381705251270469e-06,
+      "loss": 0.4305,
+      "step": 41490
+    },
+    {
+      "epoch": 7.81,
+      "grad_norm": 18.318313598632812,
+      "learning_rate": 4.3779408996800305e-06,
+      "loss": 0.2915,
+      "step": 41500
+    },
+    {
+      "epoch": 7.81,
+      "grad_norm": 14.995280265808105,
+      "learning_rate": 4.374176548089592e-06,
+      "loss": 0.4918,
+      "step": 41510
+    },
+    {
+      "epoch": 7.81,
+      "grad_norm": 8.48552417755127,
+      "learning_rate": 4.3704121964991534e-06,
+      "loss": 0.502,
+      "step": 41520
+    },
+    {
+      "epoch": 7.82,
+      "grad_norm": 6.212337017059326,
+      "learning_rate": 4.366647844908715e-06,
+      "loss": 0.5833,
+      "step": 41530
+    },
+    {
+      "epoch": 7.82,
+      "grad_norm": 7.40939474105835,
+      "learning_rate": 4.362883493318276e-06,
+      "loss": 0.4578,
+      "step": 41540
+    },
+    {
+      "epoch": 7.82,
+      "grad_norm": 7.82789945602417,
+      "learning_rate": 4.359119141727838e-06,
+      "loss": 0.4524,
+      "step": 41550
+    },
+    {
+      "epoch": 7.82,
+      "grad_norm": 20.18194580078125,
+      "learning_rate": 4.355354790137399e-06,
+      "loss": 0.7412,
+      "step": 41560
+    },
+    {
+      "epoch": 7.82,
+      "grad_norm": 1.9940485954284668,
+      "learning_rate": 4.351590438546961e-06,
+      "loss": 0.3181,
+      "step": 41570
+    },
+    {
+      "epoch": 7.83,
+      "grad_norm": 4.829334259033203,
+      "learning_rate": 4.347826086956522e-06,
+      "loss": 0.19,
+      "step": 41580
+    },
+    {
+      "epoch": 7.83,
+      "grad_norm": 0.15990076959133148,
+      "learning_rate": 4.344061735366084e-06,
+      "loss": 0.2216,
+      "step": 41590
+    },
+    {
+      "epoch": 7.83,
+      "grad_norm": 6.221506595611572,
+      "learning_rate": 4.340297383775645e-06,
+      "loss": 0.389,
+      "step": 41600
+    },
+    {
+      "epoch": 7.83,
+      "grad_norm": 8.496952056884766,
+      "learning_rate": 4.336533032185207e-06,
+      "loss": 0.5286,
+      "step": 41610
+    },
+    {
+      "epoch": 7.83,
+      "grad_norm": 23.715831756591797,
+      "learning_rate": 4.332768680594768e-06,
+      "loss": 0.4569,
+      "step": 41620
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 5.149649143218994,
+      "learning_rate": 4.3290043290043295e-06,
+      "loss": 0.4774,
+      "step": 41630
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 17.749088287353516,
+      "learning_rate": 4.325239977413891e-06,
+      "loss": 0.514,
+      "step": 41640
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 1.3218231201171875,
+      "learning_rate": 4.3214756258234524e-06,
+      "loss": 0.2201,
+      "step": 41650
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 10.610811233520508,
+      "learning_rate": 4.317711274233014e-06,
+      "loss": 0.5015,
+      "step": 41660
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 30.114301681518555,
+      "learning_rate": 4.313946922642575e-06,
+      "loss": 0.456,
+      "step": 41670
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 6.920668601989746,
+      "learning_rate": 4.310182571052137e-06,
+      "loss": 0.4653,
+      "step": 41680
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 0.0986671969294548,
+      "learning_rate": 4.306418219461698e-06,
+      "loss": 0.4564,
+      "step": 41690
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 10.98770523071289,
+      "learning_rate": 4.302653867871259e-06,
+      "loss": 0.4449,
+      "step": 41700
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 7.872487545013428,
+      "learning_rate": 4.298889516280821e-06,
+      "loss": 0.3511,
+      "step": 41710
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 0.12211643904447556,
+      "learning_rate": 4.295125164690383e-06,
+      "loss": 0.3521,
+      "step": 41720
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 2.3334715366363525,
+      "learning_rate": 4.291360813099944e-06,
+      "loss": 0.4879,
+      "step": 41730
+    },
+    {
+      "epoch": 7.86,
+      "grad_norm": 1.1513813734054565,
+      "learning_rate": 4.287596461509506e-06,
+      "loss": 0.2625,
+      "step": 41740
+    },
+    {
+      "epoch": 7.86,
+      "grad_norm": 3.9860312938690186,
+      "learning_rate": 4.283832109919067e-06,
+      "loss": 0.3835,
+      "step": 41750
+    },
+    {
+      "epoch": 7.86,
+      "grad_norm": 16.243440628051758,
+      "learning_rate": 4.2800677583286285e-06,
+      "loss": 0.4477,
+      "step": 41760
+    },
+    {
+      "epoch": 7.86,
+      "grad_norm": 23.092260360717773,
+      "learning_rate": 4.27630340673819e-06,
+      "loss": 0.4379,
+      "step": 41770
+    },
+    {
+      "epoch": 7.86,
+      "grad_norm": 0.4055626094341278,
+      "learning_rate": 4.272539055147751e-06,
+      "loss": 0.5019,
+      "step": 41780
+    },
+    {
+      "epoch": 7.87,
+      "grad_norm": 15.21871566772461,
+      "learning_rate": 4.268774703557312e-06,
+      "loss": 0.4496,
+      "step": 41790
+    },
+    {
+      "epoch": 7.87,
+      "grad_norm": 22.797069549560547,
+      "learning_rate": 4.265010351966874e-06,
+      "loss": 0.5968,
+      "step": 41800
+    },
+    {
+      "epoch": 7.87,
+      "grad_norm": 0.028871312737464905,
+      "learning_rate": 4.261246000376436e-06,
+      "loss": 0.5744,
+      "step": 41810
+    },
+    {
+      "epoch": 7.87,
+      "grad_norm": 14.592629432678223,
+      "learning_rate": 4.257481648785997e-06,
+      "loss": 0.2576,
+      "step": 41820
+    },
+    {
+      "epoch": 7.87,
+      "grad_norm": 50.91252899169922,
+      "learning_rate": 4.253717297195558e-06,
+      "loss": 0.4805,
+      "step": 41830
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": 0.8403691649436951,
+      "learning_rate": 4.249952945605119e-06,
+      "loss": 0.3705,
+      "step": 41840
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": 0.6822836995124817,
+      "learning_rate": 4.246188594014682e-06,
+      "loss": 0.2433,
+      "step": 41850
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": 4.622705936431885,
+      "learning_rate": 4.242424242424243e-06,
+      "loss": 0.6058,
+      "step": 41860
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": 17.88176155090332,
+      "learning_rate": 4.2386598908338046e-06,
+      "loss": 0.5387,
+      "step": 41870
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": 0.5473073124885559,
+      "learning_rate": 4.234895539243365e-06,
+      "loss": 0.6625,
+      "step": 41880
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": 0.17052659392356873,
+      "learning_rate": 4.2311311876529275e-06,
+      "loss": 0.3257,
+      "step": 41890
+    },
+    {
+      "epoch": 7.89,
+      "grad_norm": 0.7636673450469971,
+      "learning_rate": 4.227366836062489e-06,
+      "loss": 0.3624,
+      "step": 41900
+    },
+    {
+      "epoch": 7.89,
+      "grad_norm": 20.62749481201172,
+      "learning_rate": 4.22360248447205e-06,
+      "loss": 0.4521,
+      "step": 41910
+    },
+    {
+      "epoch": 7.89,
+      "grad_norm": 6.845425128936768,
+      "learning_rate": 4.219838132881611e-06,
+      "loss": 0.3049,
+      "step": 41920
+    },
+    {
+      "epoch": 7.89,
+      "grad_norm": 26.27695655822754,
+      "learning_rate": 4.2160737812911725e-06,
+      "loss": 0.699,
+      "step": 41930
+    },
+    {
+      "epoch": 7.89,
+      "grad_norm": 10.913443565368652,
+      "learning_rate": 4.212309429700735e-06,
+      "loss": 0.7243,
+      "step": 41940
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 29.593870162963867,
+      "learning_rate": 4.208545078110296e-06,
+      "loss": 0.5249,
+      "step": 41950
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 7.848869323730469,
+      "learning_rate": 4.204780726519857e-06,
+      "loss": 0.4757,
+      "step": 41960
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 10.693510055541992,
+      "learning_rate": 4.201016374929418e-06,
+      "loss": 0.4779,
+      "step": 41970
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 2.0626046657562256,
+      "learning_rate": 4.19725202333898e-06,
+      "loss": 0.412,
+      "step": 41980
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 7.025002479553223,
+      "learning_rate": 4.193487671748542e-06,
+      "loss": 0.2752,
+      "step": 41990
+    },
+    {
+      "epoch": 7.91,
+      "grad_norm": 23.475351333618164,
+      "learning_rate": 4.1897233201581036e-06,
+      "loss": 0.4691,
+      "step": 42000
+    },
+    {
+      "epoch": 7.91,
+      "grad_norm": 17.180540084838867,
+      "learning_rate": 4.185958968567664e-06,
+      "loss": 0.3662,
+      "step": 42010
+    },
+    {
+      "epoch": 7.91,
+      "grad_norm": 30.194869995117188,
+      "learning_rate": 4.182194616977226e-06,
+      "loss": 0.3635,
+      "step": 42020
+    },
+    {
+      "epoch": 7.91,
+      "grad_norm": 0.9849434494972229,
+      "learning_rate": 4.178430265386788e-06,
+      "loss": 0.5017,
+      "step": 42030
+    },
+    {
+      "epoch": 7.91,
+      "grad_norm": 0.7494462132453918,
+      "learning_rate": 4.174665913796349e-06,
+      "loss": 0.3769,
+      "step": 42040
+    },
+    {
+      "epoch": 7.91,
+      "grad_norm": 15.633011817932129,
+      "learning_rate": 4.17090156220591e-06,
+      "loss": 0.4248,
+      "step": 42050
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 0.03347090259194374,
+      "learning_rate": 4.1671372106154715e-06,
+      "loss": 0.6362,
+      "step": 42060
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 2.5611062049865723,
+      "learning_rate": 4.163372859025033e-06,
+      "loss": 0.5368,
+      "step": 42070
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 3.9073028564453125,
+      "learning_rate": 4.159608507434595e-06,
+      "loss": 0.6604,
+      "step": 42080
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 30.027442932128906,
+      "learning_rate": 4.155844155844157e-06,
+      "loss": 0.521,
+      "step": 42090
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 22.11733055114746,
+      "learning_rate": 4.152079804253717e-06,
+      "loss": 0.5561,
+      "step": 42100
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 13.414361000061035,
+      "learning_rate": 4.148315452663279e-06,
+      "loss": 0.5497,
+      "step": 42110
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 0.3089528977870941,
+      "learning_rate": 4.14455110107284e-06,
+      "loss": 0.3887,
+      "step": 42120
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 31.682811737060547,
+      "learning_rate": 4.1407867494824025e-06,
+      "loss": 0.7196,
+      "step": 42130
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 43.80996322631836,
+      "learning_rate": 4.137022397891963e-06,
+      "loss": 0.3509,
+      "step": 42140
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 3.049875020980835,
+      "learning_rate": 4.133258046301525e-06,
+      "loss": 0.4552,
+      "step": 42150
+    },
+    {
+      "epoch": 7.94,
+      "grad_norm": 22.50640296936035,
+      "learning_rate": 4.129493694711086e-06,
+      "loss": 0.298,
+      "step": 42160
+    },
+    {
+      "epoch": 7.94,
+      "grad_norm": 19.20767593383789,
+      "learning_rate": 4.125729343120648e-06,
+      "loss": 0.3963,
+      "step": 42170
+    },
+    {
+      "epoch": 7.94,
+      "grad_norm": 15.141514778137207,
+      "learning_rate": 4.121964991530209e-06,
+      "loss": 0.4502,
+      "step": 42180
+    },
+    {
+      "epoch": 7.94,
+      "grad_norm": 17.997087478637695,
+      "learning_rate": 4.1182006399397705e-06,
+      "loss": 0.2375,
+      "step": 42190
+    },
+    {
+      "epoch": 7.94,
+      "grad_norm": 1.2107799053192139,
+      "learning_rate": 4.114436288349332e-06,
+      "loss": 0.4342,
+      "step": 42200
+    },
+    {
+      "epoch": 7.94,
+      "grad_norm": 17.25642967224121,
+      "learning_rate": 4.110671936758893e-06,
+      "loss": 0.599,
+      "step": 42210
+    },
+    {
+      "epoch": 7.95,
+      "grad_norm": 23.210981369018555,
+      "learning_rate": 4.106907585168456e-06,
+      "loss": 0.6262,
+      "step": 42220
+    },
+    {
+      "epoch": 7.95,
+      "grad_norm": 18.647939682006836,
+      "learning_rate": 4.103143233578016e-06,
+      "loss": 0.6408,
+      "step": 42230
+    },
+    {
+      "epoch": 7.95,
+      "grad_norm": 24.76263999938965,
+      "learning_rate": 4.099378881987578e-06,
+      "loss": 0.4443,
+      "step": 42240
+    },
+    {
+      "epoch": 7.95,
+      "grad_norm": 6.9933013916015625,
+      "learning_rate": 4.095614530397139e-06,
+      "loss": 0.6158,
+      "step": 42250
+    },
+    {
+      "epoch": 7.95,
+      "grad_norm": 0.15161870419979095,
+      "learning_rate": 4.091850178806701e-06,
+      "loss": 0.2637,
+      "step": 42260
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 5.794905185699463,
+      "learning_rate": 4.088085827216262e-06,
+      "loss": 1.0222,
+      "step": 42270
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 37.5270881652832,
+      "learning_rate": 4.084321475625824e-06,
+      "loss": 0.3388,
+      "step": 42280
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 20.47008514404297,
+      "learning_rate": 4.080557124035385e-06,
+      "loss": 0.4998,
+      "step": 42290
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 10.392271041870117,
+      "learning_rate": 4.0767927724449465e-06,
+      "loss": 0.2468,
+      "step": 42300
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 22.866073608398438,
+      "learning_rate": 4.073028420854508e-06,
+      "loss": 0.361,
+      "step": 42310
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 3.4224941730499268,
+      "learning_rate": 4.0692640692640695e-06,
+      "loss": 0.5056,
+      "step": 42320
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 15.538722038269043,
+      "learning_rate": 4.065499717673631e-06,
+      "loss": 0.5327,
+      "step": 42330
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 17.44469451904297,
+      "learning_rate": 4.061735366083192e-06,
+      "loss": 0.628,
+      "step": 42340
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 24.630922317504883,
+      "learning_rate": 4.057971014492754e-06,
+      "loss": 0.5497,
+      "step": 42350
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 2.710216760635376,
+      "learning_rate": 4.054206662902315e-06,
+      "loss": 0.2458,
+      "step": 42360
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 2.76000714302063,
+      "learning_rate": 4.050442311311877e-06,
+      "loss": 0.4532,
+      "step": 42370
+    },
+    {
+      "epoch": 7.98,
+      "grad_norm": 20.97554588317871,
+      "learning_rate": 4.046677959721438e-06,
+      "loss": 0.559,
+      "step": 42380
+    },
+    {
+      "epoch": 7.98,
+      "grad_norm": 53.717735290527344,
+      "learning_rate": 4.042913608131e-06,
+      "loss": 0.8006,
+      "step": 42390
+    },
+    {
+      "epoch": 7.98,
+      "grad_norm": 11.499624252319336,
+      "learning_rate": 4.039149256540561e-06,
+      "loss": 0.593,
+      "step": 42400
+    },
+    {
+      "epoch": 7.98,
+      "grad_norm": 1.2157829999923706,
+      "learning_rate": 4.035384904950123e-06,
+      "loss": 0.4126,
+      "step": 42410
+    },
+    {
+      "epoch": 7.98,
+      "grad_norm": 6.117697715759277,
+      "learning_rate": 4.031620553359684e-06,
+      "loss": 0.4347,
+      "step": 42420
+    },
+    {
+      "epoch": 7.99,
+      "grad_norm": 6.3184051513671875,
+      "learning_rate": 4.0278562017692455e-06,
+      "loss": 0.2998,
+      "step": 42430
+    },
+    {
+      "epoch": 7.99,
+      "grad_norm": 20.870182037353516,
+      "learning_rate": 4.024091850178807e-06,
+      "loss": 0.6725,
+      "step": 42440
+    },
+    {
+      "epoch": 7.99,
+      "grad_norm": 17.98334503173828,
+      "learning_rate": 4.0203274985883684e-06,
+      "loss": 0.3782,
+      "step": 42450
+    },
+    {
+      "epoch": 7.99,
+      "grad_norm": 7.8092193603515625,
+      "learning_rate": 4.01656314699793e-06,
+      "loss": 0.3603,
+      "step": 42460
+    },
+    {
+      "epoch": 7.99,
+      "grad_norm": 1.0869126319885254,
+      "learning_rate": 4.012798795407491e-06,
+      "loss": 0.4338,
+      "step": 42470
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 4.888917446136475,
+      "learning_rate": 4.009034443817053e-06,
+      "loss": 0.3939,
+      "step": 42480
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.40398913621902466,
+      "learning_rate": 4.005270092226614e-06,
+      "loss": 0.6347,
+      "step": 42490
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 17.665483474731445,
+      "learning_rate": 4.001505740636176e-06,
+      "loss": 0.6128,
+      "step": 42500
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.9261333333333334,
+      "eval_loss": 0.2983257472515106,
+      "eval_runtime": 51.1739,
+      "eval_samples_per_second": 146.559,
+      "eval_steps_per_second": 18.33,
+      "step": 42504
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 8.759658813476562,
+      "learning_rate": 3.997741389045737e-06,
+      "loss": 0.6272,
+      "step": 42510
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 1.319584846496582,
+      "learning_rate": 3.993977037455299e-06,
+      "loss": 0.4546,
+      "step": 42520
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 16.56199836730957,
+      "learning_rate": 3.99021268586486e-06,
+      "loss": 0.7,
+      "step": 42530
+    },
+    {
+      "epoch": 8.01,
+      "grad_norm": 2.113156318664551,
+      "learning_rate": 3.986448334274422e-06,
+      "loss": 0.2512,
+      "step": 42540
+    },
+    {
+      "epoch": 8.01,
+      "grad_norm": 5.923486709594727,
+      "learning_rate": 3.982683982683983e-06,
+      "loss": 0.4087,
+      "step": 42550
+    },
+    {
+      "epoch": 8.01,
+      "grad_norm": 10.338817596435547,
+      "learning_rate": 3.9789196310935445e-06,
+      "loss": 0.4118,
+      "step": 42560
+    },
+    {
+      "epoch": 8.01,
+      "grad_norm": 16.17352294921875,
+      "learning_rate": 3.975155279503106e-06,
+      "loss": 0.6563,
+      "step": 42570
+    },
+    {
+      "epoch": 8.01,
+      "grad_norm": 12.729879379272461,
+      "learning_rate": 3.9713909279126674e-06,
+      "loss": 0.5468,
+      "step": 42580
+    },
+    {
+      "epoch": 8.02,
+      "grad_norm": 40.52555847167969,
+      "learning_rate": 3.967626576322229e-06,
+      "loss": 0.3984,
+      "step": 42590
+    },
+    {
+      "epoch": 8.02,
+      "grad_norm": 22.066959381103516,
+      "learning_rate": 3.96386222473179e-06,
+      "loss": 0.6228,
+      "step": 42600
+    },
+    {
+      "epoch": 8.02,
+      "grad_norm": 10.428549766540527,
+      "learning_rate": 3.960097873141352e-06,
+      "loss": 0.5729,
+      "step": 42610
+    },
+    {
+      "epoch": 8.02,
+      "grad_norm": 9.915628433227539,
+      "learning_rate": 3.956333521550913e-06,
+      "loss": 0.3231,
+      "step": 42620
+    },
+    {
+      "epoch": 8.02,
+      "grad_norm": 33.11208724975586,
+      "learning_rate": 3.952569169960475e-06,
+      "loss": 0.3779,
+      "step": 42630
+    },
+    {
+      "epoch": 8.03,
+      "grad_norm": 4.810709476470947,
+      "learning_rate": 3.948804818370036e-06,
+      "loss": 0.6932,
+      "step": 42640
+    },
+    {
+      "epoch": 8.03,
+      "grad_norm": 21.556655883789062,
+      "learning_rate": 3.945040466779598e-06,
+      "loss": 0.3488,
+      "step": 42650
+    },
+    {
+      "epoch": 8.03,
+      "grad_norm": 12.891266822814941,
+      "learning_rate": 3.941276115189159e-06,
+      "loss": 0.4167,
+      "step": 42660
+    },
+    {
+      "epoch": 8.03,
+      "grad_norm": 0.5069437026977539,
+      "learning_rate": 3.9375117635987206e-06,
+      "loss": 0.2696,
+      "step": 42670
+    },
+    {
+      "epoch": 8.03,
+      "grad_norm": 8.775614738464355,
+      "learning_rate": 3.933747412008282e-06,
+      "loss": 0.5591,
+      "step": 42680
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 6.012180805206299,
+      "learning_rate": 3.9299830604178435e-06,
+      "loss": 0.5263,
+      "step": 42690
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 7.999334812164307,
+      "learning_rate": 3.926218708827404e-06,
+      "loss": 0.3107,
+      "step": 42700
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 12.329937934875488,
+      "learning_rate": 3.922454357236966e-06,
+      "loss": 0.3968,
+      "step": 42710
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 16.49465560913086,
+      "learning_rate": 3.918690005646528e-06,
+      "loss": 0.6177,
+      "step": 42720
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 21.259458541870117,
+      "learning_rate": 3.914925654056089e-06,
+      "loss": 0.5283,
+      "step": 42730
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 17.6918888092041,
+      "learning_rate": 3.911161302465651e-06,
+      "loss": 0.2778,
+      "step": 42740
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 8.05215072631836,
+      "learning_rate": 3.907396950875212e-06,
+      "loss": 0.4152,
+      "step": 42750
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 17.829729080200195,
+      "learning_rate": 3.903632599284774e-06,
+      "loss": 0.2527,
+      "step": 42760
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 28.24970054626465,
+      "learning_rate": 3.899868247694335e-06,
+      "loss": 0.3636,
+      "step": 42770
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 13.935927391052246,
+      "learning_rate": 3.896103896103897e-06,
+      "loss": 0.2224,
+      "step": 42780
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 21.196792602539062,
+      "learning_rate": 3.892339544513457e-06,
+      "loss": 0.5058,
+      "step": 42790
+    },
+    {
+      "epoch": 8.06,
+      "grad_norm": 18.93883514404297,
+      "learning_rate": 3.8885751929230196e-06,
+      "loss": 0.4206,
+      "step": 42800
+    },
+    {
+      "epoch": 8.06,
+      "grad_norm": 0.547015368938446,
+      "learning_rate": 3.884810841332581e-06,
+      "loss": 0.6522,
+      "step": 42810
+    },
+    {
+      "epoch": 8.06,
+      "grad_norm": 9.596744537353516,
+      "learning_rate": 3.8810464897421425e-06,
+      "loss": 0.6818,
+      "step": 42820
+    },
+    {
+      "epoch": 8.06,
+      "grad_norm": 17.509187698364258,
+      "learning_rate": 3.877282138151704e-06,
+      "loss": 0.3802,
+      "step": 42830
+    },
+    {
+      "epoch": 8.06,
+      "grad_norm": 25.433029174804688,
+      "learning_rate": 3.8735177865612646e-06,
+      "loss": 0.2126,
+      "step": 42840
+    },
+    {
+      "epoch": 8.07,
+      "grad_norm": 41.067604064941406,
+      "learning_rate": 3.869753434970827e-06,
+      "loss": 0.4666,
+      "step": 42850
+    },
+    {
+      "epoch": 8.07,
+      "grad_norm": 14.244887351989746,
+      "learning_rate": 3.865989083380388e-06,
+      "loss": 0.4267,
+      "step": 42860
+    },
+    {
+      "epoch": 8.07,
+      "grad_norm": 19.82988929748535,
+      "learning_rate": 3.86222473178995e-06,
+      "loss": 0.6026,
+      "step": 42870
+    },
+    {
+      "epoch": 8.07,
+      "grad_norm": 4.339291572570801,
+      "learning_rate": 3.85846038019951e-06,
+      "loss": 0.2404,
+      "step": 42880
+    },
+    {
+      "epoch": 8.07,
+      "grad_norm": 17.403106689453125,
+      "learning_rate": 3.854696028609073e-06,
+      "loss": 0.6288,
+      "step": 42890
+    },
+    {
+      "epoch": 8.07,
+      "grad_norm": 44.81716537475586,
+      "learning_rate": 3.850931677018634e-06,
+      "loss": 0.4931,
+      "step": 42900
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 6.8778510093688965,
+      "learning_rate": 3.847167325428196e-06,
+      "loss": 0.4437,
+      "step": 42910
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 0.9462854862213135,
+      "learning_rate": 3.843402973837756e-06,
+      "loss": 0.3791,
+      "step": 42920
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 13.075630187988281,
+      "learning_rate": 3.839638622247318e-06,
+      "loss": 0.4265,
+      "step": 42930
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 34.70516586303711,
+      "learning_rate": 3.83587427065688e-06,
+      "loss": 0.3808,
+      "step": 42940
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 0.2037682831287384,
+      "learning_rate": 3.8321099190664415e-06,
+      "loss": 0.3715,
+      "step": 42950
+    },
+    {
+      "epoch": 8.09,
+      "grad_norm": 1.690575361251831,
+      "learning_rate": 3.828345567476003e-06,
+      "loss": 0.3384,
+      "step": 42960
+    },
+    {
+      "epoch": 8.09,
+      "grad_norm": 4.726039886474609,
+      "learning_rate": 3.8245812158855635e-06,
+      "loss": 0.3805,
+      "step": 42970
+    },
+    {
+      "epoch": 8.09,
+      "grad_norm": 10.409626007080078,
+      "learning_rate": 3.820816864295126e-06,
+      "loss": 0.2682,
+      "step": 42980
+    },
+    {
+      "epoch": 8.09,
+      "grad_norm": 12.783897399902344,
+      "learning_rate": 3.817052512704687e-06,
+      "loss": 0.6112,
+      "step": 42990
+    },
+    {
+      "epoch": 8.09,
+      "grad_norm": 0.31685903668403625,
+      "learning_rate": 3.8132881611142488e-06,
+      "loss": 0.4119,
+      "step": 43000
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 43.89032745361328,
+      "learning_rate": 3.80952380952381e-06,
+      "loss": 0.4113,
+      "step": 43010
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 3.8857526779174805,
+      "learning_rate": 3.8057594579333713e-06,
+      "loss": 0.6077,
+      "step": 43020
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 7.122863292694092,
+      "learning_rate": 3.8019951063429327e-06,
+      "loss": 0.1793,
+      "step": 43030
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 21.782251358032227,
+      "learning_rate": 3.798230754752494e-06,
+      "loss": 0.5502,
+      "step": 43040
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 12.728435516357422,
+      "learning_rate": 3.7944664031620552e-06,
+      "loss": 0.6658,
+      "step": 43050
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 20.162574768066406,
+      "learning_rate": 3.790702051571617e-06,
+      "loss": 0.5913,
+      "step": 43060
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 0.8549222946166992,
+      "learning_rate": 3.7869376999811786e-06,
+      "loss": 0.2046,
+      "step": 43070
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 19.518613815307617,
+      "learning_rate": 3.78317334839074e-06,
+      "loss": 0.4695,
+      "step": 43080
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 8.266684532165527,
+      "learning_rate": 3.7794089968003015e-06,
+      "loss": 0.4126,
+      "step": 43090
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 21.11467170715332,
+      "learning_rate": 3.7756446452098625e-06,
+      "loss": 0.6289,
+      "step": 43100
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 25.66094970703125,
+      "learning_rate": 3.7718802936194244e-06,
+      "loss": 0.2911,
+      "step": 43110
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 23.889019012451172,
+      "learning_rate": 3.768115942028986e-06,
+      "loss": 0.6902,
+      "step": 43120
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 11.190979957580566,
+      "learning_rate": 3.7643515904385473e-06,
+      "loss": 0.4301,
+      "step": 43130
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 1.0701913833618164,
+      "learning_rate": 3.7605872388481084e-06,
+      "loss": 0.356,
+      "step": 43140
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 9.612865447998047,
+      "learning_rate": 3.7568228872576703e-06,
+      "loss": 0.487,
+      "step": 43150
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 0.3060283064842224,
+      "learning_rate": 3.7530585356672317e-06,
+      "loss": 0.2213,
+      "step": 43160
+    },
+    {
+      "epoch": 8.13,
+      "grad_norm": 1.2015522718429565,
+      "learning_rate": 3.749294184076793e-06,
+      "loss": 0.4213,
+      "step": 43170
+    },
+    {
+      "epoch": 8.13,
+      "grad_norm": 9.978540420532227,
+      "learning_rate": 3.7455298324863542e-06,
+      "loss": 0.6853,
+      "step": 43180
+    },
+    {
+      "epoch": 8.13,
+      "grad_norm": 10.406660079956055,
+      "learning_rate": 3.7417654808959157e-06,
+      "loss": 0.4558,
+      "step": 43190
+    },
+    {
+      "epoch": 8.13,
+      "grad_norm": 0.04582936689257622,
+      "learning_rate": 3.7380011293054776e-06,
+      "loss": 0.468,
+      "step": 43200
+    },
+    {
+      "epoch": 8.13,
+      "grad_norm": 4.475496292114258,
+      "learning_rate": 3.734236777715039e-06,
+      "loss": 0.5617,
+      "step": 43210
+    },
+    {
+      "epoch": 8.13,
+      "grad_norm": 11.952781677246094,
+      "learning_rate": 3.7304724261246005e-06,
+      "loss": 0.4628,
+      "step": 43220
+    },
+    {
+      "epoch": 8.14,
+      "grad_norm": 23.401905059814453,
+      "learning_rate": 3.7267080745341615e-06,
+      "loss": 0.3671,
+      "step": 43230
+    },
+    {
+      "epoch": 8.14,
+      "grad_norm": 11.260015487670898,
+      "learning_rate": 3.722943722943723e-06,
+      "loss": 0.5275,
+      "step": 43240
+    },
+    {
+      "epoch": 8.14,
+      "grad_norm": 7.514461994171143,
+      "learning_rate": 3.719179371353285e-06,
+      "loss": 0.6261,
+      "step": 43250
+    },
+    {
+      "epoch": 8.14,
+      "grad_norm": 13.519734382629395,
+      "learning_rate": 3.7154150197628463e-06,
+      "loss": 0.5934,
+      "step": 43260
+    },
+    {
+      "epoch": 8.14,
+      "grad_norm": 26.910083770751953,
+      "learning_rate": 3.7116506681724074e-06,
+      "loss": 0.7417,
+      "step": 43270
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 8.547293663024902,
+      "learning_rate": 3.707886316581969e-06,
+      "loss": 0.2699,
+      "step": 43280
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 0.09624161571264267,
+      "learning_rate": 3.7041219649915307e-06,
+      "loss": 0.5396,
+      "step": 43290
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 10.193184852600098,
+      "learning_rate": 3.700357613401092e-06,
+      "loss": 0.6458,
+      "step": 43300
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 4.950165271759033,
+      "learning_rate": 3.6965932618106532e-06,
+      "loss": 0.3573,
+      "step": 43310
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 20.236759185791016,
+      "learning_rate": 3.6928289102202147e-06,
+      "loss": 0.5585,
+      "step": 43320
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 7.94288969039917,
+      "learning_rate": 3.689064558629776e-06,
+      "loss": 0.5142,
+      "step": 43330
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 4.145201206207275,
+      "learning_rate": 3.685300207039338e-06,
+      "loss": 0.3554,
+      "step": 43340
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 14.267728805541992,
+      "learning_rate": 3.6815358554488995e-06,
+      "loss": 0.3611,
+      "step": 43350
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 0.9065011143684387,
+      "learning_rate": 3.6777715038584605e-06,
+      "loss": 0.2953,
+      "step": 43360
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 9.731773376464844,
+      "learning_rate": 3.674007152268022e-06,
+      "loss": 0.5442,
+      "step": 43370
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 23.468883514404297,
+      "learning_rate": 3.670242800677584e-06,
+      "loss": 0.6701,
+      "step": 43380
+    },
+    {
+      "epoch": 8.17,
+      "grad_norm": 9.353164672851562,
+      "learning_rate": 3.6664784490871453e-06,
+      "loss": 0.4046,
+      "step": 43390
+    },
+    {
+      "epoch": 8.17,
+      "grad_norm": 10.205562591552734,
+      "learning_rate": 3.6627140974967064e-06,
+      "loss": 0.3539,
+      "step": 43400
+    },
+    {
+      "epoch": 8.17,
+      "grad_norm": 7.568580150604248,
+      "learning_rate": 3.658949745906268e-06,
+      "loss": 0.7413,
+      "step": 43410
+    },
+    {
+      "epoch": 8.17,
+      "grad_norm": 0.29625949263572693,
+      "learning_rate": 3.6551853943158293e-06,
+      "loss": 0.3291,
+      "step": 43420
+    },
+    {
+      "epoch": 8.17,
+      "grad_norm": 8.475521087646484,
+      "learning_rate": 3.651421042725391e-06,
+      "loss": 0.239,
+      "step": 43430
+    },
+    {
+      "epoch": 8.18,
+      "grad_norm": 21.264280319213867,
+      "learning_rate": 3.6476566911349526e-06,
+      "loss": 0.6242,
+      "step": 43440
+    },
+    {
+      "epoch": 8.18,
+      "grad_norm": 7.065421104431152,
+      "learning_rate": 3.6438923395445137e-06,
+      "loss": 0.4781,
+      "step": 43450
+    },
+    {
+      "epoch": 8.18,
+      "grad_norm": 11.760064125061035,
+      "learning_rate": 3.640127987954075e-06,
+      "loss": 0.2676,
+      "step": 43460
+    },
+    {
+      "epoch": 8.18,
+      "grad_norm": 0.0757002905011177,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 0.4388,
+      "step": 43470
+    },
+    {
+      "epoch": 8.18,
+      "grad_norm": 6.151671886444092,
+      "learning_rate": 3.6325992847731985e-06,
+      "loss": 0.5118,
+      "step": 43480
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 0.9541302919387817,
+      "learning_rate": 3.6288349331827595e-06,
+      "loss": 0.7873,
+      "step": 43490
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 0.43069586157798767,
+      "learning_rate": 3.625070581592321e-06,
+      "loss": 0.4353,
+      "step": 43500
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 20.052248001098633,
+      "learning_rate": 3.6213062300018824e-06,
+      "loss": 0.4903,
+      "step": 43510
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 8.033267974853516,
+      "learning_rate": 3.6175418784114443e-06,
+      "loss": 0.5027,
+      "step": 43520
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 12.000602722167969,
+      "learning_rate": 3.613777526821005e-06,
+      "loss": 0.4379,
+      "step": 43530
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 0.025615401566028595,
+      "learning_rate": 3.610013175230567e-06,
+      "loss": 0.362,
+      "step": 43540
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 0.06578720360994339,
+      "learning_rate": 3.6062488236401283e-06,
+      "loss": 0.4048,
+      "step": 43550
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 19.945615768432617,
+      "learning_rate": 3.6024844720496897e-06,
+      "loss": 0.4744,
+      "step": 43560
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 1.860599398612976,
+      "learning_rate": 3.5987201204592516e-06,
+      "loss": 0.3348,
+      "step": 43570
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 0.18426920473575592,
+      "learning_rate": 3.5949557688688127e-06,
+      "loss": 0.3012,
+      "step": 43580
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 18.49991798400879,
+      "learning_rate": 3.591191417278374e-06,
+      "loss": 0.5334,
+      "step": 43590
+    },
+    {
+      "epoch": 8.21,
+      "grad_norm": 7.533271312713623,
+      "learning_rate": 3.5874270656879356e-06,
+      "loss": 0.269,
+      "step": 43600
+    },
+    {
+      "epoch": 8.21,
+      "grad_norm": 31.758325576782227,
+      "learning_rate": 3.583662714097497e-06,
+      "loss": 0.6352,
+      "step": 43610
+    },
+    {
+      "epoch": 8.21,
+      "grad_norm": 24.36396026611328,
+      "learning_rate": 3.579898362507058e-06,
+      "loss": 0.3615,
+      "step": 43620
+    },
+    {
+      "epoch": 8.21,
+      "grad_norm": 0.032064277678728104,
+      "learning_rate": 3.57613401091662e-06,
+      "loss": 0.7085,
+      "step": 43630
+    },
+    {
+      "epoch": 8.21,
+      "grad_norm": 10.80846881866455,
+      "learning_rate": 3.5723696593261814e-06,
+      "loss": 0.6805,
+      "step": 43640
+    },
+    {
+      "epoch": 8.22,
+      "grad_norm": 7.788702964782715,
+      "learning_rate": 3.568605307735743e-06,
+      "loss": 0.6523,
+      "step": 43650
+    },
+    {
+      "epoch": 8.22,
+      "grad_norm": 6.078163146972656,
+      "learning_rate": 3.564840956145304e-06,
+      "loss": 0.3491,
+      "step": 43660
+    },
+    {
+      "epoch": 8.22,
+      "grad_norm": 34.41270065307617,
+      "learning_rate": 3.561076604554866e-06,
+      "loss": 0.3832,
+      "step": 43670
+    },
+    {
+      "epoch": 8.22,
+      "grad_norm": 0.11715062707662582,
+      "learning_rate": 3.5573122529644273e-06,
+      "loss": 0.4516,
+      "step": 43680
+    },
+    {
+      "epoch": 8.22,
+      "grad_norm": 25.467695236206055,
+      "learning_rate": 3.5535479013739887e-06,
+      "loss": 0.3377,
+      "step": 43690
+    },
+    {
+      "epoch": 8.23,
+      "grad_norm": 0.11090195924043655,
+      "learning_rate": 3.54978354978355e-06,
+      "loss": 0.4038,
+      "step": 43700
+    },
+    {
+      "epoch": 8.23,
+      "grad_norm": 12.662226676940918,
+      "learning_rate": 3.5460191981931112e-06,
+      "loss": 0.379,
+      "step": 43710
+    },
+    {
+      "epoch": 8.23,
+      "grad_norm": 5.7085652351379395,
+      "learning_rate": 3.542254846602673e-06,
+      "loss": 0.3542,
+      "step": 43720
+    },
+    {
+      "epoch": 8.23,
+      "grad_norm": 15.676560401916504,
+      "learning_rate": 3.5384904950122346e-06,
+      "loss": 0.4124,
+      "step": 43730
+    },
+    {
+      "epoch": 8.23,
+      "grad_norm": 34.08810806274414,
+      "learning_rate": 3.534726143421796e-06,
+      "loss": 0.3937,
+      "step": 43740
+    },
+    {
+      "epoch": 8.23,
+      "grad_norm": 9.684732437133789,
+      "learning_rate": 3.530961791831357e-06,
+      "loss": 0.2242,
+      "step": 43750
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 8.743865013122559,
+      "learning_rate": 3.5271974402409185e-06,
+      "loss": 0.3141,
+      "step": 43760
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 9.374190330505371,
+      "learning_rate": 3.5234330886504804e-06,
+      "loss": 0.3993,
+      "step": 43770
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 6.62257719039917,
+      "learning_rate": 3.519668737060042e-06,
+      "loss": 0.4389,
+      "step": 43780
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 126.70000457763672,
+      "learning_rate": 3.515904385469603e-06,
+      "loss": 0.3082,
+      "step": 43790
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 16.551645278930664,
+      "learning_rate": 3.5121400338791644e-06,
+      "loss": 0.3682,
+      "step": 43800
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 0.19622400403022766,
+      "learning_rate": 3.5083756822887262e-06,
+      "loss": 0.389,
+      "step": 43810
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 11.304600715637207,
+      "learning_rate": 3.5046113306982877e-06,
+      "loss": 0.4126,
+      "step": 43820
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 13.223969459533691,
+      "learning_rate": 3.500846979107849e-06,
+      "loss": 0.3159,
+      "step": 43830
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 0.057929929345846176,
+      "learning_rate": 3.49708262751741e-06,
+      "loss": 0.6223,
+      "step": 43840
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 28.158432006835938,
+      "learning_rate": 3.4933182759269717e-06,
+      "loss": 0.3654,
+      "step": 43850
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 0.9253088235855103,
+      "learning_rate": 3.4895539243365336e-06,
+      "loss": 0.5879,
+      "step": 43860
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 25.360929489135742,
+      "learning_rate": 3.485789572746095e-06,
+      "loss": 0.5482,
+      "step": 43870
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 25.015487670898438,
+      "learning_rate": 3.482025221155656e-06,
+      "loss": 0.5676,
+      "step": 43880
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 25.207012176513672,
+      "learning_rate": 3.4782608695652175e-06,
+      "loss": 0.5568,
+      "step": 43890
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 29.287954330444336,
+      "learning_rate": 3.474496517974779e-06,
+      "loss": 0.5971,
+      "step": 43900
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 0.47074535489082336,
+      "learning_rate": 3.470732166384341e-06,
+      "loss": 0.3605,
+      "step": 43910
+    },
+    {
+      "epoch": 8.27,
+      "grad_norm": 7.249814510345459,
+      "learning_rate": 3.466967814793902e-06,
+      "loss": 0.4114,
+      "step": 43920
+    },
+    {
+      "epoch": 8.27,
+      "grad_norm": 16.79249382019043,
+      "learning_rate": 3.4632034632034634e-06,
+      "loss": 0.4512,
+      "step": 43930
+    },
+    {
+      "epoch": 8.27,
+      "grad_norm": 20.726879119873047,
+      "learning_rate": 3.459439111613025e-06,
+      "loss": 0.4038,
+      "step": 43940
+    },
+    {
+      "epoch": 8.27,
+      "grad_norm": 46.845523834228516,
+      "learning_rate": 3.4556747600225867e-06,
+      "loss": 0.4775,
+      "step": 43950
+    },
+    {
+      "epoch": 8.27,
+      "grad_norm": 18.067689895629883,
+      "learning_rate": 3.451910408432148e-06,
+      "loss": 0.4175,
+      "step": 43960
+    },
+    {
+      "epoch": 8.28,
+      "grad_norm": 42.61848831176758,
+      "learning_rate": 3.448146056841709e-06,
+      "loss": 0.6115,
+      "step": 43970
+    },
+    {
+      "epoch": 8.28,
+      "grad_norm": 6.6008687019348145,
+      "learning_rate": 3.4443817052512707e-06,
+      "loss": 0.3097,
+      "step": 43980
+    },
+    {
+      "epoch": 8.28,
+      "grad_norm": 27.931385040283203,
+      "learning_rate": 3.440617353660832e-06,
+      "loss": 0.3939,
+      "step": 43990
+    },
+    {
+      "epoch": 8.28,
+      "grad_norm": 17.77448844909668,
+      "learning_rate": 3.436853002070394e-06,
+      "loss": 0.299,
+      "step": 44000
+    },
+    {
+      "epoch": 8.28,
+      "grad_norm": 8.479036331176758,
+      "learning_rate": 3.433088650479955e-06,
+      "loss": 0.4371,
+      "step": 44010
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 12.390946388244629,
+      "learning_rate": 3.4293242988895165e-06,
+      "loss": 0.2047,
+      "step": 44020
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 12.151750564575195,
+      "learning_rate": 3.425559947299078e-06,
+      "loss": 0.5079,
+      "step": 44030
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 17.92153549194336,
+      "learning_rate": 3.4217955957086394e-06,
+      "loss": 0.543,
+      "step": 44040
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 8.07863998413086,
+      "learning_rate": 3.4180312441182005e-06,
+      "loss": 0.5345,
+      "step": 44050
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 0.12737701833248138,
+      "learning_rate": 3.4142668925277623e-06,
+      "loss": 0.7064,
+      "step": 44060
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 10.987157821655273,
+      "learning_rate": 3.410502540937324e-06,
+      "loss": 0.6823,
+      "step": 44070
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 12.050490379333496,
+      "learning_rate": 3.4067381893468853e-06,
+      "loss": 0.5315,
+      "step": 44080
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 17.466691970825195,
+      "learning_rate": 3.402973837756447e-06,
+      "loss": 0.5675,
+      "step": 44090
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 19.85411262512207,
+      "learning_rate": 3.399209486166008e-06,
+      "loss": 0.3984,
+      "step": 44100
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 12.507843971252441,
+      "learning_rate": 3.3954451345755696e-06,
+      "loss": 0.413,
+      "step": 44110
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 3.8064775466918945,
+      "learning_rate": 3.391680782985131e-06,
+      "loss": 0.4874,
+      "step": 44120
+    },
+    {
+      "epoch": 8.31,
+      "grad_norm": 20.31875228881836,
+      "learning_rate": 3.3879164313946926e-06,
+      "loss": 0.5857,
+      "step": 44130
+    },
+    {
+      "epoch": 8.31,
+      "grad_norm": 0.29422613978385925,
+      "learning_rate": 3.3841520798042536e-06,
+      "loss": 0.3885,
+      "step": 44140
+    },
+    {
+      "epoch": 8.31,
+      "grad_norm": 13.232914924621582,
+      "learning_rate": 3.3803877282138155e-06,
+      "loss": 0.2175,
+      "step": 44150
+    },
+    {
+      "epoch": 8.31,
+      "grad_norm": 0.48562586307525635,
+      "learning_rate": 3.376623376623377e-06,
+      "loss": 0.2482,
+      "step": 44160
+    },
+    {
+      "epoch": 8.31,
+      "grad_norm": 18.276710510253906,
+      "learning_rate": 3.3728590250329384e-06,
+      "loss": 0.5239,
+      "step": 44170
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 1.2684160470962524,
+      "learning_rate": 3.3690946734425e-06,
+      "loss": 0.2591,
+      "step": 44180
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 11.328255653381348,
+      "learning_rate": 3.365330321852061e-06,
+      "loss": 0.3205,
+      "step": 44190
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 8.913219451904297,
+      "learning_rate": 3.361565970261623e-06,
+      "loss": 0.4611,
+      "step": 44200
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 1.8168461322784424,
+      "learning_rate": 3.3578016186711843e-06,
+      "loss": 0.2763,
+      "step": 44210
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 9.876787185668945,
+      "learning_rate": 3.3540372670807457e-06,
+      "loss": 0.3158,
+      "step": 44220
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 26.141977310180664,
+      "learning_rate": 3.3502729154903068e-06,
+      "loss": 0.519,
+      "step": 44230
+    },
+    {
+      "epoch": 8.33,
+      "grad_norm": 0.03895975649356842,
+      "learning_rate": 3.3465085638998686e-06,
+      "loss": 0.3647,
+      "step": 44240
+    },
+    {
+      "epoch": 8.33,
+      "grad_norm": 28.49066162109375,
+      "learning_rate": 3.34274421230943e-06,
+      "loss": 0.5051,
+      "step": 44250
+    },
+    {
+      "epoch": 8.33,
+      "grad_norm": 48.42610168457031,
+      "learning_rate": 3.3389798607189916e-06,
+      "loss": 0.5039,
+      "step": 44260
+    },
+    {
+      "epoch": 8.33,
+      "grad_norm": 33.13661575317383,
+      "learning_rate": 3.3352155091285526e-06,
+      "loss": 0.6969,
+      "step": 44270
+    },
+    {
+      "epoch": 8.33,
+      "grad_norm": 24.356779098510742,
+      "learning_rate": 3.331451157538114e-06,
+      "loss": 0.646,
+      "step": 44280
+    },
+    {
+      "epoch": 8.34,
+      "grad_norm": 26.351469039916992,
+      "learning_rate": 3.327686805947676e-06,
+      "loss": 0.4839,
+      "step": 44290
+    },
+    {
+      "epoch": 8.34,
+      "grad_norm": 3.9066474437713623,
+      "learning_rate": 3.3239224543572374e-06,
+      "loss": 0.4166,
+      "step": 44300
+    },
+    {
+      "epoch": 8.34,
+      "grad_norm": 0.12449698895215988,
+      "learning_rate": 3.320158102766799e-06,
+      "loss": 0.213,
+      "step": 44310
+    },
+    {
+      "epoch": 8.34,
+      "grad_norm": 6.33676290512085,
+      "learning_rate": 3.31639375117636e-06,
+      "loss": 0.3959,
+      "step": 44320
+    },
+    {
+      "epoch": 8.34,
+      "grad_norm": 12.31187629699707,
+      "learning_rate": 3.3126293995859214e-06,
+      "loss": 0.6402,
+      "step": 44330
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 9.686624526977539,
+      "learning_rate": 3.3088650479954832e-06,
+      "loss": 0.4905,
+      "step": 44340
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 15.566709518432617,
+      "learning_rate": 3.3051006964050447e-06,
+      "loss": 0.3201,
+      "step": 44350
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 19.786592483520508,
+      "learning_rate": 3.3013363448146057e-06,
+      "loss": 0.4149,
+      "step": 44360
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 43.112449645996094,
+      "learning_rate": 3.297571993224167e-06,
+      "loss": 0.5619,
+      "step": 44370
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 32.787532806396484,
+      "learning_rate": 3.293807641633729e-06,
+      "loss": 0.4769,
+      "step": 44380
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 21.896869659423828,
+      "learning_rate": 3.2900432900432905e-06,
+      "loss": 0.3075,
+      "step": 44390
+    },
+    {
+      "epoch": 8.36,
+      "grad_norm": 2.6966474056243896,
+      "learning_rate": 3.2862789384528516e-06,
+      "loss": 0.6221,
+      "step": 44400
+    },
+    {
+      "epoch": 8.36,
+      "grad_norm": 0.19632263481616974,
+      "learning_rate": 3.282514586862413e-06,
+      "loss": 0.4079,
+      "step": 44410
+    },
+    {
+      "epoch": 8.36,
+      "grad_norm": 23.635337829589844,
+      "learning_rate": 3.2787502352719745e-06,
+      "loss": 0.3567,
+      "step": 44420
+    },
+    {
+      "epoch": 8.36,
+      "grad_norm": 13.236377716064453,
+      "learning_rate": 3.2749858836815364e-06,
+      "loss": 0.3691,
+      "step": 44430
+    },
+    {
+      "epoch": 8.36,
+      "grad_norm": 19.794361114501953,
+      "learning_rate": 3.271221532091098e-06,
+      "loss": 0.5006,
+      "step": 44440
+    },
+    {
+      "epoch": 8.37,
+      "grad_norm": 0.1621648520231247,
+      "learning_rate": 3.267457180500659e-06,
+      "loss": 0.6175,
+      "step": 44450
+    },
+    {
+      "epoch": 8.37,
+      "grad_norm": 0.08422687649726868,
+      "learning_rate": 3.2636928289102203e-06,
+      "loss": 0.4767,
+      "step": 44460
+    },
+    {
+      "epoch": 8.37,
+      "grad_norm": 6.153237819671631,
+      "learning_rate": 3.259928477319782e-06,
+      "loss": 0.3885,
+      "step": 44470
+    },
+    {
+      "epoch": 8.37,
+      "grad_norm": 18.663501739501953,
+      "learning_rate": 3.2561641257293437e-06,
+      "loss": 0.3928,
+      "step": 44480
+    },
+    {
+      "epoch": 8.37,
+      "grad_norm": 12.590583801269531,
+      "learning_rate": 3.2523997741389047e-06,
+      "loss": 0.6693,
+      "step": 44490
+    },
+    {
+      "epoch": 8.38,
+      "grad_norm": 0.5464389324188232,
+      "learning_rate": 3.248635422548466e-06,
+      "loss": 0.4404,
+      "step": 44500
+    },
+    {
+      "epoch": 8.38,
+      "grad_norm": 2.3248085975646973,
+      "learning_rate": 3.2448710709580277e-06,
+      "loss": 0.6251,
+      "step": 44510
+    },
+    {
+      "epoch": 8.38,
+      "grad_norm": 27.805416107177734,
+      "learning_rate": 3.2411067193675895e-06,
+      "loss": 0.1487,
+      "step": 44520
+    },
+    {
+      "epoch": 8.38,
+      "grad_norm": 31.039339065551758,
+      "learning_rate": 3.2373423677771506e-06,
+      "loss": 0.4554,
+      "step": 44530
+    },
+    {
+      "epoch": 8.38,
+      "grad_norm": 20.5913143157959,
+      "learning_rate": 3.233578016186712e-06,
+      "loss": 0.3836,
+      "step": 44540
+    },
+    {
+      "epoch": 8.39,
+      "grad_norm": 43.38515853881836,
+      "learning_rate": 3.2298136645962735e-06,
+      "loss": 0.4318,
+      "step": 44550
+    },
+    {
+      "epoch": 8.39,
+      "grad_norm": 12.96084976196289,
+      "learning_rate": 3.226049313005835e-06,
+      "loss": 0.3663,
+      "step": 44560
+    },
+    {
+      "epoch": 8.39,
+      "grad_norm": 8.591901779174805,
+      "learning_rate": 3.222284961415397e-06,
+      "loss": 0.4199,
+      "step": 44570
+    },
+    {
+      "epoch": 8.39,
+      "grad_norm": 14.201598167419434,
+      "learning_rate": 3.218520609824958e-06,
+      "loss": 0.6438,
+      "step": 44580
+    },
+    {
+      "epoch": 8.39,
+      "grad_norm": 20.398212432861328,
+      "learning_rate": 3.2147562582345193e-06,
+      "loss": 0.2309,
+      "step": 44590
+    },
+    {
+      "epoch": 8.39,
+      "grad_norm": 4.721048355102539,
+      "learning_rate": 3.210991906644081e-06,
+      "loss": 0.5085,
+      "step": 44600
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 11.639357566833496,
+      "learning_rate": 3.2072275550536427e-06,
+      "loss": 0.5171,
+      "step": 44610
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 9.085111618041992,
+      "learning_rate": 3.2034632034632033e-06,
+      "loss": 0.4878,
+      "step": 44620
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 14.544848442077637,
+      "learning_rate": 3.199698851872765e-06,
+      "loss": 0.4514,
+      "step": 44630
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 3.9646859169006348,
+      "learning_rate": 3.1959345002823266e-06,
+      "loss": 0.4601,
+      "step": 44640
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 25.853208541870117,
+      "learning_rate": 3.192170148691888e-06,
+      "loss": 0.5069,
+      "step": 44650
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 0.13335028290748596,
+      "learning_rate": 3.188405797101449e-06,
+      "loss": 0.4518,
+      "step": 44660
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 4.05998420715332,
+      "learning_rate": 3.184641445511011e-06,
+      "loss": 0.2396,
+      "step": 44670
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 15.973206520080566,
+      "learning_rate": 3.1808770939205725e-06,
+      "loss": 0.3994,
+      "step": 44680
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 5.575290679931641,
+      "learning_rate": 3.177112742330134e-06,
+      "loss": 0.8244,
+      "step": 44690
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 78.56654357910156,
+      "learning_rate": 3.1733483907396954e-06,
+      "loss": 0.8038,
+      "step": 44700
+    },
+    {
+      "epoch": 8.42,
+      "grad_norm": 0.16877728700637817,
+      "learning_rate": 3.1695840391492564e-06,
+      "loss": 0.3286,
+      "step": 44710
+    },
+    {
+      "epoch": 8.42,
+      "grad_norm": 9.68952751159668,
+      "learning_rate": 3.1658196875588183e-06,
+      "loss": 0.5651,
+      "step": 44720
+    },
+    {
+      "epoch": 8.42,
+      "grad_norm": 47.49652862548828,
+      "learning_rate": 3.1620553359683798e-06,
+      "loss": 0.3635,
+      "step": 44730
+    },
+    {
+      "epoch": 8.42,
+      "grad_norm": 19.1754207611084,
+      "learning_rate": 3.1582909843779412e-06,
+      "loss": 0.5277,
+      "step": 44740
+    },
+    {
+      "epoch": 8.42,
+      "grad_norm": 21.708391189575195,
+      "learning_rate": 3.1545266327875023e-06,
+      "loss": 0.4516,
+      "step": 44750
+    },
+    {
+      "epoch": 8.42,
+      "grad_norm": 2.858497381210327,
+      "learning_rate": 3.1507622811970637e-06,
+      "loss": 0.5113,
+      "step": 44760
+    },
+    {
+      "epoch": 8.43,
+      "grad_norm": 14.189351081848145,
+      "learning_rate": 3.1469979296066256e-06,
+      "loss": 0.3809,
+      "step": 44770
+    },
+    {
+      "epoch": 8.43,
+      "grad_norm": 15.705812454223633,
+      "learning_rate": 3.143233578016187e-06,
+      "loss": 0.4126,
+      "step": 44780
+    },
+    {
+      "epoch": 8.43,
+      "grad_norm": 66.25598907470703,
+      "learning_rate": 3.1394692264257485e-06,
+      "loss": 0.6253,
+      "step": 44790
+    },
+    {
+      "epoch": 8.43,
+      "grad_norm": 7.330162048339844,
+      "learning_rate": 3.1357048748353096e-06,
+      "loss": 0.5187,
+      "step": 44800
+    },
+    {
+      "epoch": 8.43,
+      "grad_norm": 5.743100166320801,
+      "learning_rate": 3.1319405232448715e-06,
+      "loss": 0.4984,
+      "step": 44810
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 8.980222702026367,
+      "learning_rate": 3.128176171654433e-06,
+      "loss": 0.4996,
+      "step": 44820
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 0.5218448042869568,
+      "learning_rate": 3.1244118200639944e-06,
+      "loss": 0.5344,
+      "step": 44830
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 1.2095797061920166,
+      "learning_rate": 3.1206474684735554e-06,
+      "loss": 0.2706,
+      "step": 44840
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 44.43700408935547,
+      "learning_rate": 3.116883116883117e-06,
+      "loss": 0.4895,
+      "step": 44850
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 5.458186626434326,
+      "learning_rate": 3.1131187652926788e-06,
+      "loss": 0.3371,
+      "step": 44860
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 6.718016624450684,
+      "learning_rate": 3.1093544137022402e-06,
+      "loss": 0.2024,
+      "step": 44870
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 0.143607959151268,
+      "learning_rate": 3.1055900621118013e-06,
+      "loss": 0.2257,
+      "step": 44880
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 10.796164512634277,
+      "learning_rate": 3.1018257105213627e-06,
+      "loss": 0.6509,
+      "step": 44890
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 9.629598617553711,
+      "learning_rate": 3.0980613589309246e-06,
+      "loss": 0.5034,
+      "step": 44900
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 0.05811280012130737,
+      "learning_rate": 3.094297007340486e-06,
+      "loss": 0.4932,
+      "step": 44910
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 9.126168251037598,
+      "learning_rate": 3.0905326557500475e-06,
+      "loss": 0.4464,
+      "step": 44920
+    },
+    {
+      "epoch": 8.46,
+      "grad_norm": 15.563265800476074,
+      "learning_rate": 3.0867683041596086e-06,
+      "loss": 0.4157,
+      "step": 44930
+    },
+    {
+      "epoch": 8.46,
+      "grad_norm": 14.518123626708984,
+      "learning_rate": 3.08300395256917e-06,
+      "loss": 0.3727,
+      "step": 44940
+    },
+    {
+      "epoch": 8.46,
+      "grad_norm": 5.1271138191223145,
+      "learning_rate": 3.079239600978732e-06,
+      "loss": 0.6179,
+      "step": 44950
+    },
+    {
+      "epoch": 8.46,
+      "grad_norm": 0.39322352409362793,
+      "learning_rate": 3.0754752493882934e-06,
+      "loss": 0.2365,
+      "step": 44960
+    },
+    {
+      "epoch": 8.46,
+      "grad_norm": 3.110018014907837,
+      "learning_rate": 3.0717108977978544e-06,
+      "loss": 0.4637,
+      "step": 44970
+    },
+    {
+      "epoch": 8.47,
+      "grad_norm": 0.4475475251674652,
+      "learning_rate": 3.067946546207416e-06,
+      "loss": 0.4693,
+      "step": 44980
+    },
+    {
+      "epoch": 8.47,
+      "grad_norm": 31.466312408447266,
+      "learning_rate": 3.0641821946169773e-06,
+      "loss": 0.6135,
+      "step": 44990
+    },
+    {
+      "epoch": 8.47,
+      "grad_norm": 8.0161714553833,
+      "learning_rate": 3.0604178430265392e-06,
+      "loss": 0.4308,
+      "step": 45000
+    },
+    {
+      "epoch": 8.47,
+      "grad_norm": 0.21481114625930786,
+      "learning_rate": 3.0566534914361003e-06,
+      "loss": 0.3452,
+      "step": 45010
+    },
+    {
+      "epoch": 8.47,
+      "grad_norm": 6.250048637390137,
+      "learning_rate": 3.0528891398456617e-06,
+      "loss": 0.3198,
+      "step": 45020
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 14.705049514770508,
+      "learning_rate": 3.049124788255223e-06,
+      "loss": 0.5674,
+      "step": 45030
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 13.137300491333008,
+      "learning_rate": 3.045360436664785e-06,
+      "loss": 0.4053,
+      "step": 45040
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 5.976210594177246,
+      "learning_rate": 3.0415960850743465e-06,
+      "loss": 0.5598,
+      "step": 45050
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 23.372941970825195,
+      "learning_rate": 3.0378317334839076e-06,
+      "loss": 0.9162,
+      "step": 45060
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 16.597759246826172,
+      "learning_rate": 3.034067381893469e-06,
+      "loss": 0.4671,
+      "step": 45070
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 21.51179313659668,
+      "learning_rate": 3.0303030303030305e-06,
+      "loss": 0.5754,
+      "step": 45080
+    },
+    {
+      "epoch": 8.49,
+      "grad_norm": 1.9006558656692505,
+      "learning_rate": 3.0265386787125924e-06,
+      "loss": 0.4043,
+      "step": 45090
+    },
+    {
+      "epoch": 8.49,
+      "grad_norm": 12.068243980407715,
+      "learning_rate": 3.0227743271221534e-06,
+      "loss": 0.3859,
+      "step": 45100
+    },
+    {
+      "epoch": 8.49,
+      "grad_norm": 7.060889720916748,
+      "learning_rate": 3.019009975531715e-06,
+      "loss": 0.3461,
+      "step": 45110
+    },
+    {
+      "epoch": 8.49,
+      "grad_norm": 8.987565040588379,
+      "learning_rate": 3.0152456239412763e-06,
+      "loss": 0.5619,
+      "step": 45120
+    },
+    {
+      "epoch": 8.49,
+      "grad_norm": 7.495594501495361,
+      "learning_rate": 3.011481272350838e-06,
+      "loss": 0.2997,
+      "step": 45130
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 19.95897102355957,
+      "learning_rate": 3.007716920760399e-06,
+      "loss": 0.5481,
+      "step": 45140
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 13.875322341918945,
+      "learning_rate": 3.0039525691699607e-06,
+      "loss": 0.4323,
+      "step": 45150
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 13.79182243347168,
+      "learning_rate": 3.000188217579522e-06,
+      "loss": 0.2615,
+      "step": 45160
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 1.8631694316864014,
+      "learning_rate": 2.9964238659890836e-06,
+      "loss": 0.7032,
+      "step": 45170
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 5.367626190185547,
+      "learning_rate": 2.9926595143986455e-06,
+      "loss": 0.509,
+      "step": 45180
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 6.53688907623291,
+      "learning_rate": 2.9888951628082066e-06,
+      "loss": 0.4337,
+      "step": 45190
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 17.974353790283203,
+      "learning_rate": 2.985130811217768e-06,
+      "loss": 0.7414,
+      "step": 45200
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 17.594377517700195,
+      "learning_rate": 2.9813664596273295e-06,
+      "loss": 0.5211,
+      "step": 45210
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 29.985471725463867,
+      "learning_rate": 2.977602108036891e-06,
+      "loss": 0.6543,
+      "step": 45220
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 0.37220674753189087,
+      "learning_rate": 2.973837756446452e-06,
+      "loss": 0.183,
+      "step": 45230
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 51.8664436340332,
+      "learning_rate": 2.970073404856014e-06,
+      "loss": 0.4546,
+      "step": 45240
+    },
+    {
+      "epoch": 8.52,
+      "grad_norm": 0.12566237151622772,
+      "learning_rate": 2.9663090532655753e-06,
+      "loss": 0.6928,
+      "step": 45250
+    },
+    {
+      "epoch": 8.52,
+      "grad_norm": 27.250438690185547,
+      "learning_rate": 2.9625447016751368e-06,
+      "loss": 0.6348,
+      "step": 45260
+    },
+    {
+      "epoch": 8.52,
+      "grad_norm": 1.0455703735351562,
+      "learning_rate": 2.958780350084698e-06,
+      "loss": 0.51,
+      "step": 45270
+    },
+    {
+      "epoch": 8.52,
+      "grad_norm": 18.566070556640625,
+      "learning_rate": 2.9550159984942593e-06,
+      "loss": 0.4247,
+      "step": 45280
+    },
+    {
+      "epoch": 8.52,
+      "grad_norm": 22.149303436279297,
+      "learning_rate": 2.951251646903821e-06,
+      "loss": 0.5396,
+      "step": 45290
+    },
+    {
+      "epoch": 8.53,
+      "grad_norm": 7.4732561111450195,
+      "learning_rate": 2.9474872953133826e-06,
+      "loss": 0.6225,
+      "step": 45300
+    },
+    {
+      "epoch": 8.53,
+      "grad_norm": 21.741275787353516,
+      "learning_rate": 2.943722943722944e-06,
+      "loss": 0.7131,
+      "step": 45310
+    },
+    {
+      "epoch": 8.53,
+      "grad_norm": 3.925509214401245,
+      "learning_rate": 2.939958592132505e-06,
+      "loss": 0.3896,
+      "step": 45320
+    },
+    {
+      "epoch": 8.53,
+      "grad_norm": 0.22670765221118927,
+      "learning_rate": 2.936194240542067e-06,
+      "loss": 0.2686,
+      "step": 45330
+    },
+    {
+      "epoch": 8.53,
+      "grad_norm": 9.519122123718262,
+      "learning_rate": 2.9324298889516285e-06,
+      "loss": 0.5187,
+      "step": 45340
+    },
+    {
+      "epoch": 8.54,
+      "grad_norm": 1.0992772579193115,
+      "learning_rate": 2.92866553736119e-06,
+      "loss": 0.2774,
+      "step": 45350
+    },
+    {
+      "epoch": 8.54,
+      "grad_norm": 16.670631408691406,
+      "learning_rate": 2.924901185770751e-06,
+      "loss": 0.7732,
+      "step": 45360
+    },
+    {
+      "epoch": 8.54,
+      "grad_norm": 0.2704927623271942,
+      "learning_rate": 2.9211368341803124e-06,
+      "loss": 0.3941,
+      "step": 45370
+    },
+    {
+      "epoch": 8.54,
+      "grad_norm": 0.42437541484832764,
+      "learning_rate": 2.9173724825898743e-06,
+      "loss": 0.689,
+      "step": 45380
+    },
+    {
+      "epoch": 8.54,
+      "grad_norm": 1.1275883913040161,
+      "learning_rate": 2.9136081309994358e-06,
+      "loss": 0.3139,
+      "step": 45390
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 19.983110427856445,
+      "learning_rate": 2.9098437794089972e-06,
+      "loss": 0.5787,
+      "step": 45400
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 17.412151336669922,
+      "learning_rate": 2.9060794278185583e-06,
+      "loss": 0.4265,
+      "step": 45410
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 1.204263687133789,
+      "learning_rate": 2.9023150762281197e-06,
+      "loss": 0.8414,
+      "step": 45420
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 18.92685890197754,
+      "learning_rate": 2.8985507246376816e-06,
+      "loss": 0.5895,
+      "step": 45430
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 20.907590866088867,
+      "learning_rate": 2.894786373047243e-06,
+      "loss": 0.6537,
+      "step": 45440
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 0.030812498182058334,
+      "learning_rate": 2.891022021456804e-06,
+      "loss": 0.8201,
+      "step": 45450
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 19.68631935119629,
+      "learning_rate": 2.8872576698663656e-06,
+      "loss": 0.27,
+      "step": 45460
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 0.13562491536140442,
+      "learning_rate": 2.8834933182759275e-06,
+      "loss": 0.7518,
+      "step": 45470
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 13.175758361816406,
+      "learning_rate": 2.879728966685489e-06,
+      "loss": 0.2638,
+      "step": 45480
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 38.72014617919922,
+      "learning_rate": 2.87596461509505e-06,
+      "loss": 0.6186,
+      "step": 45490
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 5.701416969299316,
+      "learning_rate": 2.8722002635046114e-06,
+      "loss": 0.4767,
+      "step": 45500
+    },
+    {
+      "epoch": 8.57,
+      "grad_norm": 6.803305625915527,
+      "learning_rate": 2.868435911914173e-06,
+      "loss": 0.3493,
+      "step": 45510
+    },
+    {
+      "epoch": 8.57,
+      "grad_norm": 34.136417388916016,
+      "learning_rate": 2.8646715603237348e-06,
+      "loss": 0.3706,
+      "step": 45520
+    },
+    {
+      "epoch": 8.57,
+      "grad_norm": 15.489117622375488,
+      "learning_rate": 2.8609072087332962e-06,
+      "loss": 0.5064,
+      "step": 45530
+    },
+    {
+      "epoch": 8.57,
+      "grad_norm": 5.791557788848877,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.4265,
+      "step": 45540
+    },
+    {
+      "epoch": 8.57,
+      "grad_norm": 5.909194469451904,
+      "learning_rate": 2.8533785055524187e-06,
+      "loss": 0.4563,
+      "step": 45550
+    },
+    {
+      "epoch": 8.58,
+      "grad_norm": 21.65177345275879,
+      "learning_rate": 2.84961415396198e-06,
+      "loss": 0.3891,
+      "step": 45560
+    },
+    {
+      "epoch": 8.58,
+      "grad_norm": 3.773228645324707,
+      "learning_rate": 2.845849802371542e-06,
+      "loss": 0.501,
+      "step": 45570
+    },
+    {
+      "epoch": 8.58,
+      "grad_norm": 9.513245582580566,
+      "learning_rate": 2.842085450781103e-06,
+      "loss": 0.4636,
+      "step": 45580
+    },
+    {
+      "epoch": 8.58,
+      "grad_norm": 8.289735794067383,
+      "learning_rate": 2.8383210991906646e-06,
+      "loss": 0.4985,
+      "step": 45590
+    },
+    {
+      "epoch": 8.58,
+      "grad_norm": 0.9724116921424866,
+      "learning_rate": 2.834556747600226e-06,
+      "loss": 0.5116,
+      "step": 45600
+    },
+    {
+      "epoch": 8.58,
+      "grad_norm": 21.775800704956055,
+      "learning_rate": 2.830792396009788e-06,
+      "loss": 0.5441,
+      "step": 45610
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 48.54513931274414,
+      "learning_rate": 2.827028044419349e-06,
+      "loss": 0.3705,
+      "step": 45620
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 0.5952854752540588,
+      "learning_rate": 2.8232636928289104e-06,
+      "loss": 0.3866,
+      "step": 45630
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 12.538860321044922,
+      "learning_rate": 2.819499341238472e-06,
+      "loss": 0.4496,
+      "step": 45640
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 46.39492416381836,
+      "learning_rate": 2.8157349896480333e-06,
+      "loss": 0.3195,
+      "step": 45650
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 13.821070671081543,
+      "learning_rate": 2.811970638057595e-06,
+      "loss": 0.2962,
+      "step": 45660
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 6.4973225593566895,
+      "learning_rate": 2.8082062864671562e-06,
+      "loss": 0.4642,
+      "step": 45670
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 9.997817039489746,
+      "learning_rate": 2.8044419348767177e-06,
+      "loss": 0.2811,
+      "step": 45680
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 25.398149490356445,
+      "learning_rate": 2.800677583286279e-06,
+      "loss": 0.5049,
+      "step": 45690
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 22.4417667388916,
+      "learning_rate": 2.7969132316958406e-06,
+      "loss": 0.5439,
+      "step": 45700
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 19.63978385925293,
+      "learning_rate": 2.7931488801054017e-06,
+      "loss": 0.5902,
+      "step": 45710
+    },
+    {
+      "epoch": 8.61,
+      "grad_norm": 0.3071557581424713,
+      "learning_rate": 2.7893845285149635e-06,
+      "loss": 0.3403,
+      "step": 45720
+    },
+    {
+      "epoch": 8.61,
+      "grad_norm": 35.829811096191406,
+      "learning_rate": 2.785620176924525e-06,
+      "loss": 0.3616,
+      "step": 45730
+    },
+    {
+      "epoch": 8.61,
+      "grad_norm": 14.984370231628418,
+      "learning_rate": 2.7818558253340865e-06,
+      "loss": 0.7215,
+      "step": 45740
+    },
+    {
+      "epoch": 8.61,
+      "grad_norm": 8.33390998840332,
+      "learning_rate": 2.7780914737436475e-06,
+      "loss": 0.6237,
+      "step": 45750
+    },
+    {
+      "epoch": 8.61,
+      "grad_norm": 22.571699142456055,
+      "learning_rate": 2.7743271221532094e-06,
+      "loss": 0.6739,
+      "step": 45760
+    },
+    {
+      "epoch": 8.61,
+      "grad_norm": 17.001461029052734,
+      "learning_rate": 2.770562770562771e-06,
+      "loss": 0.2547,
+      "step": 45770
+    },
+    {
+      "epoch": 8.62,
+      "grad_norm": 19.975543975830078,
+      "learning_rate": 2.7667984189723323e-06,
+      "loss": 0.5098,
+      "step": 45780
+    },
+    {
+      "epoch": 8.62,
+      "grad_norm": 8.409460067749023,
+      "learning_rate": 2.7630340673818938e-06,
+      "loss": 0.407,
+      "step": 45790
+    },
+    {
+      "epoch": 8.62,
+      "grad_norm": 11.9586820602417,
+      "learning_rate": 2.759269715791455e-06,
+      "loss": 0.6268,
+      "step": 45800
+    },
+    {
+      "epoch": 8.62,
+      "grad_norm": 25.05490493774414,
+      "learning_rate": 2.7555053642010167e-06,
+      "loss": 0.3221,
+      "step": 45810
+    },
+    {
+      "epoch": 8.62,
+      "grad_norm": 16.877796173095703,
+      "learning_rate": 2.751741012610578e-06,
+      "loss": 0.3368,
+      "step": 45820
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 37.895652770996094,
+      "learning_rate": 2.7479766610201396e-06,
+      "loss": 0.5498,
+      "step": 45830
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 16.549964904785156,
+      "learning_rate": 2.7442123094297007e-06,
+      "loss": 0.6585,
+      "step": 45840
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 23.333738327026367,
+      "learning_rate": 2.740447957839262e-06,
+      "loss": 0.4209,
+      "step": 45850
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 17.801307678222656,
+      "learning_rate": 2.736683606248824e-06,
+      "loss": 0.3118,
+      "step": 45860
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 4.754730224609375,
+      "learning_rate": 2.7329192546583855e-06,
+      "loss": 0.3706,
+      "step": 45870
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 12.892595291137695,
+      "learning_rate": 2.7291549030679465e-06,
+      "loss": 0.5805,
+      "step": 45880
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 16.939796447753906,
+      "learning_rate": 2.725390551477508e-06,
+      "loss": 0.5106,
+      "step": 45890
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 16.29544448852539,
+      "learning_rate": 2.72162619988707e-06,
+      "loss": 0.4779,
+      "step": 45900
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 28.001670837402344,
+      "learning_rate": 2.7178618482966313e-06,
+      "loss": 0.1302,
+      "step": 45910
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 0.1437859982252121,
+      "learning_rate": 2.7140974967061928e-06,
+      "loss": 0.4312,
+      "step": 45920
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 18.679183959960938,
+      "learning_rate": 2.710333145115754e-06,
+      "loss": 0.2055,
+      "step": 45930
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 19.28763198852539,
+      "learning_rate": 2.7065687935253153e-06,
+      "loss": 0.828,
+      "step": 45940
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 77.62793731689453,
+      "learning_rate": 2.702804441934877e-06,
+      "loss": 0.5013,
+      "step": 45950
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 21.08930015563965,
+      "learning_rate": 2.6990400903444386e-06,
+      "loss": 0.5488,
+      "step": 45960
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 25.826997756958008,
+      "learning_rate": 2.6952757387539996e-06,
+      "loss": 0.3493,
+      "step": 45970
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 4.74186897277832,
+      "learning_rate": 2.691511387163561e-06,
+      "loss": 0.5234,
+      "step": 45980
+    },
+    {
+      "epoch": 8.66,
+      "grad_norm": 8.202046394348145,
+      "learning_rate": 2.687747035573123e-06,
+      "loss": 0.5571,
+      "step": 45990
+    },
+    {
+      "epoch": 8.66,
+      "grad_norm": 19.04949188232422,
+      "learning_rate": 2.6839826839826844e-06,
+      "loss": 0.4959,
+      "step": 46000
+    },
+    {
+      "epoch": 8.66,
+      "grad_norm": 10.953725814819336,
+      "learning_rate": 2.6802183323922455e-06,
+      "loss": 0.5771,
+      "step": 46010
+    },
+    {
+      "epoch": 8.66,
+      "grad_norm": 1.0045043230056763,
+      "learning_rate": 2.676453980801807e-06,
+      "loss": 0.3382,
+      "step": 46020
+    },
+    {
+      "epoch": 8.66,
+      "grad_norm": 0.04298378899693489,
+      "learning_rate": 2.6726896292113684e-06,
+      "loss": 0.4584,
+      "step": 46030
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 0.0919141098856926,
+      "learning_rate": 2.6689252776209303e-06,
+      "loss": 0.5683,
+      "step": 46040
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 27.5792293548584,
+      "learning_rate": 2.6651609260304918e-06,
+      "loss": 0.5817,
+      "step": 46050
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 5.107207775115967,
+      "learning_rate": 2.661396574440053e-06,
+      "loss": 0.3344,
+      "step": 46060
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 0.08540444076061249,
+      "learning_rate": 2.6576322228496142e-06,
+      "loss": 0.4315,
+      "step": 46070
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 6.959580898284912,
+      "learning_rate": 2.6538678712591757e-06,
+      "loss": 0.7812,
+      "step": 46080
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 21.82856559753418,
+      "learning_rate": 2.6501035196687376e-06,
+      "loss": 0.4448,
+      "step": 46090
+    },
+    {
+      "epoch": 8.68,
+      "grad_norm": 17.987428665161133,
+      "learning_rate": 2.6463391680782986e-06,
+      "loss": 0.5252,
+      "step": 46100
+    },
+    {
+      "epoch": 8.68,
+      "grad_norm": 29.11836814880371,
+      "learning_rate": 2.64257481648786e-06,
+      "loss": 0.5055,
+      "step": 46110
+    },
+    {
+      "epoch": 8.68,
+      "grad_norm": 10.755464553833008,
+      "learning_rate": 2.6388104648974216e-06,
+      "loss": 0.2324,
+      "step": 46120
+    },
+    {
+      "epoch": 8.68,
+      "grad_norm": 20.402507781982422,
+      "learning_rate": 2.6350461133069834e-06,
+      "loss": 0.3029,
+      "step": 46130
+    },
+    {
+      "epoch": 8.68,
+      "grad_norm": 4.016020774841309,
+      "learning_rate": 2.631281761716545e-06,
+      "loss": 0.3907,
+      "step": 46140
+    },
+    {
+      "epoch": 8.69,
+      "grad_norm": 14.392875671386719,
+      "learning_rate": 2.627517410126106e-06,
+      "loss": 0.5084,
+      "step": 46150
+    },
+    {
+      "epoch": 8.69,
+      "grad_norm": 1.0872794389724731,
+      "learning_rate": 2.6237530585356674e-06,
+      "loss": 0.658,
+      "step": 46160
+    },
+    {
+      "epoch": 8.69,
+      "grad_norm": 5.119357109069824,
+      "learning_rate": 2.619988706945229e-06,
+      "loss": 0.3549,
+      "step": 46170
+    },
+    {
+      "epoch": 8.69,
+      "grad_norm": 0.06370334327220917,
+      "learning_rate": 2.6162243553547907e-06,
+      "loss": 0.4486,
+      "step": 46180
+    },
+    {
+      "epoch": 8.69,
+      "grad_norm": 8.075560569763184,
+      "learning_rate": 2.6124600037643518e-06,
+      "loss": 0.4195,
+      "step": 46190
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 27.81960678100586,
+      "learning_rate": 2.6086956521739132e-06,
+      "loss": 0.7454,
+      "step": 46200
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 16.909664154052734,
+      "learning_rate": 2.6049313005834747e-06,
+      "loss": 0.5945,
+      "step": 46210
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 1.0878101587295532,
+      "learning_rate": 2.601166948993036e-06,
+      "loss": 0.3299,
+      "step": 46220
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 19.037450790405273,
+      "learning_rate": 2.597402597402597e-06,
+      "loss": 0.338,
+      "step": 46230
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 14.639829635620117,
+      "learning_rate": 2.593638245812159e-06,
+      "loss": 0.425,
+      "step": 46240
+    },
+    {
+      "epoch": 8.71,
+      "grad_norm": 0.7264255881309509,
+      "learning_rate": 2.5898738942217205e-06,
+      "loss": 0.3609,
+      "step": 46250
+    },
+    {
+      "epoch": 8.71,
+      "grad_norm": 8.234590530395508,
+      "learning_rate": 2.586109542631282e-06,
+      "loss": 0.4549,
+      "step": 46260
+    },
+    {
+      "epoch": 8.71,
+      "grad_norm": 9.959664344787598,
+      "learning_rate": 2.582345191040844e-06,
+      "loss": 0.3512,
+      "step": 46270
+    },
+    {
+      "epoch": 8.71,
+      "grad_norm": 0.06254367530345917,
+      "learning_rate": 2.578580839450405e-06,
+      "loss": 0.2918,
+      "step": 46280
+    },
+    {
+      "epoch": 8.71,
+      "grad_norm": 2.8327713012695312,
+      "learning_rate": 2.5748164878599664e-06,
+      "loss": 0.5665,
+      "step": 46290
+    },
+    {
+      "epoch": 8.71,
+      "grad_norm": 6.886368751525879,
+      "learning_rate": 2.571052136269528e-06,
+      "loss": 0.4747,
+      "step": 46300
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 10.857604026794434,
+      "learning_rate": 2.5672877846790893e-06,
+      "loss": 0.7911,
+      "step": 46310
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 17.762399673461914,
+      "learning_rate": 2.5635234330886503e-06,
+      "loss": 0.2114,
+      "step": 46320
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 9.51750659942627,
+      "learning_rate": 2.5597590814982122e-06,
+      "loss": 0.2595,
+      "step": 46330
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 9.107878684997559,
+      "learning_rate": 2.5559947299077737e-06,
+      "loss": 0.514,
+      "step": 46340
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 0.17868681252002716,
+      "learning_rate": 2.552230378317335e-06,
+      "loss": 0.3083,
+      "step": 46350
+    },
+    {
+      "epoch": 8.73,
+      "grad_norm": 0.14755873382091522,
+      "learning_rate": 2.548466026726896e-06,
+      "loss": 0.6239,
+      "step": 46360
+    },
+    {
+      "epoch": 8.73,
+      "grad_norm": 4.853813648223877,
+      "learning_rate": 2.5447016751364576e-06,
+      "loss": 0.4185,
+      "step": 46370
+    },
+    {
+      "epoch": 8.73,
+      "grad_norm": 20.324190139770508,
+      "learning_rate": 2.5409373235460195e-06,
+      "loss": 0.6337,
+      "step": 46380
+    },
+    {
+      "epoch": 8.73,
+      "grad_norm": 0.38305485248565674,
+      "learning_rate": 2.537172971955581e-06,
+      "loss": 0.3451,
+      "step": 46390
+    },
+    {
+      "epoch": 8.73,
+      "grad_norm": 15.589489936828613,
+      "learning_rate": 2.5334086203651425e-06,
+      "loss": 0.7042,
+      "step": 46400
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 35.21918869018555,
+      "learning_rate": 2.5296442687747035e-06,
+      "loss": 0.6307,
+      "step": 46410
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 0.9967852830886841,
+      "learning_rate": 2.5258799171842654e-06,
+      "loss": 0.457,
+      "step": 46420
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 18.615461349487305,
+      "learning_rate": 2.522115565593827e-06,
+      "loss": 0.4258,
+      "step": 46430
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 7.5972981452941895,
+      "learning_rate": 2.5183512140033883e-06,
+      "loss": 0.3654,
+      "step": 46440
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 0.10862822085618973,
+      "learning_rate": 2.5145868624129493e-06,
+      "loss": 0.386,
+      "step": 46450
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 13.601619720458984,
+      "learning_rate": 2.510822510822511e-06,
+      "loss": 0.4652,
+      "step": 46460
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 11.241927146911621,
+      "learning_rate": 2.5070581592320727e-06,
+      "loss": 0.2974,
+      "step": 46470
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 4.576162338256836,
+      "learning_rate": 2.503293807641634e-06,
+      "loss": 0.4565,
+      "step": 46480
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 1.5839916467666626,
+      "learning_rate": 2.4995294560511956e-06,
+      "loss": 0.3777,
+      "step": 46490
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 25.34798812866211,
+      "learning_rate": 2.4957651044607566e-06,
+      "loss": 0.4831,
+      "step": 46500
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 0.46405714750289917,
+      "learning_rate": 2.492000752870318e-06,
+      "loss": 0.3859,
+      "step": 46510
+    },
+    {
+      "epoch": 8.76,
+      "grad_norm": 12.911664962768555,
+      "learning_rate": 2.48823640127988e-06,
+      "loss": 0.4369,
+      "step": 46520
+    },
+    {
+      "epoch": 8.76,
+      "grad_norm": 10.62582778930664,
+      "learning_rate": 2.484472049689441e-06,
+      "loss": 0.2246,
+      "step": 46530
+    },
+    {
+      "epoch": 8.76,
+      "grad_norm": 15.52188777923584,
+      "learning_rate": 2.480707698099003e-06,
+      "loss": 0.3334,
+      "step": 46540
+    },
+    {
+      "epoch": 8.76,
+      "grad_norm": 8.674419403076172,
+      "learning_rate": 2.476943346508564e-06,
+      "loss": 0.4668,
+      "step": 46550
+    },
+    {
+      "epoch": 8.76,
+      "grad_norm": 17.727657318115234,
+      "learning_rate": 2.473178994918126e-06,
+      "loss": 0.4731,
+      "step": 46560
+    },
+    {
+      "epoch": 8.77,
+      "grad_norm": 11.835549354553223,
+      "learning_rate": 2.469414643327687e-06,
+      "loss": 0.4562,
+      "step": 46570
+    },
+    {
+      "epoch": 8.77,
+      "grad_norm": 5.226273536682129,
+      "learning_rate": 2.4656502917372483e-06,
+      "loss": 0.4373,
+      "step": 46580
+    },
+    {
+      "epoch": 8.77,
+      "grad_norm": 16.807971954345703,
+      "learning_rate": 2.4618859401468098e-06,
+      "loss": 0.5285,
+      "step": 46590
+    },
+    {
+      "epoch": 8.77,
+      "grad_norm": 11.890029907226562,
+      "learning_rate": 2.4581215885563712e-06,
+      "loss": 0.3881,
+      "step": 46600
+    },
+    {
+      "epoch": 8.77,
+      "grad_norm": 7.499889373779297,
+      "learning_rate": 2.4543572369659327e-06,
+      "loss": 0.3529,
+      "step": 46610
+    },
+    {
+      "epoch": 8.77,
+      "grad_norm": 2.0104305744171143,
+      "learning_rate": 2.450592885375494e-06,
+      "loss": 0.5556,
+      "step": 46620
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 1.1118526458740234,
+      "learning_rate": 2.446828533785056e-06,
+      "loss": 0.3103,
+      "step": 46630
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 10.730484008789062,
+      "learning_rate": 2.443064182194617e-06,
+      "loss": 0.4211,
+      "step": 46640
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 0.4229322075843811,
+      "learning_rate": 2.4392998306041785e-06,
+      "loss": 0.7197,
+      "step": 46650
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 7.243539810180664,
+      "learning_rate": 2.43553547901374e-06,
+      "loss": 0.2981,
+      "step": 46660
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 0.10506740212440491,
+      "learning_rate": 2.4317711274233015e-06,
+      "loss": 0.4107,
+      "step": 46670
+    },
+    {
+      "epoch": 8.79,
+      "grad_norm": 0.9739378690719604,
+      "learning_rate": 2.428006775832863e-06,
+      "loss": 0.4228,
+      "step": 46680
+    },
+    {
+      "epoch": 8.79,
+      "grad_norm": 21.565500259399414,
+      "learning_rate": 2.4242424242424244e-06,
+      "loss": 0.554,
+      "step": 46690
+    },
+    {
+      "epoch": 8.79,
+      "grad_norm": 8.53973388671875,
+      "learning_rate": 2.420478072651986e-06,
+      "loss": 0.7312,
+      "step": 46700
+    },
+    {
+      "epoch": 8.79,
+      "grad_norm": 15.285102844238281,
+      "learning_rate": 2.4167137210615473e-06,
+      "loss": 0.4223,
+      "step": 46710
+    },
+    {
+      "epoch": 8.79,
+      "grad_norm": 27.030920028686523,
+      "learning_rate": 2.4129493694711088e-06,
+      "loss": 0.3982,
+      "step": 46720
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 1.0703567266464233,
+      "learning_rate": 2.4091850178806702e-06,
+      "loss": 0.5266,
+      "step": 46730
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 1.5598318576812744,
+      "learning_rate": 2.4054206662902317e-06,
+      "loss": 0.4147,
+      "step": 46740
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 5.893131732940674,
+      "learning_rate": 2.401656314699793e-06,
+      "loss": 0.9074,
+      "step": 46750
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 0.11686406284570694,
+      "learning_rate": 2.3978919631093546e-06,
+      "loss": 0.3176,
+      "step": 46760
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 17.671430587768555,
+      "learning_rate": 2.394127611518916e-06,
+      "loss": 0.6666,
+      "step": 46770
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 6.053145408630371,
+      "learning_rate": 2.3903632599284775e-06,
+      "loss": 0.453,
+      "step": 46780
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 17.766340255737305,
+      "learning_rate": 2.386598908338039e-06,
+      "loss": 0.292,
+      "step": 46790
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 7.656116485595703,
+      "learning_rate": 2.3828345567476005e-06,
+      "loss": 0.4534,
+      "step": 46800
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 9.28414249420166,
+      "learning_rate": 2.379070205157162e-06,
+      "loss": 0.2725,
+      "step": 46810
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 0.04184451699256897,
+      "learning_rate": 2.3753058535667234e-06,
+      "loss": 0.4548,
+      "step": 46820
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 51.41383743286133,
+      "learning_rate": 2.371541501976285e-06,
+      "loss": 0.6366,
+      "step": 46830
+    },
+    {
+      "epoch": 8.82,
+      "grad_norm": 0.9687474370002747,
+      "learning_rate": 2.3677771503858463e-06,
+      "loss": 0.1807,
+      "step": 46840
+    },
+    {
+      "epoch": 8.82,
+      "grad_norm": 10.670173645019531,
+      "learning_rate": 2.3640127987954078e-06,
+      "loss": 0.2247,
+      "step": 46850
+    },
+    {
+      "epoch": 8.82,
+      "grad_norm": 25.308748245239258,
+      "learning_rate": 2.3602484472049692e-06,
+      "loss": 0.4082,
+      "step": 46860
+    },
+    {
+      "epoch": 8.82,
+      "grad_norm": 6.221033573150635,
+      "learning_rate": 2.3564840956145303e-06,
+      "loss": 0.4089,
+      "step": 46870
+    },
+    {
+      "epoch": 8.82,
+      "grad_norm": 6.021878719329834,
+      "learning_rate": 2.352719744024092e-06,
+      "loss": 0.5282,
+      "step": 46880
+    },
+    {
+      "epoch": 8.83,
+      "grad_norm": 4.055378437042236,
+      "learning_rate": 2.3489553924336536e-06,
+      "loss": 0.3669,
+      "step": 46890
+    },
+    {
+      "epoch": 8.83,
+      "grad_norm": 2.8399837017059326,
+      "learning_rate": 2.345191040843215e-06,
+      "loss": 0.4622,
+      "step": 46900
+    },
+    {
+      "epoch": 8.83,
+      "grad_norm": 13.155150413513184,
+      "learning_rate": 2.3414266892527765e-06,
+      "loss": 0.3145,
+      "step": 46910
+    },
+    {
+      "epoch": 8.83,
+      "grad_norm": 6.751850605010986,
+      "learning_rate": 2.337662337662338e-06,
+      "loss": 0.6429,
+      "step": 46920
+    },
+    {
+      "epoch": 8.83,
+      "grad_norm": 25.934751510620117,
+      "learning_rate": 2.3338979860718994e-06,
+      "loss": 0.3962,
+      "step": 46930
+    },
+    {
+      "epoch": 8.83,
+      "grad_norm": 0.046511393040418625,
+      "learning_rate": 2.3301336344814605e-06,
+      "loss": 0.6207,
+      "step": 46940
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 26.9798641204834,
+      "learning_rate": 2.3263692828910224e-06,
+      "loss": 0.4355,
+      "step": 46950
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 9.000409126281738,
+      "learning_rate": 2.3226049313005834e-06,
+      "loss": 0.5159,
+      "step": 46960
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 26.408952713012695,
+      "learning_rate": 2.3188405797101453e-06,
+      "loss": 0.3787,
+      "step": 46970
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 53.70569610595703,
+      "learning_rate": 2.3150762281197063e-06,
+      "loss": 0.5017,
+      "step": 46980
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 24.432947158813477,
+      "learning_rate": 2.311311876529268e-06,
+      "loss": 0.2867,
+      "step": 46990
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 0.2684187591075897,
+      "learning_rate": 2.3075475249388297e-06,
+      "loss": 0.3527,
+      "step": 47000
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 7.198283672332764,
+      "learning_rate": 2.3037831733483907e-06,
+      "loss": 0.4833,
+      "step": 47010
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 15.57585334777832,
+      "learning_rate": 2.3000188217579526e-06,
+      "loss": 0.3947,
+      "step": 47020
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 34.16131591796875,
+      "learning_rate": 2.2962544701675136e-06,
+      "loss": 0.4851,
+      "step": 47030
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 1.565393090248108,
+      "learning_rate": 2.2924901185770755e-06,
+      "loss": 0.5857,
+      "step": 47040
+    },
+    {
+      "epoch": 8.86,
+      "grad_norm": 43.33597946166992,
+      "learning_rate": 2.2887257669866366e-06,
+      "loss": 0.6679,
+      "step": 47050
+    },
+    {
+      "epoch": 8.86,
+      "grad_norm": 11.061119079589844,
+      "learning_rate": 2.2849614153961984e-06,
+      "loss": 0.5468,
+      "step": 47060
+    },
+    {
+      "epoch": 8.86,
+      "grad_norm": 0.20168103277683258,
+      "learning_rate": 2.2811970638057595e-06,
+      "loss": 0.5677,
+      "step": 47070
+    },
+    {
+      "epoch": 8.86,
+      "grad_norm": 5.765427112579346,
+      "learning_rate": 2.277432712215321e-06,
+      "loss": 0.6151,
+      "step": 47080
+    },
+    {
+      "epoch": 8.86,
+      "grad_norm": 14.285442352294922,
+      "learning_rate": 2.2736683606248824e-06,
+      "loss": 0.2915,
+      "step": 47090
+    },
+    {
+      "epoch": 8.87,
+      "grad_norm": 11.015250205993652,
+      "learning_rate": 2.269904009034444e-06,
+      "loss": 0.4167,
+      "step": 47100
+    },
+    {
+      "epoch": 8.87,
+      "grad_norm": 42.790672302246094,
+      "learning_rate": 2.2661396574440053e-06,
+      "loss": 0.632,
+      "step": 47110
+    },
+    {
+      "epoch": 8.87,
+      "grad_norm": 23.58126449584961,
+      "learning_rate": 2.2623753058535668e-06,
+      "loss": 0.5522,
+      "step": 47120
+    },
+    {
+      "epoch": 8.87,
+      "grad_norm": 0.10688672214746475,
+      "learning_rate": 2.2586109542631287e-06,
+      "loss": 0.5342,
+      "step": 47130
+    },
+    {
+      "epoch": 8.87,
+      "grad_norm": 22.022897720336914,
+      "learning_rate": 2.2548466026726897e-06,
+      "loss": 0.5221,
+      "step": 47140
+    },
+    {
+      "epoch": 8.87,
+      "grad_norm": 3.2762904167175293,
+      "learning_rate": 2.251082251082251e-06,
+      "loss": 0.508,
+      "step": 47150
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 8.347319602966309,
+      "learning_rate": 2.2473178994918126e-06,
+      "loss": 0.3269,
+      "step": 47160
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 0.2670442461967468,
+      "learning_rate": 2.243553547901374e-06,
+      "loss": 0.4172,
+      "step": 47170
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 22.307165145874023,
+      "learning_rate": 2.2397891963109355e-06,
+      "loss": 0.6191,
+      "step": 47180
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 12.482026100158691,
+      "learning_rate": 2.236024844720497e-06,
+      "loss": 0.4978,
+      "step": 47190
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 12.081169128417969,
+      "learning_rate": 2.2322604931300585e-06,
+      "loss": 0.4355,
+      "step": 47200
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 16.33040428161621,
+      "learning_rate": 2.22849614153962e-06,
+      "loss": 0.5954,
+      "step": 47210
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 15.850767135620117,
+      "learning_rate": 2.2247317899491814e-06,
+      "loss": 0.4529,
+      "step": 47220
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 39.77294158935547,
+      "learning_rate": 2.220967438358743e-06,
+      "loss": 0.2148,
+      "step": 47230
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 18.885601043701172,
+      "learning_rate": 2.2172030867683043e-06,
+      "loss": 0.4973,
+      "step": 47240
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 16.801286697387695,
+      "learning_rate": 2.2134387351778658e-06,
+      "loss": 0.4566,
+      "step": 47250
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 18.179574966430664,
+      "learning_rate": 2.2096743835874272e-06,
+      "loss": 0.7612,
+      "step": 47260
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 3.8040480613708496,
+      "learning_rate": 2.2059100319969887e-06,
+      "loss": 0.4395,
+      "step": 47270
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 26.792110443115234,
+      "learning_rate": 2.20214568040655e-06,
+      "loss": 0.6617,
+      "step": 47280
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 10.858789443969727,
+      "learning_rate": 2.1983813288161116e-06,
+      "loss": 0.4822,
+      "step": 47290
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 0.04407741501927376,
+      "learning_rate": 2.194616977225673e-06,
+      "loss": 0.6258,
+      "step": 47300
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 23.669815063476562,
+      "learning_rate": 2.1908526256352345e-06,
+      "loss": 0.4182,
+      "step": 47310
+    },
+    {
+      "epoch": 8.91,
+      "grad_norm": 15.9193115234375,
+      "learning_rate": 2.187088274044796e-06,
+      "loss": 0.3994,
+      "step": 47320
+    },
+    {
+      "epoch": 8.91,
+      "grad_norm": 35.726600646972656,
+      "learning_rate": 2.1833239224543575e-06,
+      "loss": 0.2649,
+      "step": 47330
+    },
+    {
+      "epoch": 8.91,
+      "grad_norm": 23.043758392333984,
+      "learning_rate": 2.179559570863919e-06,
+      "loss": 0.3575,
+      "step": 47340
+    },
+    {
+      "epoch": 8.91,
+      "grad_norm": 1.2536325454711914,
+      "learning_rate": 2.1757952192734804e-06,
+      "loss": 0.3955,
+      "step": 47350
+    },
+    {
+      "epoch": 8.91,
+      "grad_norm": 19.012310028076172,
+      "learning_rate": 2.172030867683042e-06,
+      "loss": 0.5392,
+      "step": 47360
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 7.179901599884033,
+      "learning_rate": 2.1682665160926033e-06,
+      "loss": 0.5012,
+      "step": 47370
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 15.418367385864258,
+      "learning_rate": 2.1645021645021648e-06,
+      "loss": 0.5293,
+      "step": 47380
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 25.173019409179688,
+      "learning_rate": 2.1607378129117262e-06,
+      "loss": 0.4274,
+      "step": 47390
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 1.2999225854873657,
+      "learning_rate": 2.1569734613212877e-06,
+      "loss": 0.3006,
+      "step": 47400
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 0.06721743196249008,
+      "learning_rate": 2.153209109730849e-06,
+      "loss": 0.5072,
+      "step": 47410
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 43.51065444946289,
+      "learning_rate": 2.1494447581404106e-06,
+      "loss": 0.5687,
+      "step": 47420
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 1.2883092164993286,
+      "learning_rate": 2.145680406549972e-06,
+      "loss": 0.2178,
+      "step": 47430
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 3.8370001316070557,
+      "learning_rate": 2.1419160549595335e-06,
+      "loss": 0.4656,
+      "step": 47440
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 1.2734487056732178,
+      "learning_rate": 2.138151703369095e-06,
+      "loss": 0.3782,
+      "step": 47450
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 7.262458324432373,
+      "learning_rate": 2.134387351778656e-06,
+      "loss": 0.4573,
+      "step": 47460
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 33.35980987548828,
+      "learning_rate": 2.130623000188218e-06,
+      "loss": 0.3935,
+      "step": 47470
+    },
+    {
+      "epoch": 8.94,
+      "grad_norm": 5.109274387359619,
+      "learning_rate": 2.126858648597779e-06,
+      "loss": 0.4141,
+      "step": 47480
+    },
+    {
+      "epoch": 8.94,
+      "grad_norm": 5.794544219970703,
+      "learning_rate": 2.123094297007341e-06,
+      "loss": 0.4369,
+      "step": 47490
+    },
+    {
+      "epoch": 8.94,
+      "grad_norm": 1.0066646337509155,
+      "learning_rate": 2.1193299454169023e-06,
+      "loss": 0.6685,
+      "step": 47500
+    },
+    {
+      "epoch": 8.94,
+      "grad_norm": 11.47095012664795,
+      "learning_rate": 2.1155655938264637e-06,
+      "loss": 0.3829,
+      "step": 47510
+    },
+    {
+      "epoch": 8.94,
+      "grad_norm": 14.297951698303223,
+      "learning_rate": 2.111801242236025e-06,
+      "loss": 0.4893,
+      "step": 47520
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 11.840413093566895,
+      "learning_rate": 2.1080368906455862e-06,
+      "loss": 0.7612,
+      "step": 47530
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 38.1298713684082,
+      "learning_rate": 2.104272539055148e-06,
+      "loss": 0.6616,
+      "step": 47540
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 25.78653907775879,
+      "learning_rate": 2.100508187464709e-06,
+      "loss": 0.2917,
+      "step": 47550
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 19.186622619628906,
+      "learning_rate": 2.096743835874271e-06,
+      "loss": 0.5563,
+      "step": 47560
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 6.531424522399902,
+      "learning_rate": 2.092979484283832e-06,
+      "loss": 0.3453,
+      "step": 47570
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 0.02262808382511139,
+      "learning_rate": 2.089215132693394e-06,
+      "loss": 0.4988,
+      "step": 47580
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 18.038076400756836,
+      "learning_rate": 2.085450781102955e-06,
+      "loss": 0.5999,
+      "step": 47590
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 12.460607528686523,
+      "learning_rate": 2.0816864295125165e-06,
+      "loss": 0.7356,
+      "step": 47600
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 45.871124267578125,
+      "learning_rate": 2.0779220779220784e-06,
+      "loss": 0.2332,
+      "step": 47610
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 6.987379550933838,
+      "learning_rate": 2.0741577263316394e-06,
+      "loss": 0.4788,
+      "step": 47620
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 27.808523178100586,
+      "learning_rate": 2.0703933747412013e-06,
+      "loss": 0.6439,
+      "step": 47630
+    },
+    {
+      "epoch": 8.97,
+      "grad_norm": 15.022832870483398,
+      "learning_rate": 2.0666290231507623e-06,
+      "loss": 0.4583,
+      "step": 47640
+    },
+    {
+      "epoch": 8.97,
+      "grad_norm": 1.9445959329605103,
+      "learning_rate": 2.062864671560324e-06,
+      "loss": 0.6531,
+      "step": 47650
+    },
+    {
+      "epoch": 8.97,
+      "grad_norm": 19.67501449584961,
+      "learning_rate": 2.0591003199698852e-06,
+      "loss": 0.5846,
+      "step": 47660
+    },
+    {
+      "epoch": 8.97,
+      "grad_norm": 4.834009647369385,
+      "learning_rate": 2.0553359683794467e-06,
+      "loss": 0.3116,
+      "step": 47670
+    },
+    {
+      "epoch": 8.97,
+      "grad_norm": 41.817848205566406,
+      "learning_rate": 2.051571616789008e-06,
+      "loss": 0.5561,
+      "step": 47680
+    },
+    {
+      "epoch": 8.98,
+      "grad_norm": 10.177002906799316,
+      "learning_rate": 2.0478072651985696e-06,
+      "loss": 0.4742,
+      "step": 47690
+    },
+    {
+      "epoch": 8.98,
+      "grad_norm": 9.804389953613281,
+      "learning_rate": 2.044042913608131e-06,
+      "loss": 0.5149,
+      "step": 47700
+    },
+    {
+      "epoch": 8.98,
+      "grad_norm": 0.5583279132843018,
+      "learning_rate": 2.0402785620176925e-06,
+      "loss": 0.1706,
+      "step": 47710
+    },
+    {
+      "epoch": 8.98,
+      "grad_norm": 17.16972541809082,
+      "learning_rate": 2.036514210427254e-06,
+      "loss": 0.5996,
+      "step": 47720
+    },
+    {
+      "epoch": 8.98,
+      "grad_norm": 12.238551139831543,
+      "learning_rate": 2.0327498588368155e-06,
+      "loss": 0.3095,
+      "step": 47730
+    },
+    {
+      "epoch": 8.99,
+      "grad_norm": 23.57866859436035,
+      "learning_rate": 2.028985507246377e-06,
+      "loss": 0.332,
+      "step": 47740
+    },
+    {
+      "epoch": 8.99,
+      "grad_norm": 40.92672348022461,
+      "learning_rate": 2.0252211556559384e-06,
+      "loss": 0.6045,
+      "step": 47750
+    },
+    {
+      "epoch": 8.99,
+      "grad_norm": 0.26052126288414,
+      "learning_rate": 2.0214568040655e-06,
+      "loss": 0.5199,
+      "step": 47760
+    },
+    {
+      "epoch": 8.99,
+      "grad_norm": 14.104440689086914,
+      "learning_rate": 2.0176924524750613e-06,
+      "loss": 0.3415,
+      "step": 47770
+    },
+    {
+      "epoch": 8.99,
+      "grad_norm": 5.497945785522461,
+      "learning_rate": 2.0139281008846228e-06,
+      "loss": 0.3376,
+      "step": 47780
+    },
+    {
+      "epoch": 8.99,
+      "grad_norm": 16.819019317626953,
+      "learning_rate": 2.0101637492941842e-06,
+      "loss": 0.2294,
+      "step": 47790
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 26.409992218017578,
+      "learning_rate": 2.0063993977037457e-06,
+      "loss": 0.4863,
+      "step": 47800
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 37.679100036621094,
+      "learning_rate": 2.002635046113307e-06,
+      "loss": 0.3794,
+      "step": 47810
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.926,
+      "eval_loss": 0.29454880952835083,
+      "eval_runtime": 51.0107,
+      "eval_samples_per_second": 147.028,
+      "eval_steps_per_second": 18.388,
+      "step": 47817
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.07574218511581421,
+      "learning_rate": 1.9988706945228686e-06,
+      "loss": 0.384,
+      "step": 47820
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.08334751427173615,
+      "learning_rate": 1.99510634293243e-06,
+      "loss": 0.3326,
+      "step": 47830
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 22.99428367614746,
+      "learning_rate": 1.9913419913419915e-06,
+      "loss": 0.3852,
+      "step": 47840
+    },
+    {
+      "epoch": 9.01,
+      "grad_norm": 2.2758443355560303,
+      "learning_rate": 1.987577639751553e-06,
+      "loss": 0.3487,
+      "step": 47850
+    },
+    {
+      "epoch": 9.01,
+      "grad_norm": 16.36058807373047,
+      "learning_rate": 1.9838132881611144e-06,
+      "loss": 0.732,
+      "step": 47860
+    },
+    {
+      "epoch": 9.01,
+      "grad_norm": 18.320423126220703,
+      "learning_rate": 1.980048936570676e-06,
+      "loss": 0.5723,
+      "step": 47870
+    },
+    {
+      "epoch": 9.01,
+      "grad_norm": 0.30943912267684937,
+      "learning_rate": 1.9762845849802374e-06,
+      "loss": 0.5259,
+      "step": 47880
+    },
+    {
+      "epoch": 9.01,
+      "grad_norm": 10.592155456542969,
+      "learning_rate": 1.972520233389799e-06,
+      "loss": 0.357,
+      "step": 47890
+    },
+    {
+      "epoch": 9.02,
+      "grad_norm": 20.643972396850586,
+      "learning_rate": 1.9687558817993603e-06,
+      "loss": 0.4431,
+      "step": 47900
+    },
+    {
+      "epoch": 9.02,
+      "grad_norm": 3.8286397457122803,
+      "learning_rate": 1.9649915302089217e-06,
+      "loss": 0.4602,
+      "step": 47910
+    },
+    {
+      "epoch": 9.02,
+      "grad_norm": 13.922072410583496,
+      "learning_rate": 1.961227178618483e-06,
+      "loss": 0.4626,
+      "step": 47920
+    },
+    {
+      "epoch": 9.02,
+      "grad_norm": 14.62498950958252,
+      "learning_rate": 1.9574628270280447e-06,
+      "loss": 0.1669,
+      "step": 47930
+    },
+    {
+      "epoch": 9.02,
+      "grad_norm": 13.237974166870117,
+      "learning_rate": 1.953698475437606e-06,
+      "loss": 0.3758,
+      "step": 47940
+    },
+    {
+      "epoch": 9.03,
+      "grad_norm": 47.09726333618164,
+      "learning_rate": 1.9499341238471676e-06,
+      "loss": 0.6936,
+      "step": 47950
+    },
+    {
+      "epoch": 9.03,
+      "grad_norm": 7.44305419921875,
+      "learning_rate": 1.9461697722567286e-06,
+      "loss": 0.5368,
+      "step": 47960
+    },
+    {
+      "epoch": 9.03,
+      "grad_norm": 10.614571571350098,
+      "learning_rate": 1.9424054206662905e-06,
+      "loss": 0.3991,
+      "step": 47970
+    },
+    {
+      "epoch": 9.03,
+      "grad_norm": 14.058085441589355,
+      "learning_rate": 1.938641069075852e-06,
+      "loss": 0.2912,
+      "step": 47980
+    },
+    {
+      "epoch": 9.03,
+      "grad_norm": 10.969157218933105,
+      "learning_rate": 1.9348767174854134e-06,
+      "loss": 0.4087,
+      "step": 47990
+    },
+    {
+      "epoch": 9.03,
+      "grad_norm": 10.787178039550781,
+      "learning_rate": 1.931112365894975e-06,
+      "loss": 0.4769,
+      "step": 48000
+    },
+    {
+      "epoch": 9.04,
+      "grad_norm": 15.043909072875977,
+      "learning_rate": 1.9273480143045364e-06,
+      "loss": 0.1376,
+      "step": 48010
+    },
+    {
+      "epoch": 9.04,
+      "grad_norm": 1.5361169576644897,
+      "learning_rate": 1.923583662714098e-06,
+      "loss": 0.4009,
+      "step": 48020
+    },
+    {
+      "epoch": 9.04,
+      "grad_norm": 9.03320598602295,
+      "learning_rate": 1.919819311123659e-06,
+      "loss": 0.6094,
+      "step": 48030
+    },
+    {
+      "epoch": 9.04,
+      "grad_norm": 0.36295774579048157,
+      "learning_rate": 1.9160549595332207e-06,
+      "loss": 0.282,
+      "step": 48040
+    },
+    {
+      "epoch": 9.04,
+      "grad_norm": 5.759310722351074,
+      "learning_rate": 1.9122906079427818e-06,
+      "loss": 0.2708,
+      "step": 48050
+    },
+    {
+      "epoch": 9.05,
+      "grad_norm": 0.02767598256468773,
+      "learning_rate": 1.9085262563523437e-06,
+      "loss": 0.5305,
+      "step": 48060
+    },
+    {
+      "epoch": 9.05,
+      "grad_norm": 1.4058023691177368,
+      "learning_rate": 1.904761904761905e-06,
+      "loss": 0.4482,
+      "step": 48070
+    },
+    {
+      "epoch": 9.05,
+      "grad_norm": 0.3036022186279297,
+      "learning_rate": 1.9009975531714664e-06,
+      "loss": 0.5453,
+      "step": 48080
+    },
+    {
+      "epoch": 9.05,
+      "grad_norm": 4.839325904846191,
+      "learning_rate": 1.8972332015810276e-06,
+      "loss": 0.3741,
+      "step": 48090
+    },
+    {
+      "epoch": 9.05,
+      "grad_norm": 14.845388412475586,
+      "learning_rate": 1.8934688499905893e-06,
+      "loss": 0.5632,
+      "step": 48100
+    },
+    {
+      "epoch": 9.06,
+      "grad_norm": 16.649734497070312,
+      "learning_rate": 1.8897044984001508e-06,
+      "loss": 0.1853,
+      "step": 48110
+    },
+    {
+      "epoch": 9.06,
+      "grad_norm": 25.03996467590332,
+      "learning_rate": 1.8859401468097122e-06,
+      "loss": 0.3492,
+      "step": 48120
+    },
+    {
+      "epoch": 9.06,
+      "grad_norm": 30.69152069091797,
+      "learning_rate": 1.8821757952192737e-06,
+      "loss": 0.4833,
+      "step": 48130
+    },
+    {
+      "epoch": 9.06,
+      "grad_norm": 13.534026145935059,
+      "learning_rate": 1.8784114436288351e-06,
+      "loss": 0.3608,
+      "step": 48140
+    },
+    {
+      "epoch": 9.06,
+      "grad_norm": 23.936994552612305,
+      "learning_rate": 1.8746470920383966e-06,
+      "loss": 0.4873,
+      "step": 48150
+    },
+    {
+      "epoch": 9.06,
+      "grad_norm": 41.1656608581543,
+      "learning_rate": 1.8708827404479578e-06,
+      "loss": 0.683,
+      "step": 48160
+    },
+    {
+      "epoch": 9.07,
+      "grad_norm": 12.103029251098633,
+      "learning_rate": 1.8671183888575195e-06,
+      "loss": 0.3607,
+      "step": 48170
+    },
+    {
+      "epoch": 9.07,
+      "grad_norm": 0.34992146492004395,
+      "learning_rate": 1.8633540372670808e-06,
+      "loss": 0.3565,
+      "step": 48180
+    },
+    {
+      "epoch": 9.07,
+      "grad_norm": 5.494448661804199,
+      "learning_rate": 1.8595896856766424e-06,
+      "loss": 0.3514,
+      "step": 48190
+    },
+    {
+      "epoch": 9.07,
+      "grad_norm": 19.909494400024414,
+      "learning_rate": 1.8558253340862037e-06,
+      "loss": 0.3663,
+      "step": 48200
+    },
+    {
+      "epoch": 9.07,
+      "grad_norm": 23.300514221191406,
+      "learning_rate": 1.8520609824957654e-06,
+      "loss": 0.6781,
+      "step": 48210
+    },
+    {
+      "epoch": 9.08,
+      "grad_norm": 10.3643798828125,
+      "learning_rate": 1.8482966309053266e-06,
+      "loss": 0.4268,
+      "step": 48220
+    },
+    {
+      "epoch": 9.08,
+      "grad_norm": 23.494760513305664,
+      "learning_rate": 1.844532279314888e-06,
+      "loss": 0.4987,
+      "step": 48230
+    },
+    {
+      "epoch": 9.08,
+      "grad_norm": 3.709768772125244,
+      "learning_rate": 1.8407679277244497e-06,
+      "loss": 0.3542,
+      "step": 48240
+    },
+    {
+      "epoch": 9.08,
+      "grad_norm": 0.16881948709487915,
+      "learning_rate": 1.837003576134011e-06,
+      "loss": 0.5134,
+      "step": 48250
+    },
+    {
+      "epoch": 9.08,
+      "grad_norm": 9.188733100891113,
+      "learning_rate": 1.8332392245435727e-06,
+      "loss": 0.6763,
+      "step": 48260
+    },
+    {
+      "epoch": 9.09,
+      "grad_norm": 12.953929901123047,
+      "learning_rate": 1.829474872953134e-06,
+      "loss": 0.3074,
+      "step": 48270
+    },
+    {
+      "epoch": 9.09,
+      "grad_norm": 6.657861232757568,
+      "learning_rate": 1.8257105213626956e-06,
+      "loss": 0.3391,
+      "step": 48280
+    },
+    {
+      "epoch": 9.09,
+      "grad_norm": 22.9435977935791,
+      "learning_rate": 1.8219461697722568e-06,
+      "loss": 0.7275,
+      "step": 48290
+    },
+    {
+      "epoch": 9.09,
+      "grad_norm": 25.13018035888672,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 0.1599,
+      "step": 48300
+    },
+    {
+      "epoch": 9.09,
+      "grad_norm": 16.501449584960938,
+      "learning_rate": 1.8144174665913798e-06,
+      "loss": 0.577,
+      "step": 48310
+    },
+    {
+      "epoch": 9.09,
+      "grad_norm": 39.59345626831055,
+      "learning_rate": 1.8106531150009412e-06,
+      "loss": 0.7677,
+      "step": 48320
+    },
+    {
+      "epoch": 9.1,
+      "grad_norm": 14.423561096191406,
+      "learning_rate": 1.8068887634105025e-06,
+      "loss": 0.5273,
+      "step": 48330
+    },
+    {
+      "epoch": 9.1,
+      "grad_norm": 24.364534378051758,
+      "learning_rate": 1.8031244118200641e-06,
+      "loss": 0.2518,
+      "step": 48340
+    },
+    {
+      "epoch": 9.1,
+      "grad_norm": 12.524968147277832,
+      "learning_rate": 1.7993600602296258e-06,
+      "loss": 0.6041,
+      "step": 48350
+    },
+    {
+      "epoch": 9.1,
+      "grad_norm": 23.22210693359375,
+      "learning_rate": 1.795595708639187e-06,
+      "loss": 0.6273,
+      "step": 48360
+    },
+    {
+      "epoch": 9.1,
+      "grad_norm": 9.35206413269043,
+      "learning_rate": 1.7918313570487485e-06,
+      "loss": 0.363,
+      "step": 48370
+    },
+    {
+      "epoch": 9.11,
+      "grad_norm": 11.951454162597656,
+      "learning_rate": 1.78806700545831e-06,
+      "loss": 0.2939,
+      "step": 48380
+    },
+    {
+      "epoch": 9.11,
+      "grad_norm": 8.72339153289795,
+      "learning_rate": 1.7843026538678714e-06,
+      "loss": 0.61,
+      "step": 48390
+    },
+    {
+      "epoch": 9.11,
+      "grad_norm": 36.56248092651367,
+      "learning_rate": 1.780538302277433e-06,
+      "loss": 0.5281,
+      "step": 48400
+    },
+    {
+      "epoch": 9.11,
+      "grad_norm": 13.832367897033691,
+      "learning_rate": 1.7767739506869944e-06,
+      "loss": 0.2781,
+      "step": 48410
+    },
+    {
+      "epoch": 9.11,
+      "grad_norm": 24.075450897216797,
+      "learning_rate": 1.7730095990965556e-06,
+      "loss": 0.6961,
+      "step": 48420
+    },
+    {
+      "epoch": 9.12,
+      "grad_norm": 2.9399142265319824,
+      "learning_rate": 1.7692452475061173e-06,
+      "loss": 0.2952,
+      "step": 48430
+    },
+    {
+      "epoch": 9.12,
+      "grad_norm": 16.432899475097656,
+      "learning_rate": 1.7654808959156785e-06,
+      "loss": 0.4107,
+      "step": 48440
+    },
+    {
+      "epoch": 9.12,
+      "grad_norm": 21.72768211364746,
+      "learning_rate": 1.7617165443252402e-06,
+      "loss": 0.7116,
+      "step": 48450
+    },
+    {
+      "epoch": 9.12,
+      "grad_norm": 20.015758514404297,
+      "learning_rate": 1.7579521927348015e-06,
+      "loss": 0.8156,
+      "step": 48460
+    },
+    {
+      "epoch": 9.12,
+      "grad_norm": 7.275087833404541,
+      "learning_rate": 1.7541878411443631e-06,
+      "loss": 0.7025,
+      "step": 48470
+    },
+    {
+      "epoch": 9.12,
+      "grad_norm": 59.83738708496094,
+      "learning_rate": 1.7504234895539246e-06,
+      "loss": 0.363,
+      "step": 48480
+    },
+    {
+      "epoch": 9.13,
+      "grad_norm": 19.102800369262695,
+      "learning_rate": 1.7466591379634858e-06,
+      "loss": 0.6994,
+      "step": 48490
+    },
+    {
+      "epoch": 9.13,
+      "grad_norm": 37.69584274291992,
+      "learning_rate": 1.7428947863730475e-06,
+      "loss": 0.6699,
+      "step": 48500
+    },
+    {
+      "epoch": 9.13,
+      "grad_norm": 7.24312686920166,
+      "learning_rate": 1.7391304347826088e-06,
+      "loss": 0.9191,
+      "step": 48510
+    },
+    {
+      "epoch": 9.13,
+      "grad_norm": 28.769311904907227,
+      "learning_rate": 1.7353660831921704e-06,
+      "loss": 0.3994,
+      "step": 48520
+    },
+    {
+      "epoch": 9.13,
+      "grad_norm": 0.12715600430965424,
+      "learning_rate": 1.7316017316017317e-06,
+      "loss": 0.3204,
+      "step": 48530
+    },
+    {
+      "epoch": 9.14,
+      "grad_norm": 29.197851181030273,
+      "learning_rate": 1.7278373800112933e-06,
+      "loss": 0.6342,
+      "step": 48540
+    },
+    {
+      "epoch": 9.14,
+      "grad_norm": 4.771313667297363,
+      "learning_rate": 1.7240730284208546e-06,
+      "loss": 0.4141,
+      "step": 48550
+    },
+    {
+      "epoch": 9.14,
+      "grad_norm": 10.536064147949219,
+      "learning_rate": 1.720308676830416e-06,
+      "loss": 0.407,
+      "step": 48560
+    },
+    {
+      "epoch": 9.14,
+      "grad_norm": 19.451534271240234,
+      "learning_rate": 1.7165443252399775e-06,
+      "loss": 0.4267,
+      "step": 48570
+    },
+    {
+      "epoch": 9.14,
+      "grad_norm": 15.74153995513916,
+      "learning_rate": 1.712779973649539e-06,
+      "loss": 0.3457,
+      "step": 48580
+    },
+    {
+      "epoch": 9.15,
+      "grad_norm": 8.152856826782227,
+      "learning_rate": 1.7090156220591002e-06,
+      "loss": 0.2276,
+      "step": 48590
+    },
+    {
+      "epoch": 9.15,
+      "grad_norm": 6.447235584259033,
+      "learning_rate": 1.705251270468662e-06,
+      "loss": 0.2864,
+      "step": 48600
+    },
+    {
+      "epoch": 9.15,
+      "grad_norm": 15.62842845916748,
+      "learning_rate": 1.7014869188782236e-06,
+      "loss": 0.5409,
+      "step": 48610
+    },
+    {
+      "epoch": 9.15,
+      "grad_norm": 7.044529438018799,
+      "learning_rate": 1.6977225672877848e-06,
+      "loss": 0.6466,
+      "step": 48620
+    },
+    {
+      "epoch": 9.15,
+      "grad_norm": 17.935876846313477,
+      "learning_rate": 1.6939582156973463e-06,
+      "loss": 0.4593,
+      "step": 48630
+    },
+    {
+      "epoch": 9.15,
+      "grad_norm": 18.33387565612793,
+      "learning_rate": 1.6901938641069077e-06,
+      "loss": 0.5601,
+      "step": 48640
+    },
+    {
+      "epoch": 9.16,
+      "grad_norm": 11.096664428710938,
+      "learning_rate": 1.6864295125164692e-06,
+      "loss": 0.7234,
+      "step": 48650
+    },
+    {
+      "epoch": 9.16,
+      "grad_norm": 31.238327026367188,
+      "learning_rate": 1.6826651609260305e-06,
+      "loss": 0.496,
+      "step": 48660
+    },
+    {
+      "epoch": 9.16,
+      "grad_norm": 16.450227737426758,
+      "learning_rate": 1.6789008093355921e-06,
+      "loss": 0.4201,
+      "step": 48670
+    },
+    {
+      "epoch": 9.16,
+      "grad_norm": 7.500368118286133,
+      "learning_rate": 1.6751364577451534e-06,
+      "loss": 0.3319,
+      "step": 48680
+    },
+    {
+      "epoch": 9.16,
+      "grad_norm": 3.6829004287719727,
+      "learning_rate": 1.671372106154715e-06,
+      "loss": 0.5277,
+      "step": 48690
+    },
+    {
+      "epoch": 9.17,
+      "grad_norm": 5.226105213165283,
+      "learning_rate": 1.6676077545642763e-06,
+      "loss": 0.4659,
+      "step": 48700
+    },
+    {
+      "epoch": 9.17,
+      "grad_norm": 17.099241256713867,
+      "learning_rate": 1.663843402973838e-06,
+      "loss": 0.1683,
+      "step": 48710
+    },
+    {
+      "epoch": 9.17,
+      "grad_norm": 19.311145782470703,
+      "learning_rate": 1.6600790513833994e-06,
+      "loss": 0.3761,
+      "step": 48720
+    },
+    {
+      "epoch": 9.17,
+      "grad_norm": 11.647774696350098,
+      "learning_rate": 1.6563146997929607e-06,
+      "loss": 0.4523,
+      "step": 48730
+    },
+    {
+      "epoch": 9.17,
+      "grad_norm": 18.899429321289062,
+      "learning_rate": 1.6525503482025224e-06,
+      "loss": 0.4971,
+      "step": 48740
+    },
+    {
+      "epoch": 9.18,
+      "grad_norm": 0.5511684417724609,
+      "learning_rate": 1.6487859966120836e-06,
+      "loss": 0.2433,
+      "step": 48750
+    },
+    {
+      "epoch": 9.18,
+      "grad_norm": 7.991838455200195,
+      "learning_rate": 1.6450216450216453e-06,
+      "loss": 0.4022,
+      "step": 48760
+    },
+    {
+      "epoch": 9.18,
+      "grad_norm": 25.697017669677734,
+      "learning_rate": 1.6412572934312065e-06,
+      "loss": 0.4417,
+      "step": 48770
+    },
+    {
+      "epoch": 9.18,
+      "grad_norm": 12.145305633544922,
+      "learning_rate": 1.6374929418407682e-06,
+      "loss": 0.4612,
+      "step": 48780
+    },
+    {
+      "epoch": 9.18,
+      "grad_norm": 5.5945892333984375,
+      "learning_rate": 1.6337285902503294e-06,
+      "loss": 0.3146,
+      "step": 48790
+    },
+    {
+      "epoch": 9.19,
+      "grad_norm": 12.325715065002441,
+      "learning_rate": 1.629964238659891e-06,
+      "loss": 0.6039,
+      "step": 48800
+    },
+    {
+      "epoch": 9.19,
+      "grad_norm": 5.670255184173584,
+      "learning_rate": 1.6261998870694524e-06,
+      "loss": 0.5651,
+      "step": 48810
+    },
+    {
+      "epoch": 9.19,
+      "grad_norm": 80.41706848144531,
+      "learning_rate": 1.6224355354790138e-06,
+      "loss": 0.5711,
+      "step": 48820
+    },
+    {
+      "epoch": 9.19,
+      "grad_norm": 8.065768241882324,
+      "learning_rate": 1.6186711838885753e-06,
+      "loss": 0.6124,
+      "step": 48830
+    },
+    {
+      "epoch": 9.19,
+      "grad_norm": 0.024990417063236237,
+      "learning_rate": 1.6149068322981367e-06,
+      "loss": 0.269,
+      "step": 48840
+    },
+    {
+      "epoch": 9.19,
+      "grad_norm": 44.296573638916016,
+      "learning_rate": 1.6111424807076984e-06,
+      "loss": 0.4364,
+      "step": 48850
+    },
+    {
+      "epoch": 9.2,
+      "grad_norm": 6.221458435058594,
+      "learning_rate": 1.6073781291172597e-06,
+      "loss": 0.2955,
+      "step": 48860
+    },
+    {
+      "epoch": 9.2,
+      "grad_norm": 17.889265060424805,
+      "learning_rate": 1.6036137775268213e-06,
+      "loss": 0.3683,
+      "step": 48870
+    },
+    {
+      "epoch": 9.2,
+      "grad_norm": 9.34254264831543,
+      "learning_rate": 1.5998494259363826e-06,
+      "loss": 0.5364,
+      "step": 48880
+    },
+    {
+      "epoch": 9.2,
+      "grad_norm": 0.1648782640695572,
+      "learning_rate": 1.596085074345944e-06,
+      "loss": 0.3695,
+      "step": 48890
+    },
+    {
+      "epoch": 9.2,
+      "grad_norm": 1.0630319118499756,
+      "learning_rate": 1.5923207227555055e-06,
+      "loss": 0.3158,
+      "step": 48900
+    },
+    {
+      "epoch": 9.21,
+      "grad_norm": 28.75983428955078,
+      "learning_rate": 1.588556371165067e-06,
+      "loss": 0.3781,
+      "step": 48910
+    },
+    {
+      "epoch": 9.21,
+      "grad_norm": 4.615087509155273,
+      "learning_rate": 1.5847920195746282e-06,
+      "loss": 0.3918,
+      "step": 48920
+    },
+    {
+      "epoch": 9.21,
+      "grad_norm": 15.744338035583496,
+      "learning_rate": 1.5810276679841899e-06,
+      "loss": 0.153,
+      "step": 48930
+    },
+    {
+      "epoch": 9.21,
+      "grad_norm": 5.790966510772705,
+      "learning_rate": 1.5772633163937511e-06,
+      "loss": 0.3341,
+      "step": 48940
+    },
+    {
+      "epoch": 9.21,
+      "grad_norm": 0.05763714388012886,
+      "learning_rate": 1.5734989648033128e-06,
+      "loss": 0.4415,
+      "step": 48950
+    },
+    {
+      "epoch": 9.22,
+      "grad_norm": 15.090495109558105,
+      "learning_rate": 1.5697346132128743e-06,
+      "loss": 0.5551,
+      "step": 48960
+    },
+    {
+      "epoch": 9.22,
+      "grad_norm": 14.567575454711914,
+      "learning_rate": 1.5659702616224357e-06,
+      "loss": 0.7791,
+      "step": 48970
+    },
+    {
+      "epoch": 9.22,
+      "grad_norm": 5.712708473205566,
+      "learning_rate": 1.5622059100319972e-06,
+      "loss": 0.3734,
+      "step": 48980
+    },
+    {
+      "epoch": 9.22,
+      "grad_norm": 64.14569091796875,
+      "learning_rate": 1.5584415584415584e-06,
+      "loss": 0.4575,
+      "step": 48990
+    },
+    {
+      "epoch": 9.22,
+      "grad_norm": 20.8106632232666,
+      "learning_rate": 1.5546772068511201e-06,
+      "loss": 0.4518,
+      "step": 49000
+    },
+    {
+      "epoch": 9.22,
+      "grad_norm": 9.944857597351074,
+      "learning_rate": 1.5509128552606814e-06,
+      "loss": 0.4911,
+      "step": 49010
+    },
+    {
+      "epoch": 9.23,
+      "grad_norm": 8.021162986755371,
+      "learning_rate": 1.547148503670243e-06,
+      "loss": 0.6,
+      "step": 49020
+    },
+    {
+      "epoch": 9.23,
+      "grad_norm": 13.243287086486816,
+      "learning_rate": 1.5433841520798043e-06,
+      "loss": 0.507,
+      "step": 49030
+    },
+    {
+      "epoch": 9.23,
+      "grad_norm": 32.78733825683594,
+      "learning_rate": 1.539619800489366e-06,
+      "loss": 0.5617,
+      "step": 49040
+    },
+    {
+      "epoch": 9.23,
+      "grad_norm": 3.1205811500549316,
+      "learning_rate": 1.5358554488989272e-06,
+      "loss": 0.4243,
+      "step": 49050
+    },
+    {
+      "epoch": 9.23,
+      "grad_norm": 14.635000228881836,
+      "learning_rate": 1.5320910973084887e-06,
+      "loss": 0.5009,
+      "step": 49060
+    },
+    {
+      "epoch": 9.24,
+      "grad_norm": 18.14510154724121,
+      "learning_rate": 1.5283267457180501e-06,
+      "loss": 0.3399,
+      "step": 49070
+    },
+    {
+      "epoch": 9.24,
+      "grad_norm": 0.11989280581474304,
+      "learning_rate": 1.5245623941276116e-06,
+      "loss": 0.3495,
+      "step": 49080
+    },
+    {
+      "epoch": 9.24,
+      "grad_norm": 24.832670211791992,
+      "learning_rate": 1.5207980425371733e-06,
+      "loss": 0.3462,
+      "step": 49090
+    },
+    {
+      "epoch": 9.24,
+      "grad_norm": 12.079398155212402,
+      "learning_rate": 1.5170336909467345e-06,
+      "loss": 0.3772,
+      "step": 49100
+    },
+    {
+      "epoch": 9.24,
+      "grad_norm": 0.3867708146572113,
+      "learning_rate": 1.5132693393562962e-06,
+      "loss": 0.4475,
+      "step": 49110
+    },
+    {
+      "epoch": 9.25,
+      "grad_norm": 26.119632720947266,
+      "learning_rate": 1.5095049877658574e-06,
+      "loss": 0.6135,
+      "step": 49120
+    },
+    {
+      "epoch": 9.25,
+      "grad_norm": 1.460374116897583,
+      "learning_rate": 1.505740636175419e-06,
+      "loss": 0.3183,
+      "step": 49130
+    },
+    {
+      "epoch": 9.25,
+      "grad_norm": 5.709151268005371,
+      "learning_rate": 1.5019762845849804e-06,
+      "loss": 0.6727,
+      "step": 49140
+    },
+    {
+      "epoch": 9.25,
+      "grad_norm": 3.2548608779907227,
+      "learning_rate": 1.4982119329945418e-06,
+      "loss": 0.3298,
+      "step": 49150
+    },
+    {
+      "epoch": 9.25,
+      "grad_norm": 0.8151334524154663,
+      "learning_rate": 1.4944475814041033e-06,
+      "loss": 0.4324,
+      "step": 49160
+    },
+    {
+      "epoch": 9.25,
+      "grad_norm": 8.766486167907715,
+      "learning_rate": 1.4906832298136647e-06,
+      "loss": 0.5849,
+      "step": 49170
+    },
+    {
+      "epoch": 9.26,
+      "grad_norm": 12.971768379211426,
+      "learning_rate": 1.486918878223226e-06,
+      "loss": 0.5602,
+      "step": 49180
+    },
+    {
+      "epoch": 9.26,
+      "grad_norm": 0.07424553483724594,
+      "learning_rate": 1.4831545266327877e-06,
+      "loss": 0.3006,
+      "step": 49190
+    },
+    {
+      "epoch": 9.26,
+      "grad_norm": 40.09004592895508,
+      "learning_rate": 1.479390175042349e-06,
+      "loss": 0.4148,
+      "step": 49200
+    },
+    {
+      "epoch": 9.26,
+      "grad_norm": 40.092464447021484,
+      "learning_rate": 1.4756258234519106e-06,
+      "loss": 0.312,
+      "step": 49210
+    },
+    {
+      "epoch": 9.26,
+      "grad_norm": 28.163589477539062,
+      "learning_rate": 1.471861471861472e-06,
+      "loss": 0.281,
+      "step": 49220
+    },
+    {
+      "epoch": 9.27,
+      "grad_norm": 12.769076347351074,
+      "learning_rate": 1.4680971202710335e-06,
+      "loss": 0.5183,
+      "step": 49230
+    },
+    {
+      "epoch": 9.27,
+      "grad_norm": 2.987441301345825,
+      "learning_rate": 1.464332768680595e-06,
+      "loss": 0.5749,
+      "step": 49240
+    },
+    {
+      "epoch": 9.27,
+      "grad_norm": 6.323493003845215,
+      "learning_rate": 1.4605684170901562e-06,
+      "loss": 0.4274,
+      "step": 49250
+    },
+    {
+      "epoch": 9.27,
+      "grad_norm": 10.778115272521973,
+      "learning_rate": 1.4568040654997179e-06,
+      "loss": 0.5296,
+      "step": 49260
+    },
+    {
+      "epoch": 9.27,
+      "grad_norm": 0.0840461477637291,
+      "learning_rate": 1.4530397139092791e-06,
+      "loss": 0.3179,
+      "step": 49270
+    },
+    {
+      "epoch": 9.28,
+      "grad_norm": 0.9376565217971802,
+      "learning_rate": 1.4492753623188408e-06,
+      "loss": 0.2639,
+      "step": 49280
+    },
+    {
+      "epoch": 9.28,
+      "grad_norm": 13.167097091674805,
+      "learning_rate": 1.445511010728402e-06,
+      "loss": 0.3673,
+      "step": 49290
+    },
+    {
+      "epoch": 9.28,
+      "grad_norm": 0.04004925489425659,
+      "learning_rate": 1.4417466591379637e-06,
+      "loss": 0.349,
+      "step": 49300
+    },
+    {
+      "epoch": 9.28,
+      "grad_norm": 21.248655319213867,
+      "learning_rate": 1.437982307547525e-06,
+      "loss": 0.6205,
+      "step": 49310
+    },
+    {
+      "epoch": 9.28,
+      "grad_norm": 1.5469751358032227,
+      "learning_rate": 1.4342179559570864e-06,
+      "loss": 0.3818,
+      "step": 49320
+    },
+    {
+      "epoch": 9.28,
+      "grad_norm": 0.027218230068683624,
+      "learning_rate": 1.4304536043666481e-06,
+      "loss": 0.3784,
+      "step": 49330
+    },
+    {
+      "epoch": 9.29,
+      "grad_norm": 21.65924072265625,
+      "learning_rate": 1.4266892527762094e-06,
+      "loss": 0.4164,
+      "step": 49340
+    },
+    {
+      "epoch": 9.29,
+      "grad_norm": 32.658966064453125,
+      "learning_rate": 1.422924901185771e-06,
+      "loss": 0.4811,
+      "step": 49350
+    },
+    {
+      "epoch": 9.29,
+      "grad_norm": 0.4044090509414673,
+      "learning_rate": 1.4191605495953323e-06,
+      "loss": 0.3344,
+      "step": 49360
+    },
+    {
+      "epoch": 9.29,
+      "grad_norm": 25.323366165161133,
+      "learning_rate": 1.415396198004894e-06,
+      "loss": 0.2475,
+      "step": 49370
+    },
+    {
+      "epoch": 9.29,
+      "grad_norm": 10.626187324523926,
+      "learning_rate": 1.4116318464144552e-06,
+      "loss": 0.4904,
+      "step": 49380
+    },
+    {
+      "epoch": 9.3,
+      "grad_norm": 11.54336166381836,
+      "learning_rate": 1.4078674948240167e-06,
+      "loss": 0.4524,
+      "step": 49390
+    },
+    {
+      "epoch": 9.3,
+      "grad_norm": 26.418764114379883,
+      "learning_rate": 1.4041031432335781e-06,
+      "loss": 0.5615,
+      "step": 49400
+    },
+    {
+      "epoch": 9.3,
+      "grad_norm": 3.740413188934326,
+      "learning_rate": 1.4003387916431396e-06,
+      "loss": 0.1174,
+      "step": 49410
+    },
+    {
+      "epoch": 9.3,
+      "grad_norm": 15.656806945800781,
+      "learning_rate": 1.3965744400527008e-06,
+      "loss": 0.5396,
+      "step": 49420
+    },
+    {
+      "epoch": 9.3,
+      "grad_norm": 0.4741308391094208,
+      "learning_rate": 1.3928100884622625e-06,
+      "loss": 0.5084,
+      "step": 49430
+    },
+    {
+      "epoch": 9.31,
+      "grad_norm": 3.0225412845611572,
+      "learning_rate": 1.3890457368718238e-06,
+      "loss": 0.4074,
+      "step": 49440
+    },
+    {
+      "epoch": 9.31,
+      "grad_norm": 0.20880745351314545,
+      "learning_rate": 1.3852813852813854e-06,
+      "loss": 0.2169,
+      "step": 49450
+    },
+    {
+      "epoch": 9.31,
+      "grad_norm": 0.055619094520807266,
+      "learning_rate": 1.3815170336909469e-06,
+      "loss": 0.2984,
+      "step": 49460
+    },
+    {
+      "epoch": 9.31,
+      "grad_norm": 0.15215404331684113,
+      "learning_rate": 1.3777526821005083e-06,
+      "loss": 0.2482,
+      "step": 49470
+    },
+    {
+      "epoch": 9.31,
+      "grad_norm": 5.14688777923584,
+      "learning_rate": 1.3739883305100698e-06,
+      "loss": 0.285,
+      "step": 49480
+    },
+    {
+      "epoch": 9.31,
+      "grad_norm": 19.21902084350586,
+      "learning_rate": 1.370223978919631e-06,
+      "loss": 0.4919,
+      "step": 49490
+    },
+    {
+      "epoch": 9.32,
+      "grad_norm": 23.251541137695312,
+      "learning_rate": 1.3664596273291927e-06,
+      "loss": 0.4524,
+      "step": 49500
+    },
+    {
+      "epoch": 9.32,
+      "grad_norm": 5.6620588302612305,
+      "learning_rate": 1.362695275738754e-06,
+      "loss": 0.3456,
+      "step": 49510
+    },
+    {
+      "epoch": 9.32,
+      "grad_norm": 0.2966133654117584,
+      "learning_rate": 1.3589309241483157e-06,
+      "loss": 0.4254,
+      "step": 49520
+    },
+    {
+      "epoch": 9.32,
+      "grad_norm": 1.9271348714828491,
+      "learning_rate": 1.355166572557877e-06,
+      "loss": 0.75,
+      "step": 49530
+    },
+    {
+      "epoch": 9.32,
+      "grad_norm": 47.30406188964844,
+      "learning_rate": 1.3514022209674386e-06,
+      "loss": 0.2837,
+      "step": 49540
+    },
+    {
+      "epoch": 9.33,
+      "grad_norm": 6.934784889221191,
+      "learning_rate": 1.3476378693769998e-06,
+      "loss": 0.1063,
+      "step": 49550
+    },
+    {
+      "epoch": 9.33,
+      "grad_norm": 0.4049472212791443,
+      "learning_rate": 1.3438735177865615e-06,
+      "loss": 0.2103,
+      "step": 49560
+    },
+    {
+      "epoch": 9.33,
+      "grad_norm": 8.255983352661133,
+      "learning_rate": 1.3401091661961227e-06,
+      "loss": 0.4509,
+      "step": 49570
+    },
+    {
+      "epoch": 9.33,
+      "grad_norm": 14.6072416305542,
+      "learning_rate": 1.3363448146056842e-06,
+      "loss": 0.6148,
+      "step": 49580
+    },
+    {
+      "epoch": 9.33,
+      "grad_norm": 0.24846920371055603,
+      "learning_rate": 1.3325804630152459e-06,
+      "loss": 0.2994,
+      "step": 49590
+    },
+    {
+      "epoch": 9.34,
+      "grad_norm": 10.933160781860352,
+      "learning_rate": 1.3288161114248071e-06,
+      "loss": 0.3762,
+      "step": 49600
+    },
+    {
+      "epoch": 9.34,
+      "grad_norm": 19.94293975830078,
+      "learning_rate": 1.3250517598343688e-06,
+      "loss": 0.6265,
+      "step": 49610
+    },
+    {
+      "epoch": 9.34,
+      "grad_norm": 0.2474469691514969,
+      "learning_rate": 1.32128740824393e-06,
+      "loss": 0.5879,
+      "step": 49620
+    },
+    {
+      "epoch": 9.34,
+      "grad_norm": 4.398937225341797,
+      "learning_rate": 1.3175230566534917e-06,
+      "loss": 0.4399,
+      "step": 49630
+    },
+    {
+      "epoch": 9.34,
+      "grad_norm": 0.6015844941139221,
+      "learning_rate": 1.313758705063053e-06,
+      "loss": 0.2968,
+      "step": 49640
+    },
+    {
+      "epoch": 9.35,
+      "grad_norm": 0.09437728673219681,
+      "learning_rate": 1.3099943534726144e-06,
+      "loss": 0.4135,
+      "step": 49650
+    },
+    {
+      "epoch": 9.35,
+      "grad_norm": 26.453380584716797,
+      "learning_rate": 1.3062300018821759e-06,
+      "loss": 0.3049,
+      "step": 49660
+    },
+    {
+      "epoch": 9.35,
+      "grad_norm": 1.22520911693573,
+      "learning_rate": 1.3024656502917373e-06,
+      "loss": 0.4276,
+      "step": 49670
+    },
+    {
+      "epoch": 9.35,
+      "grad_norm": 26.06831169128418,
+      "learning_rate": 1.2987012987012986e-06,
+      "loss": 0.4042,
+      "step": 49680
+    },
+    {
+      "epoch": 9.35,
+      "grad_norm": 34.11174392700195,
+      "learning_rate": 1.2949369471108603e-06,
+      "loss": 0.2479,
+      "step": 49690
+    },
+    {
+      "epoch": 9.35,
+      "grad_norm": 6.261451721191406,
+      "learning_rate": 1.291172595520422e-06,
+      "loss": 0.4995,
+      "step": 49700
+    },
+    {
+      "epoch": 9.36,
+      "grad_norm": 25.64111328125,
+      "learning_rate": 1.2874082439299832e-06,
+      "loss": 0.4061,
+      "step": 49710
+    },
+    {
+      "epoch": 9.36,
+      "grad_norm": 12.963571548461914,
+      "learning_rate": 1.2836438923395447e-06,
+      "loss": 0.3339,
+      "step": 49720
+    },
+    {
+      "epoch": 9.36,
+      "grad_norm": 22.063447952270508,
+      "learning_rate": 1.2798795407491061e-06,
+      "loss": 0.5355,
+      "step": 49730
+    },
+    {
+      "epoch": 9.36,
+      "grad_norm": 5.219394207000732,
+      "learning_rate": 1.2761151891586676e-06,
+      "loss": 0.3802,
+      "step": 49740
+    },
+    {
+      "epoch": 9.36,
+      "grad_norm": 1.1139286756515503,
+      "learning_rate": 1.2723508375682288e-06,
+      "loss": 0.1823,
+      "step": 49750
+    },
+    {
+      "epoch": 9.37,
+      "grad_norm": 36.65229034423828,
+      "learning_rate": 1.2685864859777905e-06,
+      "loss": 0.4732,
+      "step": 49760
+    },
+    {
+      "epoch": 9.37,
+      "grad_norm": 1.4162871837615967,
+      "learning_rate": 1.2648221343873517e-06,
+      "loss": 0.255,
+      "step": 49770
+    },
+    {
+      "epoch": 9.37,
+      "grad_norm": 1.9202029705047607,
+      "learning_rate": 1.2610577827969134e-06,
+      "loss": 0.5858,
+      "step": 49780
+    },
+    {
+      "epoch": 9.37,
+      "grad_norm": 0.02834530733525753,
+      "learning_rate": 1.2572934312064747e-06,
+      "loss": 0.5202,
+      "step": 49790
+    },
+    {
+      "epoch": 9.37,
+      "grad_norm": 8.294178009033203,
+      "learning_rate": 1.2535290796160363e-06,
+      "loss": 0.1606,
+      "step": 49800
+    },
+    {
+      "epoch": 9.38,
+      "grad_norm": 10.036473274230957,
+      "learning_rate": 1.2497647280255978e-06,
+      "loss": 0.2661,
+      "step": 49810
+    },
+    {
+      "epoch": 9.38,
+      "grad_norm": 0.019691968336701393,
+      "learning_rate": 1.246000376435159e-06,
+      "loss": 0.4195,
+      "step": 49820
+    },
+    {
+      "epoch": 9.38,
+      "grad_norm": 14.037164688110352,
+      "learning_rate": 1.2422360248447205e-06,
+      "loss": 0.3664,
+      "step": 49830
+    },
+    {
+      "epoch": 9.38,
+      "grad_norm": 20.38290786743164,
+      "learning_rate": 1.238471673254282e-06,
+      "loss": 0.4841,
+      "step": 49840
+    },
+    {
+      "epoch": 9.38,
+      "grad_norm": 6.970554351806641,
+      "learning_rate": 1.2347073216638434e-06,
+      "loss": 0.2273,
+      "step": 49850
+    },
+    {
+      "epoch": 9.38,
+      "grad_norm": 22.626224517822266,
+      "learning_rate": 1.2309429700734049e-06,
+      "loss": 0.482,
+      "step": 49860
+    },
+    {
+      "epoch": 9.39,
+      "grad_norm": 17.559032440185547,
+      "learning_rate": 1.2271786184829664e-06,
+      "loss": 0.4027,
+      "step": 49870
+    },
+    {
+      "epoch": 9.39,
+      "grad_norm": 7.236875057220459,
+      "learning_rate": 1.223414266892528e-06,
+      "loss": 0.278,
+      "step": 49880
+    },
+    {
+      "epoch": 9.39,
+      "grad_norm": 10.33447265625,
+      "learning_rate": 1.2196499153020893e-06,
+      "loss": 0.8987,
+      "step": 49890
+    },
+    {
+      "epoch": 9.39,
+      "grad_norm": 15.287224769592285,
+      "learning_rate": 1.2158855637116507e-06,
+      "loss": 0.4645,
+      "step": 49900
+    },
+    {
+      "epoch": 9.39,
+      "grad_norm": 0.09853526204824448,
+      "learning_rate": 1.2121212121212122e-06,
+      "loss": 0.7842,
+      "step": 49910
+    },
+    {
+      "epoch": 9.4,
+      "grad_norm": 7.803390979766846,
+      "learning_rate": 1.2083568605307737e-06,
+      "loss": 0.5503,
+      "step": 49920
+    },
+    {
+      "epoch": 9.4,
+      "grad_norm": 34.554622650146484,
+      "learning_rate": 1.2045925089403351e-06,
+      "loss": 0.3182,
+      "step": 49930
+    },
+    {
+      "epoch": 9.4,
+      "grad_norm": 0.17311842739582062,
+      "learning_rate": 1.2008281573498966e-06,
+      "loss": 0.5634,
+      "step": 49940
+    },
+    {
+      "epoch": 9.4,
+      "grad_norm": 0.07609983533620834,
+      "learning_rate": 1.197063805759458e-06,
+      "loss": 0.3777,
+      "step": 49950
+    },
+    {
+      "epoch": 9.4,
+      "grad_norm": 0.10443427413702011,
+      "learning_rate": 1.1932994541690195e-06,
+      "loss": 0.3804,
+      "step": 49960
+    },
+    {
+      "epoch": 9.41,
+      "grad_norm": 1.417898178100586,
+      "learning_rate": 1.189535102578581e-06,
+      "loss": 0.3145,
+      "step": 49970
+    },
+    {
+      "epoch": 9.41,
+      "grad_norm": 0.049046795815229416,
+      "learning_rate": 1.1857707509881424e-06,
+      "loss": 0.5672,
+      "step": 49980
+    },
+    {
+      "epoch": 9.41,
+      "grad_norm": 0.9617329835891724,
+      "learning_rate": 1.1820063993977039e-06,
+      "loss": 0.5691,
+      "step": 49990
+    },
+    {
+      "epoch": 9.41,
+      "grad_norm": 10.128259658813477,
+      "learning_rate": 1.1782420478072651e-06,
+      "loss": 0.5053,
+      "step": 50000
+    },
+    {
+      "epoch": 9.41,
+      "grad_norm": 5.14787483215332,
+      "learning_rate": 1.1744776962168268e-06,
+      "loss": 0.5338,
+      "step": 50010
+    },
+    {
+      "epoch": 9.41,
+      "grad_norm": 0.5360021591186523,
+      "learning_rate": 1.1707133446263883e-06,
+      "loss": 0.3056,
+      "step": 50020
+    },
+    {
+      "epoch": 9.42,
+      "grad_norm": 10.969987869262695,
+      "learning_rate": 1.1669489930359497e-06,
+      "loss": 0.4361,
+      "step": 50030
+    },
+    {
+      "epoch": 9.42,
+      "grad_norm": 26.70008659362793,
+      "learning_rate": 1.1631846414455112e-06,
+      "loss": 0.3068,
+      "step": 50040
+    },
+    {
+      "epoch": 9.42,
+      "grad_norm": 10.523760795593262,
+      "learning_rate": 1.1594202898550726e-06,
+      "loss": 0.3849,
+      "step": 50050
+    },
+    {
+      "epoch": 9.42,
+      "grad_norm": 10.677010536193848,
+      "learning_rate": 1.155655938264634e-06,
+      "loss": 0.4415,
+      "step": 50060
+    },
+    {
+      "epoch": 9.42,
+      "grad_norm": 7.661884307861328,
+      "learning_rate": 1.1518915866741954e-06,
+      "loss": 0.0457,
+      "step": 50070
+    },
+    {
+      "epoch": 9.43,
+      "grad_norm": 15.285215377807617,
+      "learning_rate": 1.1481272350837568e-06,
+      "loss": 0.2535,
+      "step": 50080
+    },
+    {
+      "epoch": 9.43,
+      "grad_norm": 42.20638656616211,
+      "learning_rate": 1.1443628834933183e-06,
+      "loss": 0.4552,
+      "step": 50090
+    },
+    {
+      "epoch": 9.43,
+      "grad_norm": 35.964088439941406,
+      "learning_rate": 1.1405985319028797e-06,
+      "loss": 0.589,
+      "step": 50100
+    },
+    {
+      "epoch": 9.43,
+      "grad_norm": 2.185014486312866,
+      "learning_rate": 1.1368341803124412e-06,
+      "loss": 0.6969,
+      "step": 50110
+    },
+    {
+      "epoch": 9.43,
+      "grad_norm": 0.04032859951257706,
+      "learning_rate": 1.1330698287220027e-06,
+      "loss": 0.3934,
+      "step": 50120
+    },
+    {
+      "epoch": 9.44,
+      "grad_norm": 62.85636520385742,
+      "learning_rate": 1.1293054771315643e-06,
+      "loss": 0.4421,
+      "step": 50130
+    },
+    {
+      "epoch": 9.44,
+      "grad_norm": 11.45727252960205,
+      "learning_rate": 1.1255411255411256e-06,
+      "loss": 0.5068,
+      "step": 50140
+    },
+    {
+      "epoch": 9.44,
+      "grad_norm": 0.05326225981116295,
+      "learning_rate": 1.121776773950687e-06,
+      "loss": 0.3979,
+      "step": 50150
+    },
+    {
+      "epoch": 9.44,
+      "grad_norm": 16.516693115234375,
+      "learning_rate": 1.1180124223602485e-06,
+      "loss": 0.2904,
+      "step": 50160
+    },
+    {
+      "epoch": 9.44,
+      "grad_norm": 20.180761337280273,
+      "learning_rate": 1.11424807076981e-06,
+      "loss": 0.3479,
+      "step": 50170
+    },
+    {
+      "epoch": 9.44,
+      "grad_norm": 7.504915714263916,
+      "learning_rate": 1.1104837191793714e-06,
+      "loss": 0.4464,
+      "step": 50180
+    },
+    {
+      "epoch": 9.45,
+      "grad_norm": 8.153180122375488,
+      "learning_rate": 1.1067193675889329e-06,
+      "loss": 0.4644,
+      "step": 50190
+    },
+    {
+      "epoch": 9.45,
+      "grad_norm": 0.3960212171077728,
+      "learning_rate": 1.1029550159984943e-06,
+      "loss": 0.3971,
+      "step": 50200
+    },
+    {
+      "epoch": 9.45,
+      "grad_norm": 20.968027114868164,
+      "learning_rate": 1.0991906644080558e-06,
+      "loss": 0.2904,
+      "step": 50210
+    },
+    {
+      "epoch": 9.45,
+      "grad_norm": 28.1608943939209,
+      "learning_rate": 1.0954263128176173e-06,
+      "loss": 0.3588,
+      "step": 50220
+    },
+    {
+      "epoch": 9.45,
+      "grad_norm": 15.617292404174805,
+      "learning_rate": 1.0916619612271787e-06,
+      "loss": 0.4479,
+      "step": 50230
+    },
+    {
+      "epoch": 9.46,
+      "grad_norm": 0.6835996508598328,
+      "learning_rate": 1.0878976096367402e-06,
+      "loss": 0.3493,
+      "step": 50240
+    },
+    {
+      "epoch": 9.46,
+      "grad_norm": 11.191716194152832,
+      "learning_rate": 1.0841332580463016e-06,
+      "loss": 0.3882,
+      "step": 50250
+    },
+    {
+      "epoch": 9.46,
+      "grad_norm": 4.108872890472412,
+      "learning_rate": 1.0803689064558631e-06,
+      "loss": 0.4316,
+      "step": 50260
+    },
+    {
+      "epoch": 9.46,
+      "grad_norm": 6.614520072937012,
+      "learning_rate": 1.0766045548654246e-06,
+      "loss": 0.7346,
+      "step": 50270
+    },
+    {
+      "epoch": 9.46,
+      "grad_norm": 11.33967113494873,
+      "learning_rate": 1.072840203274986e-06,
+      "loss": 0.5321,
+      "step": 50280
+    },
+    {
+      "epoch": 9.47,
+      "grad_norm": 4.6037492752075195,
+      "learning_rate": 1.0690758516845475e-06,
+      "loss": 0.2836,
+      "step": 50290
+    },
+    {
+      "epoch": 9.47,
+      "grad_norm": 13.846996307373047,
+      "learning_rate": 1.065311500094109e-06,
+      "loss": 0.5464,
+      "step": 50300
+    },
+    {
+      "epoch": 9.47,
+      "grad_norm": 0.2525990307331085,
+      "learning_rate": 1.0615471485036704e-06,
+      "loss": 0.3565,
+      "step": 50310
+    },
+    {
+      "epoch": 9.47,
+      "grad_norm": 15.951204299926758,
+      "learning_rate": 1.0577827969132319e-06,
+      "loss": 0.4183,
+      "step": 50320
+    },
+    {
+      "epoch": 9.47,
+      "grad_norm": 5.232489585876465,
+      "learning_rate": 1.0540184453227931e-06,
+      "loss": 0.3319,
+      "step": 50330
+    },
+    {
+      "epoch": 9.47,
+      "grad_norm": 0.023767894133925438,
+      "learning_rate": 1.0502540937323546e-06,
+      "loss": 0.5855,
+      "step": 50340
+    },
+    {
+      "epoch": 9.48,
+      "grad_norm": 0.39152446389198303,
+      "learning_rate": 1.046489742141916e-06,
+      "loss": 0.3209,
+      "step": 50350
+    },
+    {
+      "epoch": 9.48,
+      "grad_norm": 15.136631965637207,
+      "learning_rate": 1.0427253905514775e-06,
+      "loss": 0.4202,
+      "step": 50360
+    },
+    {
+      "epoch": 9.48,
+      "grad_norm": 5.339528560638428,
+      "learning_rate": 1.0389610389610392e-06,
+      "loss": 0.2623,
+      "step": 50370
+    },
+    {
+      "epoch": 9.48,
+      "grad_norm": 5.4745097160339355,
+      "learning_rate": 1.0351966873706006e-06,
+      "loss": 0.317,
+      "step": 50380
+    },
+    {
+      "epoch": 9.48,
+      "grad_norm": 6.607236385345459,
+      "learning_rate": 1.031432335780162e-06,
+      "loss": 0.4808,
+      "step": 50390
+    },
+    {
+      "epoch": 9.49,
+      "grad_norm": 16.006132125854492,
+      "learning_rate": 1.0276679841897233e-06,
+      "loss": 0.195,
+      "step": 50400
+    },
+    {
+      "epoch": 9.49,
+      "grad_norm": 3.6388096809387207,
+      "learning_rate": 1.0239036325992848e-06,
+      "loss": 0.3289,
+      "step": 50410
+    },
+    {
+      "epoch": 9.49,
+      "grad_norm": 11.516218185424805,
+      "learning_rate": 1.0201392810088463e-06,
+      "loss": 0.4734,
+      "step": 50420
+    },
+    {
+      "epoch": 9.49,
+      "grad_norm": 30.199953079223633,
+      "learning_rate": 1.0163749294184077e-06,
+      "loss": 0.5322,
+      "step": 50430
+    },
+    {
+      "epoch": 9.49,
+      "grad_norm": 2.062737464904785,
+      "learning_rate": 1.0126105778279692e-06,
+      "loss": 0.2732,
+      "step": 50440
+    },
+    {
+      "epoch": 9.5,
+      "grad_norm": 0.2747786343097687,
+      "learning_rate": 1.0088462262375306e-06,
+      "loss": 0.389,
+      "step": 50450
+    },
+    {
+      "epoch": 9.5,
+      "grad_norm": 23.68992042541504,
+      "learning_rate": 1.0050818746470921e-06,
+      "loss": 0.7078,
+      "step": 50460
+    },
+    {
+      "epoch": 9.5,
+      "grad_norm": 10.194795608520508,
+      "learning_rate": 1.0013175230566536e-06,
+      "loss": 0.3264,
+      "step": 50470
+    },
+    {
+      "epoch": 9.5,
+      "grad_norm": 21.266271591186523,
+      "learning_rate": 9.97553171466215e-07,
+      "loss": 0.5062,
+      "step": 50480
+    },
+    {
+      "epoch": 9.5,
+      "grad_norm": 19.471515655517578,
+      "learning_rate": 9.937888198757765e-07,
+      "loss": 0.5086,
+      "step": 50490
+    },
+    {
+      "epoch": 9.5,
+      "grad_norm": 17.782821655273438,
+      "learning_rate": 9.90024468285338e-07,
+      "loss": 0.4353,
+      "step": 50500
+    },
+    {
+      "epoch": 9.51,
+      "grad_norm": 12.496779441833496,
+      "learning_rate": 9.862601166948994e-07,
+      "loss": 0.4012,
+      "step": 50510
+    },
+    {
+      "epoch": 9.51,
+      "grad_norm": 36.61666488647461,
+      "learning_rate": 9.824957651044609e-07,
+      "loss": 0.6408,
+      "step": 50520
+    },
+    {
+      "epoch": 9.51,
+      "grad_norm": 21.555734634399414,
+      "learning_rate": 9.787314135140223e-07,
+      "loss": 0.3065,
+      "step": 50530
+    },
+    {
+      "epoch": 9.51,
+      "grad_norm": 19.377365112304688,
+      "learning_rate": 9.749670619235838e-07,
+      "loss": 0.249,
+      "step": 50540
+    },
+    {
+      "epoch": 9.51,
+      "grad_norm": 20.095693588256836,
+      "learning_rate": 9.712027103331453e-07,
+      "loss": 0.3249,
+      "step": 50550
+    },
+    {
+      "epoch": 9.52,
+      "grad_norm": 18.395681381225586,
+      "learning_rate": 9.674383587427067e-07,
+      "loss": 0.3726,
+      "step": 50560
+    },
+    {
+      "epoch": 9.52,
+      "grad_norm": 0.07585670053958893,
+      "learning_rate": 9.636740071522682e-07,
+      "loss": 0.5645,
+      "step": 50570
+    },
+    {
+      "epoch": 9.52,
+      "grad_norm": 19.90555763244629,
+      "learning_rate": 9.599096555618294e-07,
+      "loss": 0.4348,
+      "step": 50580
+    },
+    {
+      "epoch": 9.52,
+      "grad_norm": 12.60806655883789,
+      "learning_rate": 9.561453039713909e-07,
+      "loss": 0.5422,
+      "step": 50590
+    },
+    {
+      "epoch": 9.52,
+      "grad_norm": 6.923961162567139,
+      "learning_rate": 9.523809523809525e-07,
+      "loss": 0.5583,
+      "step": 50600
+    },
+    {
+      "epoch": 9.53,
+      "grad_norm": 27.019386291503906,
+      "learning_rate": 9.486166007905138e-07,
+      "loss": 0.4477,
+      "step": 50610
+    },
+    {
+      "epoch": 9.53,
+      "grad_norm": 22.43336296081543,
+      "learning_rate": 9.448522492000754e-07,
+      "loss": 0.2844,
+      "step": 50620
+    },
+    {
+      "epoch": 9.53,
+      "grad_norm": 22.106285095214844,
+      "learning_rate": 9.410878976096368e-07,
+      "loss": 0.4775,
+      "step": 50630
+    },
+    {
+      "epoch": 9.53,
+      "grad_norm": 14.707420349121094,
+      "learning_rate": 9.373235460191983e-07,
+      "loss": 0.3398,
+      "step": 50640
+    },
+    {
+      "epoch": 9.53,
+      "grad_norm": 3.7978923320770264,
+      "learning_rate": 9.335591944287598e-07,
+      "loss": 0.3592,
+      "step": 50650
+    },
+    {
+      "epoch": 9.54,
+      "grad_norm": 16.250072479248047,
+      "learning_rate": 9.297948428383212e-07,
+      "loss": 0.5696,
+      "step": 50660
+    },
+    {
+      "epoch": 9.54,
+      "grad_norm": 5.815196990966797,
+      "learning_rate": 9.260304912478827e-07,
+      "loss": 0.2838,
+      "step": 50670
+    },
+    {
+      "epoch": 9.54,
+      "grad_norm": 0.15209926664829254,
+      "learning_rate": 9.22266139657444e-07,
+      "loss": 0.4138,
+      "step": 50680
+    },
+    {
+      "epoch": 9.54,
+      "grad_norm": 1.1849756240844727,
+      "learning_rate": 9.185017880670055e-07,
+      "loss": 0.5292,
+      "step": 50690
+    },
+    {
+      "epoch": 9.54,
+      "grad_norm": 22.73065948486328,
+      "learning_rate": 9.14737436476567e-07,
+      "loss": 0.367,
+      "step": 50700
+    },
+    {
+      "epoch": 9.54,
+      "grad_norm": 5.767063140869141,
+      "learning_rate": 9.109730848861284e-07,
+      "loss": 0.2446,
+      "step": 50710
+    },
+    {
+      "epoch": 9.55,
+      "grad_norm": 0.5526207089424133,
+      "learning_rate": 9.072087332956899e-07,
+      "loss": 0.3764,
+      "step": 50720
+    },
+    {
+      "epoch": 9.55,
+      "grad_norm": 13.495482444763184,
+      "learning_rate": 9.034443817052512e-07,
+      "loss": 0.4672,
+      "step": 50730
+    },
+    {
+      "epoch": 9.55,
+      "grad_norm": 3.9866769313812256,
+      "learning_rate": 8.996800301148129e-07,
+      "loss": 0.5576,
+      "step": 50740
+    },
+    {
+      "epoch": 9.55,
+      "grad_norm": 25.318235397338867,
+      "learning_rate": 8.959156785243743e-07,
+      "loss": 0.3028,
+      "step": 50750
+    },
+    {
+      "epoch": 9.55,
+      "grad_norm": 7.270644187927246,
+      "learning_rate": 8.921513269339357e-07,
+      "loss": 0.4609,
+      "step": 50760
+    },
+    {
+      "epoch": 9.56,
+      "grad_norm": 1.1293132305145264,
+      "learning_rate": 8.883869753434972e-07,
+      "loss": 0.3103,
+      "step": 50770
+    },
+    {
+      "epoch": 9.56,
+      "grad_norm": 2.1888303756713867,
+      "learning_rate": 8.846226237530586e-07,
+      "loss": 0.151,
+      "step": 50780
+    },
+    {
+      "epoch": 9.56,
+      "grad_norm": 16.56170082092285,
+      "learning_rate": 8.808582721626201e-07,
+      "loss": 0.4869,
+      "step": 50790
+    },
+    {
+      "epoch": 9.56,
+      "grad_norm": 15.505925178527832,
+      "learning_rate": 8.770939205721816e-07,
+      "loss": 0.6466,
+      "step": 50800
+    },
+    {
+      "epoch": 9.56,
+      "grad_norm": 37.347965240478516,
+      "learning_rate": 8.733295689817429e-07,
+      "loss": 0.5309,
+      "step": 50810
+    },
+    {
+      "epoch": 9.57,
+      "grad_norm": 9.028761863708496,
+      "learning_rate": 8.695652173913044e-07,
+      "loss": 0.4925,
+      "step": 50820
+    },
+    {
+      "epoch": 9.57,
+      "grad_norm": 18.40884780883789,
+      "learning_rate": 8.658008658008658e-07,
+      "loss": 0.4293,
+      "step": 50830
+    },
+    {
+      "epoch": 9.57,
+      "grad_norm": 20.18854331970215,
+      "learning_rate": 8.620365142104273e-07,
+      "loss": 0.4924,
+      "step": 50840
+    },
+    {
+      "epoch": 9.57,
+      "grad_norm": 17.72684097290039,
+      "learning_rate": 8.582721626199888e-07,
+      "loss": 0.3837,
+      "step": 50850
+    },
+    {
+      "epoch": 9.57,
+      "grad_norm": 8.206231117248535,
+      "learning_rate": 8.545078110295501e-07,
+      "loss": 0.463,
+      "step": 50860
+    },
+    {
+      "epoch": 9.57,
+      "grad_norm": 0.24733124673366547,
+      "learning_rate": 8.507434594391118e-07,
+      "loss": 0.5556,
+      "step": 50870
+    },
+    {
+      "epoch": 9.58,
+      "grad_norm": 21.102048873901367,
+      "learning_rate": 8.469791078486731e-07,
+      "loss": 0.3003,
+      "step": 50880
+    },
+    {
+      "epoch": 9.58,
+      "grad_norm": 9.614502906799316,
+      "learning_rate": 8.432147562582346e-07,
+      "loss": 0.4534,
+      "step": 50890
+    },
+    {
+      "epoch": 9.58,
+      "grad_norm": 0.05175158008933067,
+      "learning_rate": 8.394504046677961e-07,
+      "loss": 0.3368,
+      "step": 50900
+    },
+    {
+      "epoch": 9.58,
+      "grad_norm": 18.473926544189453,
+      "learning_rate": 8.356860530773575e-07,
+      "loss": 0.6013,
+      "step": 50910
+    },
+    {
+      "epoch": 9.58,
+      "grad_norm": 18.408864974975586,
+      "learning_rate": 8.31921701486919e-07,
+      "loss": 0.4173,
+      "step": 50920
+    },
+    {
+      "epoch": 9.59,
+      "grad_norm": 27.239898681640625,
+      "learning_rate": 8.281573498964803e-07,
+      "loss": 0.4574,
+      "step": 50930
+    },
+    {
+      "epoch": 9.59,
+      "grad_norm": 6.1605753898620605,
+      "learning_rate": 8.243929983060418e-07,
+      "loss": 0.6228,
+      "step": 50940
+    },
+    {
+      "epoch": 9.59,
+      "grad_norm": 4.264019012451172,
+      "learning_rate": 8.206286467156033e-07,
+      "loss": 0.2077,
+      "step": 50950
+    },
+    {
+      "epoch": 9.59,
+      "grad_norm": 9.811421394348145,
+      "learning_rate": 8.168642951251647e-07,
+      "loss": 0.4051,
+      "step": 50960
+    },
+    {
+      "epoch": 9.59,
+      "grad_norm": 5.31566858291626,
+      "learning_rate": 8.130999435347262e-07,
+      "loss": 0.37,
+      "step": 50970
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 0.32057228684425354,
+      "learning_rate": 8.093355919442876e-07,
+      "loss": 0.6192,
+      "step": 50980
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 5.027352809906006,
+      "learning_rate": 8.055712403538492e-07,
+      "loss": 0.7235,
+      "step": 50990
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 0.16423563659191132,
+      "learning_rate": 8.018068887634107e-07,
+      "loss": 0.3614,
+      "step": 51000
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 7.00469970703125,
+      "learning_rate": 7.98042537172972e-07,
+      "loss": 0.2669,
+      "step": 51010
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 13.766214370727539,
+      "learning_rate": 7.942781855825335e-07,
+      "loss": 0.3001,
+      "step": 51020
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 15.029500007629395,
+      "learning_rate": 7.905138339920949e-07,
+      "loss": 0.5506,
+      "step": 51030
+    },
+    {
+      "epoch": 9.61,
+      "grad_norm": 18.616168975830078,
+      "learning_rate": 7.867494824016564e-07,
+      "loss": 0.8327,
+      "step": 51040
+    },
+    {
+      "epoch": 9.61,
+      "grad_norm": 1.1364197731018066,
+      "learning_rate": 7.829851308112179e-07,
+      "loss": 0.3129,
+      "step": 51050
+    },
+    {
+      "epoch": 9.61,
+      "grad_norm": 29.443477630615234,
+      "learning_rate": 7.792207792207792e-07,
+      "loss": 0.5467,
+      "step": 51060
+    },
+    {
+      "epoch": 9.61,
+      "grad_norm": 34.73600387573242,
+      "learning_rate": 7.754564276303407e-07,
+      "loss": 0.4792,
+      "step": 51070
+    },
+    {
+      "epoch": 9.61,
+      "grad_norm": 0.39562925696372986,
+      "learning_rate": 7.716920760399021e-07,
+      "loss": 0.4972,
+      "step": 51080
+    },
+    {
+      "epoch": 9.62,
+      "grad_norm": 21.628002166748047,
+      "learning_rate": 7.679277244494636e-07,
+      "loss": 0.6476,
+      "step": 51090
+    },
+    {
+      "epoch": 9.62,
+      "grad_norm": 0.8809748888015747,
+      "learning_rate": 7.641633728590251e-07,
+      "loss": 0.5343,
+      "step": 51100
+    },
+    {
+      "epoch": 9.62,
+      "grad_norm": 28.88115882873535,
+      "learning_rate": 7.603990212685866e-07,
+      "loss": 0.3758,
+      "step": 51110
+    },
+    {
+      "epoch": 9.62,
+      "grad_norm": 8.242012977600098,
+      "learning_rate": 7.566346696781481e-07,
+      "loss": 0.3154,
+      "step": 51120
+    },
+    {
+      "epoch": 9.62,
+      "grad_norm": 0.03372791409492493,
+      "learning_rate": 7.528703180877094e-07,
+      "loss": 0.4118,
+      "step": 51130
+    },
+    {
+      "epoch": 9.63,
+      "grad_norm": 3.514137029647827,
+      "learning_rate": 7.491059664972709e-07,
+      "loss": 0.3894,
+      "step": 51140
+    },
+    {
+      "epoch": 9.63,
+      "grad_norm": 0.49889636039733887,
+      "learning_rate": 7.453416149068324e-07,
+      "loss": 0.3594,
+      "step": 51150
+    },
+    {
+      "epoch": 9.63,
+      "grad_norm": 0.3000124394893646,
+      "learning_rate": 7.415772633163938e-07,
+      "loss": 0.3072,
+      "step": 51160
+    },
+    {
+      "epoch": 9.63,
+      "grad_norm": 0.06081683188676834,
+      "learning_rate": 7.378129117259553e-07,
+      "loss": 0.3226,
+      "step": 51170
+    },
+    {
+      "epoch": 9.63,
+      "grad_norm": 5.96708345413208,
+      "learning_rate": 7.340485601355168e-07,
+      "loss": 0.4402,
+      "step": 51180
+    },
+    {
+      "epoch": 9.63,
+      "grad_norm": 0.35737451910972595,
+      "learning_rate": 7.302842085450781e-07,
+      "loss": 0.7907,
+      "step": 51190
+    },
+    {
+      "epoch": 9.64,
+      "grad_norm": 4.599313735961914,
+      "learning_rate": 7.265198569546396e-07,
+      "loss": 0.1227,
+      "step": 51200
+    },
+    {
+      "epoch": 9.64,
+      "grad_norm": 4.307950973510742,
+      "learning_rate": 7.22755505364201e-07,
+      "loss": 0.2469,
+      "step": 51210
+    },
+    {
+      "epoch": 9.64,
+      "grad_norm": 22.732894897460938,
+      "learning_rate": 7.189911537737625e-07,
+      "loss": 0.4584,
+      "step": 51220
+    },
+    {
+      "epoch": 9.64,
+      "grad_norm": 21.445600509643555,
+      "learning_rate": 7.152268021833241e-07,
+      "loss": 0.355,
+      "step": 51230
+    },
+    {
+      "epoch": 9.64,
+      "grad_norm": 7.818153381347656,
+      "learning_rate": 7.114624505928855e-07,
+      "loss": 0.3512,
+      "step": 51240
+    },
+    {
+      "epoch": 9.65,
+      "grad_norm": 11.798916816711426,
+      "learning_rate": 7.07698099002447e-07,
+      "loss": 0.4811,
+      "step": 51250
+    },
+    {
+      "epoch": 9.65,
+      "grad_norm": 21.753093719482422,
+      "learning_rate": 7.039337474120083e-07,
+      "loss": 0.3904,
+      "step": 51260
+    },
+    {
+      "epoch": 9.65,
+      "grad_norm": 9.064457893371582,
+      "learning_rate": 7.001693958215698e-07,
+      "loss": 0.3594,
+      "step": 51270
+    },
+    {
+      "epoch": 9.65,
+      "grad_norm": 0.17999929189682007,
+      "learning_rate": 6.964050442311313e-07,
+      "loss": 0.5093,
+      "step": 51280
+    },
+    {
+      "epoch": 9.65,
+      "grad_norm": 27.64767837524414,
+      "learning_rate": 6.926406926406927e-07,
+      "loss": 0.4248,
+      "step": 51290
+    },
+    {
+      "epoch": 9.66,
+      "grad_norm": 13.119711875915527,
+      "learning_rate": 6.888763410502542e-07,
+      "loss": 0.2889,
+      "step": 51300
+    },
+    {
+      "epoch": 9.66,
+      "grad_norm": 17.546010971069336,
+      "learning_rate": 6.851119894598155e-07,
+      "loss": 0.402,
+      "step": 51310
+    },
+    {
+      "epoch": 9.66,
+      "grad_norm": 3.10264253616333,
+      "learning_rate": 6.81347637869377e-07,
+      "loss": 0.3537,
+      "step": 51320
+    },
+    {
+      "epoch": 9.66,
+      "grad_norm": 34.127784729003906,
+      "learning_rate": 6.775832862789384e-07,
+      "loss": 0.4739,
+      "step": 51330
+    },
+    {
+      "epoch": 9.66,
+      "grad_norm": 20.95803451538086,
+      "learning_rate": 6.738189346884999e-07,
+      "loss": 0.2397,
+      "step": 51340
+    },
+    {
+      "epoch": 9.66,
+      "grad_norm": 20.92197036743164,
+      "learning_rate": 6.700545830980614e-07,
+      "loss": 0.7235,
+      "step": 51350
+    },
+    {
+      "epoch": 9.67,
+      "grad_norm": 0.09385684132575989,
+      "learning_rate": 6.662902315076229e-07,
+      "loss": 0.612,
+      "step": 51360
+    },
+    {
+      "epoch": 9.67,
+      "grad_norm": 11.531614303588867,
+      "learning_rate": 6.625258799171844e-07,
+      "loss": 0.1948,
+      "step": 51370
+    },
+    {
+      "epoch": 9.67,
+      "grad_norm": 15.86019515991211,
+      "learning_rate": 6.587615283267459e-07,
+      "loss": 0.2973,
+      "step": 51380
+    },
+    {
+      "epoch": 9.67,
+      "grad_norm": 12.996696472167969,
+      "learning_rate": 6.549971767363072e-07,
+      "loss": 0.3127,
+      "step": 51390
+    },
+    {
+      "epoch": 9.67,
+      "grad_norm": 12.861435890197754,
+      "learning_rate": 6.512328251458687e-07,
+      "loss": 0.5937,
+      "step": 51400
+    },
+    {
+      "epoch": 9.68,
+      "grad_norm": 8.911890029907227,
+      "learning_rate": 6.474684735554301e-07,
+      "loss": 0.364,
+      "step": 51410
+    },
+    {
+      "epoch": 9.68,
+      "grad_norm": 24.71923828125,
+      "learning_rate": 6.437041219649916e-07,
+      "loss": 0.6153,
+      "step": 51420
+    },
+    {
+      "epoch": 9.68,
+      "grad_norm": 17.451587677001953,
+      "learning_rate": 6.399397703745531e-07,
+      "loss": 0.6571,
+      "step": 51430
+    },
+    {
+      "epoch": 9.68,
+      "grad_norm": 20.86277198791504,
+      "learning_rate": 6.361754187841144e-07,
+      "loss": 0.3892,
+      "step": 51440
+    },
+    {
+      "epoch": 9.68,
+      "grad_norm": 32.68037414550781,
+      "learning_rate": 6.324110671936759e-07,
+      "loss": 0.2615,
+      "step": 51450
+    },
+    {
+      "epoch": 9.69,
+      "grad_norm": 25.63720703125,
+      "learning_rate": 6.286467156032373e-07,
+      "loss": 0.3267,
+      "step": 51460
+    },
+    {
+      "epoch": 9.69,
+      "grad_norm": 5.5767645835876465,
+      "learning_rate": 6.248823640127989e-07,
+      "loss": 0.2016,
+      "step": 51470
+    },
+    {
+      "epoch": 9.69,
+      "grad_norm": 27.157485961914062,
+      "learning_rate": 6.211180124223603e-07,
+      "loss": 0.6425,
+      "step": 51480
+    },
+    {
+      "epoch": 9.69,
+      "grad_norm": 8.377914428710938,
+      "learning_rate": 6.173536608319217e-07,
+      "loss": 0.2027,
+      "step": 51490
+    },
+    {
+      "epoch": 9.69,
+      "grad_norm": 21.33051300048828,
+      "learning_rate": 6.135893092414832e-07,
+      "loss": 0.2035,
+      "step": 51500
+    },
+    {
+      "epoch": 9.7,
+      "grad_norm": 17.3638916015625,
+      "learning_rate": 6.098249576510446e-07,
+      "loss": 0.1943,
+      "step": 51510
+    },
+    {
+      "epoch": 9.7,
+      "grad_norm": 0.3384857475757599,
+      "learning_rate": 6.060606060606061e-07,
+      "loss": 0.1849,
+      "step": 51520
+    },
+    {
+      "epoch": 9.7,
+      "grad_norm": 1.3415716886520386,
+      "learning_rate": 6.022962544701676e-07,
+      "loss": 0.5632,
+      "step": 51530
+    },
+    {
+      "epoch": 9.7,
+      "grad_norm": 30.25799560546875,
+      "learning_rate": 5.98531902879729e-07,
+      "loss": 0.43,
+      "step": 51540
+    },
+    {
+      "epoch": 9.7,
+      "grad_norm": 32.7429084777832,
+      "learning_rate": 5.947675512892905e-07,
+      "loss": 0.4043,
+      "step": 51550
+    },
+    {
+      "epoch": 9.7,
+      "grad_norm": 17.513532638549805,
+      "learning_rate": 5.910031996988519e-07,
+      "loss": 0.4679,
+      "step": 51560
+    },
+    {
+      "epoch": 9.71,
+      "grad_norm": 25.66455841064453,
+      "learning_rate": 5.872388481084134e-07,
+      "loss": 0.3619,
+      "step": 51570
+    },
+    {
+      "epoch": 9.71,
+      "grad_norm": 20.97486114501953,
+      "learning_rate": 5.834744965179749e-07,
+      "loss": 0.6867,
+      "step": 51580
+    },
+    {
+      "epoch": 9.71,
+      "grad_norm": 7.469673156738281,
+      "learning_rate": 5.797101449275363e-07,
+      "loss": 0.3848,
+      "step": 51590
+    },
+    {
+      "epoch": 9.71,
+      "grad_norm": 32.72283935546875,
+      "learning_rate": 5.759457933370977e-07,
+      "loss": 0.3702,
+      "step": 51600
+    },
+    {
+      "epoch": 9.71,
+      "grad_norm": 7.483311176300049,
+      "learning_rate": 5.721814417466591e-07,
+      "loss": 0.29,
+      "step": 51610
+    },
+    {
+      "epoch": 9.72,
+      "grad_norm": 12.17415714263916,
+      "learning_rate": 5.684170901562206e-07,
+      "loss": 0.3262,
+      "step": 51620
+    },
+    {
+      "epoch": 9.72,
+      "grad_norm": 18.75547218322754,
+      "learning_rate": 5.646527385657822e-07,
+      "loss": 0.4394,
+      "step": 51630
+    },
+    {
+      "epoch": 9.72,
+      "grad_norm": 12.477697372436523,
+      "learning_rate": 5.608883869753435e-07,
+      "loss": 0.4251,
+      "step": 51640
+    },
+    {
+      "epoch": 9.72,
+      "grad_norm": 15.879676818847656,
+      "learning_rate": 5.57124035384905e-07,
+      "loss": 0.4689,
+      "step": 51650
+    },
+    {
+      "epoch": 9.72,
+      "grad_norm": 31.482437133789062,
+      "learning_rate": 5.533596837944664e-07,
+      "loss": 0.525,
+      "step": 51660
+    },
+    {
+      "epoch": 9.73,
+      "grad_norm": 9.857966423034668,
+      "learning_rate": 5.495953322040279e-07,
+      "loss": 0.313,
+      "step": 51670
+    },
+    {
+      "epoch": 9.73,
+      "grad_norm": 0.783555269241333,
+      "learning_rate": 5.458309806135894e-07,
+      "loss": 0.5831,
+      "step": 51680
+    },
+    {
+      "epoch": 9.73,
+      "grad_norm": 0.14815597236156464,
+      "learning_rate": 5.420666290231508e-07,
+      "loss": 0.5301,
+      "step": 51690
+    },
+    {
+      "epoch": 9.73,
+      "grad_norm": 4.29640531539917,
+      "learning_rate": 5.383022774327123e-07,
+      "loss": 0.4009,
+      "step": 51700
+    },
+    {
+      "epoch": 9.73,
+      "grad_norm": 14.452727317810059,
+      "learning_rate": 5.345379258422737e-07,
+      "loss": 0.3563,
+      "step": 51710
+    },
+    {
+      "epoch": 9.73,
+      "grad_norm": 1.4909330606460571,
+      "learning_rate": 5.307735742518352e-07,
+      "loss": 0.3447,
+      "step": 51720
+    },
+    {
+      "epoch": 9.74,
+      "grad_norm": 0.06365931034088135,
+      "learning_rate": 5.270092226613966e-07,
+      "loss": 0.1957,
+      "step": 51730
+    },
+    {
+      "epoch": 9.74,
+      "grad_norm": 1.6421184539794922,
+      "learning_rate": 5.23244871070958e-07,
+      "loss": 0.5981,
+      "step": 51740
+    },
+    {
+      "epoch": 9.74,
+      "grad_norm": 0.5028958320617676,
+      "learning_rate": 5.194805194805196e-07,
+      "loss": 0.3268,
+      "step": 51750
+    },
+    {
+      "epoch": 9.74,
+      "grad_norm": 4.667960166931152,
+      "learning_rate": 5.15716167890081e-07,
+      "loss": 0.1761,
+      "step": 51760
+    },
+    {
+      "epoch": 9.74,
+      "grad_norm": 1.6822726726531982,
+      "learning_rate": 5.119518162996424e-07,
+      "loss": 0.5087,
+      "step": 51770
+    },
+    {
+      "epoch": 9.75,
+      "grad_norm": 0.5095930099487305,
+      "learning_rate": 5.081874647092039e-07,
+      "loss": 0.6654,
+      "step": 51780
+    },
+    {
+      "epoch": 9.75,
+      "grad_norm": 0.034996677190065384,
+      "learning_rate": 5.044231131187653e-07,
+      "loss": 0.22,
+      "step": 51790
+    },
+    {
+      "epoch": 9.75,
+      "grad_norm": 0.16648948192596436,
+      "learning_rate": 5.006587615283268e-07,
+      "loss": 0.1263,
+      "step": 51800
+    },
+    {
+      "epoch": 9.75,
+      "grad_norm": 32.999053955078125,
+      "learning_rate": 4.968944099378882e-07,
+      "loss": 0.2778,
+      "step": 51810
+    },
+    {
+      "epoch": 9.75,
+      "grad_norm": 54.918148040771484,
+      "learning_rate": 4.931300583474497e-07,
+      "loss": 0.5931,
+      "step": 51820
+    },
+    {
+      "epoch": 9.76,
+      "grad_norm": 28.86396026611328,
+      "learning_rate": 4.893657067570112e-07,
+      "loss": 0.2479,
+      "step": 51830
+    },
+    {
+      "epoch": 9.76,
+      "grad_norm": 33.95524597167969,
+      "learning_rate": 4.856013551665726e-07,
+      "loss": 0.2287,
+      "step": 51840
+    },
+    {
+      "epoch": 9.76,
+      "grad_norm": 20.653039932250977,
+      "learning_rate": 4.818370035761341e-07,
+      "loss": 0.4423,
+      "step": 51850
+    },
+    {
+      "epoch": 9.76,
+      "grad_norm": 13.929657936096191,
+      "learning_rate": 4.780726519856954e-07,
+      "loss": 0.5723,
+      "step": 51860
+    },
+    {
+      "epoch": 9.76,
+      "grad_norm": 7.612604141235352,
+      "learning_rate": 4.743083003952569e-07,
+      "loss": 0.6222,
+      "step": 51870
+    },
+    {
+      "epoch": 9.76,
+      "grad_norm": 14.222421646118164,
+      "learning_rate": 4.705439488048184e-07,
+      "loss": 0.3785,
+      "step": 51880
+    },
+    {
+      "epoch": 9.77,
+      "grad_norm": 0.42615869641304016,
+      "learning_rate": 4.667795972143799e-07,
+      "loss": 0.4963,
+      "step": 51890
+    },
+    {
+      "epoch": 9.77,
+      "grad_norm": 7.262185096740723,
+      "learning_rate": 4.6301524562394134e-07,
+      "loss": 0.6145,
+      "step": 51900
+    },
+    {
+      "epoch": 9.77,
+      "grad_norm": 11.753568649291992,
+      "learning_rate": 4.5925089403350275e-07,
+      "loss": 0.5361,
+      "step": 51910
+    },
+    {
+      "epoch": 9.77,
+      "grad_norm": 0.5254884362220764,
+      "learning_rate": 4.554865424430642e-07,
+      "loss": 0.6324,
+      "step": 51920
+    },
+    {
+      "epoch": 9.77,
+      "grad_norm": 0.03477970510721207,
+      "learning_rate": 4.517221908526256e-07,
+      "loss": 0.401,
+      "step": 51930
+    },
+    {
+      "epoch": 9.78,
+      "grad_norm": 15.322552680969238,
+      "learning_rate": 4.4795783926218713e-07,
+      "loss": 0.5914,
+      "step": 51940
+    },
+    {
+      "epoch": 9.78,
+      "grad_norm": 8.4584379196167,
+      "learning_rate": 4.441934876717486e-07,
+      "loss": 0.4317,
+      "step": 51950
+    },
+    {
+      "epoch": 9.78,
+      "grad_norm": 15.443913459777832,
+      "learning_rate": 4.4042913608131005e-07,
+      "loss": 0.4615,
+      "step": 51960
+    },
+    {
+      "epoch": 9.78,
+      "grad_norm": 6.165141582489014,
+      "learning_rate": 4.3666478449087146e-07,
+      "loss": 0.3485,
+      "step": 51970
+    },
+    {
+      "epoch": 9.78,
+      "grad_norm": 22.44362449645996,
+      "learning_rate": 4.329004329004329e-07,
+      "loss": 0.493,
+      "step": 51980
+    },
+    {
+      "epoch": 9.79,
+      "grad_norm": 0.9760726690292358,
+      "learning_rate": 4.291360813099944e-07,
+      "loss": 0.2653,
+      "step": 51990
+    },
+    {
+      "epoch": 9.79,
+      "grad_norm": 0.020068148151040077,
+      "learning_rate": 4.253717297195559e-07,
+      "loss": 0.2711,
+      "step": 52000
+    },
+    {
+      "epoch": 9.79,
+      "grad_norm": 14.025554656982422,
+      "learning_rate": 4.216073781291173e-07,
+      "loss": 0.4435,
+      "step": 52010
+    },
+    {
+      "epoch": 9.79,
+      "grad_norm": 0.10281354933977127,
+      "learning_rate": 4.1784302653867876e-07,
+      "loss": 0.2737,
+      "step": 52020
+    },
+    {
+      "epoch": 9.79,
+      "grad_norm": 17.83829116821289,
+      "learning_rate": 4.1407867494824017e-07,
+      "loss": 0.5229,
+      "step": 52030
+    },
+    {
+      "epoch": 9.79,
+      "grad_norm": 0.09223546832799911,
+      "learning_rate": 4.1031432335780163e-07,
+      "loss": 0.4226,
+      "step": 52040
+    },
+    {
+      "epoch": 9.8,
+      "grad_norm": 0.02461303025484085,
+      "learning_rate": 4.065499717673631e-07,
+      "loss": 0.2307,
+      "step": 52050
+    },
+    {
+      "epoch": 9.8,
+      "grad_norm": 12.551608085632324,
+      "learning_rate": 4.027856201769246e-07,
+      "loss": 0.1268,
+      "step": 52060
+    },
+    {
+      "epoch": 9.8,
+      "grad_norm": 12.722687721252441,
+      "learning_rate": 3.99021268586486e-07,
+      "loss": 0.5292,
+      "step": 52070
+    },
+    {
+      "epoch": 9.8,
+      "grad_norm": 7.628570556640625,
+      "learning_rate": 3.9525691699604747e-07,
+      "loss": 0.2919,
+      "step": 52080
+    },
+    {
+      "epoch": 9.8,
+      "grad_norm": 7.446393013000488,
+      "learning_rate": 3.9149256540560893e-07,
+      "loss": 0.426,
+      "step": 52090
+    },
+    {
+      "epoch": 9.81,
+      "grad_norm": 24.474695205688477,
+      "learning_rate": 3.8772821381517034e-07,
+      "loss": 0.4843,
+      "step": 52100
+    },
+    {
+      "epoch": 9.81,
+      "grad_norm": 14.59511661529541,
+      "learning_rate": 3.839638622247318e-07,
+      "loss": 0.4363,
+      "step": 52110
+    },
+    {
+      "epoch": 9.81,
+      "grad_norm": 0.08716907352209091,
+      "learning_rate": 3.801995106342933e-07,
+      "loss": 0.3595,
+      "step": 52120
+    },
+    {
+      "epoch": 9.81,
+      "grad_norm": 32.84730529785156,
+      "learning_rate": 3.764351590438547e-07,
+      "loss": 0.5217,
+      "step": 52130
+    },
+    {
+      "epoch": 9.81,
+      "grad_norm": 1.3554562330245972,
+      "learning_rate": 3.726708074534162e-07,
+      "loss": 0.2144,
+      "step": 52140
+    },
+    {
+      "epoch": 9.82,
+      "grad_norm": 14.616629600524902,
+      "learning_rate": 3.6890645586297765e-07,
+      "loss": 0.4909,
+      "step": 52150
+    },
+    {
+      "epoch": 9.82,
+      "grad_norm": 15.903406143188477,
+      "learning_rate": 3.6514210427253905e-07,
+      "loss": 0.6216,
+      "step": 52160
+    },
+    {
+      "epoch": 9.82,
+      "grad_norm": 10.447539329528809,
+      "learning_rate": 3.613777526821005e-07,
+      "loss": 0.3794,
+      "step": 52170
+    },
+    {
+      "epoch": 9.82,
+      "grad_norm": 5.395421981811523,
+      "learning_rate": 3.5761340109166203e-07,
+      "loss": 0.4563,
+      "step": 52180
+    },
+    {
+      "epoch": 9.82,
+      "grad_norm": 2.901611328125,
+      "learning_rate": 3.538490495012235e-07,
+      "loss": 0.3466,
+      "step": 52190
+    },
+    {
+      "epoch": 9.82,
+      "grad_norm": 7.655229568481445,
+      "learning_rate": 3.500846979107849e-07,
+      "loss": 0.5837,
+      "step": 52200
+    },
+    {
+      "epoch": 9.83,
+      "grad_norm": 46.457275390625,
+      "learning_rate": 3.4632034632034636e-07,
+      "loss": 0.3432,
+      "step": 52210
+    },
+    {
+      "epoch": 9.83,
+      "grad_norm": 0.17484161257743835,
+      "learning_rate": 3.4255599472990776e-07,
+      "loss": 0.3191,
+      "step": 52220
+    },
+    {
+      "epoch": 9.83,
+      "grad_norm": 21.340837478637695,
+      "learning_rate": 3.387916431394692e-07,
+      "loss": 0.9333,
+      "step": 52230
+    },
+    {
+      "epoch": 9.83,
+      "grad_norm": 9.154580116271973,
+      "learning_rate": 3.350272915490307e-07,
+      "loss": 0.2877,
+      "step": 52240
+    },
+    {
+      "epoch": 9.83,
+      "grad_norm": 31.7108097076416,
+      "learning_rate": 3.312629399585922e-07,
+      "loss": 0.5454,
+      "step": 52250
+    },
+    {
+      "epoch": 9.84,
+      "grad_norm": 0.7132735252380371,
+      "learning_rate": 3.274985883681536e-07,
+      "loss": 0.2906,
+      "step": 52260
+    },
+    {
+      "epoch": 9.84,
+      "grad_norm": 9.988203048706055,
+      "learning_rate": 3.2373423677771507e-07,
+      "loss": 0.3561,
+      "step": 52270
+    },
+    {
+      "epoch": 9.84,
+      "grad_norm": 20.169540405273438,
+      "learning_rate": 3.1996988518727653e-07,
+      "loss": 0.5719,
+      "step": 52280
+    },
+    {
+      "epoch": 9.84,
+      "grad_norm": 2.338794708251953,
+      "learning_rate": 3.1620553359683794e-07,
+      "loss": 0.5348,
+      "step": 52290
+    },
+    {
+      "epoch": 9.84,
+      "grad_norm": 9.757970809936523,
+      "learning_rate": 3.1244118200639945e-07,
+      "loss": 0.3251,
+      "step": 52300
+    },
+    {
+      "epoch": 9.85,
+      "grad_norm": 16.40189552307129,
+      "learning_rate": 3.0867683041596086e-07,
+      "loss": 0.4938,
+      "step": 52310
+    },
+    {
+      "epoch": 9.85,
+      "grad_norm": 0.104170061647892,
+      "learning_rate": 3.049124788255223e-07,
+      "loss": 0.473,
+      "step": 52320
+    },
+    {
+      "epoch": 9.85,
+      "grad_norm": 16.414779663085938,
+      "learning_rate": 3.011481272350838e-07,
+      "loss": 0.4607,
+      "step": 52330
+    },
+    {
+      "epoch": 9.85,
+      "grad_norm": 10.880690574645996,
+      "learning_rate": 2.9738377564464524e-07,
+      "loss": 0.5413,
+      "step": 52340
+    },
+    {
+      "epoch": 9.85,
+      "grad_norm": 0.09918955713510513,
+      "learning_rate": 2.936194240542067e-07,
+      "loss": 0.3942,
+      "step": 52350
+    },
+    {
+      "epoch": 9.86,
+      "grad_norm": 31.41875457763672,
+      "learning_rate": 2.8985507246376816e-07,
+      "loss": 0.5384,
+      "step": 52360
+    },
+    {
+      "epoch": 9.86,
+      "grad_norm": 15.386962890625,
+      "learning_rate": 2.8609072087332957e-07,
+      "loss": 0.4931,
+      "step": 52370
+    },
+    {
+      "epoch": 9.86,
+      "grad_norm": 6.390880584716797,
+      "learning_rate": 2.823263692828911e-07,
+      "loss": 0.3672,
+      "step": 52380
+    },
+    {
+      "epoch": 9.86,
+      "grad_norm": 29.264728546142578,
+      "learning_rate": 2.785620176924525e-07,
+      "loss": 0.2607,
+      "step": 52390
+    },
+    {
+      "epoch": 9.86,
+      "grad_norm": 10.851874351501465,
+      "learning_rate": 2.7479766610201395e-07,
+      "loss": 0.6167,
+      "step": 52400
+    },
+    {
+      "epoch": 9.86,
+      "grad_norm": 23.39576530456543,
+      "learning_rate": 2.710333145115754e-07,
+      "loss": 0.4207,
+      "step": 52410
+    },
+    {
+      "epoch": 9.87,
+      "grad_norm": 20.70049476623535,
+      "learning_rate": 2.6726896292113687e-07,
+      "loss": 0.502,
+      "step": 52420
+    },
+    {
+      "epoch": 9.87,
+      "grad_norm": 12.70500373840332,
+      "learning_rate": 2.635046113306983e-07,
+      "loss": 0.4077,
+      "step": 52430
+    },
+    {
+      "epoch": 9.87,
+      "grad_norm": 0.4478881359100342,
+      "learning_rate": 2.597402597402598e-07,
+      "loss": 0.184,
+      "step": 52440
+    },
+    {
+      "epoch": 9.87,
+      "grad_norm": 8.121723175048828,
+      "learning_rate": 2.559759081498212e-07,
+      "loss": 0.2444,
+      "step": 52450
+    },
+    {
+      "epoch": 9.87,
+      "grad_norm": 0.5330722332000732,
+      "learning_rate": 2.5221155655938266e-07,
+      "loss": 0.4361,
+      "step": 52460
+    },
+    {
+      "epoch": 9.88,
+      "grad_norm": 31.016155242919922,
+      "learning_rate": 2.484472049689441e-07,
+      "loss": 0.5419,
+      "step": 52470
+    },
+    {
+      "epoch": 9.88,
+      "grad_norm": 7.886919975280762,
+      "learning_rate": 2.446828533785056e-07,
+      "loss": 0.4408,
+      "step": 52480
+    },
+    {
+      "epoch": 9.88,
+      "grad_norm": 29.892057418823242,
+      "learning_rate": 2.4091850178806704e-07,
+      "loss": 0.4313,
+      "step": 52490
+    },
+    {
+      "epoch": 9.88,
+      "grad_norm": 0.040985073894262314,
+      "learning_rate": 2.3715415019762845e-07,
+      "loss": 0.3838,
+      "step": 52500
+    },
+    {
+      "epoch": 9.88,
+      "grad_norm": 20.565820693969727,
+      "learning_rate": 2.3338979860718994e-07,
+      "loss": 0.5992,
+      "step": 52510
+    },
+    {
+      "epoch": 9.89,
+      "grad_norm": 9.025721549987793,
+      "learning_rate": 2.2962544701675137e-07,
+      "loss": 0.3285,
+      "step": 52520
+    },
+    {
+      "epoch": 9.89,
+      "grad_norm": 18.053125381469727,
+      "learning_rate": 2.258610954263128e-07,
+      "loss": 0.4098,
+      "step": 52530
+    },
+    {
+      "epoch": 9.89,
+      "grad_norm": 20.12868309020996,
+      "learning_rate": 2.220967438358743e-07,
+      "loss": 0.5116,
+      "step": 52540
+    },
+    {
+      "epoch": 9.89,
+      "grad_norm": 18.908733367919922,
+      "learning_rate": 2.1833239224543573e-07,
+      "loss": 0.5385,
+      "step": 52550
+    },
+    {
+      "epoch": 9.89,
+      "grad_norm": 1.1563811302185059,
+      "learning_rate": 2.145680406549972e-07,
+      "loss": 0.5077,
+      "step": 52560
+    },
+    {
+      "epoch": 9.89,
+      "grad_norm": 4.780964374542236,
+      "learning_rate": 2.1080368906455865e-07,
+      "loss": 0.4955,
+      "step": 52570
+    },
+    {
+      "epoch": 9.9,
+      "grad_norm": 9.796252250671387,
+      "learning_rate": 2.0703933747412008e-07,
+      "loss": 0.4149,
+      "step": 52580
+    },
+    {
+      "epoch": 9.9,
+      "grad_norm": 0.043842703104019165,
+      "learning_rate": 2.0327498588368155e-07,
+      "loss": 0.2674,
+      "step": 52590
+    },
+    {
+      "epoch": 9.9,
+      "grad_norm": 17.572214126586914,
+      "learning_rate": 1.99510634293243e-07,
+      "loss": 0.3753,
+      "step": 52600
+    },
+    {
+      "epoch": 9.9,
+      "grad_norm": 1.6240875720977783,
+      "learning_rate": 1.9574628270280447e-07,
+      "loss": 0.2265,
+      "step": 52610
+    },
+    {
+      "epoch": 9.9,
+      "grad_norm": 23.174564361572266,
+      "learning_rate": 1.919819311123659e-07,
+      "loss": 0.7401,
+      "step": 52620
+    },
+    {
+      "epoch": 9.91,
+      "grad_norm": 6.317526817321777,
+      "learning_rate": 1.8821757952192736e-07,
+      "loss": 0.5374,
+      "step": 52630
+    },
+    {
+      "epoch": 9.91,
+      "grad_norm": 15.101905822753906,
+      "learning_rate": 1.8445322793148882e-07,
+      "loss": 0.3169,
+      "step": 52640
+    },
+    {
+      "epoch": 9.91,
+      "grad_norm": 12.77408218383789,
+      "learning_rate": 1.8068887634105026e-07,
+      "loss": 0.6194,
+      "step": 52650
+    },
+    {
+      "epoch": 9.91,
+      "grad_norm": 7.3459248542785645,
+      "learning_rate": 1.7692452475061174e-07,
+      "loss": 0.2904,
+      "step": 52660
+    },
+    {
+      "epoch": 9.91,
+      "grad_norm": 12.190526962280273,
+      "learning_rate": 1.7316017316017318e-07,
+      "loss": 0.4855,
+      "step": 52670
+    },
+    {
+      "epoch": 9.92,
+      "grad_norm": 7.406374931335449,
+      "learning_rate": 1.693958215697346e-07,
+      "loss": 0.2343,
+      "step": 52680
+    },
+    {
+      "epoch": 9.92,
+      "grad_norm": 0.23926931619644165,
+      "learning_rate": 1.656314699792961e-07,
+      "loss": 0.2742,
+      "step": 52690
+    },
+    {
+      "epoch": 9.92,
+      "grad_norm": 2.192960739135742,
+      "learning_rate": 1.6186711838885753e-07,
+      "loss": 0.3677,
+      "step": 52700
+    },
+    {
+      "epoch": 9.92,
+      "grad_norm": 12.674443244934082,
+      "learning_rate": 1.5810276679841897e-07,
+      "loss": 0.4027,
+      "step": 52710
+    },
+    {
+      "epoch": 9.92,
+      "grad_norm": 11.900289535522461,
+      "learning_rate": 1.5433841520798043e-07,
+      "loss": 0.3339,
+      "step": 52720
+    },
+    {
+      "epoch": 9.92,
+      "grad_norm": 14.372079849243164,
+      "learning_rate": 1.505740636175419e-07,
+      "loss": 0.5075,
+      "step": 52730
+    },
+    {
+      "epoch": 9.93,
+      "grad_norm": 2.849031925201416,
+      "learning_rate": 1.4680971202710335e-07,
+      "loss": 0.1418,
+      "step": 52740
+    },
+    {
+      "epoch": 9.93,
+      "grad_norm": 5.773308753967285,
+      "learning_rate": 1.4304536043666478e-07,
+      "loss": 0.4116,
+      "step": 52750
+    },
+    {
+      "epoch": 9.93,
+      "grad_norm": 10.172654151916504,
+      "learning_rate": 1.3928100884622625e-07,
+      "loss": 0.4328,
+      "step": 52760
+    },
+    {
+      "epoch": 9.93,
+      "grad_norm": 13.555858612060547,
+      "learning_rate": 1.355166572557877e-07,
+      "loss": 0.4646,
+      "step": 52770
+    },
+    {
+      "epoch": 9.93,
+      "grad_norm": 7.2199177742004395,
+      "learning_rate": 1.3175230566534914e-07,
+      "loss": 0.4224,
+      "step": 52780
+    },
+    {
+      "epoch": 9.94,
+      "grad_norm": 21.8734130859375,
+      "learning_rate": 1.279879540749106e-07,
+      "loss": 0.3808,
+      "step": 52790
+    },
+    {
+      "epoch": 9.94,
+      "grad_norm": 16.765634536743164,
+      "learning_rate": 1.2422360248447206e-07,
+      "loss": 0.615,
+      "step": 52800
+    },
+    {
+      "epoch": 9.94,
+      "grad_norm": 13.97728157043457,
+      "learning_rate": 1.2045925089403352e-07,
+      "loss": 0.516,
+      "step": 52810
+    },
+    {
+      "epoch": 9.94,
+      "grad_norm": 17.42995834350586,
+      "learning_rate": 1.1669489930359497e-07,
+      "loss": 0.3234,
+      "step": 52820
+    },
+    {
+      "epoch": 9.94,
+      "grad_norm": 0.2923338711261749,
+      "learning_rate": 1.129305477131564e-07,
+      "loss": 0.2994,
+      "step": 52830
+    },
+    {
+      "epoch": 9.95,
+      "grad_norm": 7.7803497314453125,
+      "learning_rate": 1.0916619612271786e-07,
+      "loss": 0.3944,
+      "step": 52840
+    },
+    {
+      "epoch": 9.95,
+      "grad_norm": 27.952133178710938,
+      "learning_rate": 1.0540184453227933e-07,
+      "loss": 0.3718,
+      "step": 52850
+    },
+    {
+      "epoch": 9.95,
+      "grad_norm": 34.63023376464844,
+      "learning_rate": 1.0163749294184077e-07,
+      "loss": 0.7521,
+      "step": 52860
+    },
+    {
+      "epoch": 9.95,
+      "grad_norm": 24.050615310668945,
+      "learning_rate": 9.787314135140223e-08,
+      "loss": 0.4687,
+      "step": 52870
+    },
+    {
+      "epoch": 9.95,
+      "grad_norm": 0.22404804825782776,
+      "learning_rate": 9.410878976096368e-08,
+      "loss": 0.3324,
+      "step": 52880
+    },
+    {
+      "epoch": 9.95,
+      "grad_norm": 8.329122543334961,
+      "learning_rate": 9.034443817052513e-08,
+      "loss": 0.5473,
+      "step": 52890
+    },
+    {
+      "epoch": 9.96,
+      "grad_norm": 2.2988033294677734,
+      "learning_rate": 8.658008658008659e-08,
+      "loss": 0.1834,
+      "step": 52900
+    },
+    {
+      "epoch": 9.96,
+      "grad_norm": 16.856712341308594,
+      "learning_rate": 8.281573498964805e-08,
+      "loss": 0.3419,
+      "step": 52910
+    },
+    {
+      "epoch": 9.96,
+      "grad_norm": 22.70846176147461,
+      "learning_rate": 7.905138339920948e-08,
+      "loss": 0.4666,
+      "step": 52920
+    },
+    {
+      "epoch": 9.96,
+      "grad_norm": 8.477212905883789,
+      "learning_rate": 7.528703180877094e-08,
+      "loss": 0.1898,
+      "step": 52930
+    },
+    {
+      "epoch": 9.96,
+      "grad_norm": 7.554108619689941,
+      "learning_rate": 7.152268021833239e-08,
+      "loss": 0.4636,
+      "step": 52940
+    },
+    {
+      "epoch": 9.97,
+      "grad_norm": 35.77024841308594,
+      "learning_rate": 6.775832862789385e-08,
+      "loss": 0.5227,
+      "step": 52950
+    },
+    {
+      "epoch": 9.97,
+      "grad_norm": 0.06595637649297714,
+      "learning_rate": 6.39939770374553e-08,
+      "loss": 0.3577,
+      "step": 52960
+    },
+    {
+      "epoch": 9.97,
+      "grad_norm": 19.368412017822266,
+      "learning_rate": 6.022962544701676e-08,
+      "loss": 0.5107,
+      "step": 52970
+    },
+    {
+      "epoch": 9.97,
+      "grad_norm": 0.05418463796377182,
+      "learning_rate": 5.64652738565782e-08,
+      "loss": 0.4543,
+      "step": 52980
+    },
+    {
+      "epoch": 9.97,
+      "grad_norm": 10.112632751464844,
+      "learning_rate": 5.270092226613966e-08,
+      "loss": 0.6876,
+      "step": 52990
+    },
+    {
+      "epoch": 9.98,
+      "grad_norm": 22.078157424926758,
+      "learning_rate": 4.893657067570112e-08,
+      "loss": 0.505,
+      "step": 53000
+    },
+    {
+      "epoch": 9.98,
+      "grad_norm": 7.689653396606445,
+      "learning_rate": 4.5172219085262564e-08,
+      "loss": 0.4111,
+      "step": 53010
+    },
+    {
+      "epoch": 9.98,
+      "grad_norm": 17.780681610107422,
+      "learning_rate": 4.1407867494824025e-08,
+      "loss": 0.6067,
+      "step": 53020
+    },
+    {
+      "epoch": 9.98,
+      "grad_norm": 54.270999908447266,
+      "learning_rate": 3.764351590438547e-08,
+      "loss": 0.3095,
+      "step": 53030
+    },
+    {
+      "epoch": 9.98,
+      "grad_norm": 12.391640663146973,
+      "learning_rate": 3.3879164313946926e-08,
+      "loss": 0.6079,
+      "step": 53040
+    },
+    {
+      "epoch": 9.98,
+      "grad_norm": 15.050522804260254,
+      "learning_rate": 3.011481272350838e-08,
+      "loss": 0.6605,
+      "step": 53050
+    },
+    {
+      "epoch": 9.99,
+      "grad_norm": 10.04753303527832,
+      "learning_rate": 2.635046113306983e-08,
+      "loss": 0.5057,
+      "step": 53060
+    },
+    {
+      "epoch": 9.99,
+      "grad_norm": 14.447115898132324,
+      "learning_rate": 2.2586109542631282e-08,
+      "loss": 0.5889,
+      "step": 53070
+    },
+    {
+      "epoch": 9.99,
+      "grad_norm": 15.600353240966797,
+      "learning_rate": 1.8821757952192736e-08,
+      "loss": 0.3044,
+      "step": 53080
+    },
+    {
+      "epoch": 9.99,
+      "grad_norm": 41.5399284362793,
+      "learning_rate": 1.505740636175419e-08,
+      "loss": 0.3476,
+      "step": 53090
+    },
+    {
+      "epoch": 9.99,
+      "grad_norm": 0.05904613435268402,
+      "learning_rate": 1.1293054771315641e-08,
+      "loss": 0.3915,
+      "step": 53100
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.6749778389930725,
+      "learning_rate": 7.528703180877095e-09,
+      "loss": 0.4269,
+      "step": 53110
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.3194694519042969,
+      "learning_rate": 3.7643515904385476e-09,
+      "loss": 0.2406,
+      "step": 53120
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.023369356989860535,
+      "learning_rate": 0.0,
+      "loss": 0.3274,
+      "step": 53130
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.9269333333333334,
+      "eval_loss": 0.3031522035598755,
+      "eval_runtime": 53.4027,
+      "eval_samples_per_second": 140.442,
+      "eval_steps_per_second": 17.565,
+      "step": 53130
+    },
+    {
+      "epoch": 10.0,
+      "step": 53130,
+      "total_flos": 3.29630230185984e+19,
+      "train_loss": 0.8302146609071148,
+      "train_runtime": 9599.828,
+      "train_samples_per_second": 44.272,
+      "train_steps_per_second": 5.534
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 53130,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 500,
+  "total_flos": 3.29630230185984e+19,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}