diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9645 @@
+{
+  "best_metric": 0.8666666666666667,
+  "best_model_checkpoint": "CTMAE-P2-V5-3g-S5/checkpoint-3654",
+  "epoch": 49.02,
+  "eval_steps": 500,
+  "global_step": 13050,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0007662835249042146,
+      "grad_norm": 6.515963554382324,
+      "learning_rate": 7.662835249042146e-08,
+      "loss": 0.7348,
+      "step": 10
+    },
+    {
+      "epoch": 0.0015325670498084292,
+      "grad_norm": 6.009317874908447,
+      "learning_rate": 1.5325670498084292e-07,
+      "loss": 0.8031,
+      "step": 20
+    },
+    {
+      "epoch": 0.0022988505747126436,
+      "grad_norm": 4.49639892578125,
+      "learning_rate": 2.2988505747126437e-07,
+      "loss": 0.71,
+      "step": 30
+    },
+    {
+      "epoch": 0.0030651340996168583,
+      "grad_norm": 5.10139799118042,
+      "learning_rate": 3.0651340996168583e-07,
+      "loss": 0.7344,
+      "step": 40
+    },
+    {
+      "epoch": 0.0038314176245210726,
+      "grad_norm": 6.141551971435547,
+      "learning_rate": 3.831417624521073e-07,
+      "loss": 0.7151,
+      "step": 50
+    },
+    {
+      "epoch": 0.004597701149425287,
+      "grad_norm": 4.640506267547607,
+      "learning_rate": 4.5977011494252875e-07,
+      "loss": 0.672,
+      "step": 60
+    },
+    {
+      "epoch": 0.0053639846743295016,
+      "grad_norm": 6.34081506729126,
+      "learning_rate": 5.363984674329502e-07,
+      "loss": 0.779,
+      "step": 70
+    },
+    {
+      "epoch": 0.006130268199233717,
+      "grad_norm": 5.613821983337402,
+      "learning_rate": 6.130268199233717e-07,
+      "loss": 0.7196,
+      "step": 80
+    },
+    {
+      "epoch": 0.006896551724137931,
+      "grad_norm": 6.071259021759033,
+      "learning_rate": 6.896551724137931e-07,
+      "loss": 0.7607,
+      "step": 90
+    },
+    {
+      "epoch": 0.007662835249042145,
+      "grad_norm": 5.040432453155518,
+      "learning_rate": 7.662835249042146e-07,
+      "loss": 0.6946,
+      "step": 100
+    },
+    {
+      "epoch": 0.00842911877394636,
+      "grad_norm": 5.427196502685547,
+      "learning_rate": 8.429118773946361e-07,
+      "loss": 0.7196,
+      "step": 110
+    },
+    {
+      "epoch": 0.009195402298850575,
+      "grad_norm": 5.190512180328369,
+      "learning_rate": 9.195402298850575e-07,
+      "loss": 0.7041,
+      "step": 120
+    },
+    {
+      "epoch": 0.00996168582375479,
+      "grad_norm": 4.816593647003174,
+      "learning_rate": 9.96168582375479e-07,
+      "loss": 0.6509,
+      "step": 130
+    },
+    {
+      "epoch": 0.010727969348659003,
+      "grad_norm": 4.155104637145996,
+      "learning_rate": 1.0727969348659004e-06,
+      "loss": 0.7013,
+      "step": 140
+    },
+    {
+      "epoch": 0.011494252873563218,
+      "grad_norm": 6.712978363037109,
+      "learning_rate": 1.1494252873563219e-06,
+      "loss": 0.6218,
+      "step": 150
+    },
+    {
+      "epoch": 0.012260536398467433,
+      "grad_norm": 7.192935466766357,
+      "learning_rate": 1.2260536398467433e-06,
+      "loss": 0.6437,
+      "step": 160
+    },
+    {
+      "epoch": 0.013026819923371647,
+      "grad_norm": 3.722414493560791,
+      "learning_rate": 1.3026819923371648e-06,
+      "loss": 0.6214,
+      "step": 170
+    },
+    {
+      "epoch": 0.013793103448275862,
+      "grad_norm": 5.249204635620117,
+      "learning_rate": 1.3793103448275862e-06,
+      "loss": 0.6217,
+      "step": 180
+    },
+    {
+      "epoch": 0.014559386973180077,
+      "grad_norm": 3.773120880126953,
+      "learning_rate": 1.455938697318008e-06,
+      "loss": 0.714,
+      "step": 190
+    },
+    {
+      "epoch": 0.01532567049808429,
+      "grad_norm": 4.748465538024902,
+      "learning_rate": 1.5325670498084292e-06,
+      "loss": 0.6316,
+      "step": 200
+    },
+    {
+      "epoch": 0.016091954022988506,
+      "grad_norm": 4.1444501876831055,
+      "learning_rate": 1.6091954022988506e-06,
+      "loss": 0.7874,
+      "step": 210
+    },
+    {
+      "epoch": 0.01685823754789272,
+      "grad_norm": 5.518341064453125,
+      "learning_rate": 1.6858237547892723e-06,
+      "loss": 0.8583,
+      "step": 220
+    },
+    {
+      "epoch": 0.017624521072796936,
+      "grad_norm": 5.157224655151367,
+      "learning_rate": 1.7624521072796935e-06,
+      "loss": 0.3741,
+      "step": 230
+    },
+    {
+      "epoch": 0.01839080459770115,
+      "grad_norm": 3.7469072341918945,
+      "learning_rate": 1.839080459770115e-06,
+      "loss": 1.2018,
+      "step": 240
+    },
+    {
+      "epoch": 0.019157088122605363,
+      "grad_norm": 2.4285781383514404,
+      "learning_rate": 1.9157088122605367e-06,
+      "loss": 1.0356,
+      "step": 250
+    },
+    {
+      "epoch": 0.01992337164750958,
+      "grad_norm": 3.1459391117095947,
+      "learning_rate": 1.992337164750958e-06,
+      "loss": 1.8752,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "eval_accuracy": 0.4666666666666667,
+      "eval_loss": 2.000931978225708,
+      "eval_runtime": 17.5306,
+      "eval_samples_per_second": 2.567,
+      "eval_steps_per_second": 2.567,
+      "step": 261
+    },
+    {
+      "epoch": 1.0006896551724138,
+      "grad_norm": 1.068664312362671,
+      "learning_rate": 2.0689655172413796e-06,
+      "loss": 0.8689,
+      "step": 270
+    },
+    {
+      "epoch": 1.001455938697318,
+      "grad_norm": 55.32078552246094,
+      "learning_rate": 2.145593869731801e-06,
+      "loss": 2.2527,
+      "step": 280
+    },
+    {
+      "epoch": 1.0022222222222221,
+      "grad_norm": 54.1043701171875,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 1.395,
+      "step": 290
+    },
+    {
+      "epoch": 1.0029885057471264,
+      "grad_norm": 0.6383324265480042,
+      "learning_rate": 2.2988505747126437e-06,
+      "loss": 1.3467,
+      "step": 300
+    },
+    {
+      "epoch": 1.0037547892720307,
+      "grad_norm": 1.1302201747894287,
+      "learning_rate": 2.3754789272030654e-06,
+      "loss": 3.2186,
+      "step": 310
+    },
+    {
+      "epoch": 1.004521072796935,
+      "grad_norm": 1.0692795515060425,
+      "learning_rate": 2.4521072796934867e-06,
+      "loss": 1.545,
+      "step": 320
+    },
+    {
+      "epoch": 1.0052873563218392,
+      "grad_norm": 1.788543462753296,
+      "learning_rate": 2.5287356321839083e-06,
+      "loss": 1.4408,
+      "step": 330
+    },
+    {
+      "epoch": 1.0060536398467432,
+      "grad_norm": 0.7168189287185669,
+      "learning_rate": 2.6053639846743296e-06,
+      "loss": 1.6561,
+      "step": 340
+    },
+    {
+      "epoch": 1.0068199233716475,
+      "grad_norm": 1.9753859043121338,
+      "learning_rate": 2.6819923371647512e-06,
+      "loss": 1.9563,
+      "step": 350
+    },
+    {
+      "epoch": 1.0075862068965518,
+      "grad_norm": 39.59136199951172,
+      "learning_rate": 2.7586206896551725e-06,
+      "loss": 1.5893,
+      "step": 360
+    },
+    {
+      "epoch": 1.008352490421456,
+      "grad_norm": 40.53175354003906,
+      "learning_rate": 2.835249042145594e-06,
+      "loss": 2.2763,
+      "step": 370
+    },
+    {
+      "epoch": 1.00911877394636,
+      "grad_norm": 0.5290699005126953,
+      "learning_rate": 2.911877394636016e-06,
+      "loss": 0.4305,
+      "step": 380
+    },
+    {
+      "epoch": 1.0098850574712643,
+      "grad_norm": 34.556400299072266,
+      "learning_rate": 2.988505747126437e-06,
+      "loss": 2.1726,
+      "step": 390
+    },
+    {
+      "epoch": 1.0106513409961686,
+      "grad_norm": 0.360198050737381,
+      "learning_rate": 3.0651340996168583e-06,
+      "loss": 0.4444,
+      "step": 400
+    },
+    {
+      "epoch": 1.0114176245210729,
+      "grad_norm": 37.28133773803711,
+      "learning_rate": 3.14176245210728e-06,
+      "loss": 1.5316,
+      "step": 410
+    },
+    {
+      "epoch": 1.012183908045977,
+      "grad_norm": 0.7995922565460205,
+      "learning_rate": 3.2183908045977012e-06,
+      "loss": 1.832,
+      "step": 420
+    },
+    {
+      "epoch": 1.0129501915708812,
+      "grad_norm": 0.4065437614917755,
+      "learning_rate": 3.295019157088123e-06,
+      "loss": 1.2957,
+      "step": 430
+    },
+    {
+      "epoch": 1.0137164750957854,
+      "grad_norm": 0.4343510866165161,
+      "learning_rate": 3.3716475095785446e-06,
+      "loss": 1.3974,
+      "step": 440
+    },
+    {
+      "epoch": 1.0144827586206897,
+      "grad_norm": 0.17931339144706726,
+      "learning_rate": 3.448275862068966e-06,
+      "loss": 0.009,
+      "step": 450
+    },
+    {
+      "epoch": 1.015249042145594,
+      "grad_norm": 38.98386001586914,
+      "learning_rate": 3.524904214559387e-06,
+      "loss": 2.1494,
+      "step": 460
+    },
+    {
+      "epoch": 1.016015325670498,
+      "grad_norm": 0.6379644870758057,
+      "learning_rate": 3.6015325670498087e-06,
+      "loss": 1.4487,
+      "step": 470
+    },
+    {
+      "epoch": 1.0167816091954023,
+      "grad_norm": 35.66837692260742,
+      "learning_rate": 3.67816091954023e-06,
+      "loss": 2.0168,
+      "step": 480
+    },
+    {
+      "epoch": 1.0175478927203065,
+      "grad_norm": 1.7801227569580078,
+      "learning_rate": 3.7547892720306517e-06,
+      "loss": 1.7985,
+      "step": 490
+    },
+    {
+      "epoch": 1.0183141762452108,
+      "grad_norm": 0.5001053810119629,
+      "learning_rate": 3.831417624521073e-06,
+      "loss": 0.3404,
+      "step": 500
+    },
+    {
+      "epoch": 1.0190804597701149,
+      "grad_norm": 0.29768675565719604,
+      "learning_rate": 3.908045977011495e-06,
+      "loss": 0.901,
+      "step": 510
+    },
+    {
+      "epoch": 1.0198467432950191,
+      "grad_norm": 0.17741112411022186,
+      "learning_rate": 3.984674329501916e-06,
+      "loss": 0.5365,
+      "step": 520
+    },
+    {
+      "epoch": 1.02,
+      "eval_accuracy": 0.4666666666666667,
+      "eval_loss": 2.8415279388427734,
+      "eval_runtime": 16.3662,
+      "eval_samples_per_second": 2.75,
+      "eval_steps_per_second": 2.75,
+      "step": 522
+    },
+    {
+      "epoch": 2.0006130268199236,
+      "grad_norm": 69.17105102539062,
+      "learning_rate": 4.0613026819923375e-06,
+      "loss": 2.5607,
+      "step": 530
+    },
+    {
+      "epoch": 2.0013793103448276,
+      "grad_norm": 24.83037567138672,
+      "learning_rate": 4.137931034482759e-06,
+      "loss": 2.1662,
+      "step": 540
+    },
+    {
+      "epoch": 2.0021455938697317,
+      "grad_norm": 31.966571807861328,
+      "learning_rate": 4.214559386973181e-06,
+      "loss": 0.6009,
+      "step": 550
+    },
+    {
+      "epoch": 2.002911877394636,
+      "grad_norm": 42.25388717651367,
+      "learning_rate": 4.291187739463602e-06,
+      "loss": 1.5873,
+      "step": 560
+    },
+    {
+      "epoch": 2.00367816091954,
+      "grad_norm": 0.9124504327774048,
+      "learning_rate": 4.367816091954023e-06,
+      "loss": 1.5907,
+      "step": 570
+    },
+    {
+      "epoch": 2.0044444444444443,
+      "grad_norm": 0.7654790878295898,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 1.3789,
+      "step": 580
+    },
+    {
+      "epoch": 2.0052107279693487,
+      "grad_norm": 0.4567537307739258,
+      "learning_rate": 4.521072796934866e-06,
+      "loss": 1.3995,
+      "step": 590
+    },
+    {
+      "epoch": 2.005977011494253,
+      "grad_norm": 0.9925104975700378,
+      "learning_rate": 4.5977011494252875e-06,
+      "loss": 1.995,
+      "step": 600
+    },
+    {
+      "epoch": 2.0067432950191573,
+      "grad_norm": 0.2815808653831482,
+      "learning_rate": 4.674329501915709e-06,
+      "loss": 0.3418,
+      "step": 610
+    },
+    {
+      "epoch": 2.0075095785440613,
+      "grad_norm": 0.21866410970687866,
+      "learning_rate": 4.750957854406131e-06,
+      "loss": 1.5961,
+      "step": 620
+    },
+    {
+      "epoch": 2.0082758620689654,
+      "grad_norm": 0.11550973355770111,
+      "learning_rate": 4.8275862068965525e-06,
+      "loss": 0.0053,
+      "step": 630
+    },
+    {
+      "epoch": 2.00904214559387,
+      "grad_norm": 31.14862823486328,
+      "learning_rate": 4.904214559386973e-06,
+      "loss": 2.2064,
+      "step": 640
+    },
+    {
+      "epoch": 2.009808429118774,
+      "grad_norm": 36.643001556396484,
+      "learning_rate": 4.980842911877395e-06,
+      "loss": 1.3828,
+      "step": 650
+    },
+    {
+      "epoch": 2.0105747126436784,
+      "grad_norm": 0.2615107297897339,
+      "learning_rate": 5.057471264367817e-06,
+      "loss": 0.4727,
+      "step": 660
+    },
+    {
+      "epoch": 2.0113409961685824,
+      "grad_norm": 28.37118148803711,
+      "learning_rate": 5.134099616858238e-06,
+      "loss": 0.9614,
+      "step": 670
+    },
+    {
+      "epoch": 2.0121072796934865,
+      "grad_norm": 30.042430877685547,
+      "learning_rate": 5.210727969348659e-06,
+      "loss": 2.7478,
+      "step": 680
+    },
+    {
+      "epoch": 2.012873563218391,
+      "grad_norm": 0.43138423562049866,
+      "learning_rate": 5.287356321839081e-06,
+      "loss": 0.3899,
+      "step": 690
+    },
+    {
+      "epoch": 2.013639846743295,
+      "grad_norm": 26.292491912841797,
+      "learning_rate": 5.3639846743295025e-06,
+      "loss": 1.8188,
+      "step": 700
+    },
+    {
+      "epoch": 2.014406130268199,
+      "grad_norm": 0.4923742115497589,
+      "learning_rate": 5.440613026819924e-06,
+      "loss": 1.3005,
+      "step": 710
+    },
+    {
+      "epoch": 2.0151724137931035,
+      "grad_norm": 0.6332253217697144,
+      "learning_rate": 5.517241379310345e-06,
+      "loss": 2.3829,
+      "step": 720
+    },
+    {
+      "epoch": 2.0159386973180076,
+      "grad_norm": 22.945175170898438,
+      "learning_rate": 5.593869731800766e-06,
+      "loss": 1.7472,
+      "step": 730
+    },
+    {
+      "epoch": 2.016704980842912,
+      "grad_norm": 1.0190438032150269,
+      "learning_rate": 5.670498084291188e-06,
+      "loss": 0.9494,
+      "step": 740
+    },
+    {
+      "epoch": 2.017471264367816,
+      "grad_norm": 23.87836265563965,
+      "learning_rate": 5.747126436781609e-06,
+      "loss": 1.3298,
+      "step": 750
+    },
+    {
+      "epoch": 2.01823754789272,
+      "grad_norm": 0.3306765854358673,
+      "learning_rate": 5.823754789272032e-06,
+      "loss": 0.7995,
+      "step": 760
+    },
+    {
+      "epoch": 2.0190038314176246,
+      "grad_norm": 25.512117385864258,
+      "learning_rate": 5.9003831417624525e-06,
+      "loss": 1.3456,
+      "step": 770
+    },
+    {
+      "epoch": 2.0197701149425287,
+      "grad_norm": 0.47690722346305847,
+      "learning_rate": 5.977011494252874e-06,
+      "loss": 1.8181,
+      "step": 780
+    },
+    {
+      "epoch": 2.02,
+      "eval_accuracy": 0.4666666666666667,
+      "eval_loss": 2.1330909729003906,
+      "eval_runtime": 15.2884,
+      "eval_samples_per_second": 2.943,
+      "eval_steps_per_second": 2.943,
+      "step": 783
+    },
+    {
+      "epoch": 3.000536398467433,
+      "grad_norm": 22.61998176574707,
+      "learning_rate": 6.053639846743296e-06,
+      "loss": 2.2133,
+      "step": 790
+    },
+    {
+      "epoch": 3.001302681992337,
+      "grad_norm": 0.9248330593109131,
+      "learning_rate": 6.130268199233717e-06,
+      "loss": 0.6473,
+      "step": 800
+    },
+    {
+      "epoch": 3.0020689655172412,
+      "grad_norm": 0.3790869116783142,
+      "learning_rate": 6.206896551724138e-06,
+      "loss": 1.1646,
+      "step": 810
+    },
+    {
+      "epoch": 3.0028352490421457,
+      "grad_norm": 0.31349891424179077,
+      "learning_rate": 6.28352490421456e-06,
+      "loss": 1.3614,
+      "step": 820
+    },
+    {
+      "epoch": 3.0036015325670498,
+      "grad_norm": 0.4672112464904785,
+      "learning_rate": 6.360153256704982e-06,
+      "loss": 1.2955,
+      "step": 830
+    },
+    {
+      "epoch": 3.004367816091954,
+      "grad_norm": 24.16468048095703,
+      "learning_rate": 6.4367816091954025e-06,
+      "loss": 0.8488,
+      "step": 840
+    },
+    {
+      "epoch": 3.0051340996168583,
+      "grad_norm": 0.5347825288772583,
+      "learning_rate": 6.513409961685824e-06,
+      "loss": 1.7289,
+      "step": 850
+    },
+    {
+      "epoch": 3.0059003831417623,
+      "grad_norm": 0.4466633200645447,
+      "learning_rate": 6.590038314176246e-06,
+      "loss": 0.7971,
+      "step": 860
+    },
+    {
+      "epoch": 3.006666666666667,
+      "grad_norm": 0.29940515756607056,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.2878,
+      "step": 870
+    },
+    {
+      "epoch": 3.007432950191571,
+      "grad_norm": 0.1792927086353302,
+      "learning_rate": 6.743295019157089e-06,
+      "loss": 0.0103,
+      "step": 880
+    },
+    {
+      "epoch": 3.008199233716475,
+      "grad_norm": 24.740402221679688,
+      "learning_rate": 6.81992337164751e-06,
+      "loss": 1.9771,
+      "step": 890
+    },
+    {
+      "epoch": 3.0089655172413794,
+      "grad_norm": 0.5116446018218994,
+      "learning_rate": 6.896551724137932e-06,
+      "loss": 1.3359,
+      "step": 900
+    },
+    {
+      "epoch": 3.0097318007662834,
+      "grad_norm": 0.3623947203159332,
+      "learning_rate": 6.973180076628353e-06,
+      "loss": 1.6457,
+      "step": 910
+    },
+    {
+      "epoch": 3.010498084291188,
+      "grad_norm": 19.105806350708008,
+      "learning_rate": 7.049808429118774e-06,
+      "loss": 2.4007,
+      "step": 920
+    },
+    {
+      "epoch": 3.011264367816092,
+      "grad_norm": 0.9638222455978394,
+      "learning_rate": 7.126436781609196e-06,
+      "loss": 0.3213,
+      "step": 930
+    },
+    {
+      "epoch": 3.012030651340996,
+      "grad_norm": 22.83616065979004,
+      "learning_rate": 7.2030651340996175e-06,
+      "loss": 1.9054,
+      "step": 940
+    },
+    {
+      "epoch": 3.0127969348659005,
+      "grad_norm": 0.5195755958557129,
+      "learning_rate": 7.279693486590039e-06,
+      "loss": 1.1278,
+      "step": 950
+    },
+    {
+      "epoch": 3.0135632183908045,
+      "grad_norm": 21.764245986938477,
+      "learning_rate": 7.35632183908046e-06,
+      "loss": 1.1217,
+      "step": 960
+    },
+    {
+      "epoch": 3.014329501915709,
+      "grad_norm": 20.36184310913086,
+      "learning_rate": 7.4329501915708825e-06,
+      "loss": 2.2718,
+      "step": 970
+    },
+    {
+      "epoch": 3.015095785440613,
+      "grad_norm": 0.6821834444999695,
+      "learning_rate": 7.509578544061303e-06,
+      "loss": 0.7317,
+      "step": 980
+    },
+    {
+      "epoch": 3.015862068965517,
+      "grad_norm": 0.6219151616096497,
+      "learning_rate": 7.586206896551724e-06,
+      "loss": 1.427,
+      "step": 990
+    },
+    {
+      "epoch": 3.0166283524904216,
+      "grad_norm": 0.6644460558891296,
+      "learning_rate": 7.662835249042147e-06,
+      "loss": 1.0886,
+      "step": 1000
+    },
+    {
+      "epoch": 3.0173946360153256,
+      "grad_norm": 22.19208335876465,
+      "learning_rate": 7.739463601532567e-06,
+      "loss": 2.3383,
+      "step": 1010
+    },
+    {
+      "epoch": 3.0181609195402297,
+      "grad_norm": 1.41365647315979,
+      "learning_rate": 7.81609195402299e-06,
+      "loss": 1.5415,
+      "step": 1020
+    },
+    {
+      "epoch": 3.018927203065134,
+      "grad_norm": 1.0619806051254272,
+      "learning_rate": 7.89272030651341e-06,
+      "loss": 0.8251,
+      "step": 1030
+    },
+    {
+      "epoch": 3.0196934865900382,
+      "grad_norm": 0.8489217758178711,
+      "learning_rate": 7.969348659003832e-06,
+      "loss": 0.6674,
+      "step": 1040
+    },
+    {
+      "epoch": 3.02,
+      "eval_accuracy": 0.4666666666666667,
+      "eval_loss": 2.5702672004699707,
+      "eval_runtime": 15.9903,
+      "eval_samples_per_second": 2.814,
+      "eval_steps_per_second": 2.814,
+      "step": 1044
+    },
+    {
+      "epoch": 4.000459770114943,
+      "grad_norm": 0.16821230947971344,
+      "learning_rate": 8.045977011494253e-06,
+      "loss": 0.9716,
+      "step": 1050
+    },
+    {
+      "epoch": 4.001226053639847,
+      "grad_norm": 29.10614013671875,
+      "learning_rate": 8.122605363984675e-06,
+      "loss": 0.5545,
+      "step": 1060
+    },
+    {
+      "epoch": 4.001992337164751,
+      "grad_norm": 28.9191951751709,
+      "learning_rate": 8.199233716475097e-06,
+      "loss": 1.0854,
+      "step": 1070
+    },
+    {
+      "epoch": 4.002758620689655,
+      "grad_norm": 0.3250425457954407,
+      "learning_rate": 8.275862068965518e-06,
+      "loss": 2.2773,
+      "step": 1080
+    },
+    {
+      "epoch": 4.00352490421456,
+      "grad_norm": 0.6510913968086243,
+      "learning_rate": 8.35249042145594e-06,
+      "loss": 0.7474,
+      "step": 1090
+    },
+    {
+      "epoch": 4.004291187739463,
+      "grad_norm": 23.383602142333984,
+      "learning_rate": 8.429118773946362e-06,
+      "loss": 1.6119,
+      "step": 1100
+    },
+    {
+      "epoch": 4.005057471264368,
+      "grad_norm": 18.749574661254883,
+      "learning_rate": 8.505747126436782e-06,
+      "loss": 1.812,
+      "step": 1110
+    },
+    {
+      "epoch": 4.005823754789272,
+      "grad_norm": 0.4811108112335205,
+      "learning_rate": 8.582375478927203e-06,
+      "loss": 1.2609,
+      "step": 1120
+    },
+    {
+      "epoch": 4.006590038314176,
+      "grad_norm": 2.357823133468628,
+      "learning_rate": 8.659003831417625e-06,
+      "loss": 1.449,
+      "step": 1130
+    },
+    {
+      "epoch": 4.00735632183908,
+      "grad_norm": 22.41640853881836,
+      "learning_rate": 8.735632183908047e-06,
+      "loss": 1.4203,
+      "step": 1140
+    },
+    {
+      "epoch": 4.008122605363985,
+      "grad_norm": 0.318185031414032,
+      "learning_rate": 8.812260536398468e-06,
+      "loss": 0.717,
+      "step": 1150
+    },
+    {
+      "epoch": 4.0088888888888885,
+      "grad_norm": 0.09603653103113174,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 0.9936,
+      "step": 1160
+    },
+    {
+      "epoch": 4.009655172413793,
+      "grad_norm": 0.7078070640563965,
+      "learning_rate": 8.965517241379312e-06,
+      "loss": 2.4035,
+      "step": 1170
+    },
+    {
+      "epoch": 4.0104214559386975,
+      "grad_norm": 1.1963130235671997,
+      "learning_rate": 9.042145593869732e-06,
+      "loss": 1.2541,
+      "step": 1180
+    },
+    {
+      "epoch": 4.011187739463602,
+      "grad_norm": 0.7769260406494141,
+      "learning_rate": 9.118773946360155e-06,
+      "loss": 1.5987,
+      "step": 1190
+    },
+    {
+      "epoch": 4.011954022988506,
+      "grad_norm": 1.0503462553024292,
+      "learning_rate": 9.195402298850575e-06,
+      "loss": 0.9641,
+      "step": 1200
+    },
+    {
+      "epoch": 4.01272030651341,
+      "grad_norm": 0.12320452928543091,
+      "learning_rate": 9.272030651340997e-06,
+      "loss": 0.397,
+      "step": 1210
+    },
+    {
+      "epoch": 4.0134865900383145,
+      "grad_norm": 0.14361780881881714,
+      "learning_rate": 9.348659003831418e-06,
+      "loss": 0.9065,
+      "step": 1220
+    },
+    {
+      "epoch": 4.014252873563218,
+      "grad_norm": 0.17569315433502197,
+      "learning_rate": 9.42528735632184e-06,
+      "loss": 2.1919,
+      "step": 1230
+    },
+    {
+      "epoch": 4.015019157088123,
+      "grad_norm": 0.5088914632797241,
+      "learning_rate": 9.501915708812262e-06,
+      "loss": 1.5856,
+      "step": 1240
+    },
+    {
+      "epoch": 4.015785440613027,
+      "grad_norm": 100.4439468383789,
+      "learning_rate": 9.578544061302683e-06,
+      "loss": 1.0188,
+      "step": 1250
+    },
+    {
+      "epoch": 4.016551724137931,
+      "grad_norm": 0.5644910335540771,
+      "learning_rate": 9.655172413793105e-06,
+      "loss": 1.995,
+      "step": 1260
+    },
+    {
+      "epoch": 4.017318007662835,
+      "grad_norm": 0.18410761654376984,
+      "learning_rate": 9.731800766283525e-06,
+      "loss": 0.4526,
+      "step": 1270
+    },
+    {
+      "epoch": 4.01808429118774,
+      "grad_norm": 0.1986103355884552,
+      "learning_rate": 9.808429118773947e-06,
+      "loss": 1.589,
+      "step": 1280
+    },
+    {
+      "epoch": 4.018850574712643,
+      "grad_norm": 0.7268558740615845,
+      "learning_rate": 9.885057471264368e-06,
+      "loss": 1.6857,
+      "step": 1290
+    },
+    {
+      "epoch": 4.019616858237548,
+      "grad_norm": 23.243221282958984,
+      "learning_rate": 9.96168582375479e-06,
+      "loss": 1.7586,
+      "step": 1300
+    },
+    {
+      "epoch": 4.02,
+      "eval_accuracy": 0.4666666666666667,
+      "eval_loss": 1.1412569284439087,
+      "eval_runtime": 16.4314,
+      "eval_samples_per_second": 2.739,
+      "eval_steps_per_second": 2.739,
+      "step": 1305
+    },
+    {
+      "epoch": 5.000383141762452,
+      "grad_norm": 1.1051661968231201,
+      "learning_rate": 9.995742869306088e-06,
+      "loss": 0.8011,
+      "step": 1310
+    },
+    {
+      "epoch": 5.001149425287356,
+      "grad_norm": 0.23424309492111206,
+      "learning_rate": 9.987228607918263e-06,
+      "loss": 1.2352,
+      "step": 1320
+    },
+    {
+      "epoch": 5.001915708812261,
+      "grad_norm": 0.19838300347328186,
+      "learning_rate": 9.97871434653044e-06,
+      "loss": 0.493,
+      "step": 1330
+    },
+    {
+      "epoch": 5.002681992337164,
+      "grad_norm": 0.2764905095100403,
+      "learning_rate": 9.970200085142615e-06,
+      "loss": 1.4016,
+      "step": 1340
+    },
+    {
+      "epoch": 5.003448275862069,
+      "grad_norm": 19.335851669311523,
+      "learning_rate": 9.96168582375479e-06,
+      "loss": 2.0891,
+      "step": 1350
+    },
+    {
+      "epoch": 5.004214559386973,
+      "grad_norm": 17.96892547607422,
+      "learning_rate": 9.953171562366965e-06,
+      "loss": 1.3015,
+      "step": 1360
+    },
+    {
+      "epoch": 5.004980842911878,
+      "grad_norm": 0.40493690967559814,
+      "learning_rate": 9.944657300979142e-06,
+      "loss": 0.0353,
+      "step": 1370
+    },
+    {
+      "epoch": 5.005747126436781,
+      "grad_norm": 0.46046707034111023,
+      "learning_rate": 9.936143039591317e-06,
+      "loss": 1.6454,
+      "step": 1380
+    },
+    {
+      "epoch": 5.006513409961686,
+      "grad_norm": 19.26921844482422,
+      "learning_rate": 9.927628778203492e-06,
+      "loss": 2.6308,
+      "step": 1390
+    },
+    {
+      "epoch": 5.00727969348659,
+      "grad_norm": 0.6409273147583008,
+      "learning_rate": 9.919114516815667e-06,
+      "loss": 0.3655,
+      "step": 1400
+    },
+    {
+      "epoch": 5.008045977011494,
+      "grad_norm": 20.259307861328125,
+      "learning_rate": 9.910600255427842e-06,
+      "loss": 1.8601,
+      "step": 1410
+    },
+    {
+      "epoch": 5.0088122605363985,
+      "grad_norm": 0.2982420325279236,
+      "learning_rate": 9.902085994040018e-06,
+      "loss": 0.4097,
+      "step": 1420
+    },
+    {
+      "epoch": 5.009578544061303,
+      "grad_norm": 21.23941993713379,
+      "learning_rate": 9.893571732652193e-06,
+      "loss": 1.3039,
+      "step": 1430
+    },
+    {
+      "epoch": 5.010344827586207,
+      "grad_norm": 0.37259936332702637,
+      "learning_rate": 9.885057471264368e-06,
+      "loss": 1.6146,
+      "step": 1440
+    },
+    {
+      "epoch": 5.011111111111111,
+      "grad_norm": 0.6786749958992004,
+      "learning_rate": 9.876543209876543e-06,
+      "loss": 1.3758,
+      "step": 1450
+    },
+    {
+      "epoch": 5.011877394636016,
+      "grad_norm": 0.5649729371070862,
+      "learning_rate": 9.86802894848872e-06,
+      "loss": 0.7113,
+      "step": 1460
+    },
+    {
+      "epoch": 5.012643678160919,
+      "grad_norm": 19.882408142089844,
+      "learning_rate": 9.859514687100895e-06,
+      "loss": 1.191,
+      "step": 1470
+    },
+    {
+      "epoch": 5.013409961685824,
+      "grad_norm": 0.5207841396331787,
+      "learning_rate": 9.85100042571307e-06,
+      "loss": 1.6009,
+      "step": 1480
+    },
+    {
+      "epoch": 5.014176245210728,
+      "grad_norm": 0.5753087997436523,
+      "learning_rate": 9.842486164325245e-06,
+      "loss": 1.1187,
+      "step": 1490
+    },
+    {
+      "epoch": 5.014942528735633,
+      "grad_norm": 0.377913236618042,
+      "learning_rate": 9.833971902937422e-06,
+      "loss": 1.1263,
+      "step": 1500
+    },
+    {
+      "epoch": 5.015708812260536,
+      "grad_norm": 16.884944915771484,
+      "learning_rate": 9.825457641549597e-06,
+      "loss": 1.4622,
+      "step": 1510
+    },
+    {
+      "epoch": 5.016475095785441,
+      "grad_norm": 0.6560386419296265,
+      "learning_rate": 9.816943380161772e-06,
+      "loss": 0.9987,
+      "step": 1520
+    },
+    {
+      "epoch": 5.017241379310345,
+      "grad_norm": 18.208162307739258,
+      "learning_rate": 9.808429118773947e-06,
+      "loss": 2.1104,
+      "step": 1530
+    },
+    {
+      "epoch": 5.018007662835249,
+      "grad_norm": 18.947525024414062,
+      "learning_rate": 9.799914857386122e-06,
+      "loss": 1.8238,
+      "step": 1540
+    },
+    {
+      "epoch": 5.018773946360153,
+      "grad_norm": 1.023240566253662,
+      "learning_rate": 9.791400595998298e-06,
+      "loss": 0.8421,
+      "step": 1550
+    },
+    {
+      "epoch": 5.019540229885058,
+      "grad_norm": 25.64605140686035,
+      "learning_rate": 9.782886334610473e-06,
+      "loss": 1.0731,
+      "step": 1560
+    },
+    {
+      "epoch": 5.02,
+      "eval_accuracy": 0.4666666666666667,
+      "eval_loss": 1.8861749172210693,
+      "eval_runtime": 16.269,
+      "eval_samples_per_second": 2.766,
+      "eval_steps_per_second": 2.766,
+      "step": 1566
+    },
+    {
+      "epoch": 6.000306513409962,
+      "grad_norm": 18.987686157226562,
+      "learning_rate": 9.774372073222648e-06,
+      "loss": 1.5048,
+      "step": 1570
+    },
+    {
+      "epoch": 6.001072796934866,
+      "grad_norm": 19.56184196472168,
+      "learning_rate": 9.765857811834825e-06,
+      "loss": 1.072,
+      "step": 1580
+    },
+    {
+      "epoch": 6.00183908045977,
+      "grad_norm": 0.44697844982147217,
+      "learning_rate": 9.757343550447e-06,
+      "loss": 0.7249,
+      "step": 1590
+    },
+    {
+      "epoch": 6.002605363984674,
+      "grad_norm": 0.06227334588766098,
+      "learning_rate": 9.748829289059175e-06,
+      "loss": 1.42,
+      "step": 1600
+    },
+    {
+      "epoch": 6.003371647509579,
+      "grad_norm": 0.5351743102073669,
+      "learning_rate": 9.74031502767135e-06,
+      "loss": 1.7496,
+      "step": 1610
+    },
+    {
+      "epoch": 6.0041379310344825,
+      "grad_norm": 0.6440963745117188,
+      "learning_rate": 9.731800766283525e-06,
+      "loss": 1.095,
+      "step": 1620
+    },
+    {
+      "epoch": 6.004904214559387,
+      "grad_norm": 28.36171531677246,
+      "learning_rate": 9.723286504895702e-06,
+      "loss": 1.7879,
+      "step": 1630
+    },
+    {
+      "epoch": 6.005670498084291,
+      "grad_norm": 0.16215454041957855,
+      "learning_rate": 9.714772243507877e-06,
+      "loss": 0.0226,
+      "step": 1640
+    },
+    {
+      "epoch": 6.006436781609195,
+      "grad_norm": 34.247615814208984,
+      "learning_rate": 9.706257982120052e-06,
+      "loss": 1.473,
+      "step": 1650
+    },
+    {
+      "epoch": 6.0072030651340995,
+      "grad_norm": 28.649755477905273,
+      "learning_rate": 9.697743720732228e-06,
+      "loss": 0.919,
+      "step": 1660
+    },
+    {
+      "epoch": 6.007969348659004,
+      "grad_norm": 18.813629150390625,
+      "learning_rate": 9.689229459344403e-06,
+      "loss": 2.428,
+      "step": 1670
+    },
+    {
+      "epoch": 6.008735632183908,
+      "grad_norm": 1.3150634765625,
+      "learning_rate": 9.680715197956578e-06,
+      "loss": 0.8913,
+      "step": 1680
+    },
+    {
+      "epoch": 6.009501915708812,
+      "grad_norm": 0.3356521427631378,
+      "learning_rate": 9.672200936568753e-06,
+      "loss": 0.9907,
+      "step": 1690
+    },
+    {
+      "epoch": 6.010268199233717,
+      "grad_norm": 17.486032485961914,
+      "learning_rate": 9.663686675180928e-06,
+      "loss": 1.6711,
+      "step": 1700
+    },
+    {
+      "epoch": 6.011034482758621,
+      "grad_norm": 1.6453771591186523,
+      "learning_rate": 9.655172413793105e-06,
+      "loss": 1.6673,
+      "step": 1710
+    },
+    {
+      "epoch": 6.011800766283525,
+      "grad_norm": 35.23017883300781,
+      "learning_rate": 9.64665815240528e-06,
+      "loss": 1.1546,
+      "step": 1720
+    },
+    {
+      "epoch": 6.012567049808429,
+      "grad_norm": 1.0530450344085693,
+      "learning_rate": 9.638143891017455e-06,
+      "loss": 0.0673,
+      "step": 1730
+    },
+    {
+      "epoch": 6.013333333333334,
+      "grad_norm": 44.10451126098633,
+      "learning_rate": 9.62962962962963e-06,
+      "loss": 1.704,
+      "step": 1740
+    },
+    {
+      "epoch": 6.014099616858237,
+      "grad_norm": 24.64010238647461,
+      "learning_rate": 9.621115368241805e-06,
+      "loss": 1.9377,
+      "step": 1750
+    },
+    {
+      "epoch": 6.014865900383142,
+      "grad_norm": 0.5813682675361633,
+      "learning_rate": 9.612601106853982e-06,
+      "loss": 0.7754,
+      "step": 1760
+    },
+    {
+      "epoch": 6.015632183908046,
+      "grad_norm": 0.3799228072166443,
+      "learning_rate": 9.604086845466157e-06,
+      "loss": 0.3575,
+      "step": 1770
+    },
+    {
+      "epoch": 6.01639846743295,
+      "grad_norm": 58.88032913208008,
+      "learning_rate": 9.595572584078332e-06,
+      "loss": 1.7625,
+      "step": 1780
+    },
+    {
+      "epoch": 6.017164750957854,
+      "grad_norm": 53.162986755371094,
+      "learning_rate": 9.587058322690508e-06,
+      "loss": 2.6159,
+      "step": 1790
+    },
+    {
+      "epoch": 6.017931034482759,
+      "grad_norm": 17.41657829284668,
+      "learning_rate": 9.578544061302683e-06,
+      "loss": 1.7754,
+      "step": 1800
+    },
+    {
+      "epoch": 6.018697318007663,
+      "grad_norm": 1.2673225402832031,
+      "learning_rate": 9.570029799914858e-06,
+      "loss": 0.739,
+      "step": 1810
+    },
+    {
+      "epoch": 6.019463601532567,
+      "grad_norm": 30.341552734375,
+      "learning_rate": 9.561515538527033e-06,
+      "loss": 1.3539,
+      "step": 1820
+    },
+    {
+      "epoch": 6.02,
+      "eval_accuracy": 0.4666666666666667,
+      "eval_loss": 1.82771635055542,
+      "eval_runtime": 16.5146,
+      "eval_samples_per_second": 2.725,
+      "eval_steps_per_second": 2.725,
+      "step": 1827
+    },
+    {
+      "epoch": 7.000229885057471,
+      "grad_norm": 40.3336067199707,
+      "learning_rate": 9.553001277139208e-06,
+      "loss": 0.7449,
+      "step": 1830
+    },
+    {
+      "epoch": 7.000996168582375,
+      "grad_norm": 47.77488327026367,
+      "learning_rate": 9.544487015751385e-06,
+      "loss": 1.6668,
+      "step": 1840
+    },
+    {
+      "epoch": 7.00176245210728,
+      "grad_norm": 0.17062117159366608,
+      "learning_rate": 9.53597275436356e-06,
+      "loss": 0.6863,
+      "step": 1850
+    },
+    {
+      "epoch": 7.0025287356321835,
+      "grad_norm": 52.891422271728516,
+      "learning_rate": 9.527458492975735e-06,
+      "loss": 2.1523,
+      "step": 1860
+    },
+    {
+      "epoch": 7.003295019157088,
+      "grad_norm": 1.2075984477996826,
+      "learning_rate": 9.518944231587912e-06,
+      "loss": 0.7233,
+      "step": 1870
+    },
+    {
+      "epoch": 7.0040613026819925,
+      "grad_norm": 51.95742416381836,
+      "learning_rate": 9.510429970200085e-06,
+      "loss": 1.3763,
+      "step": 1880
+    },
+    {
+      "epoch": 7.004827586206897,
+      "grad_norm": 49.27976608276367,
+      "learning_rate": 9.501915708812262e-06,
+      "loss": 1.3568,
+      "step": 1890
+    },
+    {
+      "epoch": 7.0055938697318005,
+      "grad_norm": 14.621587753295898,
+      "learning_rate": 9.493401447424437e-06,
+      "loss": 1.8932,
+      "step": 1900
+    },
+    {
+      "epoch": 7.006360153256705,
+      "grad_norm": 2.2778844833374023,
+      "learning_rate": 9.484887186036612e-06,
+      "loss": 0.7018,
+      "step": 1910
+    },
+    {
+      "epoch": 7.0071264367816095,
+      "grad_norm": 4.356560230255127,
+      "learning_rate": 9.476372924648788e-06,
+      "loss": 1.4481,
+      "step": 1920
+    },
+    {
+      "epoch": 7.007892720306513,
+      "grad_norm": 0.10934069752693176,
+      "learning_rate": 9.467858663260963e-06,
+      "loss": 0.7275,
+      "step": 1930
+    },
+    {
+      "epoch": 7.008659003831418,
+      "grad_norm": 0.4002261459827423,
+      "learning_rate": 9.459344401873138e-06,
+      "loss": 1.648,
+      "step": 1940
+    },
+    {
+      "epoch": 7.009425287356322,
+      "grad_norm": 57.19015121459961,
+      "learning_rate": 9.450830140485315e-06,
+      "loss": 1.4453,
+      "step": 1950
+    },
+    {
+      "epoch": 7.010191570881226,
+      "grad_norm": 45.90049362182617,
+      "learning_rate": 9.442315879097488e-06,
+      "loss": 1.2843,
+      "step": 1960
+    },
+    {
+      "epoch": 7.01095785440613,
+      "grad_norm": 61.82890701293945,
+      "learning_rate": 9.433801617709665e-06,
+      "loss": 1.7412,
+      "step": 1970
+    },
+    {
+      "epoch": 7.011724137931035,
+      "grad_norm": 0.06092312932014465,
+      "learning_rate": 9.42528735632184e-06,
+      "loss": 0.2979,
+      "step": 1980
+    },
+    {
+      "epoch": 7.012490421455938,
+      "grad_norm": 46.58448028564453,
+      "learning_rate": 9.416773094934015e-06,
+      "loss": 1.9555,
+      "step": 1990
+    },
+    {
+      "epoch": 7.013256704980843,
+      "grad_norm": 0.8877707719802856,
+      "learning_rate": 9.408258833546192e-06,
+      "loss": 0.6078,
+      "step": 2000
+    },
+    {
+      "epoch": 7.014022988505747,
+      "grad_norm": 1.0596059560775757,
+      "learning_rate": 9.399744572158365e-06,
+      "loss": 0.6166,
+      "step": 2010
+    },
+    {
+      "epoch": 7.014789272030652,
+      "grad_norm": 93.8338623046875,
+      "learning_rate": 9.391230310770542e-06,
+      "loss": 1.2646,
+      "step": 2020
+    },
+    {
+      "epoch": 7.015555555555555,
+      "grad_norm": 0.6872724294662476,
+      "learning_rate": 9.382716049382717e-06,
+      "loss": 0.5737,
+      "step": 2030
+    },
+    {
+      "epoch": 7.01632183908046,
+      "grad_norm": 13.280865669250488,
+      "learning_rate": 9.374201787994892e-06,
+      "loss": 0.1844,
+      "step": 2040
+    },
+    {
+      "epoch": 7.017088122605364,
+      "grad_norm": 0.06641419976949692,
+      "learning_rate": 9.365687526607068e-06,
+      "loss": 1.1597,
+      "step": 2050
+    },
+    {
+      "epoch": 7.017854406130268,
+      "grad_norm": 0.3752352297306061,
+      "learning_rate": 9.357173265219243e-06,
+      "loss": 3.2409,
+      "step": 2060
+    },
+    {
+      "epoch": 7.018620689655172,
+      "grad_norm": 8.496101379394531,
+      "learning_rate": 9.348659003831418e-06,
+      "loss": 1.3105,
+      "step": 2070
+    },
+    {
+      "epoch": 7.019386973180077,
+      "grad_norm": 57.28456497192383,
+      "learning_rate": 9.340144742443595e-06,
+      "loss": 0.8906,
+      "step": 2080
+    },
+    {
+      "epoch": 7.02,
+      "eval_accuracy": 0.4666666666666667,
+      "eval_loss": 1.7558737993240356,
+      "eval_runtime": 15.2334,
+      "eval_samples_per_second": 2.954,
+      "eval_steps_per_second": 2.954,
+      "step": 2088
+    },
+    {
+      "epoch": 8.00015325670498,
+      "grad_norm": 2.2176547050476074,
+      "learning_rate": 9.331630481055768e-06,
+      "loss": 0.786,
+      "step": 2090
+    },
+    {
+      "epoch": 8.000919540229885,
+      "grad_norm": 146.79502868652344,
+      "learning_rate": 9.323116219667945e-06,
+      "loss": 0.6585,
+      "step": 2100
+    },
+    {
+      "epoch": 8.001685823754789,
+      "grad_norm": 0.0524924211204052,
+      "learning_rate": 9.31460195828012e-06,
+      "loss": 1.4611,
+      "step": 2110
+    },
+    {
+      "epoch": 8.002452107279694,
+      "grad_norm": 0.035575300455093384,
+      "learning_rate": 9.306087696892295e-06,
+      "loss": 1.2997,
+      "step": 2120
+    },
+    {
+      "epoch": 8.003218390804598,
+      "grad_norm": 0.5172576904296875,
+      "learning_rate": 9.297573435504472e-06,
+      "loss": 2.0572,
+      "step": 2130
+    },
+    {
+      "epoch": 8.003984674329502,
+      "grad_norm": 6.22727108001709,
+      "learning_rate": 9.289059174116647e-06,
+      "loss": 0.0919,
+      "step": 2140
+    },
+    {
+      "epoch": 8.004750957854407,
+      "grad_norm": 0.8947464227676392,
+      "learning_rate": 9.280544912728822e-06,
+      "loss": 0.3653,
+      "step": 2150
+    },
+    {
+      "epoch": 8.00551724137931,
+      "grad_norm": 79.02603912353516,
+      "learning_rate": 9.272030651340997e-06,
+      "loss": 1.5212,
+      "step": 2160
+    },
+    {
+      "epoch": 8.006283524904214,
+      "grad_norm": 28.48909568786621,
+      "learning_rate": 9.263516389953172e-06,
+      "loss": 3.6352,
+      "step": 2170
+    },
+    {
+      "epoch": 8.00704980842912,
+      "grad_norm": 23.416492462158203,
+      "learning_rate": 9.255002128565348e-06,
+      "loss": 1.8813,
+      "step": 2180
+    },
+    {
+      "epoch": 8.007816091954023,
+      "grad_norm": 10.679008483886719,
+      "learning_rate": 9.246487867177523e-06,
+      "loss": 0.434,
+      "step": 2190
+    },
+    {
+      "epoch": 8.008582375478927,
+      "grad_norm": 105.9638900756836,
+      "learning_rate": 9.237973605789698e-06,
+      "loss": 0.5633,
+      "step": 2200
+    },
+    {
+      "epoch": 8.009348659003832,
+      "grad_norm": 0.22639591991901398,
+      "learning_rate": 9.229459344401875e-06,
+      "loss": 2.3839,
+      "step": 2210
+    },
+    {
+      "epoch": 8.010114942528736,
+      "grad_norm": 0.23017604649066925,
+      "learning_rate": 9.220945083014048e-06,
+      "loss": 0.7008,
+      "step": 2220
+    },
+    {
+      "epoch": 8.01088122605364,
+      "grad_norm": 127.92184448242188,
+      "learning_rate": 9.212430821626225e-06,
+      "loss": 0.949,
+      "step": 2230
+    },
+    {
+      "epoch": 8.011647509578545,
+      "grad_norm": 61.606781005859375,
+      "learning_rate": 9.2039165602384e-06,
+      "loss": 1.359,
+      "step": 2240
+    },
+    {
+      "epoch": 8.012413793103448,
+      "grad_norm": 22.344465255737305,
+      "learning_rate": 9.195402298850575e-06,
+      "loss": 0.2749,
+      "step": 2250
+    },
+    {
+      "epoch": 8.013180076628352,
+      "grad_norm": 268.2001647949219,
+      "learning_rate": 9.186888037462752e-06,
+      "loss": 1.2585,
+      "step": 2260
+    },
+    {
+      "epoch": 8.013946360153257,
+      "grad_norm": 60.801551818847656,
+      "learning_rate": 9.178373776074927e-06,
+      "loss": 1.553,
+      "step": 2270
+    },
+    {
+      "epoch": 8.01471264367816,
+      "grad_norm": 2.967214822769165,
+      "learning_rate": 9.169859514687102e-06,
+      "loss": 0.6537,
+      "step": 2280
+    },
+    {
+      "epoch": 8.015478927203064,
+      "grad_norm": 83.84700012207031,
+      "learning_rate": 9.161345253299277e-06,
+      "loss": 1.1928,
+      "step": 2290
+    },
+    {
+      "epoch": 8.01624521072797,
+      "grad_norm": 0.36373046040534973,
+      "learning_rate": 9.152830991911452e-06,
+      "loss": 2.2204,
+      "step": 2300
+    },
+    {
+      "epoch": 8.017011494252873,
+      "grad_norm": 0.08310884982347488,
+      "learning_rate": 9.144316730523628e-06,
+      "loss": 0.9423,
+      "step": 2310
+    },
+    {
+      "epoch": 8.017777777777777,
+      "grad_norm": 126.53848266601562,
+      "learning_rate": 9.135802469135803e-06,
+      "loss": 1.4315,
+      "step": 2320
+    },
+    {
+      "epoch": 8.018544061302682,
+      "grad_norm": 40.52521514892578,
+      "learning_rate": 9.127288207747978e-06,
+      "loss": 0.4683,
+      "step": 2330
+    },
+    {
+      "epoch": 8.019310344827586,
+      "grad_norm": 56.70500564575195,
+      "learning_rate": 9.118773946360155e-06,
+      "loss": 1.5706,
+      "step": 2340
+    },
+    {
+      "epoch": 8.02,
+      "eval_accuracy": 0.4666666666666667,
+      "eval_loss": 2.0013673305511475,
+      "eval_runtime": 15.4307,
+      "eval_samples_per_second": 2.916,
+      "eval_steps_per_second": 2.916,
+      "step": 2349
+    },
+    {
+      "epoch": 9.00007662835249,
+      "grad_norm": 0.04681820422410965,
+      "learning_rate": 9.110259684972328e-06,
+      "loss": 0.6411,
+      "step": 2350
+    },
+    {
+      "epoch": 9.000842911877395,
+      "grad_norm": 79.75486755371094,
+      "learning_rate": 9.101745423584505e-06,
+      "loss": 0.5805,
+      "step": 2360
+    },
+    {
+      "epoch": 9.001609195402299,
+      "grad_norm": 59.682044982910156,
+      "learning_rate": 9.09323116219668e-06,
+      "loss": 1.9657,
+      "step": 2370
+    },
+    {
+      "epoch": 9.002375478927203,
+      "grad_norm": 25.446149826049805,
+      "learning_rate": 9.084716900808855e-06,
+      "loss": 0.1048,
+      "step": 2380
+    },
+    {
+      "epoch": 9.003141762452108,
+      "grad_norm": 0.19314312934875488,
+      "learning_rate": 9.076202639421032e-06,
+      "loss": 1.7946,
+      "step": 2390
+    },
+    {
+      "epoch": 9.003908045977012,
+      "grad_norm": 46.27597427368164,
+      "learning_rate": 9.067688378033207e-06,
+      "loss": 2.5579,
+      "step": 2400
+    },
+    {
+      "epoch": 9.004674329501915,
+      "grad_norm": 51.210208892822266,
+      "learning_rate": 9.059174116645382e-06,
+      "loss": 0.7125,
+      "step": 2410
+    },
+    {
+      "epoch": 9.00544061302682,
+      "grad_norm": 10.87521743774414,
+      "learning_rate": 9.050659855257558e-06,
+      "loss": 0.6746,
+      "step": 2420
+    },
+    {
+      "epoch": 9.006206896551724,
+      "grad_norm": 13.82815933227539,
+      "learning_rate": 9.042145593869732e-06,
+      "loss": 1.0982,
+      "step": 2430
+    },
+    {
+      "epoch": 9.006973180076628,
+      "grad_norm": 34.410831451416016,
+      "learning_rate": 9.033631332481908e-06,
+      "loss": 1.1018,
+      "step": 2440
+    },
+    {
+      "epoch": 9.007739463601533,
+      "grad_norm": 0.4074358642101288,
+      "learning_rate": 9.025117071094083e-06,
+      "loss": 1.4791,
+      "step": 2450
+    },
+    {
+      "epoch": 9.008505747126437,
+      "grad_norm": 34.35459518432617,
+      "learning_rate": 9.016602809706258e-06,
+      "loss": 2.391,
+      "step": 2460
+    },
+    {
+      "epoch": 9.00927203065134,
+      "grad_norm": 9.86334228515625,
+      "learning_rate": 9.008088548318435e-06,
+      "loss": 0.5826,
+      "step": 2470
+    },
+    {
+      "epoch": 9.010038314176246,
+      "grad_norm": 0.3567091226577759,
+      "learning_rate": 8.999574286930608e-06,
+      "loss": 0.2535,
+      "step": 2480
+    },
+    {
+      "epoch": 9.01080459770115,
+      "grad_norm": 0.1000981479883194,
+      "learning_rate": 8.991060025542785e-06,
+      "loss": 0.7387,
+      "step": 2490
+    },
+    {
+      "epoch": 9.011570881226053,
+      "grad_norm": 0.6302473545074463,
+      "learning_rate": 8.98254576415496e-06,
+      "loss": 0.8229,
+      "step": 2500
+    },
+    {
+      "epoch": 9.012337164750958,
+      "grad_norm": 0.02157972939312458,
+      "learning_rate": 8.974031502767135e-06,
+      "loss": 0.9654,
+      "step": 2510
+    },
+    {
+      "epoch": 9.013103448275862,
+      "grad_norm": 0.21129199862480164,
+      "learning_rate": 8.965517241379312e-06,
+      "loss": 2.5648,
+      "step": 2520
+    },
+    {
+      "epoch": 9.013869731800765,
+      "grad_norm": 0.05354182422161102,
+      "learning_rate": 8.957002979991487e-06,
+      "loss": 0.1968,
+      "step": 2530
+    },
+    {
+      "epoch": 9.01463601532567,
+      "grad_norm": 7.898284435272217,
+      "learning_rate": 8.948488718603662e-06,
+      "loss": 1.5846,
+      "step": 2540
+    },
+    {
+      "epoch": 9.015402298850574,
+      "grad_norm": 0.8477972745895386,
+      "learning_rate": 8.939974457215838e-06,
+      "loss": 0.6042,
+      "step": 2550
+    },
+    {
+      "epoch": 9.01616858237548,
+      "grad_norm": 2.289872646331787,
+      "learning_rate": 8.931460195828012e-06,
+      "loss": 0.2621,
+      "step": 2560
+    },
+    {
+      "epoch": 9.016934865900383,
+      "grad_norm": 127.58092498779297,
+      "learning_rate": 8.922945934440188e-06,
+      "loss": 0.1349,
+      "step": 2570
+    },
+    {
+      "epoch": 9.017701149425287,
+      "grad_norm": 0.35964861512184143,
+      "learning_rate": 8.914431673052363e-06,
+      "loss": 0.2656,
+      "step": 2580
+    },
+    {
+      "epoch": 9.018467432950192,
+      "grad_norm": 97.85234069824219,
+      "learning_rate": 8.905917411664538e-06,
+      "loss": 1.4753,
+      "step": 2590
+    },
+    {
+      "epoch": 9.019233716475096,
+      "grad_norm": 0.021428819745779037,
+      "learning_rate": 8.897403150276715e-06,
+      "loss": 1.069,
+      "step": 2600
+    },
+    {
+      "epoch": 9.02,
+      "grad_norm": 0.36919301748275757,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 0.3113,
+      "step": 2610
+    },
+    {
+      "epoch": 9.02,
+      "eval_accuracy": 0.6666666666666666,
+      "eval_loss": 1.3057535886764526,
+      "eval_runtime": 16.4734,
+      "eval_samples_per_second": 2.732,
+      "eval_steps_per_second": 2.732,
+      "step": 2610
+    },
+    {
+      "epoch": 10.000766283524904,
+      "grad_norm": 0.06585513800382614,
+      "learning_rate": 8.880374627501065e-06,
+      "loss": 1.5534,
+      "step": 2620
+    },
+    {
+      "epoch": 10.001532567049809,
+      "grad_norm": 1.7465332746505737,
+      "learning_rate": 8.87186036611324e-06,
+      "loss": 1.0735,
+      "step": 2630
+    },
+    {
+      "epoch": 10.002298850574713,
+      "grad_norm": 0.041456516832113266,
+      "learning_rate": 8.863346104725415e-06,
+      "loss": 0.2433,
+      "step": 2640
+    },
+    {
+      "epoch": 10.003065134099616,
+      "grad_norm": 118.21949005126953,
+      "learning_rate": 8.854831843337592e-06,
+      "loss": 0.9212,
+      "step": 2650
+    },
+    {
+      "epoch": 10.003831417624522,
+      "grad_norm": 6.035428524017334,
+      "learning_rate": 8.846317581949767e-06,
+      "loss": 0.9281,
+      "step": 2660
+    },
+    {
+      "epoch": 10.004597701149425,
+      "grad_norm": 0.02484976500272751,
+      "learning_rate": 8.837803320561942e-06,
+      "loss": 1.6067,
+      "step": 2670
+    },
+    {
+      "epoch": 10.005363984674329,
+      "grad_norm": 4.0249433517456055,
+      "learning_rate": 8.829289059174118e-06,
+      "loss": 1.5351,
+      "step": 2680
+    },
+    {
+      "epoch": 10.006130268199234,
+      "grad_norm": 0.6469050049781799,
+      "learning_rate": 8.820774797786292e-06,
+      "loss": 0.0113,
+      "step": 2690
+    },
+    {
+      "epoch": 10.006896551724138,
+      "grad_norm": 0.02827748842537403,
+      "learning_rate": 8.812260536398468e-06,
+      "loss": 0.9929,
+      "step": 2700
+    },
+    {
+      "epoch": 10.007662835249041,
+      "grad_norm": 84.25238037109375,
+      "learning_rate": 8.803746275010643e-06,
+      "loss": 2.1007,
+      "step": 2710
+    },
+    {
+      "epoch": 10.008429118773947,
+      "grad_norm": 1244.77197265625,
+      "learning_rate": 8.795232013622818e-06,
+      "loss": 1.3255,
+      "step": 2720
+    },
+    {
+      "epoch": 10.00919540229885,
+      "grad_norm": 0.021987877786159515,
+      "learning_rate": 8.786717752234995e-06,
+      "loss": 1.5886,
+      "step": 2730
+    },
+    {
+      "epoch": 10.009961685823756,
+      "grad_norm": 0.67450350522995,
+      "learning_rate": 8.77820349084717e-06,
+      "loss": 1.1008,
+      "step": 2740
+    },
+    {
+      "epoch": 10.01072796934866,
+      "grad_norm": 9.497635841369629,
+      "learning_rate": 8.769689229459345e-06,
+      "loss": 0.5836,
+      "step": 2750
+    },
+    {
+      "epoch": 10.011494252873563,
+      "grad_norm": 0.020129255950450897,
+      "learning_rate": 8.76117496807152e-06,
+      "loss": 1.2045,
+      "step": 2760
+    },
+    {
+      "epoch": 10.012260536398468,
+      "grad_norm": 0.017797917127609253,
+      "learning_rate": 8.752660706683695e-06,
+      "loss": 0.8505,
+      "step": 2770
+    },
+    {
+      "epoch": 10.013026819923372,
+      "grad_norm": 0.5655618906021118,
+      "learning_rate": 8.744146445295872e-06,
+      "loss": 0.9844,
+      "step": 2780
+    },
+    {
+      "epoch": 10.013793103448275,
+      "grad_norm": 0.10610664635896683,
+      "learning_rate": 8.735632183908047e-06,
+      "loss": 0.4595,
+      "step": 2790
+    },
+    {
+      "epoch": 10.01455938697318,
+      "grad_norm": 66.6644515991211,
+      "learning_rate": 8.727117922520222e-06,
+      "loss": 1.4337,
+      "step": 2800
+    },
+    {
+      "epoch": 10.015325670498084,
+      "grad_norm": 196.35189819335938,
+      "learning_rate": 8.718603661132398e-06,
+      "loss": 1.218,
+      "step": 2810
+    },
+    {
+      "epoch": 10.016091954022988,
+      "grad_norm": 328.2312316894531,
+      "learning_rate": 8.710089399744572e-06,
+      "loss": 0.795,
+      "step": 2820
+    },
+    {
+      "epoch": 10.016858237547893,
+      "grad_norm": 2.4351890087127686,
+      "learning_rate": 8.701575138356748e-06,
+      "loss": 0.5883,
+      "step": 2830
+    },
+    {
+      "epoch": 10.017624521072797,
+      "grad_norm": 0.07693709433078766,
+      "learning_rate": 8.693060876968923e-06,
+      "loss": 0.3392,
+      "step": 2840
+    },
+    {
+      "epoch": 10.0183908045977,
+      "grad_norm": 0.016635416075587273,
+      "learning_rate": 8.684546615581098e-06,
+      "loss": 0.7708,
+      "step": 2850
+    },
+    {
+      "epoch": 10.019157088122606,
+      "grad_norm": 1.0986331701278687,
+      "learning_rate": 8.676032354193275e-06,
+      "loss": 0.6342,
+      "step": 2860
+    },
+    {
+      "epoch": 10.01992337164751,
+      "grad_norm": 105.34931182861328,
+      "learning_rate": 8.66751809280545e-06,
+      "loss": 1.6658,
+      "step": 2870
+    },
+    {
+      "epoch": 10.02,
+      "eval_accuracy": 0.6444444444444445,
+      "eval_loss": 1.5835341215133667,
+      "eval_runtime": 16.0892,
+      "eval_samples_per_second": 2.797,
+      "eval_steps_per_second": 2.797,
+      "step": 2871
+    },
+    {
+      "epoch": 11.000689655172414,
+      "grad_norm": 0.042755126953125,
+      "learning_rate": 8.659003831417625e-06,
+      "loss": 1.2281,
+      "step": 2880
+    },
+    {
+      "epoch": 11.001455938697317,
+      "grad_norm": 0.13419094681739807,
+      "learning_rate": 8.650489570029802e-06,
+      "loss": 0.643,
+      "step": 2890
+    },
+    {
+      "epoch": 11.002222222222223,
+      "grad_norm": 0.5722396969795227,
+      "learning_rate": 8.641975308641975e-06,
+      "loss": 1.1059,
+      "step": 2900
+    },
+    {
+      "epoch": 11.002988505747126,
+      "grad_norm": 512.3408203125,
+      "learning_rate": 8.633461047254152e-06,
+      "loss": 0.5732,
+      "step": 2910
+    },
+    {
+      "epoch": 11.00375478927203,
+      "grad_norm": 0.011800810694694519,
+      "learning_rate": 8.624946785866327e-06,
+      "loss": 1.0223,
+      "step": 2920
+    },
+    {
+      "epoch": 11.004521072796935,
+      "grad_norm": 8.030099868774414,
+      "learning_rate": 8.616432524478502e-06,
+      "loss": 0.3257,
+      "step": 2930
+    },
+    {
+      "epoch": 11.005287356321839,
+      "grad_norm": 45.093021392822266,
+      "learning_rate": 8.607918263090678e-06,
+      "loss": 1.7945,
+      "step": 2940
+    },
+    {
+      "epoch": 11.006053639846744,
+      "grad_norm": 0.22022829949855804,
+      "learning_rate": 8.599404001702853e-06,
+      "loss": 1.0616,
+      "step": 2950
+    },
+    {
+      "epoch": 11.006819923371648,
+      "grad_norm": 0.08585281670093536,
+      "learning_rate": 8.590889740315028e-06,
+      "loss": 1.644,
+      "step": 2960
+    },
+    {
+      "epoch": 11.007586206896551,
+      "grad_norm": 2.1208853721618652,
+      "learning_rate": 8.582375478927203e-06,
+      "loss": 0.6133,
+      "step": 2970
+    },
+    {
+      "epoch": 11.008352490421457,
+      "grad_norm": 91.65081024169922,
+      "learning_rate": 8.573861217539378e-06,
+      "loss": 1.0825,
+      "step": 2980
+    },
+    {
+      "epoch": 11.00911877394636,
+      "grad_norm": 3.9088544845581055,
+      "learning_rate": 8.565346956151555e-06,
+      "loss": 1.7217,
+      "step": 2990
+    },
+    {
+      "epoch": 11.009885057471264,
+      "grad_norm": 6.6397929191589355,
+      "learning_rate": 8.55683269476373e-06,
+      "loss": 1.8179,
+      "step": 3000
+    },
+    {
+      "epoch": 11.01065134099617,
+      "grad_norm": 11.316490173339844,
+      "learning_rate": 8.548318433375905e-06,
+      "loss": 0.0325,
+      "step": 3010
+    },
+    {
+      "epoch": 11.011417624521073,
+      "grad_norm": 0.28897494077682495,
+      "learning_rate": 8.539804171988082e-06,
+      "loss": 0.6603,
+      "step": 3020
+    },
+    {
+      "epoch": 11.012183908045976,
+      "grad_norm": 68.44873809814453,
+      "learning_rate": 8.531289910600255e-06,
+      "loss": 0.7484,
+      "step": 3030
+    },
+    {
+      "epoch": 11.012950191570882,
+      "grad_norm": 0.008221070282161236,
+      "learning_rate": 8.522775649212432e-06,
+      "loss": 0.0064,
+      "step": 3040
+    },
+    {
+      "epoch": 11.013716475095785,
+      "grad_norm": 189.57403564453125,
+      "learning_rate": 8.514261387824607e-06,
+      "loss": 1.7867,
+      "step": 3050
+    },
+    {
+      "epoch": 11.014482758620689,
+      "grad_norm": 0.04196823760867119,
+      "learning_rate": 8.505747126436782e-06,
+      "loss": 1.6422,
+      "step": 3060
+    },
+    {
+      "epoch": 11.015249042145594,
+      "grad_norm": 0.06809419393539429,
+      "learning_rate": 8.497232865048958e-06,
+      "loss": 0.0093,
+      "step": 3070
+    },
+    {
+      "epoch": 11.016015325670498,
+      "grad_norm": 0.6770594120025635,
+      "learning_rate": 8.488718603661133e-06,
+      "loss": 1.0171,
+      "step": 3080
+    },
+    {
+      "epoch": 11.016781609195402,
+      "grad_norm": 0.04048323631286621,
+      "learning_rate": 8.480204342273308e-06,
+      "loss": 2.187,
+      "step": 3090
+    },
+    {
+      "epoch": 11.017547892720307,
+      "grad_norm": 0.030397607013583183,
+      "learning_rate": 8.471690080885483e-06,
+      "loss": 0.5514,
+      "step": 3100
+    },
+    {
+      "epoch": 11.01831417624521,
+      "grad_norm": 32.990638732910156,
+      "learning_rate": 8.463175819497658e-06,
+      "loss": 1.2996,
+      "step": 3110
+    },
+    {
+      "epoch": 11.019080459770114,
+      "grad_norm": 0.03340402990579605,
+      "learning_rate": 8.454661558109835e-06,
+      "loss": 0.4193,
+      "step": 3120
+    },
+    {
+      "epoch": 11.01984674329502,
+      "grad_norm": 50.187522888183594,
+      "learning_rate": 8.44614729672201e-06,
+      "loss": 1.3587,
+      "step": 3130
+    },
+    {
+      "epoch": 11.02,
+      "eval_accuracy": 0.7555555555555555,
+      "eval_loss": 0.8309432864189148,
+      "eval_runtime": 16.5175,
+      "eval_samples_per_second": 2.724,
+      "eval_steps_per_second": 2.724,
+      "step": 3132
+    },
+    {
+      "epoch": 12.000613026819924,
+      "grad_norm": 0.5363060235977173,
+      "learning_rate": 8.437633035334185e-06,
+      "loss": 0.0229,
+      "step": 3140
+    },
+    {
+      "epoch": 12.001379310344827,
+      "grad_norm": 0.007902957499027252,
+      "learning_rate": 8.429118773946362e-06,
+      "loss": 1.1452,
+      "step": 3150
+    },
+    {
+      "epoch": 12.002145593869733,
+      "grad_norm": 232.3303680419922,
+      "learning_rate": 8.420604512558537e-06,
+      "loss": 1.4698,
+      "step": 3160
+    },
+    {
+      "epoch": 12.002911877394636,
+      "grad_norm": 2.575563669204712,
+      "learning_rate": 8.412090251170712e-06,
+      "loss": 1.0589,
+      "step": 3170
+    },
+    {
+      "epoch": 12.00367816091954,
+      "grad_norm": 28.641246795654297,
+      "learning_rate": 8.403575989782887e-06,
+      "loss": 1.9366,
+      "step": 3180
+    },
+    {
+      "epoch": 12.004444444444445,
+      "grad_norm": 31.25893211364746,
+      "learning_rate": 8.395061728395062e-06,
+      "loss": 0.5945,
+      "step": 3190
+    },
+    {
+      "epoch": 12.005210727969349,
+      "grad_norm": 1.1079418659210205,
+      "learning_rate": 8.386547467007238e-06,
+      "loss": 0.012,
+      "step": 3200
+    },
+    {
+      "epoch": 12.005977011494252,
+      "grad_norm": 0.012604065239429474,
+      "learning_rate": 8.378033205619413e-06,
+      "loss": 0.9649,
+      "step": 3210
+    },
+    {
+      "epoch": 12.006743295019158,
+      "grad_norm": 87.2633056640625,
+      "learning_rate": 8.369518944231588e-06,
+      "loss": 1.5178,
+      "step": 3220
+    },
+    {
+      "epoch": 12.007509578544061,
+      "grad_norm": 0.013699457980692387,
+      "learning_rate": 8.361004682843763e-06,
+      "loss": 1.3879,
+      "step": 3230
+    },
+    {
+      "epoch": 12.008275862068965,
+      "grad_norm": 52.45587158203125,
+      "learning_rate": 8.35249042145594e-06,
+      "loss": 1.9076,
+      "step": 3240
+    },
+    {
+      "epoch": 12.00904214559387,
+      "grad_norm": 41.09255599975586,
+      "learning_rate": 8.343976160068115e-06,
+      "loss": 1.284,
+      "step": 3250
+    },
+    {
+      "epoch": 12.009808429118774,
+      "grad_norm": 0.022570697590708733,
+      "learning_rate": 8.33546189868029e-06,
+      "loss": 0.1351,
+      "step": 3260
+    },
+    {
+      "epoch": 12.010574712643677,
+      "grad_norm": 0.04445415362715721,
+      "learning_rate": 8.326947637292465e-06,
+      "loss": 2.312,
+      "step": 3270
+    },
+    {
+      "epoch": 12.011340996168583,
+      "grad_norm": 0.39725983142852783,
+      "learning_rate": 8.318433375904642e-06,
+      "loss": 1.4933,
+      "step": 3280
+    },
+    {
+      "epoch": 12.012107279693486,
+      "grad_norm": 1.8836842775344849,
+      "learning_rate": 8.309919114516817e-06,
+      "loss": 0.9747,
+      "step": 3290
+    },
+    {
+      "epoch": 12.01287356321839,
+      "grad_norm": 147.71759033203125,
+      "learning_rate": 8.301404853128992e-06,
+      "loss": 1.5269,
+      "step": 3300
+    },
+    {
+      "epoch": 12.013639846743295,
+      "grad_norm": 1.9329392910003662,
+      "learning_rate": 8.292890591741167e-06,
+      "loss": 0.9688,
+      "step": 3310
+    },
+    {
+      "epoch": 12.014406130268199,
+      "grad_norm": 38.93643569946289,
+      "learning_rate": 8.284376330353342e-06,
+      "loss": 1.4943,
+      "step": 3320
+    },
+    {
+      "epoch": 12.015172413793103,
+      "grad_norm": 0.17174561321735382,
+      "learning_rate": 8.275862068965518e-06,
+      "loss": 1.0749,
+      "step": 3330
+    },
+    {
+      "epoch": 12.015938697318008,
+      "grad_norm": 31.687175750732422,
+      "learning_rate": 8.267347807577693e-06,
+      "loss": 0.9041,
+      "step": 3340
+    },
+    {
+      "epoch": 12.016704980842912,
+      "grad_norm": 0.12959319353103638,
+      "learning_rate": 8.258833546189868e-06,
+      "loss": 1.0097,
+      "step": 3350
+    },
+    {
+      "epoch": 12.017471264367815,
+      "grad_norm": 0.02093736082315445,
+      "learning_rate": 8.250319284802043e-06,
+      "loss": 1.1702,
+      "step": 3360
+    },
+    {
+      "epoch": 12.01823754789272,
+      "grad_norm": 25.50687026977539,
+      "learning_rate": 8.24180502341422e-06,
+      "loss": 0.5414,
+      "step": 3370
+    },
+    {
+      "epoch": 12.019003831417624,
+      "grad_norm": 0.06102278456091881,
+      "learning_rate": 8.233290762026395e-06,
+      "loss": 1.0346,
+      "step": 3380
+    },
+    {
+      "epoch": 12.01977011494253,
+      "grad_norm": 0.15459617972373962,
+      "learning_rate": 8.22477650063857e-06,
+      "loss": 0.31,
+      "step": 3390
+    },
+    {
+      "epoch": 12.02,
+      "eval_accuracy": 0.7777777777777778,
+      "eval_loss": 0.8154128193855286,
+      "eval_runtime": 14.3191,
+      "eval_samples_per_second": 3.143,
+      "eval_steps_per_second": 3.143,
+      "step": 3393
+    },
+    {
+      "epoch": 13.000536398467434,
+      "grad_norm": 19.95181655883789,
+      "learning_rate": 8.216262239250745e-06,
+      "loss": 1.3112,
+      "step": 3400
+    },
+    {
+      "epoch": 13.001302681992337,
+      "grad_norm": 0.020247263833880424,
+      "learning_rate": 8.207747977862922e-06,
+      "loss": 0.7465,
+      "step": 3410
+    },
+    {
+      "epoch": 13.00206896551724,
+      "grad_norm": 18.45448112487793,
+      "learning_rate": 8.199233716475097e-06,
+      "loss": 1.4082,
+      "step": 3420
+    },
+    {
+      "epoch": 13.002835249042146,
+      "grad_norm": 39.223690032958984,
+      "learning_rate": 8.190719455087272e-06,
+      "loss": 0.1986,
+      "step": 3430
+    },
+    {
+      "epoch": 13.00360153256705,
+      "grad_norm": 0.008466905914247036,
+      "learning_rate": 8.182205193699447e-06,
+      "loss": 0.7523,
+      "step": 3440
+    },
+    {
+      "epoch": 13.004367816091953,
+      "grad_norm": 0.12857244908809662,
+      "learning_rate": 8.173690932311623e-06,
+      "loss": 0.6167,
+      "step": 3450
+    },
+    {
+      "epoch": 13.005134099616859,
+      "grad_norm": 1.4279407262802124,
+      "learning_rate": 8.165176670923798e-06,
+      "loss": 0.5034,
+      "step": 3460
+    },
+    {
+      "epoch": 13.005900383141762,
+      "grad_norm": 0.1705969125032425,
+      "learning_rate": 8.156662409535973e-06,
+      "loss": 0.5446,
+      "step": 3470
+    },
+    {
+      "epoch": 13.006666666666666,
+      "grad_norm": 0.724726140499115,
+      "learning_rate": 8.148148148148148e-06,
+      "loss": 0.6268,
+      "step": 3480
+    },
+    {
+      "epoch": 13.007432950191571,
+      "grad_norm": 0.21082398295402527,
+      "learning_rate": 8.139633886760325e-06,
+      "loss": 0.3352,
+      "step": 3490
+    },
+    {
+      "epoch": 13.008199233716475,
+      "grad_norm": 234.13392639160156,
+      "learning_rate": 8.1311196253725e-06,
+      "loss": 2.6406,
+      "step": 3500
+    },
+    {
+      "epoch": 13.008965517241379,
+      "grad_norm": 4.463503837585449,
+      "learning_rate": 8.122605363984675e-06,
+      "loss": 1.2419,
+      "step": 3510
+    },
+    {
+      "epoch": 13.009731800766284,
+      "grad_norm": 6.503235816955566,
+      "learning_rate": 8.11409110259685e-06,
+      "loss": 0.349,
+      "step": 3520
+    },
+    {
+      "epoch": 13.010498084291187,
+      "grad_norm": 50.43798828125,
+      "learning_rate": 8.105576841209027e-06,
+      "loss": 1.2279,
+      "step": 3530
+    },
+    {
+      "epoch": 13.011264367816091,
+      "grad_norm": 5.821425914764404,
+      "learning_rate": 8.097062579821202e-06,
+      "loss": 1.5754,
+      "step": 3540
+    },
+    {
+      "epoch": 13.012030651340996,
+      "grad_norm": 0.020040128380060196,
+      "learning_rate": 8.088548318433377e-06,
+      "loss": 0.9977,
+      "step": 3550
+    },
+    {
+      "epoch": 13.0127969348659,
+      "grad_norm": 213.29562377929688,
+      "learning_rate": 8.080034057045552e-06,
+      "loss": 1.5818,
+      "step": 3560
+    },
+    {
+      "epoch": 13.013563218390805,
+      "grad_norm": 18.644872665405273,
+      "learning_rate": 8.071519795657727e-06,
+      "loss": 0.8922,
+      "step": 3570
+    },
+    {
+      "epoch": 13.014329501915709,
+      "grad_norm": 0.29788148403167725,
+      "learning_rate": 8.063005534269903e-06,
+      "loss": 0.4074,
+      "step": 3580
+    },
+    {
+      "epoch": 13.015095785440613,
+      "grad_norm": 0.3964133858680725,
+      "learning_rate": 8.054491272882078e-06,
+      "loss": 0.3987,
+      "step": 3590
+    },
+    {
+      "epoch": 13.015862068965518,
+      "grad_norm": 0.017643410712480545,
+      "learning_rate": 8.045977011494253e-06,
+      "loss": 0.1342,
+      "step": 3600
+    },
+    {
+      "epoch": 13.016628352490422,
+      "grad_norm": 0.29650014638900757,
+      "learning_rate": 8.037462750106428e-06,
+      "loss": 1.569,
+      "step": 3610
+    },
+    {
+      "epoch": 13.017394636015325,
+      "grad_norm": 3.4459073543548584,
+      "learning_rate": 8.028948488718605e-06,
+      "loss": 1.8939,
+      "step": 3620
+    },
+    {
+      "epoch": 13.01816091954023,
+      "grad_norm": 957.4302978515625,
+      "learning_rate": 8.02043422733078e-06,
+      "loss": 1.4765,
+      "step": 3630
+    },
+    {
+      "epoch": 13.018927203065134,
+      "grad_norm": 0.47024571895599365,
+      "learning_rate": 8.011919965942955e-06,
+      "loss": 0.9642,
+      "step": 3640
+    },
+    {
+      "epoch": 13.019693486590038,
+      "grad_norm": 1.1732593774795532,
+      "learning_rate": 8.00340570455513e-06,
+      "loss": 1.2834,
+      "step": 3650
+    },
+    {
+      "epoch": 13.02,
+      "eval_accuracy": 0.8666666666666667,
+      "eval_loss": 0.5090304613113403,
+      "eval_runtime": 15.0562,
+      "eval_samples_per_second": 2.989,
+      "eval_steps_per_second": 2.989,
+      "step": 3654
+    },
+    {
+      "epoch": 14.000459770114942,
+      "grad_norm": 594.275634765625,
+      "learning_rate": 7.994891443167307e-06,
+      "loss": 0.8765,
+      "step": 3660
+    },
+    {
+      "epoch": 14.001226053639847,
+      "grad_norm": 0.0192561112344265,
+      "learning_rate": 7.986377181779482e-06,
+      "loss": 1.1006,
+      "step": 3670
+    },
+    {
+      "epoch": 14.00199233716475,
+      "grad_norm": 0.007181501016020775,
+      "learning_rate": 7.977862920391657e-06,
+      "loss": 0.4418,
+      "step": 3680
+    },
+    {
+      "epoch": 14.002758620689654,
+      "grad_norm": 0.016903869807720184,
+      "learning_rate": 7.969348659003832e-06,
+      "loss": 0.4686,
+      "step": 3690
+    },
+    {
+      "epoch": 14.00352490421456,
+      "grad_norm": 0.25686225295066833,
+      "learning_rate": 7.960834397616007e-06,
+      "loss": 0.9965,
+      "step": 3700
+    },
+    {
+      "epoch": 14.004291187739463,
+      "grad_norm": 0.7215161919593811,
+      "learning_rate": 7.952320136228183e-06,
+      "loss": 0.006,
+      "step": 3710
+    },
+    {
+      "epoch": 14.005057471264367,
+      "grad_norm": 19.98381996154785,
+      "learning_rate": 7.943805874840358e-06,
+      "loss": 1.5662,
+      "step": 3720
+    },
+    {
+      "epoch": 14.005823754789272,
+      "grad_norm": 0.016891516745090485,
+      "learning_rate": 7.935291613452533e-06,
+      "loss": 1.9538,
+      "step": 3730
+    },
+    {
+      "epoch": 14.006590038314176,
+      "grad_norm": 1.6335327625274658,
+      "learning_rate": 7.92677735206471e-06,
+      "loss": 0.6845,
+      "step": 3740
+    },
+    {
+      "epoch": 14.007356321839081,
+      "grad_norm": 74.78628540039062,
+      "learning_rate": 7.918263090676885e-06,
+      "loss": 0.7214,
+      "step": 3750
+    },
+    {
+      "epoch": 14.008122605363985,
+      "grad_norm": 0.01958787813782692,
+      "learning_rate": 7.90974882928906e-06,
+      "loss": 0.5193,
+      "step": 3760
+    },
+    {
+      "epoch": 14.008888888888889,
+      "grad_norm": 948.969970703125,
+      "learning_rate": 7.901234567901235e-06,
+      "loss": 1.4058,
+      "step": 3770
+    },
+    {
+      "epoch": 14.009655172413794,
+      "grad_norm": 0.07399129867553711,
+      "learning_rate": 7.89272030651341e-06,
+      "loss": 1.5285,
+      "step": 3780
+    },
+    {
+      "epoch": 14.010421455938697,
+      "grad_norm": 0.11624588072299957,
+      "learning_rate": 7.884206045125587e-06,
+      "loss": 1.6973,
+      "step": 3790
+    },
+    {
+      "epoch": 14.011187739463601,
+      "grad_norm": 1.2204347848892212,
+      "learning_rate": 7.875691783737762e-06,
+      "loss": 1.1063,
+      "step": 3800
+    },
+    {
+      "epoch": 14.011954022988506,
+      "grad_norm": 0.1173257902264595,
+      "learning_rate": 7.867177522349937e-06,
+      "loss": 0.9653,
+      "step": 3810
+    },
+    {
+      "epoch": 14.01272030651341,
+      "grad_norm": 0.05164877697825432,
+      "learning_rate": 7.858663260962112e-06,
+      "loss": 1.7888,
+      "step": 3820
+    },
+    {
+      "epoch": 14.013486590038314,
+      "grad_norm": 151.31732177734375,
+      "learning_rate": 7.850148999574287e-06,
+      "loss": 0.9606,
+      "step": 3830
+    },
+    {
+      "epoch": 14.014252873563219,
+      "grad_norm": 0.1222265437245369,
+      "learning_rate": 7.841634738186463e-06,
+      "loss": 0.2589,
+      "step": 3840
+    },
+    {
+      "epoch": 14.015019157088123,
+      "grad_norm": 0.30466440320014954,
+      "learning_rate": 7.833120476798638e-06,
+      "loss": 0.5544,
+      "step": 3850
+    },
+    {
+      "epoch": 14.015785440613026,
+      "grad_norm": 0.015721678733825684,
+      "learning_rate": 7.824606215410813e-06,
+      "loss": 0.6975,
+      "step": 3860
+    },
+    {
+      "epoch": 14.016551724137932,
+      "grad_norm": 96.57806396484375,
+      "learning_rate": 7.81609195402299e-06,
+      "loss": 0.6042,
+      "step": 3870
+    },
+    {
+      "epoch": 14.017318007662835,
+      "grad_norm": 0.08435537666082382,
+      "learning_rate": 7.807577692635165e-06,
+      "loss": 0.1693,
+      "step": 3880
+    },
+    {
+      "epoch": 14.018084291187739,
+      "grad_norm": 0.6997672915458679,
+      "learning_rate": 7.79906343124734e-06,
+      "loss": 1.8493,
+      "step": 3890
+    },
+    {
+      "epoch": 14.018850574712644,
+      "grad_norm": 0.7452917098999023,
+      "learning_rate": 7.790549169859515e-06,
+      "loss": 1.0341,
+      "step": 3900
+    },
+    {
+      "epoch": 14.019616858237548,
+      "grad_norm": 31.302000045776367,
+      "learning_rate": 7.78203490847169e-06,
+      "loss": 0.7111,
+      "step": 3910
+    },
+    {
+      "epoch": 14.02,
+      "eval_accuracy": 0.6888888888888889,
+      "eval_loss": 1.2499884366989136,
+      "eval_runtime": 14.7662,
+      "eval_samples_per_second": 3.048,
+      "eval_steps_per_second": 3.048,
+      "step": 3915
+    },
+    {
+      "epoch": 15.000383141762452,
+      "grad_norm": 0.09833050519227982,
+      "learning_rate": 7.773520647083867e-06,
+      "loss": 2.1868,
+      "step": 3920
+    },
+    {
+      "epoch": 15.001149425287357,
+      "grad_norm": 0.0771474689245224,
+      "learning_rate": 7.765006385696042e-06,
+      "loss": 0.8732,
+      "step": 3930
+    },
+    {
+      "epoch": 15.00191570881226,
+      "grad_norm": 0.12298612296581268,
+      "learning_rate": 7.756492124308217e-06,
+      "loss": 1.2974,
+      "step": 3940
+    },
+    {
+      "epoch": 15.002681992337164,
+      "grad_norm": 386.5329284667969,
+      "learning_rate": 7.747977862920393e-06,
+      "loss": 0.1457,
+      "step": 3950
+    },
+    {
+      "epoch": 15.00344827586207,
+      "grad_norm": 0.1593717783689499,
+      "learning_rate": 7.739463601532567e-06,
+      "loss": 1.3177,
+      "step": 3960
+    },
+    {
+      "epoch": 15.004214559386973,
+      "grad_norm": 5.293508052825928,
+      "learning_rate": 7.730949340144743e-06,
+      "loss": 0.5551,
+      "step": 3970
+    },
+    {
+      "epoch": 15.004980842911877,
+      "grad_norm": 408.6733093261719,
+      "learning_rate": 7.722435078756918e-06,
+      "loss": 1.4917,
+      "step": 3980
+    },
+    {
+      "epoch": 15.005747126436782,
+      "grad_norm": 137.54627990722656,
+      "learning_rate": 7.713920817369093e-06,
+      "loss": 0.4599,
+      "step": 3990
+    },
+    {
+      "epoch": 15.006513409961686,
+      "grad_norm": 335.1684875488281,
+      "learning_rate": 7.70540655598127e-06,
+      "loss": 1.0811,
+      "step": 4000
+    },
+    {
+      "epoch": 15.00727969348659,
+      "grad_norm": 19.96000099182129,
+      "learning_rate": 7.696892294593445e-06,
+      "loss": 1.2689,
+      "step": 4010
+    },
+    {
+      "epoch": 15.008045977011495,
+      "grad_norm": 5.857684135437012,
+      "learning_rate": 7.68837803320562e-06,
+      "loss": 1.0879,
+      "step": 4020
+    },
+    {
+      "epoch": 15.008812260536398,
+      "grad_norm": 0.012607713229954243,
+      "learning_rate": 7.679863771817797e-06,
+      "loss": 0.3736,
+      "step": 4030
+    },
+    {
+      "epoch": 15.009578544061302,
+      "grad_norm": 2.0022695064544678,
+      "learning_rate": 7.67134951042997e-06,
+      "loss": 1.3587,
+      "step": 4040
+    },
+    {
+      "epoch": 15.010344827586207,
+      "grad_norm": 0.08656860142946243,
+      "learning_rate": 7.662835249042147e-06,
+      "loss": 0.3155,
+      "step": 4050
+    },
+    {
+      "epoch": 15.011111111111111,
+      "grad_norm": 0.09074855595827103,
+      "learning_rate": 7.654320987654322e-06,
+      "loss": 0.0547,
+      "step": 4060
+    },
+    {
+      "epoch": 15.011877394636015,
+      "grad_norm": 4.4385223388671875,
+      "learning_rate": 7.645806726266497e-06,
+      "loss": 3.1175,
+      "step": 4070
+    },
+    {
+      "epoch": 15.01264367816092,
+      "grad_norm": 0.578523576259613,
+      "learning_rate": 7.637292464878673e-06,
+      "loss": 1.1603,
+      "step": 4080
+    },
+    {
+      "epoch": 15.013409961685824,
+      "grad_norm": 31.44670295715332,
+      "learning_rate": 7.6287782034908475e-06,
+      "loss": 1.1973,
+      "step": 4090
+    },
+    {
+      "epoch": 15.014176245210727,
+      "grad_norm": 0.009388735517859459,
+      "learning_rate": 7.620263942103023e-06,
+      "loss": 1.1828,
+      "step": 4100
+    },
+    {
+      "epoch": 15.014942528735633,
+      "grad_norm": 0.02472112700343132,
+      "learning_rate": 7.611749680715198e-06,
+      "loss": 0.357,
+      "step": 4110
+    },
+    {
+      "epoch": 15.015708812260536,
+      "grad_norm": 54.052547454833984,
+      "learning_rate": 7.603235419327374e-06,
+      "loss": 2.107,
+      "step": 4120
+    },
+    {
+      "epoch": 15.01647509578544,
+      "grad_norm": 0.5136244297027588,
+      "learning_rate": 7.59472115793955e-06,
+      "loss": 0.1768,
+      "step": 4130
+    },
+    {
+      "epoch": 15.017241379310345,
+      "grad_norm": 0.0888374000787735,
+      "learning_rate": 7.586206896551724e-06,
+      "loss": 0.6062,
+      "step": 4140
+    },
+    {
+      "epoch": 15.018007662835249,
+      "grad_norm": 0.32292622327804565,
+      "learning_rate": 7.5776926351639e-06,
+      "loss": 0.7942,
+      "step": 4150
+    },
+    {
+      "epoch": 15.018773946360152,
+      "grad_norm": 54.13147735595703,
+      "learning_rate": 7.569178373776076e-06,
+      "loss": 1.2646,
+      "step": 4160
+    },
+    {
+      "epoch": 15.019540229885058,
+      "grad_norm": 72.40896606445312,
+      "learning_rate": 7.560664112388251e-06,
+      "loss": 2.3551,
+      "step": 4170
+    },
+    {
+      "epoch": 15.02,
+      "eval_accuracy": 0.8222222222222222,
+      "eval_loss": 0.6881113648414612,
+      "eval_runtime": 15.5764,
+      "eval_samples_per_second": 2.889,
+      "eval_steps_per_second": 2.889,
+      "step": 4176
+    },
+    {
+      "epoch": 16.00030651340996,
+      "grad_norm": 0.17236869037151337,
+      "learning_rate": 7.552149851000427e-06,
+      "loss": 0.9726,
+      "step": 4180
+    },
+    {
+      "epoch": 16.001072796934867,
+      "grad_norm": 0.37512490153312683,
+      "learning_rate": 7.543635589612601e-06,
+      "loss": 0.0073,
+      "step": 4190
+    },
+    {
+      "epoch": 16.00183908045977,
+      "grad_norm": 341.52520751953125,
+      "learning_rate": 7.535121328224777e-06,
+      "loss": 1.8988,
+      "step": 4200
+    },
+    {
+      "epoch": 16.002605363984674,
+      "grad_norm": 43.17485427856445,
+      "learning_rate": 7.5266070668369525e-06,
+      "loss": 1.7032,
+      "step": 4210
+    },
+    {
+      "epoch": 16.003371647509578,
+      "grad_norm": 0.2590516209602356,
+      "learning_rate": 7.5180928054491275e-06,
+      "loss": 1.2217,
+      "step": 4220
+    },
+    {
+      "epoch": 16.00413793103448,
+      "grad_norm": 0.7092514038085938,
+      "learning_rate": 7.509578544061303e-06,
+      "loss": 0.9297,
+      "step": 4230
+    },
+    {
+      "epoch": 16.00490421455939,
+      "grad_norm": 0.04314722120761871,
+      "learning_rate": 7.501064282673479e-06,
+      "loss": 0.4931,
+      "step": 4240
+    },
+    {
+      "epoch": 16.005670498084292,
+      "grad_norm": 0.051585156470537186,
+      "learning_rate": 7.492550021285654e-06,
+      "loss": 1.5563,
+      "step": 4250
+    },
+    {
+      "epoch": 16.006436781609196,
+      "grad_norm": 1.28795325756073,
+      "learning_rate": 7.48403575989783e-06,
+      "loss": 1.452,
+      "step": 4260
+    },
+    {
+      "epoch": 16.0072030651341,
+      "grad_norm": 403.7496032714844,
+      "learning_rate": 7.475521498510004e-06,
+      "loss": 1.7767,
+      "step": 4270
+    },
+    {
+      "epoch": 16.007969348659003,
+      "grad_norm": 0.0961364358663559,
+      "learning_rate": 7.46700723712218e-06,
+      "loss": 0.4419,
+      "step": 4280
+    },
+    {
+      "epoch": 16.008735632183907,
+      "grad_norm": 0.14619210362434387,
+      "learning_rate": 7.458492975734356e-06,
+      "loss": 0.9539,
+      "step": 4290
+    },
+    {
+      "epoch": 16.009501915708814,
+      "grad_norm": 47.26947021484375,
+      "learning_rate": 7.449978714346531e-06,
+      "loss": 1.6533,
+      "step": 4300
+    },
+    {
+      "epoch": 16.010268199233717,
+      "grad_norm": 280.16326904296875,
+      "learning_rate": 7.441464452958707e-06,
+      "loss": 0.8886,
+      "step": 4310
+    },
+    {
+      "epoch": 16.01103448275862,
+      "grad_norm": 51.095741271972656,
+      "learning_rate": 7.4329501915708825e-06,
+      "loss": 0.7588,
+      "step": 4320
+    },
+    {
+      "epoch": 16.011800766283525,
+      "grad_norm": 2.6517045497894287,
+      "learning_rate": 7.4244359301830575e-06,
+      "loss": 0.1651,
+      "step": 4330
+    },
+    {
+      "epoch": 16.01256704980843,
+      "grad_norm": 0.10823695361614227,
+      "learning_rate": 7.4159216687952325e-06,
+      "loss": 0.8843,
+      "step": 4340
+    },
+    {
+      "epoch": 16.013333333333332,
+      "grad_norm": 0.05993505194783211,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 1.3364,
+      "step": 4350
+    },
+    {
+      "epoch": 16.01409961685824,
+      "grad_norm": 0.5145869255065918,
+      "learning_rate": 7.398893146019583e-06,
+      "loss": 0.9984,
+      "step": 4360
+    },
+    {
+      "epoch": 16.014865900383143,
+      "grad_norm": 3.6809394359588623,
+      "learning_rate": 7.390378884631759e-06,
+      "loss": 1.3562,
+      "step": 4370
+    },
+    {
+      "epoch": 16.015632183908046,
+      "grad_norm": 516.924560546875,
+      "learning_rate": 7.381864623243934e-06,
+      "loss": 0.4271,
+      "step": 4380
+    },
+    {
+      "epoch": 16.01639846743295,
+      "grad_norm": 1.5104776620864868,
+      "learning_rate": 7.37335036185611e-06,
+      "loss": 0.8018,
+      "step": 4390
+    },
+    {
+      "epoch": 16.017164750957853,
+      "grad_norm": 0.01973959617316723,
+      "learning_rate": 7.364836100468284e-06,
+      "loss": 0.9224,
+      "step": 4400
+    },
+    {
+      "epoch": 16.017931034482757,
+      "grad_norm": 2.8555283546447754,
+      "learning_rate": 7.35632183908046e-06,
+      "loss": 0.0167,
+      "step": 4410
+    },
+    {
+      "epoch": 16.018697318007664,
+      "grad_norm": 0.00982819963246584,
+      "learning_rate": 7.347807577692636e-06,
+      "loss": 0.7395,
+      "step": 4420
+    },
+    {
+      "epoch": 16.019463601532568,
+      "grad_norm": 0.015840256586670876,
+      "learning_rate": 7.339293316304811e-06,
+      "loss": 0.2734,
+      "step": 4430
+    },
+    {
+      "epoch": 16.02,
+      "eval_accuracy": 0.8444444444444444,
+      "eval_loss": 0.4506371021270752,
+      "eval_runtime": 13.6245,
+      "eval_samples_per_second": 3.303,
+      "eval_steps_per_second": 3.303,
+      "step": 4437
+    },
+    {
+      "epoch": 17.000229885057472,
+      "grad_norm": 1.947767972946167,
+      "learning_rate": 7.330779054916987e-06,
+      "loss": 0.0567,
+      "step": 4440
+    },
+    {
+      "epoch": 17.000996168582375,
+      "grad_norm": 383.13385009765625,
+      "learning_rate": 7.3222647935291625e-06,
+      "loss": 1.4525,
+      "step": 4450
+    },
+    {
+      "epoch": 17.00176245210728,
+      "grad_norm": 0.005475871730595827,
+      "learning_rate": 7.3137505321413375e-06,
+      "loss": 0.2238,
+      "step": 4460
+    },
+    {
+      "epoch": 17.002528735632183,
+      "grad_norm": 199.94427490234375,
+      "learning_rate": 7.305236270753513e-06,
+      "loss": 0.395,
+      "step": 4470
+    },
+    {
+      "epoch": 17.00329501915709,
+      "grad_norm": 0.12597985565662384,
+      "learning_rate": 7.2967220093656875e-06,
+      "loss": 1.2407,
+      "step": 4480
+    },
+    {
+      "epoch": 17.004061302681993,
+      "grad_norm": 22.926143646240234,
+      "learning_rate": 7.288207747977863e-06,
+      "loss": 1.3336,
+      "step": 4490
+    },
+    {
+      "epoch": 17.004827586206897,
+      "grad_norm": 2.7508955001831055,
+      "learning_rate": 7.279693486590039e-06,
+      "loss": 1.735,
+      "step": 4500
+    },
+    {
+      "epoch": 17.0055938697318,
+      "grad_norm": 0.7520158886909485,
+      "learning_rate": 7.271179225202214e-06,
+      "loss": 0.5005,
+      "step": 4510
+    },
+    {
+      "epoch": 17.006360153256704,
+      "grad_norm": 172.81605529785156,
+      "learning_rate": 7.26266496381439e-06,
+      "loss": 1.0526,
+      "step": 4520
+    },
+    {
+      "epoch": 17.007126436781608,
+      "grad_norm": 2.283644199371338,
+      "learning_rate": 7.254150702426566e-06,
+      "loss": 1.2169,
+      "step": 4530
+    },
+    {
+      "epoch": 17.007892720306515,
+      "grad_norm": 0.5311365127563477,
+      "learning_rate": 7.24563644103874e-06,
+      "loss": 0.5474,
+      "step": 4540
+    },
+    {
+      "epoch": 17.00865900383142,
+      "grad_norm": 0.0861196294426918,
+      "learning_rate": 7.237122179650916e-06,
+      "loss": 1.158,
+      "step": 4550
+    },
+    {
+      "epoch": 17.009425287356322,
+      "grad_norm": 0.42244112491607666,
+      "learning_rate": 7.228607918263091e-06,
+      "loss": 1.3332,
+      "step": 4560
+    },
+    {
+      "epoch": 17.010191570881226,
+      "grad_norm": 0.16180934011936188,
+      "learning_rate": 7.220093656875267e-06,
+      "loss": 0.0296,
+      "step": 4570
+    },
+    {
+      "epoch": 17.01095785440613,
+      "grad_norm": 529.973388671875,
+      "learning_rate": 7.2115793954874425e-06,
+      "loss": 0.9939,
+      "step": 4580
+    },
+    {
+      "epoch": 17.011724137931033,
+      "grad_norm": 0.008947962895035744,
+      "learning_rate": 7.2030651340996175e-06,
+      "loss": 0.3492,
+      "step": 4590
+    },
+    {
+      "epoch": 17.01249042145594,
+      "grad_norm": 0.0053781550377607346,
+      "learning_rate": 7.194550872711793e-06,
+      "loss": 0.8263,
+      "step": 4600
+    },
+    {
+      "epoch": 17.013256704980844,
+      "grad_norm": 106.14468383789062,
+      "learning_rate": 7.1860366113239675e-06,
+      "loss": 1.6513,
+      "step": 4610
+    },
+    {
+      "epoch": 17.014022988505747,
+      "grad_norm": 0.2938723564147949,
+      "learning_rate": 7.177522349936143e-06,
+      "loss": 0.447,
+      "step": 4620
+    },
+    {
+      "epoch": 17.01478927203065,
+      "grad_norm": 2.80863881111145,
+      "learning_rate": 7.169008088548319e-06,
+      "loss": 1.0175,
+      "step": 4630
+    },
+    {
+      "epoch": 17.015555555555554,
+      "grad_norm": 57.34475326538086,
+      "learning_rate": 7.160493827160494e-06,
+      "loss": 0.5093,
+      "step": 4640
+    },
+    {
+      "epoch": 17.016321839080458,
+      "grad_norm": 0.034363433718681335,
+      "learning_rate": 7.15197956577267e-06,
+      "loss": 0.584,
+      "step": 4650
+    },
+    {
+      "epoch": 17.017088122605365,
+      "grad_norm": 0.12468564510345459,
+      "learning_rate": 7.143465304384846e-06,
+      "loss": 0.3767,
+      "step": 4660
+    },
+    {
+      "epoch": 17.01785440613027,
+      "grad_norm": 0.005793506279587746,
+      "learning_rate": 7.13495104299702e-06,
+      "loss": 0.7981,
+      "step": 4670
+    },
+    {
+      "epoch": 17.018620689655172,
+      "grad_norm": 0.0034271469339728355,
+      "learning_rate": 7.126436781609196e-06,
+      "loss": 0.6918,
+      "step": 4680
+    },
+    {
+      "epoch": 17.019386973180076,
+      "grad_norm": 0.31982433795928955,
+      "learning_rate": 7.117922520221371e-06,
+      "loss": 0.9675,
+      "step": 4690
+    },
+    {
+      "epoch": 17.02,
+      "eval_accuracy": 0.6666666666666666,
+      "eval_loss": 1.7514704465866089,
+      "eval_runtime": 13.3064,
+      "eval_samples_per_second": 3.382,
+      "eval_steps_per_second": 3.382,
+      "step": 4698
+    },
+    {
+      "epoch": 18.00015325670498,
+      "grad_norm": 0.16283735632896423,
+      "learning_rate": 7.109408258833547e-06,
+      "loss": 1.3484,
+      "step": 4700
+    },
+    {
+      "epoch": 18.000919540229884,
+      "grad_norm": 16.916929244995117,
+      "learning_rate": 7.1008939974457225e-06,
+      "loss": 2.394,
+      "step": 4710
+    },
+    {
+      "epoch": 18.00168582375479,
+      "grad_norm": 16.609804153442383,
+      "learning_rate": 7.0923797360578975e-06,
+      "loss": 0.3868,
+      "step": 4720
+    },
+    {
+      "epoch": 18.002452107279694,
+      "grad_norm": 16.240642547607422,
+      "learning_rate": 7.083865474670073e-06,
+      "loss": 0.7722,
+      "step": 4730
+    },
+    {
+      "epoch": 18.003218390804598,
+      "grad_norm": 1.8989708423614502,
+      "learning_rate": 7.075351213282249e-06,
+      "loss": 0.2564,
+      "step": 4740
+    },
+    {
+      "epoch": 18.0039846743295,
+      "grad_norm": 0.0033782178070396185,
+      "learning_rate": 7.066836951894423e-06,
+      "loss": 0.6987,
+      "step": 4750
+    },
+    {
+      "epoch": 18.004750957854405,
+      "grad_norm": 18.385026931762695,
+      "learning_rate": 7.058322690506599e-06,
+      "loss": 2.9738,
+      "step": 4760
+    },
+    {
+      "epoch": 18.00551724137931,
+      "grad_norm": 0.004265069495886564,
+      "learning_rate": 7.049808429118774e-06,
+      "loss": 1.5315,
+      "step": 4770
+    },
+    {
+      "epoch": 18.006283524904216,
+      "grad_norm": 0.07541969418525696,
+      "learning_rate": 7.04129416773095e-06,
+      "loss": 1.4052,
+      "step": 4780
+    },
+    {
+      "epoch": 18.00704980842912,
+      "grad_norm": 0.004919661208987236,
+      "learning_rate": 7.032779906343126e-06,
+      "loss": 0.7981,
+      "step": 4790
+    },
+    {
+      "epoch": 18.007816091954023,
+      "grad_norm": 0.38469570875167847,
+      "learning_rate": 7.0242656449553e-06,
+      "loss": 0.0221,
+      "step": 4800
+    },
+    {
+      "epoch": 18.008582375478927,
+      "grad_norm": 0.6571864485740662,
+      "learning_rate": 7.015751383567476e-06,
+      "loss": 0.341,
+      "step": 4810
+    },
+    {
+      "epoch": 18.00934865900383,
+      "grad_norm": 18.751163482666016,
+      "learning_rate": 7.007237122179652e-06,
+      "loss": 2.1029,
+      "step": 4820
+    },
+    {
+      "epoch": 18.010114942528734,
+      "grad_norm": 67.68429565429688,
+      "learning_rate": 6.998722860791827e-06,
+      "loss": 0.6903,
+      "step": 4830
+    },
+    {
+      "epoch": 18.01088122605364,
+      "grad_norm": 0.42845791578292847,
+      "learning_rate": 6.9902085994040025e-06,
+      "loss": 0.8219,
+      "step": 4840
+    },
+    {
+      "epoch": 18.011647509578545,
+      "grad_norm": 41.950721740722656,
+      "learning_rate": 6.9816943380161775e-06,
+      "loss": 0.8313,
+      "step": 4850
+    },
+    {
+      "epoch": 18.01241379310345,
+      "grad_norm": 16.151935577392578,
+      "learning_rate": 6.973180076628353e-06,
+      "loss": 1.5668,
+      "step": 4860
+    },
+    {
+      "epoch": 18.013180076628352,
+      "grad_norm": 0.09513141959905624,
+      "learning_rate": 6.964665815240529e-06,
+      "loss": 0.0103,
+      "step": 4870
+    },
+    {
+      "epoch": 18.013946360153255,
+      "grad_norm": 54.980491638183594,
+      "learning_rate": 6.956151553852703e-06,
+      "loss": 0.6221,
+      "step": 4880
+    },
+    {
+      "epoch": 18.014712643678163,
+      "grad_norm": 0.5550600290298462,
+      "learning_rate": 6.947637292464879e-06,
+      "loss": 1.1071,
+      "step": 4890
+    },
+    {
+      "epoch": 18.015478927203066,
+      "grad_norm": 0.00374088017269969,
+      "learning_rate": 6.939123031077054e-06,
+      "loss": 0.6851,
+      "step": 4900
+    },
+    {
+      "epoch": 18.01624521072797,
+      "grad_norm": 0.05271366983652115,
+      "learning_rate": 6.93060876968923e-06,
+      "loss": 0.0099,
+      "step": 4910
+    },
+    {
+      "epoch": 18.017011494252873,
+      "grad_norm": 0.3116820752620697,
+      "learning_rate": 6.922094508301406e-06,
+      "loss": 0.4008,
+      "step": 4920
+    },
+    {
+      "epoch": 18.017777777777777,
+      "grad_norm": 6.082730770111084,
+      "learning_rate": 6.913580246913581e-06,
+      "loss": 1.5001,
+      "step": 4930
+    },
+    {
+      "epoch": 18.01854406130268,
+      "grad_norm": 0.003171339863911271,
+      "learning_rate": 6.905065985525757e-06,
+      "loss": 1.2366,
+      "step": 4940
+    },
+    {
+      "epoch": 18.019310344827588,
+      "grad_norm": 0.35565054416656494,
+      "learning_rate": 6.896551724137932e-06,
+      "loss": 0.449,
+      "step": 4950
+    },
+    {
+      "epoch": 18.02,
+      "eval_accuracy": 0.7777777777777778,
+      "eval_loss": 0.7239938378334045,
+      "eval_runtime": 13.2593,
+      "eval_samples_per_second": 3.394,
+      "eval_steps_per_second": 3.394,
+      "step": 4959
+    },
+    {
+      "epoch": 19.000076628352492,
+      "grad_norm": 173.56851196289062,
+      "learning_rate": 6.888037462750107e-06,
+      "loss": 0.5259,
+      "step": 4960
+    },
+    {
+      "epoch": 19.000842911877395,
+      "grad_norm": 25.518754959106445,
+      "learning_rate": 6.8795232013622825e-06,
+      "loss": 0.8176,
+      "step": 4970
+    },
+    {
+      "epoch": 19.0016091954023,
+      "grad_norm": 0.004466726444661617,
+      "learning_rate": 6.8710089399744575e-06,
+      "loss": 0.3623,
+      "step": 4980
+    },
+    {
+      "epoch": 19.002375478927203,
+      "grad_norm": 1.6072368621826172,
+      "learning_rate": 6.862494678586633e-06,
+      "loss": 0.7608,
+      "step": 4990
+    },
+    {
+      "epoch": 19.003141762452106,
+      "grad_norm": 262.5456848144531,
+      "learning_rate": 6.853980417198809e-06,
+      "loss": 1.0434,
+      "step": 5000
+    },
+    {
+      "epoch": 19.00390804597701,
+      "grad_norm": 0.3474750220775604,
+      "learning_rate": 6.845466155810983e-06,
+      "loss": 2.76,
+      "step": 5010
+    },
+    {
+      "epoch": 19.004674329501917,
+      "grad_norm": 0.3191026449203491,
+      "learning_rate": 6.836951894423159e-06,
+      "loss": 0.921,
+      "step": 5020
+    },
+    {
+      "epoch": 19.00544061302682,
+      "grad_norm": 0.008138914592564106,
+      "learning_rate": 6.828437633035335e-06,
+      "loss": 1.0089,
+      "step": 5030
+    },
+    {
+      "epoch": 19.006206896551724,
+      "grad_norm": 226.68228149414062,
+      "learning_rate": 6.81992337164751e-06,
+      "loss": 0.466,
+      "step": 5040
+    },
+    {
+      "epoch": 19.006973180076628,
+      "grad_norm": 141.62277221679688,
+      "learning_rate": 6.811409110259686e-06,
+      "loss": 0.4943,
+      "step": 5050
+    },
+    {
+      "epoch": 19.00773946360153,
+      "grad_norm": 0.542086660861969,
+      "learning_rate": 6.802894848871861e-06,
+      "loss": 0.1911,
+      "step": 5060
+    },
+    {
+      "epoch": 19.00850574712644,
+      "grad_norm": 0.004684785380959511,
+      "learning_rate": 6.794380587484037e-06,
+      "loss": 1.7487,
+      "step": 5070
+    },
+    {
+      "epoch": 19.009272030651342,
+      "grad_norm": 0.17809058725833893,
+      "learning_rate": 6.7858663260962125e-06,
+      "loss": 0.4984,
+      "step": 5080
+    },
+    {
+      "epoch": 19.010038314176246,
+      "grad_norm": 3.824796438217163,
+      "learning_rate": 6.777352064708387e-06,
+      "loss": 0.4295,
+      "step": 5090
+    },
+    {
+      "epoch": 19.01080459770115,
+      "grad_norm": 18.227113723754883,
+      "learning_rate": 6.7688378033205625e-06,
+      "loss": 1.0123,
+      "step": 5100
+    },
+    {
+      "epoch": 19.011570881226053,
+      "grad_norm": 0.016013886779546738,
+      "learning_rate": 6.760323541932738e-06,
+      "loss": 0.5572,
+      "step": 5110
+    },
+    {
+      "epoch": 19.012337164750956,
+      "grad_norm": 0.010713410563766956,
+      "learning_rate": 6.751809280544913e-06,
+      "loss": 0.0082,
+      "step": 5120
+    },
+    {
+      "epoch": 19.013103448275864,
+      "grad_norm": 0.2963060140609741,
+      "learning_rate": 6.743295019157089e-06,
+      "loss": 0.823,
+      "step": 5130
+    },
+    {
+      "epoch": 19.013869731800767,
+      "grad_norm": 435.67083740234375,
+      "learning_rate": 6.734780757769263e-06,
+      "loss": 0.8881,
+      "step": 5140
+    },
+    {
+      "epoch": 19.01463601532567,
+      "grad_norm": 0.31719154119491577,
+      "learning_rate": 6.726266496381439e-06,
+      "loss": 0.405,
+      "step": 5150
+    },
+    {
+      "epoch": 19.015402298850574,
+      "grad_norm": 9.507041931152344,
+      "learning_rate": 6.717752234993615e-06,
+      "loss": 0.9877,
+      "step": 5160
+    },
+    {
+      "epoch": 19.016168582375478,
+      "grad_norm": 0.00443949643522501,
+      "learning_rate": 6.70923797360579e-06,
+      "loss": 0.0049,
+      "step": 5170
+    },
+    {
+      "epoch": 19.01693486590038,
+      "grad_norm": 17.073001861572266,
+      "learning_rate": 6.700723712217966e-06,
+      "loss": 1.0038,
+      "step": 5180
+    },
+    {
+      "epoch": 19.01770114942529,
+      "grad_norm": 0.0035155299119651318,
+      "learning_rate": 6.692209450830141e-06,
+      "loss": 0.3384,
+      "step": 5190
+    },
+    {
+      "epoch": 19.018467432950192,
+      "grad_norm": 0.05622873082756996,
+      "learning_rate": 6.683695189442317e-06,
+      "loss": 0.4446,
+      "step": 5200
+    },
+    {
+      "epoch": 19.019233716475096,
+      "grad_norm": 118.16485595703125,
+      "learning_rate": 6.6751809280544925e-06,
+      "loss": 1.8194,
+      "step": 5210
+    },
+    {
+      "epoch": 19.02,
+      "grad_norm": 0.19169829785823822,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.5843,
+      "step": 5220
+    },
+    {
+      "epoch": 19.02,
+      "eval_accuracy": 0.7777777777777778,
+      "eval_loss": 0.9560596942901611,
+      "eval_runtime": 14.2022,
+      "eval_samples_per_second": 3.169,
+      "eval_steps_per_second": 3.169,
+      "step": 5220
+    },
+    {
+      "epoch": 20.000766283524904,
+      "grad_norm": 138.09426879882812,
+      "learning_rate": 6.6581524052788425e-06,
+      "loss": 0.6963,
+      "step": 5230
+    },
+    {
+      "epoch": 20.001532567049807,
+      "grad_norm": 0.012549187056720257,
+      "learning_rate": 6.649638143891018e-06,
+      "loss": 0.6626,
+      "step": 5240
+    },
+    {
+      "epoch": 20.002298850574714,
+      "grad_norm": 0.004213971551507711,
+      "learning_rate": 6.641123882503193e-06,
+      "loss": 0.7115,
+      "step": 5250
+    },
+    {
+      "epoch": 20.003065134099618,
+      "grad_norm": 0.5979669690132141,
+      "learning_rate": 6.632609621115369e-06,
+      "loss": 0.9221,
+      "step": 5260
+    },
+    {
+      "epoch": 20.00383141762452,
+      "grad_norm": 0.29207003116607666,
+      "learning_rate": 6.624095359727543e-06,
+      "loss": 0.5816,
+      "step": 5270
+    },
+    {
+      "epoch": 20.004597701149425,
+      "grad_norm": 28.108585357666016,
+      "learning_rate": 6.615581098339719e-06,
+      "loss": 0.8503,
+      "step": 5280
+    },
+    {
+      "epoch": 20.00536398467433,
+      "grad_norm": 0.2536892890930176,
+      "learning_rate": 6.607066836951895e-06,
+      "loss": 0.4296,
+      "step": 5290
+    },
+    {
+      "epoch": 20.006130268199232,
+      "grad_norm": 0.2590063512325287,
+      "learning_rate": 6.59855257556407e-06,
+      "loss": 0.265,
+      "step": 5300
+    },
+    {
+      "epoch": 20.00689655172414,
+      "grad_norm": 0.3933335244655609,
+      "learning_rate": 6.590038314176246e-06,
+      "loss": 0.8066,
+      "step": 5310
+    },
+    {
+      "epoch": 20.007662835249043,
+      "grad_norm": 0.00344213773496449,
+      "learning_rate": 6.581524052788422e-06,
+      "loss": 0.3928,
+      "step": 5320
+    },
+    {
+      "epoch": 20.008429118773947,
+      "grad_norm": 0.003421328729018569,
+      "learning_rate": 6.573009791400597e-06,
+      "loss": 0.5046,
+      "step": 5330
+    },
+    {
+      "epoch": 20.00919540229885,
+      "grad_norm": 119.35005187988281,
+      "learning_rate": 6.5644955300127725e-06,
+      "loss": 0.6248,
+      "step": 5340
+    },
+    {
+      "epoch": 20.009961685823754,
+      "grad_norm": 1.8184597492218018,
+      "learning_rate": 6.555981268624947e-06,
+      "loss": 0.5608,
+      "step": 5350
+    },
+    {
+      "epoch": 20.010727969348657,
+      "grad_norm": 0.0024780328385531902,
+      "learning_rate": 6.5474670072371225e-06,
+      "loss": 0.6801,
+      "step": 5360
+    },
+    {
+      "epoch": 20.011494252873565,
+      "grad_norm": 0.004241599701344967,
+      "learning_rate": 6.538952745849298e-06,
+      "loss": 0.8765,
+      "step": 5370
+    },
+    {
+      "epoch": 20.01226053639847,
+      "grad_norm": 162.39012145996094,
+      "learning_rate": 6.530438484461473e-06,
+      "loss": 1.77,
+      "step": 5380
+    },
+    {
+      "epoch": 20.013026819923372,
+      "grad_norm": 0.0034117966424673796,
+      "learning_rate": 6.521924223073649e-06,
+      "loss": 1.6807,
+      "step": 5390
+    },
+    {
+      "epoch": 20.013793103448275,
+      "grad_norm": 0.011221050284802914,
+      "learning_rate": 6.513409961685824e-06,
+      "loss": 1.3964,
+      "step": 5400
+    },
+    {
+      "epoch": 20.01455938697318,
+      "grad_norm": 0.10529179871082306,
+      "learning_rate": 6.504895700297999e-06,
+      "loss": 1.1688,
+      "step": 5410
+    },
+    {
+      "epoch": 20.015325670498083,
+      "grad_norm": 0.2718166410923004,
+      "learning_rate": 6.496381438910175e-06,
+      "loss": 0.4018,
+      "step": 5420
+    },
+    {
+      "epoch": 20.01609195402299,
+      "grad_norm": 0.010549083352088928,
+      "learning_rate": 6.48786717752235e-06,
+      "loss": 0.6929,
+      "step": 5430
+    },
+    {
+      "epoch": 20.016858237547893,
+      "grad_norm": 0.9952169060707092,
+      "learning_rate": 6.479352916134526e-06,
+      "loss": 0.1293,
+      "step": 5440
+    },
+    {
+      "epoch": 20.017624521072797,
+      "grad_norm": 137.77610778808594,
+      "learning_rate": 6.470838654746702e-06,
+      "loss": 0.8465,
+      "step": 5450
+    },
+    {
+      "epoch": 20.0183908045977,
+      "grad_norm": 0.004174103494733572,
+      "learning_rate": 6.462324393358877e-06,
+      "loss": 0.9037,
+      "step": 5460
+    },
+    {
+      "epoch": 20.019157088122604,
+      "grad_norm": 0.019375838339328766,
+      "learning_rate": 6.4538101319710525e-06,
+      "loss": 0.5878,
+      "step": 5470
+    },
+    {
+      "epoch": 20.01992337164751,
+      "grad_norm": 0.01853746734559536,
+      "learning_rate": 6.445295870583227e-06,
+      "loss": 1.0949,
+      "step": 5480
+    },
+    {
+      "epoch": 20.02,
+      "eval_accuracy": 0.6888888888888889,
+      "eval_loss": 1.2866085767745972,
+      "eval_runtime": 13.2844,
+      "eval_samples_per_second": 3.387,
+      "eval_steps_per_second": 3.387,
+      "step": 5481
+    },
+    {
+      "epoch": 21.000689655172415,
+      "grad_norm": 0.010936838574707508,
+      "learning_rate": 6.4367816091954025e-06,
+      "loss": 1.8286,
+      "step": 5490
+    },
+    {
+      "epoch": 21.00145593869732,
+      "grad_norm": 0.0537358820438385,
+      "learning_rate": 6.428267347807578e-06,
+      "loss": 1.1297,
+      "step": 5500
+    },
+    {
+      "epoch": 21.002222222222223,
+      "grad_norm": 0.018108276650309563,
+      "learning_rate": 6.419753086419753e-06,
+      "loss": 0.2811,
+      "step": 5510
+    },
+    {
+      "epoch": 21.002988505747126,
+      "grad_norm": 0.00682029128074646,
+      "learning_rate": 6.411238825031929e-06,
+      "loss": 0.8071,
+      "step": 5520
+    },
+    {
+      "epoch": 21.00375478927203,
+      "grad_norm": 0.05539993569254875,
+      "learning_rate": 6.402724563644105e-06,
+      "loss": 0.9488,
+      "step": 5530
+    },
+    {
+      "epoch": 21.004521072796933,
+      "grad_norm": 0.2652181386947632,
+      "learning_rate": 6.39421030225628e-06,
+      "loss": 0.8627,
+      "step": 5540
+    },
+    {
+      "epoch": 21.00528735632184,
+      "grad_norm": 581.5076904296875,
+      "learning_rate": 6.385696040868455e-06,
+      "loss": 0.5053,
+      "step": 5550
+    },
+    {
+      "epoch": 21.006053639846744,
+      "grad_norm": 0.4918936789035797,
+      "learning_rate": 6.37718177948063e-06,
+      "loss": 0.0138,
+      "step": 5560
+    },
+    {
+      "epoch": 21.006819923371648,
+      "grad_norm": 0.37181881070137024,
+      "learning_rate": 6.368667518092806e-06,
+      "loss": 0.0198,
+      "step": 5570
+    },
+    {
+      "epoch": 21.00758620689655,
+      "grad_norm": 0.009263670071959496,
+      "learning_rate": 6.360153256704982e-06,
+      "loss": 1.2818,
+      "step": 5580
+    },
+    {
+      "epoch": 21.008352490421455,
+      "grad_norm": 0.0035500682424753904,
+      "learning_rate": 6.351638995317157e-06,
+      "loss": 0.6626,
+      "step": 5590
+    },
+    {
+      "epoch": 21.00911877394636,
+      "grad_norm": 391.8262023925781,
+      "learning_rate": 6.3431247339293325e-06,
+      "loss": 0.776,
+      "step": 5600
+    },
+    {
+      "epoch": 21.009885057471266,
+      "grad_norm": 227.4144287109375,
+      "learning_rate": 6.334610472541508e-06,
+      "loss": 1.1783,
+      "step": 5610
+    },
+    {
+      "epoch": 21.01065134099617,
+      "grad_norm": 0.007147862110286951,
+      "learning_rate": 6.3260962111536825e-06,
+      "loss": 0.0036,
+      "step": 5620
+    },
+    {
+      "epoch": 21.011417624521073,
+      "grad_norm": 0.0029464103281497955,
+      "learning_rate": 6.317581949765858e-06,
+      "loss": 1.3983,
+      "step": 5630
+    },
+    {
+      "epoch": 21.012183908045976,
+      "grad_norm": 0.2004404515028,
+      "learning_rate": 6.309067688378033e-06,
+      "loss": 1.3843,
+      "step": 5640
+    },
+    {
+      "epoch": 21.01295019157088,
+      "grad_norm": 0.06141708418726921,
+      "learning_rate": 6.300553426990209e-06,
+      "loss": 0.0749,
+      "step": 5650
+    },
+    {
+      "epoch": 21.013716475095784,
+      "grad_norm": 0.10852473974227905,
+      "learning_rate": 6.292039165602385e-06,
+      "loss": 0.256,
+      "step": 5660
+    },
+    {
+      "epoch": 21.01448275862069,
+      "grad_norm": 0.00618616072461009,
+      "learning_rate": 6.28352490421456e-06,
+      "loss": 0.4822,
+      "step": 5670
+    },
+    {
+      "epoch": 21.015249042145594,
+      "grad_norm": 348.1380615234375,
+      "learning_rate": 6.275010642826736e-06,
+      "loss": 1.542,
+      "step": 5680
+    },
+    {
+      "epoch": 21.016015325670498,
+      "grad_norm": 59.00828552246094,
+      "learning_rate": 6.26649638143891e-06,
+      "loss": 0.3801,
+      "step": 5690
+    },
+    {
+      "epoch": 21.0167816091954,
+      "grad_norm": 32.972965240478516,
+      "learning_rate": 6.257982120051086e-06,
+      "loss": 1.4585,
+      "step": 5700
+    },
+    {
+      "epoch": 21.017547892720305,
+      "grad_norm": 0.1276509165763855,
+      "learning_rate": 6.249467858663262e-06,
+      "loss": 0.5012,
+      "step": 5710
+    },
+    {
+      "epoch": 21.018314176245212,
+      "grad_norm": 0.27684059739112854,
+      "learning_rate": 6.240953597275437e-06,
+      "loss": 1.6095,
+      "step": 5720
+    },
+    {
+      "epoch": 21.019080459770116,
+      "grad_norm": 66.66431427001953,
+      "learning_rate": 6.2324393358876125e-06,
+      "loss": 1.5397,
+      "step": 5730
+    },
+    {
+      "epoch": 21.01984674329502,
+      "grad_norm": 2.397763967514038,
+      "learning_rate": 6.223925074499788e-06,
+      "loss": 1.2073,
+      "step": 5740
+    },
+    {
+      "epoch": 21.02,
+      "eval_accuracy": 0.7555555555555555,
+      "eval_loss": 1.033568024635315,
+      "eval_runtime": 14.5617,
+      "eval_samples_per_second": 3.09,
+      "eval_steps_per_second": 3.09,
+      "step": 5742
+    },
+    {
+      "epoch": 22.000613026819924,
+      "grad_norm": 0.05683393031358719,
+      "learning_rate": 6.2154108131119625e-06,
+      "loss": 0.4272,
+      "step": 5750
+    },
+    {
+      "epoch": 22.001379310344827,
+      "grad_norm": 0.4416787326335907,
+      "learning_rate": 6.206896551724138e-06,
+      "loss": 0.75,
+      "step": 5760
+    },
+    {
+      "epoch": 22.00214559386973,
+      "grad_norm": 0.009222766384482384,
+      "learning_rate": 6.198382290336313e-06,
+      "loss": 0.0079,
+      "step": 5770
+    },
+    {
+      "epoch": 22.002911877394634,
+      "grad_norm": 25.15428352355957,
+      "learning_rate": 6.189868028948489e-06,
+      "loss": 0.2564,
+      "step": 5780
+    },
+    {
+      "epoch": 22.00367816091954,
+      "grad_norm": 0.09567564725875854,
+      "learning_rate": 6.181353767560665e-06,
+      "loss": 0.9887,
+      "step": 5790
+    },
+    {
+      "epoch": 22.004444444444445,
+      "grad_norm": 0.02086619660258293,
+      "learning_rate": 6.17283950617284e-06,
+      "loss": 0.5132,
+      "step": 5800
+    },
+    {
+      "epoch": 22.00521072796935,
+      "grad_norm": 0.1771589070558548,
+      "learning_rate": 6.164325244785016e-06,
+      "loss": 0.623,
+      "step": 5810
+    },
+    {
+      "epoch": 22.005977011494252,
+      "grad_norm": 0.08775811642408371,
+      "learning_rate": 6.155810983397192e-06,
+      "loss": 0.0024,
+      "step": 5820
+    },
+    {
+      "epoch": 22.006743295019156,
+      "grad_norm": 0.9292959570884705,
+      "learning_rate": 6.147296722009366e-06,
+      "loss": 1.7215,
+      "step": 5830
+    },
+    {
+      "epoch": 22.00750957854406,
+      "grad_norm": 0.1804896742105484,
+      "learning_rate": 6.138782460621542e-06,
+      "loss": 1.1822,
+      "step": 5840
+    },
+    {
+      "epoch": 22.008275862068967,
+      "grad_norm": 0.0041860733181238174,
+      "learning_rate": 6.130268199233717e-06,
+      "loss": 1.1366,
+      "step": 5850
+    },
+    {
+      "epoch": 22.00904214559387,
+      "grad_norm": 0.029360786080360413,
+      "learning_rate": 6.1217539378458925e-06,
+      "loss": 0.8683,
+      "step": 5860
+    },
+    {
+      "epoch": 22.009808429118774,
+      "grad_norm": 0.3557896614074707,
+      "learning_rate": 6.113239676458068e-06,
+      "loss": 0.8134,
+      "step": 5870
+    },
+    {
+      "epoch": 22.010574712643677,
+      "grad_norm": 0.0051303138025105,
+      "learning_rate": 6.1047254150702425e-06,
+      "loss": 0.4578,
+      "step": 5880
+    },
+    {
+      "epoch": 22.01134099616858,
+      "grad_norm": 0.017394129186868668,
+      "learning_rate": 6.096211153682418e-06,
+      "loss": 0.9201,
+      "step": 5890
+    },
+    {
+      "epoch": 22.01210727969349,
+      "grad_norm": 0.4005459249019623,
+      "learning_rate": 6.087696892294594e-06,
+      "loss": 0.6393,
+      "step": 5900
+    },
+    {
+      "epoch": 22.012873563218392,
+      "grad_norm": 0.012193914502859116,
+      "learning_rate": 6.079182630906769e-06,
+      "loss": 0.8868,
+      "step": 5910
+    },
+    {
+      "epoch": 22.013639846743295,
+      "grad_norm": 0.42486461997032166,
+      "learning_rate": 6.070668369518945e-06,
+      "loss": 0.0113,
+      "step": 5920
+    },
+    {
+      "epoch": 22.0144061302682,
+      "grad_norm": 0.20244956016540527,
+      "learning_rate": 6.06215410813112e-06,
+      "loss": 0.0046,
+      "step": 5930
+    },
+    {
+      "epoch": 22.015172413793103,
+      "grad_norm": 0.005920723546296358,
+      "learning_rate": 6.053639846743296e-06,
+      "loss": 0.2765,
+      "step": 5940
+    },
+    {
+      "epoch": 22.015938697318006,
+      "grad_norm": 0.14310921728610992,
+      "learning_rate": 6.045125585355472e-06,
+      "loss": 1.2898,
+      "step": 5950
+    },
+    {
+      "epoch": 22.016704980842913,
+      "grad_norm": 0.08543712645769119,
+      "learning_rate": 6.036611323967646e-06,
+      "loss": 0.4782,
+      "step": 5960
+    },
+    {
+      "epoch": 22.017471264367817,
+      "grad_norm": 0.10066840052604675,
+      "learning_rate": 6.028097062579822e-06,
+      "loss": 0.2598,
+      "step": 5970
+    },
+    {
+      "epoch": 22.01823754789272,
+      "grad_norm": 0.3465437591075897,
+      "learning_rate": 6.019582801191997e-06,
+      "loss": 0.6391,
+      "step": 5980
+    },
+    {
+      "epoch": 22.019003831417624,
+      "grad_norm": 0.9225018620491028,
+      "learning_rate": 6.0110685398041725e-06,
+      "loss": 1.2283,
+      "step": 5990
+    },
+    {
+      "epoch": 22.019770114942528,
+      "grad_norm": 219.1271514892578,
+      "learning_rate": 6.002554278416348e-06,
+      "loss": 1.3534,
+      "step": 6000
+    },
+    {
+      "epoch": 22.02,
+      "eval_accuracy": 0.7111111111111111,
+      "eval_loss": 1.8028539419174194,
+      "eval_runtime": 15.1916,
+      "eval_samples_per_second": 2.962,
+      "eval_steps_per_second": 2.962,
+      "step": 6003
+    },
+    {
+      "epoch": 23.000536398467432,
+      "grad_norm": 289.6174011230469,
+      "learning_rate": 5.9940400170285225e-06,
+      "loss": 1.1603,
+      "step": 6010
+    },
+    {
+      "epoch": 23.001302681992335,
+      "grad_norm": 0.2584246098995209,
+      "learning_rate": 5.985525755640698e-06,
+      "loss": 2.4426,
+      "step": 6020
+    },
+    {
+      "epoch": 23.002068965517243,
+      "grad_norm": 217.7529754638672,
+      "learning_rate": 5.977011494252874e-06,
+      "loss": 0.1134,
+      "step": 6030
+    },
+    {
+      "epoch": 23.002835249042146,
+      "grad_norm": 0.3803919851779938,
+      "learning_rate": 5.968497232865049e-06,
+      "loss": 0.2302,
+      "step": 6040
+    },
+    {
+      "epoch": 23.00360153256705,
+      "grad_norm": 0.02779291570186615,
+      "learning_rate": 5.959982971477225e-06,
+      "loss": 0.5589,
+      "step": 6050
+    },
+    {
+      "epoch": 23.004367816091953,
+      "grad_norm": 0.044463906437158585,
+      "learning_rate": 5.9514687100894e-06,
+      "loss": 0.4939,
+      "step": 6060
+    },
+    {
+      "epoch": 23.005134099616857,
+      "grad_norm": 0.021367818117141724,
+      "learning_rate": 5.942954448701576e-06,
+      "loss": 1.4178,
+      "step": 6070
+    },
+    {
+      "epoch": 23.005900383141764,
+      "grad_norm": 0.23603151738643646,
+      "learning_rate": 5.934440187313752e-06,
+      "loss": 0.6045,
+      "step": 6080
+    },
+    {
+      "epoch": 23.006666666666668,
+      "grad_norm": 0.0922141820192337,
+      "learning_rate": 5.925925925925926e-06,
+      "loss": 0.7727,
+      "step": 6090
+    },
+    {
+      "epoch": 23.00743295019157,
+      "grad_norm": 6.932680606842041,
+      "learning_rate": 5.917411664538102e-06,
+      "loss": 0.0075,
+      "step": 6100
+    },
+    {
+      "epoch": 23.008199233716475,
+      "grad_norm": 0.014019292779266834,
+      "learning_rate": 5.9088974031502775e-06,
+      "loss": 0.8208,
+      "step": 6110
+    },
+    {
+      "epoch": 23.00896551724138,
+      "grad_norm": 275.9669494628906,
+      "learning_rate": 5.9003831417624525e-06,
+      "loss": 0.9227,
+      "step": 6120
+    },
+    {
+      "epoch": 23.009731800766282,
+      "grad_norm": 0.7556512951850891,
+      "learning_rate": 5.891868880374628e-06,
+      "loss": 0.4839,
+      "step": 6130
+    },
+    {
+      "epoch": 23.01049808429119,
+      "grad_norm": 0.13979677855968475,
+      "learning_rate": 5.883354618986803e-06,
+      "loss": 0.0037,
+      "step": 6140
+    },
+    {
+      "epoch": 23.011264367816093,
+      "grad_norm": 71.52664184570312,
+      "learning_rate": 5.874840357598979e-06,
+      "loss": 0.708,
+      "step": 6150
+    },
+    {
+      "epoch": 23.012030651340996,
+      "grad_norm": 0.011060028336942196,
+      "learning_rate": 5.866326096211154e-06,
+      "loss": 0.5177,
+      "step": 6160
+    },
+    {
+      "epoch": 23.0127969348659,
+      "grad_norm": 53.9853401184082,
+      "learning_rate": 5.857811834823329e-06,
+      "loss": 1.2794,
+      "step": 6170
+    },
+    {
+      "epoch": 23.013563218390804,
+      "grad_norm": 0.004570251330733299,
+      "learning_rate": 5.849297573435505e-06,
+      "loss": 0.5029,
+      "step": 6180
+    },
+    {
+      "epoch": 23.014329501915707,
+      "grad_norm": 0.02976115606725216,
+      "learning_rate": 5.84078331204768e-06,
+      "loss": 1.0682,
+      "step": 6190
+    },
+    {
+      "epoch": 23.015095785440614,
+      "grad_norm": 0.003304552286863327,
+      "learning_rate": 5.832269050659856e-06,
+      "loss": 0.0027,
+      "step": 6200
+    },
+    {
+      "epoch": 23.015862068965518,
+      "grad_norm": 0.004245653282850981,
+      "learning_rate": 5.823754789272032e-06,
+      "loss": 0.5402,
+      "step": 6210
+    },
+    {
+      "epoch": 23.01662835249042,
+      "grad_norm": 0.0026863988023251295,
+      "learning_rate": 5.815240527884206e-06,
+      "loss": 0.1568,
+      "step": 6220
+    },
+    {
+      "epoch": 23.017394636015325,
+      "grad_norm": 0.047919392585754395,
+      "learning_rate": 5.806726266496382e-06,
+      "loss": 0.0018,
+      "step": 6230
+    },
+    {
+      "epoch": 23.01816091954023,
+      "grad_norm": 0.13500118255615234,
+      "learning_rate": 5.7982120051085575e-06,
+      "loss": 1.5147,
+      "step": 6240
+    },
+    {
+      "epoch": 23.018927203065132,
+      "grad_norm": 664.8700561523438,
+      "learning_rate": 5.7896977437207325e-06,
+      "loss": 0.5652,
+      "step": 6250
+    },
+    {
+      "epoch": 23.01969348659004,
+      "grad_norm": 0.010892790742218494,
+      "learning_rate": 5.781183482332908e-06,
+      "loss": 0.0423,
+      "step": 6260
+    },
+    {
+      "epoch": 23.02,
+      "eval_accuracy": 0.7111111111111111,
+      "eval_loss": 1.457066297531128,
+      "eval_runtime": 16.4944,
+      "eval_samples_per_second": 2.728,
+      "eval_steps_per_second": 2.728,
+      "step": 6264
+    },
+    {
+      "epoch": 24.000459770114944,
+      "grad_norm": 54.31410598754883,
+      "learning_rate": 5.772669220945083e-06,
+      "loss": 0.7064,
+      "step": 6270
+    },
+    {
+      "epoch": 24.001226053639847,
+      "grad_norm": 755.3297119140625,
+      "learning_rate": 5.764154959557259e-06,
+      "loss": 1.8701,
+      "step": 6280
+    },
+    {
+      "epoch": 24.00199233716475,
+      "grad_norm": 0.004258429165929556,
+      "learning_rate": 5.755640698169435e-06,
+      "loss": 1.2701,
+      "step": 6290
+    },
+    {
+      "epoch": 24.002758620689654,
+      "grad_norm": 0.0329936183989048,
+      "learning_rate": 5.747126436781609e-06,
+      "loss": 2.0358,
+      "step": 6300
+    },
+    {
+      "epoch": 24.003524904214558,
+      "grad_norm": 0.7360894680023193,
+      "learning_rate": 5.738612175393785e-06,
+      "loss": 0.4042,
+      "step": 6310
+    },
+    {
+      "epoch": 24.004291187739465,
+      "grad_norm": 0.011302406899631023,
+      "learning_rate": 5.730097914005961e-06,
+      "loss": 1.0555,
+      "step": 6320
+    },
+    {
+      "epoch": 24.00505747126437,
+      "grad_norm": 5.05350399017334,
+      "learning_rate": 5.721583652618136e-06,
+      "loss": 0.7022,
+      "step": 6330
+    },
+    {
+      "epoch": 24.005823754789272,
+      "grad_norm": 557.7550048828125,
+      "learning_rate": 5.713069391230312e-06,
+      "loss": 0.6111,
+      "step": 6340
+    },
+    {
+      "epoch": 24.006590038314176,
+      "grad_norm": 2.5530898571014404,
+      "learning_rate": 5.704555129842486e-06,
+      "loss": 0.4165,
+      "step": 6350
+    },
+    {
+      "epoch": 24.00735632183908,
+      "grad_norm": 0.43523475527763367,
+      "learning_rate": 5.696040868454662e-06,
+      "loss": 0.7824,
+      "step": 6360
+    },
+    {
+      "epoch": 24.008122605363983,
+      "grad_norm": 49.91386032104492,
+      "learning_rate": 5.6875266070668375e-06,
+      "loss": 2.0053,
+      "step": 6370
+    },
+    {
+      "epoch": 24.00888888888889,
+      "grad_norm": 0.011766809970140457,
+      "learning_rate": 5.6790123456790125e-06,
+      "loss": 0.4156,
+      "step": 6380
+    },
+    {
+      "epoch": 24.009655172413794,
+      "grad_norm": 0.0064444891177117825,
+      "learning_rate": 5.670498084291188e-06,
+      "loss": 1.0425,
+      "step": 6390
+    },
+    {
+      "epoch": 24.010421455938697,
+      "grad_norm": 365.0169372558594,
+      "learning_rate": 5.661983822903364e-06,
+      "loss": 1.0558,
+      "step": 6400
+    },
+    {
+      "epoch": 24.0111877394636,
+      "grad_norm": 1.042160153388977,
+      "learning_rate": 5.653469561515539e-06,
+      "loss": 0.3502,
+      "step": 6410
+    },
+    {
+      "epoch": 24.011954022988505,
+      "grad_norm": 135.4849395751953,
+      "learning_rate": 5.644955300127715e-06,
+      "loss": 1.4494,
+      "step": 6420
+    },
+    {
+      "epoch": 24.01272030651341,
+      "grad_norm": 0.0020956031512469053,
+      "learning_rate": 5.636441038739889e-06,
+      "loss": 0.9078,
+      "step": 6430
+    },
+    {
+      "epoch": 24.013486590038315,
+      "grad_norm": 0.004369780886918306,
+      "learning_rate": 5.627926777352065e-06,
+      "loss": 0.0083,
+      "step": 6440
+    },
+    {
+      "epoch": 24.01425287356322,
+      "grad_norm": 0.1496662199497223,
+      "learning_rate": 5.619412515964241e-06,
+      "loss": 0.4498,
+      "step": 6450
+    },
+    {
+      "epoch": 24.015019157088123,
+      "grad_norm": 0.19664430618286133,
+      "learning_rate": 5.610898254576416e-06,
+      "loss": 0.4715,
+      "step": 6460
+    },
+    {
+      "epoch": 24.015785440613026,
+      "grad_norm": 2.650960683822632,
+      "learning_rate": 5.602383993188592e-06,
+      "loss": 0.5836,
+      "step": 6470
+    },
+    {
+      "epoch": 24.01655172413793,
+      "grad_norm": 0.002323655877262354,
+      "learning_rate": 5.593869731800766e-06,
+      "loss": 0.346,
+      "step": 6480
+    },
+    {
+      "epoch": 24.017318007662837,
+      "grad_norm": 0.03131331130862236,
+      "learning_rate": 5.585355470412942e-06,
+      "loss": 1.4927,
+      "step": 6490
+    },
+    {
+      "epoch": 24.01808429118774,
+      "grad_norm": 0.0025550054851919413,
+      "learning_rate": 5.5768412090251175e-06,
+      "loss": 1.1915,
+      "step": 6500
+    },
+    {
+      "epoch": 24.018850574712644,
+      "grad_norm": 811.5127563476562,
+      "learning_rate": 5.5683269476372925e-06,
+      "loss": 0.3877,
+      "step": 6510
+    },
+    {
+      "epoch": 24.019616858237548,
+      "grad_norm": 0.04266892001032829,
+      "learning_rate": 5.559812686249468e-06,
+      "loss": 1.0068,
+      "step": 6520
+    },
+    {
+      "epoch": 24.02,
+      "eval_accuracy": 0.6444444444444445,
+      "eval_loss": 1.7790253162384033,
+      "eval_runtime": 14.3666,
+      "eval_samples_per_second": 3.132,
+      "eval_steps_per_second": 3.132,
+      "step": 6525
+    },
+    {
+      "epoch": 25.000383141762452,
+      "grad_norm": 0.18075942993164062,
+      "learning_rate": 5.551298424861644e-06,
+      "loss": 0.0016,
+      "step": 6530
+    },
+    {
+      "epoch": 25.001149425287355,
+      "grad_norm": 0.08188223838806152,
+      "learning_rate": 5.542784163473819e-06,
+      "loss": 1.1096,
+      "step": 6540
+    },
+    {
+      "epoch": 25.00191570881226,
+      "grad_norm": 0.08906185626983643,
+      "learning_rate": 5.534269902085995e-06,
+      "loss": 1.2369,
+      "step": 6550
+    },
+    {
+      "epoch": 25.002681992337166,
+      "grad_norm": 90.94135284423828,
+      "learning_rate": 5.525755640698169e-06,
+      "loss": 1.1431,
+      "step": 6560
+    },
+    {
+      "epoch": 25.00344827586207,
+      "grad_norm": 0.11608108878135681,
+      "learning_rate": 5.517241379310345e-06,
+      "loss": 0.3466,
+      "step": 6570
+    },
+    {
+      "epoch": 25.004214559386973,
+      "grad_norm": 1376.1837158203125,
+      "learning_rate": 5.508727117922521e-06,
+      "loss": 0.1968,
+      "step": 6580
+    },
+    {
+      "epoch": 25.004980842911877,
+      "grad_norm": 0.018664706498384476,
+      "learning_rate": 5.500212856534696e-06,
+      "loss": 0.441,
+      "step": 6590
+    },
+    {
+      "epoch": 25.00574712643678,
+      "grad_norm": 0.007804445456713438,
+      "learning_rate": 5.491698595146872e-06,
+      "loss": 0.5534,
+      "step": 6600
+    },
+    {
+      "epoch": 25.006513409961684,
+      "grad_norm": 0.017437539994716644,
+      "learning_rate": 5.4831843337590475e-06,
+      "loss": 0.0191,
+      "step": 6610
+    },
+    {
+      "epoch": 25.00727969348659,
+      "grad_norm": 5.701467990875244,
+      "learning_rate": 5.474670072371222e-06,
+      "loss": 1.3893,
+      "step": 6620
+    },
+    {
+      "epoch": 25.008045977011495,
+      "grad_norm": 0.06029016524553299,
+      "learning_rate": 5.4661558109833975e-06,
+      "loss": 0.0441,
+      "step": 6630
+    },
+    {
+      "epoch": 25.0088122605364,
+      "grad_norm": 0.007205882575362921,
+      "learning_rate": 5.4576415495955725e-06,
+      "loss": 0.3946,
+      "step": 6640
+    },
+    {
+      "epoch": 25.009578544061302,
+      "grad_norm": 0.004521137103438377,
+      "learning_rate": 5.449127288207748e-06,
+      "loss": 0.5575,
+      "step": 6650
+    },
+    {
+      "epoch": 25.010344827586206,
+      "grad_norm": 63.423763275146484,
+      "learning_rate": 5.440613026819924e-06,
+      "loss": 0.7411,
+      "step": 6660
+    },
+    {
+      "epoch": 25.011111111111113,
+      "grad_norm": 0.05708396062254906,
+      "learning_rate": 5.432098765432099e-06,
+      "loss": 0.6057,
+      "step": 6670
+    },
+    {
+      "epoch": 25.011877394636016,
+      "grad_norm": 0.01645585335791111,
+      "learning_rate": 5.423584504044275e-06,
+      "loss": 1.1414,
+      "step": 6680
+    },
+    {
+      "epoch": 25.01264367816092,
+      "grad_norm": 0.03839023783802986,
+      "learning_rate": 5.415070242656451e-06,
+      "loss": 1.191,
+      "step": 6690
+    },
+    {
+      "epoch": 25.013409961685824,
+      "grad_norm": 0.14333435893058777,
+      "learning_rate": 5.406555981268625e-06,
+      "loss": 0.444,
+      "step": 6700
+    },
+    {
+      "epoch": 25.014176245210727,
+      "grad_norm": 0.006897382903844118,
+      "learning_rate": 5.398041719880801e-06,
+      "loss": 0.5621,
+      "step": 6710
+    },
+    {
+      "epoch": 25.01494252873563,
+      "grad_norm": 0.0030311744194477797,
+      "learning_rate": 5.389527458492976e-06,
+      "loss": 0.6271,
+      "step": 6720
+    },
+    {
+      "epoch": 25.015708812260538,
+      "grad_norm": 0.03591395169496536,
+      "learning_rate": 5.381013197105152e-06,
+      "loss": 0.4289,
+      "step": 6730
+    },
+    {
+      "epoch": 25.01647509578544,
+      "grad_norm": 130.26043701171875,
+      "learning_rate": 5.3724989357173275e-06,
+      "loss": 1.6805,
+      "step": 6740
+    },
+    {
+      "epoch": 25.017241379310345,
+      "grad_norm": 0.023178286850452423,
+      "learning_rate": 5.3639846743295025e-06,
+      "loss": 1.0406,
+      "step": 6750
+    },
+    {
+      "epoch": 25.01800766283525,
+      "grad_norm": 0.0045823934487998486,
+      "learning_rate": 5.3554704129416775e-06,
+      "loss": 0.3643,
+      "step": 6760
+    },
+    {
+      "epoch": 25.018773946360152,
+      "grad_norm": 1255.3531494140625,
+      "learning_rate": 5.3469561515538525e-06,
+      "loss": 0.3438,
+      "step": 6770
+    },
+    {
+      "epoch": 25.019540229885056,
+      "grad_norm": 0.820978581905365,
+      "learning_rate": 5.338441890166028e-06,
+      "loss": 1.5772,
+      "step": 6780
+    },
+    {
+      "epoch": 25.02,
+      "eval_accuracy": 0.6666666666666666,
+      "eval_loss": 1.7893223762512207,
+      "eval_runtime": 15.2457,
+      "eval_samples_per_second": 2.952,
+      "eval_steps_per_second": 2.952,
+      "step": 6786
+    },
+    {
+      "epoch": 26.00030651340996,
+      "grad_norm": 0.3054686188697815,
+      "learning_rate": 5.329927628778204e-06,
+      "loss": 0.7461,
+      "step": 6790
+    },
+    {
+      "epoch": 26.001072796934867,
+      "grad_norm": 0.04109475016593933,
+      "learning_rate": 5.321413367390379e-06,
+      "loss": 0.0492,
+      "step": 6800
+    },
+    {
+      "epoch": 26.00183908045977,
+      "grad_norm": 46.89008712768555,
+      "learning_rate": 5.312899106002555e-06,
+      "loss": 0.6362,
+      "step": 6810
+    },
+    {
+      "epoch": 26.002605363984674,
+      "grad_norm": 0.004781241994351149,
+      "learning_rate": 5.304384844614731e-06,
+      "loss": 1.1682,
+      "step": 6820
+    },
+    {
+      "epoch": 26.003371647509578,
+      "grad_norm": 0.21307355165481567,
+      "learning_rate": 5.295870583226905e-06,
+      "loss": 0.5552,
+      "step": 6830
+    },
+    {
+      "epoch": 26.00413793103448,
+      "grad_norm": 0.026560096070170403,
+      "learning_rate": 5.287356321839081e-06,
+      "loss": 0.7666,
+      "step": 6840
+    },
+    {
+      "epoch": 26.00490421455939,
+      "grad_norm": 0.003154867561534047,
+      "learning_rate": 5.278842060451256e-06,
+      "loss": 0.4521,
+      "step": 6850
+    },
+    {
+      "epoch": 26.005670498084292,
+      "grad_norm": 0.25655969977378845,
+      "learning_rate": 5.270327799063432e-06,
+      "loss": 0.4738,
+      "step": 6860
+    },
+    {
+      "epoch": 26.006436781609196,
+      "grad_norm": 0.053296688944101334,
+      "learning_rate": 5.2618135376756075e-06,
+      "loss": 0.0045,
+      "step": 6870
+    },
+    {
+      "epoch": 26.0072030651341,
+      "grad_norm": 0.027281196787953377,
+      "learning_rate": 5.2532992762877825e-06,
+      "loss": 0.6215,
+      "step": 6880
+    },
+    {
+      "epoch": 26.007969348659003,
+      "grad_norm": 0.02793821506202221,
+      "learning_rate": 5.244785014899958e-06,
+      "loss": 0.0015,
+      "step": 6890
+    },
+    {
+      "epoch": 26.008735632183907,
+      "grad_norm": 0.005113948602229357,
+      "learning_rate": 5.236270753512134e-06,
+      "loss": 0.0007,
+      "step": 6900
+    },
+    {
+      "epoch": 26.009501915708814,
+      "grad_norm": 0.1896917223930359,
+      "learning_rate": 5.227756492124308e-06,
+      "loss": 0.7445,
+      "step": 6910
+    },
+    {
+      "epoch": 26.010268199233717,
+      "grad_norm": 0.02426121011376381,
+      "learning_rate": 5.219242230736484e-06,
+      "loss": 0.9747,
+      "step": 6920
+    },
+    {
+      "epoch": 26.01103448275862,
+      "grad_norm": 0.15907245874404907,
+      "learning_rate": 5.210727969348659e-06,
+      "loss": 0.1155,
+      "step": 6930
+    },
+    {
+      "epoch": 26.011800766283525,
+      "grad_norm": 0.09672944247722626,
+      "learning_rate": 5.202213707960835e-06,
+      "loss": 0.0018,
+      "step": 6940
+    },
+    {
+      "epoch": 26.01256704980843,
+      "grad_norm": 0.00523536279797554,
+      "learning_rate": 5.193699446573011e-06,
+      "loss": 0.1146,
+      "step": 6950
+    },
+    {
+      "epoch": 26.013333333333332,
+      "grad_norm": 21.234874725341797,
+      "learning_rate": 5.185185185185185e-06,
+      "loss": 1.1605,
+      "step": 6960
+    },
+    {
+      "epoch": 26.01409961685824,
+      "grad_norm": 0.0017688468797132373,
+      "learning_rate": 5.176670923797361e-06,
+      "loss": 0.0008,
+      "step": 6970
+    },
+    {
+      "epoch": 26.014865900383143,
+      "grad_norm": 31.0344181060791,
+      "learning_rate": 5.168156662409536e-06,
+      "loss": 3.1422,
+      "step": 6980
+    },
+    {
+      "epoch": 26.015632183908046,
+      "grad_norm": 19.070343017578125,
+      "learning_rate": 5.159642401021712e-06,
+      "loss": 1.0949,
+      "step": 6990
+    },
+    {
+      "epoch": 26.01639846743295,
+      "grad_norm": 0.05039878934621811,
+      "learning_rate": 5.1511281396338875e-06,
+      "loss": 1.7007,
+      "step": 7000
+    },
+    {
+      "epoch": 26.017164750957853,
+      "grad_norm": 0.08749368041753769,
+      "learning_rate": 5.1426138782460625e-06,
+      "loss": 0.8494,
+      "step": 7010
+    },
+    {
+      "epoch": 26.017931034482757,
+      "grad_norm": 0.12880153954029083,
+      "learning_rate": 5.134099616858238e-06,
+      "loss": 1.417,
+      "step": 7020
+    },
+    {
+      "epoch": 26.018697318007664,
+      "grad_norm": 7.412830829620361,
+      "learning_rate": 5.125585355470414e-06,
+      "loss": 0.0061,
+      "step": 7030
+    },
+    {
+      "epoch": 26.019463601532568,
+      "grad_norm": 0.015643224120140076,
+      "learning_rate": 5.117071094082588e-06,
+      "loss": 0.8409,
+      "step": 7040
+    },
+    {
+      "epoch": 26.02,
+      "eval_accuracy": 0.6666666666666666,
+      "eval_loss": 1.6453967094421387,
+      "eval_runtime": 14.268,
+      "eval_samples_per_second": 3.154,
+      "eval_steps_per_second": 3.154,
+      "step": 7047
+    },
+    {
+      "epoch": 27.000229885057472,
+      "grad_norm": 19.965852737426758,
+      "learning_rate": 5.108556832694764e-06,
+      "loss": 0.658,
+      "step": 7050
+    },
+    {
+      "epoch": 27.000996168582375,
+      "grad_norm": 2.890854597091675,
+      "learning_rate": 5.100042571306939e-06,
+      "loss": 0.5235,
+      "step": 7060
+    },
+    {
+      "epoch": 27.00176245210728,
+      "grad_norm": 0.012044958770275116,
+      "learning_rate": 5.091528309919115e-06,
+      "loss": 0.006,
+      "step": 7070
+    },
+    {
+      "epoch": 27.002528735632183,
+      "grad_norm": 4.6264967918396,
+      "learning_rate": 5.083014048531291e-06,
+      "loss": 0.0541,
+      "step": 7080
+    },
+    {
+      "epoch": 27.00329501915709,
+      "grad_norm": 0.008120655082166195,
+      "learning_rate": 5.074499787143465e-06,
+      "loss": 1.1141,
+      "step": 7090
+    },
+    {
+      "epoch": 27.004061302681993,
+      "grad_norm": 0.02046016789972782,
+      "learning_rate": 5.065985525755641e-06,
+      "loss": 0.0064,
+      "step": 7100
+    },
+    {
+      "epoch": 27.004827586206897,
+      "grad_norm": 0.04679565131664276,
+      "learning_rate": 5.057471264367817e-06,
+      "loss": 0.4713,
+      "step": 7110
+    },
+    {
+      "epoch": 27.0055938697318,
+      "grad_norm": 0.003998655825853348,
+      "learning_rate": 5.048957002979992e-06,
+      "loss": 0.8596,
+      "step": 7120
+    },
+    {
+      "epoch": 27.006360153256704,
+      "grad_norm": 0.0444590263068676,
+      "learning_rate": 5.0404427415921675e-06,
+      "loss": 1.0469,
+      "step": 7130
+    },
+    {
+      "epoch": 27.007126436781608,
+      "grad_norm": 0.021894382312893867,
+      "learning_rate": 5.0319284802043425e-06,
+      "loss": 0.0378,
+      "step": 7140
+    },
+    {
+      "epoch": 27.007892720306515,
+      "grad_norm": 57.658607482910156,
+      "learning_rate": 5.023414218816518e-06,
+      "loss": 2.1019,
+      "step": 7150
+    },
+    {
+      "epoch": 27.00865900383142,
+      "grad_norm": 0.034019336104393005,
+      "learning_rate": 5.014899957428694e-06,
+      "loss": 0.4746,
+      "step": 7160
+    },
+    {
+      "epoch": 27.009425287356322,
+      "grad_norm": 0.005097901914268732,
+      "learning_rate": 5.006385696040868e-06,
+      "loss": 0.0269,
+      "step": 7170
+    },
+    {
+      "epoch": 27.010191570881226,
+      "grad_norm": 0.5953840017318726,
+      "learning_rate": 4.997871434653044e-06,
+      "loss": 1.6232,
+      "step": 7180
+    },
+    {
+      "epoch": 27.01095785440613,
+      "grad_norm": 0.003489244729280472,
+      "learning_rate": 4.98935717326522e-06,
+      "loss": 0.7787,
+      "step": 7190
+    },
+    {
+      "epoch": 27.011724137931033,
+      "grad_norm": 809.7433471679688,
+      "learning_rate": 4.980842911877395e-06,
+      "loss": 1.401,
+      "step": 7200
+    },
+    {
+      "epoch": 27.01249042145594,
+      "grad_norm": 0.12849143147468567,
+      "learning_rate": 4.972328650489571e-06,
+      "loss": 0.4929,
+      "step": 7210
+    },
+    {
+      "epoch": 27.013256704980844,
+      "grad_norm": 0.3912018835544586,
+      "learning_rate": 4.963814389101746e-06,
+      "loss": 0.0422,
+      "step": 7220
+    },
+    {
+      "epoch": 27.014022988505747,
+      "grad_norm": 0.024638323113322258,
+      "learning_rate": 4.955300127713921e-06,
+      "loss": 1.1046,
+      "step": 7230
+    },
+    {
+      "epoch": 27.01478927203065,
+      "grad_norm": 2.2969040870666504,
+      "learning_rate": 4.946785866326097e-06,
+      "loss": 1.0911,
+      "step": 7240
+    },
+    {
+      "epoch": 27.015555555555554,
+      "grad_norm": 0.24452875554561615,
+      "learning_rate": 4.938271604938272e-06,
+      "loss": 0.5244,
+      "step": 7250
+    },
+    {
+      "epoch": 27.016321839080458,
+      "grad_norm": 0.004459705203771591,
+      "learning_rate": 4.9297573435504475e-06,
+      "loss": 0.4651,
+      "step": 7260
+    },
+    {
+      "epoch": 27.017088122605365,
+      "grad_norm": 0.05164725333452225,
+      "learning_rate": 4.9212430821626225e-06,
+      "loss": 0.0037,
+      "step": 7270
+    },
+    {
+      "epoch": 27.01785440613027,
+      "grad_norm": 0.25497305393218994,
+      "learning_rate": 4.912728820774798e-06,
+      "loss": 0.5484,
+      "step": 7280
+    },
+    {
+      "epoch": 27.018620689655172,
+      "grad_norm": 116.34848022460938,
+      "learning_rate": 4.904214559386973e-06,
+      "loss": 0.0093,
+      "step": 7290
+    },
+    {
+      "epoch": 27.019386973180076,
+      "grad_norm": 0.22834795713424683,
+      "learning_rate": 4.895700297999149e-06,
+      "loss": 0.6828,
+      "step": 7300
+    },
+    {
+      "epoch": 27.02,
+      "eval_accuracy": 0.6888888888888889,
+      "eval_loss": 1.852113962173462,
+      "eval_runtime": 15.2767,
+      "eval_samples_per_second": 2.946,
+      "eval_steps_per_second": 2.946,
+      "step": 7308
+    },
+    {
+      "epoch": 28.00015325670498,
+      "grad_norm": 0.01499945018440485,
+      "learning_rate": 4.887186036611324e-06,
+      "loss": 1.3009,
+      "step": 7310
+    },
+    {
+      "epoch": 28.000919540229884,
+      "grad_norm": 20.283266067504883,
+      "learning_rate": 4.8786717752235e-06,
+      "loss": 1.0598,
+      "step": 7320
+    },
+    {
+      "epoch": 28.00168582375479,
+      "grad_norm": 680.4053344726562,
+      "learning_rate": 4.870157513835675e-06,
+      "loss": 0.9413,
+      "step": 7330
+    },
+    {
+      "epoch": 28.002452107279694,
+      "grad_norm": 0.0037998862098902464,
+      "learning_rate": 4.861643252447851e-06,
+      "loss": 1.6265,
+      "step": 7340
+    },
+    {
+      "epoch": 28.003218390804598,
+      "grad_norm": 0.028214074671268463,
+      "learning_rate": 4.853128991060026e-06,
+      "loss": 0.2507,
+      "step": 7350
+    },
+    {
+      "epoch": 28.0039846743295,
+      "grad_norm": 0.27702319622039795,
+      "learning_rate": 4.844614729672202e-06,
+      "loss": 0.7148,
+      "step": 7360
+    },
+    {
+      "epoch": 28.004750957854405,
+      "grad_norm": 88.49553680419922,
+      "learning_rate": 4.836100468284377e-06,
+      "loss": 1.1467,
+      "step": 7370
+    },
+    {
+      "epoch": 28.00551724137931,
+      "grad_norm": 195.29783630371094,
+      "learning_rate": 4.8275862068965525e-06,
+      "loss": 1.0426,
+      "step": 7380
+    },
+    {
+      "epoch": 28.006283524904216,
+      "grad_norm": 0.013184229843318462,
+      "learning_rate": 4.8190719455087275e-06,
+      "loss": 0.6131,
+      "step": 7390
+    },
+    {
+      "epoch": 28.00704980842912,
+      "grad_norm": 824.2332763671875,
+      "learning_rate": 4.8105576841209025e-06,
+      "loss": 0.5226,
+      "step": 7400
+    },
+    {
+      "epoch": 28.007816091954023,
+      "grad_norm": 0.014120840467512608,
+      "learning_rate": 4.802043422733078e-06,
+      "loss": 0.6207,
+      "step": 7410
+    },
+    {
+      "epoch": 28.008582375478927,
+      "grad_norm": 0.0181321669369936,
+      "learning_rate": 4.793529161345254e-06,
+      "loss": 0.5579,
+      "step": 7420
+    },
+    {
+      "epoch": 28.00934865900383,
+      "grad_norm": 0.033630553632974625,
+      "learning_rate": 4.785014899957429e-06,
+      "loss": 0.2249,
+      "step": 7430
+    },
+    {
+      "epoch": 28.010114942528734,
+      "grad_norm": 0.004489063750952482,
+      "learning_rate": 4.776500638569604e-06,
+      "loss": 0.4761,
+      "step": 7440
+    },
+    {
+      "epoch": 28.01088122605364,
+      "grad_norm": 0.01676010712981224,
+      "learning_rate": 4.76798637718178e-06,
+      "loss": 0.0105,
+      "step": 7450
+    },
+    {
+      "epoch": 28.011647509578545,
+      "grad_norm": 0.0805138424038887,
+      "learning_rate": 4.759472115793956e-06,
+      "loss": 0.786,
+      "step": 7460
+    },
+    {
+      "epoch": 28.01241379310345,
+      "grad_norm": 0.4894491136074066,
+      "learning_rate": 4.750957854406131e-06,
+      "loss": 0.4835,
+      "step": 7470
+    },
+    {
+      "epoch": 28.013180076628352,
+      "grad_norm": 0.00567967863753438,
+      "learning_rate": 4.742443593018306e-06,
+      "loss": 0.92,
+      "step": 7480
+    },
+    {
+      "epoch": 28.013946360153255,
+      "grad_norm": 0.13989655673503876,
+      "learning_rate": 4.733929331630482e-06,
+      "loss": 0.0026,
+      "step": 7490
+    },
+    {
+      "epoch": 28.014712643678163,
+      "grad_norm": 0.006515056826174259,
+      "learning_rate": 4.7254150702426575e-06,
+      "loss": 0.0017,
+      "step": 7500
+    },
+    {
+      "epoch": 28.015478927203066,
+      "grad_norm": 0.013979991897940636,
+      "learning_rate": 4.7169008088548325e-06,
+      "loss": 0.8258,
+      "step": 7510
+    },
+    {
+      "epoch": 28.01624521072797,
+      "grad_norm": 0.00985270831733942,
+      "learning_rate": 4.7083865474670075e-06,
+      "loss": 1.0633,
+      "step": 7520
+    },
+    {
+      "epoch": 28.017011494252873,
+      "grad_norm": 0.5813823342323303,
+      "learning_rate": 4.6998722860791825e-06,
+      "loss": 0.882,
+      "step": 7530
+    },
+    {
+      "epoch": 28.017777777777777,
+      "grad_norm": 0.03663960471749306,
+      "learning_rate": 4.691358024691358e-06,
+      "loss": 0.9276,
+      "step": 7540
+    },
+    {
+      "epoch": 28.01854406130268,
+      "grad_norm": 0.22019168734550476,
+      "learning_rate": 4.682843763303534e-06,
+      "loss": 0.5073,
+      "step": 7550
+    },
+    {
+      "epoch": 28.019310344827588,
+      "grad_norm": 0.007242546882480383,
+      "learning_rate": 4.674329501915709e-06,
+      "loss": 0.5191,
+      "step": 7560
+    },
+    {
+      "epoch": 28.02,
+      "eval_accuracy": 0.7555555555555555,
+      "eval_loss": 1.2733657360076904,
+      "eval_runtime": 14.1698,
+      "eval_samples_per_second": 3.176,
+      "eval_steps_per_second": 3.176,
+      "step": 7569
+    },
+    {
+      "epoch": 29.000076628352492,
+      "grad_norm": 0.017437102273106575,
+      "learning_rate": 4.665815240527884e-06,
+      "loss": 0.2045,
+      "step": 7570
+    },
+    {
+      "epoch": 29.000842911877395,
+      "grad_norm": 0.7838713526725769,
+      "learning_rate": 4.65730097914006e-06,
+      "loss": 0.0042,
+      "step": 7580
+    },
+    {
+      "epoch": 29.0016091954023,
+      "grad_norm": 0.003179553197696805,
+      "learning_rate": 4.648786717752236e-06,
+      "loss": 0.0065,
+      "step": 7590
+    },
+    {
+      "epoch": 29.002375478927203,
+      "grad_norm": 0.0024605966173112392,
+      "learning_rate": 4.640272456364411e-06,
+      "loss": 0.6129,
+      "step": 7600
+    },
+    {
+      "epoch": 29.003141762452106,
+      "grad_norm": 0.1865365356206894,
+      "learning_rate": 4.631758194976586e-06,
+      "loss": 0.6739,
+      "step": 7610
+    },
+    {
+      "epoch": 29.00390804597701,
+      "grad_norm": 0.08629602938890457,
+      "learning_rate": 4.623243933588762e-06,
+      "loss": 0.0105,
+      "step": 7620
+    },
+    {
+      "epoch": 29.004674329501917,
+      "grad_norm": 0.002063839230686426,
+      "learning_rate": 4.6147296722009375e-06,
+      "loss": 0.5017,
+      "step": 7630
+    },
+    {
+      "epoch": 29.00544061302682,
+      "grad_norm": 0.19979241490364075,
+      "learning_rate": 4.6062154108131125e-06,
+      "loss": 1.6036,
+      "step": 7640
+    },
+    {
+      "epoch": 29.006206896551724,
+      "grad_norm": 0.009347641840577126,
+      "learning_rate": 4.5977011494252875e-06,
+      "loss": 0.2936,
+      "step": 7650
+    },
+    {
+      "epoch": 29.006973180076628,
+      "grad_norm": 0.43149271607398987,
+      "learning_rate": 4.589186888037463e-06,
+      "loss": 0.4749,
+      "step": 7660
+    },
+    {
+      "epoch": 29.00773946360153,
+      "grad_norm": 0.0016137725906446576,
+      "learning_rate": 4.580672626649638e-06,
+      "loss": 0.002,
+      "step": 7670
+    },
+    {
+      "epoch": 29.00850574712644,
+      "grad_norm": 0.0033685702364891768,
+      "learning_rate": 4.572158365261814e-06,
+      "loss": 0.4395,
+      "step": 7680
+    },
+    {
+      "epoch": 29.009272030651342,
+      "grad_norm": 0.01022439356893301,
+      "learning_rate": 4.563644103873989e-06,
+      "loss": 0.0013,
+      "step": 7690
+    },
+    {
+      "epoch": 29.010038314176246,
+      "grad_norm": 0.022741427645087242,
+      "learning_rate": 4.555129842486164e-06,
+      "loss": 0.0026,
+      "step": 7700
+    },
+    {
+      "epoch": 29.01080459770115,
+      "grad_norm": 0.0013971420703455806,
+      "learning_rate": 4.54661558109834e-06,
+      "loss": 1.2997,
+      "step": 7710
+    },
+    {
+      "epoch": 29.011570881226053,
+      "grad_norm": 0.0018495945259928703,
+      "learning_rate": 4.538101319710516e-06,
+      "loss": 0.0621,
+      "step": 7720
+    },
+    {
+      "epoch": 29.012337164750956,
+      "grad_norm": 8.727816581726074,
+      "learning_rate": 4.529587058322691e-06,
+      "loss": 0.6664,
+      "step": 7730
+    },
+    {
+      "epoch": 29.013103448275864,
+      "grad_norm": 0.003989522345364094,
+      "learning_rate": 4.521072796934866e-06,
+      "loss": 0.046,
+      "step": 7740
+    },
+    {
+      "epoch": 29.013869731800767,
+      "grad_norm": 0.1331661492586136,
+      "learning_rate": 4.512558535547042e-06,
+      "loss": 0.9251,
+      "step": 7750
+    },
+    {
+      "epoch": 29.01463601532567,
+      "grad_norm": 0.17615145444869995,
+      "learning_rate": 4.5040442741592175e-06,
+      "loss": 0.4809,
+      "step": 7760
+    },
+    {
+      "epoch": 29.015402298850574,
+      "grad_norm": 0.0029606192838400602,
+      "learning_rate": 4.4955300127713925e-06,
+      "loss": 0.3072,
+      "step": 7770
+    },
+    {
+      "epoch": 29.016168582375478,
+      "grad_norm": 1.035951852798462,
+      "learning_rate": 4.4870157513835675e-06,
+      "loss": 0.0023,
+      "step": 7780
+    },
+    {
+      "epoch": 29.01693486590038,
+      "grad_norm": 88.19917297363281,
+      "learning_rate": 4.478501489995743e-06,
+      "loss": 1.0988,
+      "step": 7790
+    },
+    {
+      "epoch": 29.01770114942529,
+      "grad_norm": 0.0017601007129997015,
+      "learning_rate": 4.469987228607919e-06,
+      "loss": 0.4938,
+      "step": 7800
+    },
+    {
+      "epoch": 29.018467432950192,
+      "grad_norm": 0.001365689910016954,
+      "learning_rate": 4.461472967220094e-06,
+      "loss": 1.212,
+      "step": 7810
+    },
+    {
+      "epoch": 29.019233716475096,
+      "grad_norm": 4.030818939208984,
+      "learning_rate": 4.452958705832269e-06,
+      "loss": 1.0075,
+      "step": 7820
+    },
+    {
+      "epoch": 29.02,
+      "grad_norm": 0.0018792204791679978,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.4537,
+      "step": 7830
+    },
+    {
+      "epoch": 29.02,
+      "eval_accuracy": 0.7111111111111111,
+      "eval_loss": 1.8098787069320679,
+      "eval_runtime": 16.2348,
+      "eval_samples_per_second": 2.772,
+      "eval_steps_per_second": 2.772,
+      "step": 7830
+    },
+    {
+      "epoch": 30.000766283524904,
+      "grad_norm": 0.0031461918260902166,
+      "learning_rate": 4.43593018305662e-06,
+      "loss": 0.7998,
+      "step": 7840
+    },
+    {
+      "epoch": 30.001532567049807,
+      "grad_norm": 0.21870797872543335,
+      "learning_rate": 4.427415921668796e-06,
+      "loss": 0.006,
+      "step": 7850
+    },
+    {
+      "epoch": 30.002298850574714,
+      "grad_norm": 0.04204126074910164,
+      "learning_rate": 4.418901660280971e-06,
+      "loss": 0.0008,
+      "step": 7860
+    },
+    {
+      "epoch": 30.003065134099618,
+      "grad_norm": 0.053357142955064774,
+      "learning_rate": 4.410387398893146e-06,
+      "loss": 0.6211,
+      "step": 7870
+    },
+    {
+      "epoch": 30.00383141762452,
+      "grad_norm": 0.0014875102788209915,
+      "learning_rate": 4.401873137505322e-06,
+      "loss": 0.588,
+      "step": 7880
+    },
+    {
+      "epoch": 30.004597701149425,
+      "grad_norm": 0.14551614224910736,
+      "learning_rate": 4.3933588761174975e-06,
+      "loss": 0.434,
+      "step": 7890
+    },
+    {
+      "epoch": 30.00536398467433,
+      "grad_norm": 0.003355887020006776,
+      "learning_rate": 4.3848446147296725e-06,
+      "loss": 0.4216,
+      "step": 7900
+    },
+    {
+      "epoch": 30.006130268199232,
+      "grad_norm": 235.74745178222656,
+      "learning_rate": 4.3763303533418475e-06,
+      "loss": 1.7108,
+      "step": 7910
+    },
+    {
+      "epoch": 30.00689655172414,
+      "grad_norm": 0.4113011658191681,
+      "learning_rate": 4.367816091954023e-06,
+      "loss": 0.9043,
+      "step": 7920
+    },
+    {
+      "epoch": 30.007662835249043,
+      "grad_norm": 359.8921813964844,
+      "learning_rate": 4.359301830566199e-06,
+      "loss": 0.5145,
+      "step": 7930
+    },
+    {
+      "epoch": 30.008429118773947,
+      "grad_norm": 5.985605239868164,
+      "learning_rate": 4.350787569178374e-06,
+      "loss": 0.0965,
+      "step": 7940
+    },
+    {
+      "epoch": 30.00919540229885,
+      "grad_norm": 0.07877785712480545,
+      "learning_rate": 4.342273307790549e-06,
+      "loss": 0.577,
+      "step": 7950
+    },
+    {
+      "epoch": 30.009961685823754,
+      "grad_norm": 0.01540356408804655,
+      "learning_rate": 4.333759046402725e-06,
+      "loss": 0.2528,
+      "step": 7960
+    },
+    {
+      "epoch": 30.010727969348657,
+      "grad_norm": 0.06074034050107002,
+      "learning_rate": 4.325244785014901e-06,
+      "loss": 0.0008,
+      "step": 7970
+    },
+    {
+      "epoch": 30.011494252873565,
+      "grad_norm": 0.17096516489982605,
+      "learning_rate": 4.316730523627076e-06,
+      "loss": 0.0174,
+      "step": 7980
+    },
+    {
+      "epoch": 30.01226053639847,
+      "grad_norm": 0.005554437171667814,
+      "learning_rate": 4.308216262239251e-06,
+      "loss": 0.5229,
+      "step": 7990
+    },
+    {
+      "epoch": 30.013026819923372,
+      "grad_norm": 0.231408953666687,
+      "learning_rate": 4.299702000851427e-06,
+      "loss": 0.938,
+      "step": 8000
+    },
+    {
+      "epoch": 30.013793103448275,
+      "grad_norm": 0.001429896685294807,
+      "learning_rate": 4.291187739463602e-06,
+      "loss": 1.4188,
+      "step": 8010
+    },
+    {
+      "epoch": 30.01455938697318,
+      "grad_norm": 0.17686352133750916,
+      "learning_rate": 4.2826734780757775e-06,
+      "loss": 0.5479,
+      "step": 8020
+    },
+    {
+      "epoch": 30.015325670498083,
+      "grad_norm": 0.00278076040558517,
+      "learning_rate": 4.2741592166879525e-06,
+      "loss": 0.0121,
+      "step": 8030
+    },
+    {
+      "epoch": 30.01609195402299,
+      "grad_norm": 306.07781982421875,
+      "learning_rate": 4.2656449553001275e-06,
+      "loss": 0.5206,
+      "step": 8040
+    },
+    {
+      "epoch": 30.016858237547893,
+      "grad_norm": 0.002114189090207219,
+      "learning_rate": 4.257130693912303e-06,
+      "loss": 0.6451,
+      "step": 8050
+    },
+    {
+      "epoch": 30.017624521072797,
+      "grad_norm": 0.0014462157851085067,
+      "learning_rate": 4.248616432524479e-06,
+      "loss": 0.591,
+      "step": 8060
+    },
+    {
+      "epoch": 30.0183908045977,
+      "grad_norm": 0.09331923723220825,
+      "learning_rate": 4.240102171136654e-06,
+      "loss": 1.0715,
+      "step": 8070
+    },
+    {
+      "epoch": 30.019157088122604,
+      "grad_norm": 153.68565368652344,
+      "learning_rate": 4.231587909748829e-06,
+      "loss": 2.1039,
+      "step": 8080
+    },
+    {
+      "epoch": 30.01992337164751,
+      "grad_norm": 0.002513670129701495,
+      "learning_rate": 4.223073648361005e-06,
+      "loss": 0.003,
+      "step": 8090
+    },
+    {
+      "epoch": 30.02,
+      "eval_accuracy": 0.7333333333333333,
+      "eval_loss": 1.5859893560409546,
+      "eval_runtime": 15.4753,
+      "eval_samples_per_second": 2.908,
+      "eval_steps_per_second": 2.908,
+      "step": 8091
+    },
+    {
+      "epoch": 31.000689655172415,
+      "grad_norm": 0.5041216611862183,
+      "learning_rate": 4.214559386973181e-06,
+      "loss": 0.4334,
+      "step": 8100
+    },
+    {
+      "epoch": 31.00145593869732,
+      "grad_norm": 0.0025321398861706257,
+      "learning_rate": 4.206045125585356e-06,
+      "loss": 0.0029,
+      "step": 8110
+    },
+    {
+      "epoch": 31.002222222222223,
+      "grad_norm": 0.005120292771607637,
+      "learning_rate": 4.197530864197531e-06,
+      "loss": 0.4062,
+      "step": 8120
+    },
+    {
+      "epoch": 31.002988505747126,
+      "grad_norm": 72.58399963378906,
+      "learning_rate": 4.189016602809707e-06,
+      "loss": 0.0169,
+      "step": 8130
+    },
+    {
+      "epoch": 31.00375478927203,
+      "grad_norm": 0.001410025986842811,
+      "learning_rate": 4.180502341421882e-06,
+      "loss": 0.4409,
+      "step": 8140
+    },
+    {
+      "epoch": 31.004521072796933,
+      "grad_norm": 0.07055753469467163,
+      "learning_rate": 4.1719880800340575e-06,
+      "loss": 0.0572,
+      "step": 8150
+    },
+    {
+      "epoch": 31.00528735632184,
+      "grad_norm": 0.011598301120102406,
+      "learning_rate": 4.1634738186462325e-06,
+      "loss": 0.4805,
+      "step": 8160
+    },
+    {
+      "epoch": 31.006053639846744,
+      "grad_norm": 0.003230190370231867,
+      "learning_rate": 4.154959557258408e-06,
+      "loss": 0.0171,
+      "step": 8170
+    },
+    {
+      "epoch": 31.006819923371648,
+      "grad_norm": 482.70013427734375,
+      "learning_rate": 4.146445295870583e-06,
+      "loss": 0.8875,
+      "step": 8180
+    },
+    {
+      "epoch": 31.00758620689655,
+      "grad_norm": 0.0022959550842642784,
+      "learning_rate": 4.137931034482759e-06,
+      "loss": 1.376,
+      "step": 8190
+    },
+    {
+      "epoch": 31.008352490421455,
+      "grad_norm": 141.5695343017578,
+      "learning_rate": 4.129416773094934e-06,
+      "loss": 1.0537,
+      "step": 8200
+    },
+    {
+      "epoch": 31.00911877394636,
+      "grad_norm": 0.06153375282883644,
+      "learning_rate": 4.12090251170711e-06,
+      "loss": 0.5831,
+      "step": 8210
+    },
+    {
+      "epoch": 31.009885057471266,
+      "grad_norm": 0.09612569957971573,
+      "learning_rate": 4.112388250319285e-06,
+      "loss": 1.1069,
+      "step": 8220
+    },
+    {
+      "epoch": 31.01065134099617,
+      "grad_norm": 431.49713134765625,
+      "learning_rate": 4.103873988931461e-06,
+      "loss": 0.8904,
+      "step": 8230
+    },
+    {
+      "epoch": 31.011417624521073,
+      "grad_norm": 0.008050643838942051,
+      "learning_rate": 4.095359727543636e-06,
+      "loss": 0.0011,
+      "step": 8240
+    },
+    {
+      "epoch": 31.012183908045976,
+      "grad_norm": 0.009870373643934727,
+      "learning_rate": 4.086845466155812e-06,
+      "loss": 0.2455,
+      "step": 8250
+    },
+    {
+      "epoch": 31.01295019157088,
+      "grad_norm": 0.027511747553944588,
+      "learning_rate": 4.078331204767987e-06,
+      "loss": 0.0028,
+      "step": 8260
+    },
+    {
+      "epoch": 31.013716475095784,
+      "grad_norm": 0.011268673464655876,
+      "learning_rate": 4.0698169433801625e-06,
+      "loss": 0.5757,
+      "step": 8270
+    },
+    {
+      "epoch": 31.01448275862069,
+      "grad_norm": 0.17219111323356628,
+      "learning_rate": 4.0613026819923375e-06,
+      "loss": 1.1817,
+      "step": 8280
+    },
+    {
+      "epoch": 31.015249042145594,
+      "grad_norm": 0.0020057205110788345,
+      "learning_rate": 4.052788420604513e-06,
+      "loss": 0.0356,
+      "step": 8290
+    },
+    {
+      "epoch": 31.016015325670498,
+      "grad_norm": 0.03642117604613304,
+      "learning_rate": 4.044274159216688e-06,
+      "loss": 0.011,
+      "step": 8300
+    },
+    {
+      "epoch": 31.0167816091954,
+      "grad_norm": 0.18827003240585327,
+      "learning_rate": 4.035759897828863e-06,
+      "loss": 1.0117,
+      "step": 8310
+    },
+    {
+      "epoch": 31.017547892720305,
+      "grad_norm": 0.004165468737483025,
+      "learning_rate": 4.027245636441039e-06,
+      "loss": 0.6789,
+      "step": 8320
+    },
+    {
+      "epoch": 31.018314176245212,
+      "grad_norm": 0.22859685122966766,
+      "learning_rate": 4.018731375053214e-06,
+      "loss": 0.6094,
+      "step": 8330
+    },
+    {
+      "epoch": 31.019080459770116,
+      "grad_norm": 0.32649898529052734,
+      "learning_rate": 4.01021711366539e-06,
+      "loss": 0.7456,
+      "step": 8340
+    },
+    {
+      "epoch": 31.01984674329502,
+      "grad_norm": 0.15983377397060394,
+      "learning_rate": 4.001702852277565e-06,
+      "loss": 0.0004,
+      "step": 8350
+    },
+    {
+      "epoch": 31.02,
+      "eval_accuracy": 0.6444444444444445,
+      "eval_loss": 2.2568414211273193,
+      "eval_runtime": 16.4461,
+      "eval_samples_per_second": 2.736,
+      "eval_steps_per_second": 2.736,
+      "step": 8352
+    },
+    {
+      "epoch": 32.00061302681992,
+      "grad_norm": 0.07776673138141632,
+      "learning_rate": 3.993188590889741e-06,
+      "loss": 0.0006,
+      "step": 8360
+    },
+    {
+      "epoch": 32.00137931034483,
+      "grad_norm": 0.06379430741071701,
+      "learning_rate": 3.984674329501916e-06,
+      "loss": 0.0012,
+      "step": 8370
+    },
+    {
+      "epoch": 32.002145593869734,
+      "grad_norm": 0.016460496932268143,
+      "learning_rate": 3.976160068114092e-06,
+      "loss": 1.0225,
+      "step": 8380
+    },
+    {
+      "epoch": 32.00291187739464,
+      "grad_norm": 0.055286090821027756,
+      "learning_rate": 3.967645806726267e-06,
+      "loss": 1.295,
+      "step": 8390
+    },
+    {
+      "epoch": 32.00367816091954,
+      "grad_norm": 0.01606907695531845,
+      "learning_rate": 3.9591315453384425e-06,
+      "loss": 0.0004,
+      "step": 8400
+    },
+    {
+      "epoch": 32.004444444444445,
+      "grad_norm": 0.3932900130748749,
+      "learning_rate": 3.9506172839506175e-06,
+      "loss": 0.0016,
+      "step": 8410
+    },
+    {
+      "epoch": 32.00521072796935,
+      "grad_norm": 0.6772463321685791,
+      "learning_rate": 3.942103022562793e-06,
+      "loss": 0.5876,
+      "step": 8420
+    },
+    {
+      "epoch": 32.00597701149425,
+      "grad_norm": 194.08460998535156,
+      "learning_rate": 3.933588761174968e-06,
+      "loss": 0.5877,
+      "step": 8430
+    },
+    {
+      "epoch": 32.006743295019156,
+      "grad_norm": 0.01706990785896778,
+      "learning_rate": 3.925074499787143e-06,
+      "loss": 0.5995,
+      "step": 8440
+    },
+    {
+      "epoch": 32.00750957854406,
+      "grad_norm": 0.0012355584185570478,
+      "learning_rate": 3.916560238399319e-06,
+      "loss": 0.8429,
+      "step": 8450
+    },
+    {
+      "epoch": 32.00827586206896,
+      "grad_norm": 0.0049840230494737625,
+      "learning_rate": 3.908045977011495e-06,
+      "loss": 0.449,
+      "step": 8460
+    },
+    {
+      "epoch": 32.00904214559387,
+      "grad_norm": 0.004344654735177755,
+      "learning_rate": 3.89953171562367e-06,
+      "loss": 0.0006,
+      "step": 8470
+    },
+    {
+      "epoch": 32.00980842911878,
+      "grad_norm": 2.9717156887054443,
+      "learning_rate": 3.891017454235845e-06,
+      "loss": 0.3473,
+      "step": 8480
+    },
+    {
+      "epoch": 32.01057471264368,
+      "grad_norm": 0.026496445760130882,
+      "learning_rate": 3.882503192848021e-06,
+      "loss": 0.0017,
+      "step": 8490
+    },
+    {
+      "epoch": 32.011340996168585,
+      "grad_norm": 0.02829962782561779,
+      "learning_rate": 3.873988931460197e-06,
+      "loss": 0.0012,
+      "step": 8500
+    },
+    {
+      "epoch": 32.01210727969349,
+      "grad_norm": 0.10174506902694702,
+      "learning_rate": 3.865474670072372e-06,
+      "loss": 0.5407,
+      "step": 8510
+    },
+    {
+      "epoch": 32.01287356321839,
+      "grad_norm": 0.0033128669019788504,
+      "learning_rate": 3.856960408684547e-06,
+      "loss": 0.0004,
+      "step": 8520
+    },
+    {
+      "epoch": 32.013639846743295,
+      "grad_norm": 0.003256371710449457,
+      "learning_rate": 3.8484461472967225e-06,
+      "loss": 0.0009,
+      "step": 8530
+    },
+    {
+      "epoch": 32.0144061302682,
+      "grad_norm": 0.14455612003803253,
+      "learning_rate": 3.839931885908898e-06,
+      "loss": 0.6465,
+      "step": 8540
+    },
+    {
+      "epoch": 32.0151724137931,
+      "grad_norm": 0.05306978151202202,
+      "learning_rate": 3.831417624521073e-06,
+      "loss": 0.0009,
+      "step": 8550
+    },
+    {
+      "epoch": 32.015938697318006,
+      "grad_norm": 0.031177254393696785,
+      "learning_rate": 3.822903363133248e-06,
+      "loss": 0.7795,
+      "step": 8560
+    },
+    {
+      "epoch": 32.01670498084291,
+      "grad_norm": 0.01676277443766594,
+      "learning_rate": 3.8143891017454237e-06,
+      "loss": 0.0016,
+      "step": 8570
+    },
+    {
+      "epoch": 32.01747126436781,
+      "grad_norm": 0.011179128661751747,
+      "learning_rate": 3.805874840357599e-06,
+      "loss": 0.0007,
+      "step": 8580
+    },
+    {
+      "epoch": 32.01823754789272,
+      "grad_norm": 0.03544120863080025,
+      "learning_rate": 3.797360578969775e-06,
+      "loss": 0.7343,
+      "step": 8590
+    },
+    {
+      "epoch": 32.01900383141763,
+      "grad_norm": 0.0019313012016937137,
+      "learning_rate": 3.78884631758195e-06,
+      "loss": 1.6009,
+      "step": 8600
+    },
+    {
+      "epoch": 32.01977011494253,
+      "grad_norm": 378.3231506347656,
+      "learning_rate": 3.7803320561941254e-06,
+      "loss": 0.1452,
+      "step": 8610
+    },
+    {
+      "epoch": 32.02,
+      "eval_accuracy": 0.6444444444444445,
+      "eval_loss": 2.4112446308135986,
+      "eval_runtime": 13.281,
+      "eval_samples_per_second": 3.388,
+      "eval_steps_per_second": 3.388,
+      "step": 8613
+    },
+    {
+      "epoch": 33.00053639846743,
+      "grad_norm": 0.15281549096107483,
+      "learning_rate": 3.7718177948063004e-06,
+      "loss": 0.0016,
+      "step": 8620
+    },
+    {
+      "epoch": 33.001302681992335,
+      "grad_norm": 0.053307775408029556,
+      "learning_rate": 3.7633035334184762e-06,
+      "loss": 0.2357,
+      "step": 8630
+    },
+    {
+      "epoch": 33.00206896551724,
+      "grad_norm": 0.09021246433258057,
+      "learning_rate": 3.7547892720306517e-06,
+      "loss": 0.0004,
+      "step": 8640
+    },
+    {
+      "epoch": 33.00283524904214,
+      "grad_norm": 0.0028194631449878216,
+      "learning_rate": 3.746275010642827e-06,
+      "loss": 0.5114,
+      "step": 8650
+    },
+    {
+      "epoch": 33.00360153256705,
+      "grad_norm": 0.0028222943656146526,
+      "learning_rate": 3.737760749255002e-06,
+      "loss": 0.001,
+      "step": 8660
+    },
+    {
+      "epoch": 33.00436781609196,
+      "grad_norm": 0.0010228869505226612,
+      "learning_rate": 3.729246487867178e-06,
+      "loss": 0.0011,
+      "step": 8670
+    },
+    {
+      "epoch": 33.00513409961686,
+      "grad_norm": 0.062401916831731796,
+      "learning_rate": 3.7207322264793533e-06,
+      "loss": 0.5505,
+      "step": 8680
+    },
+    {
+      "epoch": 33.005900383141764,
+      "grad_norm": 362.3966979980469,
+      "learning_rate": 3.7122179650915287e-06,
+      "loss": 0.4197,
+      "step": 8690
+    },
+    {
+      "epoch": 33.00666666666667,
+      "grad_norm": 0.028478719294071198,
+      "learning_rate": 3.7037037037037037e-06,
+      "loss": 0.4369,
+      "step": 8700
+    },
+    {
+      "epoch": 33.00743295019157,
+      "grad_norm": 0.0008925276342779398,
+      "learning_rate": 3.6951894423158796e-06,
+      "loss": 0.0018,
+      "step": 8710
+    },
+    {
+      "epoch": 33.008199233716475,
+      "grad_norm": 69.93021392822266,
+      "learning_rate": 3.686675180928055e-06,
+      "loss": 0.5998,
+      "step": 8720
+    },
+    {
+      "epoch": 33.00896551724138,
+      "grad_norm": 0.008064580149948597,
+      "learning_rate": 3.67816091954023e-06,
+      "loss": 0.0008,
+      "step": 8730
+    },
+    {
+      "epoch": 33.00973180076628,
+      "grad_norm": 0.0009439765708521008,
+      "learning_rate": 3.6696466581524054e-06,
+      "loss": 0.001,
+      "step": 8740
+    },
+    {
+      "epoch": 33.010498084291186,
+      "grad_norm": 0.0008714428986422718,
+      "learning_rate": 3.6611323967645812e-06,
+      "loss": 0.2434,
+      "step": 8750
+    },
+    {
+      "epoch": 33.01126436781609,
+      "grad_norm": 0.0016324096359312534,
+      "learning_rate": 3.6526181353767567e-06,
+      "loss": 0.0026,
+      "step": 8760
+    },
+    {
+      "epoch": 33.01203065134099,
+      "grad_norm": 0.2354644238948822,
+      "learning_rate": 3.6441038739889317e-06,
+      "loss": 0.947,
+      "step": 8770
+    },
+    {
+      "epoch": 33.012796934865904,
+      "grad_norm": 0.017056241631507874,
+      "learning_rate": 3.635589612601107e-06,
+      "loss": 0.5597,
+      "step": 8780
+    },
+    {
+      "epoch": 33.01356321839081,
+      "grad_norm": 0.007571222726255655,
+      "learning_rate": 3.627075351213283e-06,
+      "loss": 0.5969,
+      "step": 8790
+    },
+    {
+      "epoch": 33.01432950191571,
+      "grad_norm": 0.06942231208086014,
+      "learning_rate": 3.618561089825458e-06,
+      "loss": 0.2976,
+      "step": 8800
+    },
+    {
+      "epoch": 33.015095785440614,
+      "grad_norm": 0.0020972101483494043,
+      "learning_rate": 3.6100468284376333e-06,
+      "loss": 0.7729,
+      "step": 8810
+    },
+    {
+      "epoch": 33.01586206896552,
+      "grad_norm": 0.0008585660834796727,
+      "learning_rate": 3.6015325670498087e-06,
+      "loss": 1.0393,
+      "step": 8820
+    },
+    {
+      "epoch": 33.01662835249042,
+      "grad_norm": 0.03751160949468613,
+      "learning_rate": 3.5930183056619837e-06,
+      "loss": 0.0003,
+      "step": 8830
+    },
+    {
+      "epoch": 33.017394636015325,
+      "grad_norm": 0.002867656061425805,
+      "learning_rate": 3.5845040442741596e-06,
+      "loss": 0.5826,
+      "step": 8840
+    },
+    {
+      "epoch": 33.01816091954023,
+      "grad_norm": 0.0013876117300242186,
+      "learning_rate": 3.575989782886335e-06,
+      "loss": 0.6495,
+      "step": 8850
+    },
+    {
+      "epoch": 33.01892720306513,
+      "grad_norm": 983.3112182617188,
+      "learning_rate": 3.56747552149851e-06,
+      "loss": 1.1592,
+      "step": 8860
+    },
+    {
+      "epoch": 33.019693486590036,
+      "grad_norm": 0.011666815727949142,
+      "learning_rate": 3.5589612601106854e-06,
+      "loss": 0.3815,
+      "step": 8870
+    },
+    {
+      "epoch": 33.02,
+      "eval_accuracy": 0.7555555555555555,
+      "eval_loss": 1.3678969144821167,
+      "eval_runtime": 13.8216,
+      "eval_samples_per_second": 3.256,
+      "eval_steps_per_second": 3.256,
+      "step": 8874
+    },
+    {
+      "epoch": 34.000459770114944,
+      "grad_norm": 0.2762603461742401,
+      "learning_rate": 3.5504469987228612e-06,
+      "loss": 3.2451,
+      "step": 8880
+    },
+    {
+      "epoch": 34.00122605363985,
+      "grad_norm": 0.004093986004590988,
+      "learning_rate": 3.5419327373350367e-06,
+      "loss": 0.2555,
+      "step": 8890
+    },
+    {
+      "epoch": 34.00199233716475,
+      "grad_norm": 0.003161295084282756,
+      "learning_rate": 3.5334184759472117e-06,
+      "loss": 0.0051,
+      "step": 8900
+    },
+    {
+      "epoch": 34.002758620689654,
+      "grad_norm": 0.046276550740003586,
+      "learning_rate": 3.524904214559387e-06,
+      "loss": 0.3972,
+      "step": 8910
+    },
+    {
+      "epoch": 34.00352490421456,
+      "grad_norm": 0.0007882455829530954,
+      "learning_rate": 3.516389953171563e-06,
+      "loss": 0.0008,
+      "step": 8920
+    },
+    {
+      "epoch": 34.00429118773946,
+      "grad_norm": 0.00730425538495183,
+      "learning_rate": 3.507875691783738e-06,
+      "loss": 0.7386,
+      "step": 8930
+    },
+    {
+      "epoch": 34.005057471264365,
+      "grad_norm": 0.18467597663402557,
+      "learning_rate": 3.4993614303959133e-06,
+      "loss": 1.5343,
+      "step": 8940
+    },
+    {
+      "epoch": 34.00582375478927,
+      "grad_norm": 0.19425460696220398,
+      "learning_rate": 3.4908471690080887e-06,
+      "loss": 0.001,
+      "step": 8950
+    },
+    {
+      "epoch": 34.00659003831418,
+      "grad_norm": 0.5217742323875427,
+      "learning_rate": 3.4823329076202646e-06,
+      "loss": 0.0153,
+      "step": 8960
+    },
+    {
+      "epoch": 34.00735632183908,
+      "grad_norm": 0.013114755041897297,
+      "learning_rate": 3.4738186462324396e-06,
+      "loss": 2.0398,
+      "step": 8970
+    },
+    {
+      "epoch": 34.00812260536399,
+      "grad_norm": 0.0025911638513207436,
+      "learning_rate": 3.465304384844615e-06,
+      "loss": 0.0116,
+      "step": 8980
+    },
+    {
+      "epoch": 34.00888888888889,
+      "grad_norm": 0.002555274870246649,
+      "learning_rate": 3.4567901234567904e-06,
+      "loss": 0.7117,
+      "step": 8990
+    },
+    {
+      "epoch": 34.009655172413794,
+      "grad_norm": 0.0019624708220362663,
+      "learning_rate": 3.448275862068966e-06,
+      "loss": 0.0025,
+      "step": 9000
+    },
+    {
+      "epoch": 34.0104214559387,
+      "grad_norm": 0.006993942428380251,
+      "learning_rate": 3.4397616006811412e-06,
+      "loss": 0.752,
+      "step": 9010
+    },
+    {
+      "epoch": 34.0111877394636,
+      "grad_norm": 0.12625227868556976,
+      "learning_rate": 3.4312473392933167e-06,
+      "loss": 0.2687,
+      "step": 9020
+    },
+    {
+      "epoch": 34.011954022988505,
+      "grad_norm": 0.0011391348671168089,
+      "learning_rate": 3.4227330779054917e-06,
+      "loss": 0.5354,
+      "step": 9030
+    },
+    {
+      "epoch": 34.01272030651341,
+      "grad_norm": 0.3375532031059265,
+      "learning_rate": 3.4142188165176675e-06,
+      "loss": 0.001,
+      "step": 9040
+    },
+    {
+      "epoch": 34.01348659003831,
+      "grad_norm": 0.0027614478021860123,
+      "learning_rate": 3.405704555129843e-06,
+      "loss": 1.7076,
+      "step": 9050
+    },
+    {
+      "epoch": 34.014252873563215,
+      "grad_norm": 7.6758856773376465,
+      "learning_rate": 3.3971902937420183e-06,
+      "loss": 0.008,
+      "step": 9060
+    },
+    {
+      "epoch": 34.01501915708812,
+      "grad_norm": 0.001212018309161067,
+      "learning_rate": 3.3886760323541933e-06,
+      "loss": 0.0027,
+      "step": 9070
+    },
+    {
+      "epoch": 34.01578544061303,
+      "grad_norm": 0.0031360953580588102,
+      "learning_rate": 3.380161770966369e-06,
+      "loss": 0.0019,
+      "step": 9080
+    },
+    {
+      "epoch": 34.01655172413793,
+      "grad_norm": 1.0964499711990356,
+      "learning_rate": 3.3716475095785446e-06,
+      "loss": 0.0009,
+      "step": 9090
+    },
+    {
+      "epoch": 34.01731800766284,
+      "grad_norm": 135.07989501953125,
+      "learning_rate": 3.3631332481907196e-06,
+      "loss": 2.0915,
+      "step": 9100
+    },
+    {
+      "epoch": 34.01808429118774,
+      "grad_norm": 0.005994447972625494,
+      "learning_rate": 3.354618986802895e-06,
+      "loss": 0.5322,
+      "step": 9110
+    },
+    {
+      "epoch": 34.018850574712644,
+      "grad_norm": 0.001170378876850009,
+      "learning_rate": 3.3461047254150704e-06,
+      "loss": 0.0008,
+      "step": 9120
+    },
+    {
+      "epoch": 34.01961685823755,
+      "grad_norm": 0.0009761180845089257,
+      "learning_rate": 3.3375904640272463e-06,
+      "loss": 0.0013,
+      "step": 9130
+    },
+    {
+      "epoch": 34.02,
+      "eval_accuracy": 0.7111111111111111,
+      "eval_loss": 1.8306384086608887,
+      "eval_runtime": 14.4419,
+      "eval_samples_per_second": 3.116,
+      "eval_steps_per_second": 3.116,
+      "step": 9135
+    },
+    {
+      "epoch": 35.000383141762455,
+      "grad_norm": 0.008362943306565285,
+      "learning_rate": 3.3290762026394212e-06,
+      "loss": 0.0008,
+      "step": 9140
+    },
+    {
+      "epoch": 35.00114942528736,
+      "grad_norm": 0.09023687988519669,
+      "learning_rate": 3.3205619412515967e-06,
+      "loss": 0.0005,
+      "step": 9150
+    },
+    {
+      "epoch": 35.00191570881226,
+      "grad_norm": 0.012076799757778645,
+      "learning_rate": 3.3120476798637717e-06,
+      "loss": 0.0592,
+      "step": 9160
+    },
+    {
+      "epoch": 35.002681992337166,
+      "grad_norm": 0.0008539219270460308,
+      "learning_rate": 3.3035334184759475e-06,
+      "loss": 0.0013,
+      "step": 9170
+    },
+    {
+      "epoch": 35.00344827586207,
+      "grad_norm": 6.743452072143555,
+      "learning_rate": 3.295019157088123e-06,
+      "loss": 1.3463,
+      "step": 9180
+    },
+    {
+      "epoch": 35.00421455938697,
+      "grad_norm": 0.0008903272682800889,
+      "learning_rate": 3.2865048957002983e-06,
+      "loss": 0.0009,
+      "step": 9190
+    },
+    {
+      "epoch": 35.00498084291188,
+      "grad_norm": 0.0033099281135946512,
+      "learning_rate": 3.2779906343124733e-06,
+      "loss": 0.0095,
+      "step": 9200
+    },
+    {
+      "epoch": 35.00574712643678,
+      "grad_norm": 0.0029596216045320034,
+      "learning_rate": 3.269476372924649e-06,
+      "loss": 0.5017,
+      "step": 9210
+    },
+    {
+      "epoch": 35.006513409961684,
+      "grad_norm": 0.1439450979232788,
+      "learning_rate": 3.2609621115368246e-06,
+      "loss": 0.7239,
+      "step": 9220
+    },
+    {
+      "epoch": 35.00727969348659,
+      "grad_norm": 0.05335940048098564,
+      "learning_rate": 3.2524478501489996e-06,
+      "loss": 0.0105,
+      "step": 9230
+    },
+    {
+      "epoch": 35.00804597701149,
+      "grad_norm": 0.08966932445764542,
+      "learning_rate": 3.243933588761175e-06,
+      "loss": 0.6246,
+      "step": 9240
+    },
+    {
+      "epoch": 35.008812260536395,
+      "grad_norm": 0.13768485188484192,
+      "learning_rate": 3.235419327373351e-06,
+      "loss": 0.0014,
+      "step": 9250
+    },
+    {
+      "epoch": 35.009578544061306,
+      "grad_norm": 0.003539847442880273,
+      "learning_rate": 3.2269050659855262e-06,
+      "loss": 0.4629,
+      "step": 9260
+    },
+    {
+      "epoch": 35.01034482758621,
+      "grad_norm": 0.17655432224273682,
+      "learning_rate": 3.2183908045977012e-06,
+      "loss": 1.1958,
+      "step": 9270
+    },
+    {
+      "epoch": 35.01111111111111,
+      "grad_norm": 0.3544051945209503,
+      "learning_rate": 3.2098765432098767e-06,
+      "loss": 0.0111,
+      "step": 9280
+    },
+    {
+      "epoch": 35.01187739463602,
+      "grad_norm": 0.2912541627883911,
+      "learning_rate": 3.2013622818220525e-06,
+      "loss": 1.2761,
+      "step": 9290
+    },
+    {
+      "epoch": 35.01264367816092,
+      "grad_norm": 0.005516159813851118,
+      "learning_rate": 3.1928480204342275e-06,
+      "loss": 0.0014,
+      "step": 9300
+    },
+    {
+      "epoch": 35.013409961685824,
+      "grad_norm": 25.71188735961914,
+      "learning_rate": 3.184333759046403e-06,
+      "loss": 0.5261,
+      "step": 9310
+    },
+    {
+      "epoch": 35.01417624521073,
+      "grad_norm": 0.0013126012636348605,
+      "learning_rate": 3.1758194976585783e-06,
+      "loss": 0.077,
+      "step": 9320
+    },
+    {
+      "epoch": 35.01494252873563,
+      "grad_norm": 49.9453125,
+      "learning_rate": 3.167305236270754e-06,
+      "loss": 0.4977,
+      "step": 9330
+    },
+    {
+      "epoch": 35.015708812260534,
+      "grad_norm": 0.0009205365204252303,
+      "learning_rate": 3.158790974882929e-06,
+      "loss": 0.5102,
+      "step": 9340
+    },
+    {
+      "epoch": 35.01647509578544,
+      "grad_norm": 0.06953918188810349,
+      "learning_rate": 3.1502767134951046e-06,
+      "loss": 0.0009,
+      "step": 9350
+    },
+    {
+      "epoch": 35.01724137931034,
+      "grad_norm": 0.04685080796480179,
+      "learning_rate": 3.14176245210728e-06,
+      "loss": 0.6047,
+      "step": 9360
+    },
+    {
+      "epoch": 35.01800766283525,
+      "grad_norm": 54.803558349609375,
+      "learning_rate": 3.133248190719455e-06,
+      "loss": 1.22,
+      "step": 9370
+    },
+    {
+      "epoch": 35.018773946360156,
+      "grad_norm": 0.0024438260588794947,
+      "learning_rate": 3.124733929331631e-06,
+      "loss": 0.0024,
+      "step": 9380
+    },
+    {
+      "epoch": 35.01954022988506,
+      "grad_norm": 0.004958088044077158,
+      "learning_rate": 3.1162196679438062e-06,
+      "loss": 0.7655,
+      "step": 9390
+    },
+    {
+      "epoch": 35.02,
+      "eval_accuracy": 0.7333333333333333,
+      "eval_loss": 1.4607685804367065,
+      "eval_runtime": 14.1654,
+      "eval_samples_per_second": 3.177,
+      "eval_steps_per_second": 3.177,
+      "step": 9396
+    },
+    {
+      "epoch": 36.00030651340996,
+      "grad_norm": 0.061240918934345245,
+      "learning_rate": 3.1077054065559812e-06,
+      "loss": 0.0131,
+      "step": 9400
+    },
+    {
+      "epoch": 36.001072796934864,
+      "grad_norm": 0.004116494674235582,
+      "learning_rate": 3.0991911451681567e-06,
+      "loss": 0.0026,
+      "step": 9410
+    },
+    {
+      "epoch": 36.00183908045977,
+      "grad_norm": 0.0022276234813034534,
+      "learning_rate": 3.0906768837803325e-06,
+      "loss": 0.6602,
+      "step": 9420
+    },
+    {
+      "epoch": 36.00260536398467,
+      "grad_norm": 0.007051907479763031,
+      "learning_rate": 3.082162622392508e-06,
+      "loss": 0.6238,
+      "step": 9430
+    },
+    {
+      "epoch": 36.00337164750958,
+      "grad_norm": 0.010487622581422329,
+      "learning_rate": 3.073648361004683e-06,
+      "loss": 0.0005,
+      "step": 9440
+    },
+    {
+      "epoch": 36.004137931034485,
+      "grad_norm": 0.020656408742070198,
+      "learning_rate": 3.0651340996168583e-06,
+      "loss": 0.0003,
+      "step": 9450
+    },
+    {
+      "epoch": 36.00490421455939,
+      "grad_norm": 0.0014071677578613162,
+      "learning_rate": 3.056619838229034e-06,
+      "loss": 0.0008,
+      "step": 9460
+    },
+    {
+      "epoch": 36.00567049808429,
+      "grad_norm": 0.04767834395170212,
+      "learning_rate": 3.048105576841209e-06,
+      "loss": 0.6376,
+      "step": 9470
+    },
+    {
+      "epoch": 36.006436781609196,
+      "grad_norm": 0.001168592949397862,
+      "learning_rate": 3.0395913154533846e-06,
+      "loss": 1.2768,
+      "step": 9480
+    },
+    {
+      "epoch": 36.0072030651341,
+      "grad_norm": 0.12005781382322311,
+      "learning_rate": 3.03107705406556e-06,
+      "loss": 0.7032,
+      "step": 9490
+    },
+    {
+      "epoch": 36.007969348659,
+      "grad_norm": 0.049370937049388885,
+      "learning_rate": 3.022562792677736e-06,
+      "loss": 0.3523,
+      "step": 9500
+    },
+    {
+      "epoch": 36.00873563218391,
+      "grad_norm": 25.666242599487305,
+      "learning_rate": 3.014048531289911e-06,
+      "loss": 0.5111,
+      "step": 9510
+    },
+    {
+      "epoch": 36.00950191570881,
+      "grad_norm": 0.0010242098942399025,
+      "learning_rate": 3.0055342699020862e-06,
+      "loss": 0.9859,
+      "step": 9520
+    },
+    {
+      "epoch": 36.010268199233714,
+      "grad_norm": 0.020110201090574265,
+      "learning_rate": 2.9970200085142612e-06,
+      "loss": 0.0011,
+      "step": 9530
+    },
+    {
+      "epoch": 36.01103448275862,
+      "grad_norm": 0.12330694496631622,
+      "learning_rate": 2.988505747126437e-06,
+      "loss": 0.0268,
+      "step": 9540
+    },
+    {
+      "epoch": 36.01180076628353,
+      "grad_norm": 0.0015601320192217827,
+      "learning_rate": 2.9799914857386125e-06,
+      "loss": 0.0023,
+      "step": 9550
+    },
+    {
+      "epoch": 36.01256704980843,
+      "grad_norm": 0.0010449588298797607,
+      "learning_rate": 2.971477224350788e-06,
+      "loss": 0.0034,
+      "step": 9560
+    },
+    {
+      "epoch": 36.013333333333335,
+      "grad_norm": 60.46844482421875,
+      "learning_rate": 2.962962962962963e-06,
+      "loss": 1.1134,
+      "step": 9570
+    },
+    {
+      "epoch": 36.01409961685824,
+      "grad_norm": 0.13516788184642792,
+      "learning_rate": 2.9544487015751387e-06,
+      "loss": 1.1393,
+      "step": 9580
+    },
+    {
+      "epoch": 36.01486590038314,
+      "grad_norm": 0.19734834134578705,
+      "learning_rate": 2.945934440187314e-06,
+      "loss": 1.1429,
+      "step": 9590
+    },
+    {
+      "epoch": 36.015632183908046,
+      "grad_norm": 0.04803425073623657,
+      "learning_rate": 2.9374201787994896e-06,
+      "loss": 0.8387,
+      "step": 9600
+    },
+    {
+      "epoch": 36.01639846743295,
+      "grad_norm": 0.004391372203826904,
+      "learning_rate": 2.9289059174116646e-06,
+      "loss": 1.5079,
+      "step": 9610
+    },
+    {
+      "epoch": 36.01716475095785,
+      "grad_norm": 0.04461931064724922,
+      "learning_rate": 2.92039165602384e-06,
+      "loss": 0.0154,
+      "step": 9620
+    },
+    {
+      "epoch": 36.01793103448276,
+      "grad_norm": 0.16307049989700317,
+      "learning_rate": 2.911877394636016e-06,
+      "loss": 0.0026,
+      "step": 9630
+    },
+    {
+      "epoch": 36.01869731800766,
+      "grad_norm": 0.17401742935180664,
+      "learning_rate": 2.903363133248191e-06,
+      "loss": 0.6489,
+      "step": 9640
+    },
+    {
+      "epoch": 36.019463601532564,
+      "grad_norm": 0.0012012688675895333,
+      "learning_rate": 2.8948488718603662e-06,
+      "loss": 0.003,
+      "step": 9650
+    },
+    {
+      "epoch": 36.02,
+      "eval_accuracy": 0.6666666666666666,
+      "eval_loss": 2.202925205230713,
+      "eval_runtime": 13.8555,
+      "eval_samples_per_second": 3.248,
+      "eval_steps_per_second": 3.248,
+      "step": 9657
+    },
+    {
+      "epoch": 37.00022988505747,
+      "grad_norm": 0.23111657798290253,
+      "learning_rate": 2.8863346104725417e-06,
+      "loss": 0.0025,
+      "step": 9660
+    },
+    {
+      "epoch": 37.000996168582375,
+      "grad_norm": 0.0007817544392310083,
+      "learning_rate": 2.8778203490847175e-06,
+      "loss": 0.4859,
+      "step": 9670
+    },
+    {
+      "epoch": 37.00176245210728,
+      "grad_norm": 0.0477173775434494,
+      "learning_rate": 2.8693060876968925e-06,
+      "loss": 0.0008,
+      "step": 9680
+    },
+    {
+      "epoch": 37.00252873563218,
+      "grad_norm": 0.000743559910915792,
+      "learning_rate": 2.860791826309068e-06,
+      "loss": 0.0004,
+      "step": 9690
+    },
+    {
+      "epoch": 37.003295019157086,
+      "grad_norm": 0.000858695711940527,
+      "learning_rate": 2.852277564921243e-06,
+      "loss": 0.6213,
+      "step": 9700
+    },
+    {
+      "epoch": 37.00406130268199,
+      "grad_norm": 2.1517531871795654,
+      "learning_rate": 2.8437633035334187e-06,
+      "loss": 0.9624,
+      "step": 9710
+    },
+    {
+      "epoch": 37.00482758620689,
+      "grad_norm": 0.012443887069821358,
+      "learning_rate": 2.835249042145594e-06,
+      "loss": 0.0015,
+      "step": 9720
+    },
+    {
+      "epoch": 37.005593869731804,
+      "grad_norm": 0.006380707025527954,
+      "learning_rate": 2.8267347807577696e-06,
+      "loss": 0.4598,
+      "step": 9730
+    },
+    {
+      "epoch": 37.00636015325671,
+      "grad_norm": 0.0008849736768752337,
+      "learning_rate": 2.8182205193699446e-06,
+      "loss": 0.0001,
+      "step": 9740
+    },
+    {
+      "epoch": 37.00712643678161,
+      "grad_norm": 0.0005933063803240657,
+      "learning_rate": 2.8097062579821204e-06,
+      "loss": 0.0007,
+      "step": 9750
+    },
+    {
+      "epoch": 37.007892720306515,
+      "grad_norm": 0.0010139941005036235,
+      "learning_rate": 2.801191996594296e-06,
+      "loss": 1.2024,
+      "step": 9760
+    },
+    {
+      "epoch": 37.00865900383142,
+      "grad_norm": 0.0027134341653436422,
+      "learning_rate": 2.792677735206471e-06,
+      "loss": 1.7367,
+      "step": 9770
+    },
+    {
+      "epoch": 37.00942528735632,
+      "grad_norm": 0.001102240988984704,
+      "learning_rate": 2.7841634738186462e-06,
+      "loss": 0.7939,
+      "step": 9780
+    },
+    {
+      "epoch": 37.010191570881226,
+      "grad_norm": 0.013465874828398228,
+      "learning_rate": 2.775649212430822e-06,
+      "loss": 0.1877,
+      "step": 9790
+    },
+    {
+      "epoch": 37.01095785440613,
+      "grad_norm": 0.004637483507394791,
+      "learning_rate": 2.7671349510429975e-06,
+      "loss": 0.0006,
+      "step": 9800
+    },
+    {
+      "epoch": 37.01172413793103,
+      "grad_norm": 0.0016687975730746984,
+      "learning_rate": 2.7586206896551725e-06,
+      "loss": 0.0005,
+      "step": 9810
+    },
+    {
+      "epoch": 37.01249042145594,
+      "grad_norm": 0.10679076611995697,
+      "learning_rate": 2.750106428267348e-06,
+      "loss": 0.0824,
+      "step": 9820
+    },
+    {
+      "epoch": 37.01325670498084,
+      "grad_norm": 0.008856571279466152,
+      "learning_rate": 2.7415921668795238e-06,
+      "loss": 0.0004,
+      "step": 9830
+    },
+    {
+      "epoch": 37.014022988505744,
+      "grad_norm": 0.0008473598281852901,
+      "learning_rate": 2.7330779054916987e-06,
+      "loss": 0.0011,
+      "step": 9840
+    },
+    {
+      "epoch": 37.014789272030654,
+      "grad_norm": 0.11508283019065857,
+      "learning_rate": 2.724563644103874e-06,
+      "loss": 1.1716,
+      "step": 9850
+    },
+    {
+      "epoch": 37.01555555555556,
+      "grad_norm": 0.03232954069972038,
+      "learning_rate": 2.7160493827160496e-06,
+      "loss": 0.0005,
+      "step": 9860
+    },
+    {
+      "epoch": 37.01632183908046,
+      "grad_norm": 0.2793327569961548,
+      "learning_rate": 2.7075351213282254e-06,
+      "loss": 0.0017,
+      "step": 9870
+    },
+    {
+      "epoch": 37.017088122605365,
+      "grad_norm": 0.0020207474008202553,
+      "learning_rate": 2.6990208599404004e-06,
+      "loss": 0.0006,
+      "step": 9880
+    },
+    {
+      "epoch": 37.01785440613027,
+      "grad_norm": 0.0022292647045105696,
+      "learning_rate": 2.690506598552576e-06,
+      "loss": 0.559,
+      "step": 9890
+    },
+    {
+      "epoch": 37.01862068965517,
+      "grad_norm": 0.007433728780597448,
+      "learning_rate": 2.6819923371647512e-06,
+      "loss": 0.0557,
+      "step": 9900
+    },
+    {
+      "epoch": 37.019386973180076,
+      "grad_norm": 0.0006471008528023958,
+      "learning_rate": 2.6734780757769262e-06,
+      "loss": 0.0246,
+      "step": 9910
+    },
+    {
+      "epoch": 37.02,
+      "eval_accuracy": 0.6222222222222222,
+      "eval_loss": 2.758582592010498,
+      "eval_runtime": 15.3967,
+      "eval_samples_per_second": 2.923,
+      "eval_steps_per_second": 2.923,
+      "step": 9918
+    },
+    {
+      "epoch": 38.000153256704984,
+      "grad_norm": 1.0717862844467163,
+      "learning_rate": 2.664963814389102e-06,
+      "loss": 1.7363,
+      "step": 9920
+    },
+    {
+      "epoch": 38.00091954022989,
+      "grad_norm": 0.17203135788440704,
+      "learning_rate": 2.6564495530012775e-06,
+      "loss": 0.0015,
+      "step": 9930
+    },
+    {
+      "epoch": 38.00168582375479,
+      "grad_norm": 0.11133809387683868,
+      "learning_rate": 2.6479352916134525e-06,
+      "loss": 0.0013,
+      "step": 9940
+    },
+    {
+      "epoch": 38.002452107279694,
+      "grad_norm": 0.5325710773468018,
+      "learning_rate": 2.639421030225628e-06,
+      "loss": 1.3669,
+      "step": 9950
+    },
+    {
+      "epoch": 38.0032183908046,
+      "grad_norm": 0.7260966897010803,
+      "learning_rate": 2.6309067688378037e-06,
+      "loss": 0.0061,
+      "step": 9960
+    },
+    {
+      "epoch": 38.0039846743295,
+      "grad_norm": 22.296585083007812,
+      "learning_rate": 2.622392507449979e-06,
+      "loss": 0.0036,
+      "step": 9970
+    },
+    {
+      "epoch": 38.004750957854405,
+      "grad_norm": 0.11299914866685867,
+      "learning_rate": 2.613878246062154e-06,
+      "loss": 0.0022,
+      "step": 9980
+    },
+    {
+      "epoch": 38.00551724137931,
+      "grad_norm": 0.049726832658052444,
+      "learning_rate": 2.6053639846743296e-06,
+      "loss": 1.2186,
+      "step": 9990
+    },
+    {
+      "epoch": 38.00628352490421,
+      "grad_norm": 0.0019447283120825887,
+      "learning_rate": 2.5968497232865054e-06,
+      "loss": 0.508,
+      "step": 10000
+    },
+    {
+      "epoch": 38.007049808429116,
+      "grad_norm": 0.0006838354747742414,
+      "learning_rate": 2.5883354618986804e-06,
+      "loss": 1.5742,
+      "step": 10010
+    },
+    {
+      "epoch": 38.00781609195402,
+      "grad_norm": 1236.9111328125,
+      "learning_rate": 2.579821200510856e-06,
+      "loss": 1.7882,
+      "step": 10020
+    },
+    {
+      "epoch": 38.00858237547893,
+      "grad_norm": 0.12140507996082306,
+      "learning_rate": 2.5713069391230312e-06,
+      "loss": 0.0014,
+      "step": 10030
+    },
+    {
+      "epoch": 38.009348659003834,
+      "grad_norm": 0.00671731773763895,
+      "learning_rate": 2.562792677735207e-06,
+      "loss": 0.001,
+      "step": 10040
+    },
+    {
+      "epoch": 38.01011494252874,
+      "grad_norm": 0.0007894644513726234,
+      "learning_rate": 2.554278416347382e-06,
+      "loss": 0.6358,
+      "step": 10050
+    },
+    {
+      "epoch": 38.01088122605364,
+      "grad_norm": 0.18793921172618866,
+      "learning_rate": 2.5457641549595575e-06,
+      "loss": 0.2218,
+      "step": 10060
+    },
+    {
+      "epoch": 38.011647509578545,
+      "grad_norm": 0.00469109695404768,
+      "learning_rate": 2.5372498935717325e-06,
+      "loss": 0.0006,
+      "step": 10070
+    },
+    {
+      "epoch": 38.01241379310345,
+      "grad_norm": 0.037202171981334686,
+      "learning_rate": 2.5287356321839083e-06,
+      "loss": 0.0056,
+      "step": 10080
+    },
+    {
+      "epoch": 38.01318007662835,
+      "grad_norm": 0.06974373012781143,
+      "learning_rate": 2.5202213707960837e-06,
+      "loss": 0.0065,
+      "step": 10090
+    },
+    {
+      "epoch": 38.013946360153255,
+      "grad_norm": 0.0019093331648036838,
+      "learning_rate": 2.511707109408259e-06,
+      "loss": 0.5021,
+      "step": 10100
+    },
+    {
+      "epoch": 38.01471264367816,
+      "grad_norm": 0.0015057099517434835,
+      "learning_rate": 2.503192848020434e-06,
+      "loss": 0.0011,
+      "step": 10110
+    },
+    {
+      "epoch": 38.01547892720306,
+      "grad_norm": 0.0014377710176631808,
+      "learning_rate": 2.49467858663261e-06,
+      "loss": 1.1023,
+      "step": 10120
+    },
+    {
+      "epoch": 38.016245210727966,
+      "grad_norm": 0.12007944285869598,
+      "learning_rate": 2.4861643252447854e-06,
+      "loss": 0.0129,
+      "step": 10130
+    },
+    {
+      "epoch": 38.01701149425288,
+      "grad_norm": 0.10494039952754974,
+      "learning_rate": 2.4776500638569604e-06,
+      "loss": 0.3025,
+      "step": 10140
+    },
+    {
+      "epoch": 38.01777777777778,
+      "grad_norm": 183.83294677734375,
+      "learning_rate": 2.469135802469136e-06,
+      "loss": 0.5541,
+      "step": 10150
+    },
+    {
+      "epoch": 38.018544061302684,
+      "grad_norm": 0.016142094507813454,
+      "learning_rate": 2.4606215410813112e-06,
+      "loss": 0.0807,
+      "step": 10160
+    },
+    {
+      "epoch": 38.01931034482759,
+      "grad_norm": 0.007817224599421024,
+      "learning_rate": 2.4521072796934867e-06,
+      "loss": 0.0007,
+      "step": 10170
+    },
+    {
+      "epoch": 38.02,
+      "eval_accuracy": 0.6444444444444445,
+      "eval_loss": 2.6803932189941406,
+      "eval_runtime": 16.4696,
+      "eval_samples_per_second": 2.732,
+      "eval_steps_per_second": 2.732,
+      "step": 10179
+    },
+    {
+      "epoch": 39.00007662835249,
+      "grad_norm": 0.011514992453157902,
+      "learning_rate": 2.443593018305662e-06,
+      "loss": 0.711,
+      "step": 10180
+    },
+    {
+      "epoch": 39.00084291187739,
+      "grad_norm": 0.038958631455898285,
+      "learning_rate": 2.4350787569178375e-06,
+      "loss": 1.1963,
+      "step": 10190
+    },
+    {
+      "epoch": 39.001609195402295,
+      "grad_norm": 0.062485817819833755,
+      "learning_rate": 2.426564495530013e-06,
+      "loss": 0.0074,
+      "step": 10200
+    },
+    {
+      "epoch": 39.002375478927206,
+      "grad_norm": 25.11356544494629,
+      "learning_rate": 2.4180502341421883e-06,
+      "loss": 0.0049,
+      "step": 10210
+    },
+    {
+      "epoch": 39.00314176245211,
+      "grad_norm": 0.00820657704025507,
+      "learning_rate": 2.4095359727543637e-06,
+      "loss": 0.5552,
+      "step": 10220
+    },
+    {
+      "epoch": 39.00390804597701,
+      "grad_norm": 558.9811401367188,
+      "learning_rate": 2.401021711366539e-06,
+      "loss": 0.5175,
+      "step": 10230
+    },
+    {
+      "epoch": 39.00467432950192,
+      "grad_norm": 0.001662991475313902,
+      "learning_rate": 2.3925074499787146e-06,
+      "loss": 0.8054,
+      "step": 10240
+    },
+    {
+      "epoch": 39.00544061302682,
+      "grad_norm": 32.78825378417969,
+      "learning_rate": 2.38399318859089e-06,
+      "loss": 0.6772,
+      "step": 10250
+    },
+    {
+      "epoch": 39.006206896551724,
+      "grad_norm": 0.01901569776237011,
+      "learning_rate": 2.3754789272030654e-06,
+      "loss": 0.3458,
+      "step": 10260
+    },
+    {
+      "epoch": 39.00697318007663,
+      "grad_norm": 0.8498271703720093,
+      "learning_rate": 2.366964665815241e-06,
+      "loss": 0.473,
+      "step": 10270
+    },
+    {
+      "epoch": 39.00773946360153,
+      "grad_norm": 434.0633544921875,
+      "learning_rate": 2.3584504044274162e-06,
+      "loss": 1.7739,
+      "step": 10280
+    },
+    {
+      "epoch": 39.008505747126435,
+      "grad_norm": 0.05720654875040054,
+      "learning_rate": 2.3499361430395912e-06,
+      "loss": 0.0005,
+      "step": 10290
+    },
+    {
+      "epoch": 39.00927203065134,
+      "grad_norm": 0.29777073860168457,
+      "learning_rate": 2.341421881651767e-06,
+      "loss": 0.001,
+      "step": 10300
+    },
+    {
+      "epoch": 39.01003831417624,
+      "grad_norm": 0.0018811736954376101,
+      "learning_rate": 2.332907620263942e-06,
+      "loss": 0.6039,
+      "step": 10310
+    },
+    {
+      "epoch": 39.01080459770115,
+      "grad_norm": 0.0006688185385428369,
+      "learning_rate": 2.324393358876118e-06,
+      "loss": 0.6483,
+      "step": 10320
+    },
+    {
+      "epoch": 39.011570881226056,
+      "grad_norm": 0.0013783015310764313,
+      "learning_rate": 2.315879097488293e-06,
+      "loss": 0.0058,
+      "step": 10330
+    },
+    {
+      "epoch": 39.01233716475096,
+      "grad_norm": 0.0050174882635474205,
+      "learning_rate": 2.3073648361004688e-06,
+      "loss": 0.0005,
+      "step": 10340
+    },
+    {
+      "epoch": 39.013103448275864,
+      "grad_norm": 0.02628183178603649,
+      "learning_rate": 2.2988505747126437e-06,
+      "loss": 0.4835,
+      "step": 10350
+    },
+    {
+      "epoch": 39.01386973180077,
+      "grad_norm": 0.18247047066688538,
+      "learning_rate": 2.290336313324819e-06,
+      "loss": 0.3041,
+      "step": 10360
+    },
+    {
+      "epoch": 39.01463601532567,
+      "grad_norm": 0.042854152619838715,
+      "learning_rate": 2.2818220519369946e-06,
+      "loss": 0.4872,
+      "step": 10370
+    },
+    {
+      "epoch": 39.015402298850574,
+      "grad_norm": 0.0007316448027268052,
+      "learning_rate": 2.27330779054917e-06,
+      "loss": 0.5496,
+      "step": 10380
+    },
+    {
+      "epoch": 39.01616858237548,
+      "grad_norm": 0.06810470670461655,
+      "learning_rate": 2.2647935291613454e-06,
+      "loss": 0.0009,
+      "step": 10390
+    },
+    {
+      "epoch": 39.01693486590038,
+      "grad_norm": 1.610714077949524,
+      "learning_rate": 2.256279267773521e-06,
+      "loss": 0.3855,
+      "step": 10400
+    },
+    {
+      "epoch": 39.017701149425285,
+      "grad_norm": 0.0006430544308386743,
+      "learning_rate": 2.2477650063856962e-06,
+      "loss": 0.2692,
+      "step": 10410
+    },
+    {
+      "epoch": 39.01846743295019,
+      "grad_norm": 0.009712583385407925,
+      "learning_rate": 2.2392507449978717e-06,
+      "loss": 0.6717,
+      "step": 10420
+    },
+    {
+      "epoch": 39.01923371647509,
+      "grad_norm": 0.0006527869845740497,
+      "learning_rate": 2.230736483610047e-06,
+      "loss": 0.0016,
+      "step": 10430
+    },
+    {
+      "epoch": 39.02,
+      "grad_norm": 0.001249029883183539,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 0.5967,
+      "step": 10440
+    },
+    {
+      "epoch": 39.02,
+      "eval_accuracy": 0.6444444444444445,
+      "eval_loss": 2.596872568130493,
+      "eval_runtime": 14.1307,
+      "eval_samples_per_second": 3.185,
+      "eval_steps_per_second": 3.185,
+      "step": 10440
+    },
+    {
+      "epoch": 40.000766283524904,
+      "grad_norm": 0.0007803108892403543,
+      "learning_rate": 2.213707960834398e-06,
+      "loss": 0.8372,
+      "step": 10450
+    },
+    {
+      "epoch": 40.00153256704981,
+      "grad_norm": 0.0010762476595118642,
+      "learning_rate": 2.205193699446573e-06,
+      "loss": 0.0006,
+      "step": 10460
+    },
+    {
+      "epoch": 40.00229885057471,
+      "grad_norm": 0.004534687381237745,
+      "learning_rate": 2.1966794380587487e-06,
+      "loss": 0.6515,
+      "step": 10470
+    },
+    {
+      "epoch": 40.003065134099614,
+      "grad_norm": 0.0006635275203734636,
+      "learning_rate": 2.1881651766709237e-06,
+      "loss": 0.5376,
+      "step": 10480
+    },
+    {
+      "epoch": 40.00383141762452,
+      "grad_norm": 0.007556573953479528,
+      "learning_rate": 2.1796509152830996e-06,
+      "loss": 0.0631,
+      "step": 10490
+    },
+    {
+      "epoch": 40.00459770114943,
+      "grad_norm": 0.0008903385605663061,
+      "learning_rate": 2.1711366538952746e-06,
+      "loss": 0.6217,
+      "step": 10500
+    },
+    {
+      "epoch": 40.00536398467433,
+      "grad_norm": 0.13304153084754944,
+      "learning_rate": 2.1626223925074504e-06,
+      "loss": 0.0064,
+      "step": 10510
+    },
+    {
+      "epoch": 40.006130268199236,
+      "grad_norm": 0.1292513608932495,
+      "learning_rate": 2.1541081311196254e-06,
+      "loss": 0.2007,
+      "step": 10520
+    },
+    {
+      "epoch": 40.00689655172414,
+      "grad_norm": 0.001017851522192359,
+      "learning_rate": 2.145593869731801e-06,
+      "loss": 0.9946,
+      "step": 10530
+    },
+    {
+      "epoch": 40.00766283524904,
+      "grad_norm": 0.028370540589094162,
+      "learning_rate": 2.1370796083439762e-06,
+      "loss": 0.0004,
+      "step": 10540
+    },
+    {
+      "epoch": 40.00842911877395,
+      "grad_norm": 0.006354500073939562,
+      "learning_rate": 2.1285653469561517e-06,
+      "loss": 0.0455,
+      "step": 10550
+    },
+    {
+      "epoch": 40.00919540229885,
+      "grad_norm": 0.0005885363207198679,
+      "learning_rate": 2.120051085568327e-06,
+      "loss": 0.7072,
+      "step": 10560
+    },
+    {
+      "epoch": 40.009961685823754,
+      "grad_norm": 1.9478639364242554,
+      "learning_rate": 2.1115368241805025e-06,
+      "loss": 1.1553,
+      "step": 10570
+    },
+    {
+      "epoch": 40.01072796934866,
+      "grad_norm": 0.0016932955477386713,
+      "learning_rate": 2.103022562792678e-06,
+      "loss": 0.0008,
+      "step": 10580
+    },
+    {
+      "epoch": 40.01149425287356,
+      "grad_norm": 2.8671131134033203,
+      "learning_rate": 2.0945083014048533e-06,
+      "loss": 0.6408,
+      "step": 10590
+    },
+    {
+      "epoch": 40.012260536398465,
+      "grad_norm": 0.0005292880814522505,
+      "learning_rate": 2.0859940400170287e-06,
+      "loss": 0.0005,
+      "step": 10600
+    },
+    {
+      "epoch": 40.01302681992337,
+      "grad_norm": 0.17220841348171234,
+      "learning_rate": 2.077479778629204e-06,
+      "loss": 0.0007,
+      "step": 10610
+    },
+    {
+      "epoch": 40.01379310344828,
+      "grad_norm": 0.0016101287910714746,
+      "learning_rate": 2.0689655172413796e-06,
+      "loss": 0.0255,
+      "step": 10620
+    },
+    {
+      "epoch": 40.01455938697318,
+      "grad_norm": 0.09431593865156174,
+      "learning_rate": 2.060451255853555e-06,
+      "loss": 0.0056,
+      "step": 10630
+    },
+    {
+      "epoch": 40.015325670498086,
+      "grad_norm": 0.0005647214711643755,
+      "learning_rate": 2.0519369944657304e-06,
+      "loss": 0.0008,
+      "step": 10640
+    },
+    {
+      "epoch": 40.01609195402299,
+      "grad_norm": 0.0007362478063441813,
+      "learning_rate": 2.043422733077906e-06,
+      "loss": 0.4844,
+      "step": 10650
+    },
+    {
+      "epoch": 40.01685823754789,
+      "grad_norm": 0.0007467013783752918,
+      "learning_rate": 2.0349084716900813e-06,
+      "loss": 0.3836,
+      "step": 10660
+    },
+    {
+      "epoch": 40.0176245210728,
+      "grad_norm": 0.0005726607050746679,
+      "learning_rate": 2.0263942103022567e-06,
+      "loss": 0.0002,
+      "step": 10670
+    },
+    {
+      "epoch": 40.0183908045977,
+      "grad_norm": 0.030910883098840714,
+      "learning_rate": 2.0178799489144317e-06,
+      "loss": 0.0352,
+      "step": 10680
+    },
+    {
+      "epoch": 40.019157088122604,
+      "grad_norm": 0.001673361286520958,
+      "learning_rate": 2.009365687526607e-06,
+      "loss": 0.2036,
+      "step": 10690
+    },
+    {
+      "epoch": 40.01992337164751,
+      "grad_norm": 0.0005115558160468936,
+      "learning_rate": 2.0008514261387825e-06,
+      "loss": 0.0006,
+      "step": 10700
+    },
+    {
+      "epoch": 40.02,
+      "eval_accuracy": 0.6444444444444445,
+      "eval_loss": 2.638057231903076,
+      "eval_runtime": 14.1734,
+      "eval_samples_per_second": 3.175,
+      "eval_steps_per_second": 3.175,
+      "step": 10701
+    },
+    {
+      "epoch": 41.000689655172415,
+      "grad_norm": 0.013981521129608154,
+      "learning_rate": 1.992337164750958e-06,
+      "loss": 0.0002,
+      "step": 10710
+    },
+    {
+      "epoch": 41.00145593869732,
+      "grad_norm": 0.0017696694703772664,
+      "learning_rate": 1.9838229033631333e-06,
+      "loss": 0.0005,
+      "step": 10720
+    },
+    {
+      "epoch": 41.00222222222222,
+      "grad_norm": 0.0005197998834773898,
+      "learning_rate": 1.9753086419753087e-06,
+      "loss": 0.6036,
+      "step": 10730
+    },
+    {
+      "epoch": 41.002988505747126,
+      "grad_norm": 0.0005476609221659601,
+      "learning_rate": 1.966794380587484e-06,
+      "loss": 0.5618,
+      "step": 10740
+    },
+    {
+      "epoch": 41.00375478927203,
+      "grad_norm": 0.0012840996496379375,
+      "learning_rate": 1.9582801191996596e-06,
+      "loss": 0.0005,
+      "step": 10750
+    },
+    {
+      "epoch": 41.00452107279693,
+      "grad_norm": 0.0009479990694671869,
+      "learning_rate": 1.949765857811835e-06,
+      "loss": 0.0036,
+      "step": 10760
+    },
+    {
+      "epoch": 41.00528735632184,
+      "grad_norm": 0.04873271659016609,
+      "learning_rate": 1.9412515964240104e-06,
+      "loss": 0.1735,
+      "step": 10770
+    },
+    {
+      "epoch": 41.00605363984674,
+      "grad_norm": 0.14039330184459686,
+      "learning_rate": 1.932737335036186e-06,
+      "loss": 0.0056,
+      "step": 10780
+    },
+    {
+      "epoch": 41.006819923371644,
+      "grad_norm": 0.005437355488538742,
+      "learning_rate": 1.9242230736483612e-06,
+      "loss": 0.0154,
+      "step": 10790
+    },
+    {
+      "epoch": 41.007586206896555,
+      "grad_norm": 0.0047526853159070015,
+      "learning_rate": 1.9157088122605367e-06,
+      "loss": 0.3819,
+      "step": 10800
+    },
+    {
+      "epoch": 41.00835249042146,
+      "grad_norm": 0.007671783212572336,
+      "learning_rate": 1.9071945508727119e-06,
+      "loss": 0.0091,
+      "step": 10810
+    },
+    {
+      "epoch": 41.00911877394636,
+      "grad_norm": 1302.8955078125,
+      "learning_rate": 1.8986802894848875e-06,
+      "loss": 0.714,
+      "step": 10820
+    },
+    {
+      "epoch": 41.009885057471266,
+      "grad_norm": 0.0007232970674522221,
+      "learning_rate": 1.8901660280970627e-06,
+      "loss": 0.0,
+      "step": 10830
+    },
+    {
+      "epoch": 41.01065134099617,
+      "grad_norm": 0.0005007084691897035,
+      "learning_rate": 1.8816517667092381e-06,
+      "loss": 0.0005,
+      "step": 10840
+    },
+    {
+      "epoch": 41.01141762452107,
+      "grad_norm": 0.008196084760129452,
+      "learning_rate": 1.8731375053214135e-06,
+      "loss": 0.2376,
+      "step": 10850
+    },
+    {
+      "epoch": 41.01218390804598,
+      "grad_norm": 0.07016335427761078,
+      "learning_rate": 1.864623243933589e-06,
+      "loss": 1.0228,
+      "step": 10860
+    },
+    {
+      "epoch": 41.01295019157088,
+      "grad_norm": 0.10909241437911987,
+      "learning_rate": 1.8561089825457644e-06,
+      "loss": 0.297,
+      "step": 10870
+    },
+    {
+      "epoch": 41.013716475095784,
+      "grad_norm": 0.057047039270401,
+      "learning_rate": 1.8475947211579398e-06,
+      "loss": 0.001,
+      "step": 10880
+    },
+    {
+      "epoch": 41.01448275862069,
+      "grad_norm": 0.0005653087282553315,
+      "learning_rate": 1.839080459770115e-06,
+      "loss": 0.6173,
+      "step": 10890
+    },
+    {
+      "epoch": 41.01524904214559,
+      "grad_norm": 0.033106788992881775,
+      "learning_rate": 1.8305661983822906e-06,
+      "loss": 0.0074,
+      "step": 10900
+    },
+    {
+      "epoch": 41.0160153256705,
+      "grad_norm": 0.0005475867656059563,
+      "learning_rate": 1.8220519369944658e-06,
+      "loss": 0.0063,
+      "step": 10910
+    },
+    {
+      "epoch": 41.016781609195405,
+      "grad_norm": 0.053291477262973785,
+      "learning_rate": 1.8135376756066415e-06,
+      "loss": 0.0004,
+      "step": 10920
+    },
+    {
+      "epoch": 41.01754789272031,
+      "grad_norm": 69.3751449584961,
+      "learning_rate": 1.8050234142188167e-06,
+      "loss": 0.9711,
+      "step": 10930
+    },
+    {
+      "epoch": 41.01831417624521,
+      "grad_norm": 0.09939111769199371,
+      "learning_rate": 1.7965091528309919e-06,
+      "loss": 0.0019,
+      "step": 10940
+    },
+    {
+      "epoch": 41.019080459770116,
+      "grad_norm": 0.14862337708473206,
+      "learning_rate": 1.7879948914431675e-06,
+      "loss": 0.0394,
+      "step": 10950
+    },
+    {
+      "epoch": 41.01984674329502,
+      "grad_norm": 0.0012591223930940032,
+      "learning_rate": 1.7794806300553427e-06,
+      "loss": 0.0004,
+      "step": 10960
+    },
+    {
+      "epoch": 41.02,
+      "eval_accuracy": 0.6222222222222222,
+      "eval_loss": 2.9590911865234375,
+      "eval_runtime": 15.8526,
+      "eval_samples_per_second": 2.839,
+      "eval_steps_per_second": 2.839,
+      "step": 10962
+    },
+    {
+      "epoch": 42.00061302681992,
+      "grad_norm": 0.001878397073596716,
+      "learning_rate": 1.7709663686675183e-06,
+      "loss": 0.9958,
+      "step": 10970
+    },
+    {
+      "epoch": 42.00137931034483,
+      "grad_norm": 0.009673736989498138,
+      "learning_rate": 1.7624521072796935e-06,
+      "loss": 0.1777,
+      "step": 10980
+    },
+    {
+      "epoch": 42.002145593869734,
+      "grad_norm": 0.001056146458722651,
+      "learning_rate": 1.753937845891869e-06,
+      "loss": 0.0004,
+      "step": 10990
+    },
+    {
+      "epoch": 42.00291187739464,
+      "grad_norm": 0.0020266796927899122,
+      "learning_rate": 1.7454235845040444e-06,
+      "loss": 0.0007,
+      "step": 11000
+    },
+    {
+      "epoch": 42.00367816091954,
+      "grad_norm": 0.0008849030709825456,
+      "learning_rate": 1.7369093231162198e-06,
+      "loss": 0.0004,
+      "step": 11010
+    },
+    {
+      "epoch": 42.004444444444445,
+      "grad_norm": 0.0004981319652870297,
+      "learning_rate": 1.7283950617283952e-06,
+      "loss": 0.0052,
+      "step": 11020
+    },
+    {
+      "epoch": 42.00521072796935,
+      "grad_norm": 0.001915338565595448,
+      "learning_rate": 1.7198808003405706e-06,
+      "loss": 0.6275,
+      "step": 11030
+    },
+    {
+      "epoch": 42.00597701149425,
+      "grad_norm": 0.0005534188239835203,
+      "learning_rate": 1.7113665389527458e-06,
+      "loss": 0.0006,
+      "step": 11040
+    },
+    {
+      "epoch": 42.006743295019156,
+      "grad_norm": 397.24444580078125,
+      "learning_rate": 1.7028522775649215e-06,
+      "loss": 0.0332,
+      "step": 11050
+    },
+    {
+      "epoch": 42.00750957854406,
+      "grad_norm": 0.002358200028538704,
+      "learning_rate": 1.6943380161770967e-06,
+      "loss": 0.0001,
+      "step": 11060
+    },
+    {
+      "epoch": 42.00827586206896,
+      "grad_norm": 0.09730993211269379,
+      "learning_rate": 1.6858237547892723e-06,
+      "loss": 0.0003,
+      "step": 11070
+    },
+    {
+      "epoch": 42.00904214559387,
+      "grad_norm": 0.002618571277707815,
+      "learning_rate": 1.6773094934014475e-06,
+      "loss": 0.7586,
+      "step": 11080
+    },
+    {
+      "epoch": 42.00980842911878,
+      "grad_norm": 0.010873343795537949,
+      "learning_rate": 1.6687952320136231e-06,
+      "loss": 0.0015,
+      "step": 11090
+    },
+    {
+      "epoch": 42.01057471264368,
+      "grad_norm": 0.0019932216964662075,
+      "learning_rate": 1.6602809706257983e-06,
+      "loss": 0.1034,
+      "step": 11100
+    },
+    {
+      "epoch": 42.011340996168585,
+      "grad_norm": 0.0022048912942409515,
+      "learning_rate": 1.6517667092379737e-06,
+      "loss": 0.1463,
+      "step": 11110
+    },
+    {
+      "epoch": 42.01210727969349,
+      "grad_norm": 6.6127095222473145,
+      "learning_rate": 1.6432524478501492e-06,
+      "loss": 0.0007,
+      "step": 11120
+    },
+    {
+      "epoch": 42.01287356321839,
+      "grad_norm": 1.7690030336380005,
+      "learning_rate": 1.6347381864623246e-06,
+      "loss": 0.6631,
+      "step": 11130
+    },
+    {
+      "epoch": 42.013639846743295,
+      "grad_norm": 0.0005693206912837923,
+      "learning_rate": 1.6262239250744998e-06,
+      "loss": 0.0003,
+      "step": 11140
+    },
+    {
+      "epoch": 42.0144061302682,
+      "grad_norm": 0.0005519501864910126,
+      "learning_rate": 1.6177096636866754e-06,
+      "loss": 0.5764,
+      "step": 11150
+    },
+    {
+      "epoch": 42.0151724137931,
+      "grad_norm": 0.0006714254850521684,
+      "learning_rate": 1.6091954022988506e-06,
+      "loss": 0.6308,
+      "step": 11160
+    },
+    {
+      "epoch": 42.015938697318006,
+      "grad_norm": 0.0019181851530447602,
+      "learning_rate": 1.6006811409110262e-06,
+      "loss": 0.0038,
+      "step": 11170
+    },
+    {
+      "epoch": 42.01670498084291,
+      "grad_norm": 0.0021259617060422897,
+      "learning_rate": 1.5921668795232015e-06,
+      "loss": 0.0024,
+      "step": 11180
+    },
+    {
+      "epoch": 42.01747126436781,
+      "grad_norm": 0.0013196076033636928,
+      "learning_rate": 1.583652618135377e-06,
+      "loss": 0.5949,
+      "step": 11190
+    },
+    {
+      "epoch": 42.01823754789272,
+      "grad_norm": 0.0004218143585603684,
+      "learning_rate": 1.5751383567475523e-06,
+      "loss": 0.0001,
+      "step": 11200
+    },
+    {
+      "epoch": 42.01900383141763,
+      "grad_norm": 0.00047586203436367214,
+      "learning_rate": 1.5666240953597275e-06,
+      "loss": 0.5043,
+      "step": 11210
+    },
+    {
+      "epoch": 42.01977011494253,
+      "grad_norm": 0.07100889831781387,
+      "learning_rate": 1.5581098339719031e-06,
+      "loss": 0.0004,
+      "step": 11220
+    },
+    {
+      "epoch": 42.02,
+      "eval_accuracy": 0.7111111111111111,
+      "eval_loss": 2.12404727935791,
+      "eval_runtime": 16.4348,
+      "eval_samples_per_second": 2.738,
+      "eval_steps_per_second": 2.738,
+      "step": 11223
+    },
+    {
+      "epoch": 43.00053639846743,
+      "grad_norm": 0.0005287248059175909,
+      "learning_rate": 1.5495955725840783e-06,
+      "loss": 0.6208,
+      "step": 11230
+    },
+    {
+      "epoch": 43.001302681992335,
+      "grad_norm": 0.015506441704928875,
+      "learning_rate": 1.541081311196254e-06,
+      "loss": 0.0003,
+      "step": 11240
+    },
+    {
+      "epoch": 43.00206896551724,
+      "grad_norm": 272.2698974609375,
+      "learning_rate": 1.5325670498084292e-06,
+      "loss": 1.0295,
+      "step": 11250
+    },
+    {
+      "epoch": 43.00283524904214,
+      "grad_norm": 0.05146227031946182,
+      "learning_rate": 1.5240527884206046e-06,
+      "loss": 0.6084,
+      "step": 11260
+    },
+    {
+      "epoch": 43.00360153256705,
+      "grad_norm": 103.00202178955078,
+      "learning_rate": 1.51553852703278e-06,
+      "loss": 1.2964,
+      "step": 11270
+    },
+    {
+      "epoch": 43.00436781609196,
+      "grad_norm": 0.0007857238524593413,
+      "learning_rate": 1.5070242656449554e-06,
+      "loss": 0.0002,
+      "step": 11280
+    },
+    {
+      "epoch": 43.00513409961686,
+      "grad_norm": 0.0006366973393596709,
+      "learning_rate": 1.4985100042571306e-06,
+      "loss": 0.0002,
+      "step": 11290
+    },
+    {
+      "epoch": 43.005900383141764,
+      "grad_norm": 0.006167873274534941,
+      "learning_rate": 1.4899957428693062e-06,
+      "loss": 0.0002,
+      "step": 11300
+    },
+    {
+      "epoch": 43.00666666666667,
+      "grad_norm": 0.009739157743752003,
+      "learning_rate": 1.4814814814814815e-06,
+      "loss": 0.7216,
+      "step": 11310
+    },
+    {
+      "epoch": 43.00743295019157,
+      "grad_norm": 1272.1881103515625,
+      "learning_rate": 1.472967220093657e-06,
+      "loss": 0.066,
+      "step": 11320
+    },
+    {
+      "epoch": 43.008199233716475,
+      "grad_norm": 0.0007570978486910462,
+      "learning_rate": 1.4644529587058323e-06,
+      "loss": 0.5015,
+      "step": 11330
+    },
+    {
+      "epoch": 43.00896551724138,
+      "grad_norm": 0.0004982962273061275,
+      "learning_rate": 1.455938697318008e-06,
+      "loss": 0.5787,
+      "step": 11340
+    },
+    {
+      "epoch": 43.00973180076628,
+      "grad_norm": 0.02334807999432087,
+      "learning_rate": 1.4474244359301831e-06,
+      "loss": 0.0005,
+      "step": 11350
+    },
+    {
+      "epoch": 43.010498084291186,
+      "grad_norm": 0.000609906914178282,
+      "learning_rate": 1.4389101745423588e-06,
+      "loss": 0.0006,
+      "step": 11360
+    },
+    {
+      "epoch": 43.01126436781609,
+      "grad_norm": 0.04427050054073334,
+      "learning_rate": 1.430395913154534e-06,
+      "loss": 0.6396,
+      "step": 11370
+    },
+    {
+      "epoch": 43.01203065134099,
+      "grad_norm": 0.11915020644664764,
+      "learning_rate": 1.4218816517667094e-06,
+      "loss": 1.1079,
+      "step": 11380
+    },
+    {
+      "epoch": 43.012796934865904,
+      "grad_norm": 176.30267333984375,
+      "learning_rate": 1.4133673903788848e-06,
+      "loss": 0.4117,
+      "step": 11390
+    },
+    {
+      "epoch": 43.01356321839081,
+      "grad_norm": 0.0005373156745918095,
+      "learning_rate": 1.4048531289910602e-06,
+      "loss": 0.0022,
+      "step": 11400
+    },
+    {
+      "epoch": 43.01432950191571,
+      "grad_norm": 0.014790279790759087,
+      "learning_rate": 1.3963388676032354e-06,
+      "loss": 0.0007,
+      "step": 11410
+    },
+    {
+      "epoch": 43.015095785440614,
+      "grad_norm": 0.0004909559502266347,
+      "learning_rate": 1.387824606215411e-06,
+      "loss": 1.1803,
+      "step": 11420
+    },
+    {
+      "epoch": 43.01586206896552,
+      "grad_norm": 0.0029413867741823196,
+      "learning_rate": 1.3793103448275862e-06,
+      "loss": 0.0001,
+      "step": 11430
+    },
+    {
+      "epoch": 43.01662835249042,
+      "grad_norm": 0.0030794162303209305,
+      "learning_rate": 1.3707960834397619e-06,
+      "loss": 0.8496,
+      "step": 11440
+    },
+    {
+      "epoch": 43.017394636015325,
+      "grad_norm": 0.10320518910884857,
+      "learning_rate": 1.362281822051937e-06,
+      "loss": 1.1705,
+      "step": 11450
+    },
+    {
+      "epoch": 43.01816091954023,
+      "grad_norm": 0.002483441261574626,
+      "learning_rate": 1.3537675606641127e-06,
+      "loss": 0.6525,
+      "step": 11460
+    },
+    {
+      "epoch": 43.01892720306513,
+      "grad_norm": 0.000820959045086056,
+      "learning_rate": 1.345253299276288e-06,
+      "loss": 0.0315,
+      "step": 11470
+    },
+    {
+      "epoch": 43.019693486590036,
+      "grad_norm": 0.0005174216348677874,
+      "learning_rate": 1.3367390378884631e-06,
+      "loss": 1.109,
+      "step": 11480
+    },
+    {
+      "epoch": 43.02,
+      "eval_accuracy": 0.6,
+      "eval_loss": 2.763355016708374,
+      "eval_runtime": 16.1842,
+      "eval_samples_per_second": 2.78,
+      "eval_steps_per_second": 2.78,
+      "step": 11484
+    },
+    {
+      "epoch": 44.000459770114944,
+      "grad_norm": 0.0006952774128876626,
+      "learning_rate": 1.3282247765006387e-06,
+      "loss": 0.0002,
+      "step": 11490
+    },
+    {
+      "epoch": 44.00122605363985,
+      "grad_norm": 0.02722359634935856,
+      "learning_rate": 1.319710515112814e-06,
+      "loss": 1.6376,
+      "step": 11500
+    },
+    {
+      "epoch": 44.00199233716475,
+      "grad_norm": 0.006522133480757475,
+      "learning_rate": 1.3111962537249896e-06,
+      "loss": 0.0,
+      "step": 11510
+    },
+    {
+      "epoch": 44.002758620689654,
+      "grad_norm": 0.03956187888979912,
+      "learning_rate": 1.3026819923371648e-06,
+      "loss": 0.0214,
+      "step": 11520
+    },
+    {
+      "epoch": 44.00352490421456,
+      "grad_norm": 0.11104047298431396,
+      "learning_rate": 1.2941677309493402e-06,
+      "loss": 0.7209,
+      "step": 11530
+    },
+    {
+      "epoch": 44.00429118773946,
+      "grad_norm": 0.0005099502741359174,
+      "learning_rate": 1.2856534695615156e-06,
+      "loss": 0.3281,
+      "step": 11540
+    },
+    {
+      "epoch": 44.005057471264365,
+      "grad_norm": 0.0005218337755650282,
+      "learning_rate": 1.277139208173691e-06,
+      "loss": 0.0006,
+      "step": 11550
+    },
+    {
+      "epoch": 44.00582375478927,
+      "grad_norm": 0.0004927856498397887,
+      "learning_rate": 1.2686249467858662e-06,
+      "loss": 0.0002,
+      "step": 11560
+    },
+    {
+      "epoch": 44.00659003831418,
+      "grad_norm": 12.864944458007812,
+      "learning_rate": 1.2601106853980419e-06,
+      "loss": 0.0027,
+      "step": 11570
+    },
+    {
+      "epoch": 44.00735632183908,
+      "grad_norm": 0.004535001236945391,
+      "learning_rate": 1.251596424010217e-06,
+      "loss": 0.0004,
+      "step": 11580
+    },
+    {
+      "epoch": 44.00812260536399,
+      "grad_norm": 0.04480987787246704,
+      "learning_rate": 1.2430821626223927e-06,
+      "loss": 0.0016,
+      "step": 11590
+    },
+    {
+      "epoch": 44.00888888888889,
+      "grad_norm": 10.11349868774414,
+      "learning_rate": 1.234567901234568e-06,
+      "loss": 0.0017,
+      "step": 11600
+    },
+    {
+      "epoch": 44.009655172413794,
+      "grad_norm": 0.000782816088758409,
+      "learning_rate": 1.2260536398467433e-06,
+      "loss": 0.3825,
+      "step": 11610
+    },
+    {
+      "epoch": 44.0104214559387,
+      "grad_norm": 0.0004511185979936272,
+      "learning_rate": 1.2175393784589187e-06,
+      "loss": 1.0973,
+      "step": 11620
+    },
+    {
+      "epoch": 44.0111877394636,
+      "grad_norm": 0.0004485390963964164,
+      "learning_rate": 1.2090251170710942e-06,
+      "loss": 0.0009,
+      "step": 11630
+    },
+    {
+      "epoch": 44.011954022988505,
+      "grad_norm": 0.0005191498785279691,
+      "learning_rate": 1.2005108556832696e-06,
+      "loss": 0.0009,
+      "step": 11640
+    },
+    {
+      "epoch": 44.01272030651341,
+      "grad_norm": 0.0012321961112320423,
+      "learning_rate": 1.191996594295445e-06,
+      "loss": 0.0004,
+      "step": 11650
+    },
+    {
+      "epoch": 44.01348659003831,
+      "grad_norm": 0.006523193791508675,
+      "learning_rate": 1.1834823329076204e-06,
+      "loss": 0.541,
+      "step": 11660
+    },
+    {
+      "epoch": 44.014252873563215,
+      "grad_norm": 0.001767021487466991,
+      "learning_rate": 1.1749680715197956e-06,
+      "loss": 0.0001,
+      "step": 11670
+    },
+    {
+      "epoch": 44.01501915708812,
+      "grad_norm": 0.0005826118867844343,
+      "learning_rate": 1.166453810131971e-06,
+      "loss": 0.0003,
+      "step": 11680
+    },
+    {
+      "epoch": 44.01578544061303,
+      "grad_norm": 0.0010437779128551483,
+      "learning_rate": 1.1579395487441465e-06,
+      "loss": 0.0001,
+      "step": 11690
+    },
+    {
+      "epoch": 44.01655172413793,
+      "grad_norm": 0.0007777928840368986,
+      "learning_rate": 1.1494252873563219e-06,
+      "loss": 0.3824,
+      "step": 11700
+    },
+    {
+      "epoch": 44.01731800766284,
+      "grad_norm": 0.0009876244002953172,
+      "learning_rate": 1.1409110259684973e-06,
+      "loss": 0.0002,
+      "step": 11710
+    },
+    {
+      "epoch": 44.01808429118774,
+      "grad_norm": 0.007304656784981489,
+      "learning_rate": 1.1323967645806727e-06,
+      "loss": 0.55,
+      "step": 11720
+    },
+    {
+      "epoch": 44.018850574712644,
+      "grad_norm": 0.12421630322933197,
+      "learning_rate": 1.1238825031928481e-06,
+      "loss": 0.0326,
+      "step": 11730
+    },
+    {
+      "epoch": 44.01961685823755,
+      "grad_norm": 2.9402077198028564,
+      "learning_rate": 1.1153682418050235e-06,
+      "loss": 1.074,
+      "step": 11740
+    },
+    {
+      "epoch": 44.02,
+      "eval_accuracy": 0.7111111111111111,
+      "eval_loss": 2.1594574451446533,
+      "eval_runtime": 16.432,
+      "eval_samples_per_second": 2.739,
+      "eval_steps_per_second": 2.739,
+      "step": 11745
+    },
+    {
+      "epoch": 45.000383141762455,
+      "grad_norm": 0.0022075134329497814,
+      "learning_rate": 1.106853980417199e-06,
+      "loss": 0.0002,
+      "step": 11750
+    },
+    {
+      "epoch": 45.00114942528736,
+      "grad_norm": 0.0017397621413692832,
+      "learning_rate": 1.0983397190293744e-06,
+      "loss": 0.5031,
+      "step": 11760
+    },
+    {
+      "epoch": 45.00191570881226,
+      "grad_norm": 0.0005939102848060429,
+      "learning_rate": 1.0898254576415498e-06,
+      "loss": 0.0004,
+      "step": 11770
+    },
+    {
+      "epoch": 45.002681992337166,
+      "grad_norm": 0.00116265332326293,
+      "learning_rate": 1.0813111962537252e-06,
+      "loss": 0.0004,
+      "step": 11780
+    },
+    {
+      "epoch": 45.00344827586207,
+      "grad_norm": 0.0006234461907297373,
+      "learning_rate": 1.0727969348659004e-06,
+      "loss": 0.7825,
+      "step": 11790
+    },
+    {
+      "epoch": 45.00421455938697,
+      "grad_norm": 806.7313842773438,
+      "learning_rate": 1.0642826734780758e-06,
+      "loss": 0.3269,
+      "step": 11800
+    },
+    {
+      "epoch": 45.00498084291188,
+      "grad_norm": 0.00044658681144937873,
+      "learning_rate": 1.0557684120902512e-06,
+      "loss": 0.4785,
+      "step": 11810
+    },
+    {
+      "epoch": 45.00574712643678,
+      "grad_norm": 0.0012156810844317079,
+      "learning_rate": 1.0472541507024267e-06,
+      "loss": 0.5853,
+      "step": 11820
+    },
+    {
+      "epoch": 45.006513409961684,
+      "grad_norm": 0.00045517139369621873,
+      "learning_rate": 1.038739889314602e-06,
+      "loss": 0.0044,
+      "step": 11830
+    },
+    {
+      "epoch": 45.00727969348659,
+      "grad_norm": 0.0006456039263866842,
+      "learning_rate": 1.0302256279267775e-06,
+      "loss": 0.0009,
+      "step": 11840
+    },
+    {
+      "epoch": 45.00804597701149,
+      "grad_norm": 0.0011488504242151976,
+      "learning_rate": 1.021711366538953e-06,
+      "loss": 0.7269,
+      "step": 11850
+    },
+    {
+      "epoch": 45.008812260536395,
+      "grad_norm": 0.0012480664299800992,
+      "learning_rate": 1.0131971051511283e-06,
+      "loss": 0.7995,
+      "step": 11860
+    },
+    {
+      "epoch": 45.009578544061306,
+      "grad_norm": 0.09593397378921509,
+      "learning_rate": 1.0046828437633035e-06,
+      "loss": 0.6449,
+      "step": 11870
+    },
+    {
+      "epoch": 45.01034482758621,
+      "grad_norm": 0.44798505306243896,
+      "learning_rate": 9.96168582375479e-07,
+      "loss": 0.0003,
+      "step": 11880
+    },
+    {
+      "epoch": 45.01111111111111,
+      "grad_norm": 0.004326379392296076,
+      "learning_rate": 9.876543209876544e-07,
+      "loss": 0.4509,
+      "step": 11890
+    },
+    {
+      "epoch": 45.01187739463602,
+      "grad_norm": 0.010555016808211803,
+      "learning_rate": 9.791400595998298e-07,
+      "loss": 0.0005,
+      "step": 11900
+    },
+    {
+      "epoch": 45.01264367816092,
+      "grad_norm": 0.0004949315916746855,
+      "learning_rate": 9.706257982120052e-07,
+      "loss": 0.0003,
+      "step": 11910
+    },
+    {
+      "epoch": 45.013409961685824,
+      "grad_norm": 0.001112055848352611,
+      "learning_rate": 9.621115368241806e-07,
+      "loss": 0.0002,
+      "step": 11920
+    },
+    {
+      "epoch": 45.01417624521073,
+      "grad_norm": 0.1513567864894867,
+      "learning_rate": 9.535972754363559e-07,
+      "loss": 0.001,
+      "step": 11930
+    },
+    {
+      "epoch": 45.01494252873563,
+      "grad_norm": 0.003296300768852234,
+      "learning_rate": 9.450830140485314e-07,
+      "loss": 0.0001,
+      "step": 11940
+    },
+    {
+      "epoch": 45.015708812260534,
+      "grad_norm": 0.07278383523225784,
+      "learning_rate": 9.365687526607068e-07,
+      "loss": 0.9561,
+      "step": 11950
+    },
+    {
+      "epoch": 45.01647509578544,
+      "grad_norm": 0.0007136106723919511,
+      "learning_rate": 9.280544912728822e-07,
+      "loss": 0.0005,
+      "step": 11960
+    },
+    {
+      "epoch": 45.01724137931034,
+      "grad_norm": 0.015829024836421013,
+      "learning_rate": 9.195402298850575e-07,
+      "loss": 0.7246,
+      "step": 11970
+    },
+    {
+      "epoch": 45.01800766283525,
+      "grad_norm": 334.7724914550781,
+      "learning_rate": 9.110259684972329e-07,
+      "loss": 0.4681,
+      "step": 11980
+    },
+    {
+      "epoch": 45.018773946360156,
+      "grad_norm": 0.00398367689922452,
+      "learning_rate": 9.025117071094083e-07,
+      "loss": 0.7101,
+      "step": 11990
+    },
+    {
+      "epoch": 45.01954022988506,
+      "grad_norm": 0.04738711565732956,
+      "learning_rate": 8.939974457215837e-07,
+      "loss": 0.0016,
+      "step": 12000
+    },
+    {
+      "epoch": 45.02,
+      "eval_accuracy": 0.7555555555555555,
+      "eval_loss": 1.853808879852295,
+      "eval_runtime": 16.4639,
+      "eval_samples_per_second": 2.733,
+      "eval_steps_per_second": 2.733,
+      "step": 12006
+    },
+    {
+      "epoch": 46.00030651340996,
+      "grad_norm": 0.0007470714044757187,
+      "learning_rate": 8.854831843337592e-07,
+      "loss": 0.0004,
+      "step": 12010
+    },
+    {
+      "epoch": 46.001072796934864,
+      "grad_norm": 0.008812682703137398,
+      "learning_rate": 8.769689229459345e-07,
+      "loss": 0.0366,
+      "step": 12020
+    },
+    {
+      "epoch": 46.00183908045977,
+      "grad_norm": 0.08954495936632156,
+      "learning_rate": 8.684546615581099e-07,
+      "loss": 0.0008,
+      "step": 12030
+    },
+    {
+      "epoch": 46.00260536398467,
+      "grad_norm": 0.016129348427057266,
+      "learning_rate": 8.599404001702853e-07,
+      "loss": 0.0002,
+      "step": 12040
+    },
+    {
+      "epoch": 46.00337164750958,
+      "grad_norm": 0.015925360843539238,
+      "learning_rate": 8.514261387824607e-07,
+      "loss": 0.0001,
+      "step": 12050
+    },
+    {
+      "epoch": 46.004137931034485,
+      "grad_norm": 0.004947807639837265,
+      "learning_rate": 8.429118773946361e-07,
+      "loss": 0.0003,
+      "step": 12060
+    },
+    {
+      "epoch": 46.00490421455939,
+      "grad_norm": 0.038208961486816406,
+      "learning_rate": 8.343976160068116e-07,
+      "loss": 0.5025,
+      "step": 12070
+    },
+    {
+      "epoch": 46.00567049808429,
+      "grad_norm": 0.05429494380950928,
+      "learning_rate": 8.258833546189869e-07,
+      "loss": 0.7446,
+      "step": 12080
+    },
+    {
+      "epoch": 46.006436781609196,
+      "grad_norm": 0.005066359415650368,
+      "learning_rate": 8.173690932311623e-07,
+      "loss": 0.0002,
+      "step": 12090
+    },
+    {
+      "epoch": 46.0072030651341,
+      "grad_norm": 0.0029733364935964346,
+      "learning_rate": 8.088548318433377e-07,
+      "loss": 0.5126,
+      "step": 12100
+    },
+    {
+      "epoch": 46.007969348659,
+      "grad_norm": 0.30708321928977966,
+      "learning_rate": 8.003405704555131e-07,
+      "loss": 0.0009,
+      "step": 12110
+    },
+    {
+      "epoch": 46.00873563218391,
+      "grad_norm": 0.006712548900395632,
+      "learning_rate": 7.918263090676885e-07,
+      "loss": 1.0584,
+      "step": 12120
+    },
+    {
+      "epoch": 46.00950191570881,
+      "grad_norm": 0.0009609467815607786,
+      "learning_rate": 7.833120476798637e-07,
+      "loss": 0.0413,
+      "step": 12130
+    },
+    {
+      "epoch": 46.010268199233714,
+      "grad_norm": 0.008377721533179283,
+      "learning_rate": 7.747977862920392e-07,
+      "loss": 0.0002,
+      "step": 12140
+    },
+    {
+      "epoch": 46.01103448275862,
+      "grad_norm": 0.0004620867548510432,
+      "learning_rate": 7.662835249042146e-07,
+      "loss": 0.0002,
+      "step": 12150
+    },
+    {
+      "epoch": 46.01180076628353,
+      "grad_norm": 0.002079889178276062,
+      "learning_rate": 7.5776926351639e-07,
+      "loss": 0.0005,
+      "step": 12160
+    },
+    {
+      "epoch": 46.01256704980843,
+      "grad_norm": 0.0010952705051749945,
+      "learning_rate": 7.492550021285653e-07,
+      "loss": 1.1892,
+      "step": 12170
+    },
+    {
+      "epoch": 46.013333333333335,
+      "grad_norm": 0.000436866917880252,
+      "learning_rate": 7.407407407407407e-07,
+      "loss": 0.8613,
+      "step": 12180
+    },
+    {
+      "epoch": 46.01409961685824,
+      "grad_norm": 0.023280123248696327,
+      "learning_rate": 7.322264793529161e-07,
+      "loss": 0.001,
+      "step": 12190
+    },
+    {
+      "epoch": 46.01486590038314,
+      "grad_norm": 0.0006040579755790532,
+      "learning_rate": 7.237122179650916e-07,
+      "loss": 0.0006,
+      "step": 12200
+    },
+    {
+      "epoch": 46.015632183908046,
+      "grad_norm": 1269.834228515625,
+      "learning_rate": 7.15197956577267e-07,
+      "loss": 0.3661,
+      "step": 12210
+    },
+    {
+      "epoch": 46.01639846743295,
+      "grad_norm": 0.028524765744805336,
+      "learning_rate": 7.066836951894424e-07,
+      "loss": 0.0003,
+      "step": 12220
+    },
+    {
+      "epoch": 46.01716475095785,
+      "grad_norm": 0.006299199536442757,
+      "learning_rate": 6.981694338016177e-07,
+      "loss": 0.6344,
+      "step": 12230
+    },
+    {
+      "epoch": 46.01793103448276,
+      "grad_norm": 0.0004130868474021554,
+      "learning_rate": 6.896551724137931e-07,
+      "loss": 0.606,
+      "step": 12240
+    },
+    {
+      "epoch": 46.01869731800766,
+      "grad_norm": 0.006240964867174625,
+      "learning_rate": 6.811409110259685e-07,
+      "loss": 0.0001,
+      "step": 12250
+    },
+    {
+      "epoch": 46.019463601532564,
+      "grad_norm": 0.004726331681013107,
+      "learning_rate": 6.72626649638144e-07,
+      "loss": 0.0002,
+      "step": 12260
+    },
+    {
+      "epoch": 46.02,
+      "eval_accuracy": 0.7555555555555555,
+      "eval_loss": 1.9418870210647583,
+      "eval_runtime": 15.2295,
+      "eval_samples_per_second": 2.955,
+      "eval_steps_per_second": 2.955,
+      "step": 12267
+    },
+    {
+      "epoch": 47.00022988505747,
+      "grad_norm": 0.0046911342069506645,
+      "learning_rate": 6.641123882503194e-07,
+      "loss": 0.0022,
+      "step": 12270
+    },
+    {
+      "epoch": 47.000996168582375,
+      "grad_norm": 36.75137710571289,
+      "learning_rate": 6.555981268624948e-07,
+      "loss": 1.1365,
+      "step": 12280
+    },
+    {
+      "epoch": 47.00176245210728,
+      "grad_norm": 0.03164047747850418,
+      "learning_rate": 6.470838654746701e-07,
+      "loss": 0.0012,
+      "step": 12290
+    },
+    {
+      "epoch": 47.00252873563218,
+      "grad_norm": 114.06194305419922,
+      "learning_rate": 6.385696040868455e-07,
+      "loss": 0.5134,
+      "step": 12300
+    },
+    {
+      "epoch": 47.003295019157086,
+      "grad_norm": 0.004502067808061838,
+      "learning_rate": 6.300553426990209e-07,
+      "loss": 0.0009,
+      "step": 12310
+    },
+    {
+      "epoch": 47.00406130268199,
+      "grad_norm": 0.004849694203585386,
+      "learning_rate": 6.215410813111964e-07,
+      "loss": 0.6976,
+      "step": 12320
+    },
+    {
+      "epoch": 47.00482758620689,
+      "grad_norm": 0.0008688484085723758,
+      "learning_rate": 6.130268199233717e-07,
+      "loss": 0.4239,
+      "step": 12330
+    },
+    {
+      "epoch": 47.005593869731804,
+      "grad_norm": 0.0006198923219926655,
+      "learning_rate": 6.045125585355471e-07,
+      "loss": 0.5128,
+      "step": 12340
+    },
+    {
+      "epoch": 47.00636015325671,
+      "grad_norm": 0.0004461881471797824,
+      "learning_rate": 5.959982971477225e-07,
+      "loss": 0.0003,
+      "step": 12350
+    },
+    {
+      "epoch": 47.00712643678161,
+      "grad_norm": 0.005352925043553114,
+      "learning_rate": 5.874840357598978e-07,
+      "loss": 0.0004,
+      "step": 12360
+    },
+    {
+      "epoch": 47.007892720306515,
+      "grad_norm": 0.00048328927368856966,
+      "learning_rate": 5.789697743720732e-07,
+      "loss": 0.0001,
+      "step": 12370
+    },
+    {
+      "epoch": 47.00865900383142,
+      "grad_norm": 0.0007524097454734147,
+      "learning_rate": 5.704555129842486e-07,
+      "loss": 0.0016,
+      "step": 12380
+    },
+    {
+      "epoch": 47.00942528735632,
+      "grad_norm": 17.501802444458008,
+      "learning_rate": 5.619412515964241e-07,
+      "loss": 0.0019,
+      "step": 12390
+    },
+    {
+      "epoch": 47.010191570881226,
+      "grad_norm": 0.0005570445209741592,
+      "learning_rate": 5.534269902085995e-07,
+      "loss": 0.0004,
+      "step": 12400
+    },
+    {
+      "epoch": 47.01095785440613,
+      "grad_norm": 0.002129359170794487,
+      "learning_rate": 5.449127288207749e-07,
+      "loss": 0.0003,
+      "step": 12410
+    },
+    {
+      "epoch": 47.01172413793103,
+      "grad_norm": 0.0004879856714978814,
+      "learning_rate": 5.363984674329502e-07,
+      "loss": 0.0002,
+      "step": 12420
+    },
+    {
+      "epoch": 47.01249042145594,
+      "grad_norm": 0.00046909274533391,
+      "learning_rate": 5.278842060451256e-07,
+      "loss": 0.3502,
+      "step": 12430
+    },
+    {
+      "epoch": 47.01325670498084,
+      "grad_norm": 0.03282985836267471,
+      "learning_rate": 5.19369944657301e-07,
+      "loss": 0.4477,
+      "step": 12440
+    },
+    {
+      "epoch": 47.014022988505744,
+      "grad_norm": 0.000536035280674696,
+      "learning_rate": 5.108556832694765e-07,
+      "loss": 0.0009,
+      "step": 12450
+    },
+    {
+      "epoch": 47.014789272030654,
+      "grad_norm": 0.0028050511609762907,
+      "learning_rate": 5.023414218816518e-07,
+      "loss": 0.0005,
+      "step": 12460
+    },
+    {
+      "epoch": 47.01555555555556,
+      "grad_norm": 0.12039543688297272,
+      "learning_rate": 4.938271604938272e-07,
+      "loss": 0.0112,
+      "step": 12470
+    },
+    {
+      "epoch": 47.01632183908046,
+      "grad_norm": 0.11689059436321259,
+      "learning_rate": 4.853128991060026e-07,
+      "loss": 0.0003,
+      "step": 12480
+    },
+    {
+      "epoch": 47.017088122605365,
+      "grad_norm": 0.004974637646228075,
+      "learning_rate": 4.7679863771817797e-07,
+      "loss": 0.0002,
+      "step": 12490
+    },
+    {
+      "epoch": 47.01785440613027,
+      "grad_norm": 0.00041797617450356483,
+      "learning_rate": 4.682843763303534e-07,
+      "loss": 0.0017,
+      "step": 12500
+    },
+    {
+      "epoch": 47.01862068965517,
+      "grad_norm": 0.018261730670928955,
+      "learning_rate": 4.5977011494252875e-07,
+      "loss": 0.0022,
+      "step": 12510
+    },
+    {
+      "epoch": 47.019386973180076,
+      "grad_norm": 0.0004435388545971364,
+      "learning_rate": 4.5125585355470417e-07,
+      "loss": 0.0005,
+      "step": 12520
+    },
+    {
+      "epoch": 47.02,
+      "eval_accuracy": 0.7777777777777778,
+      "eval_loss": 1.8378126621246338,
+      "eval_runtime": 13.2167,
+      "eval_samples_per_second": 3.405,
+      "eval_steps_per_second": 3.405,
+      "step": 12528
+    },
+    {
+      "epoch": 48.000153256704984,
+      "grad_norm": 0.0004780022718477994,
+      "learning_rate": 4.427415921668796e-07,
+      "loss": 0.0002,
+      "step": 12530
+    },
+    {
+      "epoch": 48.00091954022989,
+      "grad_norm": 0.0005424380651675165,
+      "learning_rate": 4.3422733077905495e-07,
+      "loss": 0.2688,
+      "step": 12540
+    },
+    {
+      "epoch": 48.00168582375479,
+      "grad_norm": 0.0015980154275894165,
+      "learning_rate": 4.2571306939123036e-07,
+      "loss": 0.0003,
+      "step": 12550
+    },
+    {
+      "epoch": 48.002452107279694,
+      "grad_norm": 0.0005452589248307049,
+      "learning_rate": 4.171988080034058e-07,
+      "loss": 0.0006,
+      "step": 12560
+    },
+    {
+      "epoch": 48.0032183908046,
+      "grad_norm": 0.0007816003053449094,
+      "learning_rate": 4.0868454661558115e-07,
+      "loss": 0.0001,
+      "step": 12570
+    },
+    {
+      "epoch": 48.0039846743295,
+      "grad_norm": 0.0005313084693625569,
+      "learning_rate": 4.0017028522775656e-07,
+      "loss": 0.7947,
+      "step": 12580
+    },
+    {
+      "epoch": 48.004750957854405,
+      "grad_norm": 0.04077745974063873,
+      "learning_rate": 3.916560238399319e-07,
+      "loss": 0.6254,
+      "step": 12590
+    },
+    {
+      "epoch": 48.00551724137931,
+      "grad_norm": 0.0005556725664064288,
+      "learning_rate": 3.831417624521073e-07,
+      "loss": 0.0001,
+      "step": 12600
+    },
+    {
+      "epoch": 48.00628352490421,
+      "grad_norm": 0.003675886895507574,
+      "learning_rate": 3.7462750106428265e-07,
+      "loss": 1.3723,
+      "step": 12610
+    },
+    {
+      "epoch": 48.007049808429116,
+      "grad_norm": 16.252607345581055,
+      "learning_rate": 3.6611323967645807e-07,
+      "loss": 0.712,
+      "step": 12620
+    },
+    {
+      "epoch": 48.00781609195402,
+      "grad_norm": 0.0017170652281492949,
+      "learning_rate": 3.575989782886335e-07,
+      "loss": 0.0001,
+      "step": 12630
+    },
+    {
+      "epoch": 48.00858237547893,
+      "grad_norm": 0.0009851758368313313,
+      "learning_rate": 3.4908471690080885e-07,
+      "loss": 0.0005,
+      "step": 12640
+    },
+    {
+      "epoch": 48.009348659003834,
+      "grad_norm": 170.0392608642578,
+      "learning_rate": 3.4057045551298427e-07,
+      "loss": 0.6978,
+      "step": 12650
+    },
+    {
+      "epoch": 48.01011494252874,
+      "grad_norm": 0.000492373073939234,
+      "learning_rate": 3.320561941251597e-07,
+      "loss": 0.0001,
+      "step": 12660
+    },
+    {
+      "epoch": 48.01088122605364,
+      "grad_norm": 0.0003946028009522706,
+      "learning_rate": 3.2354193273733505e-07,
+      "loss": 0.0001,
+      "step": 12670
+    },
+    {
+      "epoch": 48.011647509578545,
+      "grad_norm": 0.008205920457839966,
+      "learning_rate": 3.1502767134951047e-07,
+      "loss": 0.0002,
+      "step": 12680
+    },
+    {
+      "epoch": 48.01241379310345,
+      "grad_norm": 0.003728834679350257,
+      "learning_rate": 3.0651340996168583e-07,
+      "loss": 0.0002,
+      "step": 12690
+    },
+    {
+      "epoch": 48.01318007662835,
+      "grad_norm": 0.0004123710095882416,
+      "learning_rate": 2.9799914857386125e-07,
+      "loss": 0.433,
+      "step": 12700
+    },
+    {
+      "epoch": 48.013946360153255,
+      "grad_norm": 0.011552422307431698,
+      "learning_rate": 2.894848871860366e-07,
+      "loss": 0.0003,
+      "step": 12710
+    },
+    {
+      "epoch": 48.01471264367816,
+      "grad_norm": 0.4225425124168396,
+      "learning_rate": 2.8097062579821203e-07,
+      "loss": 0.0005,
+      "step": 12720
+    },
+    {
+      "epoch": 48.01547892720306,
+      "grad_norm": 0.23171181976795197,
+      "learning_rate": 2.7245636441038745e-07,
+      "loss": 1.3635,
+      "step": 12730
+    },
+    {
+      "epoch": 48.016245210727966,
+      "grad_norm": 0.0025505023077130318,
+      "learning_rate": 2.639421030225628e-07,
+      "loss": 0.0002,
+      "step": 12740
+    },
+    {
+      "epoch": 48.01701149425288,
+      "grad_norm": 0.0010226961458101869,
+      "learning_rate": 2.5542784163473823e-07,
+      "loss": 0.4759,
+      "step": 12750
+    },
+    {
+      "epoch": 48.01777777777778,
+      "grad_norm": 0.0028585917316377163,
+      "learning_rate": 2.469135802469136e-07,
+      "loss": 0.0005,
+      "step": 12760
+    },
+    {
+      "epoch": 48.018544061302684,
+      "grad_norm": 0.0008389436989091337,
+      "learning_rate": 2.3839931885908898e-07,
+      "loss": 0.0002,
+      "step": 12770
+    },
+    {
+      "epoch": 48.01931034482759,
+      "grad_norm": 0.015469871461391449,
+      "learning_rate": 2.2988505747126437e-07,
+      "loss": 0.0001,
+      "step": 12780
+    },
+    {
+      "epoch": 48.02,
+      "eval_accuracy": 0.7333333333333333,
+      "eval_loss": 2.0624947547912598,
+      "eval_runtime": 13.2804,
+      "eval_samples_per_second": 3.388,
+      "eval_steps_per_second": 3.388,
+      "step": 12789
+    },
+    {
+      "epoch": 49.00007662835249,
+      "grad_norm": 0.0008100093109533191,
+      "learning_rate": 2.213707960834398e-07,
+      "loss": 0.0009,
+      "step": 12790
+    },
+    {
+      "epoch": 49.00084291187739,
+      "grad_norm": 0.0005122015136294067,
+      "learning_rate": 2.1285653469561518e-07,
+      "loss": 0.0006,
+      "step": 12800
+    },
+    {
+      "epoch": 49.001609195402295,
+      "grad_norm": 0.003411679295822978,
+      "learning_rate": 2.0434227330779057e-07,
+      "loss": 0.6112,
+      "step": 12810
+    },
+    {
+      "epoch": 49.002375478927206,
+      "grad_norm": 0.00041588529711589217,
+      "learning_rate": 1.9582801191996594e-07,
+      "loss": 0.6942,
+      "step": 12820
+    },
+    {
+      "epoch": 49.00314176245211,
+      "grad_norm": 0.0009530864772386849,
+      "learning_rate": 1.8731375053214133e-07,
+      "loss": 0.8652,
+      "step": 12830
+    },
+    {
+      "epoch": 49.00390804597701,
+      "grad_norm": 1160.8369140625,
+      "learning_rate": 1.7879948914431674e-07,
+      "loss": 0.9585,
+      "step": 12840
+    },
+    {
+      "epoch": 49.00467432950192,
+      "grad_norm": 0.24398158490657806,
+      "learning_rate": 1.7028522775649214e-07,
+      "loss": 0.0002,
+      "step": 12850
+    },
+    {
+      "epoch": 49.00544061302682,
+      "grad_norm": 0.001491599716246128,
+      "learning_rate": 1.6177096636866753e-07,
+      "loss": 0.0011,
+      "step": 12860
+    },
+    {
+      "epoch": 49.006206896551724,
+      "grad_norm": 0.013720017857849598,
+      "learning_rate": 1.5325670498084292e-07,
+      "loss": 0.0002,
+      "step": 12870
+    },
+    {
+      "epoch": 49.00697318007663,
+      "grad_norm": 0.006028338335454464,
+      "learning_rate": 1.447424435930183e-07,
+      "loss": 0.0006,
+      "step": 12880
+    },
+    {
+      "epoch": 49.00773946360153,
+      "grad_norm": 0.000989487743936479,
+      "learning_rate": 1.3622818220519372e-07,
+      "loss": 0.0002,
+      "step": 12890
+    },
+    {
+      "epoch": 49.008505747126435,
+      "grad_norm": 0.0008877341751940548,
+      "learning_rate": 1.2771392081736911e-07,
+      "loss": 0.7048,
+      "step": 12900
+    },
+    {
+      "epoch": 49.00927203065134,
+      "grad_norm": 0.0015821445267647505,
+      "learning_rate": 1.1919965942954449e-07,
+      "loss": 0.0001,
+      "step": 12910
+    },
+    {
+      "epoch": 49.01003831417624,
+      "grad_norm": 0.0005364256794564426,
+      "learning_rate": 1.106853980417199e-07,
+      "loss": 0.0005,
+      "step": 12920
+    },
+    {
+      "epoch": 49.01080459770115,
+      "grad_norm": 0.0021387161687016487,
+      "learning_rate": 1.0217113665389529e-07,
+      "loss": 0.6235,
+      "step": 12930
+    },
+    {
+      "epoch": 49.011570881226056,
+      "grad_norm": 0.0005515425582416356,
+      "learning_rate": 9.365687526607066e-08,
+      "loss": 0.0001,
+      "step": 12940
+    },
+    {
+      "epoch": 49.01233716475096,
+      "grad_norm": 0.46449440717697144,
+      "learning_rate": 8.514261387824607e-08,
+      "loss": 0.0002,
+      "step": 12950
+    },
+    {
+      "epoch": 49.013103448275864,
+      "grad_norm": 0.002538820030167699,
+      "learning_rate": 7.662835249042146e-08,
+      "loss": 0.406,
+      "step": 12960
+    },
+    {
+      "epoch": 49.01386973180077,
+      "grad_norm": 0.00038823066279292107,
+      "learning_rate": 6.811409110259686e-08,
+      "loss": 0.0527,
+      "step": 12970
+    },
+    {
+      "epoch": 49.01463601532567,
+      "grad_norm": 0.32803836464881897,
+      "learning_rate": 5.9599829714772246e-08,
+      "loss": 0.0039,
+      "step": 12980
+    },
+    {
+      "epoch": 49.015402298850574,
+      "grad_norm": 0.0092452522367239,
+      "learning_rate": 5.108556832694764e-08,
+      "loss": 0.0001,
+      "step": 12990
+    },
+    {
+      "epoch": 49.01616858237548,
+      "grad_norm": 0.039832472801208496,
+      "learning_rate": 4.2571306939123034e-08,
+      "loss": 0.0313,
+      "step": 13000
+    },
+    {
+      "epoch": 49.01693486590038,
+      "grad_norm": 0.4108397960662842,
+      "learning_rate": 3.405704555129843e-08,
+      "loss": 0.0005,
+      "step": 13010
+    },
+    {
+      "epoch": 49.017701149425285,
+      "grad_norm": 0.00046747943270020187,
+      "learning_rate": 2.554278416347382e-08,
+      "loss": 0.0004,
+      "step": 13020
+    },
+    {
+      "epoch": 49.01846743295019,
+      "grad_norm": 0.0007605894934386015,
+      "learning_rate": 1.7028522775649215e-08,
+      "loss": 0.0832,
+      "step": 13030
+    },
+    {
+      "epoch": 49.01923371647509,
+      "grad_norm": 0.0008342632208950818,
+      "learning_rate": 8.514261387824608e-09,
+      "loss": 0.0001,
+      "step": 13040
+    },
+    {
+      "epoch": 49.02,
+      "grad_norm": 0.03882928565144539,
+      "learning_rate": 0.0,
+      "loss": 0.59,
+      "step": 13050
+    },
+    {
+      "epoch": 49.02,
+      "eval_accuracy": 0.7333333333333333,
+      "eval_loss": 2.0340445041656494,
+      "eval_runtime": 15.8259,
+      "eval_samples_per_second": 2.843,
+      "eval_steps_per_second": 2.843,
+      "step": 13050
+    },
+    {
+      "epoch": 49.02,
+      "step": 13050,
+      "total_flos": 5.730289341462282e+19,
+      "train_loss": 0.7221503470706512,
+      "train_runtime": 9500.8761,
+      "train_samples_per_second": 1.374,
+      "train_steps_per_second": 1.374
+    },
+    {
+      "epoch": 49.02,
+      "eval_accuracy": 0.8666666666666667,
+      "eval_loss": 0.5090304017066956,
+      "eval_runtime": 13.1433,
+      "eval_samples_per_second": 3.424,
+      "eval_steps_per_second": 3.424,
+      "step": 13050
+    },
+    {
+      "epoch": 49.02,
+      "eval_accuracy": 0.8666666666666667,
+      "eval_loss": 0.5090304017066956,
+      "eval_runtime": 13.1401,
+      "eval_samples_per_second": 3.425,
+      "eval_steps_per_second": 3.425,
+      "step": 13050
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 13050,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.730289341462282e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}