diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,10849 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.7806973218797373,
+  "eval_steps": 500,
+  "global_step": 1545,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005053057099545225,
+      "grad_norm": 1.6384633779525757,
+      "learning_rate": 0.0,
+      "loss": 0.4704,
+      "step": 1
+    },
+    {
+      "epoch": 0.001010611419909045,
+      "grad_norm": 2.194593906402588,
+      "learning_rate": 6.41025641025641e-07,
+      "loss": 0.4286,
+      "step": 2
+    },
+    {
+      "epoch": 0.0015159171298635675,
+      "grad_norm": 3.6220381259918213,
+      "learning_rate": 1.282051282051282e-06,
+      "loss": 0.8297,
+      "step": 3
+    },
+    {
+      "epoch": 0.00202122283981809,
+      "grad_norm": 1.929093837738037,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 0.6096,
+      "step": 4
+    },
+    {
+      "epoch": 0.0025265285497726125,
+      "grad_norm": 3.1244475841522217,
+      "learning_rate": 2.564102564102564e-06,
+      "loss": 0.8738,
+      "step": 5
+    },
+    {
+      "epoch": 0.003031834259727135,
+      "grad_norm": 3.0733749866485596,
+      "learning_rate": 3.205128205128205e-06,
+      "loss": 0.668,
+      "step": 6
+    },
+    {
+      "epoch": 0.0035371399696816574,
+      "grad_norm": 1.7501100301742554,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 0.5054,
+      "step": 7
+    },
+    {
+      "epoch": 0.00404244567963618,
+      "grad_norm": 2.8912413120269775,
+      "learning_rate": 4.487179487179488e-06,
+      "loss": 0.8734,
+      "step": 8
+    },
+    {
+      "epoch": 0.004547751389590703,
+      "grad_norm": 1.4918310642242432,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 0.8562,
+      "step": 9
+    },
+    {
+      "epoch": 0.005053057099545225,
+      "grad_norm": 2.0631442070007324,
+      "learning_rate": 5.76923076923077e-06,
+      "loss": 0.7117,
+      "step": 10
+    },
+    {
+      "epoch": 0.0055583628094997475,
+      "grad_norm": 1.4029316902160645,
+      "learning_rate": 6.41025641025641e-06,
+      "loss": 0.5705,
+      "step": 11
+    },
+    {
+      "epoch": 0.00606366851945427,
+      "grad_norm": 1.4444737434387207,
+      "learning_rate": 7.051282051282052e-06,
+      "loss": 0.661,
+      "step": 12
+    },
+    {
+      "epoch": 0.006568974229408792,
+      "grad_norm": 1.2494786977767944,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 0.7558,
+      "step": 13
+    },
+    {
+      "epoch": 0.007074279939363315,
+      "grad_norm": 1.3939992189407349,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.5186,
+      "step": 14
+    },
+    {
+      "epoch": 0.007579585649317837,
+      "grad_norm": 1.0526946783065796,
+      "learning_rate": 8.974358974358976e-06,
+      "loss": 0.543,
+      "step": 15
+    },
+    {
+      "epoch": 0.00808489135927236,
+      "grad_norm": 0.7698296904563904,
+      "learning_rate": 9.615384615384616e-06,
+      "loss": 0.366,
+      "step": 16
+    },
+    {
+      "epoch": 0.008590197069226882,
+      "grad_norm": 1.036993145942688,
+      "learning_rate": 1.0256410256410256e-05,
+      "loss": 0.4858,
+      "step": 17
+    },
+    {
+      "epoch": 0.009095502779181405,
+      "grad_norm": 1.1180205345153809,
+      "learning_rate": 1.0897435897435898e-05,
+      "loss": 0.5841,
+      "step": 18
+    },
+    {
+      "epoch": 0.009600808489135927,
+      "grad_norm": 0.8241035342216492,
+      "learning_rate": 1.153846153846154e-05,
+      "loss": 0.5563,
+      "step": 19
+    },
+    {
+      "epoch": 0.01010611419909045,
+      "grad_norm": 0.6611151695251465,
+      "learning_rate": 1.217948717948718e-05,
+      "loss": 0.3478,
+      "step": 20
+    },
+    {
+      "epoch": 0.010611419909044972,
+      "grad_norm": 0.4365403950214386,
+      "learning_rate": 1.282051282051282e-05,
+      "loss": 0.2442,
+      "step": 21
+    },
+    {
+      "epoch": 0.011116725618999495,
+      "grad_norm": 0.7351874709129333,
+      "learning_rate": 1.3461538461538462e-05,
+      "loss": 0.3456,
+      "step": 22
+    },
+    {
+      "epoch": 0.011622031328954016,
+      "grad_norm": 0.55842524766922,
+      "learning_rate": 1.4102564102564104e-05,
+      "loss": 0.3809,
+      "step": 23
+    },
+    {
+      "epoch": 0.01212733703890854,
+      "grad_norm": 0.9409593343734741,
+      "learning_rate": 1.4743589743589745e-05,
+      "loss": 0.4769,
+      "step": 24
+    },
+    {
+      "epoch": 0.012632642748863061,
+      "grad_norm": 0.803429901599884,
+      "learning_rate": 1.5384615384615387e-05,
+      "loss": 0.5099,
+      "step": 25
+    },
+    {
+      "epoch": 0.013137948458817585,
+      "grad_norm": 0.9147859215736389,
+      "learning_rate": 1.602564102564103e-05,
+      "loss": 0.5725,
+      "step": 26
+    },
+    {
+      "epoch": 0.013643254168772108,
+      "grad_norm": 0.6797801852226257,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.3108,
+      "step": 27
+    },
+    {
+      "epoch": 0.01414855987872663,
+      "grad_norm": 0.6329852342605591,
+      "learning_rate": 1.730769230769231e-05,
+      "loss": 0.4665,
+      "step": 28
+    },
+    {
+      "epoch": 0.014653865588681153,
+      "grad_norm": 0.49528101086616516,
+      "learning_rate": 1.794871794871795e-05,
+      "loss": 0.3252,
+      "step": 29
+    },
+    {
+      "epoch": 0.015159171298635674,
+      "grad_norm": 0.6480412483215332,
+      "learning_rate": 1.858974358974359e-05,
+      "loss": 0.4001,
+      "step": 30
+    },
+    {
+      "epoch": 0.015664477008590198,
+      "grad_norm": 1.0041990280151367,
+      "learning_rate": 1.923076923076923e-05,
+      "loss": 0.635,
+      "step": 31
+    },
+    {
+      "epoch": 0.01616978271854472,
+      "grad_norm": 0.7632623314857483,
+      "learning_rate": 1.987179487179487e-05,
+      "loss": 0.3599,
+      "step": 32
+    },
+    {
+      "epoch": 0.01667508842849924,
+      "grad_norm": 0.8572860360145569,
+      "learning_rate": 2.0512820512820512e-05,
+      "loss": 0.4534,
+      "step": 33
+    },
+    {
+      "epoch": 0.017180394138453764,
+      "grad_norm": 0.9279261231422424,
+      "learning_rate": 2.1153846153846154e-05,
+      "loss": 0.4895,
+      "step": 34
+    },
+    {
+      "epoch": 0.017685699848408287,
+      "grad_norm": 0.5363617539405823,
+      "learning_rate": 2.1794871794871795e-05,
+      "loss": 0.284,
+      "step": 35
+    },
+    {
+      "epoch": 0.01819100555836281,
+      "grad_norm": 0.6156304478645325,
+      "learning_rate": 2.2435897435897437e-05,
+      "loss": 0.3305,
+      "step": 36
+    },
+    {
+      "epoch": 0.01869631126831733,
+      "grad_norm": 0.8967059850692749,
+      "learning_rate": 2.307692307692308e-05,
+      "loss": 0.5963,
+      "step": 37
+    },
+    {
+      "epoch": 0.019201616978271854,
+      "grad_norm": 0.8374812006950378,
+      "learning_rate": 2.3717948717948718e-05,
+      "loss": 0.3155,
+      "step": 38
+    },
+    {
+      "epoch": 0.019706922688226377,
+      "grad_norm": 0.7348405718803406,
+      "learning_rate": 2.435897435897436e-05,
+      "loss": 0.3585,
+      "step": 39
+    },
+    {
+      "epoch": 0.0202122283981809,
+      "grad_norm": 0.7170049548149109,
+      "learning_rate": 2.5e-05,
+      "loss": 0.3285,
+      "step": 40
+    },
+    {
+      "epoch": 0.020717534108135423,
+      "grad_norm": 0.7275897860527039,
+      "learning_rate": 2.564102564102564e-05,
+      "loss": 0.3612,
+      "step": 41
+    },
+    {
+      "epoch": 0.021222839818089943,
+      "grad_norm": 0.7469742298126221,
+      "learning_rate": 2.6282051282051285e-05,
+      "loss": 0.3367,
+      "step": 42
+    },
+    {
+      "epoch": 0.021728145528044467,
+      "grad_norm": 0.6935989856719971,
+      "learning_rate": 2.6923076923076923e-05,
+      "loss": 0.2455,
+      "step": 43
+    },
+    {
+      "epoch": 0.02223345123799899,
+      "grad_norm": 0.7077957391738892,
+      "learning_rate": 2.756410256410257e-05,
+      "loss": 0.4161,
+      "step": 44
+    },
+    {
+      "epoch": 0.022738756947953513,
+      "grad_norm": 0.38968655467033386,
+      "learning_rate": 2.8205128205128207e-05,
+      "loss": 0.1841,
+      "step": 45
+    },
+    {
+      "epoch": 0.023244062657908033,
+      "grad_norm": 0.9259281158447266,
+      "learning_rate": 2.8846153846153845e-05,
+      "loss": 0.4498,
+      "step": 46
+    },
+    {
+      "epoch": 0.023749368367862556,
+      "grad_norm": 1.0964964628219604,
+      "learning_rate": 2.948717948717949e-05,
+      "loss": 0.5429,
+      "step": 47
+    },
+    {
+      "epoch": 0.02425467407781708,
+      "grad_norm": 0.4879132807254791,
+      "learning_rate": 3.012820512820513e-05,
+      "loss": 0.2142,
+      "step": 48
+    },
+    {
+      "epoch": 0.024759979787771603,
+      "grad_norm": 0.6453592777252197,
+      "learning_rate": 3.0769230769230774e-05,
+      "loss": 0.2836,
+      "step": 49
+    },
+    {
+      "epoch": 0.025265285497726123,
+      "grad_norm": 0.5058504343032837,
+      "learning_rate": 3.141025641025641e-05,
+      "loss": 0.2157,
+      "step": 50
+    },
+    {
+      "epoch": 0.025770591207680646,
+      "grad_norm": 0.8575659990310669,
+      "learning_rate": 3.205128205128206e-05,
+      "loss": 0.6295,
+      "step": 51
+    },
+    {
+      "epoch": 0.02627589691763517,
+      "grad_norm": 0.48030632734298706,
+      "learning_rate": 3.269230769230769e-05,
+      "loss": 0.2874,
+      "step": 52
+    },
+    {
+      "epoch": 0.026781202627589692,
+      "grad_norm": 0.8366153240203857,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.4651,
+      "step": 53
+    },
+    {
+      "epoch": 0.027286508337544216,
+      "grad_norm": 0.5325554609298706,
+      "learning_rate": 3.397435897435898e-05,
+      "loss": 0.2285,
+      "step": 54
+    },
+    {
+      "epoch": 0.027791814047498736,
+      "grad_norm": 0.764580488204956,
+      "learning_rate": 3.461538461538462e-05,
+      "loss": 0.428,
+      "step": 55
+    },
+    {
+      "epoch": 0.02829711975745326,
+      "grad_norm": 0.8604835867881775,
+      "learning_rate": 3.525641025641026e-05,
+      "loss": 0.5092,
+      "step": 56
+    },
+    {
+      "epoch": 0.028802425467407782,
+      "grad_norm": 2.482994794845581,
+      "learning_rate": 3.58974358974359e-05,
+      "loss": 0.4473,
+      "step": 57
+    },
+    {
+      "epoch": 0.029307731177362305,
+      "grad_norm": 0.8809303045272827,
+      "learning_rate": 3.653846153846154e-05,
+      "loss": 0.5762,
+      "step": 58
+    },
+    {
+      "epoch": 0.029813036887316825,
+      "grad_norm": 0.43776047229766846,
+      "learning_rate": 3.717948717948718e-05,
+      "loss": 0.1808,
+      "step": 59
+    },
+    {
+      "epoch": 0.03031834259727135,
+      "grad_norm": 0.6836262941360474,
+      "learning_rate": 3.782051282051282e-05,
+      "loss": 0.2987,
+      "step": 60
+    },
+    {
+      "epoch": 0.030823648307225872,
+      "grad_norm": 0.5660818815231323,
+      "learning_rate": 3.846153846153846e-05,
+      "loss": 0.2198,
+      "step": 61
+    },
+    {
+      "epoch": 0.031328954017180395,
+      "grad_norm": 0.4436876177787781,
+      "learning_rate": 3.9102564102564105e-05,
+      "loss": 0.1963,
+      "step": 62
+    },
+    {
+      "epoch": 0.03183425972713492,
+      "grad_norm": 0.7230072617530823,
+      "learning_rate": 3.974358974358974e-05,
+      "loss": 0.4539,
+      "step": 63
+    },
+    {
+      "epoch": 0.03233956543708944,
+      "grad_norm": 0.9129766225814819,
+      "learning_rate": 4.038461538461539e-05,
+      "loss": 0.3737,
+      "step": 64
+    },
+    {
+      "epoch": 0.032844871147043965,
+      "grad_norm": 1.1869587898254395,
+      "learning_rate": 4.1025641025641023e-05,
+      "loss": 0.696,
+      "step": 65
+    },
+    {
+      "epoch": 0.03335017685699848,
+      "grad_norm": 0.7398460507392883,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.5932,
+      "step": 66
+    },
+    {
+      "epoch": 0.033855482566953005,
+      "grad_norm": 0.7067481875419617,
+      "learning_rate": 4.230769230769231e-05,
+      "loss": 0.3213,
+      "step": 67
+    },
+    {
+      "epoch": 0.03436078827690753,
+      "grad_norm": 0.9902649521827698,
+      "learning_rate": 4.294871794871795e-05,
+      "loss": 0.416,
+      "step": 68
+    },
+    {
+      "epoch": 0.03486609398686205,
+      "grad_norm": 0.42111849784851074,
+      "learning_rate": 4.358974358974359e-05,
+      "loss": 0.2465,
+      "step": 69
+    },
+    {
+      "epoch": 0.035371399696816574,
+      "grad_norm": 0.9634829759597778,
+      "learning_rate": 4.423076923076923e-05,
+      "loss": 0.5916,
+      "step": 70
+    },
+    {
+      "epoch": 0.0358767054067711,
+      "grad_norm": 0.8240818977355957,
+      "learning_rate": 4.4871794871794874e-05,
+      "loss": 0.4181,
+      "step": 71
+    },
+    {
+      "epoch": 0.03638201111672562,
+      "grad_norm": 0.7216759324073792,
+      "learning_rate": 4.5512820512820516e-05,
+      "loss": 0.4966,
+      "step": 72
+    },
+    {
+      "epoch": 0.036887316826680144,
+      "grad_norm": 0.7505428194999695,
+      "learning_rate": 4.615384615384616e-05,
+      "loss": 0.3799,
+      "step": 73
+    },
+    {
+      "epoch": 0.03739262253663466,
+      "grad_norm": 1.2560772895812988,
+      "learning_rate": 4.67948717948718e-05,
+      "loss": 0.4691,
+      "step": 74
+    },
+    {
+      "epoch": 0.037897928246589184,
+      "grad_norm": 0.7000345587730408,
+      "learning_rate": 4.7435897435897435e-05,
+      "loss": 0.2898,
+      "step": 75
+    },
+    {
+      "epoch": 0.03840323395654371,
+      "grad_norm": 0.6096041202545166,
+      "learning_rate": 4.8076923076923084e-05,
+      "loss": 0.2929,
+      "step": 76
+    },
+    {
+      "epoch": 0.03890853966649823,
+      "grad_norm": 0.6045755743980408,
+      "learning_rate": 4.871794871794872e-05,
+      "loss": 0.282,
+      "step": 77
+    },
+    {
+      "epoch": 0.039413845376452754,
+      "grad_norm": 0.9146702885627747,
+      "learning_rate": 4.935897435897436e-05,
+      "loss": 0.4637,
+      "step": 78
+    },
+    {
+      "epoch": 0.03991915108640728,
+      "grad_norm": 0.9550428986549377,
+      "learning_rate": 5e-05,
+      "loss": 0.3772,
+      "step": 79
+    },
+    {
+      "epoch": 0.0404244567963618,
+      "grad_norm": 0.7160158157348633,
+      "learning_rate": 4.999998018159948e-05,
+      "loss": 0.3099,
+      "step": 80
+    },
+    {
+      "epoch": 0.040929762506316324,
+      "grad_norm": 0.6506277322769165,
+      "learning_rate": 4.999992072642933e-05,
+      "loss": 0.3599,
+      "step": 81
+    },
+    {
+      "epoch": 0.04143506821627085,
+      "grad_norm": 0.8049689531326294,
+      "learning_rate": 4.999982163458383e-05,
+      "loss": 0.3184,
+      "step": 82
+    },
+    {
+      "epoch": 0.04194037392622536,
+      "grad_norm": 1.0008819103240967,
+      "learning_rate": 4.999968290622007e-05,
+      "loss": 0.4903,
+      "step": 83
+    },
+    {
+      "epoch": 0.04244567963617989,
+      "grad_norm": 0.6113778352737427,
+      "learning_rate": 4.999950454155801e-05,
+      "loss": 0.2561,
+      "step": 84
+    },
+    {
+      "epoch": 0.04295098534613441,
+      "grad_norm": 1.4032593965530396,
+      "learning_rate": 4.999928654088044e-05,
+      "loss": 0.3617,
+      "step": 85
+    },
+    {
+      "epoch": 0.04345629105608893,
+      "grad_norm": 0.5701761841773987,
+      "learning_rate": 4.9999028904532996e-05,
+      "loss": 0.3419,
+      "step": 86
+    },
+    {
+      "epoch": 0.043961596766043456,
+      "grad_norm": 0.8148670792579651,
+      "learning_rate": 4.9998731632924144e-05,
+      "loss": 0.2048,
+      "step": 87
+    },
+    {
+      "epoch": 0.04446690247599798,
+      "grad_norm": 1.7208861112594604,
+      "learning_rate": 4.999839472652522e-05,
+      "loss": 0.4322,
+      "step": 88
+    },
+    {
+      "epoch": 0.0449722081859525,
+      "grad_norm": 2.0720438957214355,
+      "learning_rate": 4.9998018185870356e-05,
+      "loss": 0.2536,
+      "step": 89
+    },
+    {
+      "epoch": 0.045477513895907026,
+      "grad_norm": 1.085567831993103,
+      "learning_rate": 4.999760201155657e-05,
+      "loss": 0.5578,
+      "step": 90
+    },
+    {
+      "epoch": 0.04598281960586155,
+      "grad_norm": 0.5517830848693848,
+      "learning_rate": 4.999714620424367e-05,
+      "loss": 0.2543,
+      "step": 91
+    },
+    {
+      "epoch": 0.046488125315816066,
+      "grad_norm": 0.6557161808013916,
+      "learning_rate": 4.999665076465434e-05,
+      "loss": 0.4033,
+      "step": 92
+    },
+    {
+      "epoch": 0.04699343102577059,
+      "grad_norm": 0.8143996000289917,
+      "learning_rate": 4.9996115693574086e-05,
+      "loss": 0.3873,
+      "step": 93
+    },
+    {
+      "epoch": 0.04749873673572511,
+      "grad_norm": 0.9482316970825195,
+      "learning_rate": 4.999554099185124e-05,
+      "loss": 0.3902,
+      "step": 94
+    },
+    {
+      "epoch": 0.048004042445679636,
+      "grad_norm": 0.6633082032203674,
+      "learning_rate": 4.999492666039699e-05,
+      "loss": 0.3041,
+      "step": 95
+    },
+    {
+      "epoch": 0.04850934815563416,
+      "grad_norm": 1.107414960861206,
+      "learning_rate": 4.999427270018533e-05,
+      "loss": 0.4195,
+      "step": 96
+    },
+    {
+      "epoch": 0.04901465386558868,
+      "grad_norm": 0.4633142948150635,
+      "learning_rate": 4.99935791122531e-05,
+      "loss": 0.22,
+      "step": 97
+    },
+    {
+      "epoch": 0.049519959575543206,
+      "grad_norm": 0.7195435166358948,
+      "learning_rate": 4.999284589769996e-05,
+      "loss": 0.2776,
+      "step": 98
+    },
+    {
+      "epoch": 0.05002526528549773,
+      "grad_norm": 0.38064494729042053,
+      "learning_rate": 4.9992073057688403e-05,
+      "loss": 0.2126,
+      "step": 99
+    },
+    {
+      "epoch": 0.050530570995452245,
+      "grad_norm": 0.4273965060710907,
+      "learning_rate": 4.9991260593443745e-05,
+      "loss": 0.2464,
+      "step": 100
+    },
+    {
+      "epoch": 0.05103587670540677,
+      "grad_norm": 0.7326411008834839,
+      "learning_rate": 4.9990408506254134e-05,
+      "loss": 0.3473,
+      "step": 101
+    },
+    {
+      "epoch": 0.05154118241536129,
+      "grad_norm": 0.6799935102462769,
+      "learning_rate": 4.998951679747051e-05,
+      "loss": 0.3258,
+      "step": 102
+    },
+    {
+      "epoch": 0.052046488125315815,
+      "grad_norm": 0.6898297071456909,
+      "learning_rate": 4.998858546850668e-05,
+      "loss": 0.6305,
+      "step": 103
+    },
+    {
+      "epoch": 0.05255179383527034,
+      "grad_norm": 0.5631976127624512,
+      "learning_rate": 4.998761452083922e-05,
+      "loss": 0.3241,
+      "step": 104
+    },
+    {
+      "epoch": 0.05305709954522486,
+      "grad_norm": 0.4945349395275116,
+      "learning_rate": 4.998660395600754e-05,
+      "loss": 0.3002,
+      "step": 105
+    },
+    {
+      "epoch": 0.053562405255179385,
+      "grad_norm": 0.6329333186149597,
+      "learning_rate": 4.998555377561387e-05,
+      "loss": 0.3074,
+      "step": 106
+    },
+    {
+      "epoch": 0.05406771096513391,
+      "grad_norm": 0.532968282699585,
+      "learning_rate": 4.9984463981323246e-05,
+      "loss": 0.334,
+      "step": 107
+    },
+    {
+      "epoch": 0.05457301667508843,
+      "grad_norm": 0.7290483713150024,
+      "learning_rate": 4.998333457486349e-05,
+      "loss": 0.2753,
+      "step": 108
+    },
+    {
+      "epoch": 0.05507832238504295,
+      "grad_norm": 0.7623900771141052,
+      "learning_rate": 4.998216555802526e-05,
+      "loss": 0.4656,
+      "step": 109
+    },
+    {
+      "epoch": 0.05558362809499747,
+      "grad_norm": 0.6192227005958557,
+      "learning_rate": 4.9980956932661996e-05,
+      "loss": 0.258,
+      "step": 110
+    },
+    {
+      "epoch": 0.056088933804951994,
+      "grad_norm": 1.0014296770095825,
+      "learning_rate": 4.9979708700689934e-05,
+      "loss": 0.5037,
+      "step": 111
+    },
+    {
+      "epoch": 0.05659423951490652,
+      "grad_norm": 0.5563094615936279,
+      "learning_rate": 4.9978420864088115e-05,
+      "loss": 0.3379,
+      "step": 112
+    },
+    {
+      "epoch": 0.05709954522486104,
+      "grad_norm": 0.9516816139221191,
+      "learning_rate": 4.997709342489837e-05,
+      "loss": 0.4447,
+      "step": 113
+    },
+    {
+      "epoch": 0.057604850934815564,
+      "grad_norm": 0.7699498534202576,
+      "learning_rate": 4.997572638522531e-05,
+      "loss": 0.5482,
+      "step": 114
+    },
+    {
+      "epoch": 0.05811015664477009,
+      "grad_norm": 0.623145580291748,
+      "learning_rate": 4.9974319747236344e-05,
+      "loss": 0.3454,
+      "step": 115
+    },
+    {
+      "epoch": 0.05861546235472461,
+      "grad_norm": 0.62673419713974,
+      "learning_rate": 4.997287351316165e-05,
+      "loss": 0.4063,
+      "step": 116
+    },
+    {
+      "epoch": 0.059120768064679134,
+      "grad_norm": 0.6477133631706238,
+      "learning_rate": 4.997138768529421e-05,
+      "loss": 0.2508,
+      "step": 117
+    },
+    {
+      "epoch": 0.05962607377463365,
+      "grad_norm": 0.650576114654541,
+      "learning_rate": 4.9969862265989745e-05,
+      "loss": 0.3157,
+      "step": 118
+    },
+    {
+      "epoch": 0.060131379484588174,
+      "grad_norm": 0.6956886053085327,
+      "learning_rate": 4.9968297257666764e-05,
+      "loss": 0.2984,
+      "step": 119
+    },
+    {
+      "epoch": 0.0606366851945427,
+      "grad_norm": 1.090057611465454,
+      "learning_rate": 4.996669266280655e-05,
+      "loss": 0.5217,
+      "step": 120
+    },
+    {
+      "epoch": 0.06114199090449722,
+      "grad_norm": 0.6649771928787231,
+      "learning_rate": 4.996504848395314e-05,
+      "loss": 0.4094,
+      "step": 121
+    },
+    {
+      "epoch": 0.061647296614451744,
+      "grad_norm": 1.0914398431777954,
+      "learning_rate": 4.9963364723713344e-05,
+      "loss": 0.7907,
+      "step": 122
+    },
+    {
+      "epoch": 0.06215260232440627,
+      "grad_norm": 1.1362191438674927,
+      "learning_rate": 4.9961641384756705e-05,
+      "loss": 0.5202,
+      "step": 123
+    },
+    {
+      "epoch": 0.06265790803436079,
+      "grad_norm": 0.6446903347969055,
+      "learning_rate": 4.9959878469815534e-05,
+      "loss": 0.5434,
+      "step": 124
+    },
+    {
+      "epoch": 0.06316321374431531,
+      "grad_norm": 0.5553262233734131,
+      "learning_rate": 4.9958075981684884e-05,
+      "loss": 0.2622,
+      "step": 125
+    },
+    {
+      "epoch": 0.06366851945426984,
+      "grad_norm": 0.7034339904785156,
+      "learning_rate": 4.995623392322254e-05,
+      "loss": 0.4785,
+      "step": 126
+    },
+    {
+      "epoch": 0.06417382516422436,
+      "grad_norm": 0.8382132649421692,
+      "learning_rate": 4.995435229734905e-05,
+      "loss": 0.3813,
+      "step": 127
+    },
+    {
+      "epoch": 0.06467913087417888,
+      "grad_norm": 0.8590944409370422,
+      "learning_rate": 4.995243110704766e-05,
+      "loss": 0.6019,
+      "step": 128
+    },
+    {
+      "epoch": 0.0651844365841334,
+      "grad_norm": 0.8058075904846191,
+      "learning_rate": 4.9950470355364386e-05,
+      "loss": 0.4251,
+      "step": 129
+    },
+    {
+      "epoch": 0.06568974229408793,
+      "grad_norm": 0.5873439908027649,
+      "learning_rate": 4.994847004540793e-05,
+      "loss": 0.4432,
+      "step": 130
+    },
+    {
+      "epoch": 0.06619504800404244,
+      "grad_norm": 0.6679672598838806,
+      "learning_rate": 4.994643018034973e-05,
+      "loss": 0.4053,
+      "step": 131
+    },
+    {
+      "epoch": 0.06670035371399696,
+      "grad_norm": 0.8931362628936768,
+      "learning_rate": 4.994435076342394e-05,
+      "loss": 0.5092,
+      "step": 132
+    },
+    {
+      "epoch": 0.06720565942395149,
+      "grad_norm": 0.4417133033275604,
+      "learning_rate": 4.994223179792741e-05,
+      "loss": 0.2055,
+      "step": 133
+    },
+    {
+      "epoch": 0.06771096513390601,
+      "grad_norm": 0.3722580671310425,
+      "learning_rate": 4.9940073287219703e-05,
+      "loss": 0.2183,
+      "step": 134
+    },
+    {
+      "epoch": 0.06821627084386053,
+      "grad_norm": 1.3628448247909546,
+      "learning_rate": 4.9937875234723075e-05,
+      "loss": 0.4615,
+      "step": 135
+    },
+    {
+      "epoch": 0.06872157655381506,
+      "grad_norm": 0.6365114450454712,
+      "learning_rate": 4.993563764392248e-05,
+      "loss": 0.3071,
+      "step": 136
+    },
+    {
+      "epoch": 0.06922688226376958,
+      "grad_norm": 0.5635626912117004,
+      "learning_rate": 4.9933360518365566e-05,
+      "loss": 0.3514,
+      "step": 137
+    },
+    {
+      "epoch": 0.0697321879737241,
+      "grad_norm": 0.7801100611686707,
+      "learning_rate": 4.993104386166263e-05,
+      "loss": 0.3682,
+      "step": 138
+    },
+    {
+      "epoch": 0.07023749368367863,
+      "grad_norm": 0.7181437611579895,
+      "learning_rate": 4.9928687677486685e-05,
+      "loss": 0.3864,
+      "step": 139
+    },
+    {
+      "epoch": 0.07074279939363315,
+      "grad_norm": 0.7251635193824768,
+      "learning_rate": 4.992629196957339e-05,
+      "loss": 0.4515,
+      "step": 140
+    },
+    {
+      "epoch": 0.07124810510358767,
+      "grad_norm": 0.45708200335502625,
+      "learning_rate": 4.992385674172107e-05,
+      "loss": 0.2097,
+      "step": 141
+    },
+    {
+      "epoch": 0.0717534108135422,
+      "grad_norm": 0.5553367733955383,
+      "learning_rate": 4.9921381997790705e-05,
+      "loss": 0.246,
+      "step": 142
+    },
+    {
+      "epoch": 0.07225871652349672,
+      "grad_norm": 0.6126506924629211,
+      "learning_rate": 4.991886774170594e-05,
+      "loss": 0.2903,
+      "step": 143
+    },
+    {
+      "epoch": 0.07276402223345124,
+      "grad_norm": 0.8143837451934814,
+      "learning_rate": 4.991631397745307e-05,
+      "loss": 0.3333,
+      "step": 144
+    },
+    {
+      "epoch": 0.07326932794340577,
+      "grad_norm": 1.035240888595581,
+      "learning_rate": 4.991372070908099e-05,
+      "loss": 0.6579,
+      "step": 145
+    },
+    {
+      "epoch": 0.07377463365336029,
+      "grad_norm": 0.3664172887802124,
+      "learning_rate": 4.9911087940701274e-05,
+      "loss": 0.1993,
+      "step": 146
+    },
+    {
+      "epoch": 0.07427993936331481,
+      "grad_norm": 0.42232248187065125,
+      "learning_rate": 4.9908415676488094e-05,
+      "loss": 0.211,
+      "step": 147
+    },
+    {
+      "epoch": 0.07478524507326932,
+      "grad_norm": 0.7374719381332397,
+      "learning_rate": 4.990570392067826e-05,
+      "loss": 0.2841,
+      "step": 148
+    },
+    {
+      "epoch": 0.07529055078322384,
+      "grad_norm": 0.7315521240234375,
+      "learning_rate": 4.990295267757117e-05,
+      "loss": 0.4779,
+      "step": 149
+    },
+    {
+      "epoch": 0.07579585649317837,
+      "grad_norm": 0.8291402459144592,
+      "learning_rate": 4.9900161951528854e-05,
+      "loss": 0.5315,
+      "step": 150
+    },
+    {
+      "epoch": 0.07630116220313289,
+      "grad_norm": 0.9879725575447083,
+      "learning_rate": 4.9897331746975926e-05,
+      "loss": 0.5184,
+      "step": 151
+    },
+    {
+      "epoch": 0.07680646791308741,
+      "grad_norm": 0.6654685139656067,
+      "learning_rate": 4.98944620683996e-05,
+      "loss": 0.4268,
+      "step": 152
+    },
+    {
+      "epoch": 0.07731177362304194,
+      "grad_norm": 0.6749838590621948,
+      "learning_rate": 4.989155292034966e-05,
+      "loss": 0.3266,
+      "step": 153
+    },
+    {
+      "epoch": 0.07781707933299646,
+      "grad_norm": 0.5096444487571716,
+      "learning_rate": 4.98886043074385e-05,
+      "loss": 0.4308,
+      "step": 154
+    },
+    {
+      "epoch": 0.07832238504295098,
+      "grad_norm": 0.5235171914100647,
+      "learning_rate": 4.988561623434105e-05,
+      "loss": 0.3358,
+      "step": 155
+    },
+    {
+      "epoch": 0.07882769075290551,
+      "grad_norm": 0.39268213510513306,
+      "learning_rate": 4.988258870579481e-05,
+      "loss": 0.2653,
+      "step": 156
+    },
+    {
+      "epoch": 0.07933299646286003,
+      "grad_norm": 0.6037962436676025,
+      "learning_rate": 4.987952172659985e-05,
+      "loss": 0.1633,
+      "step": 157
+    },
+    {
+      "epoch": 0.07983830217281455,
+      "grad_norm": 0.4751133918762207,
+      "learning_rate": 4.9876415301618787e-05,
+      "loss": 0.2796,
+      "step": 158
+    },
+    {
+      "epoch": 0.08034360788276908,
+      "grad_norm": 0.7282652854919434,
+      "learning_rate": 4.987326943577675e-05,
+      "loss": 0.5218,
+      "step": 159
+    },
+    {
+      "epoch": 0.0808489135927236,
+      "grad_norm": 0.42741066217422485,
+      "learning_rate": 4.987008413406144e-05,
+      "loss": 0.1733,
+      "step": 160
+    },
+    {
+      "epoch": 0.08135421930267812,
+      "grad_norm": 0.5225434899330139,
+      "learning_rate": 4.9866859401523056e-05,
+      "loss": 0.1923,
+      "step": 161
+    },
+    {
+      "epoch": 0.08185952501263265,
+      "grad_norm": 0.5940921902656555,
+      "learning_rate": 4.986359524327433e-05,
+      "loss": 0.2548,
+      "step": 162
+    },
+    {
+      "epoch": 0.08236483072258717,
+      "grad_norm": 1.0237605571746826,
+      "learning_rate": 4.986029166449048e-05,
+      "loss": 0.3913,
+      "step": 163
+    },
+    {
+      "epoch": 0.0828701364325417,
+      "grad_norm": 0.44926929473876953,
+      "learning_rate": 4.985694867040924e-05,
+      "loss": 0.2304,
+      "step": 164
+    },
+    {
+      "epoch": 0.0833754421424962,
+      "grad_norm": 0.968037486076355,
+      "learning_rate": 4.9853566266330855e-05,
+      "loss": 0.4391,
+      "step": 165
+    },
+    {
+      "epoch": 0.08388074785245073,
+      "grad_norm": 0.6649966835975647,
+      "learning_rate": 4.9850144457618e-05,
+      "loss": 0.2662,
+      "step": 166
+    },
+    {
+      "epoch": 0.08438605356240525,
+      "grad_norm": 0.5354177951812744,
+      "learning_rate": 4.984668324969588e-05,
+      "loss": 0.223,
+      "step": 167
+    },
+    {
+      "epoch": 0.08489135927235977,
+      "grad_norm": 0.9306359887123108,
+      "learning_rate": 4.9843182648052136e-05,
+      "loss": 0.5532,
+      "step": 168
+    },
+    {
+      "epoch": 0.0853966649823143,
+      "grad_norm": 0.2804405987262726,
+      "learning_rate": 4.983964265823687e-05,
+      "loss": 0.1795,
+      "step": 169
+    },
+    {
+      "epoch": 0.08590197069226882,
+      "grad_norm": 0.6046093702316284,
+      "learning_rate": 4.9836063285862636e-05,
+      "loss": 0.3467,
+      "step": 170
+    },
+    {
+      "epoch": 0.08640727640222334,
+      "grad_norm": 0.3996451199054718,
+      "learning_rate": 4.983244453660445e-05,
+      "loss": 0.2116,
+      "step": 171
+    },
+    {
+      "epoch": 0.08691258211217787,
+      "grad_norm": 0.5045155882835388,
+      "learning_rate": 4.98287864161997e-05,
+      "loss": 0.2028,
+      "step": 172
+    },
+    {
+      "epoch": 0.08741788782213239,
+      "grad_norm": 0.7425013780593872,
+      "learning_rate": 4.982508893044827e-05,
+      "loss": 0.4183,
+      "step": 173
+    },
+    {
+      "epoch": 0.08792319353208691,
+      "grad_norm": 0.4544399380683899,
+      "learning_rate": 4.98213520852124e-05,
+      "loss": 0.2321,
+      "step": 174
+    },
+    {
+      "epoch": 0.08842849924204144,
+      "grad_norm": 0.7836821675300598,
+      "learning_rate": 4.981757588641676e-05,
+      "loss": 0.4456,
+      "step": 175
+    },
+    {
+      "epoch": 0.08893380495199596,
+      "grad_norm": 0.24565193057060242,
+      "learning_rate": 4.981376034004841e-05,
+      "loss": 0.1439,
+      "step": 176
+    },
+    {
+      "epoch": 0.08943911066195048,
+      "grad_norm": 0.3137032687664032,
+      "learning_rate": 4.980990545215678e-05,
+      "loss": 0.1513,
+      "step": 177
+    },
+    {
+      "epoch": 0.089944416371905,
+      "grad_norm": 0.6917077302932739,
+      "learning_rate": 4.98060112288537e-05,
+      "loss": 0.3292,
+      "step": 178
+    },
+    {
+      "epoch": 0.09044972208185953,
+      "grad_norm": 0.6623265147209167,
+      "learning_rate": 4.9802077676313354e-05,
+      "loss": 0.381,
+      "step": 179
+    },
+    {
+      "epoch": 0.09095502779181405,
+      "grad_norm": 0.6027669906616211,
+      "learning_rate": 4.979810480077227e-05,
+      "loss": 0.3129,
+      "step": 180
+    },
+    {
+      "epoch": 0.09146033350176858,
+      "grad_norm": 0.4134594798088074,
+      "learning_rate": 4.979409260852933e-05,
+      "loss": 0.1588,
+      "step": 181
+    },
+    {
+      "epoch": 0.0919656392117231,
+      "grad_norm": 0.33261945843696594,
+      "learning_rate": 4.979004110594576e-05,
+      "loss": 0.1806,
+      "step": 182
+    },
+    {
+      "epoch": 0.09247094492167761,
+      "grad_norm": 1.0738641023635864,
+      "learning_rate": 4.97859502994451e-05,
+      "loss": 0.4609,
+      "step": 183
+    },
+    {
+      "epoch": 0.09297625063163213,
+      "grad_norm": 1.1859742403030396,
+      "learning_rate": 4.978182019551321e-05,
+      "loss": 0.5995,
+      "step": 184
+    },
+    {
+      "epoch": 0.09348155634158666,
+      "grad_norm": 0.5822703838348389,
+      "learning_rate": 4.977765080069825e-05,
+      "loss": 0.2734,
+      "step": 185
+    },
+    {
+      "epoch": 0.09398686205154118,
+      "grad_norm": 0.9306544065475464,
+      "learning_rate": 4.977344212161069e-05,
+      "loss": 0.5029,
+      "step": 186
+    },
+    {
+      "epoch": 0.0944921677614957,
+      "grad_norm": 0.3687601685523987,
+      "learning_rate": 4.976919416492327e-05,
+      "loss": 0.2571,
+      "step": 187
+    },
+    {
+      "epoch": 0.09499747347145023,
+      "grad_norm": 0.7644825577735901,
+      "learning_rate": 4.9764906937370995e-05,
+      "loss": 0.4293,
+      "step": 188
+    },
+    {
+      "epoch": 0.09550277918140475,
+      "grad_norm": 0.6048409342765808,
+      "learning_rate": 4.976058044575116e-05,
+      "loss": 0.252,
+      "step": 189
+    },
+    {
+      "epoch": 0.09600808489135927,
+      "grad_norm": 0.89262455701828,
+      "learning_rate": 4.975621469692329e-05,
+      "loss": 0.4889,
+      "step": 190
+    },
+    {
+      "epoch": 0.0965133906013138,
+      "grad_norm": 0.7101221680641174,
+      "learning_rate": 4.9751809697809145e-05,
+      "loss": 0.2739,
+      "step": 191
+    },
+    {
+      "epoch": 0.09701869631126832,
+      "grad_norm": 0.29569289088249207,
+      "learning_rate": 4.974736545539275e-05,
+      "loss": 0.1573,
+      "step": 192
+    },
+    {
+      "epoch": 0.09752400202122284,
+      "grad_norm": 0.5595278739929199,
+      "learning_rate": 4.9742881976720314e-05,
+      "loss": 0.335,
+      "step": 193
+    },
+    {
+      "epoch": 0.09802930773117736,
+      "grad_norm": 0.6488416790962219,
+      "learning_rate": 4.973835926890027e-05,
+      "loss": 0.4142,
+      "step": 194
+    },
+    {
+      "epoch": 0.09853461344113189,
+      "grad_norm": 0.6213271021842957,
+      "learning_rate": 4.973379733910325e-05,
+      "loss": 0.3829,
+      "step": 195
+    },
+    {
+      "epoch": 0.09903991915108641,
+      "grad_norm": 0.5019063353538513,
+      "learning_rate": 4.972919619456205e-05,
+      "loss": 0.275,
+      "step": 196
+    },
+    {
+      "epoch": 0.09954522486104093,
+      "grad_norm": 0.9575361609458923,
+      "learning_rate": 4.972455584257168e-05,
+      "loss": 0.507,
+      "step": 197
+    },
+    {
+      "epoch": 0.10005053057099546,
+      "grad_norm": 0.36353740096092224,
+      "learning_rate": 4.9719876290489264e-05,
+      "loss": 0.2864,
+      "step": 198
+    },
+    {
+      "epoch": 0.10055583628094998,
+      "grad_norm": 0.46957746148109436,
+      "learning_rate": 4.971515754573412e-05,
+      "loss": 0.2227,
+      "step": 199
+    },
+    {
+      "epoch": 0.10106114199090449,
+      "grad_norm": 0.7227563858032227,
+      "learning_rate": 4.9710399615787676e-05,
+      "loss": 0.5132,
+      "step": 200
+    },
+    {
+      "epoch": 0.10156644770085901,
+      "grad_norm": 0.4742310643196106,
+      "learning_rate": 4.9705602508193496e-05,
+      "loss": 0.1944,
+      "step": 201
+    },
+    {
+      "epoch": 0.10207175341081354,
+      "grad_norm": 0.7540463209152222,
+      "learning_rate": 4.970076623055727e-05,
+      "loss": 0.2761,
+      "step": 202
+    },
+    {
+      "epoch": 0.10257705912076806,
+      "grad_norm": 0.7352909445762634,
+      "learning_rate": 4.9695890790546765e-05,
+      "loss": 0.486,
+      "step": 203
+    },
+    {
+      "epoch": 0.10308236483072258,
+      "grad_norm": 0.6061909794807434,
+      "learning_rate": 4.969097619589187e-05,
+      "loss": 0.5126,
+      "step": 204
+    },
+    {
+      "epoch": 0.10358767054067711,
+      "grad_norm": 0.8726075291633606,
+      "learning_rate": 4.968602245438453e-05,
+      "loss": 0.5034,
+      "step": 205
+    },
+    {
+      "epoch": 0.10409297625063163,
+      "grad_norm": 0.9859480857849121,
+      "learning_rate": 4.968102957387877e-05,
+      "loss": 0.4496,
+      "step": 206
+    },
+    {
+      "epoch": 0.10459828196058615,
+      "grad_norm": 0.5186076164245605,
+      "learning_rate": 4.967599756229065e-05,
+      "loss": 0.3507,
+      "step": 207
+    },
+    {
+      "epoch": 0.10510358767054068,
+      "grad_norm": 0.4728350043296814,
+      "learning_rate": 4.967092642759829e-05,
+      "loss": 0.1672,
+      "step": 208
+    },
+    {
+      "epoch": 0.1056088933804952,
+      "grad_norm": 0.6087246537208557,
+      "learning_rate": 4.966581617784184e-05,
+      "loss": 0.3796,
+      "step": 209
+    },
+    {
+      "epoch": 0.10611419909044972,
+      "grad_norm": 0.8859485983848572,
+      "learning_rate": 4.966066682112345e-05,
+      "loss": 0.5639,
+      "step": 210
+    },
+    {
+      "epoch": 0.10661950480040425,
+      "grad_norm": 0.8429753184318542,
+      "learning_rate": 4.9655478365607276e-05,
+      "loss": 0.2432,
+      "step": 211
+    },
+    {
+      "epoch": 0.10712481051035877,
+      "grad_norm": 0.609099805355072,
+      "learning_rate": 4.9650250819519485e-05,
+      "loss": 0.3056,
+      "step": 212
+    },
+    {
+      "epoch": 0.1076301162203133,
+      "grad_norm": 0.40738436579704285,
+      "learning_rate": 4.964498419114819e-05,
+      "loss": 0.1862,
+      "step": 213
+    },
+    {
+      "epoch": 0.10813542193026782,
+      "grad_norm": 0.7260504364967346,
+      "learning_rate": 4.963967848884349e-05,
+      "loss": 0.5646,
+      "step": 214
+    },
+    {
+      "epoch": 0.10864072764022234,
+      "grad_norm": 1.013758659362793,
+      "learning_rate": 4.963433372101743e-05,
+      "loss": 0.4497,
+      "step": 215
+    },
+    {
+      "epoch": 0.10914603335017686,
+      "grad_norm": 0.41159892082214355,
+      "learning_rate": 4.962894989614398e-05,
+      "loss": 0.3157,
+      "step": 216
+    },
+    {
+      "epoch": 0.10965133906013137,
+      "grad_norm": 0.42557021975517273,
+      "learning_rate": 4.9623527022759054e-05,
+      "loss": 0.1879,
+      "step": 217
+    },
+    {
+      "epoch": 0.1101566447700859,
+      "grad_norm": 0.7539249658584595,
+      "learning_rate": 4.9618065109460465e-05,
+      "loss": 0.4206,
+      "step": 218
+    },
+    {
+      "epoch": 0.11066195048004042,
+      "grad_norm": 0.6201863884925842,
+      "learning_rate": 4.961256416490792e-05,
+      "loss": 0.2961,
+      "step": 219
+    },
+    {
+      "epoch": 0.11116725618999494,
+      "grad_norm": 0.7610240578651428,
+      "learning_rate": 4.9607024197823014e-05,
+      "loss": 0.4731,
+      "step": 220
+    },
+    {
+      "epoch": 0.11167256189994947,
+      "grad_norm": 0.4443597197532654,
+      "learning_rate": 4.9601445216989206e-05,
+      "loss": 0.2716,
+      "step": 221
+    },
+    {
+      "epoch": 0.11217786760990399,
+      "grad_norm": 0.47122883796691895,
+      "learning_rate": 4.9595827231251824e-05,
+      "loss": 0.1896,
+      "step": 222
+    },
+    {
+      "epoch": 0.11268317331985851,
+      "grad_norm": 0.6130548715591431,
+      "learning_rate": 4.959017024951802e-05,
+      "loss": 0.3805,
+      "step": 223
+    },
+    {
+      "epoch": 0.11318847902981304,
+      "grad_norm": 0.6293324828147888,
+      "learning_rate": 4.958447428075678e-05,
+      "loss": 0.5559,
+      "step": 224
+    },
+    {
+      "epoch": 0.11369378473976756,
+      "grad_norm": 0.5696532726287842,
+      "learning_rate": 4.957873933399891e-05,
+      "loss": 0.2836,
+      "step": 225
+    },
+    {
+      "epoch": 0.11419909044972208,
+      "grad_norm": 0.5803974270820618,
+      "learning_rate": 4.9572965418336995e-05,
+      "loss": 0.3246,
+      "step": 226
+    },
+    {
+      "epoch": 0.1147043961596766,
+      "grad_norm": 0.6962433457374573,
+      "learning_rate": 4.956715254292543e-05,
+      "loss": 0.4295,
+      "step": 227
+    },
+    {
+      "epoch": 0.11520970186963113,
+      "grad_norm": 0.8711988925933838,
+      "learning_rate": 4.9561300716980365e-05,
+      "loss": 0.6052,
+      "step": 228
+    },
+    {
+      "epoch": 0.11571500757958565,
+      "grad_norm": 0.39029747247695923,
+      "learning_rate": 4.95554099497797e-05,
+      "loss": 0.2225,
+      "step": 229
+    },
+    {
+      "epoch": 0.11622031328954018,
+      "grad_norm": 0.7048918008804321,
+      "learning_rate": 4.9549480250663086e-05,
+      "loss": 0.3617,
+      "step": 230
+    },
+    {
+      "epoch": 0.1167256189994947,
+      "grad_norm": 0.8495561480522156,
+      "learning_rate": 4.954351162903188e-05,
+      "loss": 0.47,
+      "step": 231
+    },
+    {
+      "epoch": 0.11723092470944922,
+      "grad_norm": 0.8110531568527222,
+      "learning_rate": 4.953750409434919e-05,
+      "loss": 0.5883,
+      "step": 232
+    },
+    {
+      "epoch": 0.11773623041940375,
+      "grad_norm": 0.5132966637611389,
+      "learning_rate": 4.953145765613978e-05,
+      "loss": 0.2866,
+      "step": 233
+    },
+    {
+      "epoch": 0.11824153612935827,
+      "grad_norm": 0.6017918586730957,
+      "learning_rate": 4.952537232399012e-05,
+      "loss": 0.3651,
+      "step": 234
+    },
+    {
+      "epoch": 0.11874684183931278,
+      "grad_norm": 0.5894491672515869,
+      "learning_rate": 4.951924810754831e-05,
+      "loss": 0.2754,
+      "step": 235
+    },
+    {
+      "epoch": 0.1192521475492673,
+      "grad_norm": 0.46862131357192993,
+      "learning_rate": 4.9513085016524153e-05,
+      "loss": 0.3216,
+      "step": 236
+    },
+    {
+      "epoch": 0.11975745325922182,
+      "grad_norm": 0.6514838933944702,
+      "learning_rate": 4.9506883060689035e-05,
+      "loss": 0.2915,
+      "step": 237
+    },
+    {
+      "epoch": 0.12026275896917635,
+      "grad_norm": 1.0483429431915283,
+      "learning_rate": 4.950064224987599e-05,
+      "loss": 0.6246,
+      "step": 238
+    },
+    {
+      "epoch": 0.12076806467913087,
+      "grad_norm": 0.3560936748981476,
+      "learning_rate": 4.949436259397966e-05,
+      "loss": 0.2231,
+      "step": 239
+    },
+    {
+      "epoch": 0.1212733703890854,
+      "grad_norm": 0.7140908241271973,
+      "learning_rate": 4.948804410295625e-05,
+      "loss": 0.8288,
+      "step": 240
+    },
+    {
+      "epoch": 0.12177867609903992,
+      "grad_norm": 0.8638331294059753,
+      "learning_rate": 4.9481686786823554e-05,
+      "loss": 0.4715,
+      "step": 241
+    },
+    {
+      "epoch": 0.12228398180899444,
+      "grad_norm": 0.8147590160369873,
+      "learning_rate": 4.9475290655660926e-05,
+      "loss": 0.3322,
+      "step": 242
+    },
+    {
+      "epoch": 0.12278928751894896,
+      "grad_norm": 1.1923398971557617,
+      "learning_rate": 4.946885571960925e-05,
+      "loss": 0.4714,
+      "step": 243
+    },
+    {
+      "epoch": 0.12329459322890349,
+      "grad_norm": 0.4850132167339325,
+      "learning_rate": 4.946238198887093e-05,
+      "loss": 0.2975,
+      "step": 244
+    },
+    {
+      "epoch": 0.12379989893885801,
+      "grad_norm": 0.4749908447265625,
+      "learning_rate": 4.9455869473709895e-05,
+      "loss": 0.3996,
+      "step": 245
+    },
+    {
+      "epoch": 0.12430520464881253,
+      "grad_norm": 0.4123810827732086,
+      "learning_rate": 4.9449318184451545e-05,
+      "loss": 0.1817,
+      "step": 246
+    },
+    {
+      "epoch": 0.12481051035876706,
+      "grad_norm": 0.603053629398346,
+      "learning_rate": 4.944272813148277e-05,
+      "loss": 0.4783,
+      "step": 247
+    },
+    {
+      "epoch": 0.12531581606872158,
+      "grad_norm": 0.2952713668346405,
+      "learning_rate": 4.943609932525193e-05,
+      "loss": 0.1619,
+      "step": 248
+    },
+    {
+      "epoch": 0.1258211217786761,
+      "grad_norm": 0.7722076773643494,
+      "learning_rate": 4.942943177626879e-05,
+      "loss": 0.4847,
+      "step": 249
+    },
+    {
+      "epoch": 0.12632642748863063,
+      "grad_norm": 0.4615485966205597,
+      "learning_rate": 4.942272549510457e-05,
+      "loss": 0.2843,
+      "step": 250
+    },
+    {
+      "epoch": 0.12683173319858515,
+      "grad_norm": 0.501020073890686,
+      "learning_rate": 4.94159804923919e-05,
+      "loss": 0.2411,
+      "step": 251
+    },
+    {
+      "epoch": 0.12733703890853967,
+      "grad_norm": 0.9735652208328247,
+      "learning_rate": 4.9409196778824776e-05,
+      "loss": 0.546,
+      "step": 252
+    },
+    {
+      "epoch": 0.1278423446184942,
+      "grad_norm": 0.9183477163314819,
+      "learning_rate": 4.94023743651586e-05,
+      "loss": 0.5422,
+      "step": 253
+    },
+    {
+      "epoch": 0.12834765032844872,
+      "grad_norm": 0.25691601634025574,
+      "learning_rate": 4.939551326221012e-05,
+      "loss": 0.1477,
+      "step": 254
+    },
+    {
+      "epoch": 0.12885295603840324,
+      "grad_norm": 1.1621726751327515,
+      "learning_rate": 4.938861348085742e-05,
+      "loss": 0.518,
+      "step": 255
+    },
+    {
+      "epoch": 0.12935826174835777,
+      "grad_norm": 1.0264886617660522,
+      "learning_rate": 4.9381675032039896e-05,
+      "loss": 0.513,
+      "step": 256
+    },
+    {
+      "epoch": 0.1298635674583123,
+      "grad_norm": 0.46496647596359253,
+      "learning_rate": 4.9374697926758275e-05,
+      "loss": 0.2556,
+      "step": 257
+    },
+    {
+      "epoch": 0.1303688731682668,
+      "grad_norm": 0.2897842526435852,
+      "learning_rate": 4.936768217607457e-05,
+      "loss": 0.1476,
+      "step": 258
+    },
+    {
+      "epoch": 0.13087417887822134,
+      "grad_norm": 1.0029571056365967,
+      "learning_rate": 4.9360627791112046e-05,
+      "loss": 0.6114,
+      "step": 259
+    },
+    {
+      "epoch": 0.13137948458817586,
+      "grad_norm": 0.42095255851745605,
+      "learning_rate": 4.935353478305524e-05,
+      "loss": 0.3028,
+      "step": 260
+    },
+    {
+      "epoch": 0.13188479029813036,
+      "grad_norm": 1.4107495546340942,
+      "learning_rate": 4.9346403163149924e-05,
+      "loss": 0.677,
+      "step": 261
+    },
+    {
+      "epoch": 0.13239009600808488,
+      "grad_norm": 0.33536162972450256,
+      "learning_rate": 4.9339232942703067e-05,
+      "loss": 0.1942,
+      "step": 262
+    },
+    {
+      "epoch": 0.1328954017180394,
+      "grad_norm": 0.4370989203453064,
+      "learning_rate": 4.933202413308286e-05,
+      "loss": 0.3653,
+      "step": 263
+    },
+    {
+      "epoch": 0.13340070742799393,
+      "grad_norm": 1.1874058246612549,
+      "learning_rate": 4.932477674571867e-05,
+      "loss": 0.5772,
+      "step": 264
+    },
+    {
+      "epoch": 0.13390601313794845,
+      "grad_norm": 0.6043871641159058,
+      "learning_rate": 4.931749079210103e-05,
+      "loss": 0.2519,
+      "step": 265
+    },
+    {
+      "epoch": 0.13441131884790297,
+      "grad_norm": 0.5337929129600525,
+      "learning_rate": 4.931016628378161e-05,
+      "loss": 0.2081,
+      "step": 266
+    },
+    {
+      "epoch": 0.1349166245578575,
+      "grad_norm": 0.6960159540176392,
+      "learning_rate": 4.930280323237321e-05,
+      "loss": 0.3724,
+      "step": 267
+    },
+    {
+      "epoch": 0.13542193026781202,
+      "grad_norm": 0.3179565966129303,
+      "learning_rate": 4.9295401649549755e-05,
+      "loss": 0.2135,
+      "step": 268
+    },
+    {
+      "epoch": 0.13592723597776654,
+      "grad_norm": 0.4810411036014557,
+      "learning_rate": 4.9287961547046234e-05,
+      "loss": 0.2307,
+      "step": 269
+    },
+    {
+      "epoch": 0.13643254168772107,
+      "grad_norm": 0.4444022476673126,
+      "learning_rate": 4.928048293665873e-05,
+      "loss": 0.3267,
+      "step": 270
+    },
+    {
+      "epoch": 0.1369378473976756,
+      "grad_norm": 0.5844589471817017,
+      "learning_rate": 4.9272965830244366e-05,
+      "loss": 0.2969,
+      "step": 271
+    },
+    {
+      "epoch": 0.1374431531076301,
+      "grad_norm": 0.37751492857933044,
+      "learning_rate": 4.9265410239721313e-05,
+      "loss": 0.2331,
+      "step": 272
+    },
+    {
+      "epoch": 0.13794845881758463,
+      "grad_norm": 0.9071863293647766,
+      "learning_rate": 4.925781617706874e-05,
+      "loss": 0.486,
+      "step": 273
+    },
+    {
+      "epoch": 0.13845376452753916,
+      "grad_norm": 0.4862212538719177,
+      "learning_rate": 4.925018365432682e-05,
+      "loss": 0.234,
+      "step": 274
+    },
+    {
+      "epoch": 0.13895907023749368,
+      "grad_norm": 0.2530370056629181,
+      "learning_rate": 4.92425126835967e-05,
+      "loss": 0.1656,
+      "step": 275
+    },
+    {
+      "epoch": 0.1394643759474482,
+      "grad_norm": 0.8576244711875916,
+      "learning_rate": 4.923480327704051e-05,
+      "loss": 0.6208,
+      "step": 276
+    },
+    {
+      "epoch": 0.13996968165740273,
+      "grad_norm": 0.6219656467437744,
+      "learning_rate": 4.922705544688127e-05,
+      "loss": 0.3705,
+      "step": 277
+    },
+    {
+      "epoch": 0.14047498736735725,
+      "grad_norm": 0.3833385705947876,
+      "learning_rate": 4.9219269205402976e-05,
+      "loss": 0.2777,
+      "step": 278
+    },
+    {
+      "epoch": 0.14098029307731177,
+      "grad_norm": 0.5600672960281372,
+      "learning_rate": 4.921144456495048e-05,
+      "loss": 0.5572,
+      "step": 279
+    },
+    {
+      "epoch": 0.1414855987872663,
+      "grad_norm": 0.5854461193084717,
+      "learning_rate": 4.9203581537929536e-05,
+      "loss": 0.3957,
+      "step": 280
+    },
+    {
+      "epoch": 0.14199090449722082,
+      "grad_norm": 0.6858914494514465,
+      "learning_rate": 4.919568013680676e-05,
+      "loss": 0.3994,
+      "step": 281
+    },
+    {
+      "epoch": 0.14249621020717534,
+      "grad_norm": 0.8274922370910645,
+      "learning_rate": 4.918774037410958e-05,
+      "loss": 0.3523,
+      "step": 282
+    },
+    {
+      "epoch": 0.14300151591712987,
+      "grad_norm": 0.423542320728302,
+      "learning_rate": 4.9179762262426286e-05,
+      "loss": 0.2194,
+      "step": 283
+    },
+    {
+      "epoch": 0.1435068216270844,
+      "grad_norm": 0.7165141105651855,
+      "learning_rate": 4.917174581440595e-05,
+      "loss": 0.4029,
+      "step": 284
+    },
+    {
+      "epoch": 0.14401212733703891,
+      "grad_norm": 0.900047242641449,
+      "learning_rate": 4.916369104275842e-05,
+      "loss": 0.5053,
+      "step": 285
+    },
+    {
+      "epoch": 0.14451743304699344,
+      "grad_norm": 0.466221421957016,
+      "learning_rate": 4.9155597960254316e-05,
+      "loss": 0.3247,
+      "step": 286
+    },
+    {
+      "epoch": 0.14502273875694796,
+      "grad_norm": 0.6081900596618652,
+      "learning_rate": 4.914746657972499e-05,
+      "loss": 0.3749,
+      "step": 287
+    },
+    {
+      "epoch": 0.14552804446690248,
+      "grad_norm": 0.8832564949989319,
+      "learning_rate": 4.913929691406253e-05,
+      "loss": 0.3952,
+      "step": 288
+    },
+    {
+      "epoch": 0.146033350176857,
+      "grad_norm": 0.6159337759017944,
+      "learning_rate": 4.91310889762197e-05,
+      "loss": 0.4868,
+      "step": 289
+    },
+    {
+      "epoch": 0.14653865588681153,
+      "grad_norm": 0.6993011832237244,
+      "learning_rate": 4.9122842779209954e-05,
+      "loss": 0.4032,
+      "step": 290
+    },
+    {
+      "epoch": 0.14704396159676605,
+      "grad_norm": 0.7520043253898621,
+      "learning_rate": 4.911455833610742e-05,
+      "loss": 0.2641,
+      "step": 291
+    },
+    {
+      "epoch": 0.14754926730672058,
+      "grad_norm": 0.4298335313796997,
+      "learning_rate": 4.910623566004684e-05,
+      "loss": 0.1774,
+      "step": 292
+    },
+    {
+      "epoch": 0.1480545730166751,
+      "grad_norm": 0.5507643818855286,
+      "learning_rate": 4.909787476422358e-05,
+      "loss": 0.305,
+      "step": 293
+    },
+    {
+      "epoch": 0.14855987872662962,
+      "grad_norm": 0.6847674250602722,
+      "learning_rate": 4.9089475661893625e-05,
+      "loss": 0.3482,
+      "step": 294
+    },
+    {
+      "epoch": 0.14906518443658415,
+      "grad_norm": 0.48942941427230835,
+      "learning_rate": 4.908103836637351e-05,
+      "loss": 0.2464,
+      "step": 295
+    },
+    {
+      "epoch": 0.14957049014653864,
+      "grad_norm": 0.39289000630378723,
+      "learning_rate": 4.907256289104032e-05,
+      "loss": 0.195,
+      "step": 296
+    },
+    {
+      "epoch": 0.15007579585649317,
+      "grad_norm": 0.599053144454956,
+      "learning_rate": 4.906404924933169e-05,
+      "loss": 0.2337,
+      "step": 297
+    },
+    {
+      "epoch": 0.1505811015664477,
+      "grad_norm": 0.6427844762802124,
+      "learning_rate": 4.9055497454745755e-05,
+      "loss": 0.3626,
+      "step": 298
+    },
+    {
+      "epoch": 0.1510864072764022,
+      "grad_norm": 0.875948429107666,
+      "learning_rate": 4.904690752084117e-05,
+      "loss": 0.4159,
+      "step": 299
+    },
+    {
+      "epoch": 0.15159171298635674,
+      "grad_norm": 0.7781713604927063,
+      "learning_rate": 4.9038279461237016e-05,
+      "loss": 0.3178,
+      "step": 300
+    },
+    {
+      "epoch": 0.15209701869631126,
+      "grad_norm": 0.5986030697822571,
+      "learning_rate": 4.902961328961283e-05,
+      "loss": 0.5787,
+      "step": 301
+    },
+    {
+      "epoch": 0.15260232440626578,
+      "grad_norm": 0.6488214135169983,
+      "learning_rate": 4.902090901970861e-05,
+      "loss": 0.3557,
+      "step": 302
+    },
+    {
+      "epoch": 0.1531076301162203,
+      "grad_norm": 0.6705836057662964,
+      "learning_rate": 4.9012166665324714e-05,
+      "loss": 0.306,
+      "step": 303
+    },
+    {
+      "epoch": 0.15361293582617483,
+      "grad_norm": 0.5052811503410339,
+      "learning_rate": 4.900338624032191e-05,
+      "loss": 0.2413,
+      "step": 304
+    },
+    {
+      "epoch": 0.15411824153612935,
+      "grad_norm": 0.5678831934928894,
+      "learning_rate": 4.899456775862131e-05,
+      "loss": 0.3372,
+      "step": 305
+    },
+    {
+      "epoch": 0.15462354724608388,
+      "grad_norm": 0.4097347855567932,
+      "learning_rate": 4.898571123420438e-05,
+      "loss": 0.2792,
+      "step": 306
+    },
+    {
+      "epoch": 0.1551288529560384,
+      "grad_norm": 0.3428628444671631,
+      "learning_rate": 4.897681668111288e-05,
+      "loss": 0.2215,
+      "step": 307
+    },
+    {
+      "epoch": 0.15563415866599292,
+      "grad_norm": 0.5401535630226135,
+      "learning_rate": 4.896788411344888e-05,
+      "loss": 0.2768,
+      "step": 308
+    },
+    {
+      "epoch": 0.15613946437594745,
+      "grad_norm": 0.65704745054245,
+      "learning_rate": 4.895891354537472e-05,
+      "loss": 0.4244,
+      "step": 309
+    },
+    {
+      "epoch": 0.15664477008590197,
+      "grad_norm": 0.8244349360466003,
+      "learning_rate": 4.894990499111297e-05,
+      "loss": 0.4721,
+      "step": 310
+    },
+    {
+      "epoch": 0.1571500757958565,
+      "grad_norm": 1.0010887384414673,
+      "learning_rate": 4.894085846494647e-05,
+      "loss": 0.6647,
+      "step": 311
+    },
+    {
+      "epoch": 0.15765538150581102,
+      "grad_norm": 0.5577598214149475,
+      "learning_rate": 4.893177398121821e-05,
+      "loss": 0.458,
+      "step": 312
+    },
+    {
+      "epoch": 0.15816068721576554,
+      "grad_norm": 0.9697132110595703,
+      "learning_rate": 4.8922651554331386e-05,
+      "loss": 0.583,
+      "step": 313
+    },
+    {
+      "epoch": 0.15866599292572006,
+      "grad_norm": 0.7918589115142822,
+      "learning_rate": 4.891349119874936e-05,
+      "loss": 0.3947,
+      "step": 314
+    },
+    {
+      "epoch": 0.15917129863567459,
+      "grad_norm": 0.5606548190116882,
+      "learning_rate": 4.890429292899562e-05,
+      "loss": 0.2518,
+      "step": 315
+    },
+    {
+      "epoch": 0.1596766043456291,
+      "grad_norm": 0.46425801515579224,
+      "learning_rate": 4.889505675965376e-05,
+      "loss": 0.231,
+      "step": 316
+    },
+    {
+      "epoch": 0.16018191005558363,
+      "grad_norm": 0.5073766708374023,
+      "learning_rate": 4.888578270536747e-05,
+      "loss": 0.404,
+      "step": 317
+    },
+    {
+      "epoch": 0.16068721576553815,
+      "grad_norm": 0.8138315677642822,
+      "learning_rate": 4.88764707808405e-05,
+      "loss": 0.3759,
+      "step": 318
+    },
+    {
+      "epoch": 0.16119252147549268,
+      "grad_norm": 0.6912230849266052,
+      "learning_rate": 4.886712100083665e-05,
+      "loss": 0.4989,
+      "step": 319
+    },
+    {
+      "epoch": 0.1616978271854472,
+      "grad_norm": 0.43325480818748474,
+      "learning_rate": 4.8857733380179735e-05,
+      "loss": 0.3023,
+      "step": 320
+    },
+    {
+      "epoch": 0.16220313289540172,
+      "grad_norm": 0.4167453944683075,
+      "learning_rate": 4.8848307933753566e-05,
+      "loss": 0.2701,
+      "step": 321
+    },
+    {
+      "epoch": 0.16270843860535625,
+      "grad_norm": 0.3875753581523895,
+      "learning_rate": 4.883884467650193e-05,
+      "loss": 0.2084,
+      "step": 322
+    },
+    {
+      "epoch": 0.16321374431531077,
+      "grad_norm": 0.4990168511867523,
+      "learning_rate": 4.882934362342854e-05,
+      "loss": 0.6131,
+      "step": 323
+    },
+    {
+      "epoch": 0.1637190500252653,
+      "grad_norm": 0.5788955092430115,
+      "learning_rate": 4.881980478959707e-05,
+      "loss": 0.2931,
+      "step": 324
+    },
+    {
+      "epoch": 0.16422435573521982,
+      "grad_norm": 0.7873169183731079,
+      "learning_rate": 4.881022819013106e-05,
+      "loss": 0.3825,
+      "step": 325
+    },
+    {
+      "epoch": 0.16472966144517434,
+      "grad_norm": 0.30970779061317444,
+      "learning_rate": 4.8800613840213936e-05,
+      "loss": 0.1879,
+      "step": 326
+    },
+    {
+      "epoch": 0.16523496715512886,
+      "grad_norm": 0.43223828077316284,
+      "learning_rate": 4.8790961755089006e-05,
+      "loss": 0.194,
+      "step": 327
+    },
+    {
+      "epoch": 0.1657402728650834,
+      "grad_norm": 0.4357849061489105,
+      "learning_rate": 4.878127195005936e-05,
+      "loss": 0.2503,
+      "step": 328
+    },
+    {
+      "epoch": 0.1662455785750379,
+      "grad_norm": 0.6348159313201904,
+      "learning_rate": 4.8771544440487916e-05,
+      "loss": 0.2907,
+      "step": 329
+    },
+    {
+      "epoch": 0.1667508842849924,
+      "grad_norm": 0.4712270200252533,
+      "learning_rate": 4.8761779241797366e-05,
+      "loss": 0.2412,
+      "step": 330
+    },
+    {
+      "epoch": 0.16725618999494693,
+      "grad_norm": 0.7354241609573364,
+      "learning_rate": 4.875197636947018e-05,
+      "loss": 0.4536,
+      "step": 331
+    },
+    {
+      "epoch": 0.16776149570490145,
+      "grad_norm": 0.3180365264415741,
+      "learning_rate": 4.874213583904851e-05,
+      "loss": 0.1503,
+      "step": 332
+    },
+    {
+      "epoch": 0.16826680141485598,
+      "grad_norm": 0.9528758525848389,
+      "learning_rate": 4.873225766613425e-05,
+      "loss": 0.4599,
+      "step": 333
+    },
+    {
+      "epoch": 0.1687721071248105,
+      "grad_norm": 0.5695481300354004,
+      "learning_rate": 4.872234186638898e-05,
+      "loss": 0.3238,
+      "step": 334
+    },
+    {
+      "epoch": 0.16927741283476502,
+      "grad_norm": 0.9408870935440063,
+      "learning_rate": 4.871238845553391e-05,
+      "loss": 0.4594,
+      "step": 335
+    },
+    {
+      "epoch": 0.16978271854471955,
+      "grad_norm": 0.5320512652397156,
+      "learning_rate": 4.87023974493499e-05,
+      "loss": 0.3218,
+      "step": 336
+    },
+    {
+      "epoch": 0.17028802425467407,
+      "grad_norm": 0.6941995024681091,
+      "learning_rate": 4.8692368863677416e-05,
+      "loss": 0.4405,
+      "step": 337
+    },
+    {
+      "epoch": 0.1707933299646286,
+      "grad_norm": 0.5261353850364685,
+      "learning_rate": 4.8682302714416494e-05,
+      "loss": 0.2217,
+      "step": 338
+    },
+    {
+      "epoch": 0.17129863567458312,
+      "grad_norm": 0.6723947525024414,
+      "learning_rate": 4.8672199017526726e-05,
+      "loss": 0.3233,
+      "step": 339
+    },
+    {
+      "epoch": 0.17180394138453764,
+      "grad_norm": 0.499994695186615,
+      "learning_rate": 4.866205778902725e-05,
+      "loss": 0.2603,
+      "step": 340
+    },
+    {
+      "epoch": 0.17230924709449216,
+      "grad_norm": 0.3436490595340729,
+      "learning_rate": 4.86518790449967e-05,
+      "loss": 0.2763,
+      "step": 341
+    },
+    {
+      "epoch": 0.17281455280444669,
+      "grad_norm": 0.6315925717353821,
+      "learning_rate": 4.8641662801573185e-05,
+      "loss": 0.2521,
+      "step": 342
+    },
+    {
+      "epoch": 0.1733198585144012,
+      "grad_norm": 0.5178197026252747,
+      "learning_rate": 4.863140907495428e-05,
+      "loss": 0.222,
+      "step": 343
+    },
+    {
+      "epoch": 0.17382516422435573,
+      "grad_norm": 0.5388314127922058,
+      "learning_rate": 4.862111788139697e-05,
+      "loss": 0.3894,
+      "step": 344
+    },
+    {
+      "epoch": 0.17433046993431026,
+      "grad_norm": 0.7556846737861633,
+      "learning_rate": 4.861078923721767e-05,
+      "loss": 0.3644,
+      "step": 345
+    },
+    {
+      "epoch": 0.17483577564426478,
+      "grad_norm": 0.827375590801239,
+      "learning_rate": 4.860042315879214e-05,
+      "loss": 0.5613,
+      "step": 346
+    },
+    {
+      "epoch": 0.1753410813542193,
+      "grad_norm": 0.550095796585083,
+      "learning_rate": 4.8590019662555534e-05,
+      "loss": 0.3149,
+      "step": 347
+    },
+    {
+      "epoch": 0.17584638706417383,
+      "grad_norm": 0.41516733169555664,
+      "learning_rate": 4.857957876500227e-05,
+      "loss": 0.254,
+      "step": 348
+    },
+    {
+      "epoch": 0.17635169277412835,
+      "grad_norm": 0.6708664298057556,
+      "learning_rate": 4.8569100482686125e-05,
+      "loss": 0.7031,
+      "step": 349
+    },
+    {
+      "epoch": 0.17685699848408287,
+      "grad_norm": 0.8377763032913208,
+      "learning_rate": 4.855858483222011e-05,
+      "loss": 0.4918,
+      "step": 350
+    },
+    {
+      "epoch": 0.1773623041940374,
+      "grad_norm": 0.5428167581558228,
+      "learning_rate": 4.8548031830276496e-05,
+      "loss": 0.3071,
+      "step": 351
+    },
+    {
+      "epoch": 0.17786760990399192,
+      "grad_norm": 0.6892185807228088,
+      "learning_rate": 4.8537441493586786e-05,
+      "loss": 0.3797,
+      "step": 352
+    },
+    {
+      "epoch": 0.17837291561394644,
+      "grad_norm": 0.5104579925537109,
+      "learning_rate": 4.852681383894164e-05,
+      "loss": 0.3025,
+      "step": 353
+    },
+    {
+      "epoch": 0.17887822132390097,
+      "grad_norm": 0.42806196212768555,
+      "learning_rate": 4.8516148883190926e-05,
+      "loss": 0.2176,
+      "step": 354
+    },
+    {
+      "epoch": 0.1793835270338555,
+      "grad_norm": 0.6678165197372437,
+      "learning_rate": 4.850544664324363e-05,
+      "loss": 0.3057,
+      "step": 355
+    },
+    {
+      "epoch": 0.17988883274381,
+      "grad_norm": 0.33582162857055664,
+      "learning_rate": 4.849470713606785e-05,
+      "loss": 0.1869,
+      "step": 356
+    },
+    {
+      "epoch": 0.18039413845376454,
+      "grad_norm": 0.8661889433860779,
+      "learning_rate": 4.8483930378690765e-05,
+      "loss": 0.4672,
+      "step": 357
+    },
+    {
+      "epoch": 0.18089944416371906,
+      "grad_norm": 0.6204270124435425,
+      "learning_rate": 4.847311638819865e-05,
+      "loss": 0.2449,
+      "step": 358
+    },
+    {
+      "epoch": 0.18140474987367358,
+      "grad_norm": 0.8477338552474976,
+      "learning_rate": 4.8462265181736756e-05,
+      "loss": 0.3236,
+      "step": 359
+    },
+    {
+      "epoch": 0.1819100555836281,
+      "grad_norm": 0.5757856965065002,
+      "learning_rate": 4.845137677650937e-05,
+      "loss": 0.2292,
+      "step": 360
+    },
+    {
+      "epoch": 0.18241536129358263,
+      "grad_norm": 0.3746514320373535,
+      "learning_rate": 4.8440451189779776e-05,
+      "loss": 0.2001,
+      "step": 361
+    },
+    {
+      "epoch": 0.18292066700353715,
+      "grad_norm": 0.5158061981201172,
+      "learning_rate": 4.8429488438870174e-05,
+      "loss": 0.4076,
+      "step": 362
+    },
+    {
+      "epoch": 0.18342597271349168,
+      "grad_norm": 0.6773928999900818,
+      "learning_rate": 4.841848854116169e-05,
+      "loss": 0.3376,
+      "step": 363
+    },
+    {
+      "epoch": 0.1839312784234462,
+      "grad_norm": 0.6250408291816711,
+      "learning_rate": 4.840745151409437e-05,
+      "loss": 0.3262,
+      "step": 364
+    },
+    {
+      "epoch": 0.1844365841334007,
+      "grad_norm": 0.43249404430389404,
+      "learning_rate": 4.83963773751671e-05,
+      "loss": 0.2738,
+      "step": 365
+    },
+    {
+      "epoch": 0.18494188984335522,
+      "grad_norm": 0.5019465684890747,
+      "learning_rate": 4.838526614193762e-05,
+      "loss": 0.4144,
+      "step": 366
+    },
+    {
+      "epoch": 0.18544719555330974,
+      "grad_norm": 0.45221251249313354,
+      "learning_rate": 4.837411783202249e-05,
+      "loss": 0.2398,
+      "step": 367
+    },
+    {
+      "epoch": 0.18595250126326426,
+      "grad_norm": 0.36865469813346863,
+      "learning_rate": 4.836293246309702e-05,
+      "loss": 0.222,
+      "step": 368
+    },
+    {
+      "epoch": 0.1864578069732188,
+      "grad_norm": 0.43405380845069885,
+      "learning_rate": 4.8351710052895324e-05,
+      "loss": 0.2316,
+      "step": 369
+    },
+    {
+      "epoch": 0.1869631126831733,
+      "grad_norm": 0.7884693145751953,
+      "learning_rate": 4.834045061921022e-05,
+      "loss": 0.3766,
+      "step": 370
+    },
+    {
+      "epoch": 0.18746841839312783,
+      "grad_norm": 0.8879896402359009,
+      "learning_rate": 4.83291541798932e-05,
+      "loss": 0.3775,
+      "step": 371
+    },
+    {
+      "epoch": 0.18797372410308236,
+      "grad_norm": 0.9513370394706726,
+      "learning_rate": 4.831782075285448e-05,
+      "loss": 0.3324,
+      "step": 372
+    },
+    {
+      "epoch": 0.18847902981303688,
+      "grad_norm": 0.4089733064174652,
+      "learning_rate": 4.830645035606287e-05,
+      "loss": 0.1853,
+      "step": 373
+    },
+    {
+      "epoch": 0.1889843355229914,
+      "grad_norm": 0.5456333756446838,
+      "learning_rate": 4.829504300754584e-05,
+      "loss": 0.2975,
+      "step": 374
+    },
+    {
+      "epoch": 0.18948964123294593,
+      "grad_norm": 0.5223088264465332,
+      "learning_rate": 4.82835987253894e-05,
+      "loss": 0.308,
+      "step": 375
+    },
+    {
+      "epoch": 0.18999494694290045,
+      "grad_norm": 0.42592930793762207,
+      "learning_rate": 4.8272117527738146e-05,
+      "loss": 0.2699,
+      "step": 376
+    },
+    {
+      "epoch": 0.19050025265285497,
+      "grad_norm": 0.30264854431152344,
+      "learning_rate": 4.8260599432795205e-05,
+      "loss": 0.1962,
+      "step": 377
+    },
+    {
+      "epoch": 0.1910055583628095,
+      "grad_norm": 0.5884383916854858,
+      "learning_rate": 4.824904445882218e-05,
+      "loss": 0.2513,
+      "step": 378
+    },
+    {
+      "epoch": 0.19151086407276402,
+      "grad_norm": 0.5112461447715759,
+      "learning_rate": 4.823745262413917e-05,
+      "loss": 0.2216,
+      "step": 379
+    },
+    {
+      "epoch": 0.19201616978271854,
+      "grad_norm": 0.5951370000839233,
+      "learning_rate": 4.822582394712471e-05,
+      "loss": 0.2971,
+      "step": 380
+    },
+    {
+      "epoch": 0.19252147549267307,
+      "grad_norm": 0.4729789197444916,
+      "learning_rate": 4.821415844621572e-05,
+      "loss": 0.3288,
+      "step": 381
+    },
+    {
+      "epoch": 0.1930267812026276,
+      "grad_norm": 0.8325059413909912,
+      "learning_rate": 4.820245613990755e-05,
+      "loss": 0.5792,
+      "step": 382
+    },
+    {
+      "epoch": 0.1935320869125821,
+      "grad_norm": 0.6559721827507019,
+      "learning_rate": 4.819071704675386e-05,
+      "loss": 0.3209,
+      "step": 383
+    },
+    {
+      "epoch": 0.19403739262253664,
+      "grad_norm": 0.5043466091156006,
+      "learning_rate": 4.817894118536667e-05,
+      "loss": 0.2815,
+      "step": 384
+    },
+    {
+      "epoch": 0.19454269833249116,
+      "grad_norm": 0.6874819397926331,
+      "learning_rate": 4.8167128574416266e-05,
+      "loss": 0.4657,
+      "step": 385
+    },
+    {
+      "epoch": 0.19504800404244568,
+      "grad_norm": 0.7303402423858643,
+      "learning_rate": 4.8155279232631214e-05,
+      "loss": 0.4223,
+      "step": 386
+    },
+    {
+      "epoch": 0.1955533097524002,
+      "grad_norm": 0.5754918456077576,
+      "learning_rate": 4.8143393178798326e-05,
+      "loss": 0.4584,
+      "step": 387
+    },
+    {
+      "epoch": 0.19605861546235473,
+      "grad_norm": 0.613545298576355,
+      "learning_rate": 4.81314704317626e-05,
+      "loss": 0.2793,
+      "step": 388
+    },
+    {
+      "epoch": 0.19656392117230925,
+      "grad_norm": 0.49800634384155273,
+      "learning_rate": 4.811951101042722e-05,
+      "loss": 0.3728,
+      "step": 389
+    },
+    {
+      "epoch": 0.19706922688226378,
+      "grad_norm": 0.3457117974758148,
+      "learning_rate": 4.8107514933753507e-05,
+      "loss": 0.3332,
+      "step": 390
+    },
+    {
+      "epoch": 0.1975745325922183,
+      "grad_norm": 0.4752323031425476,
+      "learning_rate": 4.8095482220760914e-05,
+      "loss": 0.249,
+      "step": 391
+    },
+    {
+      "epoch": 0.19807983830217282,
+      "grad_norm": 0.6126275062561035,
+      "learning_rate": 4.808341289052697e-05,
+      "loss": 0.334,
+      "step": 392
+    },
+    {
+      "epoch": 0.19858514401212735,
+      "grad_norm": 0.7043460011482239,
+      "learning_rate": 4.807130696218726e-05,
+      "loss": 0.3765,
+      "step": 393
+    },
+    {
+      "epoch": 0.19909044972208187,
+      "grad_norm": 0.410325825214386,
+      "learning_rate": 4.805916445493538e-05,
+      "loss": 0.214,
+      "step": 394
+    },
+    {
+      "epoch": 0.1995957554320364,
+      "grad_norm": 0.794836699962616,
+      "learning_rate": 4.8046985388022955e-05,
+      "loss": 0.3224,
+      "step": 395
+    },
+    {
+      "epoch": 0.20010106114199092,
+      "grad_norm": 0.3947627544403076,
+      "learning_rate": 4.8034769780759556e-05,
+      "loss": 0.2713,
+      "step": 396
+    },
+    {
+      "epoch": 0.20060636685194544,
+      "grad_norm": 0.6008957028388977,
+      "learning_rate": 4.8022517652512674e-05,
+      "loss": 0.3625,
+      "step": 397
+    },
+    {
+      "epoch": 0.20111167256189996,
+      "grad_norm": 0.796962559223175,
+      "learning_rate": 4.801022902270773e-05,
+      "loss": 0.4219,
+      "step": 398
+    },
+    {
+      "epoch": 0.20161697827185449,
+      "grad_norm": 0.5383734107017517,
+      "learning_rate": 4.799790391082799e-05,
+      "loss": 0.2021,
+      "step": 399
+    },
+    {
+      "epoch": 0.20212228398180898,
+      "grad_norm": 0.28900346159935,
+      "learning_rate": 4.7985542336414576e-05,
+      "loss": 0.1807,
+      "step": 400
+    },
+    {
+      "epoch": 0.2026275896917635,
+      "grad_norm": 0.883991539478302,
+      "learning_rate": 4.797314431906642e-05,
+      "loss": 0.4265,
+      "step": 401
+    },
+    {
+      "epoch": 0.20313289540171803,
+      "grad_norm": 0.44878625869750977,
+      "learning_rate": 4.796070987844024e-05,
+      "loss": 0.2376,
+      "step": 402
+    },
+    {
+      "epoch": 0.20363820111167255,
+      "grad_norm": 0.3813835680484772,
+      "learning_rate": 4.7948239034250485e-05,
+      "loss": 0.1876,
+      "step": 403
+    },
+    {
+      "epoch": 0.20414350682162707,
+      "grad_norm": 0.396606981754303,
+      "learning_rate": 4.793573180626934e-05,
+      "loss": 0.2111,
+      "step": 404
+    },
+    {
+      "epoch": 0.2046488125315816,
+      "grad_norm": 0.6839037537574768,
+      "learning_rate": 4.792318821432664e-05,
+      "loss": 0.4695,
+      "step": 405
+    },
+    {
+      "epoch": 0.20515411824153612,
+      "grad_norm": 0.39374062418937683,
+      "learning_rate": 4.791060827830993e-05,
+      "loss": 0.2562,
+      "step": 406
+    },
+    {
+      "epoch": 0.20565942395149064,
+      "grad_norm": 0.5716267824172974,
+      "learning_rate": 4.789799201816433e-05,
+      "loss": 0.3251,
+      "step": 407
+    },
+    {
+      "epoch": 0.20616472966144517,
+      "grad_norm": 0.4453814625740051,
+      "learning_rate": 4.788533945389257e-05,
+      "loss": 0.2136,
+      "step": 408
+    },
+    {
+      "epoch": 0.2066700353713997,
+      "grad_norm": 0.6336759924888611,
+      "learning_rate": 4.787265060555495e-05,
+      "loss": 0.3443,
+      "step": 409
+    },
+    {
+      "epoch": 0.20717534108135421,
+      "grad_norm": 0.7093701362609863,
+      "learning_rate": 4.785992549326925e-05,
+      "loss": 0.4021,
+      "step": 410
+    },
+    {
+      "epoch": 0.20768064679130874,
+      "grad_norm": 0.38103368878364563,
+      "learning_rate": 4.784716413721082e-05,
+      "loss": 0.2219,
+      "step": 411
+    },
+    {
+      "epoch": 0.20818595250126326,
+      "grad_norm": 0.49759477376937866,
+      "learning_rate": 4.783436655761241e-05,
+      "loss": 0.2649,
+      "step": 412
+    },
+    {
+      "epoch": 0.20869125821121778,
+      "grad_norm": 0.7099438309669495,
+      "learning_rate": 4.7821532774764235e-05,
+      "loss": 0.2655,
+      "step": 413
+    },
+    {
+      "epoch": 0.2091965639211723,
+      "grad_norm": 0.6431573629379272,
+      "learning_rate": 4.780866280901389e-05,
+      "loss": 0.2983,
+      "step": 414
+    },
+    {
+      "epoch": 0.20970186963112683,
+      "grad_norm": 0.6158983111381531,
+      "learning_rate": 4.779575668076636e-05,
+      "loss": 0.2617,
+      "step": 415
+    },
+    {
+      "epoch": 0.21020717534108135,
+      "grad_norm": 0.29485249519348145,
+      "learning_rate": 4.778281441048392e-05,
+      "loss": 0.1779,
+      "step": 416
+    },
+    {
+      "epoch": 0.21071248105103588,
+      "grad_norm": 0.45584672689437866,
+      "learning_rate": 4.7769836018686214e-05,
+      "loss": 0.2727,
+      "step": 417
+    },
+    {
+      "epoch": 0.2112177867609904,
+      "grad_norm": 0.5246008634567261,
+      "learning_rate": 4.77568215259501e-05,
+      "loss": 0.2185,
+      "step": 418
+    },
+    {
+      "epoch": 0.21172309247094492,
+      "grad_norm": 0.9197020530700684,
+      "learning_rate": 4.7743770952909694e-05,
+      "loss": 0.3985,
+      "step": 419
+    },
+    {
+      "epoch": 0.21222839818089945,
+      "grad_norm": 0.6227766871452332,
+      "learning_rate": 4.773068432025632e-05,
+      "loss": 0.3579,
+      "step": 420
+    },
+    {
+      "epoch": 0.21273370389085397,
+      "grad_norm": 0.42557641863822937,
+      "learning_rate": 4.771756164873847e-05,
+      "loss": 0.355,
+      "step": 421
+    },
+    {
+      "epoch": 0.2132390096008085,
+      "grad_norm": 0.2553727626800537,
+      "learning_rate": 4.770440295916176e-05,
+      "loss": 0.1222,
+      "step": 422
+    },
+    {
+      "epoch": 0.21374431531076302,
+      "grad_norm": 0.6675715446472168,
+      "learning_rate": 4.769120827238893e-05,
+      "loss": 0.31,
+      "step": 423
+    },
+    {
+      "epoch": 0.21424962102071754,
+      "grad_norm": 0.7006489634513855,
+      "learning_rate": 4.76779776093398e-05,
+      "loss": 0.5355,
+      "step": 424
+    },
+    {
+      "epoch": 0.21475492673067206,
+      "grad_norm": 0.49595072865486145,
+      "learning_rate": 4.766471099099119e-05,
+      "loss": 0.2717,
+      "step": 425
+    },
+    {
+      "epoch": 0.2152602324406266,
+      "grad_norm": 0.7701184153556824,
+      "learning_rate": 4.765140843837697e-05,
+      "loss": 0.5776,
+      "step": 426
+    },
+    {
+      "epoch": 0.2157655381505811,
+      "grad_norm": 0.5482222437858582,
+      "learning_rate": 4.763806997258798e-05,
+      "loss": 0.3752,
+      "step": 427
+    },
+    {
+      "epoch": 0.21627084386053563,
+      "grad_norm": 0.7224484086036682,
+      "learning_rate": 4.762469561477195e-05,
+      "loss": 0.4275,
+      "step": 428
+    },
+    {
+      "epoch": 0.21677614957049016,
+      "grad_norm": 0.5047935843467712,
+      "learning_rate": 4.7611285386133584e-05,
+      "loss": 0.3352,
+      "step": 429
+    },
+    {
+      "epoch": 0.21728145528044468,
+      "grad_norm": 0.7135103344917297,
+      "learning_rate": 4.7597839307934397e-05,
+      "loss": 0.4041,
+      "step": 430
+    },
+    {
+      "epoch": 0.2177867609903992,
+      "grad_norm": 0.622363805770874,
+      "learning_rate": 4.7584357401492784e-05,
+      "loss": 0.4279,
+      "step": 431
+    },
+    {
+      "epoch": 0.21829206670035373,
+      "grad_norm": 0.616588830947876,
+      "learning_rate": 4.7570839688183934e-05,
+      "loss": 0.5381,
+      "step": 432
+    },
+    {
+      "epoch": 0.21879737241030825,
+      "grad_norm": 0.4345376789569855,
+      "learning_rate": 4.7557286189439794e-05,
+      "loss": 0.2053,
+      "step": 433
+    },
+    {
+      "epoch": 0.21930267812026275,
+      "grad_norm": 0.5042337775230408,
+      "learning_rate": 4.754369692674906e-05,
+      "loss": 0.3279,
+      "step": 434
+    },
+    {
+      "epoch": 0.21980798383021727,
+      "grad_norm": 0.6724066734313965,
+      "learning_rate": 4.7530071921657134e-05,
+      "loss": 0.4296,
+      "step": 435
+    },
+    {
+      "epoch": 0.2203132895401718,
+      "grad_norm": 0.607983410358429,
+      "learning_rate": 4.751641119576607e-05,
+      "loss": 0.3913,
+      "step": 436
+    },
+    {
+      "epoch": 0.22081859525012631,
+      "grad_norm": 0.7748120427131653,
+      "learning_rate": 4.750271477073458e-05,
+      "loss": 0.2699,
+      "step": 437
+    },
+    {
+      "epoch": 0.22132390096008084,
+      "grad_norm": 0.9839958548545837,
+      "learning_rate": 4.748898266827795e-05,
+      "loss": 0.365,
+      "step": 438
+    },
+    {
+      "epoch": 0.22182920667003536,
+      "grad_norm": 0.8189884424209595,
+      "learning_rate": 4.747521491016805e-05,
+      "loss": 0.4648,
+      "step": 439
+    },
+    {
+      "epoch": 0.22233451237998988,
+      "grad_norm": 0.43248504400253296,
+      "learning_rate": 4.7461411518233286e-05,
+      "loss": 0.2829,
+      "step": 440
+    },
+    {
+      "epoch": 0.2228398180899444,
+      "grad_norm": 0.6537896990776062,
+      "learning_rate": 4.744757251435853e-05,
+      "loss": 0.4676,
+      "step": 441
+    },
+    {
+      "epoch": 0.22334512379989893,
+      "grad_norm": 0.733722448348999,
+      "learning_rate": 4.7433697920485156e-05,
+      "loss": 0.3722,
+      "step": 442
+    },
+    {
+      "epoch": 0.22385042950985345,
+      "grad_norm": 0.38846614956855774,
+      "learning_rate": 4.741978775861092e-05,
+      "loss": 0.1931,
+      "step": 443
+    },
+    {
+      "epoch": 0.22435573521980798,
+      "grad_norm": 0.5748143196105957,
+      "learning_rate": 4.7405842050790014e-05,
+      "loss": 0.3245,
+      "step": 444
+    },
+    {
+      "epoch": 0.2248610409297625,
+      "grad_norm": 0.674279510974884,
+      "learning_rate": 4.739186081913297e-05,
+      "loss": 0.2701,
+      "step": 445
+    },
+    {
+      "epoch": 0.22536634663971702,
+      "grad_norm": 0.4921652376651764,
+      "learning_rate": 4.737784408580663e-05,
+      "loss": 0.3256,
+      "step": 446
+    },
+    {
+      "epoch": 0.22587165234967155,
+      "grad_norm": 0.4046092629432678,
+      "learning_rate": 4.736379187303414e-05,
+      "loss": 0.2572,
+      "step": 447
+    },
+    {
+      "epoch": 0.22637695805962607,
+      "grad_norm": 0.935234785079956,
+      "learning_rate": 4.734970420309488e-05,
+      "loss": 0.5253,
+      "step": 448
+    },
+    {
+      "epoch": 0.2268822637695806,
+      "grad_norm": 0.44825106859207153,
+      "learning_rate": 4.733558109832447e-05,
+      "loss": 0.254,
+      "step": 449
+    },
+    {
+      "epoch": 0.22738756947953512,
+      "grad_norm": 0.45430219173431396,
+      "learning_rate": 4.732142258111468e-05,
+      "loss": 0.2459,
+      "step": 450
+    },
+    {
+      "epoch": 0.22789287518948964,
+      "grad_norm": 0.44165119528770447,
+      "learning_rate": 4.730722867391346e-05,
+      "loss": 0.23,
+      "step": 451
+    },
+    {
+      "epoch": 0.22839818089944416,
+      "grad_norm": 0.4309244155883789,
+      "learning_rate": 4.729299939922485e-05,
+      "loss": 0.2213,
+      "step": 452
+    },
+    {
+      "epoch": 0.2289034866093987,
+      "grad_norm": 0.6634194254875183,
+      "learning_rate": 4.727873477960896e-05,
+      "loss": 0.3263,
+      "step": 453
+    },
+    {
+      "epoch": 0.2294087923193532,
+      "grad_norm": 0.2821512222290039,
+      "learning_rate": 4.726443483768195e-05,
+      "loss": 0.1682,
+      "step": 454
+    },
+    {
+      "epoch": 0.22991409802930773,
+      "grad_norm": 0.7871589064598083,
+      "learning_rate": 4.7250099596115985e-05,
+      "loss": 0.4115,
+      "step": 455
+    },
+    {
+      "epoch": 0.23041940373926226,
+      "grad_norm": 0.3980810344219208,
+      "learning_rate": 4.723572907763917e-05,
+      "loss": 0.2222,
+      "step": 456
+    },
+    {
+      "epoch": 0.23092470944921678,
+      "grad_norm": 0.34110260009765625,
+      "learning_rate": 4.722132330503558e-05,
+      "loss": 0.2064,
+      "step": 457
+    },
+    {
+      "epoch": 0.2314300151591713,
+      "grad_norm": 0.5213835835456848,
+      "learning_rate": 4.7206882301145164e-05,
+      "loss": 0.3346,
+      "step": 458
+    },
+    {
+      "epoch": 0.23193532086912583,
+      "grad_norm": 0.6506749987602234,
+      "learning_rate": 4.719240608886372e-05,
+      "loss": 0.4456,
+      "step": 459
+    },
+    {
+      "epoch": 0.23244062657908035,
+      "grad_norm": 0.7075578570365906,
+      "learning_rate": 4.7177894691142874e-05,
+      "loss": 0.3906,
+      "step": 460
+    },
+    {
+      "epoch": 0.23294593228903487,
+      "grad_norm": 0.6361711025238037,
+      "learning_rate": 4.7163348130990055e-05,
+      "loss": 0.497,
+      "step": 461
+    },
+    {
+      "epoch": 0.2334512379989894,
+      "grad_norm": 0.4896164536476135,
+      "learning_rate": 4.714876643146842e-05,
+      "loss": 0.2448,
+      "step": 462
+    },
+    {
+      "epoch": 0.23395654370894392,
+      "grad_norm": 0.36150890588760376,
+      "learning_rate": 4.7134149615696846e-05,
+      "loss": 0.2344,
+      "step": 463
+    },
+    {
+      "epoch": 0.23446184941889844,
+      "grad_norm": 0.5678621530532837,
+      "learning_rate": 4.711949770684989e-05,
+      "loss": 0.594,
+      "step": 464
+    },
+    {
+      "epoch": 0.23496715512885297,
+      "grad_norm": 0.42508170008659363,
+      "learning_rate": 4.7104810728157745e-05,
+      "loss": 0.2382,
+      "step": 465
+    },
+    {
+      "epoch": 0.2354724608388075,
+      "grad_norm": 0.6501306295394897,
+      "learning_rate": 4.709008870290619e-05,
+      "loss": 0.3541,
+      "step": 466
+    },
+    {
+      "epoch": 0.235977766548762,
+      "grad_norm": 0.6006211638450623,
+      "learning_rate": 4.7075331654436606e-05,
+      "loss": 0.4746,
+      "step": 467
+    },
+    {
+      "epoch": 0.23648307225871654,
+      "grad_norm": 0.349697083234787,
+      "learning_rate": 4.706053960614586e-05,
+      "loss": 0.1804,
+      "step": 468
+    },
+    {
+      "epoch": 0.23698837796867103,
+      "grad_norm": 0.4210861325263977,
+      "learning_rate": 4.7045712581486345e-05,
+      "loss": 0.3136,
+      "step": 469
+    },
+    {
+      "epoch": 0.23749368367862556,
+      "grad_norm": 0.42396318912506104,
+      "learning_rate": 4.703085060396589e-05,
+      "loss": 0.217,
+      "step": 470
+    },
+    {
+      "epoch": 0.23799898938858008,
+      "grad_norm": 0.4881645739078522,
+      "learning_rate": 4.701595369714775e-05,
+      "loss": 0.3285,
+      "step": 471
+    },
+    {
+      "epoch": 0.2385042950985346,
+      "grad_norm": 0.33498990535736084,
+      "learning_rate": 4.7001021884650545e-05,
+      "loss": 0.2067,
+      "step": 472
+    },
+    {
+      "epoch": 0.23900960080848913,
+      "grad_norm": 0.5688101649284363,
+      "learning_rate": 4.6986055190148255e-05,
+      "loss": 0.2357,
+      "step": 473
+    },
+    {
+      "epoch": 0.23951490651844365,
+      "grad_norm": 0.4370393753051758,
+      "learning_rate": 4.697105363737014e-05,
+      "loss": 0.213,
+      "step": 474
+    },
+    {
+      "epoch": 0.24002021222839817,
+      "grad_norm": 0.31808745861053467,
+      "learning_rate": 4.695601725010077e-05,
+      "loss": 0.2219,
+      "step": 475
+    },
+    {
+      "epoch": 0.2405255179383527,
+      "grad_norm": 0.533358633518219,
+      "learning_rate": 4.694094605217989e-05,
+      "loss": 0.2173,
+      "step": 476
+    },
+    {
+      "epoch": 0.24103082364830722,
+      "grad_norm": 0.4781337380409241,
+      "learning_rate": 4.6925840067502475e-05,
+      "loss": 0.3474,
+      "step": 477
+    },
+    {
+      "epoch": 0.24153612935826174,
+      "grad_norm": 0.43365997076034546,
+      "learning_rate": 4.691069932001864e-05,
+      "loss": 0.1846,
+      "step": 478
+    },
+    {
+      "epoch": 0.24204143506821627,
+      "grad_norm": 0.6059222221374512,
+      "learning_rate": 4.689552383373361e-05,
+      "loss": 0.3881,
+      "step": 479
+    },
+    {
+      "epoch": 0.2425467407781708,
+      "grad_norm": 0.5931808352470398,
+      "learning_rate": 4.6880313632707714e-05,
+      "loss": 0.2737,
+      "step": 480
+    },
+    {
+      "epoch": 0.2430520464881253,
+      "grad_norm": 0.4886901378631592,
+      "learning_rate": 4.6865068741056276e-05,
+      "loss": 0.2409,
+      "step": 481
+    },
+    {
+      "epoch": 0.24355735219807984,
+      "grad_norm": 0.8345271348953247,
+      "learning_rate": 4.684978918294967e-05,
+      "loss": 0.3453,
+      "step": 482
+    },
+    {
+      "epoch": 0.24406265790803436,
+      "grad_norm": 0.5799844861030579,
+      "learning_rate": 4.683447498261318e-05,
+      "loss": 0.403,
+      "step": 483
+    },
+    {
+      "epoch": 0.24456796361798888,
+      "grad_norm": 0.5229687690734863,
+      "learning_rate": 4.6819126164327067e-05,
+      "loss": 0.3206,
+      "step": 484
+    },
+    {
+      "epoch": 0.2450732693279434,
+      "grad_norm": 0.8277320861816406,
+      "learning_rate": 4.680374275242644e-05,
+      "loss": 0.5465,
+      "step": 485
+    },
+    {
+      "epoch": 0.24557857503789793,
+      "grad_norm": 0.6758053302764893,
+      "learning_rate": 4.678832477130127e-05,
+      "loss": 0.3294,
+      "step": 486
+    },
+    {
+      "epoch": 0.24608388074785245,
+      "grad_norm": 0.5581978559494019,
+      "learning_rate": 4.677287224539634e-05,
+      "loss": 0.2746,
+      "step": 487
+    },
+    {
+      "epoch": 0.24658918645780697,
+      "grad_norm": 0.21748855710029602,
+      "learning_rate": 4.675738519921119e-05,
+      "loss": 0.1429,
+      "step": 488
+    },
+    {
+      "epoch": 0.2470944921677615,
+      "grad_norm": 1.0185459852218628,
+      "learning_rate": 4.6741863657300114e-05,
+      "loss": 0.4322,
+      "step": 489
+    },
+    {
+      "epoch": 0.24759979787771602,
+      "grad_norm": 0.37968578934669495,
+      "learning_rate": 4.672630764427206e-05,
+      "loss": 0.2318,
+      "step": 490
+    },
+    {
+      "epoch": 0.24810510358767054,
+      "grad_norm": 0.32278138399124146,
+      "learning_rate": 4.671071718479067e-05,
+      "loss": 0.2341,
+      "step": 491
+    },
+    {
+      "epoch": 0.24861040929762507,
+      "grad_norm": 0.49912557005882263,
+      "learning_rate": 4.669509230357418e-05,
+      "loss": 0.3063,
+      "step": 492
+    },
+    {
+      "epoch": 0.2491157150075796,
+      "grad_norm": 0.6750683188438416,
+      "learning_rate": 4.667943302539539e-05,
+      "loss": 0.3489,
+      "step": 493
+    },
+    {
+      "epoch": 0.24962102071753411,
+      "grad_norm": 0.6013567447662354,
+      "learning_rate": 4.666373937508166e-05,
+      "loss": 0.3084,
+      "step": 494
+    },
+    {
+      "epoch": 0.2501263264274886,
+      "grad_norm": 0.4183938503265381,
+      "learning_rate": 4.6648011377514834e-05,
+      "loss": 0.2177,
+      "step": 495
+    },
+    {
+      "epoch": 0.25063163213744316,
+      "grad_norm": 0.5014682412147522,
+      "learning_rate": 4.6632249057631205e-05,
+      "loss": 0.189,
+      "step": 496
+    },
+    {
+      "epoch": 0.25113693784739766,
+      "grad_norm": 0.7215374708175659,
+      "learning_rate": 4.661645244042149e-05,
+      "loss": 0.4073,
+      "step": 497
+    },
+    {
+      "epoch": 0.2516422435573522,
+      "grad_norm": 0.5658249855041504,
+      "learning_rate": 4.660062155093079e-05,
+      "loss": 0.3558,
+      "step": 498
+    },
+    {
+      "epoch": 0.2521475492673067,
+      "grad_norm": 0.5246372818946838,
+      "learning_rate": 4.6584756414258534e-05,
+      "loss": 0.2615,
+      "step": 499
+    },
+    {
+      "epoch": 0.25265285497726125,
+      "grad_norm": 0.4146447777748108,
+      "learning_rate": 4.656885705555846e-05,
+      "loss": 0.1839,
+      "step": 500
+    },
+    {
+      "epoch": 0.25315816068721575,
+      "grad_norm": 0.7092597484588623,
+      "learning_rate": 4.6552923500038544e-05,
+      "loss": 0.4383,
+      "step": 501
+    },
+    {
+      "epoch": 0.2536634663971703,
+      "grad_norm": 0.44563567638397217,
+      "learning_rate": 4.6536955772961e-05,
+      "loss": 0.2642,
+      "step": 502
+    },
+    {
+      "epoch": 0.2541687721071248,
+      "grad_norm": 0.4488155245780945,
+      "learning_rate": 4.652095389964222e-05,
+      "loss": 0.3054,
+      "step": 503
+    },
+    {
+      "epoch": 0.25467407781707935,
+      "grad_norm": 0.828281044960022,
+      "learning_rate": 4.6504917905452704e-05,
+      "loss": 0.5113,
+      "step": 504
+    },
+    {
+      "epoch": 0.25517938352703384,
+      "grad_norm": 1.007767677307129,
+      "learning_rate": 4.6488847815817096e-05,
+      "loss": 0.6539,
+      "step": 505
+    },
+    {
+      "epoch": 0.2556846892369884,
+      "grad_norm": 0.3789096176624298,
+      "learning_rate": 4.647274365621407e-05,
+      "loss": 0.277,
+      "step": 506
+    },
+    {
+      "epoch": 0.2561899949469429,
+      "grad_norm": 0.37934887409210205,
+      "learning_rate": 4.6456605452176306e-05,
+      "loss": 0.2674,
+      "step": 507
+    },
+    {
+      "epoch": 0.25669530065689744,
+      "grad_norm": 0.39194759726524353,
+      "learning_rate": 4.644043322929049e-05,
+      "loss": 0.2938,
+      "step": 508
+    },
+    {
+      "epoch": 0.25720060636685194,
+      "grad_norm": 0.5336602926254272,
+      "learning_rate": 4.642422701319723e-05,
+      "loss": 0.4041,
+      "step": 509
+    },
+    {
+      "epoch": 0.2577059120768065,
+      "grad_norm": 0.3528728783130646,
+      "learning_rate": 4.640798682959101e-05,
+      "loss": 0.2254,
+      "step": 510
+    },
+    {
+      "epoch": 0.258211217786761,
+      "grad_norm": 0.42827707529067993,
+      "learning_rate": 4.6391712704220215e-05,
+      "loss": 0.1916,
+      "step": 511
+    },
+    {
+      "epoch": 0.25871652349671553,
+      "grad_norm": 0.4275967478752136,
+      "learning_rate": 4.637540466288699e-05,
+      "loss": 0.2075,
+      "step": 512
+    },
+    {
+      "epoch": 0.25922182920667003,
+      "grad_norm": 0.8210417032241821,
+      "learning_rate": 4.6359062731447296e-05,
+      "loss": 0.5352,
+      "step": 513
+    },
+    {
+      "epoch": 0.2597271349166246,
+      "grad_norm": 0.4613182544708252,
+      "learning_rate": 4.63426869358108e-05,
+      "loss": 0.2759,
+      "step": 514
+    },
+    {
+      "epoch": 0.2602324406265791,
+      "grad_norm": 0.3744739890098572,
+      "learning_rate": 4.632627730194087e-05,
+      "loss": 0.2091,
+      "step": 515
+    },
+    {
+      "epoch": 0.2607377463365336,
+      "grad_norm": 0.43345507979393005,
+      "learning_rate": 4.630983385585452e-05,
+      "loss": 0.3114,
+      "step": 516
+    },
+    {
+      "epoch": 0.2612430520464881,
+      "grad_norm": 1.5208643674850464,
+      "learning_rate": 4.6293356623622376e-05,
+      "loss": 0.5386,
+      "step": 517
+    },
+    {
+      "epoch": 0.2617483577564427,
+      "grad_norm": 0.3538508117198944,
+      "learning_rate": 4.627684563136863e-05,
+      "loss": 0.2279,
+      "step": 518
+    },
+    {
+      "epoch": 0.26225366346639717,
+      "grad_norm": 0.5201990008354187,
+      "learning_rate": 4.6260300905271e-05,
+      "loss": 0.2865,
+      "step": 519
+    },
+    {
+      "epoch": 0.2627589691763517,
+      "grad_norm": 0.8713021874427795,
+      "learning_rate": 4.624372247156068e-05,
+      "loss": 0.5142,
+      "step": 520
+    },
+    {
+      "epoch": 0.2632642748863062,
+      "grad_norm": 0.8699602484703064,
+      "learning_rate": 4.622711035652232e-05,
+      "loss": 0.6462,
+      "step": 521
+    },
+    {
+      "epoch": 0.2637695805962607,
+      "grad_norm": 0.3401387333869934,
+      "learning_rate": 4.6210464586493963e-05,
+      "loss": 0.1582,
+      "step": 522
+    },
+    {
+      "epoch": 0.26427488630621526,
+      "grad_norm": 0.5219199061393738,
+      "learning_rate": 4.6193785187867005e-05,
+      "loss": 0.3941,
+      "step": 523
+    },
+    {
+      "epoch": 0.26478019201616976,
+      "grad_norm": 0.4327852129936218,
+      "learning_rate": 4.617707218708617e-05,
+      "loss": 0.2176,
+      "step": 524
+    },
+    {
+      "epoch": 0.2652854977261243,
+      "grad_norm": 0.4247454106807709,
+      "learning_rate": 4.6160325610649465e-05,
+      "loss": 0.1823,
+      "step": 525
+    },
+    {
+      "epoch": 0.2657908034360788,
+      "grad_norm": 0.8779489398002625,
+      "learning_rate": 4.6143545485108094e-05,
+      "loss": 0.5002,
+      "step": 526
+    },
+    {
+      "epoch": 0.26629610914603336,
+      "grad_norm": 0.7940444946289062,
+      "learning_rate": 4.6126731837066505e-05,
+      "loss": 0.334,
+      "step": 527
+    },
+    {
+      "epoch": 0.26680141485598785,
+      "grad_norm": 0.7262102365493774,
+      "learning_rate": 4.610988469318225e-05,
+      "loss": 0.3101,
+      "step": 528
+    },
+    {
+      "epoch": 0.2673067205659424,
+      "grad_norm": 0.36386728286743164,
+      "learning_rate": 4.6093004080166e-05,
+      "loss": 0.2009,
+      "step": 529
+    },
+    {
+      "epoch": 0.2678120262758969,
+      "grad_norm": 0.31724250316619873,
+      "learning_rate": 4.607609002478152e-05,
+      "loss": 0.1564,
+      "step": 530
+    },
+    {
+      "epoch": 0.26831733198585145,
+      "grad_norm": 0.9002480506896973,
+      "learning_rate": 4.605914255384554e-05,
+      "loss": 0.3996,
+      "step": 531
+    },
+    {
+      "epoch": 0.26882263769580594,
+      "grad_norm": 0.28965526819229126,
+      "learning_rate": 4.604216169422783e-05,
+      "loss": 0.1917,
+      "step": 532
+    },
+    {
+      "epoch": 0.2693279434057605,
+      "grad_norm": 0.3934868574142456,
+      "learning_rate": 4.6025147472851036e-05,
+      "loss": 0.2175,
+      "step": 533
+    },
+    {
+      "epoch": 0.269833249115715,
+      "grad_norm": 1.457297682762146,
+      "learning_rate": 4.600809991669076e-05,
+      "loss": 0.5182,
+      "step": 534
+    },
+    {
+      "epoch": 0.27033855482566954,
+      "grad_norm": 0.6048038601875305,
+      "learning_rate": 4.599101905277541e-05,
+      "loss": 0.3971,
+      "step": 535
+    },
+    {
+      "epoch": 0.27084386053562404,
+      "grad_norm": 0.4743158519268036,
+      "learning_rate": 4.597390490818623e-05,
+      "loss": 0.5295,
+      "step": 536
+    },
+    {
+      "epoch": 0.2713491662455786,
+      "grad_norm": 0.7490699291229248,
+      "learning_rate": 4.59567575100572e-05,
+      "loss": 0.3392,
+      "step": 537
+    },
+    {
+      "epoch": 0.2718544719555331,
+      "grad_norm": 0.66379314661026,
+      "learning_rate": 4.5939576885575065e-05,
+      "loss": 0.2745,
+      "step": 538
+    },
+    {
+      "epoch": 0.27235977766548763,
+      "grad_norm": 0.3140435218811035,
+      "learning_rate": 4.59223630619792e-05,
+      "loss": 0.2245,
+      "step": 539
+    },
+    {
+      "epoch": 0.27286508337544213,
+      "grad_norm": 0.6745867133140564,
+      "learning_rate": 4.590511606656165e-05,
+      "loss": 0.3131,
+      "step": 540
+    },
+    {
+      "epoch": 0.2733703890853967,
+      "grad_norm": 0.7853817939758301,
+      "learning_rate": 4.588783592666704e-05,
+      "loss": 0.5628,
+      "step": 541
+    },
+    {
+      "epoch": 0.2738756947953512,
+      "grad_norm": 0.6165689826011658,
+      "learning_rate": 4.5870522669692574e-05,
+      "loss": 0.2666,
+      "step": 542
+    },
+    {
+      "epoch": 0.2743810005053057,
+      "grad_norm": 0.5359889268875122,
+      "learning_rate": 4.58531763230879e-05,
+      "loss": 0.3545,
+      "step": 543
+    },
+    {
+      "epoch": 0.2748863062152602,
+      "grad_norm": 0.6092820763587952,
+      "learning_rate": 4.5835796914355195e-05,
+      "loss": 0.3448,
+      "step": 544
+    },
+    {
+      "epoch": 0.2753916119252148,
+      "grad_norm": 0.6095251441001892,
+      "learning_rate": 4.5818384471049006e-05,
+      "loss": 0.2423,
+      "step": 545
+    },
+    {
+      "epoch": 0.27589691763516927,
+      "grad_norm": 0.5478048920631409,
+      "learning_rate": 4.5800939020776286e-05,
+      "loss": 0.3708,
+      "step": 546
+    },
+    {
+      "epoch": 0.2764022233451238,
+      "grad_norm": 0.3708425462245941,
+      "learning_rate": 4.5783460591196305e-05,
+      "loss": 0.228,
+      "step": 547
+    },
+    {
+      "epoch": 0.2769075290550783,
+      "grad_norm": 0.7903876304626465,
+      "learning_rate": 4.576594921002062e-05,
+      "loss": 0.3286,
+      "step": 548
+    },
+    {
+      "epoch": 0.27741283476503287,
+      "grad_norm": 0.40874460339546204,
+      "learning_rate": 4.5748404905013045e-05,
+      "loss": 0.3423,
+      "step": 549
+    },
+    {
+      "epoch": 0.27791814047498736,
+      "grad_norm": 0.2980653941631317,
+      "learning_rate": 4.5730827703989574e-05,
+      "loss": 0.2125,
+      "step": 550
+    },
+    {
+      "epoch": 0.2784234461849419,
+      "grad_norm": 0.6296504139900208,
+      "learning_rate": 4.5713217634818384e-05,
+      "loss": 0.4838,
+      "step": 551
+    },
+    {
+      "epoch": 0.2789287518948964,
+      "grad_norm": 0.3194442391395569,
+      "learning_rate": 4.569557472541973e-05,
+      "loss": 0.2049,
+      "step": 552
+    },
+    {
+      "epoch": 0.27943405760485096,
+      "grad_norm": 0.4441561698913574,
+      "learning_rate": 4.5677899003765965e-05,
+      "loss": 0.2307,
+      "step": 553
+    },
+    {
+      "epoch": 0.27993936331480546,
+      "grad_norm": 0.6459462642669678,
+      "learning_rate": 4.566019049788145e-05,
+      "loss": 0.4587,
+      "step": 554
+    },
+    {
+      "epoch": 0.28044466902476,
+      "grad_norm": 0.5939819812774658,
+      "learning_rate": 4.564244923584251e-05,
+      "loss": 0.3031,
+      "step": 555
+    },
+    {
+      "epoch": 0.2809499747347145,
+      "grad_norm": 0.5529703497886658,
+      "learning_rate": 4.562467524577744e-05,
+      "loss": 0.2677,
+      "step": 556
+    },
+    {
+      "epoch": 0.281455280444669,
+      "grad_norm": 0.8261821269989014,
+      "learning_rate": 4.56068685558664e-05,
+      "loss": 0.3718,
+      "step": 557
+    },
+    {
+      "epoch": 0.28196058615462355,
+      "grad_norm": 0.42143872380256653,
+      "learning_rate": 4.558902919434139e-05,
+      "loss": 0.3227,
+      "step": 558
+    },
+    {
+      "epoch": 0.28246589186457804,
+      "grad_norm": 0.5339514017105103,
+      "learning_rate": 4.557115718948622e-05,
+      "loss": 0.2646,
+      "step": 559
+    },
+    {
+      "epoch": 0.2829711975745326,
+      "grad_norm": 0.4582768380641937,
+      "learning_rate": 4.5553252569636466e-05,
+      "loss": 0.252,
+      "step": 560
+    },
+    {
+      "epoch": 0.2834765032844871,
+      "grad_norm": 0.4415362477302551,
+      "learning_rate": 4.5535315363179395e-05,
+      "loss": 0.2225,
+      "step": 561
+    },
+    {
+      "epoch": 0.28398180899444164,
+      "grad_norm": 0.6997858881950378,
+      "learning_rate": 4.551734559855394e-05,
+      "loss": 0.4024,
+      "step": 562
+    },
+    {
+      "epoch": 0.28448711470439614,
+      "grad_norm": 0.34175369143486023,
+      "learning_rate": 4.549934330425067e-05,
+      "loss": 0.2033,
+      "step": 563
+    },
+    {
+      "epoch": 0.2849924204143507,
+      "grad_norm": 0.589993953704834,
+      "learning_rate": 4.548130850881171e-05,
+      "loss": 0.3179,
+      "step": 564
+    },
+    {
+      "epoch": 0.2854977261243052,
+      "grad_norm": 0.4499538242816925,
+      "learning_rate": 4.546324124083073e-05,
+      "loss": 0.3204,
+      "step": 565
+    },
+    {
+      "epoch": 0.28600303183425974,
+      "grad_norm": 0.33767470717430115,
+      "learning_rate": 4.544514152895289e-05,
+      "loss": 0.1882,
+      "step": 566
+    },
+    {
+      "epoch": 0.28650833754421423,
+      "grad_norm": 0.2902340292930603,
+      "learning_rate": 4.542700940187476e-05,
+      "loss": 0.1537,
+      "step": 567
+    },
+    {
+      "epoch": 0.2870136432541688,
+      "grad_norm": 0.9272497892379761,
+      "learning_rate": 4.5408844888344315e-05,
+      "loss": 0.6278,
+      "step": 568
+    },
+    {
+      "epoch": 0.2875189489641233,
+      "grad_norm": 0.4656500816345215,
+      "learning_rate": 4.5390648017160904e-05,
+      "loss": 0.2201,
+      "step": 569
+    },
+    {
+      "epoch": 0.28802425467407783,
+      "grad_norm": 0.6928651928901672,
+      "learning_rate": 4.5372418817175146e-05,
+      "loss": 0.3212,
+      "step": 570
+    },
+    {
+      "epoch": 0.2885295603840323,
+      "grad_norm": 0.40516453981399536,
+      "learning_rate": 4.535415731728892e-05,
+      "loss": 0.2295,
+      "step": 571
+    },
+    {
+      "epoch": 0.2890348660939869,
+      "grad_norm": 0.8511061072349548,
+      "learning_rate": 4.533586354645534e-05,
+      "loss": 0.6501,
+      "step": 572
+    },
+    {
+      "epoch": 0.28954017180394137,
+      "grad_norm": 0.3017459809780121,
+      "learning_rate": 4.531753753367867e-05,
+      "loss": 0.1855,
+      "step": 573
+    },
+    {
+      "epoch": 0.2900454775138959,
+      "grad_norm": 0.6671580076217651,
+      "learning_rate": 4.5299179308014266e-05,
+      "loss": 0.5083,
+      "step": 574
+    },
+    {
+      "epoch": 0.2905507832238504,
+      "grad_norm": 0.323030561208725,
+      "learning_rate": 4.52807888985686e-05,
+      "loss": 0.228,
+      "step": 575
+    },
+    {
+      "epoch": 0.29105608893380497,
+      "grad_norm": 0.5347647666931152,
+      "learning_rate": 4.526236633449916e-05,
+      "loss": 0.2377,
+      "step": 576
+    },
+    {
+      "epoch": 0.29156139464375946,
+      "grad_norm": 0.4077557325363159,
+      "learning_rate": 4.524391164501439e-05,
+      "loss": 0.2062,
+      "step": 577
+    },
+    {
+      "epoch": 0.292066700353714,
+      "grad_norm": 0.2940428853034973,
+      "learning_rate": 4.522542485937369e-05,
+      "loss": 0.2001,
+      "step": 578
+    },
+    {
+      "epoch": 0.2925720060636685,
+      "grad_norm": 0.48638758063316345,
+      "learning_rate": 4.520690600688734e-05,
+      "loss": 0.4731,
+      "step": 579
+    },
+    {
+      "epoch": 0.29307731177362306,
+      "grad_norm": 0.6385740637779236,
+      "learning_rate": 4.5188355116916465e-05,
+      "loss": 0.3074,
+      "step": 580
+    },
+    {
+      "epoch": 0.29358261748357756,
+      "grad_norm": 0.6961050033569336,
+      "learning_rate": 4.516977221887299e-05,
+      "loss": 0.4279,
+      "step": 581
+    },
+    {
+      "epoch": 0.2940879231935321,
+      "grad_norm": 0.3082885444164276,
+      "learning_rate": 4.515115734221956e-05,
+      "loss": 0.1595,
+      "step": 582
+    },
+    {
+      "epoch": 0.2945932289034866,
+      "grad_norm": 0.27871477603912354,
+      "learning_rate": 4.5132510516469553e-05,
+      "loss": 0.3318,
+      "step": 583
+    },
+    {
+      "epoch": 0.29509853461344115,
+      "grad_norm": 0.464947909116745,
+      "learning_rate": 4.5113831771187e-05,
+      "loss": 0.2083,
+      "step": 584
+    },
+    {
+      "epoch": 0.29560384032339565,
+      "grad_norm": 0.6374766230583191,
+      "learning_rate": 4.509512113598652e-05,
+      "loss": 0.3364,
+      "step": 585
+    },
+    {
+      "epoch": 0.2961091460333502,
+      "grad_norm": 0.6303216814994812,
+      "learning_rate": 4.507637864053329e-05,
+      "loss": 0.4063,
+      "step": 586
+    },
+    {
+      "epoch": 0.2966144517433047,
+      "grad_norm": 0.7377685904502869,
+      "learning_rate": 4.5057604314543033e-05,
+      "loss": 0.3406,
+      "step": 587
+    },
+    {
+      "epoch": 0.29711975745325925,
+      "grad_norm": 0.6760829091072083,
+      "learning_rate": 4.503879818778191e-05,
+      "loss": 0.3499,
+      "step": 588
+    },
+    {
+      "epoch": 0.29762506316321374,
+      "grad_norm": 0.8320351839065552,
+      "learning_rate": 4.501996029006651e-05,
+      "loss": 0.3893,
+      "step": 589
+    },
+    {
+      "epoch": 0.2981303688731683,
+      "grad_norm": 0.657802402973175,
+      "learning_rate": 4.500109065126379e-05,
+      "loss": 0.4104,
+      "step": 590
+    },
+    {
+      "epoch": 0.2986356745831228,
+      "grad_norm": 0.4237852394580841,
+      "learning_rate": 4.4982189301291037e-05,
+      "loss": 0.2241,
+      "step": 591
+    },
+    {
+      "epoch": 0.2991409802930773,
+      "grad_norm": 0.4519661068916321,
+      "learning_rate": 4.496325627011581e-05,
+      "loss": 0.2499,
+      "step": 592
+    },
+    {
+      "epoch": 0.29964628600303184,
+      "grad_norm": 0.9270481467247009,
+      "learning_rate": 4.49442915877559e-05,
+      "loss": 0.3713,
+      "step": 593
+    },
+    {
+      "epoch": 0.30015159171298633,
+      "grad_norm": 1.016149640083313,
+      "learning_rate": 4.492529528427929e-05,
+      "loss": 0.8942,
+      "step": 594
+    },
+    {
+      "epoch": 0.3006568974229409,
+      "grad_norm": 0.617546558380127,
+      "learning_rate": 4.490626738980408e-05,
+      "loss": 0.361,
+      "step": 595
+    },
+    {
+      "epoch": 0.3011622031328954,
+      "grad_norm": 0.873954176902771,
+      "learning_rate": 4.488720793449847e-05,
+      "loss": 0.3791,
+      "step": 596
+    },
+    {
+      "epoch": 0.30166750884284993,
+      "grad_norm": 0.4339880347251892,
+      "learning_rate": 4.486811694858069e-05,
+      "loss": 0.197,
+      "step": 597
+    },
+    {
+      "epoch": 0.3021728145528044,
+      "grad_norm": 0.8034185171127319,
+      "learning_rate": 4.484899446231896e-05,
+      "loss": 0.3023,
+      "step": 598
+    },
+    {
+      "epoch": 0.302678120262759,
+      "grad_norm": 0.6054482460021973,
+      "learning_rate": 4.4829840506031454e-05,
+      "loss": 0.2276,
+      "step": 599
+    },
+    {
+      "epoch": 0.30318342597271347,
+      "grad_norm": 0.37860724329948425,
+      "learning_rate": 4.481065511008623e-05,
+      "loss": 0.2614,
+      "step": 600
+    },
+    {
+      "epoch": 0.303688731682668,
+      "grad_norm": 0.6753828525543213,
+      "learning_rate": 4.47914383049012e-05,
+      "loss": 0.4661,
+      "step": 601
+    },
+    {
+      "epoch": 0.3041940373926225,
+      "grad_norm": 0.8002330660820007,
+      "learning_rate": 4.477219012094407e-05,
+      "loss": 0.7009,
+      "step": 602
+    },
+    {
+      "epoch": 0.30469934310257707,
+      "grad_norm": 0.45732784271240234,
+      "learning_rate": 4.475291058873231e-05,
+      "loss": 0.2619,
+      "step": 603
+    },
+    {
+      "epoch": 0.30520464881253156,
+      "grad_norm": 0.6500813364982605,
+      "learning_rate": 4.473359973883305e-05,
+      "loss": 0.6237,
+      "step": 604
+    },
+    {
+      "epoch": 0.3057099545224861,
+      "grad_norm": 0.4476456642150879,
+      "learning_rate": 4.471425760186313e-05,
+      "loss": 0.2877,
+      "step": 605
+    },
+    {
+      "epoch": 0.3062152602324406,
+      "grad_norm": 0.2967967689037323,
+      "learning_rate": 4.4694884208488955e-05,
+      "loss": 0.2686,
+      "step": 606
+    },
+    {
+      "epoch": 0.30672056594239516,
+      "grad_norm": 0.320054829120636,
+      "learning_rate": 4.4675479589426496e-05,
+      "loss": 0.1553,
+      "step": 607
+    },
+    {
+      "epoch": 0.30722587165234966,
+      "grad_norm": 0.7721436619758606,
+      "learning_rate": 4.465604377544124e-05,
+      "loss": 0.3458,
+      "step": 608
+    },
+    {
+      "epoch": 0.3077311773623042,
+      "grad_norm": 0.25658470392227173,
+      "learning_rate": 4.463657679734813e-05,
+      "loss": 0.1976,
+      "step": 609
+    },
+    {
+      "epoch": 0.3082364830722587,
+      "grad_norm": 0.5959826707839966,
+      "learning_rate": 4.4617078686011506e-05,
+      "loss": 0.3751,
+      "step": 610
+    },
+    {
+      "epoch": 0.30874178878221326,
+      "grad_norm": 0.7740558981895447,
+      "learning_rate": 4.459754947234508e-05,
+      "loss": 0.5662,
+      "step": 611
+    },
+    {
+      "epoch": 0.30924709449216775,
+      "grad_norm": 0.4466153085231781,
+      "learning_rate": 4.4577989187311886e-05,
+      "loss": 0.2398,
+      "step": 612
+    },
+    {
+      "epoch": 0.3097524002021223,
+      "grad_norm": 0.40263572335243225,
+      "learning_rate": 4.4558397861924195e-05,
+      "loss": 0.2314,
+      "step": 613
+    },
+    {
+      "epoch": 0.3102577059120768,
+      "grad_norm": 0.5410189032554626,
+      "learning_rate": 4.453877552724352e-05,
+      "loss": 0.3845,
+      "step": 614
+    },
+    {
+      "epoch": 0.31076301162203135,
+      "grad_norm": 0.32013383507728577,
+      "learning_rate": 4.451912221438051e-05,
+      "loss": 0.1808,
+      "step": 615
+    },
+    {
+      "epoch": 0.31126831733198584,
+      "grad_norm": 0.8941938877105713,
+      "learning_rate": 4.449943795449494e-05,
+      "loss": 0.4481,
+      "step": 616
+    },
+    {
+      "epoch": 0.3117736230419404,
+      "grad_norm": 0.5344945788383484,
+      "learning_rate": 4.4479722778795675e-05,
+      "loss": 0.3771,
+      "step": 617
+    },
+    {
+      "epoch": 0.3122789287518949,
+      "grad_norm": 0.5631364583969116,
+      "learning_rate": 4.4459976718540555e-05,
+      "loss": 0.3271,
+      "step": 618
+    },
+    {
+      "epoch": 0.31278423446184944,
+      "grad_norm": 0.33273768424987793,
+      "learning_rate": 4.444019980503641e-05,
+      "loss": 0.1694,
+      "step": 619
+    },
+    {
+      "epoch": 0.31328954017180394,
+      "grad_norm": 0.7633798122406006,
+      "learning_rate": 4.4420392069638994e-05,
+      "loss": 0.3704,
+      "step": 620
+    },
+    {
+      "epoch": 0.3137948458817585,
+      "grad_norm": 0.45242956280708313,
+      "learning_rate": 4.44005535437529e-05,
+      "loss": 0.1935,
+      "step": 621
+    },
+    {
+      "epoch": 0.314300151591713,
+      "grad_norm": 0.8318801522254944,
+      "learning_rate": 4.438068425883158e-05,
+      "loss": 0.3171,
+      "step": 622
+    },
+    {
+      "epoch": 0.31480545730166754,
+      "grad_norm": 0.5340676307678223,
+      "learning_rate": 4.4360784246377205e-05,
+      "loss": 0.2594,
+      "step": 623
+    },
+    {
+      "epoch": 0.31531076301162203,
+      "grad_norm": 0.9593076705932617,
+      "learning_rate": 4.434085353794071e-05,
+      "loss": 0.813,
+      "step": 624
+    },
+    {
+      "epoch": 0.3158160687215765,
+      "grad_norm": 0.7513559460639954,
+      "learning_rate": 4.432089216512166e-05,
+      "loss": 0.3647,
+      "step": 625
+    },
+    {
+      "epoch": 0.3163213744315311,
+      "grad_norm": 0.5096583366394043,
+      "learning_rate": 4.430090015956826e-05,
+      "loss": 0.2323,
+      "step": 626
+    },
+    {
+      "epoch": 0.3168266801414856,
+      "grad_norm": 0.587281346321106,
+      "learning_rate": 4.4280877552977276e-05,
+      "loss": 0.2932,
+      "step": 627
+    },
+    {
+      "epoch": 0.3173319858514401,
+      "grad_norm": 0.575646698474884,
+      "learning_rate": 4.426082437709399e-05,
+      "loss": 0.532,
+      "step": 628
+    },
+    {
+      "epoch": 0.3178372915613946,
+      "grad_norm": 0.38853687047958374,
+      "learning_rate": 4.4240740663712155e-05,
+      "loss": 0.1897,
+      "step": 629
+    },
+    {
+      "epoch": 0.31834259727134917,
+      "grad_norm": 0.7989168167114258,
+      "learning_rate": 4.4220626444673926e-05,
+      "loss": 0.4079,
+      "step": 630
+    },
+    {
+      "epoch": 0.31884790298130367,
+      "grad_norm": 0.783780574798584,
+      "learning_rate": 4.420048175186985e-05,
+      "loss": 0.3934,
+      "step": 631
+    },
+    {
+      "epoch": 0.3193532086912582,
+      "grad_norm": 0.7664374113082886,
+      "learning_rate": 4.418030661723877e-05,
+      "loss": 0.2589,
+      "step": 632
+    },
+    {
+      "epoch": 0.3198585144012127,
+      "grad_norm": 0.9156386852264404,
+      "learning_rate": 4.416010107276779e-05,
+      "loss": 0.3592,
+      "step": 633
+    },
+    {
+      "epoch": 0.32036382011116726,
+      "grad_norm": 0.4596467912197113,
+      "learning_rate": 4.413986515049224e-05,
+      "loss": 0.3141,
+      "step": 634
+    },
+    {
+      "epoch": 0.32086912582112176,
+      "grad_norm": 0.5608931183815002,
+      "learning_rate": 4.4119598882495606e-05,
+      "loss": 0.2723,
+      "step": 635
+    },
+    {
+      "epoch": 0.3213744315310763,
+      "grad_norm": 0.6327331066131592,
+      "learning_rate": 4.40993023009095e-05,
+      "loss": 0.273,
+      "step": 636
+    },
+    {
+      "epoch": 0.3218797372410308,
+      "grad_norm": 0.46783167123794556,
+      "learning_rate": 4.4078975437913575e-05,
+      "loss": 0.3164,
+      "step": 637
+    },
+    {
+      "epoch": 0.32238504295098536,
+      "grad_norm": 0.9565727710723877,
+      "learning_rate": 4.4058618325735506e-05,
+      "loss": 0.5771,
+      "step": 638
+    },
+    {
+      "epoch": 0.32289034866093985,
+      "grad_norm": 0.8507360816001892,
+      "learning_rate": 4.403823099665093e-05,
+      "loss": 0.5183,
+      "step": 639
+    },
+    {
+      "epoch": 0.3233956543708944,
+      "grad_norm": 0.8383386135101318,
+      "learning_rate": 4.401781348298338e-05,
+      "loss": 0.4447,
+      "step": 640
+    },
+    {
+      "epoch": 0.3239009600808489,
+      "grad_norm": 0.4394550621509552,
+      "learning_rate": 4.3997365817104255e-05,
+      "loss": 0.2804,
+      "step": 641
+    },
+    {
+      "epoch": 0.32440626579080345,
+      "grad_norm": 0.3480008840560913,
+      "learning_rate": 4.397688803143276e-05,
+      "loss": 0.2214,
+      "step": 642
+    },
+    {
+      "epoch": 0.32491157150075795,
+      "grad_norm": 0.7126933336257935,
+      "learning_rate": 4.395638015843585e-05,
+      "loss": 0.5511,
+      "step": 643
+    },
+    {
+      "epoch": 0.3254168772107125,
+      "grad_norm": 0.681785523891449,
+      "learning_rate": 4.393584223062819e-05,
+      "loss": 0.5339,
+      "step": 644
+    },
+    {
+      "epoch": 0.325922182920667,
+      "grad_norm": 0.4552098214626312,
+      "learning_rate": 4.391527428057207e-05,
+      "loss": 0.3332,
+      "step": 645
+    },
+    {
+      "epoch": 0.32642748863062154,
+      "grad_norm": 0.3405364751815796,
+      "learning_rate": 4.3894676340877425e-05,
+      "loss": 0.1796,
+      "step": 646
+    },
+    {
+      "epoch": 0.32693279434057604,
+      "grad_norm": 0.3657936155796051,
+      "learning_rate": 4.387404844420171e-05,
+      "loss": 0.2149,
+      "step": 647
+    },
+    {
+      "epoch": 0.3274381000505306,
+      "grad_norm": 0.7833860516548157,
+      "learning_rate": 4.385339062324986e-05,
+      "loss": 0.5498,
+      "step": 648
+    },
+    {
+      "epoch": 0.3279434057604851,
+      "grad_norm": 0.7392014265060425,
+      "learning_rate": 4.3832702910774294e-05,
+      "loss": 0.424,
+      "step": 649
+    },
+    {
+      "epoch": 0.32844871147043964,
+      "grad_norm": 0.37213972210884094,
+      "learning_rate": 4.381198533957478e-05,
+      "loss": 0.2245,
+      "step": 650
+    },
+    {
+      "epoch": 0.32895401718039413,
+      "grad_norm": 0.7969143986701965,
+      "learning_rate": 4.379123794249846e-05,
+      "loss": 0.5756,
+      "step": 651
+    },
+    {
+      "epoch": 0.3294593228903487,
+      "grad_norm": 0.46545347571372986,
+      "learning_rate": 4.377046075243976e-05,
+      "loss": 0.2736,
+      "step": 652
+    },
+    {
+      "epoch": 0.3299646286003032,
+      "grad_norm": 0.7078908681869507,
+      "learning_rate": 4.3749653802340315e-05,
+      "loss": 0.454,
+      "step": 653
+    },
+    {
+      "epoch": 0.33046993431025773,
+      "grad_norm": 0.43309125304222107,
+      "learning_rate": 4.372881712518898e-05,
+      "loss": 0.2874,
+      "step": 654
+    },
+    {
+      "epoch": 0.3309752400202122,
+      "grad_norm": 0.8200904726982117,
+      "learning_rate": 4.3707950754021715e-05,
+      "loss": 0.4872,
+      "step": 655
+    },
+    {
+      "epoch": 0.3314805457301668,
+      "grad_norm": 0.46345481276512146,
+      "learning_rate": 4.368705472192157e-05,
+      "loss": 0.3902,
+      "step": 656
+    },
+    {
+      "epoch": 0.33198585144012127,
+      "grad_norm": 0.6028680205345154,
+      "learning_rate": 4.366612906201862e-05,
+      "loss": 0.3987,
+      "step": 657
+    },
+    {
+      "epoch": 0.3324911571500758,
+      "grad_norm": 1.018164873123169,
+      "learning_rate": 4.364517380748991e-05,
+      "loss": 0.5398,
+      "step": 658
+    },
+    {
+      "epoch": 0.3329964628600303,
+      "grad_norm": 0.5439718961715698,
+      "learning_rate": 4.362418899155941e-05,
+      "loss": 0.2593,
+      "step": 659
+    },
+    {
+      "epoch": 0.3335017685699848,
+      "grad_norm": 0.5061584115028381,
+      "learning_rate": 4.360317464749798e-05,
+      "loss": 0.1962,
+      "step": 660
+    },
+    {
+      "epoch": 0.33400707427993936,
+      "grad_norm": 0.5780636072158813,
+      "learning_rate": 4.358213080862324e-05,
+      "loss": 0.2549,
+      "step": 661
+    },
+    {
+      "epoch": 0.33451237998989386,
+      "grad_norm": 0.6662091016769409,
+      "learning_rate": 4.356105750829963e-05,
+      "loss": 0.4128,
+      "step": 662
+    },
+    {
+      "epoch": 0.3350176856998484,
+      "grad_norm": 0.473575234413147,
+      "learning_rate": 4.3539954779938276e-05,
+      "loss": 0.2602,
+      "step": 663
+    },
+    {
+      "epoch": 0.3355229914098029,
+      "grad_norm": 0.6012740135192871,
+      "learning_rate": 4.351882265699696e-05,
+      "loss": 0.5341,
+      "step": 664
+    },
+    {
+      "epoch": 0.33602829711975746,
+      "grad_norm": 0.48335346579551697,
+      "learning_rate": 4.3497661172980074e-05,
+      "loss": 0.2869,
+      "step": 665
+    },
+    {
+      "epoch": 0.33653360282971195,
+      "grad_norm": 0.3564821481704712,
+      "learning_rate": 4.347647036143856e-05,
+      "loss": 0.2183,
+      "step": 666
+    },
+    {
+      "epoch": 0.3370389085396665,
+      "grad_norm": 0.4883384704589844,
+      "learning_rate": 4.345525025596986e-05,
+      "loss": 0.5619,
+      "step": 667
+    },
+    {
+      "epoch": 0.337544214249621,
+      "grad_norm": 0.31306517124176025,
+      "learning_rate": 4.343400089021785e-05,
+      "loss": 0.1676,
+      "step": 668
+    },
+    {
+      "epoch": 0.33804951995957555,
+      "grad_norm": 0.7773547172546387,
+      "learning_rate": 4.341272229787281e-05,
+      "loss": 0.3485,
+      "step": 669
+    },
+    {
+      "epoch": 0.33855482566953005,
+      "grad_norm": 0.349680632352829,
+      "learning_rate": 4.339141451267136e-05,
+      "loss": 0.2042,
+      "step": 670
+    },
+    {
+      "epoch": 0.3390601313794846,
+      "grad_norm": 0.8842143416404724,
+      "learning_rate": 4.337007756839638e-05,
+      "loss": 0.4101,
+      "step": 671
+    },
+    {
+      "epoch": 0.3395654370894391,
+      "grad_norm": 0.4938321113586426,
+      "learning_rate": 4.334871149887701e-05,
+      "loss": 0.2784,
+      "step": 672
+    },
+    {
+      "epoch": 0.34007074279939364,
+      "grad_norm": 0.6494061350822449,
+      "learning_rate": 4.332731633798857e-05,
+      "loss": 0.4212,
+      "step": 673
+    },
+    {
+      "epoch": 0.34057604850934814,
+      "grad_norm": 0.33829325437545776,
+      "learning_rate": 4.3305892119652457e-05,
+      "loss": 0.2634,
+      "step": 674
+    },
+    {
+      "epoch": 0.3410813542193027,
+      "grad_norm": 0.5153347849845886,
+      "learning_rate": 4.32844388778362e-05,
+      "loss": 0.235,
+      "step": 675
+    },
+    {
+      "epoch": 0.3415866599292572,
+      "grad_norm": 0.4803626835346222,
+      "learning_rate": 4.326295664655329e-05,
+      "loss": 0.3195,
+      "step": 676
+    },
+    {
+      "epoch": 0.34209196563921174,
+      "grad_norm": 0.5081639289855957,
+      "learning_rate": 4.3241445459863225e-05,
+      "loss": 0.2381,
+      "step": 677
+    },
+    {
+      "epoch": 0.34259727134916623,
+      "grad_norm": 0.5200014710426331,
+      "learning_rate": 4.321990535187138e-05,
+      "loss": 0.3903,
+      "step": 678
+    },
+    {
+      "epoch": 0.3431025770591208,
+      "grad_norm": 0.5185570120811462,
+      "learning_rate": 4.319833635672899e-05,
+      "loss": 0.1852,
+      "step": 679
+    },
+    {
+      "epoch": 0.3436078827690753,
+      "grad_norm": 0.38153204321861267,
+      "learning_rate": 4.317673850863311e-05,
+      "loss": 0.1491,
+      "step": 680
+    },
+    {
+      "epoch": 0.34411318847902983,
+      "grad_norm": 0.4061324894428253,
+      "learning_rate": 4.31551118418265e-05,
+      "loss": 0.3238,
+      "step": 681
+    },
+    {
+      "epoch": 0.3446184941889843,
+      "grad_norm": 0.45140764117240906,
+      "learning_rate": 4.313345639059766e-05,
+      "loss": 0.2929,
+      "step": 682
+    },
+    {
+      "epoch": 0.3451237998989389,
+      "grad_norm": 0.5511850714683533,
+      "learning_rate": 4.3111772189280686e-05,
+      "loss": 0.2723,
+      "step": 683
+    },
+    {
+      "epoch": 0.34562910560889337,
+      "grad_norm": 0.5179259777069092,
+      "learning_rate": 4.309005927225528e-05,
+      "loss": 0.2211,
+      "step": 684
+    },
+    {
+      "epoch": 0.3461344113188479,
+      "grad_norm": 0.49272972345352173,
+      "learning_rate": 4.306831767394666e-05,
+      "loss": 0.2261,
+      "step": 685
+    },
+    {
+      "epoch": 0.3466397170288024,
+      "grad_norm": 0.6372472047805786,
+      "learning_rate": 4.3046547428825524e-05,
+      "loss": 0.3174,
+      "step": 686
+    },
+    {
+      "epoch": 0.34714502273875697,
+      "grad_norm": 1.0002518892288208,
+      "learning_rate": 4.302474857140799e-05,
+      "loss": 0.4862,
+      "step": 687
+    },
+    {
+      "epoch": 0.34765032844871147,
+      "grad_norm": 0.46908167004585266,
+      "learning_rate": 4.3002921136255535e-05,
+      "loss": 0.2217,
+      "step": 688
+    },
+    {
+      "epoch": 0.348155634158666,
+      "grad_norm": 0.31772932410240173,
+      "learning_rate": 4.298106515797496e-05,
+      "loss": 0.2601,
+      "step": 689
+    },
+    {
+      "epoch": 0.3486609398686205,
+      "grad_norm": 0.40609073638916016,
+      "learning_rate": 4.295918067121828e-05,
+      "loss": 0.2816,
+      "step": 690
+    },
+    {
+      "epoch": 0.34916624557857506,
+      "grad_norm": 0.5628025531768799,
+      "learning_rate": 4.293726771068275e-05,
+      "loss": 0.2602,
+      "step": 691
+    },
+    {
+      "epoch": 0.34967155128852956,
+      "grad_norm": 0.42692917585372925,
+      "learning_rate": 4.2915326311110766e-05,
+      "loss": 0.2569,
+      "step": 692
+    },
+    {
+      "epoch": 0.3501768569984841,
+      "grad_norm": 0.41802138090133667,
+      "learning_rate": 4.289335650728979e-05,
+      "loss": 0.3457,
+      "step": 693
+    },
+    {
+      "epoch": 0.3506821627084386,
+      "grad_norm": 0.5327426195144653,
+      "learning_rate": 4.287135833405235e-05,
+      "loss": 0.5488,
+      "step": 694
+    },
+    {
+      "epoch": 0.3511874684183931,
+      "grad_norm": 0.36314961314201355,
+      "learning_rate": 4.2849331826275905e-05,
+      "loss": 0.3411,
+      "step": 695
+    },
+    {
+      "epoch": 0.35169277412834765,
+      "grad_norm": 0.3764992356300354,
+      "learning_rate": 4.282727701888289e-05,
+      "loss": 0.2655,
+      "step": 696
+    },
+    {
+      "epoch": 0.35219807983830215,
+      "grad_norm": 0.4138118028640747,
+      "learning_rate": 4.280519394684059e-05,
+      "loss": 0.2747,
+      "step": 697
+    },
+    {
+      "epoch": 0.3527033855482567,
+      "grad_norm": 0.31555989384651184,
+      "learning_rate": 4.2783082645161073e-05,
+      "loss": 0.2322,
+      "step": 698
+    },
+    {
+      "epoch": 0.3532086912582112,
+      "grad_norm": 0.5712180733680725,
+      "learning_rate": 4.276094314890122e-05,
+      "loss": 0.3735,
+      "step": 699
+    },
+    {
+      "epoch": 0.35371399696816574,
+      "grad_norm": 0.764113187789917,
+      "learning_rate": 4.273877549316256e-05,
+      "loss": 0.4034,
+      "step": 700
+    },
+    {
+      "epoch": 0.35421930267812024,
+      "grad_norm": 0.8156772255897522,
+      "learning_rate": 4.27165797130913e-05,
+      "loss": 0.3585,
+      "step": 701
+    },
+    {
+      "epoch": 0.3547246083880748,
+      "grad_norm": 0.5966439843177795,
+      "learning_rate": 4.269435584387823e-05,
+      "loss": 0.413,
+      "step": 702
+    },
+    {
+      "epoch": 0.3552299140980293,
+      "grad_norm": 0.621375560760498,
+      "learning_rate": 4.267210392075867e-05,
+      "loss": 0.4242,
+      "step": 703
+    },
+    {
+      "epoch": 0.35573521980798384,
+      "grad_norm": 0.6661776304244995,
+      "learning_rate": 4.2649823979012426e-05,
+      "loss": 0.3709,
+      "step": 704
+    },
+    {
+      "epoch": 0.35624052551793833,
+      "grad_norm": 0.8590171933174133,
+      "learning_rate": 4.262751605396372e-05,
+      "loss": 0.4686,
+      "step": 705
+    },
+    {
+      "epoch": 0.3567458312278929,
+      "grad_norm": 0.5522347092628479,
+      "learning_rate": 4.260518018098114e-05,
+      "loss": 0.2676,
+      "step": 706
+    },
+    {
+      "epoch": 0.3572511369378474,
+      "grad_norm": 0.5579246282577515,
+      "learning_rate": 4.25828163954776e-05,
+      "loss": 0.3184,
+      "step": 707
+    },
+    {
+      "epoch": 0.35775644264780193,
+      "grad_norm": 0.6939714550971985,
+      "learning_rate": 4.2560424732910246e-05,
+      "loss": 0.3246,
+      "step": 708
+    },
+    {
+      "epoch": 0.3582617483577564,
+      "grad_norm": 0.5740989446640015,
+      "learning_rate": 4.2538005228780426e-05,
+      "loss": 0.261,
+      "step": 709
+    },
+    {
+      "epoch": 0.358767054067711,
+      "grad_norm": 0.5999382734298706,
+      "learning_rate": 4.251555791863365e-05,
+      "loss": 0.3403,
+      "step": 710
+    },
+    {
+      "epoch": 0.3592723597776655,
+      "grad_norm": 0.7810065746307373,
+      "learning_rate": 4.2493082838059506e-05,
+      "loss": 0.2779,
+      "step": 711
+    },
+    {
+      "epoch": 0.35977766548762,
+      "grad_norm": 0.6728970408439636,
+      "learning_rate": 4.2470580022691584e-05,
+      "loss": 0.3938,
+      "step": 712
+    },
+    {
+      "epoch": 0.3602829711975745,
+      "grad_norm": 0.42731353640556335,
+      "learning_rate": 4.2448049508207495e-05,
+      "loss": 0.3949,
+      "step": 713
+    },
+    {
+      "epoch": 0.36078827690752907,
+      "grad_norm": 0.5872007608413696,
+      "learning_rate": 4.242549133032872e-05,
+      "loss": 0.2941,
+      "step": 714
+    },
+    {
+      "epoch": 0.36129358261748357,
+      "grad_norm": 0.7026878595352173,
+      "learning_rate": 4.2402905524820625e-05,
+      "loss": 0.3886,
+      "step": 715
+    },
+    {
+      "epoch": 0.3617988883274381,
+      "grad_norm": 0.8333798050880432,
+      "learning_rate": 4.238029212749237e-05,
+      "loss": 0.4645,
+      "step": 716
+    },
+    {
+      "epoch": 0.3623041940373926,
+      "grad_norm": 0.4653269946575165,
+      "learning_rate": 4.235765117419688e-05,
+      "loss": 0.2918,
+      "step": 717
+    },
+    {
+      "epoch": 0.36280949974734716,
+      "grad_norm": 0.46014687418937683,
+      "learning_rate": 4.233498270083074e-05,
+      "loss": 0.1886,
+      "step": 718
+    },
+    {
+      "epoch": 0.36331480545730166,
+      "grad_norm": 0.5073797106742859,
+      "learning_rate": 4.231228674333417e-05,
+      "loss": 0.3901,
+      "step": 719
+    },
+    {
+      "epoch": 0.3638201111672562,
+      "grad_norm": 0.4070146679878235,
+      "learning_rate": 4.2289563337691004e-05,
+      "loss": 0.3156,
+      "step": 720
+    },
+    {
+      "epoch": 0.3643254168772107,
+      "grad_norm": 0.5377737283706665,
+      "learning_rate": 4.226681251992854e-05,
+      "loss": 0.3167,
+      "step": 721
+    },
+    {
+      "epoch": 0.36483072258716526,
+      "grad_norm": 0.590934693813324,
+      "learning_rate": 4.224403432611758e-05,
+      "loss": 0.2468,
+      "step": 722
+    },
+    {
+      "epoch": 0.36533602829711975,
+      "grad_norm": 0.47819361090660095,
+      "learning_rate": 4.22212287923723e-05,
+      "loss": 0.2129,
+      "step": 723
+    },
+    {
+      "epoch": 0.3658413340070743,
+      "grad_norm": 0.39465945959091187,
+      "learning_rate": 4.219839595485026e-05,
+      "loss": 0.1701,
+      "step": 724
+    },
+    {
+      "epoch": 0.3663466397170288,
+      "grad_norm": 0.5328222513198853,
+      "learning_rate": 4.2175535849752254e-05,
+      "loss": 0.2442,
+      "step": 725
+    },
+    {
+      "epoch": 0.36685194542698335,
+      "grad_norm": 0.4375799596309662,
+      "learning_rate": 4.215264851332236e-05,
+      "loss": 0.2072,
+      "step": 726
+    },
+    {
+      "epoch": 0.36735725113693785,
+      "grad_norm": 0.5246615409851074,
+      "learning_rate": 4.212973398184781e-05,
+      "loss": 0.2764,
+      "step": 727
+    },
+    {
+      "epoch": 0.3678625568468924,
+      "grad_norm": 0.5006076097488403,
+      "learning_rate": 4.210679229165894e-05,
+      "loss": 0.2225,
+      "step": 728
+    },
+    {
+      "epoch": 0.3683678625568469,
+      "grad_norm": 0.2846226990222931,
+      "learning_rate": 4.2083823479129175e-05,
+      "loss": 0.147,
+      "step": 729
+    },
+    {
+      "epoch": 0.3688731682668014,
+      "grad_norm": 0.4136352837085724,
+      "learning_rate": 4.206082758067491e-05,
+      "loss": 0.1903,
+      "step": 730
+    },
+    {
+      "epoch": 0.36937847397675594,
+      "grad_norm": 0.8793293833732605,
+      "learning_rate": 4.203780463275551e-05,
+      "loss": 0.4302,
+      "step": 731
+    },
+    {
+      "epoch": 0.36988377968671043,
+      "grad_norm": 0.45669475197792053,
+      "learning_rate": 4.2014754671873205e-05,
+      "loss": 0.2598,
+      "step": 732
+    },
+    {
+      "epoch": 0.370389085396665,
+      "grad_norm": 0.6066158413887024,
+      "learning_rate": 4.199167773457308e-05,
+      "loss": 0.2745,
+      "step": 733
+    },
+    {
+      "epoch": 0.3708943911066195,
+      "grad_norm": 0.6452160477638245,
+      "learning_rate": 4.196857385744295e-05,
+      "loss": 0.2735,
+      "step": 734
+    },
+    {
+      "epoch": 0.37139969681657403,
+      "grad_norm": 0.9064080715179443,
+      "learning_rate": 4.194544307711338e-05,
+      "loss": 0.5442,
+      "step": 735
+    },
+    {
+      "epoch": 0.3719050025265285,
+      "grad_norm": 0.5397379398345947,
+      "learning_rate": 4.192228543025759e-05,
+      "loss": 0.3049,
+      "step": 736
+    },
+    {
+      "epoch": 0.3724103082364831,
+      "grad_norm": 0.7189819812774658,
+      "learning_rate": 4.189910095359135e-05,
+      "loss": 0.4373,
+      "step": 737
+    },
+    {
+      "epoch": 0.3729156139464376,
+      "grad_norm": 0.4282380938529968,
+      "learning_rate": 4.187588968387303e-05,
+      "loss": 0.2361,
+      "step": 738
+    },
+    {
+      "epoch": 0.3734209196563921,
+      "grad_norm": 0.6697452068328857,
+      "learning_rate": 4.185265165790343e-05,
+      "loss": 0.7482,
+      "step": 739
+    },
+    {
+      "epoch": 0.3739262253663466,
+      "grad_norm": 0.8984920978546143,
+      "learning_rate": 4.182938691252579e-05,
+      "loss": 0.3867,
+      "step": 740
+    },
+    {
+      "epoch": 0.37443153107630117,
+      "grad_norm": 0.8725920915603638,
+      "learning_rate": 4.180609548462573e-05,
+      "loss": 0.3112,
+      "step": 741
+    },
+    {
+      "epoch": 0.37493683678625567,
+      "grad_norm": 0.548840343952179,
+      "learning_rate": 4.178277741113114e-05,
+      "loss": 0.2397,
+      "step": 742
+    },
+    {
+      "epoch": 0.3754421424962102,
+      "grad_norm": 0.38334935903549194,
+      "learning_rate": 4.175943272901218e-05,
+      "loss": 0.2286,
+      "step": 743
+    },
+    {
+      "epoch": 0.3759474482061647,
+      "grad_norm": 0.3433682918548584,
+      "learning_rate": 4.173606147528119e-05,
+      "loss": 0.2938,
+      "step": 744
+    },
+    {
+      "epoch": 0.37645275391611926,
+      "grad_norm": 0.3817785680294037,
+      "learning_rate": 4.171266368699265e-05,
+      "loss": 0.1541,
+      "step": 745
+    },
+    {
+      "epoch": 0.37695805962607376,
+      "grad_norm": 0.3005577325820923,
+      "learning_rate": 4.1689239401243086e-05,
+      "loss": 0.1874,
+      "step": 746
+    },
+    {
+      "epoch": 0.3774633653360283,
+      "grad_norm": 0.41856083273887634,
+      "learning_rate": 4.166578865517105e-05,
+      "loss": 0.3474,
+      "step": 747
+    },
+    {
+      "epoch": 0.3779686710459828,
+      "grad_norm": 0.6588180661201477,
+      "learning_rate": 4.164231148595705e-05,
+      "loss": 0.3855,
+      "step": 748
+    },
+    {
+      "epoch": 0.37847397675593736,
+      "grad_norm": 0.5970372557640076,
+      "learning_rate": 4.1618807930823474e-05,
+      "loss": 0.3084,
+      "step": 749
+    },
+    {
+      "epoch": 0.37897928246589185,
+      "grad_norm": 0.29243141412734985,
+      "learning_rate": 4.159527802703456e-05,
+      "loss": 0.1611,
+      "step": 750
+    },
+    {
+      "epoch": 0.3794845881758464,
+      "grad_norm": 0.5415188670158386,
+      "learning_rate": 4.1571721811896304e-05,
+      "loss": 0.2595,
+      "step": 751
+    },
+    {
+      "epoch": 0.3799898938858009,
+      "grad_norm": 0.33659324049949646,
+      "learning_rate": 4.154813932275643e-05,
+      "loss": 0.1616,
+      "step": 752
+    },
+    {
+      "epoch": 0.38049519959575545,
+      "grad_norm": 0.542412281036377,
+      "learning_rate": 4.152453059700431e-05,
+      "loss": 0.5679,
+      "step": 753
+    },
+    {
+      "epoch": 0.38100050530570995,
+      "grad_norm": 0.43469804525375366,
+      "learning_rate": 4.150089567207094e-05,
+      "loss": 0.1785,
+      "step": 754
+    },
+    {
+      "epoch": 0.3815058110156645,
+      "grad_norm": 0.27857232093811035,
+      "learning_rate": 4.14772345854288e-05,
+      "loss": 0.1647,
+      "step": 755
+    },
+    {
+      "epoch": 0.382011116725619,
+      "grad_norm": 0.45017367601394653,
+      "learning_rate": 4.145354737459191e-05,
+      "loss": 0.2412,
+      "step": 756
+    },
+    {
+      "epoch": 0.38251642243557354,
+      "grad_norm": 0.6093769669532776,
+      "learning_rate": 4.142983407711567e-05,
+      "loss": 0.3251,
+      "step": 757
+    },
+    {
+      "epoch": 0.38302172814552804,
+      "grad_norm": 0.9887945652008057,
+      "learning_rate": 4.1406094730596844e-05,
+      "loss": 0.4258,
+      "step": 758
+    },
+    {
+      "epoch": 0.3835270338554826,
+      "grad_norm": 0.3608369827270508,
+      "learning_rate": 4.138232937267351e-05,
+      "loss": 0.1477,
+      "step": 759
+    },
+    {
+      "epoch": 0.3840323395654371,
+      "grad_norm": 0.6515466570854187,
+      "learning_rate": 4.135853804102497e-05,
+      "loss": 0.346,
+      "step": 760
+    },
+    {
+      "epoch": 0.38453764527539164,
+      "grad_norm": 0.6651821732521057,
+      "learning_rate": 4.133472077337173e-05,
+      "loss": 0.4113,
+      "step": 761
+    },
+    {
+      "epoch": 0.38504295098534613,
+      "grad_norm": 0.43307211995124817,
+      "learning_rate": 4.131087760747538e-05,
+      "loss": 0.1571,
+      "step": 762
+    },
+    {
+      "epoch": 0.3855482566953007,
+      "grad_norm": 0.928577721118927,
+      "learning_rate": 4.128700858113862e-05,
+      "loss": 0.4672,
+      "step": 763
+    },
+    {
+      "epoch": 0.3860535624052552,
+      "grad_norm": 0.5100760459899902,
+      "learning_rate": 4.1263113732205106e-05,
+      "loss": 0.3695,
+      "step": 764
+    },
+    {
+      "epoch": 0.3865588681152097,
+      "grad_norm": 0.5143333673477173,
+      "learning_rate": 4.123919309855945e-05,
+      "loss": 0.2731,
+      "step": 765
+    },
+    {
+      "epoch": 0.3870641738251642,
+      "grad_norm": 0.705123245716095,
+      "learning_rate": 4.121524671812716e-05,
+      "loss": 0.2643,
+      "step": 766
+    },
+    {
+      "epoch": 0.3875694795351187,
+      "grad_norm": 0.44116538763046265,
+      "learning_rate": 4.1191274628874545e-05,
+      "loss": 0.2125,
+      "step": 767
+    },
+    {
+      "epoch": 0.3880747852450733,
+      "grad_norm": 1.2351047992706299,
+      "learning_rate": 4.116727686880869e-05,
+      "loss": 0.4035,
+      "step": 768
+    },
+    {
+      "epoch": 0.38858009095502777,
+      "grad_norm": 0.646878182888031,
+      "learning_rate": 4.114325347597736e-05,
+      "loss": 0.2416,
+      "step": 769
+    },
+    {
+      "epoch": 0.3890853966649823,
+      "grad_norm": 0.19294582307338715,
+      "learning_rate": 4.111920448846898e-05,
+      "loss": 0.0957,
+      "step": 770
+    },
+    {
+      "epoch": 0.3895907023749368,
+      "grad_norm": 0.45887112617492676,
+      "learning_rate": 4.1095129944412556e-05,
+      "loss": 0.4743,
+      "step": 771
+    },
+    {
+      "epoch": 0.39009600808489137,
+      "grad_norm": 0.3489443063735962,
+      "learning_rate": 4.107102988197758e-05,
+      "loss": 0.1996,
+      "step": 772
+    },
+    {
+      "epoch": 0.39060131379484586,
+      "grad_norm": 0.6400259137153625,
+      "learning_rate": 4.1046904339374056e-05,
+      "loss": 0.3377,
+      "step": 773
+    },
+    {
+      "epoch": 0.3911066195048004,
+      "grad_norm": 0.8612831830978394,
+      "learning_rate": 4.1022753354852345e-05,
+      "loss": 0.3707,
+      "step": 774
+    },
+    {
+      "epoch": 0.3916119252147549,
+      "grad_norm": 0.6197051405906677,
+      "learning_rate": 4.099857696670315e-05,
+      "loss": 0.3053,
+      "step": 775
+    },
+    {
+      "epoch": 0.39211723092470946,
+      "grad_norm": 0.5743224620819092,
+      "learning_rate": 4.0974375213257475e-05,
+      "loss": 0.3574,
+      "step": 776
+    },
+    {
+      "epoch": 0.39262253663466395,
+      "grad_norm": 0.5291123986244202,
+      "learning_rate": 4.0950148132886504e-05,
+      "loss": 0.2928,
+      "step": 777
+    },
+    {
+      "epoch": 0.3931278423446185,
+      "grad_norm": 0.9359875917434692,
+      "learning_rate": 4.0925895764001615e-05,
+      "loss": 0.5865,
+      "step": 778
+    },
+    {
+      "epoch": 0.393633148054573,
+      "grad_norm": 0.40809711813926697,
+      "learning_rate": 4.090161814505424e-05,
+      "loss": 0.2414,
+      "step": 779
+    },
+    {
+      "epoch": 0.39413845376452755,
+      "grad_norm": 0.3236285448074341,
+      "learning_rate": 4.087731531453588e-05,
+      "loss": 0.1779,
+      "step": 780
+    },
+    {
+      "epoch": 0.39464375947448205,
+      "grad_norm": 0.8675460815429688,
+      "learning_rate": 4.0852987310978e-05,
+      "loss": 0.4552,
+      "step": 781
+    },
+    {
+      "epoch": 0.3951490651844366,
+      "grad_norm": 0.3517732620239258,
+      "learning_rate": 4.082863417295195e-05,
+      "loss": 0.1672,
+      "step": 782
+    },
+    {
+      "epoch": 0.3956543708943911,
+      "grad_norm": 0.5657680630683899,
+      "learning_rate": 4.080425593906896e-05,
+      "loss": 0.3311,
+      "step": 783
+    },
+    {
+      "epoch": 0.39615967660434565,
+      "grad_norm": 0.7513840198516846,
+      "learning_rate": 4.077985264798003e-05,
+      "loss": 0.3397,
+      "step": 784
+    },
+    {
+      "epoch": 0.39666498231430014,
+      "grad_norm": 0.41284772753715515,
+      "learning_rate": 4.075542433837591e-05,
+      "loss": 0.1583,
+      "step": 785
+    },
+    {
+      "epoch": 0.3971702880242547,
+      "grad_norm": 0.41029801964759827,
+      "learning_rate": 4.0730971048987e-05,
+      "loss": 0.3041,
+      "step": 786
+    },
+    {
+      "epoch": 0.3976755937342092,
+      "grad_norm": 0.7941258549690247,
+      "learning_rate": 4.070649281858329e-05,
+      "loss": 0.3464,
+      "step": 787
+    },
+    {
+      "epoch": 0.39818089944416374,
+      "grad_norm": 1.075421929359436,
+      "learning_rate": 4.068198968597435e-05,
+      "loss": 0.5832,
+      "step": 788
+    },
+    {
+      "epoch": 0.39868620515411823,
+      "grad_norm": 0.46112877130508423,
+      "learning_rate": 4.0657461690009195e-05,
+      "loss": 0.322,
+      "step": 789
+    },
+    {
+      "epoch": 0.3991915108640728,
+      "grad_norm": 0.528905987739563,
+      "learning_rate": 4.063290886957629e-05,
+      "loss": 0.5137,
+      "step": 790
+    },
+    {
+      "epoch": 0.3996968165740273,
+      "grad_norm": 0.43017876148223877,
+      "learning_rate": 4.0608331263603437e-05,
+      "loss": 0.2687,
+      "step": 791
+    },
+    {
+      "epoch": 0.40020212228398183,
+      "grad_norm": 0.49231600761413574,
+      "learning_rate": 4.058372891105775e-05,
+      "loss": 0.2438,
+      "step": 792
+    },
+    {
+      "epoch": 0.4007074279939363,
+      "grad_norm": 0.7696442008018494,
+      "learning_rate": 4.055910185094557e-05,
+      "loss": 0.3306,
+      "step": 793
+    },
+    {
+      "epoch": 0.4012127337038909,
+      "grad_norm": 0.6930801272392273,
+      "learning_rate": 4.0534450122312404e-05,
+      "loss": 0.6515,
+      "step": 794
+    },
+    {
+      "epoch": 0.4017180394138454,
+      "grad_norm": 0.28239572048187256,
+      "learning_rate": 4.0509773764242886e-05,
+      "loss": 0.2526,
+      "step": 795
+    },
+    {
+      "epoch": 0.4022233451237999,
+      "grad_norm": 0.5681050419807434,
+      "learning_rate": 4.048507281586069e-05,
+      "loss": 0.2447,
+      "step": 796
+    },
+    {
+      "epoch": 0.4027286508337544,
+      "grad_norm": 0.3820132911205292,
+      "learning_rate": 4.046034731632847e-05,
+      "loss": 0.2624,
+      "step": 797
+    },
+    {
+      "epoch": 0.40323395654370897,
+      "grad_norm": 0.9789735674858093,
+      "learning_rate": 4.043559730484784e-05,
+      "loss": 0.5155,
+      "step": 798
+    },
+    {
+      "epoch": 0.40373926225366347,
+      "grad_norm": 0.4344816505908966,
+      "learning_rate": 4.0410822820659224e-05,
+      "loss": 0.2145,
+      "step": 799
+    },
+    {
+      "epoch": 0.40424456796361796,
+      "grad_norm": 0.33764201402664185,
+      "learning_rate": 4.0386023903041895e-05,
+      "loss": 0.2634,
+      "step": 800
+    },
+    {
+      "epoch": 0.4047498736735725,
+      "grad_norm": 0.2780141830444336,
+      "learning_rate": 4.036120059131382e-05,
+      "loss": 0.1559,
+      "step": 801
+    },
+    {
+      "epoch": 0.405255179383527,
+      "grad_norm": 0.5865265130996704,
+      "learning_rate": 4.0336352924831674e-05,
+      "loss": 0.3824,
+      "step": 802
+    },
+    {
+      "epoch": 0.40576048509348156,
+      "grad_norm": 0.5158589482307434,
+      "learning_rate": 4.0311480942990756e-05,
+      "loss": 0.3323,
+      "step": 803
+    },
+    {
+      "epoch": 0.40626579080343606,
+      "grad_norm": 0.5034775733947754,
+      "learning_rate": 4.0286584685224885e-05,
+      "loss": 0.2576,
+      "step": 804
+    },
+    {
+      "epoch": 0.4067710965133906,
+      "grad_norm": 0.5502609610557556,
+      "learning_rate": 4.026166419100638e-05,
+      "loss": 0.3489,
+      "step": 805
+    },
+    {
+      "epoch": 0.4072764022233451,
+      "grad_norm": 0.4883863031864166,
+      "learning_rate": 4.023671949984598e-05,
+      "loss": 0.2665,
+      "step": 806
+    },
+    {
+      "epoch": 0.40778170793329965,
+      "grad_norm": 0.40211907029151917,
+      "learning_rate": 4.02117506512928e-05,
+      "loss": 0.2291,
+      "step": 807
+    },
+    {
+      "epoch": 0.40828701364325415,
+      "grad_norm": 0.35028353333473206,
+      "learning_rate": 4.0186757684934275e-05,
+      "loss": 0.2617,
+      "step": 808
+    },
+    {
+      "epoch": 0.4087923193532087,
+      "grad_norm": 0.6691428422927856,
+      "learning_rate": 4.0161740640396015e-05,
+      "loss": 0.4747,
+      "step": 809
+    },
+    {
+      "epoch": 0.4092976250631632,
+      "grad_norm": 0.3790917694568634,
+      "learning_rate": 4.0136699557341874e-05,
+      "loss": 0.1726,
+      "step": 810
+    },
+    {
+      "epoch": 0.40980293077311775,
+      "grad_norm": 0.2536380887031555,
+      "learning_rate": 4.011163447547378e-05,
+      "loss": 0.1844,
+      "step": 811
+    },
+    {
+      "epoch": 0.41030823648307224,
+      "grad_norm": 0.7773703336715698,
+      "learning_rate": 4.0086545434531717e-05,
+      "loss": 0.2856,
+      "step": 812
+    },
+    {
+      "epoch": 0.4108135421930268,
+      "grad_norm": 0.36982572078704834,
+      "learning_rate": 4.0061432474293656e-05,
+      "loss": 0.1549,
+      "step": 813
+    },
+    {
+      "epoch": 0.4113188479029813,
+      "grad_norm": 0.3912838399410248,
+      "learning_rate": 4.00362956345755e-05,
+      "loss": 0.2821,
+      "step": 814
+    },
+    {
+      "epoch": 0.41182415361293584,
+      "grad_norm": 0.4090442359447479,
+      "learning_rate": 4.001113495523101e-05,
+      "loss": 0.2502,
+      "step": 815
+    },
+    {
+      "epoch": 0.41232945932289033,
+      "grad_norm": 0.860283374786377,
+      "learning_rate": 3.998595047615172e-05,
+      "loss": 0.3686,
+      "step": 816
+    },
+    {
+      "epoch": 0.4128347650328449,
+      "grad_norm": 0.3724400997161865,
+      "learning_rate": 3.9960742237266946e-05,
+      "loss": 0.1763,
+      "step": 817
+    },
+    {
+      "epoch": 0.4133400707427994,
+      "grad_norm": 0.2720755636692047,
+      "learning_rate": 3.993551027854362e-05,
+      "loss": 0.1704,
+      "step": 818
+    },
+    {
+      "epoch": 0.41384537645275393,
+      "grad_norm": 0.37592801451683044,
+      "learning_rate": 3.991025463998632e-05,
+      "loss": 0.2151,
+      "step": 819
+    },
+    {
+      "epoch": 0.41435068216270843,
+      "grad_norm": 0.5671641826629639,
+      "learning_rate": 3.9884975361637146e-05,
+      "loss": 0.5564,
+      "step": 820
+    },
+    {
+      "epoch": 0.414855987872663,
+      "grad_norm": 0.6280914545059204,
+      "learning_rate": 3.9859672483575694e-05,
+      "loss": 0.3185,
+      "step": 821
+    },
+    {
+      "epoch": 0.4153612935826175,
+      "grad_norm": 0.3018336892127991,
+      "learning_rate": 3.983434604591898e-05,
+      "loss": 0.235,
+      "step": 822
+    },
+    {
+      "epoch": 0.415866599292572,
+      "grad_norm": 0.7508260607719421,
+      "learning_rate": 3.980899608882134e-05,
+      "loss": 0.3689,
+      "step": 823
+    },
+    {
+      "epoch": 0.4163719050025265,
+      "grad_norm": 0.7853586077690125,
+      "learning_rate": 3.9783622652474436e-05,
+      "loss": 0.3986,
+      "step": 824
+    },
+    {
+      "epoch": 0.41687721071248107,
+      "grad_norm": 0.6880490779876709,
+      "learning_rate": 3.975822577710714e-05,
+      "loss": 0.3101,
+      "step": 825
+    },
+    {
+      "epoch": 0.41738251642243557,
+      "grad_norm": 0.686315655708313,
+      "learning_rate": 3.973280550298549e-05,
+      "loss": 0.3909,
+      "step": 826
+    },
+    {
+      "epoch": 0.4178878221323901,
+      "grad_norm": 0.22469963133335114,
+      "learning_rate": 3.970736187041262e-05,
+      "loss": 0.1319,
+      "step": 827
+    },
+    {
+      "epoch": 0.4183931278423446,
+      "grad_norm": 0.5535637736320496,
+      "learning_rate": 3.968189491972869e-05,
+      "loss": 0.3014,
+      "step": 828
+    },
+    {
+      "epoch": 0.41889843355229917,
+      "grad_norm": 0.5633406639099121,
+      "learning_rate": 3.9656404691310846e-05,
+      "loss": 0.3009,
+      "step": 829
+    },
+    {
+      "epoch": 0.41940373926225366,
+      "grad_norm": 0.6593132615089417,
+      "learning_rate": 3.963089122557313e-05,
+      "loss": 0.5059,
+      "step": 830
+    },
+    {
+      "epoch": 0.4199090449722082,
+      "grad_norm": 0.3774155378341675,
+      "learning_rate": 3.960535456296643e-05,
+      "loss": 0.3018,
+      "step": 831
+    },
+    {
+      "epoch": 0.4204143506821627,
+      "grad_norm": 0.39002594351768494,
+      "learning_rate": 3.957979474397841e-05,
+      "loss": 0.3974,
+      "step": 832
+    },
+    {
+      "epoch": 0.42091965639211726,
+      "grad_norm": 0.5461925268173218,
+      "learning_rate": 3.9554211809133446e-05,
+      "loss": 0.3468,
+      "step": 833
+    },
+    {
+      "epoch": 0.42142496210207175,
+      "grad_norm": 0.34177666902542114,
+      "learning_rate": 3.9528605798992566e-05,
+      "loss": 0.1479,
+      "step": 834
+    },
+    {
+      "epoch": 0.42193026781202625,
+      "grad_norm": 0.5321192741394043,
+      "learning_rate": 3.950297675415339e-05,
+      "loss": 0.3468,
+      "step": 835
+    },
+    {
+      "epoch": 0.4224355735219808,
+      "grad_norm": 0.4154065251350403,
+      "learning_rate": 3.9477324715250045e-05,
+      "loss": 0.3419,
+      "step": 836
+    },
+    {
+      "epoch": 0.4229408792319353,
+      "grad_norm": 0.4703141450881958,
+      "learning_rate": 3.945164972295313e-05,
+      "loss": 0.3884,
+      "step": 837
+    },
+    {
+      "epoch": 0.42344618494188985,
+      "grad_norm": 0.43752673268318176,
+      "learning_rate": 3.942595181796961e-05,
+      "loss": 0.2079,
+      "step": 838
+    },
+    {
+      "epoch": 0.42395149065184434,
+      "grad_norm": 0.3551587760448456,
+      "learning_rate": 3.940023104104281e-05,
+      "loss": 0.2126,
+      "step": 839
+    },
+    {
+      "epoch": 0.4244567963617989,
+      "grad_norm": 0.8521201610565186,
+      "learning_rate": 3.9374487432952304e-05,
+      "loss": 0.4272,
+      "step": 840
+    },
+    {
+      "epoch": 0.4249621020717534,
+      "grad_norm": 0.477527916431427,
+      "learning_rate": 3.934872103451386e-05,
+      "loss": 0.2675,
+      "step": 841
+    },
+    {
+      "epoch": 0.42546740778170794,
+      "grad_norm": 0.46821895241737366,
+      "learning_rate": 3.932293188657937e-05,
+      "loss": 0.2343,
+      "step": 842
+    },
+    {
+      "epoch": 0.42597271349166244,
+      "grad_norm": 0.7220411896705627,
+      "learning_rate": 3.929712003003681e-05,
+      "loss": 0.4072,
+      "step": 843
+    },
+    {
+      "epoch": 0.426478019201617,
+      "grad_norm": 0.7784841656684875,
+      "learning_rate": 3.927128550581019e-05,
+      "loss": 0.5188,
+      "step": 844
+    },
+    {
+      "epoch": 0.4269833249115715,
+      "grad_norm": 0.8209807872772217,
+      "learning_rate": 3.924542835485938e-05,
+      "loss": 0.6626,
+      "step": 845
+    },
+    {
+      "epoch": 0.42748863062152603,
+      "grad_norm": 0.4226566553115845,
+      "learning_rate": 3.9219548618180194e-05,
+      "loss": 0.2538,
+      "step": 846
+    },
+    {
+      "epoch": 0.42799393633148053,
+      "grad_norm": 0.45771926641464233,
+      "learning_rate": 3.919364633680423e-05,
+      "loss": 0.3554,
+      "step": 847
+    },
+    {
+      "epoch": 0.4284992420414351,
+      "grad_norm": 0.4026903808116913,
+      "learning_rate": 3.916772155179882e-05,
+      "loss": 0.2283,
+      "step": 848
+    },
+    {
+      "epoch": 0.4290045477513896,
+      "grad_norm": 0.474115252494812,
+      "learning_rate": 3.9141774304266995e-05,
+      "loss": 0.3667,
+      "step": 849
+    },
+    {
+      "epoch": 0.4295098534613441,
+      "grad_norm": 0.5068637132644653,
+      "learning_rate": 3.91158046353474e-05,
+      "loss": 0.3598,
+      "step": 850
+    },
+    {
+      "epoch": 0.4300151591712986,
+      "grad_norm": 0.63497394323349,
+      "learning_rate": 3.9089812586214195e-05,
+      "loss": 0.3532,
+      "step": 851
+    },
+    {
+      "epoch": 0.4305204648812532,
+      "grad_norm": 1.019968032836914,
+      "learning_rate": 3.906379819807707e-05,
+      "loss": 0.5262,
+      "step": 852
+    },
+    {
+      "epoch": 0.43102577059120767,
+      "grad_norm": 0.7702900171279907,
+      "learning_rate": 3.903776151218109e-05,
+      "loss": 0.4335,
+      "step": 853
+    },
+    {
+      "epoch": 0.4315310763011622,
+      "grad_norm": 0.4589867889881134,
+      "learning_rate": 3.901170256980671e-05,
+      "loss": 0.1948,
+      "step": 854
+    },
+    {
+      "epoch": 0.4320363820111167,
+      "grad_norm": 0.3241965174674988,
+      "learning_rate": 3.8985621412269644e-05,
+      "loss": 0.3025,
+      "step": 855
+    },
+    {
+      "epoch": 0.43254168772107127,
+      "grad_norm": 0.5903672575950623,
+      "learning_rate": 3.895951808092084e-05,
+      "loss": 0.2925,
+      "step": 856
+    },
+    {
+      "epoch": 0.43304699343102576,
+      "grad_norm": 0.42580774426460266,
+      "learning_rate": 3.893339261714642e-05,
+      "loss": 0.4694,
+      "step": 857
+    },
+    {
+      "epoch": 0.4335522991409803,
+      "grad_norm": 0.4376884698867798,
+      "learning_rate": 3.890724506236753e-05,
+      "loss": 0.2743,
+      "step": 858
+    },
+    {
+      "epoch": 0.4340576048509348,
+      "grad_norm": 0.5972741842269897,
+      "learning_rate": 3.8881075458040426e-05,
+      "loss": 0.3676,
+      "step": 859
+    },
+    {
+      "epoch": 0.43456291056088936,
+      "grad_norm": 0.6178309321403503,
+      "learning_rate": 3.8854883845656275e-05,
+      "loss": 0.3012,
+      "step": 860
+    },
+    {
+      "epoch": 0.43506821627084385,
+      "grad_norm": 0.8123751878738403,
+      "learning_rate": 3.882867026674114e-05,
+      "loss": 0.4476,
+      "step": 861
+    },
+    {
+      "epoch": 0.4355735219807984,
+      "grad_norm": 0.5609534978866577,
+      "learning_rate": 3.880243476285593e-05,
+      "loss": 0.3029,
+      "step": 862
+    },
+    {
+      "epoch": 0.4360788276907529,
+      "grad_norm": 0.5837926268577576,
+      "learning_rate": 3.877617737559629e-05,
+      "loss": 0.4239,
+      "step": 863
+    },
+    {
+      "epoch": 0.43658413340070745,
+      "grad_norm": 0.40057340264320374,
+      "learning_rate": 3.874989814659258e-05,
+      "loss": 0.1811,
+      "step": 864
+    },
+    {
+      "epoch": 0.43708943911066195,
+      "grad_norm": 0.45030128955841064,
+      "learning_rate": 3.872359711750978e-05,
+      "loss": 0.2334,
+      "step": 865
+    },
+    {
+      "epoch": 0.4375947448206165,
+      "grad_norm": 0.3968111574649811,
+      "learning_rate": 3.8697274330047427e-05,
+      "loss": 0.2225,
+      "step": 866
+    },
+    {
+      "epoch": 0.438100050530571,
+      "grad_norm": 0.6887713074684143,
+      "learning_rate": 3.8670929825939586e-05,
+      "loss": 0.5391,
+      "step": 867
+    },
+    {
+      "epoch": 0.4386053562405255,
+      "grad_norm": 1.0266462564468384,
+      "learning_rate": 3.864456364695472e-05,
+      "loss": 0.4051,
+      "step": 868
+    },
+    {
+      "epoch": 0.43911066195048004,
+      "grad_norm": 0.40764105319976807,
+      "learning_rate": 3.8618175834895664e-05,
+      "loss": 0.2342,
+      "step": 869
+    },
+    {
+      "epoch": 0.43961596766043454,
+      "grad_norm": 0.3224446475505829,
+      "learning_rate": 3.8591766431599566e-05,
+      "loss": 0.1845,
+      "step": 870
+    },
+    {
+      "epoch": 0.4401212733703891,
+      "grad_norm": 0.8195884227752686,
+      "learning_rate": 3.856533547893779e-05,
+      "loss": 0.4636,
+      "step": 871
+    },
+    {
+      "epoch": 0.4406265790803436,
+      "grad_norm": 0.4880749583244324,
+      "learning_rate": 3.853888301881587e-05,
+      "loss": 0.3212,
+      "step": 872
+    },
+    {
+      "epoch": 0.44113188479029813,
+      "grad_norm": 0.6332755088806152,
+      "learning_rate": 3.8512409093173453e-05,
+      "loss": 0.3923,
+      "step": 873
+    },
+    {
+      "epoch": 0.44163719050025263,
+      "grad_norm": 0.41636690497398376,
+      "learning_rate": 3.8485913743984206e-05,
+      "loss": 0.2354,
+      "step": 874
+    },
+    {
+      "epoch": 0.4421424962102072,
+      "grad_norm": 0.5678810477256775,
+      "learning_rate": 3.845939701325575e-05,
+      "loss": 0.2633,
+      "step": 875
+    },
+    {
+      "epoch": 0.4426478019201617,
+      "grad_norm": 0.3602360188961029,
+      "learning_rate": 3.8432858943029625e-05,
+      "loss": 0.2014,
+      "step": 876
+    },
+    {
+      "epoch": 0.4431531076301162,
+      "grad_norm": 0.48009565472602844,
+      "learning_rate": 3.840629957538121e-05,
+      "loss": 0.3181,
+      "step": 877
+    },
+    {
+      "epoch": 0.4436584133400707,
+      "grad_norm": 0.3176063001155853,
+      "learning_rate": 3.837971895241964e-05,
+      "loss": 0.1613,
+      "step": 878
+    },
+    {
+      "epoch": 0.4441637190500253,
+      "grad_norm": 0.5749449133872986,
+      "learning_rate": 3.835311711628774e-05,
+      "loss": 0.2763,
+      "step": 879
+    },
+    {
+      "epoch": 0.44466902475997977,
+      "grad_norm": 0.40219807624816895,
+      "learning_rate": 3.832649410916198e-05,
+      "loss": 0.1707,
+      "step": 880
+    },
+    {
+      "epoch": 0.4451743304699343,
+      "grad_norm": 0.6188439130783081,
+      "learning_rate": 3.8299849973252386e-05,
+      "loss": 0.2757,
+      "step": 881
+    },
+    {
+      "epoch": 0.4456796361798888,
+      "grad_norm": 0.6787897348403931,
+      "learning_rate": 3.827318475080252e-05,
+      "loss": 0.2632,
+      "step": 882
+    },
+    {
+      "epoch": 0.44618494188984337,
+      "grad_norm": 0.46739840507507324,
+      "learning_rate": 3.8246498484089306e-05,
+      "loss": 0.2511,
+      "step": 883
+    },
+    {
+      "epoch": 0.44669024759979786,
+      "grad_norm": 0.5028379559516907,
+      "learning_rate": 3.82197912154231e-05,
+      "loss": 0.3607,
+      "step": 884
+    },
+    {
+      "epoch": 0.4471955533097524,
+      "grad_norm": 0.45272600650787354,
+      "learning_rate": 3.819306298714752e-05,
+      "loss": 0.2706,
+      "step": 885
+    },
+    {
+      "epoch": 0.4477008590197069,
+      "grad_norm": 0.37481459975242615,
+      "learning_rate": 3.816631384163944e-05,
+      "loss": 0.1997,
+      "step": 886
+    },
+    {
+      "epoch": 0.44820616472966146,
+      "grad_norm": 0.7351203560829163,
+      "learning_rate": 3.813954382130886e-05,
+      "loss": 0.2668,
+      "step": 887
+    },
+    {
+      "epoch": 0.44871147043961596,
+      "grad_norm": 0.7236170768737793,
+      "learning_rate": 3.8112752968598907e-05,
+      "loss": 0.3759,
+      "step": 888
+    },
+    {
+      "epoch": 0.4492167761495705,
+      "grad_norm": 0.36974865198135376,
+      "learning_rate": 3.808594132598574e-05,
+      "loss": 0.2733,
+      "step": 889
+    },
+    {
+      "epoch": 0.449722081859525,
+      "grad_norm": 0.76992267370224,
+      "learning_rate": 3.805910893597845e-05,
+      "loss": 0.4098,
+      "step": 890
+    },
+    {
+      "epoch": 0.45022738756947955,
+      "grad_norm": 0.31903111934661865,
+      "learning_rate": 3.803225584111907e-05,
+      "loss": 0.1713,
+      "step": 891
+    },
+    {
+      "epoch": 0.45073269327943405,
+      "grad_norm": 0.7338394522666931,
+      "learning_rate": 3.8005382083982386e-05,
+      "loss": 0.7272,
+      "step": 892
+    },
+    {
+      "epoch": 0.4512379989893886,
+      "grad_norm": 0.9718958139419556,
+      "learning_rate": 3.797848770717603e-05,
+      "loss": 0.4103,
+      "step": 893
+    },
+    {
+      "epoch": 0.4517433046993431,
+      "grad_norm": 0.24202537536621094,
+      "learning_rate": 3.795157275334027e-05,
+      "loss": 0.2362,
+      "step": 894
+    },
+    {
+      "epoch": 0.45224861040929765,
+      "grad_norm": 1.0490413904190063,
+      "learning_rate": 3.7924637265148e-05,
+      "loss": 0.4398,
+      "step": 895
+    },
+    {
+      "epoch": 0.45275391611925214,
+      "grad_norm": 0.46118178963661194,
+      "learning_rate": 3.789768128530471e-05,
+      "loss": 0.3123,
+      "step": 896
+    },
+    {
+      "epoch": 0.4532592218292067,
+      "grad_norm": 0.3323919475078583,
+      "learning_rate": 3.787070485654833e-05,
+      "loss": 0.1714,
+      "step": 897
+    },
+    {
+      "epoch": 0.4537645275391612,
+      "grad_norm": 0.8963467478752136,
+      "learning_rate": 3.784370802164925e-05,
+      "loss": 0.5634,
+      "step": 898
+    },
+    {
+      "epoch": 0.45426983324911574,
+      "grad_norm": 0.9990172386169434,
+      "learning_rate": 3.781669082341018e-05,
+      "loss": 0.7088,
+      "step": 899
+    },
+    {
+      "epoch": 0.45477513895907024,
+      "grad_norm": 0.5156543850898743,
+      "learning_rate": 3.7789653304666136e-05,
+      "loss": 0.2425,
+      "step": 900
+    },
+    {
+      "epoch": 0.4552804446690248,
+      "grad_norm": 0.3596484363079071,
+      "learning_rate": 3.776259550828437e-05,
+      "loss": 0.1751,
+      "step": 901
+    },
+    {
+      "epoch": 0.4557857503789793,
+      "grad_norm": 0.7215254902839661,
+      "learning_rate": 3.773551747716423e-05,
+      "loss": 0.373,
+      "step": 902
+    },
+    {
+      "epoch": 0.4562910560889338,
+      "grad_norm": 0.36745965480804443,
+      "learning_rate": 3.77084192542372e-05,
+      "loss": 0.1979,
+      "step": 903
+    },
+    {
+      "epoch": 0.45679636179888833,
+      "grad_norm": 0.3110678791999817,
+      "learning_rate": 3.768130088246674e-05,
+      "loss": 0.1676,
+      "step": 904
+    },
+    {
+      "epoch": 0.4573016675088428,
+      "grad_norm": 0.6911036968231201,
+      "learning_rate": 3.765416240484828e-05,
+      "loss": 0.331,
+      "step": 905
+    },
+    {
+      "epoch": 0.4578069732187974,
+      "grad_norm": 0.48644575476646423,
+      "learning_rate": 3.7627003864409116e-05,
+      "loss": 0.4377,
+      "step": 906
+    },
+    {
+      "epoch": 0.45831227892875187,
+      "grad_norm": 0.5354007482528687,
+      "learning_rate": 3.759982530420835e-05,
+      "loss": 0.2489,
+      "step": 907
+    },
+    {
+      "epoch": 0.4588175846387064,
+      "grad_norm": 0.4465010464191437,
+      "learning_rate": 3.757262676733684e-05,
+      "loss": 0.2772,
+      "step": 908
+    },
+    {
+      "epoch": 0.4593228903486609,
+      "grad_norm": 0.47116124629974365,
+      "learning_rate": 3.754540829691708e-05,
+      "loss": 0.202,
+      "step": 909
+    },
+    {
+      "epoch": 0.45982819605861547,
+      "grad_norm": 0.635790228843689,
+      "learning_rate": 3.751816993610323e-05,
+      "loss": 0.2873,
+      "step": 910
+    },
+    {
+      "epoch": 0.46033350176856996,
+      "grad_norm": 0.3087238371372223,
+      "learning_rate": 3.749091172808093e-05,
+      "loss": 0.19,
+      "step": 911
+    },
+    {
+      "epoch": 0.4608388074785245,
+      "grad_norm": 0.3472057282924652,
+      "learning_rate": 3.74636337160673e-05,
+      "loss": 0.2691,
+      "step": 912
+    },
+    {
+      "epoch": 0.461344113188479,
+      "grad_norm": 0.5002533197402954,
+      "learning_rate": 3.743633594331089e-05,
+      "loss": 0.2269,
+      "step": 913
+    },
+    {
+      "epoch": 0.46184941889843356,
+      "grad_norm": 0.5435811877250671,
+      "learning_rate": 3.740901845309152e-05,
+      "loss": 0.2686,
+      "step": 914
+    },
+    {
+      "epoch": 0.46235472460838806,
+      "grad_norm": 0.5896182060241699,
+      "learning_rate": 3.738168128872033e-05,
+      "loss": 0.3024,
+      "step": 915
+    },
+    {
+      "epoch": 0.4628600303183426,
+      "grad_norm": 0.6780158281326294,
+      "learning_rate": 3.735432449353963e-05,
+      "loss": 0.3342,
+      "step": 916
+    },
+    {
+      "epoch": 0.4633653360282971,
+      "grad_norm": 0.30589649081230164,
+      "learning_rate": 3.7326948110922855e-05,
+      "loss": 0.1874,
+      "step": 917
+    },
+    {
+      "epoch": 0.46387064173825165,
+      "grad_norm": 0.729263424873352,
+      "learning_rate": 3.729955218427449e-05,
+      "loss": 0.2734,
+      "step": 918
+    },
+    {
+      "epoch": 0.46437594744820615,
+      "grad_norm": 0.4440087080001831,
+      "learning_rate": 3.7272136757029995e-05,
+      "loss": 0.2259,
+      "step": 919
+    },
+    {
+      "epoch": 0.4648812531581607,
+      "grad_norm": 0.9144408702850342,
+      "learning_rate": 3.724470187265579e-05,
+      "loss": 0.544,
+      "step": 920
+    },
+    {
+      "epoch": 0.4653865588681152,
+      "grad_norm": 0.6351197957992554,
+      "learning_rate": 3.721724757464911e-05,
+      "loss": 0.3172,
+      "step": 921
+    },
+    {
+      "epoch": 0.46589186457806975,
+      "grad_norm": 0.7643218040466309,
+      "learning_rate": 3.718977390653796e-05,
+      "loss": 0.5287,
+      "step": 922
+    },
+    {
+      "epoch": 0.46639717028802424,
+      "grad_norm": 0.5360459685325623,
+      "learning_rate": 3.7162280911881085e-05,
+      "loss": 0.1788,
+      "step": 923
+    },
+    {
+      "epoch": 0.4669024759979788,
+      "grad_norm": 0.4476550221443176,
+      "learning_rate": 3.713476863426787e-05,
+      "loss": 0.2565,
+      "step": 924
+    },
+    {
+      "epoch": 0.4674077817079333,
+      "grad_norm": 0.3940451443195343,
+      "learning_rate": 3.710723711731825e-05,
+      "loss": 0.2263,
+      "step": 925
+    },
+    {
+      "epoch": 0.46791308741788784,
+      "grad_norm": 0.5123595595359802,
+      "learning_rate": 3.7079686404682674e-05,
+      "loss": 0.2216,
+      "step": 926
+    },
+    {
+      "epoch": 0.46841839312784234,
+      "grad_norm": 0.36884805560112,
+      "learning_rate": 3.7052116540042025e-05,
+      "loss": 0.2212,
+      "step": 927
+    },
+    {
+      "epoch": 0.4689236988377969,
+      "grad_norm": 0.6213935017585754,
+      "learning_rate": 3.7024527567107564e-05,
+      "loss": 0.3156,
+      "step": 928
+    },
+    {
+      "epoch": 0.4694290045477514,
+      "grad_norm": 0.40270182490348816,
+      "learning_rate": 3.6996919529620825e-05,
+      "loss": 0.1813,
+      "step": 929
+    },
+    {
+      "epoch": 0.46993431025770593,
+      "grad_norm": 0.9014922976493835,
+      "learning_rate": 3.696929247135359e-05,
+      "loss": 0.46,
+      "step": 930
+    },
+    {
+      "epoch": 0.47043961596766043,
+      "grad_norm": 0.5056794881820679,
+      "learning_rate": 3.694164643610777e-05,
+      "loss": 0.2321,
+      "step": 931
+    },
+    {
+      "epoch": 0.470944921677615,
+      "grad_norm": 0.47349992394447327,
+      "learning_rate": 3.6913981467715406e-05,
+      "loss": 0.1869,
+      "step": 932
+    },
+    {
+      "epoch": 0.4714502273875695,
+      "grad_norm": 0.7985735535621643,
+      "learning_rate": 3.6886297610038514e-05,
+      "loss": 0.37,
+      "step": 933
+    },
+    {
+      "epoch": 0.471955533097524,
+      "grad_norm": 0.9785382151603699,
+      "learning_rate": 3.6858594906969077e-05,
+      "loss": 0.4063,
+      "step": 934
+    },
+    {
+      "epoch": 0.4724608388074785,
+      "grad_norm": 0.32149001955986023,
+      "learning_rate": 3.6830873402428964e-05,
+      "loss": 0.2423,
+      "step": 935
+    },
+    {
+      "epoch": 0.4729661445174331,
+      "grad_norm": 0.8392232060432434,
+      "learning_rate": 3.6803133140369836e-05,
+      "loss": 0.7566,
+      "step": 936
+    },
+    {
+      "epoch": 0.47347145022738757,
+      "grad_norm": 0.383240669965744,
+      "learning_rate": 3.6775374164773105e-05,
+      "loss": 0.1921,
+      "step": 937
+    },
+    {
+      "epoch": 0.47397675593734206,
+      "grad_norm": 0.6196987628936768,
+      "learning_rate": 3.6747596519649864e-05,
+      "loss": 0.4214,
+      "step": 938
+    },
+    {
+      "epoch": 0.4744820616472966,
+      "grad_norm": 0.5443540811538696,
+      "learning_rate": 3.671980024904078e-05,
+      "loss": 0.2819,
+      "step": 939
+    },
+    {
+      "epoch": 0.4749873673572511,
+      "grad_norm": 0.5839280486106873,
+      "learning_rate": 3.669198539701606e-05,
+      "loss": 0.345,
+      "step": 940
+    },
+    {
+      "epoch": 0.47549267306720566,
+      "grad_norm": 0.43209922313690186,
+      "learning_rate": 3.666415200767538e-05,
+      "loss": 0.2777,
+      "step": 941
+    },
+    {
+      "epoch": 0.47599797877716016,
+      "grad_norm": 0.4791796803474426,
+      "learning_rate": 3.6636300125147804e-05,
+      "loss": 0.2516,
+      "step": 942
+    },
+    {
+      "epoch": 0.4765032844871147,
+      "grad_norm": 0.5791419148445129,
+      "learning_rate": 3.660842979359171e-05,
+      "loss": 0.4067,
+      "step": 943
+    },
+    {
+      "epoch": 0.4770085901970692,
+      "grad_norm": 0.4313831329345703,
+      "learning_rate": 3.6580541057194726e-05,
+      "loss": 0.1971,
+      "step": 944
+    },
+    {
+      "epoch": 0.47751389590702376,
+      "grad_norm": 0.3102171719074249,
+      "learning_rate": 3.655263396017367e-05,
+      "loss": 0.2316,
+      "step": 945
+    },
+    {
+      "epoch": 0.47801920161697825,
+      "grad_norm": 0.31960973143577576,
+      "learning_rate": 3.652470854677446e-05,
+      "loss": 0.2716,
+      "step": 946
+    },
+    {
+      "epoch": 0.4785245073269328,
+      "grad_norm": 0.24606995284557343,
+      "learning_rate": 3.649676486127206e-05,
+      "loss": 0.1721,
+      "step": 947
+    },
+    {
+      "epoch": 0.4790298130368873,
+      "grad_norm": 0.5517093539237976,
+      "learning_rate": 3.64688029479704e-05,
+      "loss": 0.3408,
+      "step": 948
+    },
+    {
+      "epoch": 0.47953511874684185,
+      "grad_norm": 0.6657114028930664,
+      "learning_rate": 3.644082285120231e-05,
+      "loss": 0.5508,
+      "step": 949
+    },
+    {
+      "epoch": 0.48004042445679634,
+      "grad_norm": 0.39712589979171753,
+      "learning_rate": 3.641282461532946e-05,
+      "loss": 0.2008,
+      "step": 950
+    },
+    {
+      "epoch": 0.4805457301667509,
+      "grad_norm": 0.5426539182662964,
+      "learning_rate": 3.6384808284742254e-05,
+      "loss": 0.3242,
+      "step": 951
+    },
+    {
+      "epoch": 0.4810510358767054,
+      "grad_norm": 0.4150180518627167,
+      "learning_rate": 3.635677390385982e-05,
+      "loss": 0.2666,
+      "step": 952
+    },
+    {
+      "epoch": 0.48155634158665994,
+      "grad_norm": 0.3427567780017853,
+      "learning_rate": 3.632872151712988e-05,
+      "loss": 0.2593,
+      "step": 953
+    },
+    {
+      "epoch": 0.48206164729661444,
+      "grad_norm": 0.5839005708694458,
+      "learning_rate": 3.63006511690287e-05,
+      "loss": 0.4472,
+      "step": 954
+    },
+    {
+      "epoch": 0.482566953006569,
+      "grad_norm": 0.6474454998970032,
+      "learning_rate": 3.627256290406103e-05,
+      "loss": 0.3743,
+      "step": 955
+    },
+    {
+      "epoch": 0.4830722587165235,
+      "grad_norm": 0.361176073551178,
+      "learning_rate": 3.624445676676004e-05,
+      "loss": 0.2014,
+      "step": 956
+    },
+    {
+      "epoch": 0.48357756442647803,
+      "grad_norm": 0.6297400593757629,
+      "learning_rate": 3.621633280168723e-05,
+      "loss": 0.5208,
+      "step": 957
+    },
+    {
+      "epoch": 0.48408287013643253,
+      "grad_norm": 0.38691607117652893,
+      "learning_rate": 3.618819105343234e-05,
+      "loss": 0.3079,
+      "step": 958
+    },
+    {
+      "epoch": 0.4845881758463871,
+      "grad_norm": 0.48631709814071655,
+      "learning_rate": 3.6160031566613336e-05,
+      "loss": 0.3092,
+      "step": 959
+    },
+    {
+      "epoch": 0.4850934815563416,
+      "grad_norm": 0.30502668023109436,
+      "learning_rate": 3.6131854385876314e-05,
+      "loss": 0.2976,
+      "step": 960
+    },
+    {
+      "epoch": 0.48559878726629613,
+      "grad_norm": 0.6193214654922485,
+      "learning_rate": 3.6103659555895376e-05,
+      "loss": 0.3162,
+      "step": 961
+    },
+    {
+      "epoch": 0.4861040929762506,
+      "grad_norm": 0.5869472026824951,
+      "learning_rate": 3.607544712137265e-05,
+      "loss": 0.5059,
+      "step": 962
+    },
+    {
+      "epoch": 0.4866093986862052,
+      "grad_norm": 0.6832082867622375,
+      "learning_rate": 3.604721712703817e-05,
+      "loss": 0.2766,
+      "step": 963
+    },
+    {
+      "epoch": 0.48711470439615967,
+      "grad_norm": 0.8644431829452515,
+      "learning_rate": 3.601896961764978e-05,
+      "loss": 0.632,
+      "step": 964
+    },
+    {
+      "epoch": 0.4876200101061142,
+      "grad_norm": 0.7030072212219238,
+      "learning_rate": 3.599070463799315e-05,
+      "loss": 0.4698,
+      "step": 965
+    },
+    {
+      "epoch": 0.4881253158160687,
+      "grad_norm": 0.49395883083343506,
+      "learning_rate": 3.5962422232881585e-05,
+      "loss": 0.1997,
+      "step": 966
+    },
+    {
+      "epoch": 0.48863062152602327,
+      "grad_norm": 0.3058975338935852,
+      "learning_rate": 3.5934122447156064e-05,
+      "loss": 0.1957,
+      "step": 967
+    },
+    {
+      "epoch": 0.48913592723597776,
+      "grad_norm": 0.6280907392501831,
+      "learning_rate": 3.59058053256851e-05,
+      "loss": 0.2984,
+      "step": 968
+    },
+    {
+      "epoch": 0.4896412329459323,
+      "grad_norm": 0.3592299818992615,
+      "learning_rate": 3.5877470913364695e-05,
+      "loss": 0.1573,
+      "step": 969
+    },
+    {
+      "epoch": 0.4901465386558868,
+      "grad_norm": 0.3483614921569824,
+      "learning_rate": 3.5849119255118274e-05,
+      "loss": 0.1568,
+      "step": 970
+    },
+    {
+      "epoch": 0.49065184436584136,
+      "grad_norm": 0.2722894847393036,
+      "learning_rate": 3.58207503958966e-05,
+      "loss": 0.1799,
+      "step": 971
+    },
+    {
+      "epoch": 0.49115715007579586,
+      "grad_norm": 0.4917094111442566,
+      "learning_rate": 3.57923643806777e-05,
+      "loss": 0.2378,
+      "step": 972
+    },
+    {
+      "epoch": 0.49166245578575035,
+      "grad_norm": 0.36408787965774536,
+      "learning_rate": 3.57639612544668e-05,
+      "loss": 0.1882,
+      "step": 973
+    },
+    {
+      "epoch": 0.4921677614957049,
+      "grad_norm": 0.6604106426239014,
+      "learning_rate": 3.573554106229629e-05,
+      "loss": 0.5856,
+      "step": 974
+    },
+    {
+      "epoch": 0.4926730672056594,
+      "grad_norm": 0.6453742384910583,
+      "learning_rate": 3.5707103849225554e-05,
+      "loss": 0.5509,
+      "step": 975
+    },
+    {
+      "epoch": 0.49317837291561395,
+      "grad_norm": 0.431162565946579,
+      "learning_rate": 3.5678649660341026e-05,
+      "loss": 0.2175,
+      "step": 976
+    },
+    {
+      "epoch": 0.49368367862556845,
+      "grad_norm": 0.510044515132904,
+      "learning_rate": 3.565017854075602e-05,
+      "loss": 0.2819,
+      "step": 977
+    },
+    {
+      "epoch": 0.494188984335523,
+      "grad_norm": 0.6410518288612366,
+      "learning_rate": 3.5621690535610684e-05,
+      "loss": 0.3669,
+      "step": 978
+    },
+    {
+      "epoch": 0.4946942900454775,
+      "grad_norm": 0.2983573377132416,
+      "learning_rate": 3.559318569007198e-05,
+      "loss": 0.2469,
+      "step": 979
+    },
+    {
+      "epoch": 0.49519959575543204,
+      "grad_norm": 0.4974038898944855,
+      "learning_rate": 3.5564664049333527e-05,
+      "loss": 0.2096,
+      "step": 980
+    },
+    {
+      "epoch": 0.49570490146538654,
+      "grad_norm": 0.685854434967041,
+      "learning_rate": 3.5536125658615586e-05,
+      "loss": 0.2544,
+      "step": 981
+    },
+    {
+      "epoch": 0.4962102071753411,
+      "grad_norm": 0.5799931287765503,
+      "learning_rate": 3.5507570563164996e-05,
+      "loss": 0.3086,
+      "step": 982
+    },
+    {
+      "epoch": 0.4967155128852956,
+      "grad_norm": 0.2882770597934723,
+      "learning_rate": 3.5478998808255045e-05,
+      "loss": 0.2204,
+      "step": 983
+    },
+    {
+      "epoch": 0.49722081859525014,
+      "grad_norm": 0.40922459959983826,
+      "learning_rate": 3.545041043918546e-05,
+      "loss": 0.1892,
+      "step": 984
+    },
+    {
+      "epoch": 0.49772612430520463,
+      "grad_norm": 0.3756551146507263,
+      "learning_rate": 3.542180550128229e-05,
+      "loss": 0.2262,
+      "step": 985
+    },
+    {
+      "epoch": 0.4982314300151592,
+      "grad_norm": 0.8342236280441284,
+      "learning_rate": 3.5393184039897876e-05,
+      "loss": 0.4221,
+      "step": 986
+    },
+    {
+      "epoch": 0.4987367357251137,
+      "grad_norm": 0.7572706341743469,
+      "learning_rate": 3.5364546100410744e-05,
+      "loss": 0.3071,
+      "step": 987
+    },
+    {
+      "epoch": 0.49924204143506823,
+      "grad_norm": 0.44951027631759644,
+      "learning_rate": 3.533589172822554e-05,
+      "loss": 0.3021,
+      "step": 988
+    },
+    {
+      "epoch": 0.4997473471450227,
+      "grad_norm": 0.5929198265075684,
+      "learning_rate": 3.530722096877298e-05,
+      "loss": 0.3193,
+      "step": 989
+    },
+    {
+      "epoch": 0.5002526528549772,
+      "grad_norm": 0.6089531779289246,
+      "learning_rate": 3.527853386750974e-05,
+      "loss": 0.3985,
+      "step": 990
+    },
+    {
+      "epoch": 0.5007579585649318,
+      "grad_norm": 0.6241928935050964,
+      "learning_rate": 3.524983046991842e-05,
+      "loss": 0.4037,
+      "step": 991
+    },
+    {
+      "epoch": 0.5012632642748863,
+      "grad_norm": 0.5915480852127075,
+      "learning_rate": 3.522111082150746e-05,
+      "loss": 0.3821,
+      "step": 992
+    },
+    {
+      "epoch": 0.5017685699848409,
+      "grad_norm": 0.956665575504303,
+      "learning_rate": 3.519237496781106e-05,
+      "loss": 0.5615,
+      "step": 993
+    },
+    {
+      "epoch": 0.5022738756947953,
+      "grad_norm": 0.7047909498214722,
+      "learning_rate": 3.5163622954389105e-05,
+      "loss": 0.3062,
+      "step": 994
+    },
+    {
+      "epoch": 0.5027791814047499,
+      "grad_norm": 0.31413036584854126,
+      "learning_rate": 3.5134854826827114e-05,
+      "loss": 0.2965,
+      "step": 995
+    },
+    {
+      "epoch": 0.5032844871147044,
+      "grad_norm": 0.6101884245872498,
+      "learning_rate": 3.510607063073614e-05,
+      "loss": 0.4699,
+      "step": 996
+    },
+    {
+      "epoch": 0.503789792824659,
+      "grad_norm": 0.6464351415634155,
+      "learning_rate": 3.5077270411752746e-05,
+      "loss": 0.3672,
+      "step": 997
+    },
+    {
+      "epoch": 0.5042950985346134,
+      "grad_norm": 0.5930270552635193,
+      "learning_rate": 3.504845421553884e-05,
+      "loss": 0.1938,
+      "step": 998
+    },
+    {
+      "epoch": 0.504800404244568,
+      "grad_norm": 0.614264190196991,
+      "learning_rate": 3.501962208778172e-05,
+      "loss": 0.3049,
+      "step": 999
+    },
+    {
+      "epoch": 0.5053057099545225,
+      "grad_norm": 0.5080682635307312,
+      "learning_rate": 3.4990774074193906e-05,
+      "loss": 0.3748,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5058110156644771,
+      "grad_norm": 0.30347660183906555,
+      "learning_rate": 3.496191022051312e-05,
+      "loss": 0.2212,
+      "step": 1001
+    },
+    {
+      "epoch": 0.5063163213744315,
+      "grad_norm": 0.2643427848815918,
+      "learning_rate": 3.49330305725022e-05,
+      "loss": 0.1181,
+      "step": 1002
+    },
+    {
+      "epoch": 0.506821627084386,
+      "grad_norm": 0.6040931940078735,
+      "learning_rate": 3.490413517594901e-05,
+      "loss": 0.4396,
+      "step": 1003
+    },
+    {
+      "epoch": 0.5073269327943406,
+      "grad_norm": 0.4380238652229309,
+      "learning_rate": 3.487522407666641e-05,
+      "loss": 0.2364,
+      "step": 1004
+    },
+    {
+      "epoch": 0.5078322385042952,
+      "grad_norm": 1.0693117380142212,
+      "learning_rate": 3.4846297320492124e-05,
+      "loss": 0.6679,
+      "step": 1005
+    },
+    {
+      "epoch": 0.5083375442142496,
+      "grad_norm": 0.4763444662094116,
+      "learning_rate": 3.4817354953288715e-05,
+      "loss": 0.3052,
+      "step": 1006
+    },
+    {
+      "epoch": 0.5088428499242041,
+      "grad_norm": 0.41242119669914246,
+      "learning_rate": 3.478839702094351e-05,
+      "loss": 0.2532,
+      "step": 1007
+    },
+    {
+      "epoch": 0.5093481556341587,
+      "grad_norm": 0.41044360399246216,
+      "learning_rate": 3.47594235693685e-05,
+      "loss": 0.4045,
+      "step": 1008
+    },
+    {
+      "epoch": 0.5098534613441131,
+      "grad_norm": 0.5892174243927002,
+      "learning_rate": 3.473043464450027e-05,
+      "loss": 0.3431,
+      "step": 1009
+    },
+    {
+      "epoch": 0.5103587670540677,
+      "grad_norm": 0.5369464755058289,
+      "learning_rate": 3.470143029229996e-05,
+      "loss": 0.2737,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5108640727640222,
+      "grad_norm": 0.39442282915115356,
+      "learning_rate": 3.4672410558753155e-05,
+      "loss": 0.2475,
+      "step": 1011
+    },
+    {
+      "epoch": 0.5113693784739768,
+      "grad_norm": 0.4975070059299469,
+      "learning_rate": 3.464337548986983e-05,
+      "loss": 0.2419,
+      "step": 1012
+    },
+    {
+      "epoch": 0.5118746841839312,
+      "grad_norm": 0.6060254573822021,
+      "learning_rate": 3.4614325131684275e-05,
+      "loss": 0.387,
+      "step": 1013
+    },
+    {
+      "epoch": 0.5123799898938858,
+      "grad_norm": 0.3599434196949005,
+      "learning_rate": 3.458525953025503e-05,
+      "loss": 0.1805,
+      "step": 1014
+    },
+    {
+      "epoch": 0.5128852956038403,
+      "grad_norm": 0.457060843706131,
+      "learning_rate": 3.4556178731664776e-05,
+      "loss": 0.2104,
+      "step": 1015
+    },
+    {
+      "epoch": 0.5133906013137949,
+      "grad_norm": 0.6781561970710754,
+      "learning_rate": 3.4527082782020323e-05,
+      "loss": 0.4543,
+      "step": 1016
+    },
+    {
+      "epoch": 0.5138959070237493,
+      "grad_norm": 0.4415772259235382,
+      "learning_rate": 3.449797172745247e-05,
+      "loss": 0.2245,
+      "step": 1017
+    },
+    {
+      "epoch": 0.5144012127337039,
+      "grad_norm": 0.3077252507209778,
+      "learning_rate": 3.4468845614116e-05,
+      "loss": 0.2002,
+      "step": 1018
+    },
+    {
+      "epoch": 0.5149065184436584,
+      "grad_norm": 0.7845990657806396,
+      "learning_rate": 3.443970448818954e-05,
+      "loss": 0.5271,
+      "step": 1019
+    },
+    {
+      "epoch": 0.515411824153613,
+      "grad_norm": 0.44302085041999817,
+      "learning_rate": 3.441054839587553e-05,
+      "loss": 0.1878,
+      "step": 1020
+    },
+    {
+      "epoch": 0.5159171298635674,
+      "grad_norm": 1.1317225694656372,
+      "learning_rate": 3.438137738340014e-05,
+      "loss": 0.4688,
+      "step": 1021
+    },
+    {
+      "epoch": 0.516422435573522,
+      "grad_norm": 0.7646364569664001,
+      "learning_rate": 3.435219149701319e-05,
+      "loss": 0.4279,
+      "step": 1022
+    },
+    {
+      "epoch": 0.5169277412834765,
+      "grad_norm": 0.3174137771129608,
+      "learning_rate": 3.43229907829881e-05,
+      "loss": 0.207,
+      "step": 1023
+    },
+    {
+      "epoch": 0.5174330469934311,
+      "grad_norm": 0.25344839692115784,
+      "learning_rate": 3.429377528762177e-05,
+      "loss": 0.174,
+      "step": 1024
+    },
+    {
+      "epoch": 0.5179383527033855,
+      "grad_norm": 0.4643212854862213,
+      "learning_rate": 3.4264545057234565e-05,
+      "loss": 0.3102,
+      "step": 1025
+    },
+    {
+      "epoch": 0.5184436584133401,
+      "grad_norm": 0.3090815842151642,
+      "learning_rate": 3.423530013817019e-05,
+      "loss": 0.1688,
+      "step": 1026
+    },
+    {
+      "epoch": 0.5189489641232946,
+      "grad_norm": 0.7463130950927734,
+      "learning_rate": 3.4206040576795644e-05,
+      "loss": 0.6639,
+      "step": 1027
+    },
+    {
+      "epoch": 0.5194542698332492,
+      "grad_norm": 0.4260081648826599,
+      "learning_rate": 3.417676641950115e-05,
+      "loss": 0.1466,
+      "step": 1028
+    },
+    {
+      "epoch": 0.5199595755432036,
+      "grad_norm": 0.8411809802055359,
+      "learning_rate": 3.414747771270006e-05,
+      "loss": 0.5472,
+      "step": 1029
+    },
+    {
+      "epoch": 0.5204648812531582,
+      "grad_norm": 0.590641438961029,
+      "learning_rate": 3.411817450282881e-05,
+      "loss": 0.402,
+      "step": 1030
+    },
+    {
+      "epoch": 0.5209701869631127,
+      "grad_norm": 0.9213524460792542,
+      "learning_rate": 3.408885683634681e-05,
+      "loss": 0.5014,
+      "step": 1031
+    },
+    {
+      "epoch": 0.5214754926730673,
+      "grad_norm": 0.3281521201133728,
+      "learning_rate": 3.4059524759736404e-05,
+      "loss": 0.2153,
+      "step": 1032
+    },
+    {
+      "epoch": 0.5219807983830217,
+      "grad_norm": 0.4002426266670227,
+      "learning_rate": 3.4030178319502784e-05,
+      "loss": 0.2112,
+      "step": 1033
+    },
+    {
+      "epoch": 0.5224861040929762,
+      "grad_norm": 0.27855736017227173,
+      "learning_rate": 3.40008175621739e-05,
+      "loss": 0.1904,
+      "step": 1034
+    },
+    {
+      "epoch": 0.5229914098029308,
+      "grad_norm": 1.192247986793518,
+      "learning_rate": 3.3971442534300414e-05,
+      "loss": 0.5124,
+      "step": 1035
+    },
+    {
+      "epoch": 0.5234967155128853,
+      "grad_norm": 0.3181244730949402,
+      "learning_rate": 3.3942053282455625e-05,
+      "loss": 0.2413,
+      "step": 1036
+    },
+    {
+      "epoch": 0.5240020212228398,
+      "grad_norm": 0.5175408720970154,
+      "learning_rate": 3.3912649853235355e-05,
+      "loss": 0.2505,
+      "step": 1037
+    },
+    {
+      "epoch": 0.5245073269327943,
+      "grad_norm": 0.7706686854362488,
+      "learning_rate": 3.3883232293257926e-05,
+      "loss": 0.4528,
+      "step": 1038
+    },
+    {
+      "epoch": 0.5250126326427489,
+      "grad_norm": 0.6226242184638977,
+      "learning_rate": 3.385380064916405e-05,
+      "loss": 0.2895,
+      "step": 1039
+    },
+    {
+      "epoch": 0.5255179383527034,
+      "grad_norm": 0.6919043064117432,
+      "learning_rate": 3.382435496761679e-05,
+      "loss": 0.5499,
+      "step": 1040
+    },
+    {
+      "epoch": 0.5260232440626579,
+      "grad_norm": 0.7153246402740479,
+      "learning_rate": 3.379489529530144e-05,
+      "loss": 0.4498,
+      "step": 1041
+    },
+    {
+      "epoch": 0.5265285497726124,
+      "grad_norm": 0.3823365271091461,
+      "learning_rate": 3.376542167892548e-05,
+      "loss": 0.3495,
+      "step": 1042
+    },
+    {
+      "epoch": 0.527033855482567,
+      "grad_norm": 0.7256558537483215,
+      "learning_rate": 3.373593416521852e-05,
+      "loss": 0.4197,
+      "step": 1043
+    },
+    {
+      "epoch": 0.5275391611925214,
+      "grad_norm": 0.6454598903656006,
+      "learning_rate": 3.3706432800932184e-05,
+      "loss": 0.4144,
+      "step": 1044
+    },
+    {
+      "epoch": 0.528044466902476,
+      "grad_norm": 0.3100246489048004,
+      "learning_rate": 3.367691763284005e-05,
+      "loss": 0.1587,
+      "step": 1045
+    },
+    {
+      "epoch": 0.5285497726124305,
+      "grad_norm": 0.44669437408447266,
+      "learning_rate": 3.3647388707737606e-05,
+      "loss": 0.2715,
+      "step": 1046
+    },
+    {
+      "epoch": 0.5290550783223851,
+      "grad_norm": 0.31795749068260193,
+      "learning_rate": 3.3617846072442135e-05,
+      "loss": 0.2045,
+      "step": 1047
+    },
+    {
+      "epoch": 0.5295603840323395,
+      "grad_norm": 0.34143051505088806,
+      "learning_rate": 3.3588289773792646e-05,
+      "loss": 0.2189,
+      "step": 1048
+    },
+    {
+      "epoch": 0.5300656897422941,
+      "grad_norm": 0.5310137271881104,
+      "learning_rate": 3.3558719858649834e-05,
+      "loss": 0.2596,
+      "step": 1049
+    },
+    {
+      "epoch": 0.5305709954522486,
+      "grad_norm": 0.7323408722877502,
+      "learning_rate": 3.352913637389598e-05,
+      "loss": 0.3134,
+      "step": 1050
+    },
+    {
+      "epoch": 0.5310763011622032,
+      "grad_norm": 0.3827390968799591,
+      "learning_rate": 3.349953936643486e-05,
+      "loss": 0.2751,
+      "step": 1051
+    },
+    {
+      "epoch": 0.5315816068721576,
+      "grad_norm": 0.701563835144043,
+      "learning_rate": 3.3469928883191706e-05,
+      "loss": 0.3593,
+      "step": 1052
+    },
+    {
+      "epoch": 0.5320869125821122,
+      "grad_norm": 0.68122398853302,
+      "learning_rate": 3.344030497111312e-05,
+      "loss": 0.6736,
+      "step": 1053
+    },
+    {
+      "epoch": 0.5325922182920667,
+      "grad_norm": 0.47403842210769653,
+      "learning_rate": 3.3410667677166966e-05,
+      "loss": 0.3286,
+      "step": 1054
+    },
+    {
+      "epoch": 0.5330975240020213,
+      "grad_norm": 0.24394340813159943,
+      "learning_rate": 3.338101704834236e-05,
+      "loss": 0.161,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5336028297119757,
+      "grad_norm": 0.6971885561943054,
+      "learning_rate": 3.3351353131649546e-05,
+      "loss": 0.2709,
+      "step": 1056
+    },
+    {
+      "epoch": 0.5341081354219303,
+      "grad_norm": 0.7203572988510132,
+      "learning_rate": 3.332167597411983e-05,
+      "loss": 0.3845,
+      "step": 1057
+    },
+    {
+      "epoch": 0.5346134411318848,
+      "grad_norm": 0.8475279211997986,
+      "learning_rate": 3.3291985622805516e-05,
+      "loss": 0.436,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5351187468418394,
+      "grad_norm": 0.46801191568374634,
+      "learning_rate": 3.326228212477982e-05,
+      "loss": 0.2384,
+      "step": 1059
+    },
+    {
+      "epoch": 0.5356240525517938,
+      "grad_norm": 0.4841732680797577,
+      "learning_rate": 3.3232565527136815e-05,
+      "loss": 0.3124,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5361293582617483,
+      "grad_norm": 0.2917931377887726,
+      "learning_rate": 3.3202835876991346e-05,
+      "loss": 0.178,
+      "step": 1061
+    },
+    {
+      "epoch": 0.5366346639717029,
+      "grad_norm": 0.5136047005653381,
+      "learning_rate": 3.317309322147891e-05,
+      "loss": 0.2269,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5371399696816574,
+      "grad_norm": 0.504074215888977,
+      "learning_rate": 3.3143337607755694e-05,
+      "loss": 0.2895,
+      "step": 1063
+    },
+    {
+      "epoch": 0.5376452753916119,
+      "grad_norm": 0.44980528950691223,
+      "learning_rate": 3.311356908299836e-05,
+      "loss": 0.346,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5381505811015664,
+      "grad_norm": 0.45925334095954895,
+      "learning_rate": 3.30837876944041e-05,
+      "loss": 0.2025,
+      "step": 1065
+    },
+    {
+      "epoch": 0.538655886811521,
+      "grad_norm": 0.48478057980537415,
+      "learning_rate": 3.305399348919045e-05,
+      "loss": 0.3748,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5391611925214755,
+      "grad_norm": 0.376342236995697,
+      "learning_rate": 3.3024186514595294e-05,
+      "loss": 0.2637,
+      "step": 1067
+    },
+    {
+      "epoch": 0.53966649823143,
+      "grad_norm": 0.8223666548728943,
+      "learning_rate": 3.299436681787677e-05,
+      "loss": 0.3113,
+      "step": 1068
+    },
+    {
+      "epoch": 0.5401718039413845,
+      "grad_norm": 0.3819112181663513,
+      "learning_rate": 3.296453444631316e-05,
+      "loss": 0.2424,
+      "step": 1069
+    },
+    {
+      "epoch": 0.5406771096513391,
+      "grad_norm": 0.5420382022857666,
+      "learning_rate": 3.293468944720287e-05,
+      "loss": 0.2229,
+      "step": 1070
+    },
+    {
+      "epoch": 0.5411824153612936,
+      "grad_norm": 0.497595876455307,
+      "learning_rate": 3.290483186786428e-05,
+      "loss": 0.2529,
+      "step": 1071
+    },
+    {
+      "epoch": 0.5416877210712481,
+      "grad_norm": 0.7739394307136536,
+      "learning_rate": 3.287496175563578e-05,
+      "loss": 0.3402,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5421930267812026,
+      "grad_norm": 0.5213684439659119,
+      "learning_rate": 3.284507915787559e-05,
+      "loss": 0.2469,
+      "step": 1073
+    },
+    {
+      "epoch": 0.5426983324911572,
+      "grad_norm": 0.45709890127182007,
+      "learning_rate": 3.281518412196172e-05,
+      "loss": 0.446,
+      "step": 1074
+    },
+    {
+      "epoch": 0.5432036382011117,
+      "grad_norm": 0.47826409339904785,
+      "learning_rate": 3.2785276695291935e-05,
+      "loss": 0.3853,
+      "step": 1075
+    },
+    {
+      "epoch": 0.5437089439110662,
+      "grad_norm": 0.5849752426147461,
+      "learning_rate": 3.27553569252836e-05,
+      "loss": 0.3097,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5442142496210207,
+      "grad_norm": 0.3121519982814789,
+      "learning_rate": 3.272542485937369e-05,
+      "loss": 0.1677,
+      "step": 1077
+    },
+    {
+      "epoch": 0.5447195553309753,
+      "grad_norm": 0.5105435252189636,
+      "learning_rate": 3.269548054501865e-05,
+      "loss": 0.2058,
+      "step": 1078
+    },
+    {
+      "epoch": 0.5452248610409297,
+      "grad_norm": 0.5398941040039062,
+      "learning_rate": 3.2665524029694366e-05,
+      "loss": 0.3785,
+      "step": 1079
+    },
+    {
+      "epoch": 0.5457301667508843,
+      "grad_norm": 0.5803938508033752,
+      "learning_rate": 3.263555536089604e-05,
+      "loss": 0.4135,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5462354724608388,
+      "grad_norm": 0.4731886088848114,
+      "learning_rate": 3.260557458613817e-05,
+      "loss": 0.3249,
+      "step": 1081
+    },
+    {
+      "epoch": 0.5467407781707934,
+      "grad_norm": 0.29429805278778076,
+      "learning_rate": 3.257558175295443e-05,
+      "loss": 0.1736,
+      "step": 1082
+    },
+    {
+      "epoch": 0.5472460838807478,
+      "grad_norm": 0.4236926734447479,
+      "learning_rate": 3.254557690889762e-05,
+      "loss": 0.2494,
+      "step": 1083
+    },
+    {
+      "epoch": 0.5477513895907024,
+      "grad_norm": 0.5632703900337219,
+      "learning_rate": 3.251556010153958e-05,
+      "loss": 0.3988,
+      "step": 1084
+    },
+    {
+      "epoch": 0.5482566953006569,
+      "grad_norm": 0.590407133102417,
+      "learning_rate": 3.248553137847112e-05,
+      "loss": 0.3471,
+      "step": 1085
+    },
+    {
+      "epoch": 0.5487620010106115,
+      "grad_norm": 0.5127763748168945,
+      "learning_rate": 3.245549078730195e-05,
+      "loss": 0.3066,
+      "step": 1086
+    },
+    {
+      "epoch": 0.5492673067205659,
+      "grad_norm": 0.5116364359855652,
+      "learning_rate": 3.242543837566057e-05,
+      "loss": 0.5509,
+      "step": 1087
+    },
+    {
+      "epoch": 0.5497726124305204,
+      "grad_norm": 0.40164387226104736,
+      "learning_rate": 3.239537419119425e-05,
+      "loss": 0.1977,
+      "step": 1088
+    },
+    {
+      "epoch": 0.550277918140475,
+      "grad_norm": 0.646152675151825,
+      "learning_rate": 3.236529828156891e-05,
+      "loss": 0.4706,
+      "step": 1089
+    },
+    {
+      "epoch": 0.5507832238504295,
+      "grad_norm": 0.5561092495918274,
+      "learning_rate": 3.233521069446907e-05,
+      "loss": 0.1772,
+      "step": 1090
+    },
+    {
+      "epoch": 0.551288529560384,
+      "grad_norm": 0.5761879682540894,
+      "learning_rate": 3.230511147759775e-05,
+      "loss": 0.2085,
+      "step": 1091
+    },
+    {
+      "epoch": 0.5517938352703385,
+      "grad_norm": 0.4727552533149719,
+      "learning_rate": 3.227500067867642e-05,
+      "loss": 0.2691,
+      "step": 1092
+    },
+    {
+      "epoch": 0.5522991409802931,
+      "grad_norm": 0.9213289618492126,
+      "learning_rate": 3.2244878345444905e-05,
+      "loss": 0.4643,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5528044466902476,
+      "grad_norm": 0.8000922203063965,
+      "learning_rate": 3.221474452566133e-05,
+      "loss": 0.4779,
+      "step": 1094
+    },
+    {
+      "epoch": 0.5533097524002021,
+      "grad_norm": 0.3150307238101959,
+      "learning_rate": 3.218459926710202e-05,
+      "loss": 0.1698,
+      "step": 1095
+    },
+    {
+      "epoch": 0.5538150581101566,
+      "grad_norm": 0.5562626123428345,
+      "learning_rate": 3.215444261756144e-05,
+      "loss": 0.2622,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5543203638201112,
+      "grad_norm": 0.576187014579773,
+      "learning_rate": 3.212427462485212e-05,
+      "loss": 0.2248,
+      "step": 1097
+    },
+    {
+      "epoch": 0.5548256695300657,
+      "grad_norm": 0.4368909001350403,
+      "learning_rate": 3.209409533680457e-05,
+      "loss": 0.2079,
+      "step": 1098
+    },
+    {
+      "epoch": 0.5553309752400202,
+      "grad_norm": 0.5242764353752136,
+      "learning_rate": 3.206390480126718e-05,
+      "loss": 0.3002,
+      "step": 1099
+    },
+    {
+      "epoch": 0.5558362809499747,
+      "grad_norm": 0.5973017811775208,
+      "learning_rate": 3.203370306610624e-05,
+      "loss": 0.2916,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5563415866599293,
+      "grad_norm": 0.5519207119941711,
+      "learning_rate": 3.200349017920573e-05,
+      "loss": 0.2784,
+      "step": 1101
+    },
+    {
+      "epoch": 0.5568468923698838,
+      "grad_norm": 0.5289061069488525,
+      "learning_rate": 3.1973266188467346e-05,
+      "loss": 0.2497,
+      "step": 1102
+    },
+    {
+      "epoch": 0.5573521980798383,
+      "grad_norm": 0.5926082134246826,
+      "learning_rate": 3.1943031141810385e-05,
+      "loss": 0.3212,
+      "step": 1103
+    },
+    {
+      "epoch": 0.5578575037897928,
+      "grad_norm": 0.5896833539009094,
+      "learning_rate": 3.191278508717166e-05,
+      "loss": 0.3915,
+      "step": 1104
+    },
+    {
+      "epoch": 0.5583628094997474,
+      "grad_norm": 0.6870902180671692,
+      "learning_rate": 3.188252807250545e-05,
+      "loss": 0.3646,
+      "step": 1105
+    },
+    {
+      "epoch": 0.5588681152097019,
+      "grad_norm": 0.3682582676410675,
+      "learning_rate": 3.18522601457834e-05,
+      "loss": 0.2704,
+      "step": 1106
+    },
+    {
+      "epoch": 0.5593734209196564,
+      "grad_norm": 0.47527265548706055,
+      "learning_rate": 3.182198135499447e-05,
+      "loss": 0.2499,
+      "step": 1107
+    },
+    {
+      "epoch": 0.5598787266296109,
+      "grad_norm": 0.4023267924785614,
+      "learning_rate": 3.179169174814483e-05,
+      "loss": 0.1694,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5603840323395655,
+      "grad_norm": 0.6233897805213928,
+      "learning_rate": 3.176139137325781e-05,
+      "loss": 0.4744,
+      "step": 1109
+    },
+    {
+      "epoch": 0.56088933804952,
+      "grad_norm": 0.2687748372554779,
+      "learning_rate": 3.173108027837379e-05,
+      "loss": 0.1979,
+      "step": 1110
+    },
+    {
+      "epoch": 0.5613946437594745,
+      "grad_norm": 0.4329420328140259,
+      "learning_rate": 3.1700758511550185e-05,
+      "loss": 0.2352,
+      "step": 1111
+    },
+    {
+      "epoch": 0.561899949469429,
+      "grad_norm": 0.36293160915374756,
+      "learning_rate": 3.1670426120861314e-05,
+      "loss": 0.2382,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5624052551793836,
+      "grad_norm": 0.6872026324272156,
+      "learning_rate": 3.1640083154398306e-05,
+      "loss": 0.3837,
+      "step": 1113
+    },
+    {
+      "epoch": 0.562910560889338,
+      "grad_norm": 0.346775621175766,
+      "learning_rate": 3.160972966026911e-05,
+      "loss": 0.2396,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5634158665992925,
+      "grad_norm": 0.6277110576629639,
+      "learning_rate": 3.1579365686598336e-05,
+      "loss": 0.3142,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5639211723092471,
+      "grad_norm": 0.7550572156906128,
+      "learning_rate": 3.15489912815272e-05,
+      "loss": 0.2231,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5644264780192016,
+      "grad_norm": 0.2056116759777069,
+      "learning_rate": 3.15186064932135e-05,
+      "loss": 0.1063,
+      "step": 1117
+    },
+    {
+      "epoch": 0.5649317837291561,
+      "grad_norm": 0.7212878465652466,
+      "learning_rate": 3.148821136983144e-05,
+      "loss": 0.3422,
+      "step": 1118
+    },
+    {
+      "epoch": 0.5654370894391106,
+      "grad_norm": 0.2987580895423889,
+      "learning_rate": 3.1457805959571665e-05,
+      "loss": 0.2378,
+      "step": 1119
+    },
+    {
+      "epoch": 0.5659423951490652,
+      "grad_norm": 0.45587587356567383,
+      "learning_rate": 3.142739031064108e-05,
+      "loss": 0.2602,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5664477008590197,
+      "grad_norm": 0.943342387676239,
+      "learning_rate": 3.139696447126286e-05,
+      "loss": 0.3641,
+      "step": 1121
+    },
+    {
+      "epoch": 0.5669530065689742,
+      "grad_norm": 0.8414391279220581,
+      "learning_rate": 3.1366528489676315e-05,
+      "loss": 0.5864,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5674583122789287,
+      "grad_norm": 0.5744806528091431,
+      "learning_rate": 3.133608241413685e-05,
+      "loss": 0.3196,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5679636179888833,
+      "grad_norm": 0.7059041857719421,
+      "learning_rate": 3.130562629291586e-05,
+      "loss": 0.4094,
+      "step": 1124
+    },
+    {
+      "epoch": 0.5684689236988378,
+      "grad_norm": 0.4427710175514221,
+      "learning_rate": 3.127516017430067e-05,
+      "loss": 0.2116,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5689742294087923,
+      "grad_norm": 0.8828384280204773,
+      "learning_rate": 3.124468410659448e-05,
+      "loss": 0.3982,
+      "step": 1126
+    },
+    {
+      "epoch": 0.5694795351187468,
+      "grad_norm": 0.9311108589172363,
+      "learning_rate": 3.121419813811622e-05,
+      "loss": 0.4192,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5699848408287014,
+      "grad_norm": 0.8841239809989929,
+      "learning_rate": 3.118370231720055e-05,
+      "loss": 0.4229,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5704901465386559,
+      "grad_norm": 0.549286425113678,
+      "learning_rate": 3.1153196692197745e-05,
+      "loss": 0.3589,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5709954522486104,
+      "grad_norm": 0.4687441885471344,
+      "learning_rate": 3.112268131147361e-05,
+      "loss": 0.2813,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5715007579585649,
+      "grad_norm": 0.5466412305831909,
+      "learning_rate": 3.109215622340944e-05,
+      "loss": 0.2345,
+      "step": 1131
+    },
+    {
+      "epoch": 0.5720060636685195,
+      "grad_norm": 0.39715632796287537,
+      "learning_rate": 3.10616214764019e-05,
+      "loss": 0.2707,
+      "step": 1132
+    },
+    {
+      "epoch": 0.572511369378474,
+      "grad_norm": 0.3774857521057129,
+      "learning_rate": 3.103107711886299e-05,
+      "loss": 0.3072,
+      "step": 1133
+    },
+    {
+      "epoch": 0.5730166750884285,
+      "grad_norm": 0.5280380845069885,
+      "learning_rate": 3.1000523199219925e-05,
+      "loss": 0.2819,
+      "step": 1134
+    },
+    {
+      "epoch": 0.573521980798383,
+      "grad_norm": 0.4964953064918518,
+      "learning_rate": 3.096995976591508e-05,
+      "loss": 0.2877,
+      "step": 1135
+    },
+    {
+      "epoch": 0.5740272865083376,
+      "grad_norm": 0.4315342903137207,
+      "learning_rate": 3.093938686740594e-05,
+      "loss": 0.2174,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5745325922182921,
+      "grad_norm": 0.5171803832054138,
+      "learning_rate": 3.0908804552164985e-05,
+      "loss": 0.4131,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5750378979282466,
+      "grad_norm": 0.43991976976394653,
+      "learning_rate": 3.08782128686796e-05,
+      "loss": 0.2688,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5755432036382011,
+      "grad_norm": 0.45389074087142944,
+      "learning_rate": 3.084761186545206e-05,
+      "loss": 0.3958,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5760485093481557,
+      "grad_norm": 0.4297245740890503,
+      "learning_rate": 3.081700159099939e-05,
+      "loss": 0.2264,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5765538150581102,
+      "grad_norm": 0.6526983380317688,
+      "learning_rate": 3.0786382093853336e-05,
+      "loss": 0.5204,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5770591207680646,
+      "grad_norm": 0.46042600274086,
+      "learning_rate": 3.0755753422560254e-05,
+      "loss": 0.2465,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5775644264780192,
+      "grad_norm": 0.8101718425750732,
+      "learning_rate": 3.0725115625681035e-05,
+      "loss": 0.6224,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5780697321879738,
+      "grad_norm": 0.6587077379226685,
+      "learning_rate": 3.069446875179106e-05,
+      "loss": 0.5301,
+      "step": 1144
+    },
+    {
+      "epoch": 0.5785750378979283,
+      "grad_norm": 0.49807339906692505,
+      "learning_rate": 3.0663812849480075e-05,
+      "loss": 0.2905,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5790803436078827,
+      "grad_norm": 0.514011800289154,
+      "learning_rate": 3.063314796735218e-05,
+      "loss": 0.2377,
+      "step": 1146
+    },
+    {
+      "epoch": 0.5795856493178373,
+      "grad_norm": 0.5293000340461731,
+      "learning_rate": 3.060247415402567e-05,
+      "loss": 0.3341,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5800909550277918,
+      "grad_norm": 0.705506443977356,
+      "learning_rate": 3.057179145813302e-05,
+      "loss": 0.3912,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5805962607377463,
+      "grad_norm": 0.6125894784927368,
+      "learning_rate": 3.0541099928320804e-05,
+      "loss": 0.2844,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5811015664477008,
+      "grad_norm": 0.6330406069755554,
+      "learning_rate": 3.0510399613249556e-05,
+      "loss": 0.3338,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5816068721576554,
+      "grad_norm": 0.6614536046981812,
+      "learning_rate": 3.0479690561593793e-05,
+      "loss": 0.2598,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5821121778676099,
+      "grad_norm": 0.37762120366096497,
+      "learning_rate": 3.0448972822041843e-05,
+      "loss": 0.2355,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5826174835775644,
+      "grad_norm": 0.6358450055122375,
+      "learning_rate": 3.0418246443295823e-05,
+      "loss": 0.2811,
+      "step": 1153
+    },
+    {
+      "epoch": 0.5831227892875189,
+      "grad_norm": 0.3570258617401123,
+      "learning_rate": 3.0387511474071557e-05,
+      "loss": 0.172,
+      "step": 1154
+    },
+    {
+      "epoch": 0.5836280949974735,
+      "grad_norm": 0.3080599308013916,
+      "learning_rate": 3.0356767963098464e-05,
+      "loss": 0.166,
+      "step": 1155
+    },
+    {
+      "epoch": 0.584133400707428,
+      "grad_norm": 0.3466140031814575,
+      "learning_rate": 3.0326015959119535e-05,
+      "loss": 0.1429,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5846387064173825,
+      "grad_norm": 0.4894466996192932,
+      "learning_rate": 3.0295255510891213e-05,
+      "loss": 0.4822,
+      "step": 1157
+    },
+    {
+      "epoch": 0.585144012127337,
+      "grad_norm": 0.4862903952598572,
+      "learning_rate": 3.0264486667183322e-05,
+      "loss": 0.2294,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5856493178372916,
+      "grad_norm": 0.4425124228000641,
+      "learning_rate": 3.023370947677901e-05,
+      "loss": 0.235,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5861546235472461,
+      "grad_norm": 0.4671436846256256,
+      "learning_rate": 3.020292398847464e-05,
+      "loss": 0.2426,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5866599292572006,
+      "grad_norm": 1.0623973608016968,
+      "learning_rate": 3.0172130251079757e-05,
+      "loss": 0.5149,
+      "step": 1161
+    },
+    {
+      "epoch": 0.5871652349671551,
+      "grad_norm": 0.6807923913002014,
+      "learning_rate": 3.014132831341696e-05,
+      "loss": 0.3281,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5876705406771097,
+      "grad_norm": 0.3466711640357971,
+      "learning_rate": 3.0110518224321865e-05,
+      "loss": 0.1801,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5881758463870642,
+      "grad_norm": 0.6598278284072876,
+      "learning_rate": 3.007970003264301e-05,
+      "loss": 0.3268,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5886811520970187,
+      "grad_norm": 0.5974249243736267,
+      "learning_rate": 3.0048873787241765e-05,
+      "loss": 0.2778,
+      "step": 1165
+    },
+    {
+      "epoch": 0.5891864578069732,
+      "grad_norm": 0.35448309779167175,
+      "learning_rate": 3.0018039536992298e-05,
+      "loss": 0.1175,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5896917635169278,
+      "grad_norm": 0.6994403004646301,
+      "learning_rate": 2.9987197330781435e-05,
+      "loss": 0.2628,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5901970692268823,
+      "grad_norm": 0.5090075731277466,
+      "learning_rate": 2.9956347217508633e-05,
+      "loss": 0.2562,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5907023749368367,
+      "grad_norm": 0.29259657859802246,
+      "learning_rate": 2.9925489246085897e-05,
+      "loss": 0.1874,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5912076806467913,
+      "grad_norm": 0.30072706937789917,
+      "learning_rate": 2.989462346543766e-05,
+      "loss": 0.1683,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5917129863567459,
+      "grad_norm": 0.258619099855423,
+      "learning_rate": 2.9863749924500765e-05,
+      "loss": 0.1159,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5922182920667004,
+      "grad_norm": 0.6549366116523743,
+      "learning_rate": 2.9832868672224347e-05,
+      "loss": 0.2978,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5927235977766548,
+      "grad_norm": 0.9457119703292847,
+      "learning_rate": 2.9801979757569774e-05,
+      "loss": 0.3657,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5932289034866094,
+      "grad_norm": 0.40999117493629456,
+      "learning_rate": 2.9771083229510544e-05,
+      "loss": 0.2076,
+      "step": 1174
+    },
+    {
+      "epoch": 0.593734209196564,
+      "grad_norm": 0.629391074180603,
+      "learning_rate": 2.974017913703224e-05,
+      "loss": 0.3435,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5942395149065185,
+      "grad_norm": 0.42947840690612793,
+      "learning_rate": 2.970926752913245e-05,
+      "loss": 0.2741,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5947448206164729,
+      "grad_norm": 0.7551778554916382,
+      "learning_rate": 2.9678348454820648e-05,
+      "loss": 0.2764,
+      "step": 1177
+    },
+    {
+      "epoch": 0.5952501263264275,
+      "grad_norm": 0.4660588800907135,
+      "learning_rate": 2.9647421963118177e-05,
+      "loss": 0.1889,
+      "step": 1178
+    },
+    {
+      "epoch": 0.595755432036382,
+      "grad_norm": 0.5166782140731812,
+      "learning_rate": 2.9616488103058116e-05,
+      "loss": 0.3303,
+      "step": 1179
+    },
+    {
+      "epoch": 0.5962607377463366,
+      "grad_norm": 0.6265774965286255,
+      "learning_rate": 2.9585546923685227e-05,
+      "loss": 0.2055,
+      "step": 1180
+    },
+    {
+      "epoch": 0.596766043456291,
+      "grad_norm": 0.4520041346549988,
+      "learning_rate": 2.9554598474055905e-05,
+      "loss": 0.2479,
+      "step": 1181
+    },
+    {
+      "epoch": 0.5972713491662456,
+      "grad_norm": 0.9512003064155579,
+      "learning_rate": 2.952364280323803e-05,
+      "loss": 0.4837,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5977766548762001,
+      "grad_norm": 0.33608609437942505,
+      "learning_rate": 2.9492679960310975e-05,
+      "loss": 0.2544,
+      "step": 1183
+    },
+    {
+      "epoch": 0.5982819605861546,
+      "grad_norm": 0.49464377760887146,
+      "learning_rate": 2.9461709994365446e-05,
+      "loss": 0.1771,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5987872662961091,
+      "grad_norm": 0.7408649921417236,
+      "learning_rate": 2.9430732954503454e-05,
+      "loss": 0.3388,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5992925720060637,
+      "grad_norm": 0.6129613518714905,
+      "learning_rate": 2.9399748889838246e-05,
+      "loss": 0.3103,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5997978777160182,
+      "grad_norm": 0.5622842311859131,
+      "learning_rate": 2.9368757849494183e-05,
+      "loss": 0.3837,
+      "step": 1187
+    },
+    {
+      "epoch": 0.6003031834259727,
+      "grad_norm": 0.7855788469314575,
+      "learning_rate": 2.9337759882606684e-05,
+      "loss": 0.3053,
+      "step": 1188
+    },
+    {
+      "epoch": 0.6008084891359272,
+      "grad_norm": 0.424996942281723,
+      "learning_rate": 2.930675503832217e-05,
+      "loss": 0.2624,
+      "step": 1189
+    },
+    {
+      "epoch": 0.6013137948458818,
+      "grad_norm": 0.748625636100769,
+      "learning_rate": 2.9275743365797954e-05,
+      "loss": 0.3661,
+      "step": 1190
+    },
+    {
+      "epoch": 0.6018191005558363,
+      "grad_norm": 0.47824960947036743,
+      "learning_rate": 2.9244724914202176e-05,
+      "loss": 0.2256,
+      "step": 1191
+    },
+    {
+      "epoch": 0.6023244062657908,
+      "grad_norm": 0.5377376079559326,
+      "learning_rate": 2.9213699732713717e-05,
+      "loss": 0.2619,
+      "step": 1192
+    },
+    {
+      "epoch": 0.6028297119757453,
+      "grad_norm": 0.5620574951171875,
+      "learning_rate": 2.9182667870522147e-05,
+      "loss": 0.3563,
+      "step": 1193
+    },
+    {
+      "epoch": 0.6033350176856999,
+      "grad_norm": 0.2533213198184967,
+      "learning_rate": 2.91516293768276e-05,
+      "loss": 0.1327,
+      "step": 1194
+    },
+    {
+      "epoch": 0.6038403233956544,
+      "grad_norm": 0.4888068437576294,
+      "learning_rate": 2.9120584300840746e-05,
+      "loss": 0.2132,
+      "step": 1195
+    },
+    {
+      "epoch": 0.6043456291056089,
+      "grad_norm": 0.357085645198822,
+      "learning_rate": 2.90895326917827e-05,
+      "loss": 0.1772,
+      "step": 1196
+    },
+    {
+      "epoch": 0.6048509348155634,
+      "grad_norm": 0.30169978737831116,
+      "learning_rate": 2.9058474598884893e-05,
+      "loss": 0.202,
+      "step": 1197
+    },
+    {
+      "epoch": 0.605356240525518,
+      "grad_norm": 0.3223772644996643,
+      "learning_rate": 2.902741007138909e-05,
+      "loss": 0.1932,
+      "step": 1198
+    },
+    {
+      "epoch": 0.6058615462354725,
+      "grad_norm": 0.42703449726104736,
+      "learning_rate": 2.899633915854721e-05,
+      "loss": 0.1927,
+      "step": 1199
+    },
+    {
+      "epoch": 0.6063668519454269,
+      "grad_norm": 0.39132627844810486,
+      "learning_rate": 2.896526190962132e-05,
+      "loss": 0.2409,
+      "step": 1200
+    },
+    {
+      "epoch": 0.6068721576553815,
+      "grad_norm": 0.6942405104637146,
+      "learning_rate": 2.8934178373883543e-05,
+      "loss": 0.4429,
+      "step": 1201
+    },
+    {
+      "epoch": 0.607377463365336,
+      "grad_norm": 0.6032945513725281,
+      "learning_rate": 2.890308860061594e-05,
+      "loss": 0.4111,
+      "step": 1202
+    },
+    {
+      "epoch": 0.6078827690752906,
+      "grad_norm": 0.679614782333374,
+      "learning_rate": 2.8871992639110484e-05,
+      "loss": 0.3658,
+      "step": 1203
+    },
+    {
+      "epoch": 0.608388074785245,
+      "grad_norm": 0.714556097984314,
+      "learning_rate": 2.8840890538668952e-05,
+      "loss": 0.3826,
+      "step": 1204
+    },
+    {
+      "epoch": 0.6088933804951996,
+      "grad_norm": 0.5705816745758057,
+      "learning_rate": 2.8809782348602864e-05,
+      "loss": 0.3086,
+      "step": 1205
+    },
+    {
+      "epoch": 0.6093986862051541,
+      "grad_norm": 0.7331538200378418,
+      "learning_rate": 2.877866811823337e-05,
+      "loss": 0.4354,
+      "step": 1206
+    },
+    {
+      "epoch": 0.6099039919151087,
+      "grad_norm": 0.5332103967666626,
+      "learning_rate": 2.8747547896891218e-05,
+      "loss": 0.254,
+      "step": 1207
+    },
+    {
+      "epoch": 0.6104092976250631,
+      "grad_norm": 0.3042926490306854,
+      "learning_rate": 2.871642173391666e-05,
+      "loss": 0.2578,
+      "step": 1208
+    },
+    {
+      "epoch": 0.6109146033350177,
+      "grad_norm": 0.5534253716468811,
+      "learning_rate": 2.868528967865934e-05,
+      "loss": 0.3656,
+      "step": 1209
+    },
+    {
+      "epoch": 0.6114199090449722,
+      "grad_norm": 0.6421956419944763,
+      "learning_rate": 2.8654151780478265e-05,
+      "loss": 0.2551,
+      "step": 1210
+    },
+    {
+      "epoch": 0.6119252147549268,
+      "grad_norm": 0.5455820560455322,
+      "learning_rate": 2.862300808874171e-05,
+      "loss": 0.3176,
+      "step": 1211
+    },
+    {
+      "epoch": 0.6124305204648812,
+      "grad_norm": 0.4438186585903168,
+      "learning_rate": 2.859185865282712e-05,
+      "loss": 0.2395,
+      "step": 1212
+    },
+    {
+      "epoch": 0.6129358261748358,
+      "grad_norm": 0.3606109917163849,
+      "learning_rate": 2.8560703522121064e-05,
+      "loss": 0.2739,
+      "step": 1213
+    },
+    {
+      "epoch": 0.6134411318847903,
+      "grad_norm": 0.4127027094364166,
+      "learning_rate": 2.8529542746019115e-05,
+      "loss": 0.2952,
+      "step": 1214
+    },
+    {
+      "epoch": 0.6139464375947449,
+      "grad_norm": 0.5099959969520569,
+      "learning_rate": 2.8498376373925834e-05,
+      "loss": 0.4231,
+      "step": 1215
+    },
+    {
+      "epoch": 0.6144517433046993,
+      "grad_norm": 0.68137127161026,
+      "learning_rate": 2.8467204455254614e-05,
+      "loss": 0.2562,
+      "step": 1216
+    },
+    {
+      "epoch": 0.6149570490146539,
+      "grad_norm": 0.8698126673698425,
+      "learning_rate": 2.8436027039427667e-05,
+      "loss": 0.4673,
+      "step": 1217
+    },
+    {
+      "epoch": 0.6154623547246084,
+      "grad_norm": 0.3290408253669739,
+      "learning_rate": 2.8404844175875912e-05,
+      "loss": 0.1761,
+      "step": 1218
+    },
+    {
+      "epoch": 0.6159676604345629,
+      "grad_norm": 0.6963320970535278,
+      "learning_rate": 2.8373655914038904e-05,
+      "loss": 0.3319,
+      "step": 1219
+    },
+    {
+      "epoch": 0.6164729661445174,
+      "grad_norm": 0.32454270124435425,
+      "learning_rate": 2.8342462303364777e-05,
+      "loss": 0.1979,
+      "step": 1220
+    },
+    {
+      "epoch": 0.616978271854472,
+      "grad_norm": 0.40933623909950256,
+      "learning_rate": 2.831126339331011e-05,
+      "loss": 0.1951,
+      "step": 1221
+    },
+    {
+      "epoch": 0.6174835775644265,
+      "grad_norm": 0.665077269077301,
+      "learning_rate": 2.8280059233339913e-05,
+      "loss": 0.5259,
+      "step": 1222
+    },
+    {
+      "epoch": 0.617988883274381,
+      "grad_norm": 0.48233264684677124,
+      "learning_rate": 2.8248849872927507e-05,
+      "loss": 0.2227,
+      "step": 1223
+    },
+    {
+      "epoch": 0.6184941889843355,
+      "grad_norm": 0.3461703062057495,
+      "learning_rate": 2.8217635361554456e-05,
+      "loss": 0.1607,
+      "step": 1224
+    },
+    {
+      "epoch": 0.61899949469429,
+      "grad_norm": 0.4308650493621826,
+      "learning_rate": 2.8186415748710504e-05,
+      "loss": 0.2759,
+      "step": 1225
+    },
+    {
+      "epoch": 0.6195048004042446,
+      "grad_norm": 0.8697786331176758,
+      "learning_rate": 2.8155191083893467e-05,
+      "loss": 0.4196,
+      "step": 1226
+    },
+    {
+      "epoch": 0.620010106114199,
+      "grad_norm": 0.5697680115699768,
+      "learning_rate": 2.8123961416609174e-05,
+      "loss": 0.3709,
+      "step": 1227
+    },
+    {
+      "epoch": 0.6205154118241536,
+      "grad_norm": 0.6422898769378662,
+      "learning_rate": 2.8092726796371406e-05,
+      "loss": 0.2012,
+      "step": 1228
+    },
+    {
+      "epoch": 0.6210207175341081,
+      "grad_norm": 0.271230012178421,
+      "learning_rate": 2.806148727270176e-05,
+      "loss": 0.136,
+      "step": 1229
+    },
+    {
+      "epoch": 0.6215260232440627,
+      "grad_norm": 0.3259721100330353,
+      "learning_rate": 2.8030242895129643e-05,
+      "loss": 0.239,
+      "step": 1230
+    },
+    {
+      "epoch": 0.6220313289540171,
+      "grad_norm": 0.49231839179992676,
+      "learning_rate": 2.7998993713192128e-05,
+      "loss": 0.2351,
+      "step": 1231
+    },
+    {
+      "epoch": 0.6225366346639717,
+      "grad_norm": 0.6262175440788269,
+      "learning_rate": 2.7967739776433933e-05,
+      "loss": 0.2764,
+      "step": 1232
+    },
+    {
+      "epoch": 0.6230419403739262,
+      "grad_norm": 0.4242885410785675,
+      "learning_rate": 2.79364811344073e-05,
+      "loss": 0.1978,
+      "step": 1233
+    },
+    {
+      "epoch": 0.6235472460838808,
+      "grad_norm": 0.5172401070594788,
+      "learning_rate": 2.7905217836671916e-05,
+      "loss": 0.3884,
+      "step": 1234
+    },
+    {
+      "epoch": 0.6240525517938352,
+      "grad_norm": 0.5422953963279724,
+      "learning_rate": 2.787394993279488e-05,
+      "loss": 0.318,
+      "step": 1235
+    },
+    {
+      "epoch": 0.6245578575037898,
+      "grad_norm": 0.8409627676010132,
+      "learning_rate": 2.7842677472350577e-05,
+      "loss": 0.3753,
+      "step": 1236
+    },
+    {
+      "epoch": 0.6250631632137443,
+      "grad_norm": 0.9036238789558411,
+      "learning_rate": 2.7811400504920622e-05,
+      "loss": 0.5445,
+      "step": 1237
+    },
+    {
+      "epoch": 0.6255684689236989,
+      "grad_norm": 0.5030844807624817,
+      "learning_rate": 2.778011908009376e-05,
+      "loss": 0.2084,
+      "step": 1238
+    },
+    {
+      "epoch": 0.6260737746336533,
+      "grad_norm": 0.544114351272583,
+      "learning_rate": 2.7748833247465827e-05,
+      "loss": 0.3824,
+      "step": 1239
+    },
+    {
+      "epoch": 0.6265790803436079,
+      "grad_norm": 0.5116641521453857,
+      "learning_rate": 2.7717543056639637e-05,
+      "loss": 0.4358,
+      "step": 1240
+    },
+    {
+      "epoch": 0.6270843860535624,
+      "grad_norm": 0.49113696813583374,
+      "learning_rate": 2.7686248557224904e-05,
+      "loss": 0.2644,
+      "step": 1241
+    },
+    {
+      "epoch": 0.627589691763517,
+      "grad_norm": 0.3701680600643158,
+      "learning_rate": 2.765494979883819e-05,
+      "loss": 0.1596,
+      "step": 1242
+    },
+    {
+      "epoch": 0.6280949974734714,
+      "grad_norm": 0.5240901708602905,
+      "learning_rate": 2.762364683110279e-05,
+      "loss": 0.3679,
+      "step": 1243
+    },
+    {
+      "epoch": 0.628600303183426,
+      "grad_norm": 0.49889007210731506,
+      "learning_rate": 2.7592339703648694e-05,
+      "loss": 0.6432,
+      "step": 1244
+    },
+    {
+      "epoch": 0.6291056088933805,
+      "grad_norm": 0.3745577037334442,
+      "learning_rate": 2.7561028466112482e-05,
+      "loss": 0.154,
+      "step": 1245
+    },
+    {
+      "epoch": 0.6296109146033351,
+      "grad_norm": 0.5030989050865173,
+      "learning_rate": 2.7529713168137233e-05,
+      "loss": 0.2932,
+      "step": 1246
+    },
+    {
+      "epoch": 0.6301162203132895,
+      "grad_norm": 1.0743045806884766,
+      "learning_rate": 2.7498393859372484e-05,
+      "loss": 0.2168,
+      "step": 1247
+    },
+    {
+      "epoch": 0.6306215260232441,
+      "grad_norm": 0.46885451674461365,
+      "learning_rate": 2.746707058947412e-05,
+      "loss": 0.2967,
+      "step": 1248
+    },
+    {
+      "epoch": 0.6311268317331986,
+      "grad_norm": 0.3801758885383606,
+      "learning_rate": 2.743574340810431e-05,
+      "loss": 0.2834,
+      "step": 1249
+    },
+    {
+      "epoch": 0.631632137443153,
+      "grad_norm": 0.700106143951416,
+      "learning_rate": 2.7404412364931427e-05,
+      "loss": 0.5024,
+      "step": 1250
+    },
+    {
+      "epoch": 0.6321374431531076,
+      "grad_norm": 0.8579809069633484,
+      "learning_rate": 2.7373077509629957e-05,
+      "loss": 0.313,
+      "step": 1251
+    },
+    {
+      "epoch": 0.6326427488630622,
+      "grad_norm": 0.3703233599662781,
+      "learning_rate": 2.7341738891880452e-05,
+      "loss": 0.2095,
+      "step": 1252
+    },
+    {
+      "epoch": 0.6331480545730167,
+      "grad_norm": 0.2917376160621643,
+      "learning_rate": 2.73103965613694e-05,
+      "loss": 0.1726,
+      "step": 1253
+    },
+    {
+      "epoch": 0.6336533602829711,
+      "grad_norm": 0.5064457058906555,
+      "learning_rate": 2.7279050567789194e-05,
+      "loss": 0.2515,
+      "step": 1254
+    },
+    {
+      "epoch": 0.6341586659929257,
+      "grad_norm": 0.47626200318336487,
+      "learning_rate": 2.7247700960838034e-05,
+      "loss": 0.4029,
+      "step": 1255
+    },
+    {
+      "epoch": 0.6346639717028802,
+      "grad_norm": 0.3616740107536316,
+      "learning_rate": 2.7216347790219843e-05,
+      "loss": 0.2887,
+      "step": 1256
+    },
+    {
+      "epoch": 0.6351692774128348,
+      "grad_norm": 0.2952771484851837,
+      "learning_rate": 2.7184991105644192e-05,
+      "loss": 0.2067,
+      "step": 1257
+    },
+    {
+      "epoch": 0.6356745831227892,
+      "grad_norm": 0.2758321166038513,
+      "learning_rate": 2.715363095682623e-05,
+      "loss": 0.1615,
+      "step": 1258
+    },
+    {
+      "epoch": 0.6361798888327438,
+      "grad_norm": 0.5239960551261902,
+      "learning_rate": 2.7122267393486607e-05,
+      "loss": 0.3267,
+      "step": 1259
+    },
+    {
+      "epoch": 0.6366851945426983,
+      "grad_norm": 0.4696899652481079,
+      "learning_rate": 2.7090900465351355e-05,
+      "loss": 0.3621,
+      "step": 1260
+    },
+    {
+      "epoch": 0.6371905002526529,
+      "grad_norm": 0.40326637029647827,
+      "learning_rate": 2.7059530222151873e-05,
+      "loss": 0.1739,
+      "step": 1261
+    },
+    {
+      "epoch": 0.6376958059626073,
+      "grad_norm": 0.5829006433486938,
+      "learning_rate": 2.7028156713624814e-05,
+      "loss": 0.3423,
+      "step": 1262
+    },
+    {
+      "epoch": 0.6382011116725619,
+      "grad_norm": 0.3605864644050598,
+      "learning_rate": 2.6996779989511985e-05,
+      "loss": 0.1829,
+      "step": 1263
+    },
+    {
+      "epoch": 0.6387064173825164,
+      "grad_norm": 0.38949474692344666,
+      "learning_rate": 2.6965400099560308e-05,
+      "loss": 0.1814,
+      "step": 1264
+    },
+    {
+      "epoch": 0.639211723092471,
+      "grad_norm": 0.835013747215271,
+      "learning_rate": 2.6934017093521717e-05,
+      "loss": 0.5331,
+      "step": 1265
+    },
+    {
+      "epoch": 0.6397170288024254,
+      "grad_norm": 0.3474362790584564,
+      "learning_rate": 2.6902631021153096e-05,
+      "loss": 0.1773,
+      "step": 1266
+    },
+    {
+      "epoch": 0.64022233451238,
+      "grad_norm": 0.7530576586723328,
+      "learning_rate": 2.6871241932216195e-05,
+      "loss": 0.6231,
+      "step": 1267
+    },
+    {
+      "epoch": 0.6407276402223345,
+      "grad_norm": 0.46543627977371216,
+      "learning_rate": 2.6839849876477518e-05,
+      "loss": 0.2082,
+      "step": 1268
+    },
+    {
+      "epoch": 0.6412329459322891,
+      "grad_norm": 0.6920028328895569,
+      "learning_rate": 2.6808454903708315e-05,
+      "loss": 0.3675,
+      "step": 1269
+    },
+    {
+      "epoch": 0.6417382516422435,
+      "grad_norm": 0.639060378074646,
+      "learning_rate": 2.6777057063684406e-05,
+      "loss": 0.4416,
+      "step": 1270
+    },
+    {
+      "epoch": 0.6422435573521981,
+      "grad_norm": 0.8826596736907959,
+      "learning_rate": 2.674565640618621e-05,
+      "loss": 0.3633,
+      "step": 1271
+    },
+    {
+      "epoch": 0.6427488630621526,
+      "grad_norm": 0.6439775824546814,
+      "learning_rate": 2.6714252980998595e-05,
+      "loss": 0.3737,
+      "step": 1272
+    },
+    {
+      "epoch": 0.6432541687721072,
+      "grad_norm": 0.3467325270175934,
+      "learning_rate": 2.6682846837910796e-05,
+      "loss": 0.1777,
+      "step": 1273
+    },
+    {
+      "epoch": 0.6437594744820616,
+      "grad_norm": 1.040595293045044,
+      "learning_rate": 2.6651438026716396e-05,
+      "loss": 0.4716,
+      "step": 1274
+    },
+    {
+      "epoch": 0.6442647801920162,
+      "grad_norm": 0.6756473779678345,
+      "learning_rate": 2.6620026597213167e-05,
+      "loss": 0.3154,
+      "step": 1275
+    },
+    {
+      "epoch": 0.6447700859019707,
+      "grad_norm": 0.6702762246131897,
+      "learning_rate": 2.658861259920306e-05,
+      "loss": 0.366,
+      "step": 1276
+    },
+    {
+      "epoch": 0.6452753916119253,
+      "grad_norm": 0.34285977482795715,
+      "learning_rate": 2.6557196082492105e-05,
+      "loss": 0.1714,
+      "step": 1277
+    },
+    {
+      "epoch": 0.6457806973218797,
+      "grad_norm": 0.5991812348365784,
+      "learning_rate": 2.6525777096890286e-05,
+      "loss": 0.3714,
+      "step": 1278
+    },
+    {
+      "epoch": 0.6462860030318343,
+      "grad_norm": 0.39086848497390747,
+      "learning_rate": 2.6494355692211535e-05,
+      "loss": 0.1841,
+      "step": 1279
+    },
+    {
+      "epoch": 0.6467913087417888,
+      "grad_norm": 0.3882160186767578,
+      "learning_rate": 2.6462931918273624e-05,
+      "loss": 0.2613,
+      "step": 1280
+    },
+    {
+      "epoch": 0.6472966144517434,
+      "grad_norm": 0.7259517908096313,
+      "learning_rate": 2.643150582489805e-05,
+      "loss": 0.3005,
+      "step": 1281
+    },
+    {
+      "epoch": 0.6478019201616978,
+      "grad_norm": 0.6918402314186096,
+      "learning_rate": 2.6400077461910013e-05,
+      "loss": 0.4195,
+      "step": 1282
+    },
+    {
+      "epoch": 0.6483072258716523,
+      "grad_norm": 0.6692409515380859,
+      "learning_rate": 2.6368646879138304e-05,
+      "loss": 0.4622,
+      "step": 1283
+    },
+    {
+      "epoch": 0.6488125315816069,
+      "grad_norm": 0.6194551587104797,
+      "learning_rate": 2.633721412641524e-05,
+      "loss": 0.4571,
+      "step": 1284
+    },
+    {
+      "epoch": 0.6493178372915613,
+      "grad_norm": 0.4386284351348877,
+      "learning_rate": 2.6305779253576556e-05,
+      "loss": 0.1687,
+      "step": 1285
+    },
+    {
+      "epoch": 0.6498231430015159,
+      "grad_norm": 0.43226057291030884,
+      "learning_rate": 2.6274342310461382e-05,
+      "loss": 0.3734,
+      "step": 1286
+    },
+    {
+      "epoch": 0.6503284487114704,
+      "grad_norm": 0.3043833076953888,
+      "learning_rate": 2.624290334691209e-05,
+      "loss": 0.1776,
+      "step": 1287
+    },
+    {
+      "epoch": 0.650833754421425,
+      "grad_norm": 0.4461258053779602,
+      "learning_rate": 2.6211462412774298e-05,
+      "loss": 0.2532,
+      "step": 1288
+    },
+    {
+      "epoch": 0.6513390601313794,
+      "grad_norm": 0.5978369116783142,
+      "learning_rate": 2.6180019557896725e-05,
+      "loss": 0.4937,
+      "step": 1289
+    },
+    {
+      "epoch": 0.651844365841334,
+      "grad_norm": 0.4691097140312195,
+      "learning_rate": 2.614857483213113e-05,
+      "loss": 0.3146,
+      "step": 1290
+    },
+    {
+      "epoch": 0.6523496715512885,
+      "grad_norm": 0.48737433552742004,
+      "learning_rate": 2.6117128285332248e-05,
+      "loss": 0.2174,
+      "step": 1291
+    },
+    {
+      "epoch": 0.6528549772612431,
+      "grad_norm": 0.692764937877655,
+      "learning_rate": 2.6085679967357707e-05,
+      "loss": 0.4083,
+      "step": 1292
+    },
+    {
+      "epoch": 0.6533602829711975,
+      "grad_norm": 0.6832363605499268,
+      "learning_rate": 2.6054229928067923e-05,
+      "loss": 0.7104,
+      "step": 1293
+    },
+    {
+      "epoch": 0.6538655886811521,
+      "grad_norm": 0.2708096504211426,
+      "learning_rate": 2.6022778217326077e-05,
+      "loss": 0.1944,
+      "step": 1294
+    },
+    {
+      "epoch": 0.6543708943911066,
+      "grad_norm": 0.2859404385089874,
+      "learning_rate": 2.5991324884997952e-05,
+      "loss": 0.1835,
+      "step": 1295
+    },
+    {
+      "epoch": 0.6548762001010612,
+      "grad_norm": 0.6921172142028809,
+      "learning_rate": 2.595986998095194e-05,
+      "loss": 0.3649,
+      "step": 1296
+    },
+    {
+      "epoch": 0.6553815058110156,
+      "grad_norm": 0.270704984664917,
+      "learning_rate": 2.5928413555058904e-05,
+      "loss": 0.1674,
+      "step": 1297
+    },
+    {
+      "epoch": 0.6558868115209702,
+      "grad_norm": 0.39965078234672546,
+      "learning_rate": 2.5896955657192142e-05,
+      "loss": 0.29,
+      "step": 1298
+    },
+    {
+      "epoch": 0.6563921172309247,
+      "grad_norm": 0.6206263303756714,
+      "learning_rate": 2.5865496337227256e-05,
+      "loss": 0.2713,
+      "step": 1299
+    },
+    {
+      "epoch": 0.6568974229408793,
+      "grad_norm": 0.4320612847805023,
+      "learning_rate": 2.583403564504213e-05,
+      "loss": 0.2148,
+      "step": 1300
+    },
+    {
+      "epoch": 0.6574027286508337,
+      "grad_norm": 0.4375782012939453,
+      "learning_rate": 2.5802573630516806e-05,
+      "loss": 0.481,
+      "step": 1301
+    },
+    {
+      "epoch": 0.6579080343607883,
+      "grad_norm": 0.40101978182792664,
+      "learning_rate": 2.5771110343533428e-05,
+      "loss": 0.2242,
+      "step": 1302
+    },
+    {
+      "epoch": 0.6584133400707428,
+      "grad_norm": 0.7650478482246399,
+      "learning_rate": 2.5739645833976154e-05,
+      "loss": 0.337,
+      "step": 1303
+    },
+    {
+      "epoch": 0.6589186457806974,
+      "grad_norm": 0.6947590708732605,
+      "learning_rate": 2.5708180151731103e-05,
+      "loss": 0.4345,
+      "step": 1304
+    },
+    {
+      "epoch": 0.6594239514906518,
+      "grad_norm": 0.42338618636131287,
+      "learning_rate": 2.567671334668621e-05,
+      "loss": 0.2952,
+      "step": 1305
+    },
+    {
+      "epoch": 0.6599292572006064,
+      "grad_norm": 0.748132586479187,
+      "learning_rate": 2.5645245468731234e-05,
+      "loss": 0.8043,
+      "step": 1306
+    },
+    {
+      "epoch": 0.6604345629105609,
+      "grad_norm": 0.40081891417503357,
+      "learning_rate": 2.5613776567757595e-05,
+      "loss": 0.4187,
+      "step": 1307
+    },
+    {
+      "epoch": 0.6609398686205155,
+      "grad_norm": 0.5815805792808533,
+      "learning_rate": 2.5582306693658374e-05,
+      "loss": 0.251,
+      "step": 1308
+    },
+    {
+      "epoch": 0.6614451743304699,
+      "grad_norm": 0.4176938533782959,
+      "learning_rate": 2.555083589632818e-05,
+      "loss": 0.2369,
+      "step": 1309
+    },
+    {
+      "epoch": 0.6619504800404244,
+      "grad_norm": 0.2541157007217407,
+      "learning_rate": 2.5519364225663055e-05,
+      "loss": 0.1507,
+      "step": 1310
+    },
+    {
+      "epoch": 0.662455785750379,
+      "grad_norm": 0.5441336631774902,
+      "learning_rate": 2.5487891731560482e-05,
+      "loss": 0.2521,
+      "step": 1311
+    },
+    {
+      "epoch": 0.6629610914603336,
+      "grad_norm": 0.4862014055252075,
+      "learning_rate": 2.54564184639192e-05,
+      "loss": 0.2766,
+      "step": 1312
+    },
+    {
+      "epoch": 0.663466397170288,
+      "grad_norm": 0.7492696046829224,
+      "learning_rate": 2.5424944472639216e-05,
+      "loss": 0.4431,
+      "step": 1313
+    },
+    {
+      "epoch": 0.6639717028802425,
+      "grad_norm": 0.45778077840805054,
+      "learning_rate": 2.5393469807621645e-05,
+      "loss": 0.2535,
+      "step": 1314
+    },
+    {
+      "epoch": 0.6644770085901971,
+      "grad_norm": 0.8111942410469055,
+      "learning_rate": 2.5361994518768695e-05,
+      "loss": 0.3406,
+      "step": 1315
+    },
+    {
+      "epoch": 0.6649823143001516,
+      "grad_norm": 0.48901939392089844,
+      "learning_rate": 2.5330518655983558e-05,
+      "loss": 0.2384,
+      "step": 1316
+    },
+    {
+      "epoch": 0.6654876200101061,
+      "grad_norm": 0.67337566614151,
+      "learning_rate": 2.529904226917033e-05,
+      "loss": 0.4046,
+      "step": 1317
+    },
+    {
+      "epoch": 0.6659929257200606,
+      "grad_norm": 0.5460554957389832,
+      "learning_rate": 2.5267565408233952e-05,
+      "loss": 0.286,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6664982314300152,
+      "grad_norm": 0.34170493483543396,
+      "learning_rate": 2.5236088123080092e-05,
+      "loss": 0.2063,
+      "step": 1319
+    },
+    {
+      "epoch": 0.6670035371399696,
+      "grad_norm": 0.47587305307388306,
+      "learning_rate": 2.5204610463615115e-05,
+      "loss": 0.2657,
+      "step": 1320
+    },
+    {
+      "epoch": 0.6675088428499242,
+      "grad_norm": 0.4068621098995209,
+      "learning_rate": 2.5173132479745975e-05,
+      "loss": 0.3323,
+      "step": 1321
+    },
+    {
+      "epoch": 0.6680141485598787,
+      "grad_norm": 0.4757383167743683,
+      "learning_rate": 2.514165422138012e-05,
+      "loss": 0.3589,
+      "step": 1322
+    },
+    {
+      "epoch": 0.6685194542698333,
+      "grad_norm": 0.3725772202014923,
+      "learning_rate": 2.5110175738425462e-05,
+      "loss": 0.1664,
+      "step": 1323
+    },
+    {
+      "epoch": 0.6690247599797877,
+      "grad_norm": 0.6302722692489624,
+      "learning_rate": 2.507869708079025e-05,
+      "loss": 0.4999,
+      "step": 1324
+    },
+    {
+      "epoch": 0.6695300656897423,
+      "grad_norm": 0.44832542538642883,
+      "learning_rate": 2.5047218298383015e-05,
+      "loss": 0.3016,
+      "step": 1325
+    },
+    {
+      "epoch": 0.6700353713996968,
+      "grad_norm": 0.4571584463119507,
+      "learning_rate": 2.5015739441112484e-05,
+      "loss": 0.1797,
+      "step": 1326
+    },
+    {
+      "epoch": 0.6705406771096514,
+      "grad_norm": 0.6637290120124817,
+      "learning_rate": 2.4984260558887522e-05,
+      "loss": 0.5544,
+      "step": 1327
+    },
+    {
+      "epoch": 0.6710459828196058,
+      "grad_norm": 0.337741494178772,
+      "learning_rate": 2.495278170161699e-05,
+      "loss": 0.1652,
+      "step": 1328
+    },
+    {
+      "epoch": 0.6715512885295604,
+      "grad_norm": 0.6732521653175354,
+      "learning_rate": 2.4921302919209763e-05,
+      "loss": 0.5967,
+      "step": 1329
+    },
+    {
+      "epoch": 0.6720565942395149,
+      "grad_norm": 0.8650082945823669,
+      "learning_rate": 2.4889824261574547e-05,
+      "loss": 0.4235,
+      "step": 1330
+    },
+    {
+      "epoch": 0.6725618999494695,
+      "grad_norm": 0.7857891321182251,
+      "learning_rate": 2.4858345778619887e-05,
+      "loss": 0.5156,
+      "step": 1331
+    },
+    {
+      "epoch": 0.6730672056594239,
+      "grad_norm": 0.42698702216148376,
+      "learning_rate": 2.4826867520254034e-05,
+      "loss": 0.2091,
+      "step": 1332
+    },
+    {
+      "epoch": 0.6735725113693785,
+      "grad_norm": 0.32361331582069397,
+      "learning_rate": 2.4795389536384887e-05,
+      "loss": 0.1365,
+      "step": 1333
+    },
+    {
+      "epoch": 0.674077817079333,
+      "grad_norm": 0.7567768096923828,
+      "learning_rate": 2.4763911876919917e-05,
+      "loss": 0.4576,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6745831227892876,
+      "grad_norm": 0.2475138008594513,
+      "learning_rate": 2.4732434591766057e-05,
+      "loss": 0.1624,
+      "step": 1335
+    },
+    {
+      "epoch": 0.675088428499242,
+      "grad_norm": 0.8214117288589478,
+      "learning_rate": 2.4700957730829675e-05,
+      "loss": 0.3756,
+      "step": 1336
+    },
+    {
+      "epoch": 0.6755937342091966,
+      "grad_norm": 0.6251058578491211,
+      "learning_rate": 2.4669481344016444e-05,
+      "loss": 0.4255,
+      "step": 1337
+    },
+    {
+      "epoch": 0.6760990399191511,
+      "grad_norm": 0.45437926054000854,
+      "learning_rate": 2.4638005481231304e-05,
+      "loss": 0.5109,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6766043456291057,
+      "grad_norm": 0.7901365756988525,
+      "learning_rate": 2.4606530192378358e-05,
+      "loss": 0.4083,
+      "step": 1339
+    },
+    {
+      "epoch": 0.6771096513390601,
+      "grad_norm": 0.47076261043548584,
+      "learning_rate": 2.4575055527360793e-05,
+      "loss": 0.33,
+      "step": 1340
+    },
+    {
+      "epoch": 0.6776149570490146,
+      "grad_norm": 0.7028003334999084,
+      "learning_rate": 2.45435815360808e-05,
+      "loss": 0.4038,
+      "step": 1341
+    },
+    {
+      "epoch": 0.6781202627589692,
+      "grad_norm": 0.3465520441532135,
+      "learning_rate": 2.4512108268439524e-05,
+      "loss": 0.2881,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6786255684689237,
+      "grad_norm": 0.6936810612678528,
+      "learning_rate": 2.4480635774336947e-05,
+      "loss": 0.4991,
+      "step": 1343
+    },
+    {
+      "epoch": 0.6791308741788782,
+      "grad_norm": 0.264642596244812,
+      "learning_rate": 2.4449164103671833e-05,
+      "loss": 0.1572,
+      "step": 1344
+    },
+    {
+      "epoch": 0.6796361798888327,
+      "grad_norm": 0.6691703200340271,
+      "learning_rate": 2.441769330634163e-05,
+      "loss": 0.4967,
+      "step": 1345
+    },
+    {
+      "epoch": 0.6801414855987873,
+      "grad_norm": 0.8646639585494995,
+      "learning_rate": 2.4386223432242407e-05,
+      "loss": 0.3253,
+      "step": 1346
+    },
+    {
+      "epoch": 0.6806467913087418,
+      "grad_norm": 0.618078887462616,
+      "learning_rate": 2.435475453126877e-05,
+      "loss": 0.1929,
+      "step": 1347
+    },
+    {
+      "epoch": 0.6811520970186963,
+      "grad_norm": 0.8526009917259216,
+      "learning_rate": 2.432328665331379e-05,
+      "loss": 0.4779,
+      "step": 1348
+    },
+    {
+      "epoch": 0.6816574027286508,
+      "grad_norm": 0.8699708580970764,
+      "learning_rate": 2.4291819848268906e-05,
+      "loss": 0.3817,
+      "step": 1349
+    },
+    {
+      "epoch": 0.6821627084386054,
+      "grad_norm": 0.4946329891681671,
+      "learning_rate": 2.4260354166023848e-05,
+      "loss": 0.2458,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6826680141485599,
+      "grad_norm": 0.5863744020462036,
+      "learning_rate": 2.4228889656466578e-05,
+      "loss": 0.5891,
+      "step": 1351
+    },
+    {
+      "epoch": 0.6831733198585144,
+      "grad_norm": 0.5861146450042725,
+      "learning_rate": 2.41974263694832e-05,
+      "loss": 0.3156,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6836786255684689,
+      "grad_norm": 0.590474009513855,
+      "learning_rate": 2.416596435495787e-05,
+      "loss": 0.3012,
+      "step": 1353
+    },
+    {
+      "epoch": 0.6841839312784235,
+      "grad_norm": 0.4421933591365814,
+      "learning_rate": 2.4134503662772753e-05,
+      "loss": 0.3528,
+      "step": 1354
+    },
+    {
+      "epoch": 0.6846892369883779,
+      "grad_norm": 0.3189122974872589,
+      "learning_rate": 2.4103044342807867e-05,
+      "loss": 0.1986,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6851945426983325,
+      "grad_norm": 0.6435018181800842,
+      "learning_rate": 2.4071586444941098e-05,
+      "loss": 0.3458,
+      "step": 1356
+    },
+    {
+      "epoch": 0.685699848408287,
+      "grad_norm": 0.8088643550872803,
+      "learning_rate": 2.404013001904807e-05,
+      "loss": 0.5646,
+      "step": 1357
+    },
+    {
+      "epoch": 0.6862051541182416,
+      "grad_norm": 0.4556143879890442,
+      "learning_rate": 2.4008675115002054e-05,
+      "loss": 0.3147,
+      "step": 1358
+    },
+    {
+      "epoch": 0.686710459828196,
+      "grad_norm": 0.4295537769794464,
+      "learning_rate": 2.3977221782673933e-05,
+      "loss": 0.2078,
+      "step": 1359
+    },
+    {
+      "epoch": 0.6872157655381506,
+      "grad_norm": 0.5436996817588806,
+      "learning_rate": 2.394577007193208e-05,
+      "loss": 0.3047,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6877210712481051,
+      "grad_norm": 0.6986174583435059,
+      "learning_rate": 2.39143200326423e-05,
+      "loss": 0.4117,
+      "step": 1361
+    },
+    {
+      "epoch": 0.6882263769580597,
+      "grad_norm": 0.5140484571456909,
+      "learning_rate": 2.3882871714667754e-05,
+      "loss": 0.2067,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6887316826680141,
+      "grad_norm": 0.9210074543952942,
+      "learning_rate": 2.3851425167868874e-05,
+      "loss": 0.4738,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6892369883779687,
+      "grad_norm": 0.5711868405342102,
+      "learning_rate": 2.3819980442103288e-05,
+      "loss": 0.2715,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6897422940879232,
+      "grad_norm": 0.691596508026123,
+      "learning_rate": 2.3788537587225705e-05,
+      "loss": 0.4291,
+      "step": 1365
+    },
+    {
+      "epoch": 0.6902475997978778,
+      "grad_norm": 0.6653650999069214,
+      "learning_rate": 2.3757096653087914e-05,
+      "loss": 0.2889,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6907529055078322,
+      "grad_norm": 0.5367448329925537,
+      "learning_rate": 2.3725657689538627e-05,
+      "loss": 0.3765,
+      "step": 1367
+    },
+    {
+      "epoch": 0.6912582112177867,
+      "grad_norm": 0.4669603705406189,
+      "learning_rate": 2.3694220746423447e-05,
+      "loss": 0.2137,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6917635169277413,
+      "grad_norm": 0.5713473558425903,
+      "learning_rate": 2.3662785873584773e-05,
+      "loss": 0.3873,
+      "step": 1369
+    },
+    {
+      "epoch": 0.6922688226376958,
+      "grad_norm": 0.444577693939209,
+      "learning_rate": 2.36313531208617e-05,
+      "loss": 0.257,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6927741283476503,
+      "grad_norm": 0.44812366366386414,
+      "learning_rate": 2.3599922538089993e-05,
+      "loss": 0.2485,
+      "step": 1371
+    },
+    {
+      "epoch": 0.6932794340576048,
+      "grad_norm": 0.39372605085372925,
+      "learning_rate": 2.3568494175101958e-05,
+      "loss": 0.2552,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6937847397675594,
+      "grad_norm": 0.8668807744979858,
+      "learning_rate": 2.353706808172638e-05,
+      "loss": 0.4462,
+      "step": 1373
+    },
+    {
+      "epoch": 0.6942900454775139,
+      "grad_norm": 0.37225598096847534,
+      "learning_rate": 2.3505644307788467e-05,
+      "loss": 0.2313,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6947953511874684,
+      "grad_norm": 0.2703179717063904,
+      "learning_rate": 2.3474222903109723e-05,
+      "loss": 0.1954,
+      "step": 1375
+    },
+    {
+      "epoch": 0.6953006568974229,
+      "grad_norm": 0.46796905994415283,
+      "learning_rate": 2.34428039175079e-05,
+      "loss": 0.2335,
+      "step": 1376
+    },
+    {
+      "epoch": 0.6958059626073775,
+      "grad_norm": 0.4360145926475525,
+      "learning_rate": 2.3411387400796934e-05,
+      "loss": 0.2707,
+      "step": 1377
+    },
+    {
+      "epoch": 0.696311268317332,
+      "grad_norm": 0.4478427767753601,
+      "learning_rate": 2.3379973402786832e-05,
+      "loss": 0.2095,
+      "step": 1378
+    },
+    {
+      "epoch": 0.6968165740272865,
+      "grad_norm": 0.544204831123352,
+      "learning_rate": 2.3348561973283613e-05,
+      "loss": 0.2305,
+      "step": 1379
+    },
+    {
+      "epoch": 0.697321879737241,
+      "grad_norm": 0.7342060208320618,
+      "learning_rate": 2.3317153162089206e-05,
+      "loss": 0.3866,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6978271854471956,
+      "grad_norm": 0.33234989643096924,
+      "learning_rate": 2.328574701900141e-05,
+      "loss": 0.1789,
+      "step": 1381
+    },
+    {
+      "epoch": 0.6983324911571501,
+      "grad_norm": 0.541576623916626,
+      "learning_rate": 2.325434359381379e-05,
+      "loss": 0.2606,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6988377968671046,
+      "grad_norm": 0.5520817041397095,
+      "learning_rate": 2.3222942936315606e-05,
+      "loss": 0.2925,
+      "step": 1383
+    },
+    {
+      "epoch": 0.6993431025770591,
+      "grad_norm": 0.5592846870422363,
+      "learning_rate": 2.31915450962917e-05,
+      "loss": 0.2995,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6998484082870137,
+      "grad_norm": 0.18375256657600403,
+      "learning_rate": 2.3160150123522485e-05,
+      "loss": 0.0958,
+      "step": 1385
+    },
+    {
+      "epoch": 0.7003537139969682,
+      "grad_norm": 0.5363619923591614,
+      "learning_rate": 2.3128758067783808e-05,
+      "loss": 0.2511,
+      "step": 1386
+    },
+    {
+      "epoch": 0.7008590197069227,
+      "grad_norm": 0.4795052111148834,
+      "learning_rate": 2.30973689788469e-05,
+      "loss": 0.2263,
+      "step": 1387
+    },
+    {
+      "epoch": 0.7013643254168772,
+      "grad_norm": 0.6192831993103027,
+      "learning_rate": 2.306598290647829e-05,
+      "loss": 0.3247,
+      "step": 1388
+    },
+    {
+      "epoch": 0.7018696311268318,
+      "grad_norm": 0.4603809416294098,
+      "learning_rate": 2.30345999004397e-05,
+      "loss": 0.2794,
+      "step": 1389
+    },
+    {
+      "epoch": 0.7023749368367862,
+      "grad_norm": 0.5347074270248413,
+      "learning_rate": 2.300322001048802e-05,
+      "loss": 0.3136,
+      "step": 1390
+    },
+    {
+      "epoch": 0.7028802425467408,
+      "grad_norm": 0.6828907132148743,
+      "learning_rate": 2.2971843286375192e-05,
+      "loss": 0.2847,
+      "step": 1391
+    },
+    {
+      "epoch": 0.7033855482566953,
+      "grad_norm": 0.6386706233024597,
+      "learning_rate": 2.2940469777848123e-05,
+      "loss": 0.4388,
+      "step": 1392
+    },
+    {
+      "epoch": 0.7038908539666499,
+      "grad_norm": 0.4837972819805145,
+      "learning_rate": 2.2909099534648658e-05,
+      "loss": 0.2637,
+      "step": 1393
+    },
+    {
+      "epoch": 0.7043961596766043,
+      "grad_norm": 0.4311538338661194,
+      "learning_rate": 2.2877732606513406e-05,
+      "loss": 0.2201,
+      "step": 1394
+    },
+    {
+      "epoch": 0.7049014653865588,
+      "grad_norm": 0.7682383060455322,
+      "learning_rate": 2.2846369043173775e-05,
+      "loss": 0.3609,
+      "step": 1395
+    },
+    {
+      "epoch": 0.7054067710965134,
+      "grad_norm": 0.4595930576324463,
+      "learning_rate": 2.281500889435581e-05,
+      "loss": 0.3006,
+      "step": 1396
+    },
+    {
+      "epoch": 0.705912076806468,
+      "grad_norm": 0.45893973112106323,
+      "learning_rate": 2.278365220978016e-05,
+      "loss": 0.2403,
+      "step": 1397
+    },
+    {
+      "epoch": 0.7064173825164224,
+      "grad_norm": 0.24530524015426636,
+      "learning_rate": 2.2752299039161975e-05,
+      "loss": 0.1777,
+      "step": 1398
+    },
+    {
+      "epoch": 0.7069226882263769,
+      "grad_norm": 0.6683110594749451,
+      "learning_rate": 2.272094943221081e-05,
+      "loss": 0.3115,
+      "step": 1399
+    },
+    {
+      "epoch": 0.7074279939363315,
+      "grad_norm": 0.5461668372154236,
+      "learning_rate": 2.2689603438630606e-05,
+      "loss": 0.3772,
+      "step": 1400
+    },
+    {
+      "epoch": 0.707933299646286,
+      "grad_norm": 0.2557353079319,
+      "learning_rate": 2.2658261108119554e-05,
+      "loss": 0.1629,
+      "step": 1401
+    },
+    {
+      "epoch": 0.7084386053562405,
+      "grad_norm": 0.43094396591186523,
+      "learning_rate": 2.262692249037004e-05,
+      "loss": 0.1927,
+      "step": 1402
+    },
+    {
+      "epoch": 0.708943911066195,
+      "grad_norm": 0.43190810084342957,
+      "learning_rate": 2.259558763506858e-05,
+      "loss": 0.3081,
+      "step": 1403
+    },
+    {
+      "epoch": 0.7094492167761496,
+      "grad_norm": 0.32105064392089844,
+      "learning_rate": 2.2564256591895693e-05,
+      "loss": 0.1161,
+      "step": 1404
+    },
+    {
+      "epoch": 0.7099545224861041,
+      "grad_norm": 0.45100855827331543,
+      "learning_rate": 2.2532929410525887e-05,
+      "loss": 0.2658,
+      "step": 1405
+    },
+    {
+      "epoch": 0.7104598281960586,
+      "grad_norm": 0.41769468784332275,
+      "learning_rate": 2.2501606140627518e-05,
+      "loss": 0.3101,
+      "step": 1406
+    },
+    {
+      "epoch": 0.7109651339060131,
+      "grad_norm": 0.5544657111167908,
+      "learning_rate": 2.2470286831862766e-05,
+      "loss": 0.38,
+      "step": 1407
+    },
+    {
+      "epoch": 0.7114704396159677,
+      "grad_norm": 0.45244771242141724,
+      "learning_rate": 2.2438971533887527e-05,
+      "loss": 0.3448,
+      "step": 1408
+    },
+    {
+      "epoch": 0.7119757453259222,
+      "grad_norm": 0.3420441150665283,
+      "learning_rate": 2.240766029635131e-05,
+      "loss": 0.2201,
+      "step": 1409
+    },
+    {
+      "epoch": 0.7124810510358767,
+      "grad_norm": 0.6040711998939514,
+      "learning_rate": 2.2376353168897214e-05,
+      "loss": 0.4204,
+      "step": 1410
+    },
+    {
+      "epoch": 0.7129863567458312,
+      "grad_norm": 0.6188685894012451,
+      "learning_rate": 2.2345050201161814e-05,
+      "loss": 0.2601,
+      "step": 1411
+    },
+    {
+      "epoch": 0.7134916624557858,
+      "grad_norm": 0.4009005129337311,
+      "learning_rate": 2.2313751442775098e-05,
+      "loss": 0.2132,
+      "step": 1412
+    },
+    {
+      "epoch": 0.7139969681657403,
+      "grad_norm": 1.1570433378219604,
+      "learning_rate": 2.2282456943360375e-05,
+      "loss": 0.7633,
+      "step": 1413
+    },
+    {
+      "epoch": 0.7145022738756948,
+      "grad_norm": 0.5114752054214478,
+      "learning_rate": 2.225116675253418e-05,
+      "loss": 0.2895,
+      "step": 1414
+    },
+    {
+      "epoch": 0.7150075795856493,
+      "grad_norm": 0.3823153078556061,
+      "learning_rate": 2.2219880919906244e-05,
+      "loss": 0.2226,
+      "step": 1415
+    },
+    {
+      "epoch": 0.7155128852956039,
+      "grad_norm": 0.41869547963142395,
+      "learning_rate": 2.2188599495079384e-05,
+      "loss": 0.2079,
+      "step": 1416
+    },
+    {
+      "epoch": 0.7160181910055584,
+      "grad_norm": 0.43806976079940796,
+      "learning_rate": 2.2157322527649422e-05,
+      "loss": 0.1898,
+      "step": 1417
+    },
+    {
+      "epoch": 0.7165234967155129,
+      "grad_norm": 0.8781391382217407,
+      "learning_rate": 2.2126050067205127e-05,
+      "loss": 0.3187,
+      "step": 1418
+    },
+    {
+      "epoch": 0.7170288024254674,
+      "grad_norm": 0.5207377672195435,
+      "learning_rate": 2.209478216332809e-05,
+      "loss": 0.2541,
+      "step": 1419
+    },
+    {
+      "epoch": 0.717534108135422,
+      "grad_norm": 0.7016043663024902,
+      "learning_rate": 2.206351886559271e-05,
+      "loss": 0.2743,
+      "step": 1420
+    },
+    {
+      "epoch": 0.7180394138453765,
+      "grad_norm": 0.24266977608203888,
+      "learning_rate": 2.2032260223566072e-05,
+      "loss": 0.1328,
+      "step": 1421
+    },
+    {
+      "epoch": 0.718544719555331,
+      "grad_norm": 0.9864704608917236,
+      "learning_rate": 2.2001006286807875e-05,
+      "loss": 0.3871,
+      "step": 1422
+    },
+    {
+      "epoch": 0.7190500252652855,
+      "grad_norm": 0.3588007092475891,
+      "learning_rate": 2.1969757104870367e-05,
+      "loss": 0.143,
+      "step": 1423
+    },
+    {
+      "epoch": 0.71955533097524,
+      "grad_norm": 0.716530978679657,
+      "learning_rate": 2.1938512727298246e-05,
+      "loss": 0.3186,
+      "step": 1424
+    },
+    {
+      "epoch": 0.7200606366851945,
+      "grad_norm": 0.5047246813774109,
+      "learning_rate": 2.19072732036286e-05,
+      "loss": 0.1932,
+      "step": 1425
+    },
+    {
+      "epoch": 0.720565942395149,
+      "grad_norm": 0.6216627955436707,
+      "learning_rate": 2.1876038583390825e-05,
+      "loss": 0.4384,
+      "step": 1426
+    },
+    {
+      "epoch": 0.7210712481051036,
+      "grad_norm": 0.5591406226158142,
+      "learning_rate": 2.1844808916106536e-05,
+      "loss": 0.2485,
+      "step": 1427
+    },
+    {
+      "epoch": 0.7215765538150581,
+      "grad_norm": 0.6369313597679138,
+      "learning_rate": 2.1813584251289505e-05,
+      "loss": 0.4893,
+      "step": 1428
+    },
+    {
+      "epoch": 0.7220818595250126,
+      "grad_norm": 0.8026990294456482,
+      "learning_rate": 2.1782364638445546e-05,
+      "loss": 0.6091,
+      "step": 1429
+    },
+    {
+      "epoch": 0.7225871652349671,
+      "grad_norm": 0.45339179039001465,
+      "learning_rate": 2.17511501270725e-05,
+      "loss": 0.3557,
+      "step": 1430
+    },
+    {
+      "epoch": 0.7230924709449217,
+      "grad_norm": 0.5372389554977417,
+      "learning_rate": 2.171994076666009e-05,
+      "loss": 0.5554,
+      "step": 1431
+    },
+    {
+      "epoch": 0.7235977766548762,
+      "grad_norm": 0.45178845524787903,
+      "learning_rate": 2.168873660668989e-05,
+      "loss": 0.2379,
+      "step": 1432
+    },
+    {
+      "epoch": 0.7241030823648307,
+      "grad_norm": 0.4662405550479889,
+      "learning_rate": 2.1657537696635226e-05,
+      "loss": 0.2246,
+      "step": 1433
+    },
+    {
+      "epoch": 0.7246083880747852,
+      "grad_norm": 0.7474414706230164,
+      "learning_rate": 2.16263440859611e-05,
+      "loss": 0.4195,
+      "step": 1434
+    },
+    {
+      "epoch": 0.7251136937847398,
+      "grad_norm": 0.6440469622612,
+      "learning_rate": 2.1595155824124097e-05,
+      "loss": 0.3275,
+      "step": 1435
+    },
+    {
+      "epoch": 0.7256189994946943,
+      "grad_norm": 0.40439724922180176,
+      "learning_rate": 2.1563972960572343e-05,
+      "loss": 0.1541,
+      "step": 1436
+    },
+    {
+      "epoch": 0.7261243052046488,
+      "grad_norm": 0.7414180636405945,
+      "learning_rate": 2.1532795544745392e-05,
+      "loss": 0.3458,
+      "step": 1437
+    },
+    {
+      "epoch": 0.7266296109146033,
+      "grad_norm": 0.5826343894004822,
+      "learning_rate": 2.1501623626074175e-05,
+      "loss": 0.2742,
+      "step": 1438
+    },
+    {
+      "epoch": 0.7271349166245579,
+      "grad_norm": 0.38583678007125854,
+      "learning_rate": 2.1470457253980887e-05,
+      "loss": 0.1891,
+      "step": 1439
+    },
+    {
+      "epoch": 0.7276402223345124,
+      "grad_norm": 0.4571583867073059,
+      "learning_rate": 2.143929647787894e-05,
+      "loss": 0.2069,
+      "step": 1440
+    },
+    {
+      "epoch": 0.7281455280444669,
+      "grad_norm": 0.37782159447669983,
+      "learning_rate": 2.1408141347172877e-05,
+      "loss": 0.2598,
+      "step": 1441
+    },
+    {
+      "epoch": 0.7286508337544214,
+      "grad_norm": 1.0491501092910767,
+      "learning_rate": 2.13769919112583e-05,
+      "loss": 0.4662,
+      "step": 1442
+    },
+    {
+      "epoch": 0.729156139464376,
+      "grad_norm": 0.4999569058418274,
+      "learning_rate": 2.134584821952174e-05,
+      "loss": 0.2569,
+      "step": 1443
+    },
+    {
+      "epoch": 0.7296614451743305,
+      "grad_norm": 0.422539621591568,
+      "learning_rate": 2.131471032134067e-05,
+      "loss": 0.2263,
+      "step": 1444
+    },
+    {
+      "epoch": 0.730166750884285,
+      "grad_norm": 0.7593914866447449,
+      "learning_rate": 2.128357826608335e-05,
+      "loss": 0.3716,
+      "step": 1445
+    },
+    {
+      "epoch": 0.7306720565942395,
+      "grad_norm": 0.6366856098175049,
+      "learning_rate": 2.125245210310878e-05,
+      "loss": 0.2571,
+      "step": 1446
+    },
+    {
+      "epoch": 0.7311773623041941,
+      "grad_norm": 0.46671852469444275,
+      "learning_rate": 2.122133188176664e-05,
+      "loss": 0.4696,
+      "step": 1447
+    },
+    {
+      "epoch": 0.7316826680141486,
+      "grad_norm": 0.37286749482154846,
+      "learning_rate": 2.1190217651397145e-05,
+      "loss": 0.2091,
+      "step": 1448
+    },
+    {
+      "epoch": 0.732187973724103,
+      "grad_norm": 0.8244792819023132,
+      "learning_rate": 2.115910946133105e-05,
+      "loss": 0.2842,
+      "step": 1449
+    },
+    {
+      "epoch": 0.7326932794340576,
+      "grad_norm": 0.9506634473800659,
+      "learning_rate": 2.112800736088952e-05,
+      "loss": 0.365,
+      "step": 1450
+    },
+    {
+      "epoch": 0.7331985851440121,
+      "grad_norm": 0.9866954684257507,
+      "learning_rate": 2.1096911399384066e-05,
+      "loss": 0.3483,
+      "step": 1451
+    },
+    {
+      "epoch": 0.7337038908539667,
+      "grad_norm": 0.4005601406097412,
+      "learning_rate": 2.1065821626116466e-05,
+      "loss": 0.2033,
+      "step": 1452
+    },
+    {
+      "epoch": 0.7342091965639211,
+      "grad_norm": 0.43610894680023193,
+      "learning_rate": 2.1034738090378688e-05,
+      "loss": 0.2345,
+      "step": 1453
+    },
+    {
+      "epoch": 0.7347145022738757,
+      "grad_norm": 0.2879337668418884,
+      "learning_rate": 2.10036608414528e-05,
+      "loss": 0.1507,
+      "step": 1454
+    },
+    {
+      "epoch": 0.7352198079838302,
+      "grad_norm": 0.8920745849609375,
+      "learning_rate": 2.0972589928610917e-05,
+      "loss": 0.3903,
+      "step": 1455
+    },
+    {
+      "epoch": 0.7357251136937848,
+      "grad_norm": 0.8228932619094849,
+      "learning_rate": 2.0941525401115106e-05,
+      "loss": 0.3689,
+      "step": 1456
+    },
+    {
+      "epoch": 0.7362304194037392,
+      "grad_norm": 0.3942200839519501,
+      "learning_rate": 2.0910467308217314e-05,
+      "loss": 0.1604,
+      "step": 1457
+    },
+    {
+      "epoch": 0.7367357251136938,
+      "grad_norm": 0.47732532024383545,
+      "learning_rate": 2.0879415699159256e-05,
+      "loss": 0.3561,
+      "step": 1458
+    },
+    {
+      "epoch": 0.7372410308236483,
+      "grad_norm": 0.6839754581451416,
+      "learning_rate": 2.0848370623172405e-05,
+      "loss": 0.3294,
+      "step": 1459
+    },
+    {
+      "epoch": 0.7377463365336028,
+      "grad_norm": 0.35987260937690735,
+      "learning_rate": 2.081733212947786e-05,
+      "loss": 0.2124,
+      "step": 1460
+    },
+    {
+      "epoch": 0.7382516422435573,
+      "grad_norm": 0.5114133954048157,
+      "learning_rate": 2.078630026728628e-05,
+      "loss": 0.3365,
+      "step": 1461
+    },
+    {
+      "epoch": 0.7387569479535119,
+      "grad_norm": 0.6798128485679626,
+      "learning_rate": 2.0755275085797833e-05,
+      "loss": 0.3135,
+      "step": 1462
+    },
+    {
+      "epoch": 0.7392622536634664,
+      "grad_norm": 0.7099087238311768,
+      "learning_rate": 2.0724256634202052e-05,
+      "loss": 0.2584,
+      "step": 1463
+    },
+    {
+      "epoch": 0.7397675593734209,
+      "grad_norm": 0.6906381845474243,
+      "learning_rate": 2.069324496167783e-05,
+      "loss": 0.2981,
+      "step": 1464
+    },
+    {
+      "epoch": 0.7402728650833754,
+      "grad_norm": 0.2697136402130127,
+      "learning_rate": 2.0662240117393318e-05,
+      "loss": 0.174,
+      "step": 1465
+    },
+    {
+      "epoch": 0.74077817079333,
+      "grad_norm": 0.5975183248519897,
+      "learning_rate": 2.0631242150505827e-05,
+      "loss": 0.3088,
+      "step": 1466
+    },
+    {
+      "epoch": 0.7412834765032845,
+      "grad_norm": 0.4744568467140198,
+      "learning_rate": 2.060025111016176e-05,
+      "loss": 0.1892,
+      "step": 1467
+    },
+    {
+      "epoch": 0.741788782213239,
+      "grad_norm": 0.7941866517066956,
+      "learning_rate": 2.056926704549655e-05,
+      "loss": 0.3482,
+      "step": 1468
+    },
+    {
+      "epoch": 0.7422940879231935,
+      "grad_norm": 0.827198326587677,
+      "learning_rate": 2.0538290005634564e-05,
+      "loss": 0.3751,
+      "step": 1469
+    },
+    {
+      "epoch": 0.7427993936331481,
+      "grad_norm": 0.38413000106811523,
+      "learning_rate": 2.050732003968903e-05,
+      "loss": 0.3045,
+      "step": 1470
+    },
+    {
+      "epoch": 0.7433046993431026,
+      "grad_norm": 0.4320742189884186,
+      "learning_rate": 2.0476357196761967e-05,
+      "loss": 0.2605,
+      "step": 1471
+    },
+    {
+      "epoch": 0.743810005053057,
+      "grad_norm": 0.6644576191902161,
+      "learning_rate": 2.0445401525944107e-05,
+      "loss": 0.2997,
+      "step": 1472
+    },
+    {
+      "epoch": 0.7443153107630116,
+      "grad_norm": 0.6894829869270325,
+      "learning_rate": 2.0414453076314776e-05,
+      "loss": 0.4918,
+      "step": 1473
+    },
+    {
+      "epoch": 0.7448206164729662,
+      "grad_norm": 0.35164880752563477,
+      "learning_rate": 2.0383511896941893e-05,
+      "loss": 0.3007,
+      "step": 1474
+    },
+    {
+      "epoch": 0.7453259221829207,
+      "grad_norm": 0.6786766648292542,
+      "learning_rate": 2.0352578036881826e-05,
+      "loss": 0.4865,
+      "step": 1475
+    },
+    {
+      "epoch": 0.7458312278928751,
+      "grad_norm": 0.4968150556087494,
+      "learning_rate": 2.0321651545179348e-05,
+      "loss": 0.2479,
+      "step": 1476
+    },
+    {
+      "epoch": 0.7463365336028297,
+      "grad_norm": 0.8302887678146362,
+      "learning_rate": 2.029073247086756e-05,
+      "loss": 0.2688,
+      "step": 1477
+    },
+    {
+      "epoch": 0.7468418393127843,
+      "grad_norm": 0.44390469789505005,
+      "learning_rate": 2.0259820862967765e-05,
+      "loss": 0.3654,
+      "step": 1478
+    },
+    {
+      "epoch": 0.7473471450227388,
+      "grad_norm": 0.41298776865005493,
+      "learning_rate": 2.0228916770489466e-05,
+      "loss": 0.2194,
+      "step": 1479
+    },
+    {
+      "epoch": 0.7478524507326932,
+      "grad_norm": 0.46970030665397644,
+      "learning_rate": 2.019802024243023e-05,
+      "loss": 0.2061,
+      "step": 1480
+    },
+    {
+      "epoch": 0.7483577564426478,
+      "grad_norm": 0.510265052318573,
+      "learning_rate": 2.0167131327775652e-05,
+      "loss": 0.237,
+      "step": 1481
+    },
+    {
+      "epoch": 0.7488630621526023,
+      "grad_norm": 0.2790643870830536,
+      "learning_rate": 2.013625007549924e-05,
+      "loss": 0.2037,
+      "step": 1482
+    },
+    {
+      "epoch": 0.7493683678625569,
+      "grad_norm": 0.5583669543266296,
+      "learning_rate": 2.0105376534562344e-05,
+      "loss": 0.1887,
+      "step": 1483
+    },
+    {
+      "epoch": 0.7498736735725113,
+      "grad_norm": 0.4619552493095398,
+      "learning_rate": 2.0074510753914112e-05,
+      "loss": 0.1662,
+      "step": 1484
+    },
+    {
+      "epoch": 0.7503789792824659,
+      "grad_norm": 0.4680952727794647,
+      "learning_rate": 2.004365278249137e-05,
+      "loss": 0.3742,
+      "step": 1485
+    },
+    {
+      "epoch": 0.7508842849924204,
+      "grad_norm": 0.524354875087738,
+      "learning_rate": 2.0012802669218568e-05,
+      "loss": 0.2715,
+      "step": 1486
+    },
+    {
+      "epoch": 0.751389590702375,
+      "grad_norm": 0.21866445243358612,
+      "learning_rate": 1.998196046300771e-05,
+      "loss": 0.1283,
+      "step": 1487
+    },
+    {
+      "epoch": 0.7518948964123294,
+      "grad_norm": 0.3985232412815094,
+      "learning_rate": 1.9951126212758238e-05,
+      "loss": 0.2006,
+      "step": 1488
+    },
+    {
+      "epoch": 0.752400202122284,
+      "grad_norm": 0.4596607983112335,
+      "learning_rate": 1.9920299967356997e-05,
+      "loss": 0.4051,
+      "step": 1489
+    },
+    {
+      "epoch": 0.7529055078322385,
+      "grad_norm": 0.3508613705635071,
+      "learning_rate": 1.9889481775678134e-05,
+      "loss": 0.2223,
+      "step": 1490
+    },
+    {
+      "epoch": 0.7534108135421931,
+      "grad_norm": 0.2925485372543335,
+      "learning_rate": 1.9858671686583037e-05,
+      "loss": 0.1573,
+      "step": 1491
+    },
+    {
+      "epoch": 0.7539161192521475,
+      "grad_norm": 0.6288982033729553,
+      "learning_rate": 1.9827869748920252e-05,
+      "loss": 0.4522,
+      "step": 1492
+    },
+    {
+      "epoch": 0.7544214249621021,
+      "grad_norm": 0.6666999459266663,
+      "learning_rate": 1.9797076011525365e-05,
+      "loss": 0.5193,
+      "step": 1493
+    },
+    {
+      "epoch": 0.7549267306720566,
+      "grad_norm": 0.5883360505104065,
+      "learning_rate": 1.9766290523221e-05,
+      "loss": 0.3678,
+      "step": 1494
+    },
+    {
+      "epoch": 0.7554320363820111,
+      "grad_norm": 1.0901826620101929,
+      "learning_rate": 1.973551333281668e-05,
+      "loss": 0.638,
+      "step": 1495
+    },
+    {
+      "epoch": 0.7559373420919656,
+      "grad_norm": 0.41034024953842163,
+      "learning_rate": 1.9704744489108792e-05,
+      "loss": 0.1508,
+      "step": 1496
+    },
+    {
+      "epoch": 0.7564426478019202,
+      "grad_norm": 0.7889742851257324,
+      "learning_rate": 1.9673984040880468e-05,
+      "loss": 0.44,
+      "step": 1497
+    },
+    {
+      "epoch": 0.7569479535118747,
+      "grad_norm": 0.4004971385002136,
+      "learning_rate": 1.964323203690154e-05,
+      "loss": 0.1833,
+      "step": 1498
+    },
+    {
+      "epoch": 0.7574532592218292,
+      "grad_norm": 1.0051130056381226,
+      "learning_rate": 1.9612488525928452e-05,
+      "loss": 0.2733,
+      "step": 1499
+    },
+    {
+      "epoch": 0.7579585649317837,
+      "grad_norm": 0.49123916029930115,
+      "learning_rate": 1.958175355670418e-05,
+      "loss": 0.2622,
+      "step": 1500
+    },
+    {
+      "epoch": 0.7584638706417383,
+      "grad_norm": 0.581283688545227,
+      "learning_rate": 1.955102717795817e-05,
+      "loss": 0.1906,
+      "step": 1501
+    },
+    {
+      "epoch": 0.7589691763516928,
+      "grad_norm": 0.7873732447624207,
+      "learning_rate": 1.9520309438406217e-05,
+      "loss": 0.2679,
+      "step": 1502
+    },
+    {
+      "epoch": 0.7594744820616472,
+      "grad_norm": 0.5490037798881531,
+      "learning_rate": 1.9489600386750447e-05,
+      "loss": 0.2416,
+      "step": 1503
+    },
+    {
+      "epoch": 0.7599797877716018,
+      "grad_norm": 0.5317832231521606,
+      "learning_rate": 1.9458900071679202e-05,
+      "loss": 0.2828,
+      "step": 1504
+    },
+    {
+      "epoch": 0.7604850934815564,
+      "grad_norm": 0.6812729835510254,
+      "learning_rate": 1.9428208541866973e-05,
+      "loss": 0.3149,
+      "step": 1505
+    },
+    {
+      "epoch": 0.7609903991915109,
+      "grad_norm": 0.6650682091712952,
+      "learning_rate": 1.9397525845974336e-05,
+      "loss": 0.4121,
+      "step": 1506
+    },
+    {
+      "epoch": 0.7614957049014653,
+      "grad_norm": 0.5870944261550903,
+      "learning_rate": 1.9366852032647826e-05,
+      "loss": 0.3445,
+      "step": 1507
+    },
+    {
+      "epoch": 0.7620010106114199,
+      "grad_norm": 0.5600792765617371,
+      "learning_rate": 1.9336187150519924e-05,
+      "loss": 0.2882,
+      "step": 1508
+    },
+    {
+      "epoch": 0.7625063163213744,
+      "grad_norm": 0.45185530185699463,
+      "learning_rate": 1.9305531248208946e-05,
+      "loss": 0.1826,
+      "step": 1509
+    },
+    {
+      "epoch": 0.763011622031329,
+      "grad_norm": 0.7576361894607544,
+      "learning_rate": 1.9274884374318967e-05,
+      "loss": 0.3394,
+      "step": 1510
+    },
+    {
+      "epoch": 0.7635169277412834,
+      "grad_norm": 0.3762153089046478,
+      "learning_rate": 1.9244246577439752e-05,
+      "loss": 0.1961,
+      "step": 1511
+    },
+    {
+      "epoch": 0.764022233451238,
+      "grad_norm": 0.6146969199180603,
+      "learning_rate": 1.9213617906146663e-05,
+      "loss": 0.3267,
+      "step": 1512
+    },
+    {
+      "epoch": 0.7645275391611925,
+      "grad_norm": 0.5639180541038513,
+      "learning_rate": 1.9182998409000606e-05,
+      "loss": 0.2919,
+      "step": 1513
+    },
+    {
+      "epoch": 0.7650328448711471,
+      "grad_norm": 0.5925596952438354,
+      "learning_rate": 1.9152388134547945e-05,
+      "loss": 0.2764,
+      "step": 1514
+    },
+    {
+      "epoch": 0.7655381505811015,
+      "grad_norm": 0.41065713763237,
+      "learning_rate": 1.91217871313204e-05,
+      "loss": 0.188,
+      "step": 1515
+    },
+    {
+      "epoch": 0.7660434562910561,
+      "grad_norm": 0.38348203897476196,
+      "learning_rate": 1.909119544783503e-05,
+      "loss": 0.2152,
+      "step": 1516
+    },
+    {
+      "epoch": 0.7665487620010106,
+      "grad_norm": 0.4948267340660095,
+      "learning_rate": 1.9060613132594067e-05,
+      "loss": 0.2266,
+      "step": 1517
+    },
+    {
+      "epoch": 0.7670540677109652,
+      "grad_norm": 0.5641569495201111,
+      "learning_rate": 1.9030040234084926e-05,
+      "loss": 0.2611,
+      "step": 1518
+    },
+    {
+      "epoch": 0.7675593734209196,
+      "grad_norm": 0.4439907670021057,
+      "learning_rate": 1.8999476800780087e-05,
+      "loss": 0.2271,
+      "step": 1519
+    },
+    {
+      "epoch": 0.7680646791308742,
+      "grad_norm": 0.7098090648651123,
+      "learning_rate": 1.8968922881137012e-05,
+      "loss": 0.4192,
+      "step": 1520
+    },
+    {
+      "epoch": 0.7685699848408287,
+      "grad_norm": 0.6672466397285461,
+      "learning_rate": 1.8938378523598103e-05,
+      "loss": 0.3129,
+      "step": 1521
+    },
+    {
+      "epoch": 0.7690752905507833,
+      "grad_norm": 0.672419011592865,
+      "learning_rate": 1.8907843776590567e-05,
+      "loss": 0.3302,
+      "step": 1522
+    },
+    {
+      "epoch": 0.7695805962607377,
+      "grad_norm": 0.5676461458206177,
+      "learning_rate": 1.8877318688526393e-05,
+      "loss": 0.2522,
+      "step": 1523
+    },
+    {
+      "epoch": 0.7700859019706923,
+      "grad_norm": 0.4994024932384491,
+      "learning_rate": 1.8846803307802265e-05,
+      "loss": 0.1974,
+      "step": 1524
+    },
+    {
+      "epoch": 0.7705912076806468,
+      "grad_norm": 0.4182925820350647,
+      "learning_rate": 1.8816297682799452e-05,
+      "loss": 0.1786,
+      "step": 1525
+    },
+    {
+      "epoch": 0.7710965133906014,
+      "grad_norm": 0.6781833171844482,
+      "learning_rate": 1.8785801861883788e-05,
+      "loss": 0.4861,
+      "step": 1526
+    },
+    {
+      "epoch": 0.7716018191005558,
+      "grad_norm": 0.6011713147163391,
+      "learning_rate": 1.875531589340553e-05,
+      "loss": 0.2655,
+      "step": 1527
+    },
+    {
+      "epoch": 0.7721071248105104,
+      "grad_norm": 0.6764652729034424,
+      "learning_rate": 1.8724839825699332e-05,
+      "loss": 0.5056,
+      "step": 1528
+    },
+    {
+      "epoch": 0.7726124305204649,
+      "grad_norm": 0.5193274021148682,
+      "learning_rate": 1.869437370708415e-05,
+      "loss": 0.2291,
+      "step": 1529
+    },
+    {
+      "epoch": 0.7731177362304194,
+      "grad_norm": 0.5534316301345825,
+      "learning_rate": 1.8663917585863154e-05,
+      "loss": 0.2526,
+      "step": 1530
+    },
+    {
+      "epoch": 0.7736230419403739,
+      "grad_norm": 0.4299106299877167,
+      "learning_rate": 1.8633471510323694e-05,
+      "loss": 0.1945,
+      "step": 1531
+    },
+    {
+      "epoch": 0.7741283476503285,
+      "grad_norm": 0.5962615013122559,
+      "learning_rate": 1.8603035528737146e-05,
+      "loss": 0.3837,
+      "step": 1532
+    },
+    {
+      "epoch": 0.774633653360283,
+      "grad_norm": 0.6587054133415222,
+      "learning_rate": 1.8572609689358923e-05,
+      "loss": 0.3066,
+      "step": 1533
+    },
+    {
+      "epoch": 0.7751389590702374,
+      "grad_norm": 0.919040858745575,
+      "learning_rate": 1.854219404042834e-05,
+      "loss": 0.369,
+      "step": 1534
+    },
+    {
+      "epoch": 0.775644264780192,
+      "grad_norm": 0.4253031611442566,
+      "learning_rate": 1.8511788630168557e-05,
+      "loss": 0.199,
+      "step": 1535
+    },
+    {
+      "epoch": 0.7761495704901465,
+      "grad_norm": 0.6377377510070801,
+      "learning_rate": 1.848139350678651e-05,
+      "loss": 0.3452,
+      "step": 1536
+    },
+    {
+      "epoch": 0.7766548762001011,
+      "grad_norm": 0.9110857248306274,
+      "learning_rate": 1.8451008718472802e-05,
+      "loss": 0.3778,
+      "step": 1537
+    },
+    {
+      "epoch": 0.7771601819100555,
+      "grad_norm": 0.42227810621261597,
+      "learning_rate": 1.8420634313401673e-05,
+      "loss": 0.1541,
+      "step": 1538
+    },
+    {
+      "epoch": 0.7776654876200101,
+      "grad_norm": 0.5211955904960632,
+      "learning_rate": 1.8390270339730892e-05,
+      "loss": 0.3227,
+      "step": 1539
+    },
+    {
+      "epoch": 0.7781707933299646,
+      "grad_norm": 0.366298645734787,
+      "learning_rate": 1.835991684560169e-05,
+      "loss": 0.1747,
+      "step": 1540
+    },
+    {
+      "epoch": 0.7786760990399192,
+      "grad_norm": 0.4732970893383026,
+      "learning_rate": 1.8329573879138695e-05,
+      "loss": 0.2065,
+      "step": 1541
+    },
+    {
+      "epoch": 0.7791814047498736,
+      "grad_norm": 0.43378695845603943,
+      "learning_rate": 1.8299241488449814e-05,
+      "loss": 0.1849,
+      "step": 1542
+    },
+    {
+      "epoch": 0.7796867104598282,
+      "grad_norm": 0.6272542476654053,
+      "learning_rate": 1.826891972162621e-05,
+      "loss": 0.2282,
+      "step": 1543
+    },
+    {
+      "epoch": 0.7801920161697827,
+      "grad_norm": 0.46285495162010193,
+      "learning_rate": 1.82386086267422e-05,
+      "loss": 0.2114,
+      "step": 1544
+    },
+    {
+      "epoch": 0.7806973218797373,
+      "grad_norm": 0.5049265623092651,
+      "learning_rate": 1.8208308251855176e-05,
+      "loss": 0.5401,
+      "step": 1545
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2573,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 515,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.4075086284361073e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}