diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4501 @@
+{
+  "best_metric": 0.564625850340136,
+  "best_model_checkpoint": "/data/users/bking2/tod_zero/outputs/runs/finetune/starcoder_3b/mar_18_otj9Am5v/offline_label/bqag3oyb/checkpoint-22400",
+  "epoch": 0.7928838672910627,
+  "eval_steps": 3200,
+  "global_step": 32000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6985061168670654,
+      "learning_rate": 2.5e-06,
+      "loss": 1.1039,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.080778956413269,
+      "learning_rate": 5e-06,
+      "loss": 0.2136,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0957249402999878,
+      "learning_rate": 4.9999696912850374e-06,
+      "loss": 0.1732,
+      "step": 150
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0785242319107056,
+      "learning_rate": 4.999878765875043e-06,
+      "loss": 0.1636,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1675797700881958,
+      "learning_rate": 4.999727225974682e-06,
+      "loss": 0.1622,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1212489604949951,
+      "learning_rate": 4.999515075258341e-06,
+      "loss": 0.1633,
+      "step": 300
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9962964653968811,
+      "learning_rate": 4.999242318870029e-06,
+      "loss": 0.149,
+      "step": 350
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0880298614501953,
+      "learning_rate": 4.998908963423264e-06,
+      "loss": 0.1558,
+      "step": 400
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0359498262405396,
+      "learning_rate": 4.998515017000907e-06,
+      "loss": 0.1558,
+      "step": 450
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.019607663154602,
+      "learning_rate": 4.998060489154965e-06,
+      "loss": 0.1504,
+      "step": 500
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9753773808479309,
+      "learning_rate": 4.997545390906362e-06,
+      "loss": 0.1502,
+      "step": 550
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9316331148147583,
+      "learning_rate": 4.996969734744671e-06,
+      "loss": 0.1445,
+      "step": 600
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8932620286941528,
+      "learning_rate": 4.99633353462781e-06,
+      "loss": 0.1419,
+      "step": 650
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8961341977119446,
+      "learning_rate": 4.995636805981707e-06,
+      "loss": 0.1424,
+      "step": 700
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0110450983047485,
+      "learning_rate": 4.99487956569992e-06,
+      "loss": 0.1484,
+      "step": 750
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9437933564186096,
+      "learning_rate": 4.994061832143235e-06,
+      "loss": 0.1382,
+      "step": 800
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.836251974105835,
+      "learning_rate": 4.993183625139212e-06,
+      "loss": 0.1448,
+      "step": 850
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9206503629684448,
+      "learning_rate": 4.992244965981714e-06,
+      "loss": 0.1408,
+      "step": 900
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9278163313865662,
+      "learning_rate": 4.991245877430382e-06,
+      "loss": 0.1396,
+      "step": 950
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9222483038902283,
+      "learning_rate": 4.990186383710089e-06,
+      "loss": 0.1387,
+      "step": 1000
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9481703042984009,
+      "learning_rate": 4.9890665105103484e-06,
+      "loss": 0.1391,
+      "step": 1050
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.865689218044281,
+      "learning_rate": 4.987886284984695e-06,
+      "loss": 0.1329,
+      "step": 1100
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9181361794471741,
+      "learning_rate": 4.986645735750025e-06,
+      "loss": 0.1399,
+      "step": 1150
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.942889928817749,
+      "learning_rate": 4.985344892885899e-06,
+      "loss": 0.144,
+      "step": 1200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9355672001838684,
+      "learning_rate": 4.98398378793382e-06,
+      "loss": 0.1412,
+      "step": 1250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8571763634681702,
+      "learning_rate": 4.982562453896458e-06,
+      "loss": 0.1323,
+      "step": 1300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8957087397575378,
+      "learning_rate": 4.9810809252368615e-06,
+      "loss": 0.1356,
+      "step": 1350
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9295082688331604,
+      "learning_rate": 4.979539237877615e-06,
+      "loss": 0.1332,
+      "step": 1400
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8543279767036438,
+      "learning_rate": 4.977937429199968e-06,
+      "loss": 0.1335,
+      "step": 1450
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9446925520896912,
+      "learning_rate": 4.976275538042932e-06,
+      "loss": 0.1336,
+      "step": 1500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8666749596595764,
+      "learning_rate": 4.974553604702332e-06,
+      "loss": 0.1369,
+      "step": 1550
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9225461483001709,
+      "learning_rate": 4.972771670929841e-06,
+      "loss": 0.1326,
+      "step": 1600
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6894574165344238,
+      "learning_rate": 4.970929779931955e-06,
+      "loss": 0.1317,
+      "step": 1650
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8522433638572693,
+      "learning_rate": 4.969027976368954e-06,
+      "loss": 0.134,
+      "step": 1700
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8407069444656372,
+      "learning_rate": 4.967066306353816e-06,
+      "loss": 0.13,
+      "step": 1750
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.901373565196991,
+      "learning_rate": 4.9650448174510986e-06,
+      "loss": 0.1347,
+      "step": 1800
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8298419713973999,
+      "learning_rate": 4.9629635586757865e-06,
+      "loss": 0.1329,
+      "step": 1850
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7622634172439575,
+      "learning_rate": 4.960822580492103e-06,
+      "loss": 0.1248,
+      "step": 1900
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8928254842758179,
+      "learning_rate": 4.958621934812286e-06,
+      "loss": 0.1349,
+      "step": 1950
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.830184817314148,
+      "learning_rate": 4.95636167499533e-06,
+      "loss": 0.1247,
+      "step": 2000
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9439135193824768,
+      "learning_rate": 4.9540418558456915e-06,
+      "loss": 0.1346,
+      "step": 2050
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8274710178375244,
+      "learning_rate": 4.951662533611959e-06,
+      "loss": 0.1321,
+      "step": 2100
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8595709204673767,
+      "learning_rate": 4.9492237659854946e-06,
+      "loss": 0.131,
+      "step": 2150
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.864810049533844,
+      "learning_rate": 4.9467256120990255e-06,
+      "loss": 0.1286,
+      "step": 2200
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7884636521339417,
+      "learning_rate": 4.9441681325252215e-06,
+      "loss": 0.132,
+      "step": 2250
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7718741297721863,
+      "learning_rate": 4.941551389275217e-06,
+      "loss": 0.1293,
+      "step": 2300
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8145634531974792,
+      "learning_rate": 4.938875445797112e-06,
+      "loss": 0.1348,
+      "step": 2350
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8029401898384094,
+      "learning_rate": 4.936140366974434e-06,
+      "loss": 0.1309,
+      "step": 2400
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7573081851005554,
+      "learning_rate": 4.933346219124562e-06,
+      "loss": 0.1257,
+      "step": 2450
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8588012456893921,
+      "learning_rate": 4.93049306999712e-06,
+      "loss": 0.1342,
+      "step": 2500
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7663673758506775,
+      "learning_rate": 4.927580988772336e-06,
+      "loss": 0.1263,
+      "step": 2550
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6892865300178528,
+      "learning_rate": 4.9246100460593606e-06,
+      "loss": 0.1304,
+      "step": 2600
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7494785785675049,
+      "learning_rate": 4.92158031389456e-06,
+      "loss": 0.1258,
+      "step": 2650
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7757202982902527,
+      "learning_rate": 4.918491865739763e-06,
+      "loss": 0.1284,
+      "step": 2700
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8363822102546692,
+      "learning_rate": 4.915344776480487e-06,
+      "loss": 0.1271,
+      "step": 2750
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8193435668945312,
+      "learning_rate": 4.912139122424118e-06,
+      "loss": 0.1309,
+      "step": 2800
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7342443466186523,
+      "learning_rate": 4.908874981298058e-06,
+      "loss": 0.1319,
+      "step": 2850
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7340241074562073,
+      "learning_rate": 4.9055524322478456e-06,
+      "loss": 0.1263,
+      "step": 2900
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8974625468254089,
+      "learning_rate": 4.902171555835236e-06,
+      "loss": 0.1272,
+      "step": 2950
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8946986794471741,
+      "learning_rate": 4.8987324340362445e-06,
+      "loss": 0.1316,
+      "step": 3000
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8382922410964966,
+      "learning_rate": 4.895235150239159e-06,
+      "loss": 0.1309,
+      "step": 3050
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7496135234832764,
+      "learning_rate": 4.891679789242524e-06,
+      "loss": 0.1271,
+      "step": 3100
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9334201812744141,
+      "learning_rate": 4.8880664372530765e-06,
+      "loss": 0.1264,
+      "step": 3150
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.879549503326416,
+      "learning_rate": 4.884395181883661e-06,
+      "loss": 0.1295,
+      "step": 3200
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8264172673225403,
+      "learning_rate": 4.880666112151104e-06,
+      "loss": 0.1239,
+      "step": 3250
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7288581728935242,
+      "learning_rate": 4.876879318474056e-06,
+      "loss": 0.1308,
+      "step": 3300
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6938970685005188,
+      "learning_rate": 4.873034892670795e-06,
+      "loss": 0.124,
+      "step": 3350
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7392057776451111,
+      "learning_rate": 4.869132927957007e-06,
+      "loss": 0.129,
+      "step": 3400
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7360211610794067,
+      "learning_rate": 4.8651735189435205e-06,
+      "loss": 0.1218,
+      "step": 3450
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8656970262527466,
+      "learning_rate": 4.861156761634014e-06,
+      "loss": 0.1271,
+      "step": 3500
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7724373936653137,
+      "learning_rate": 4.857082753422691e-06,
+      "loss": 0.1266,
+      "step": 3550
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7553298473358154,
+      "learning_rate": 4.852951593091914e-06,
+      "loss": 0.1251,
+      "step": 3600
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8519923090934753,
+      "learning_rate": 4.848763380809811e-06,
+      "loss": 0.1253,
+      "step": 3650
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7823351621627808,
+      "learning_rate": 4.844518218127849e-06,
+      "loss": 0.1234,
+      "step": 3700
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.863345742225647,
+      "learning_rate": 4.840216207978368e-06,
+      "loss": 0.1298,
+      "step": 3750
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.729887068271637,
+      "learning_rate": 4.835857454672087e-06,
+      "loss": 0.1293,
+      "step": 3800
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8041313290596008,
+      "learning_rate": 4.831442063895575e-06,
+      "loss": 0.125,
+      "step": 3850
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6497385501861572,
+      "learning_rate": 4.8269701427086905e-06,
+      "loss": 0.1272,
+      "step": 3900
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7458536624908447,
+      "learning_rate": 4.822441799541979e-06,
+      "loss": 0.1251,
+      "step": 3950
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7111815810203552,
+      "learning_rate": 4.8178571441940515e-06,
+      "loss": 0.1209,
+      "step": 4000
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8121939301490784,
+      "learning_rate": 4.813216287828917e-06,
+      "loss": 0.1209,
+      "step": 4050
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6625011563301086,
+      "learning_rate": 4.808519342973289e-06,
+      "loss": 0.1277,
+      "step": 4100
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8151265978813171,
+      "learning_rate": 4.80376642351386e-06,
+      "loss": 0.1268,
+      "step": 4150
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6908511519432068,
+      "learning_rate": 4.798957644694533e-06,
+      "loss": 0.1247,
+      "step": 4200
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.653842568397522,
+      "learning_rate": 4.794093123113635e-06,
+      "loss": 0.1257,
+      "step": 4250
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8020724654197693,
+      "learning_rate": 4.789172976721082e-06,
+      "loss": 0.121,
+      "step": 4300
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7690026164054871,
+      "learning_rate": 4.7841973248155275e-06,
+      "loss": 0.1265,
+      "step": 4350
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8199232220649719,
+      "learning_rate": 4.779166288041463e-06,
+      "loss": 0.1233,
+      "step": 4400
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6871870756149292,
+      "learning_rate": 4.7740799883862966e-06,
+      "loss": 0.1252,
+      "step": 4450
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8392221927642822,
+      "learning_rate": 4.7689385491773934e-06,
+      "loss": 0.1294,
+      "step": 4500
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.678961455821991,
+      "learning_rate": 4.7637420950790855e-06,
+      "loss": 0.1205,
+      "step": 4550
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8034936785697937,
+      "learning_rate": 4.75849075208965e-06,
+      "loss": 0.1237,
+      "step": 4600
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5964626669883728,
+      "learning_rate": 4.7531846475382526e-06,
+      "loss": 0.1196,
+      "step": 4650
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7667948007583618,
+      "learning_rate": 4.7478239100818626e-06,
+      "loss": 0.1215,
+      "step": 4700
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.896281898021698,
+      "learning_rate": 4.742408669702131e-06,
+      "loss": 0.1253,
+      "step": 4750
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6606175303459167,
+      "learning_rate": 4.736939057702239e-06,
+      "loss": 0.1199,
+      "step": 4800
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7194575071334839,
+      "learning_rate": 4.731415206703714e-06,
+      "loss": 0.1275,
+      "step": 4850
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7007895112037659,
+      "learning_rate": 4.725837250643218e-06,
+      "loss": 0.1241,
+      "step": 4900
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.703335702419281,
+      "learning_rate": 4.720205324769296e-06,
+      "loss": 0.1209,
+      "step": 4950
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7912906408309937,
+      "learning_rate": 4.714519565639095e-06,
+      "loss": 0.1205,
+      "step": 5000
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7368320226669312,
+      "learning_rate": 4.708780111115058e-06,
+      "loss": 0.1267,
+      "step": 5050
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8405609726905823,
+      "learning_rate": 4.702987100361578e-06,
+      "loss": 0.1217,
+      "step": 5100
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.768998384475708,
+      "learning_rate": 4.697140673841624e-06,
+      "loss": 0.1235,
+      "step": 5150
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.730161726474762,
+      "learning_rate": 4.6912409733133365e-06,
+      "loss": 0.1234,
+      "step": 5200
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7618039846420288,
+      "learning_rate": 4.685288141826589e-06,
+      "loss": 0.1206,
+      "step": 5250
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7404894828796387,
+      "learning_rate": 4.679282323719519e-06,
+      "loss": 0.1207,
+      "step": 5300
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8315970301628113,
+      "learning_rate": 4.67322366461503e-06,
+      "loss": 0.1197,
+      "step": 5350
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8174512386322021,
+      "learning_rate": 4.66711231141726e-06,
+      "loss": 0.1246,
+      "step": 5400
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7582206130027771,
+      "learning_rate": 4.660948412308018e-06,
+      "loss": 0.1179,
+      "step": 5450
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7047674655914307,
+      "learning_rate": 4.654732116743193e-06,
+      "loss": 0.124,
+      "step": 5500
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7401306629180908,
+      "learning_rate": 4.64846357544913e-06,
+      "loss": 0.1209,
+      "step": 5550
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7297683954238892,
+      "learning_rate": 4.642142940418973e-06,
+      "loss": 0.124,
+      "step": 5600
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6836703419685364,
+      "learning_rate": 4.635770364908984e-06,
+      "loss": 0.1188,
+      "step": 5650
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6775400042533875,
+      "learning_rate": 4.629346003434822e-06,
+      "loss": 0.1211,
+      "step": 5700
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8052581548690796,
+      "learning_rate": 4.622870011767798e-06,
+      "loss": 0.1203,
+      "step": 5750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7253400087356567,
+      "learning_rate": 4.616342546931103e-06,
+      "loss": 0.1226,
+      "step": 5800
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6844501495361328,
+      "learning_rate": 4.609763767195991e-06,
+      "loss": 0.1208,
+      "step": 5850
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7401016354560852,
+      "learning_rate": 4.603133832077953e-06,
+      "loss": 0.1192,
+      "step": 5900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8344407081604004,
+      "learning_rate": 4.596452902332839e-06,
+      "loss": 0.1198,
+      "step": 5950
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6959987878799438,
+      "learning_rate": 4.589721139952964e-06,
+      "loss": 0.1156,
+      "step": 6000
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6912865042686462,
+      "learning_rate": 4.582938708163183e-06,
+      "loss": 0.1151,
+      "step": 6050
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8227457404136658,
+      "learning_rate": 4.576105771416928e-06,
+      "loss": 0.1196,
+      "step": 6100
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7104191184043884,
+      "learning_rate": 4.569222495392227e-06,
+      "loss": 0.1228,
+      "step": 6150
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.816095232963562,
+      "learning_rate": 4.562289046987679e-06,
+      "loss": 0.125,
+      "step": 6200
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7309590578079224,
+      "learning_rate": 4.555305594318414e-06,
+      "loss": 0.1229,
+      "step": 6250
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6685283184051514,
+      "learning_rate": 4.548272306712013e-06,
+      "loss": 0.1204,
+      "step": 6300
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7148935794830322,
+      "learning_rate": 4.541189354704403e-06,
+      "loss": 0.1213,
+      "step": 6350
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7746471762657166,
+      "learning_rate": 4.534056910035724e-06,
+      "loss": 0.1176,
+      "step": 6400
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6856210231781006,
+      "learning_rate": 4.5268751456461605e-06,
+      "loss": 0.1214,
+      "step": 6450
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7244325280189514,
+      "learning_rate": 4.5196442356717526e-06,
+      "loss": 0.119,
+      "step": 6500
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7556556463241577,
+      "learning_rate": 4.512364355440172e-06,
+      "loss": 0.1236,
+      "step": 6550
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7055052518844604,
+      "learning_rate": 4.505035681466472e-06,
+      "loss": 0.119,
+      "step": 6600
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6788272261619568,
+      "learning_rate": 4.497658391448803e-06,
+      "loss": 0.1189,
+      "step": 6650
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8098623752593994,
+      "learning_rate": 4.49023266426411e-06,
+      "loss": 0.1178,
+      "step": 6700
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5360668301582336,
+      "learning_rate": 4.482758679963792e-06,
+      "loss": 0.1123,
+      "step": 6750
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7503378987312317,
+      "learning_rate": 4.475236619769336e-06,
+      "loss": 0.1232,
+      "step": 6800
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8074086308479309,
+      "learning_rate": 4.4676666660679265e-06,
+      "loss": 0.1124,
+      "step": 6850
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7047454118728638,
+      "learning_rate": 4.460049002408018e-06,
+      "loss": 0.1224,
+      "step": 6900
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7454392313957214,
+      "learning_rate": 4.452383813494887e-06,
+      "loss": 0.1198,
+      "step": 6950
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7296658158302307,
+      "learning_rate": 4.444671285186155e-06,
+      "loss": 0.1191,
+      "step": 7000
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7346377372741699,
+      "learning_rate": 4.4369116044872786e-06,
+      "loss": 0.123,
+      "step": 7050
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7199191451072693,
+      "learning_rate": 4.42910495954702e-06,
+      "loss": 0.1226,
+      "step": 7100
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8274137377738953,
+      "learning_rate": 4.421251539652879e-06,
+      "loss": 0.1165,
+      "step": 7150
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6081553101539612,
+      "learning_rate": 4.413351535226507e-06,
+      "loss": 0.1205,
+      "step": 7200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6576844453811646,
+      "learning_rate": 4.4054051378190915e-06,
+      "loss": 0.1196,
+      "step": 7250
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8263300657272339,
+      "learning_rate": 4.397412540106707e-06,
+      "loss": 0.1194,
+      "step": 7300
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7084447145462036,
+      "learning_rate": 4.3893739358856465e-06,
+      "loss": 0.1167,
+      "step": 7350
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6882210969924927,
+      "learning_rate": 4.38128952006772e-06,
+      "loss": 0.1179,
+      "step": 7400
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5834758281707764,
+      "learning_rate": 4.373159488675533e-06,
+      "loss": 0.1126,
+      "step": 7450
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6505135893821716,
+      "learning_rate": 4.364984038837727e-06,
+      "loss": 0.1225,
+      "step": 7500
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6714935898780823,
+      "learning_rate": 4.356763368784207e-06,
+      "loss": 0.1178,
+      "step": 7550
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7283095121383667,
+      "learning_rate": 4.348497677841328e-06,
+      "loss": 0.1184,
+      "step": 7600
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6789336800575256,
+      "learning_rate": 4.340187166427067e-06,
+      "loss": 0.1166,
+      "step": 7650
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6609737277030945,
+      "learning_rate": 4.331832036046162e-06,
+      "loss": 0.1171,
+      "step": 7700
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8023338317871094,
+      "learning_rate": 4.323432489285223e-06,
+      "loss": 0.1169,
+      "step": 7750
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6348956227302551,
+      "learning_rate": 4.3149887298078275e-06,
+      "loss": 0.1199,
+      "step": 7800
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7053850293159485,
+      "learning_rate": 4.306500962349573e-06,
+      "loss": 0.1216,
+      "step": 7850
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7582606077194214,
+      "learning_rate": 4.2979693927131205e-06,
+      "loss": 0.1191,
+      "step": 7900
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7093271017074585,
+      "learning_rate": 4.289394227763199e-06,
+      "loss": 0.1132,
+      "step": 7950
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8069114685058594,
+      "learning_rate": 4.2807756754215926e-06,
+      "loss": 0.1163,
+      "step": 8000
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7039481997489929,
+      "learning_rate": 4.272113944662099e-06,
+      "loss": 0.1236,
+      "step": 8050
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7971010208129883,
+      "learning_rate": 4.263409245505461e-06,
+      "loss": 0.1211,
+      "step": 8100
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7244572639465332,
+      "learning_rate": 4.254661789014274e-06,
+      "loss": 0.1151,
+      "step": 8150
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7881250381469727,
+      "learning_rate": 4.2458717872878715e-06,
+      "loss": 0.1169,
+      "step": 8200
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7103307843208313,
+      "learning_rate": 4.237039453457179e-06,
+      "loss": 0.1208,
+      "step": 8250
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7782537937164307,
+      "learning_rate": 4.228165001679547e-06,
+      "loss": 0.1229,
+      "step": 8300
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6521384119987488,
+      "learning_rate": 4.219248647133559e-06,
+      "loss": 0.1203,
+      "step": 8350
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7174645662307739,
+      "learning_rate": 4.210290606013813e-06,
+      "loss": 0.116,
+      "step": 8400
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8697425723075867,
+      "learning_rate": 4.2012910955256825e-06,
+      "loss": 0.1138,
+      "step": 8450
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6629298329353333,
+      "learning_rate": 4.192250333880045e-06,
+      "loss": 0.1177,
+      "step": 8500
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6568993330001831,
+      "learning_rate": 4.183168540287995e-06,
+      "loss": 0.1177,
+      "step": 8550
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6727361083030701,
+      "learning_rate": 4.174045934955527e-06,
+      "loss": 0.1166,
+      "step": 8600
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7124900221824646,
+      "learning_rate": 4.164882739078197e-06,
+      "loss": 0.1199,
+      "step": 8650
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6333892941474915,
+      "learning_rate": 4.155679174835758e-06,
+      "loss": 0.1158,
+      "step": 8700
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6718785762786865,
+      "learning_rate": 4.146435465386776e-06,
+      "loss": 0.1155,
+      "step": 8750
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7972610592842102,
+      "learning_rate": 4.137151834863213e-06,
+      "loss": 0.1165,
+      "step": 8800
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6914675831794739,
+      "learning_rate": 4.1278285083649985e-06,
+      "loss": 0.115,
+      "step": 8850
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7176665663719177,
+      "learning_rate": 4.11846571195457e-06,
+      "loss": 0.1191,
+      "step": 8900
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6305205821990967,
+      "learning_rate": 4.1090636726513875e-06,
+      "loss": 0.1162,
+      "step": 8950
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.655022144317627,
+      "learning_rate": 4.0996226184264355e-06,
+      "loss": 0.1172,
+      "step": 9000
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7785295248031616,
+      "learning_rate": 4.090142778196692e-06,
+      "loss": 0.1188,
+      "step": 9050
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6232600212097168,
+      "learning_rate": 4.080624381819577e-06,
+      "loss": 0.1172,
+      "step": 9100
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7053920030593872,
+      "learning_rate": 4.071067660087379e-06,
+      "loss": 0.111,
+      "step": 9150
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7314987182617188,
+      "learning_rate": 4.061472844721664e-06,
+      "loss": 0.1238,
+      "step": 9200
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7953878045082092,
+      "learning_rate": 4.05184016836765e-06,
+      "loss": 0.1131,
+      "step": 9250
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.840351402759552,
+      "learning_rate": 4.042169864588571e-06,
+      "loss": 0.1117,
+      "step": 9300
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8227623701095581,
+      "learning_rate": 4.032462167860012e-06,
+      "loss": 0.1151,
+      "step": 9350
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6529046297073364,
+      "learning_rate": 4.022717313564223e-06,
+      "loss": 0.1164,
+      "step": 9400
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7209616899490356,
+      "learning_rate": 4.012935537984414e-06,
+      "loss": 0.1142,
+      "step": 9450
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6554023623466492,
+      "learning_rate": 4.0031170782990214e-06,
+      "loss": 0.1177,
+      "step": 9500
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8092223405838013,
+      "learning_rate": 3.993262172575962e-06,
+      "loss": 0.1155,
+      "step": 9550
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7681934833526611,
+      "learning_rate": 3.983371059766862e-06,
+      "loss": 0.116,
+      "step": 9600
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7766609787940979,
+      "learning_rate": 3.973443979701252e-06,
+      "loss": 0.1158,
+      "step": 9650
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6570335030555725,
+      "learning_rate": 3.963481173080768e-06,
+      "loss": 0.1151,
+      "step": 9700
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7423933744430542,
+      "learning_rate": 3.9534828814733e-06,
+      "loss": 0.1159,
+      "step": 9750
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8370651006698608,
+      "learning_rate": 3.943449347307146e-06,
+      "loss": 0.1197,
+      "step": 9800
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6548042297363281,
+      "learning_rate": 3.9333808138651265e-06,
+      "loss": 0.114,
+      "step": 9850
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7364439368247986,
+      "learning_rate": 3.923277525278691e-06,
+      "loss": 0.1129,
+      "step": 9900
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.669224202632904,
+      "learning_rate": 3.913139726521993e-06,
+      "loss": 0.1158,
+      "step": 9950
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.834517240524292,
+      "learning_rate": 3.9029676634059565e-06,
+      "loss": 0.1159,
+      "step": 10000
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6793729066848755,
+      "learning_rate": 3.89276158257231e-06,
+      "loss": 0.1139,
+      "step": 10050
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7160745859146118,
+      "learning_rate": 3.882521731487609e-06,
+      "loss": 0.1158,
+      "step": 10100
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6966678500175476,
+      "learning_rate": 3.872248358437236e-06,
+      "loss": 0.1164,
+      "step": 10150
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7358713746070862,
+      "learning_rate": 3.861941712519379e-06,
+      "loss": 0.1144,
+      "step": 10200
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7294729351997375,
+      "learning_rate": 3.8516020436389945e-06,
+      "loss": 0.1191,
+      "step": 10250
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7433848977088928,
+      "learning_rate": 3.841229602501742e-06,
+      "loss": 0.112,
+      "step": 10300
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7290607690811157,
+      "learning_rate": 3.8308246406079116e-06,
+      "loss": 0.1131,
+      "step": 10350
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7990119457244873,
+      "learning_rate": 3.820387410246324e-06,
+      "loss": 0.1155,
+      "step": 10400
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7290509939193726,
+      "learning_rate": 3.809918164488208e-06,
+      "loss": 0.1179,
+      "step": 10450
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7162163257598877,
+      "learning_rate": 3.7994171571810756e-06,
+      "loss": 0.114,
+      "step": 10500
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7141215801239014,
+      "learning_rate": 3.788884642942555e-06,
+      "loss": 0.1163,
+      "step": 10550
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7095829844474792,
+      "learning_rate": 3.7783208771542237e-06,
+      "loss": 0.1133,
+      "step": 10600
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.752619743347168,
+      "learning_rate": 3.7677261159554145e-06,
+      "loss": 0.1191,
+      "step": 10650
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7586488127708435,
+      "learning_rate": 3.757100616237006e-06,
+      "loss": 0.1183,
+      "step": 10700
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6924918293952942,
+      "learning_rate": 3.746444635635191e-06,
+      "loss": 0.113,
+      "step": 10750
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6969543695449829,
+      "learning_rate": 3.735758432525234e-06,
+      "loss": 0.1096,
+      "step": 10800
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7095456719398499,
+      "learning_rate": 3.725042266015201e-06,
+      "loss": 0.1145,
+      "step": 10850
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7116329669952393,
+      "learning_rate": 3.7142963959396805e-06,
+      "loss": 0.1104,
+      "step": 10900
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7710102200508118,
+      "learning_rate": 3.7035210828534846e-06,
+      "loss": 0.118,
+      "step": 10950
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5557773113250732,
+      "learning_rate": 3.692716588025327e-06,
+      "loss": 0.1168,
+      "step": 11000
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.769664466381073,
+      "learning_rate": 3.68188317343149e-06,
+      "loss": 0.1166,
+      "step": 11050
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6898650527000427,
+      "learning_rate": 3.671021101749476e-06,
+      "loss": 0.1154,
+      "step": 11100
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7608383297920227,
+      "learning_rate": 3.6601306363516297e-06,
+      "loss": 0.1138,
+      "step": 11150
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6353211998939514,
+      "learning_rate": 3.649212041298763e-06,
+      "loss": 0.1121,
+      "step": 11200
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7474289536476135,
+      "learning_rate": 3.638265581333742e-06,
+      "loss": 0.1169,
+      "step": 11250
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6579691171646118,
+      "learning_rate": 3.627291521875076e-06,
+      "loss": 0.1102,
+      "step": 11300
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6420250535011292,
+      "learning_rate": 3.616290129010476e-06,
+      "loss": 0.111,
+      "step": 11350
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6057188510894775,
+      "learning_rate": 3.605261669490407e-06,
+      "loss": 0.1125,
+      "step": 11400
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5284745693206787,
+      "learning_rate": 3.5942064107216183e-06,
+      "loss": 0.1051,
+      "step": 11450
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6904653906822205,
+      "learning_rate": 3.5831246207606597e-06,
+      "loss": 0.1197,
+      "step": 11500
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6652848124504089,
+      "learning_rate": 3.57201656830738e-06,
+      "loss": 0.1116,
+      "step": 11550
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7260887026786804,
+      "learning_rate": 3.5608825226984168e-06,
+      "loss": 0.1108,
+      "step": 11600
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7659937143325806,
+      "learning_rate": 3.549722753900662e-06,
+      "loss": 0.1142,
+      "step": 11650
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6617051362991333,
+      "learning_rate": 3.5385375325047167e-06,
+      "loss": 0.1126,
+      "step": 11700
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7225079536437988,
+      "learning_rate": 3.5273271297183302e-06,
+      "loss": 0.1159,
+      "step": 11750
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5886747241020203,
+      "learning_rate": 3.516091817359825e-06,
+      "loss": 0.111,
+      "step": 11800
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7085060477256775,
+      "learning_rate": 3.5048318678515052e-06,
+      "loss": 0.1123,
+      "step": 11850
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7402300834655762,
+      "learning_rate": 3.493547554213051e-06,
+      "loss": 0.1161,
+      "step": 11900
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7527924180030823,
+      "learning_rate": 3.482239150054898e-06,
+      "loss": 0.1147,
+      "step": 11950
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6737644076347351,
+      "learning_rate": 3.470906929571605e-06,
+      "loss": 0.1105,
+      "step": 12000
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7493759393692017,
+      "learning_rate": 3.459551167535205e-06,
+      "loss": 0.1145,
+      "step": 12050
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7661862373352051,
+      "learning_rate": 3.4481721392885415e-06,
+      "loss": 0.1136,
+      "step": 12100
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6885182857513428,
+      "learning_rate": 3.4367701207385944e-06,
+      "loss": 0.1122,
+      "step": 12150
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8971922993659973,
+      "learning_rate": 3.425345388349787e-06,
+      "loss": 0.108,
+      "step": 12200
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6365148425102234,
+      "learning_rate": 3.4138982191372838e-06,
+      "loss": 0.113,
+      "step": 12250
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.627581775188446,
+      "learning_rate": 3.402428890660279e-06,
+      "loss": 0.1094,
+      "step": 12300
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.73035728931427,
+      "learning_rate": 3.390937681015256e-06,
+      "loss": 0.1132,
+      "step": 12350
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5905104875564575,
+      "learning_rate": 3.379424868829254e-06,
+      "loss": 0.1126,
+      "step": 12400
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6537535786628723,
+      "learning_rate": 3.367890733253108e-06,
+      "loss": 0.1137,
+      "step": 12450
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.718180239200592,
+      "learning_rate": 3.3563355539546795e-06,
+      "loss": 0.1169,
+      "step": 12500
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6057806611061096,
+      "learning_rate": 3.3447596111120767e-06,
+      "loss": 0.1127,
+      "step": 12550
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8001242876052856,
+      "learning_rate": 3.333163185406861e-06,
+      "loss": 0.1124,
+      "step": 12600
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6936395168304443,
+      "learning_rate": 3.321546558017243e-06,
+      "loss": 0.1089,
+      "step": 12650
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.664017915725708,
+      "learning_rate": 3.309910010611259e-06,
+      "loss": 0.1137,
+      "step": 12700
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6639358997344971,
+      "learning_rate": 3.29825382533995e-06,
+      "loss": 0.1152,
+      "step": 12750
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6084735989570618,
+      "learning_rate": 3.286578284830513e-06,
+      "loss": 0.1117,
+      "step": 12800
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6548373103141785,
+      "learning_rate": 3.2748836721794514e-06,
+      "loss": 0.1146,
+      "step": 12850
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7516295313835144,
+      "learning_rate": 3.263170270945709e-06,
+      "loss": 0.1095,
+      "step": 12900
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6284793615341187,
+      "learning_rate": 3.2514383651437987e-06,
+      "loss": 0.1122,
+      "step": 12950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7250240445137024,
+      "learning_rate": 3.239688239236911e-06,
+      "loss": 0.1156,
+      "step": 13000
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6905579566955566,
+      "learning_rate": 3.2279201781300206e-06,
+      "loss": 0.1143,
+      "step": 13050
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7570493221282959,
+      "learning_rate": 3.2161344671629736e-06,
+      "loss": 0.1177,
+      "step": 13100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7020050287246704,
+      "learning_rate": 3.2043313921035747e-06,
+      "loss": 0.1054,
+      "step": 13150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7082750201225281,
+      "learning_rate": 3.1925112391406534e-06,
+      "loss": 0.1118,
+      "step": 13200
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6312862634658813,
+      "learning_rate": 3.1806742948771276e-06,
+      "loss": 0.1111,
+      "step": 13250
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6235171556472778,
+      "learning_rate": 3.168820846323053e-06,
+      "loss": 0.1082,
+      "step": 13300
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7088466882705688,
+      "learning_rate": 3.1569511808886633e-06,
+      "loss": 0.1114,
+      "step": 13350
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7077568173408508,
+      "learning_rate": 3.1450655863774053e-06,
+      "loss": 0.1128,
+      "step": 13400
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6110339760780334,
+      "learning_rate": 3.1331643509789553e-06,
+      "loss": 0.1099,
+      "step": 13450
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1900233030319214,
+      "learning_rate": 3.121247763262235e-06,
+      "loss": 0.1155,
+      "step": 13500
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7000632882118225,
+      "learning_rate": 3.1093161121684118e-06,
+      "loss": 0.1166,
+      "step": 13550
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7171856164932251,
+      "learning_rate": 3.097369687003896e-06,
+      "loss": 0.1129,
+      "step": 13600
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7927132248878479,
+      "learning_rate": 3.085408777433323e-06,
+      "loss": 0.1126,
+      "step": 13650
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7410084009170532,
+      "learning_rate": 3.0734336734725327e-06,
+      "loss": 0.116,
+      "step": 13700
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8224751949310303,
+      "learning_rate": 3.0614446654815346e-06,
+      "loss": 0.1119,
+      "step": 13750
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.790632963180542,
+      "learning_rate": 3.049442044157469e-06,
+      "loss": 0.1122,
+      "step": 13800
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6657869815826416,
+      "learning_rate": 3.0374261005275606e-06,
+      "loss": 0.1105,
+      "step": 13850
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6833150386810303,
+      "learning_rate": 3.025397125942056e-06,
+      "loss": 0.1158,
+      "step": 13900
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7890775203704834,
+      "learning_rate": 3.0133554120671653e-06,
+      "loss": 0.1158,
+      "step": 13950
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6974292397499084,
+      "learning_rate": 3.001301250877987e-06,
+      "loss": 0.1121,
+      "step": 14000
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7317948341369629,
+      "learning_rate": 2.9892349346514306e-06,
+      "loss": 0.1111,
+      "step": 14050
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8558988571166992,
+      "learning_rate": 2.977156755959126e-06,
+      "loss": 0.113,
+      "step": 14100
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6563243865966797,
+      "learning_rate": 2.9650670076603342e-06,
+      "loss": 0.1121,
+      "step": 14150
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7113382816314697,
+      "learning_rate": 2.952965982894844e-06,
+      "loss": 0.1082,
+      "step": 14200
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6555420160293579,
+      "learning_rate": 2.9408539750758625e-06,
+      "loss": 0.1123,
+      "step": 14250
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7071360945701599,
+      "learning_rate": 2.9287312778829047e-06,
+      "loss": 0.1146,
+      "step": 14300
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6903111934661865,
+      "learning_rate": 2.9165981852546688e-06,
+      "loss": 0.1151,
+      "step": 14350
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6903436779975891,
+      "learning_rate": 2.9044549913819125e-06,
+      "loss": 0.111,
+      "step": 14400
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7512865662574768,
+      "learning_rate": 2.892301990700316e-06,
+      "loss": 0.1081,
+      "step": 14450
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6243235468864441,
+      "learning_rate": 2.8801394778833475e-06,
+      "loss": 0.1101,
+      "step": 14500
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.761928141117096,
+      "learning_rate": 2.8679677478351147e-06,
+      "loss": 0.1124,
+      "step": 14550
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6263371706008911,
+      "learning_rate": 2.8557870956832135e-06,
+      "loss": 0.1107,
+      "step": 14600
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5973655581474304,
+      "learning_rate": 2.8435978167715753e-06,
+      "loss": 0.1108,
+      "step": 14650
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6588687896728516,
+      "learning_rate": 2.8314002066533053e-06,
+      "loss": 0.1109,
+      "step": 14700
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.640466034412384,
+      "learning_rate": 2.8191945610835138e-06,
+      "loss": 0.1071,
+      "step": 14750
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7230612635612488,
+      "learning_rate": 2.8069811760121463e-06,
+      "loss": 0.1079,
+      "step": 14800
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6982682347297668,
+      "learning_rate": 2.794760347576809e-06,
+      "loss": 0.1137,
+      "step": 14850
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6350643038749695,
+      "learning_rate": 2.7825323720955853e-06,
+      "loss": 0.1053,
+      "step": 14900
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7661166191101074,
+      "learning_rate": 2.7702975460598545e-06,
+      "loss": 0.1109,
+      "step": 14950
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6969360113143921,
+      "learning_rate": 2.7580561661271015e-06,
+      "loss": 0.1099,
+      "step": 15000
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8628710508346558,
+      "learning_rate": 2.7458085291137213e-06,
+      "loss": 0.1073,
+      "step": 15050
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6863730549812317,
+      "learning_rate": 2.733554931987825e-06,
+      "loss": 0.1067,
+      "step": 15100
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6106501817703247,
+      "learning_rate": 2.7212956718620404e-06,
+      "loss": 0.1083,
+      "step": 15150
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7088560461997986,
+      "learning_rate": 2.709031045986302e-06,
+      "loss": 0.1074,
+      "step": 15200
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6315382122993469,
+      "learning_rate": 2.6967613517406514e-06,
+      "loss": 0.11,
+      "step": 15250
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8497392535209656,
+      "learning_rate": 2.68448688662802e-06,
+      "loss": 0.1063,
+      "step": 15300
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7048566341400146,
+      "learning_rate": 2.6722079482670196e-06,
+      "loss": 0.1044,
+      "step": 15350
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6477850675582886,
+      "learning_rate": 2.6599248343847243e-06,
+      "loss": 0.112,
+      "step": 15400
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5776751041412354,
+      "learning_rate": 2.6476378428094523e-06,
+      "loss": 0.1085,
+      "step": 15450
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7395492792129517,
+      "learning_rate": 2.6353472714635443e-06,
+      "loss": 0.1084,
+      "step": 15500
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7547775506973267,
+      "learning_rate": 2.6230534183561385e-06,
+      "loss": 0.1131,
+      "step": 15550
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6937977075576782,
+      "learning_rate": 2.6107565815759473e-06,
+      "loss": 0.1097,
+      "step": 15600
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6883909702301025,
+      "learning_rate": 2.598457059284027e-06,
+      "loss": 0.114,
+      "step": 15650
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7592599987983704,
+      "learning_rate": 2.5861551497065497e-06,
+      "loss": 0.1093,
+      "step": 15700
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6865835189819336,
+      "learning_rate": 2.5738511511275716e-06,
+      "loss": 0.1127,
+      "step": 15750
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6286877989768982,
+      "learning_rate": 2.5615453618818033e-06,
+      "loss": 0.1143,
+      "step": 15800
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6750493049621582,
+      "learning_rate": 2.5492380803473705e-06,
+      "loss": 0.1079,
+      "step": 15850
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6905640959739685,
+      "learning_rate": 2.5369296049385837e-06,
+      "loss": 0.1096,
+      "step": 15900
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6613180041313171,
+      "learning_rate": 2.5246202340987004e-06,
+      "loss": 0.1103,
+      "step": 15950
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7040888071060181,
+      "learning_rate": 2.5123102662926912e-06,
+      "loss": 0.1038,
+      "step": 16000
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7578396797180176,
+      "learning_rate": 2.5e-06,
+      "loss": 0.1087,
+      "step": 16050
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.587769627571106,
+      "learning_rate": 2.4876897337073105e-06,
+      "loss": 0.1075,
+      "step": 16100
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.688367486000061,
+      "learning_rate": 2.475379765901301e-06,
+      "loss": 0.1088,
+      "step": 16150
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7524093985557556,
+      "learning_rate": 2.4630703950614176e-06,
+      "loss": 0.1094,
+      "step": 16200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6418857574462891,
+      "learning_rate": 2.45076191965263e-06,
+      "loss": 0.1057,
+      "step": 16250
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7005917429924011,
+      "learning_rate": 2.4384546381181975e-06,
+      "loss": 0.111,
+      "step": 16300
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5830152630805969,
+      "learning_rate": 2.4261488488724284e-06,
+      "loss": 0.1066,
+      "step": 16350
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6886410117149353,
+      "learning_rate": 2.413844850293451e-06,
+      "loss": 0.1075,
+      "step": 16400
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7370522022247314,
+      "learning_rate": 2.4015429407159746e-06,
+      "loss": 0.1081,
+      "step": 16450
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.681659996509552,
+      "learning_rate": 2.3892434184240536e-06,
+      "loss": 0.1091,
+      "step": 16500
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8225075006484985,
+      "learning_rate": 2.3769465816438627e-06,
+      "loss": 0.1119,
+      "step": 16550
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8517566919326782,
+      "learning_rate": 2.3646527285364565e-06,
+      "loss": 0.1167,
+      "step": 16600
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7776892185211182,
+      "learning_rate": 2.3523621571905485e-06,
+      "loss": 0.1091,
+      "step": 16650
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6754997968673706,
+      "learning_rate": 2.340075165615276e-06,
+      "loss": 0.107,
+      "step": 16700
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6805914640426636,
+      "learning_rate": 2.3277920517329813e-06,
+      "loss": 0.1178,
+      "step": 16750
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7494375705718994,
+      "learning_rate": 2.315513113371981e-06,
+      "loss": 0.1122,
+      "step": 16800
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7486910223960876,
+      "learning_rate": 2.303238648259349e-06,
+      "loss": 0.1109,
+      "step": 16850
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7235914468765259,
+      "learning_rate": 2.2909689540136986e-06,
+      "loss": 0.1081,
+      "step": 16900
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9256291389465332,
+      "learning_rate": 2.27870432813796e-06,
+      "loss": 0.1097,
+      "step": 16950
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7309363484382629,
+      "learning_rate": 2.2664450680121757e-06,
+      "loss": 0.1137,
+      "step": 17000
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6349431872367859,
+      "learning_rate": 2.254191470886279e-06,
+      "loss": 0.1058,
+      "step": 17050
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7812219262123108,
+      "learning_rate": 2.241943833872899e-06,
+      "loss": 0.1109,
+      "step": 17100
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7502079010009766,
+      "learning_rate": 2.2297024539401463e-06,
+      "loss": 0.1092,
+      "step": 17150
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6485741138458252,
+      "learning_rate": 2.2174676279044155e-06,
+      "loss": 0.1086,
+      "step": 17200
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6442984938621521,
+      "learning_rate": 2.2052396524231924e-06,
+      "loss": 0.1026,
+      "step": 17250
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6773645281791687,
+      "learning_rate": 2.193018823987854e-06,
+      "loss": 0.1055,
+      "step": 17300
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7892882227897644,
+      "learning_rate": 2.180805438916487e-06,
+      "loss": 0.1077,
+      "step": 17350
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8464773297309875,
+      "learning_rate": 2.1685997933466947e-06,
+      "loss": 0.1083,
+      "step": 17400
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7581765651702881,
+      "learning_rate": 2.1564021832284255e-06,
+      "loss": 0.1081,
+      "step": 17450
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7814217805862427,
+      "learning_rate": 2.1442129043167877e-06,
+      "loss": 0.1067,
+      "step": 17500
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6047621965408325,
+      "learning_rate": 2.1320322521648857e-06,
+      "loss": 0.1065,
+      "step": 17550
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7465390563011169,
+      "learning_rate": 2.119860522116653e-06,
+      "loss": 0.1058,
+      "step": 17600
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7619149684906006,
+      "learning_rate": 2.1076980092996845e-06,
+      "loss": 0.109,
+      "step": 17650
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7714558839797974,
+      "learning_rate": 2.0955450086180883e-06,
+      "loss": 0.1089,
+      "step": 17700
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6427506804466248,
+      "learning_rate": 2.083401814745332e-06,
+      "loss": 0.1086,
+      "step": 17750
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6711663603782654,
+      "learning_rate": 2.071268722117096e-06,
+      "loss": 0.107,
+      "step": 17800
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7448789477348328,
+      "learning_rate": 2.0591460249241383e-06,
+      "loss": 0.1054,
+      "step": 17850
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6524811387062073,
+      "learning_rate": 2.0470340171051567e-06,
+      "loss": 0.1135,
+      "step": 17900
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6184554100036621,
+      "learning_rate": 2.034932992339666e-06,
+      "loss": 0.1093,
+      "step": 17950
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7184391021728516,
+      "learning_rate": 2.022843244040874e-06,
+      "loss": 0.1039,
+      "step": 18000
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6897612810134888,
+      "learning_rate": 2.0107650653485707e-06,
+      "loss": 0.111,
+      "step": 18050
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7289073467254639,
+      "learning_rate": 1.998698749122014e-06,
+      "loss": 0.1083,
+      "step": 18100
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6289246678352356,
+      "learning_rate": 1.986644587932835e-06,
+      "loss": 0.1062,
+      "step": 18150
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7373926639556885,
+      "learning_rate": 1.9746028740579453e-06,
+      "loss": 0.1083,
+      "step": 18200
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6934205293655396,
+      "learning_rate": 1.96257389947244e-06,
+      "loss": 0.108,
+      "step": 18250
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6576781868934631,
+      "learning_rate": 1.9505579558425315e-06,
+      "loss": 0.105,
+      "step": 18300
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6750394105911255,
+      "learning_rate": 1.938555334518466e-06,
+      "loss": 0.1098,
+      "step": 18350
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6316762566566467,
+      "learning_rate": 1.926566326527468e-06,
+      "loss": 0.1063,
+      "step": 18400
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7245384454727173,
+      "learning_rate": 1.914591222566678e-06,
+      "loss": 0.1066,
+      "step": 18450
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6794847249984741,
+      "learning_rate": 1.9026303129961049e-06,
+      "loss": 0.1081,
+      "step": 18500
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6249060034751892,
+      "learning_rate": 1.8906838878315886e-06,
+      "loss": 0.1092,
+      "step": 18550
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7857920527458191,
+      "learning_rate": 1.878752236737765e-06,
+      "loss": 0.1088,
+      "step": 18600
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6644686460494995,
+      "learning_rate": 1.8668356490210449e-06,
+      "loss": 0.1003,
+      "step": 18650
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8369055986404419,
+      "learning_rate": 1.8549344136225946e-06,
+      "loss": 0.1091,
+      "step": 18700
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7103849649429321,
+      "learning_rate": 1.8430488191113373e-06,
+      "loss": 0.1016,
+      "step": 18750
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.655951976776123,
+      "learning_rate": 1.8311791536769485e-06,
+      "loss": 0.1063,
+      "step": 18800
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7587403059005737,
+      "learning_rate": 1.819325705122873e-06,
+      "loss": 0.1043,
+      "step": 18850
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7093736529350281,
+      "learning_rate": 1.8074887608593477e-06,
+      "loss": 0.1092,
+      "step": 18900
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5937142968177795,
+      "learning_rate": 1.7956686078964257e-06,
+      "loss": 0.1045,
+      "step": 18950
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6976079344749451,
+      "learning_rate": 1.7838655328370268e-06,
+      "loss": 0.1127,
+      "step": 19000
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6277263760566711,
+      "learning_rate": 1.7720798218699798e-06,
+      "loss": 0.1115,
+      "step": 19050
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7694005966186523,
+      "learning_rate": 1.7603117607630892e-06,
+      "loss": 0.1068,
+      "step": 19100
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6981104612350464,
+      "learning_rate": 1.7485616348562023e-06,
+      "loss": 0.1053,
+      "step": 19150
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.816888153553009,
+      "learning_rate": 1.7368297290542918e-06,
+      "loss": 0.1084,
+      "step": 19200
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7193763852119446,
+      "learning_rate": 1.72511632782055e-06,
+      "loss": 0.1056,
+      "step": 19250
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6017695069313049,
+      "learning_rate": 1.7134217151694873e-06,
+      "loss": 0.1033,
+      "step": 19300
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7289357781410217,
+      "learning_rate": 1.7017461746600506e-06,
+      "loss": 0.1092,
+      "step": 19350
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7104734182357788,
+      "learning_rate": 1.690089989388741e-06,
+      "loss": 0.1075,
+      "step": 19400
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.707067608833313,
+      "learning_rate": 1.678453441982758e-06,
+      "loss": 0.1093,
+      "step": 19450
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7596837878227234,
+      "learning_rate": 1.66683681459314e-06,
+      "loss": 0.1042,
+      "step": 19500
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7431414723396301,
+      "learning_rate": 1.6552403888879243e-06,
+      "loss": 0.1034,
+      "step": 19550
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7935433387756348,
+      "learning_rate": 1.6436644460453218e-06,
+      "loss": 0.1101,
+      "step": 19600
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7206917405128479,
+      "learning_rate": 1.6321092667468926e-06,
+      "loss": 0.104,
+      "step": 19650
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7832915782928467,
+      "learning_rate": 1.6205751311707463e-06,
+      "loss": 0.106,
+      "step": 19700
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6786001324653625,
+      "learning_rate": 1.6090623189847443e-06,
+      "loss": 0.1044,
+      "step": 19750
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.739486038684845,
+      "learning_rate": 1.5975711093397223e-06,
+      "loss": 0.1058,
+      "step": 19800
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6627418398857117,
+      "learning_rate": 1.5861017808627167e-06,
+      "loss": 0.1111,
+      "step": 19850
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7300294637680054,
+      "learning_rate": 1.574654611650214e-06,
+      "loss": 0.1076,
+      "step": 19900
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8272718191146851,
+      "learning_rate": 1.5632298792614064e-06,
+      "loss": 0.1078,
+      "step": 19950
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7260649800300598,
+      "learning_rate": 1.5518278607114585e-06,
+      "loss": 0.1068,
+      "step": 20000
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7374041080474854,
+      "learning_rate": 1.540448832464796e-06,
+      "loss": 0.1083,
+      "step": 20050
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7387140989303589,
+      "learning_rate": 1.5290930704283953e-06,
+      "loss": 0.104,
+      "step": 20100
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.784419059753418,
+      "learning_rate": 1.517760849945103e-06,
+      "loss": 0.1073,
+      "step": 20150
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6998342275619507,
+      "learning_rate": 1.5064524457869506e-06,
+      "loss": 0.1079,
+      "step": 20200
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7088004946708679,
+      "learning_rate": 1.4951681321484952e-06,
+      "loss": 0.1051,
+      "step": 20250
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8070769906044006,
+      "learning_rate": 1.4839081826401756e-06,
+      "loss": 0.1053,
+      "step": 20300
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7509734034538269,
+      "learning_rate": 1.47267287028167e-06,
+      "loss": 0.1087,
+      "step": 20350
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7456051111221313,
+      "learning_rate": 1.4614624674952843e-06,
+      "loss": 0.1058,
+      "step": 20400
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7578474879264832,
+      "learning_rate": 1.4502772460993387e-06,
+      "loss": 0.1087,
+      "step": 20450
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7839920520782471,
+      "learning_rate": 1.4391174773015836e-06,
+      "loss": 0.1081,
+      "step": 20500
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7195042371749878,
+      "learning_rate": 1.4279834316926217e-06,
+      "loss": 0.1072,
+      "step": 20550
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6250014901161194,
+      "learning_rate": 1.4168753792393413e-06,
+      "loss": 0.1052,
+      "step": 20600
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5411588549613953,
+      "learning_rate": 1.405793589278382e-06,
+      "loss": 0.1015,
+      "step": 20650
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7470389008522034,
+      "learning_rate": 1.394738330509593e-06,
+      "loss": 0.1043,
+      "step": 20700
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7059767842292786,
+      "learning_rate": 1.3837098709895246e-06,
+      "loss": 0.1054,
+      "step": 20750
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8912293314933777,
+      "learning_rate": 1.3727084781249251e-06,
+      "loss": 0.1066,
+      "step": 20800
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7008835673332214,
+      "learning_rate": 1.3617344186662585e-06,
+      "loss": 0.105,
+      "step": 20850
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8228752017021179,
+      "learning_rate": 1.3507879587012378e-06,
+      "loss": 0.1056,
+      "step": 20900
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7580632567405701,
+      "learning_rate": 1.3398693636483707e-06,
+      "loss": 0.1061,
+      "step": 20950
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5401976704597473,
+      "learning_rate": 1.328978898250525e-06,
+      "loss": 0.1054,
+      "step": 21000
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7178921103477478,
+      "learning_rate": 1.31811682656851e-06,
+      "loss": 0.1048,
+      "step": 21050
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.701740026473999,
+      "learning_rate": 1.307283411974674e-06,
+      "loss": 0.1044,
+      "step": 21100
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6609808206558228,
+      "learning_rate": 1.2964789171465164e-06,
+      "loss": 0.1033,
+      "step": 21150
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8305937051773071,
+      "learning_rate": 1.2857036040603204e-06,
+      "loss": 0.1073,
+      "step": 21200
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7639765739440918,
+      "learning_rate": 1.2749577339848007e-06,
+      "loss": 0.1092,
+      "step": 21250
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7896683216094971,
+      "learning_rate": 1.2642415674747675e-06,
+      "loss": 0.1016,
+      "step": 21300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7455915808677673,
+      "learning_rate": 1.25355536436481e-06,
+      "loss": 0.1063,
+      "step": 21350
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6086365580558777,
+      "learning_rate": 1.2428993837629943e-06,
+      "loss": 0.1059,
+      "step": 21400
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7196105122566223,
+      "learning_rate": 1.2322738840445867e-06,
+      "loss": 0.1019,
+      "step": 21450
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7521807551383972,
+      "learning_rate": 1.2216791228457778e-06,
+      "loss": 0.1079,
+      "step": 21500
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6882525682449341,
+      "learning_rate": 1.2111153570574454e-06,
+      "loss": 0.1058,
+      "step": 21550
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7660160064697266,
+      "learning_rate": 1.2005828428189256e-06,
+      "loss": 0.1048,
+      "step": 21600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7205012440681458,
+      "learning_rate": 1.1900818355117918e-06,
+      "loss": 0.1028,
+      "step": 21650
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6535946726799011,
+      "learning_rate": 1.1796125897536782e-06,
+      "loss": 0.1061,
+      "step": 21700
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7131094336509705,
+      "learning_rate": 1.1691753593920884e-06,
+      "loss": 0.1053,
+      "step": 21750
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8452409505844116,
+      "learning_rate": 1.1587703974982583e-06,
+      "loss": 0.104,
+      "step": 21800
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7793323993682861,
+      "learning_rate": 1.148397956361007e-06,
+      "loss": 0.1055,
+      "step": 21850
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6791970729827881,
+      "learning_rate": 1.1380582874806208e-06,
+      "loss": 0.1047,
+      "step": 21900
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7349908351898193,
+      "learning_rate": 1.127751641562765e-06,
+      "loss": 0.1038,
+      "step": 21950
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.61646968126297,
+      "learning_rate": 1.1174782685123919e-06,
+      "loss": 0.1047,
+      "step": 22000
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8146559000015259,
+      "learning_rate": 1.107238417427691e-06,
+      "loss": 0.1041,
+      "step": 22050
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7460653185844421,
+      "learning_rate": 1.0970323365940443e-06,
+      "loss": 0.1069,
+      "step": 22100
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7389132976531982,
+      "learning_rate": 1.0868602734780075e-06,
+      "loss": 0.1038,
+      "step": 22150
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8619258999824524,
+      "learning_rate": 1.0767224747213102e-06,
+      "loss": 0.1029,
+      "step": 22200
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.814842939376831,
+      "learning_rate": 1.0666191861348741e-06,
+      "loss": 0.1063,
+      "step": 22250
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7106545567512512,
+      "learning_rate": 1.0565506526928548e-06,
+      "loss": 0.1023,
+      "step": 22300
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7529650330543518,
+      "learning_rate": 1.0465171185267007e-06,
+      "loss": 0.1068,
+      "step": 22350
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7822823524475098,
+      "learning_rate": 1.036518826919233e-06,
+      "loss": 0.1063,
+      "step": 22400
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7902446389198303,
+      "learning_rate": 1.0265560202987474e-06,
+      "loss": 0.1038,
+      "step": 22450
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8036994338035583,
+      "learning_rate": 1.0166289402331391e-06,
+      "loss": 0.1015,
+      "step": 22500
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6801638603210449,
+      "learning_rate": 1.006737827424038e-06,
+      "loss": 0.1115,
+      "step": 22550
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.66480952501297,
+      "learning_rate": 9.9688292170098e-07,
+      "loss": 0.1067,
+      "step": 22600
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6099157929420471,
+      "learning_rate": 9.870644620155878e-07,
+      "loss": 0.1002,
+      "step": 22650
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7186796069145203,
+      "learning_rate": 9.77282686435777e-07,
+      "loss": 0.1035,
+      "step": 22700
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7684562802314758,
+      "learning_rate": 9.67537832139989e-07,
+      "loss": 0.1047,
+      "step": 22750
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.804244875907898,
+      "learning_rate": 9.578301354114292e-07,
+      "loss": 0.104,
+      "step": 22800
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7283428311347961,
+      "learning_rate": 9.481598316323504e-07,
+      "loss": 0.1053,
+      "step": 22850
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7441770434379578,
+      "learning_rate": 9.385271552783376e-07,
+      "loss": 0.1003,
+      "step": 22900
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6516032814979553,
+      "learning_rate": 9.289323399126216e-07,
+      "loss": 0.1041,
+      "step": 22950
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7602024078369141,
+      "learning_rate": 9.193756181804248e-07,
+      "loss": 0.1047,
+      "step": 23000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.70171058177948,
+      "learning_rate": 9.098572218033084e-07,
+      "loss": 0.1043,
+      "step": 23050
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6728523969650269,
+      "learning_rate": 9.003773815735644e-07,
+      "loss": 0.103,
+      "step": 23100
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7401735186576843,
+      "learning_rate": 8.90936327348613e-07,
+      "loss": 0.106,
+      "step": 23150
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7361724972724915,
+      "learning_rate": 8.815342880454312e-07,
+      "loss": 0.1044,
+      "step": 23200
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8209415674209595,
+      "learning_rate": 8.721714916350019e-07,
+      "loss": 0.1063,
+      "step": 23250
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8205540776252747,
+      "learning_rate": 8.628481651367876e-07,
+      "loss": 0.1046,
+      "step": 23300
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6642114520072937,
+      "learning_rate": 8.535645346132246e-07,
+      "loss": 0.108,
+      "step": 23350
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6761398315429688,
+      "learning_rate": 8.443208251642418e-07,
+      "loss": 0.1041,
+      "step": 23400
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.821124255657196,
+      "learning_rate": 8.351172609218033e-07,
+      "loss": 0.1036,
+      "step": 23450
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6441815495491028,
+      "learning_rate": 8.259540650444736e-07,
+      "loss": 0.1103,
+      "step": 23500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6896815299987793,
+      "learning_rate": 8.168314597120059e-07,
+      "loss": 0.1049,
+      "step": 23550
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7527472972869873,
+      "learning_rate": 8.077496661199557e-07,
+      "loss": 0.1094,
+      "step": 23600
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.788384199142456,
+      "learning_rate": 7.987089044743182e-07,
+      "loss": 0.1071,
+      "step": 23650
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7462133169174194,
+      "learning_rate": 7.897093939861878e-07,
+      "loss": 0.1071,
+      "step": 23700
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6221845746040344,
+      "learning_rate": 7.807513528664415e-07,
+      "loss": 0.1026,
+      "step": 23750
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7602658271789551,
+      "learning_rate": 7.71834998320454e-07,
+      "loss": 0.1064,
+      "step": 23800
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6582159399986267,
+      "learning_rate": 7.629605465428211e-07,
+      "loss": 0.1013,
+      "step": 23850
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9127190709114075,
+      "learning_rate": 7.541282127121291e-07,
+      "loss": 0.1037,
+      "step": 23900
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7410569787025452,
+      "learning_rate": 7.453382109857269e-07,
+      "loss": 0.1041,
+      "step": 23950
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6176531314849854,
+      "learning_rate": 7.365907544945398e-07,
+      "loss": 0.1076,
+      "step": 24000
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6940147280693054,
+      "learning_rate": 7.27886055337902e-07,
+      "loss": 0.1036,
+      "step": 24050
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7587239146232605,
+      "learning_rate": 7.192243245784075e-07,
+      "loss": 0.1047,
+      "step": 24100
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6673498153686523,
+      "learning_rate": 7.106057722368012e-07,
+      "loss": 0.1021,
+      "step": 24150
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7008479833602905,
+      "learning_rate": 7.020306072868804e-07,
+      "loss": 0.1056,
+      "step": 24200
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6808987259864807,
+      "learning_rate": 6.934990376504269e-07,
+      "loss": 0.1005,
+      "step": 24250
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6858727931976318,
+      "learning_rate": 6.850112701921735e-07,
+      "loss": 0.1043,
+      "step": 24300
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.804255485534668,
+      "learning_rate": 6.76567510714777e-07,
+      "loss": 0.1027,
+      "step": 24350
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7274004817008972,
+      "learning_rate": 6.681679639538388e-07,
+      "loss": 0.1053,
+      "step": 24400
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6244350075721741,
+      "learning_rate": 6.598128335729332e-07,
+      "loss": 0.1016,
+      "step": 24450
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8558462858200073,
+      "learning_rate": 6.515023221586722e-07,
+      "loss": 0.1015,
+      "step": 24500
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.678629457950592,
+      "learning_rate": 6.432366312157933e-07,
+      "loss": 0.1057,
+      "step": 24550
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6484153866767883,
+      "learning_rate": 6.35015961162273e-07,
+      "loss": 0.1015,
+      "step": 24600
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7038909196853638,
+      "learning_rate": 6.268405113244677e-07,
+      "loss": 0.1065,
+      "step": 24650
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8573065996170044,
+      "learning_rate": 6.187104799322805e-07,
+      "loss": 0.1029,
+      "step": 24700
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8027458190917969,
+      "learning_rate": 6.106260641143547e-07,
+      "loss": 0.1032,
+      "step": 24750
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0594960451126099,
+      "learning_rate": 6.025874598932937e-07,
+      "loss": 0.1016,
+      "step": 24800
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7134573459625244,
+      "learning_rate": 5.945948621809092e-07,
+      "loss": 0.1075,
+      "step": 24850
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.732449471950531,
+      "learning_rate": 5.866484647734935e-07,
+      "loss": 0.1025,
+      "step": 24900
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7451493740081787,
+      "learning_rate": 5.787484603471221e-07,
+      "loss": 0.1048,
+      "step": 24950
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6442447304725647,
+      "learning_rate": 5.708950404529812e-07,
+      "loss": 0.0989,
+      "step": 25000
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6639313101768494,
+      "learning_rate": 5.630883955127211e-07,
+      "loss": 0.106,
+      "step": 25050
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8314816951751709,
+      "learning_rate": 5.553287148138462e-07,
+      "loss": 0.1078,
+      "step": 25100
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6702859997749329,
+      "learning_rate": 5.47616186505113e-07,
+      "loss": 0.1063,
+      "step": 25150
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6533217430114746,
+      "learning_rate": 5.399509975919828e-07,
+      "loss": 0.1015,
+      "step": 25200
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7391800284385681,
+      "learning_rate": 5.323333339320739e-07,
+      "loss": 0.1027,
+      "step": 25250
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6083396077156067,
+      "learning_rate": 5.247633802306637e-07,
+      "loss": 0.1045,
+      "step": 25300
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8118199110031128,
+      "learning_rate": 5.172413200362092e-07,
+      "loss": 0.1021,
+      "step": 25350
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7648591995239258,
+      "learning_rate": 5.097673357358906e-07,
+      "loss": 0.1046,
+      "step": 25400
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7872638702392578,
+      "learning_rate": 5.023416085511976e-07,
+      "loss": 0.1024,
+      "step": 25450
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8803538084030151,
+      "learning_rate": 4.949643185335288e-07,
+      "loss": 0.1055,
+      "step": 25500
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7241789698600769,
+      "learning_rate": 4.876356445598279e-07,
+      "loss": 0.1049,
+      "step": 25550
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7358962893486023,
+      "learning_rate": 4.803557643282486e-07,
+      "loss": 0.1027,
+      "step": 25600
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9544288516044617,
+      "learning_rate": 4.731248543538405e-07,
+      "loss": 0.1075,
+      "step": 25650
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7101548314094543,
+      "learning_rate": 4.6594308996427696e-07,
+      "loss": 0.1051,
+      "step": 25700
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7388818264007568,
+      "learning_rate": 4.588106452955973e-07,
+      "loss": 0.1059,
+      "step": 25750
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7218101024627686,
+      "learning_rate": 4.517276932879877e-07,
+      "loss": 0.1042,
+      "step": 25800
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7213023900985718,
+      "learning_rate": 4.446944056815866e-07,
+      "loss": 0.1036,
+      "step": 25850
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7361165881156921,
+      "learning_rate": 4.377109530123216e-07,
+      "loss": 0.1019,
+      "step": 25900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7176214456558228,
+      "learning_rate": 4.307775046077739e-07,
+      "loss": 0.1054,
+      "step": 25950
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8045145273208618,
+      "learning_rate": 4.2389422858307244e-07,
+      "loss": 0.1017,
+      "step": 26000
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9221975803375244,
+      "learning_rate": 4.1706129183681834e-07,
+      "loss": 0.1016,
+      "step": 26050
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.749830961227417,
+      "learning_rate": 4.10278860047037e-07,
+      "loss": 0.1008,
+      "step": 26100
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.719508707523346,
+      "learning_rate": 4.035470976671621e-07,
+      "loss": 0.1026,
+      "step": 26150
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7058319449424744,
+      "learning_rate": 3.9686616792204677e-07,
+      "loss": 0.1009,
+      "step": 26200
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6976651549339294,
+      "learning_rate": 3.902362328040091e-07,
+      "loss": 0.1057,
+      "step": 26250
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8557724952697754,
+      "learning_rate": 3.836574530688983e-07,
+      "loss": 0.1045,
+      "step": 26300
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6408629417419434,
+      "learning_rate": 3.7712998823220243e-07,
+      "loss": 0.1025,
+      "step": 26350
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6862276196479797,
+      "learning_rate": 3.7065399656517955e-07,
+      "loss": 0.1009,
+      "step": 26400
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6881060600280762,
+      "learning_rate": 3.6422963509101626e-07,
+      "loss": 0.1016,
+      "step": 26450
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7063938975334167,
+      "learning_rate": 3.578570595810274e-07,
+      "loss": 0.1059,
+      "step": 26500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7486748695373535,
+      "learning_rate": 3.515364245508704e-07,
+      "loss": 0.1042,
+      "step": 26550
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7359747886657715,
+      "learning_rate": 3.452678832568071e-07,
+      "loss": 0.1056,
+      "step": 26600
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.666910707950592,
+      "learning_rate": 3.390515876919831e-07,
+      "loss": 0.105,
+      "step": 26650
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7717428207397461,
+      "learning_rate": 3.328876885827406e-07,
+      "loss": 0.0995,
+      "step": 26700
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6869860887527466,
+      "learning_rate": 3.267763353849704e-07,
+      "loss": 0.103,
+      "step": 26750
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7728705406188965,
+      "learning_rate": 3.207176762804814e-07,
+      "loss": 0.1002,
+      "step": 26800
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.670854389667511,
+      "learning_rate": 3.1471185817341153e-07,
+      "loss": 0.1047,
+      "step": 26850
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8178605437278748,
+      "learning_rate": 3.0875902668666386e-07,
+      "loss": 0.0986,
+      "step": 26900
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.700291097164154,
+      "learning_rate": 3.0285932615837646e-07,
+      "loss": 0.1027,
+      "step": 26950
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8143967986106873,
+      "learning_rate": 2.970128996384228e-07,
+      "loss": 0.0978,
+      "step": 27000
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5957488417625427,
+      "learning_rate": 2.9121988888494297e-07,
+      "loss": 0.1012,
+      "step": 27050
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8061670660972595,
+      "learning_rate": 2.854804343609058e-07,
+      "loss": 0.1018,
+      "step": 27100
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7996898293495178,
+      "learning_rate": 2.7979467523070484e-07,
+      "loss": 0.103,
+      "step": 27150
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.712619960308075,
+      "learning_rate": 2.741627493567822e-07,
+      "loss": 0.1058,
+      "step": 27200
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9328902959823608,
+      "learning_rate": 2.685847932962868e-07,
+      "loss": 0.1032,
+      "step": 27250
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.642899751663208,
+      "learning_rate": 2.630609422977623e-07,
+      "loss": 0.1038,
+      "step": 27300
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8721086382865906,
+      "learning_rate": 2.575913302978697e-07,
+      "loss": 0.0986,
+      "step": 27350
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6639971733093262,
+      "learning_rate": 2.5217608991813774e-07,
+      "loss": 0.1029,
+      "step": 27400
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6409977078437805,
+      "learning_rate": 2.468153524617478e-07,
+      "loss": 0.1008,
+      "step": 27450
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8069073557853699,
+      "learning_rate": 2.4150924791035037e-07,
+      "loss": 0.1038,
+      "step": 27500
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6896743774414062,
+      "learning_rate": 2.3625790492091544e-07,
+      "loss": 0.1013,
+      "step": 27550
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7691417932510376,
+      "learning_rate": 2.3106145082260777e-07,
+      "loss": 0.1015,
+      "step": 27600
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7701432704925537,
+      "learning_rate": 2.2592001161370392e-07,
+      "loss": 0.1023,
+      "step": 27650
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7143141031265259,
+      "learning_rate": 2.2083371195853797e-07,
+      "loss": 0.1034,
+      "step": 27700
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7811148762702942,
+      "learning_rate": 2.158026751844733e-07,
+      "loss": 0.1052,
+      "step": 27750
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6876391172409058,
+      "learning_rate": 2.1082702327891918e-07,
+      "loss": 0.0985,
+      "step": 27800
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7319098711013794,
+      "learning_rate": 2.0590687688636619e-07,
+      "loss": 0.1036,
+      "step": 27850
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7728010416030884,
+      "learning_rate": 2.0104235530546745e-07,
+      "loss": 0.1025,
+      "step": 27900
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7075905203819275,
+      "learning_rate": 1.9623357648614088e-07,
+      "loss": 0.1025,
+      "step": 27950
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8835753202438354,
+      "learning_rate": 1.914806570267111e-07,
+      "loss": 0.1045,
+      "step": 28000
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6914262771606445,
+      "learning_rate": 1.8678371217108387e-07,
+      "loss": 0.1012,
+      "step": 28050
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7682451605796814,
+      "learning_rate": 1.821428558059493e-07,
+      "loss": 0.0987,
+      "step": 28100
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6042952537536621,
+      "learning_rate": 1.7755820045802146e-07,
+      "loss": 0.1033,
+      "step": 28150
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9065592885017395,
+      "learning_rate": 1.7302985729131e-07,
+      "loss": 0.1067,
+      "step": 28200
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7775442004203796,
+      "learning_rate": 1.6855793610442484e-07,
+      "loss": 0.1019,
+      "step": 28250
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7581862807273865,
+      "learning_rate": 1.6414254532791357e-07,
+      "loss": 0.1043,
+      "step": 28300
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7499451041221619,
+      "learning_rate": 1.5978379202163275e-07,
+      "loss": 0.1046,
+      "step": 28350
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7929815053939819,
+      "learning_rate": 1.554817818721513e-07,
+      "loss": 0.1054,
+      "step": 28400
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8825467228889465,
+      "learning_rate": 1.51236619190189e-07,
+      "loss": 0.1034,
+      "step": 28450
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6325812339782715,
+      "learning_rate": 1.4704840690808658e-07,
+      "loss": 0.105,
+      "step": 28500
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6791960000991821,
+      "learning_rate": 1.4291724657730904e-07,
+      "loss": 0.1054,
+      "step": 28550
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.696438729763031,
+      "learning_rate": 1.3884323836598656e-07,
+      "loss": 0.1007,
+      "step": 28600
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.638707160949707,
+      "learning_rate": 1.348264810564809e-07,
+      "loss": 0.0989,
+      "step": 28650
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9399768114089966,
+      "learning_rate": 1.3086707204299415e-07,
+      "loss": 0.1041,
+      "step": 28700
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7409209609031677,
+      "learning_rate": 1.269651073292058e-07,
+      "loss": 0.1028,
+      "step": 28750
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7949596047401428,
+      "learning_rate": 1.2312068152594448e-07,
+      "loss": 0.1025,
+      "step": 28800
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8088447451591492,
+      "learning_rate": 1.1933388784889617e-07,
+      "loss": 0.1034,
+      "step": 28850
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7876179814338684,
+      "learning_rate": 1.1560481811633911e-07,
+      "loss": 0.1007,
+      "step": 28900
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6866023540496826,
+      "learning_rate": 1.1193356274692424e-07,
+      "loss": 0.1031,
+      "step": 28950
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8255860805511475,
+      "learning_rate": 1.0832021075747712e-07,
+      "loss": 0.1072,
+      "step": 29000
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7951275706291199,
+      "learning_rate": 1.047648497608414e-07,
+      "loss": 0.1029,
+      "step": 29050
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7167026996612549,
+      "learning_rate": 1.0126756596375687e-07,
+      "loss": 0.1039,
+      "step": 29100
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.75605708360672,
+      "learning_rate": 9.782844416476423e-08,
+      "loss": 0.1022,
+      "step": 29150
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.72404545545578,
+      "learning_rate": 9.444756775215446e-08,
+      "loss": 0.1044,
+      "step": 29200
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8219017386436462,
+      "learning_rate": 9.112501870194273e-08,
+      "loss": 0.1024,
+      "step": 29250
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7620697617530823,
+      "learning_rate": 8.786087757588269e-08,
+      "loss": 0.1039,
+      "step": 29300
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7205595970153809,
+      "learning_rate": 8.465522351951305e-08,
+      "loss": 0.1047,
+      "step": 29350
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7058899998664856,
+      "learning_rate": 8.150813426023752e-08,
+      "loss": 0.1031,
+      "step": 29400
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7188001871109009,
+      "learning_rate": 7.841968610544121e-08,
+      "loss": 0.1008,
+      "step": 29450
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.672663152217865,
+      "learning_rate": 7.538995394063996e-08,
+      "loss": 0.1022,
+      "step": 29500
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7694998979568481,
+      "learning_rate": 7.241901122766515e-08,
+      "loss": 0.1014,
+      "step": 29550
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6881873607635498,
+      "learning_rate": 6.950693000288056e-08,
+      "loss": 0.1003,
+      "step": 29600
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8163633942604065,
+      "learning_rate": 6.665378087543889e-08,
+      "loss": 0.1019,
+      "step": 29650
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6912686228752136,
+      "learning_rate": 6.385963302556642e-08,
+      "loss": 0.1024,
+      "step": 29700
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8492407202720642,
+      "learning_rate": 6.112455420288821e-08,
+      "loss": 0.1039,
+      "step": 29750
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.737253725528717,
+      "learning_rate": 5.844861072478336e-08,
+      "loss": 0.1051,
+      "step": 29800
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6771920919418335,
+      "learning_rate": 5.583186747477848e-08,
+      "loss": 0.0996,
+      "step": 29850
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6680823564529419,
+      "learning_rate": 5.32743879009745e-08,
+      "loss": 0.1027,
+      "step": 29900
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7818738222122192,
+      "learning_rate": 5.077623401450599e-08,
+      "loss": 0.1027,
+      "step": 29950
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7418414950370789,
+      "learning_rate": 4.8337466388040935e-08,
+      "loss": 0.1027,
+      "step": 30000
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6610372066497803,
+      "learning_rate": 4.595814415430916e-08,
+      "loss": 0.0983,
+      "step": 30050
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9139551520347595,
+      "learning_rate": 4.3638325004670134e-08,
+      "loss": 0.1096,
+      "step": 30100
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7789270877838135,
+      "learning_rate": 4.1378065187714365e-08,
+      "loss": 0.1013,
+      "step": 30150
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7473266124725342,
+      "learning_rate": 3.917741950789727e-08,
+      "loss": 0.1048,
+      "step": 30200
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7920830249786377,
+      "learning_rate": 3.703644132421386e-08,
+      "loss": 0.1003,
+      "step": 30250
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8591673970222473,
+      "learning_rate": 3.4955182548901956e-08,
+      "loss": 0.102,
+      "step": 30300
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8582797646522522,
+      "learning_rate": 3.293369364618465e-08,
+      "loss": 0.1014,
+      "step": 30350
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7193279266357422,
+      "learning_rate": 3.097202363104679e-08,
+      "loss": 0.1026,
+      "step": 30400
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8688130378723145,
+      "learning_rate": 2.9070220068045663e-08,
+      "loss": 0.1055,
+      "step": 30450
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7748174071311951,
+      "learning_rate": 2.722832907015971e-08,
+      "loss": 0.1026,
+      "step": 30500
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5935469269752502,
+      "learning_rate": 2.544639529766829e-08,
+      "loss": 0.1047,
+      "step": 30550
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6561381816864014,
+      "learning_rate": 2.3724461957068955e-08,
+      "loss": 0.1002,
+      "step": 30600
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7781614065170288,
+      "learning_rate": 2.206257080003188e-08,
+      "loss": 0.1064,
+      "step": 30650
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7374931573867798,
+      "learning_rate": 2.0460762122385124e-08,
+      "loss": 0.096,
+      "step": 30700
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6997425556182861,
+      "learning_rate": 1.8919074763138757e-08,
+      "loss": 0.101,
+      "step": 30750
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8059677481651306,
+      "learning_rate": 1.7437546103542814e-08,
+      "loss": 0.1011,
+      "step": 30800
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7518121600151062,
+      "learning_rate": 1.6016212066181368e-08,
+      "loss": 0.0995,
+      "step": 30850
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7425240278244019,
+      "learning_rate": 1.4655107114101008e-08,
+      "loss": 0.1071,
+      "step": 30900
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7220152020454407,
+      "learning_rate": 1.3354264249975379e-08,
+      "loss": 0.1011,
+      "step": 30950
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7005524635314941,
+      "learning_rate": 1.2113715015304728e-08,
+      "loss": 0.1035,
+      "step": 31000
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6991942524909973,
+      "learning_rate": 1.0933489489651783e-08,
+      "loss": 0.1073,
+      "step": 31050
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7447285652160645,
+      "learning_rate": 9.81361628991151e-09,
+      "loss": 0.1046,
+      "step": 31100
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.734191358089447,
+      "learning_rate": 8.754122569618329e-09,
+      "loss": 0.1,
+      "step": 31150
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7444181442260742,
+      "learning_rate": 7.755034018286644e-09,
+      "loss": 0.1079,
+      "step": 31200
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.723635196685791,
+      "learning_rate": 6.816374860788566e-09,
+      "loss": 0.1055,
+      "step": 31250
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6592646837234497,
+      "learning_rate": 5.938167856766319e-09,
+      "loss": 0.0998,
+      "step": 31300
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8482509255409241,
+      "learning_rate": 5.120434300080745e-09,
+      "loss": 0.1018,
+      "step": 31350
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7359779477119446,
+      "learning_rate": 4.363194018293937e-09,
+      "loss": 0.1013,
+      "step": 31400
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.704291820526123,
+      "learning_rate": 3.666465372190453e-09,
+      "loss": 0.1016,
+      "step": 31450
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6474089622497559,
+      "learning_rate": 3.030265255329623e-09,
+      "loss": 0.1018,
+      "step": 31500
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.75139319896698,
+      "learning_rate": 2.4546090936383717e-09,
+      "loss": 0.1078,
+      "step": 31550
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8348722457885742,
+      "learning_rate": 1.9395108450351308e-09,
+      "loss": 0.1052,
+      "step": 31600
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6401920914649963,
+      "learning_rate": 1.4849829990931653e-09,
+      "loss": 0.102,
+      "step": 31650
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6777997612953186,
+      "learning_rate": 1.0910365767358155e-09,
+      "loss": 0.1032,
+      "step": 31700
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.808435320854187,
+      "learning_rate": 7.576811299714326e-10,
+      "loss": 0.1034,
+      "step": 31750
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7167596817016602,
+      "learning_rate": 4.849247416599534e-10,
+      "loss": 0.1047,
+      "step": 31800
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7499010562896729,
+      "learning_rate": 2.727740253177791e-10,
+      "loss": 0.1045,
+      "step": 31850
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7393851280212402,
+      "learning_rate": 1.2123412495762543e-10,
+      "loss": 0.1063,
+      "step": 31900
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7504858374595642,
+      "learning_rate": 3.0308714963067644e-11,
+      "loss": 0.1045,
+      "step": 31950
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8671845197677612,
+      "learning_rate": 0.0,
+      "loss": 0.1015,
+      "step": 32000
+    }
+  ],
+  "logging_steps": 50,
+  "max_steps": 32000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 3200,
+  "total_flos": 2.174443399415808e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}