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": 60.0,
+  "best_model_checkpoint": "/data/users/bking2/tod_zero/outputs/runs/finetune/starcoder_3b/supervised/eiqracdg/checkpoint-28800",
+  "epoch": 0.8109888995894369,
+  "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": 1.92511785030365,
+      "learning_rate": 2.5e-06,
+      "loss": 1.186,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4528826475143433,
+      "learning_rate": 5e-06,
+      "loss": 0.287,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.285142421722412,
+      "learning_rate": 4.9999696912850374e-06,
+      "loss": 0.2339,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2731701135635376,
+      "learning_rate": 4.999878765875043e-06,
+      "loss": 0.2247,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5167465209960938,
+      "learning_rate": 4.999727225974682e-06,
+      "loss": 0.2029,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0947519540786743,
+      "learning_rate": 4.999515075258341e-06,
+      "loss": 0.2034,
+      "step": 300
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1314687728881836,
+      "learning_rate": 4.999242318870029e-06,
+      "loss": 0.1977,
+      "step": 350
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.172371506690979,
+      "learning_rate": 4.998908963423264e-06,
+      "loss": 0.1924,
+      "step": 400
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.348120093345642,
+      "learning_rate": 4.998515017000907e-06,
+      "loss": 0.189,
+      "step": 450
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0465130805969238,
+      "learning_rate": 4.998060489154965e-06,
+      "loss": 0.1863,
+      "step": 500
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0949101448059082,
+      "learning_rate": 4.997545390906362e-06,
+      "loss": 0.1894,
+      "step": 550
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2342430353164673,
+      "learning_rate": 4.996969734744671e-06,
+      "loss": 0.1842,
+      "step": 600
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2161908149719238,
+      "learning_rate": 4.99633353462781e-06,
+      "loss": 0.1885,
+      "step": 650
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.237592101097107,
+      "learning_rate": 4.995636805981707e-06,
+      "loss": 0.188,
+      "step": 700
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0181180238723755,
+      "learning_rate": 4.99487956569992e-06,
+      "loss": 0.1807,
+      "step": 750
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0869165658950806,
+      "learning_rate": 4.994061832143235e-06,
+      "loss": 0.1859,
+      "step": 800
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.115837574005127,
+      "learning_rate": 4.993183625139212e-06,
+      "loss": 0.1733,
+      "step": 850
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9784592986106873,
+      "learning_rate": 4.992244965981714e-06,
+      "loss": 0.1781,
+      "step": 900
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1208012104034424,
+      "learning_rate": 4.991245877430382e-06,
+      "loss": 0.1793,
+      "step": 950
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1315056085586548,
+      "learning_rate": 4.990186383710089e-06,
+      "loss": 0.1737,
+      "step": 1000
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9561248421669006,
+      "learning_rate": 4.9890665105103484e-06,
+      "loss": 0.1755,
+      "step": 1050
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0493712425231934,
+      "learning_rate": 4.987886284984695e-06,
+      "loss": 0.1733,
+      "step": 1100
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0467311143875122,
+      "learning_rate": 4.986645735750025e-06,
+      "loss": 0.1742,
+      "step": 1150
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.915791928768158,
+      "learning_rate": 4.985344892885899e-06,
+      "loss": 0.1722,
+      "step": 1200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0208431482315063,
+      "learning_rate": 4.98398378793382e-06,
+      "loss": 0.1722,
+      "step": 1250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0767868757247925,
+      "learning_rate": 4.982562453896458e-06,
+      "loss": 0.1726,
+      "step": 1300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9579061269760132,
+      "learning_rate": 4.9810809252368615e-06,
+      "loss": 0.1688,
+      "step": 1350
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9485280513763428,
+      "learning_rate": 4.979539237877615e-06,
+      "loss": 0.1614,
+      "step": 1400
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9779006838798523,
+      "learning_rate": 4.977937429199968e-06,
+      "loss": 0.1694,
+      "step": 1450
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9930433034896851,
+      "learning_rate": 4.976275538042932e-06,
+      "loss": 0.1694,
+      "step": 1500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1304377317428589,
+      "learning_rate": 4.974553604702332e-06,
+      "loss": 0.1697,
+      "step": 1550
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9419906139373779,
+      "learning_rate": 4.972771670929841e-06,
+      "loss": 0.1678,
+      "step": 1600
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9315632581710815,
+      "learning_rate": 4.970929779931955e-06,
+      "loss": 0.1735,
+      "step": 1650
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.015580415725708,
+      "learning_rate": 4.969027976368954e-06,
+      "loss": 0.1706,
+      "step": 1700
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.059506893157959,
+      "learning_rate": 4.967066306353816e-06,
+      "loss": 0.1729,
+      "step": 1750
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8472514152526855,
+      "learning_rate": 4.9650448174510986e-06,
+      "loss": 0.1718,
+      "step": 1800
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6632156372070312,
+      "learning_rate": 4.9629635586757865e-06,
+      "loss": 0.1667,
+      "step": 1850
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0739672183990479,
+      "learning_rate": 4.960822580492103e-06,
+      "loss": 0.1654,
+      "step": 1900
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9874502420425415,
+      "learning_rate": 4.958621934812286e-06,
+      "loss": 0.1641,
+      "step": 1950
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9233792424201965,
+      "learning_rate": 4.95636167499533e-06,
+      "loss": 0.1648,
+      "step": 2000
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8490021228790283,
+      "learning_rate": 4.9540418558456915e-06,
+      "loss": 0.1688,
+      "step": 2050
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9746788144111633,
+      "learning_rate": 4.951662533611959e-06,
+      "loss": 0.169,
+      "step": 2100
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8029181957244873,
+      "learning_rate": 4.9492237659854946e-06,
+      "loss": 0.1645,
+      "step": 2150
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1753889322280884,
+      "learning_rate": 4.9467256120990255e-06,
+      "loss": 0.1692,
+      "step": 2200
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8767470717430115,
+      "learning_rate": 4.9441681325252215e-06,
+      "loss": 0.1617,
+      "step": 2250
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9515625238418579,
+      "learning_rate": 4.941551389275217e-06,
+      "loss": 0.1599,
+      "step": 2300
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9250404238700867,
+      "learning_rate": 4.938875445797112e-06,
+      "loss": 0.1678,
+      "step": 2350
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9468141794204712,
+      "learning_rate": 4.936140366974434e-06,
+      "loss": 0.1665,
+      "step": 2400
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9202454090118408,
+      "learning_rate": 4.933346219124562e-06,
+      "loss": 0.1579,
+      "step": 2450
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9553722739219666,
+      "learning_rate": 4.93049306999712e-06,
+      "loss": 0.1706,
+      "step": 2500
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8615751266479492,
+      "learning_rate": 4.927580988772336e-06,
+      "loss": 0.168,
+      "step": 2550
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9888120293617249,
+      "learning_rate": 4.9246100460593606e-06,
+      "loss": 0.1688,
+      "step": 2600
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.915663480758667,
+      "learning_rate": 4.92158031389456e-06,
+      "loss": 0.1615,
+      "step": 2650
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9765244722366333,
+      "learning_rate": 4.918491865739763e-06,
+      "loss": 0.1571,
+      "step": 2700
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.847814679145813,
+      "learning_rate": 4.915344776480487e-06,
+      "loss": 0.1602,
+      "step": 2750
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9848460555076599,
+      "learning_rate": 4.912139122424118e-06,
+      "loss": 0.1634,
+      "step": 2800
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8652825355529785,
+      "learning_rate": 4.908874981298058e-06,
+      "loss": 0.1594,
+      "step": 2850
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9553362131118774,
+      "learning_rate": 4.9055524322478456e-06,
+      "loss": 0.1681,
+      "step": 2900
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8901177644729614,
+      "learning_rate": 4.902171555835236e-06,
+      "loss": 0.1562,
+      "step": 2950
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.047637701034546,
+      "learning_rate": 4.8987324340362445e-06,
+      "loss": 0.1601,
+      "step": 3000
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8235883116722107,
+      "learning_rate": 4.895235150239159e-06,
+      "loss": 0.1666,
+      "step": 3050
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1097720861434937,
+      "learning_rate": 4.891679789242524e-06,
+      "loss": 0.1584,
+      "step": 3100
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8984447121620178,
+      "learning_rate": 4.8880664372530765e-06,
+      "loss": 0.1612,
+      "step": 3150
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9215020537376404,
+      "learning_rate": 4.884395181883661e-06,
+      "loss": 0.1606,
+      "step": 3200
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8073440790176392,
+      "learning_rate": 4.880666112151104e-06,
+      "loss": 0.1624,
+      "step": 3250
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9907032251358032,
+      "learning_rate": 4.876879318474056e-06,
+      "loss": 0.1526,
+      "step": 3300
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0674902200698853,
+      "learning_rate": 4.873034892670795e-06,
+      "loss": 0.1628,
+      "step": 3350
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.886870801448822,
+      "learning_rate": 4.869132927957007e-06,
+      "loss": 0.1597,
+      "step": 3400
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8470545411109924,
+      "learning_rate": 4.8651735189435205e-06,
+      "loss": 0.155,
+      "step": 3450
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8206085562705994,
+      "learning_rate": 4.861156761634014e-06,
+      "loss": 0.1607,
+      "step": 3500
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9522375464439392,
+      "learning_rate": 4.857082753422691e-06,
+      "loss": 0.1622,
+      "step": 3550
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9393265843391418,
+      "learning_rate": 4.852951593091914e-06,
+      "loss": 0.1574,
+      "step": 3600
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.89833003282547,
+      "learning_rate": 4.848763380809811e-06,
+      "loss": 0.1627,
+      "step": 3650
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8664096593856812,
+      "learning_rate": 4.844518218127849e-06,
+      "loss": 0.1569,
+      "step": 3700
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8381832838058472,
+      "learning_rate": 4.840216207978368e-06,
+      "loss": 0.1585,
+      "step": 3750
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7903485298156738,
+      "learning_rate": 4.835857454672087e-06,
+      "loss": 0.1591,
+      "step": 3800
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9844456315040588,
+      "learning_rate": 4.831442063895575e-06,
+      "loss": 0.1539,
+      "step": 3850
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8709468245506287,
+      "learning_rate": 4.8269701427086905e-06,
+      "loss": 0.1588,
+      "step": 3900
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9181196093559265,
+      "learning_rate": 4.822441799541979e-06,
+      "loss": 0.1569,
+      "step": 3950
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.868691623210907,
+      "learning_rate": 4.8178571441940515e-06,
+      "loss": 0.1554,
+      "step": 4000
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9550264477729797,
+      "learning_rate": 4.813216287828917e-06,
+      "loss": 0.1595,
+      "step": 4050
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9788670539855957,
+      "learning_rate": 4.808519342973289e-06,
+      "loss": 0.158,
+      "step": 4100
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9521192908287048,
+      "learning_rate": 4.80376642351386e-06,
+      "loss": 0.1541,
+      "step": 4150
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9816685318946838,
+      "learning_rate": 4.798957644694533e-06,
+      "loss": 0.1555,
+      "step": 4200
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8248597383499146,
+      "learning_rate": 4.794093123113635e-06,
+      "loss": 0.1575,
+      "step": 4250
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8728579878807068,
+      "learning_rate": 4.789172976721082e-06,
+      "loss": 0.1582,
+      "step": 4300
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8118800520896912,
+      "learning_rate": 4.7841973248155275e-06,
+      "loss": 0.1606,
+      "step": 4350
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8186320066452026,
+      "learning_rate": 4.779166288041463e-06,
+      "loss": 0.1515,
+      "step": 4400
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8599525690078735,
+      "learning_rate": 4.7740799883862966e-06,
+      "loss": 0.1604,
+      "step": 4450
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9356303811073303,
+      "learning_rate": 4.7689385491773934e-06,
+      "loss": 0.155,
+      "step": 4500
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9057179689407349,
+      "learning_rate": 4.7637420950790855e-06,
+      "loss": 0.1561,
+      "step": 4550
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.953428328037262,
+      "learning_rate": 4.75849075208965e-06,
+      "loss": 0.1624,
+      "step": 4600
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8495259284973145,
+      "learning_rate": 4.7531846475382526e-06,
+      "loss": 0.1583,
+      "step": 4650
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8086969256401062,
+      "learning_rate": 4.7478239100818626e-06,
+      "loss": 0.1514,
+      "step": 4700
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8309988975524902,
+      "learning_rate": 4.742408669702131e-06,
+      "loss": 0.1586,
+      "step": 4750
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8857250809669495,
+      "learning_rate": 4.736939057702239e-06,
+      "loss": 0.1534,
+      "step": 4800
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9398581981658936,
+      "learning_rate": 4.731415206703714e-06,
+      "loss": 0.1522,
+      "step": 4850
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8459595441818237,
+      "learning_rate": 4.725837250643218e-06,
+      "loss": 0.1488,
+      "step": 4900
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9744529724121094,
+      "learning_rate": 4.720205324769296e-06,
+      "loss": 0.153,
+      "step": 4950
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8761635422706604,
+      "learning_rate": 4.714519565639095e-06,
+      "loss": 0.1558,
+      "step": 5000
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7558119297027588,
+      "learning_rate": 4.708780111115058e-06,
+      "loss": 0.1563,
+      "step": 5050
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9466744661331177,
+      "learning_rate": 4.702987100361578e-06,
+      "loss": 0.1578,
+      "step": 5100
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9422277808189392,
+      "learning_rate": 4.697140673841624e-06,
+      "loss": 0.147,
+      "step": 5150
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7488685846328735,
+      "learning_rate": 4.6912409733133365e-06,
+      "loss": 0.1535,
+      "step": 5200
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8875038027763367,
+      "learning_rate": 4.685288141826589e-06,
+      "loss": 0.1491,
+      "step": 5250
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1153545379638672,
+      "learning_rate": 4.679282323719519e-06,
+      "loss": 0.1523,
+      "step": 5300
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9965418577194214,
+      "learning_rate": 4.67322366461503e-06,
+      "loss": 0.1499,
+      "step": 5350
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7546530961990356,
+      "learning_rate": 4.66711231141726e-06,
+      "loss": 0.1517,
+      "step": 5400
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8970305323600769,
+      "learning_rate": 4.660948412308018e-06,
+      "loss": 0.1546,
+      "step": 5450
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0205148458480835,
+      "learning_rate": 4.654732116743193e-06,
+      "loss": 0.1532,
+      "step": 5500
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8177210092544556,
+      "learning_rate": 4.64846357544913e-06,
+      "loss": 0.1437,
+      "step": 5550
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0768853425979614,
+      "learning_rate": 4.642142940418973e-06,
+      "loss": 0.1499,
+      "step": 5600
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8934694528579712,
+      "learning_rate": 4.635770364908984e-06,
+      "loss": 0.1538,
+      "step": 5650
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9092695713043213,
+      "learning_rate": 4.629346003434822e-06,
+      "loss": 0.1507,
+      "step": 5700
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8804992437362671,
+      "learning_rate": 4.622870011767798e-06,
+      "loss": 0.1494,
+      "step": 5750
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7924209237098694,
+      "learning_rate": 4.616342546931103e-06,
+      "loss": 0.155,
+      "step": 5800
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8898280262947083,
+      "learning_rate": 4.609763767195991e-06,
+      "loss": 0.1511,
+      "step": 5850
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8630085587501526,
+      "learning_rate": 4.603133832077953e-06,
+      "loss": 0.1486,
+      "step": 5900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7191097140312195,
+      "learning_rate": 4.596452902332839e-06,
+      "loss": 0.1497,
+      "step": 5950
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.815141499042511,
+      "learning_rate": 4.589721139952964e-06,
+      "loss": 0.1503,
+      "step": 6000
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8735461831092834,
+      "learning_rate": 4.582938708163183e-06,
+      "loss": 0.1532,
+      "step": 6050
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.851352334022522,
+      "learning_rate": 4.576105771416928e-06,
+      "loss": 0.1527,
+      "step": 6100
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7596533298492432,
+      "learning_rate": 4.569222495392227e-06,
+      "loss": 0.153,
+      "step": 6150
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8966355919837952,
+      "learning_rate": 4.562289046987679e-06,
+      "loss": 0.1493,
+      "step": 6200
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8785326480865479,
+      "learning_rate": 4.555305594318414e-06,
+      "loss": 0.1458,
+      "step": 6250
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8890253305435181,
+      "learning_rate": 4.548272306712013e-06,
+      "loss": 0.1495,
+      "step": 6300
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8389493823051453,
+      "learning_rate": 4.541189354704403e-06,
+      "loss": 0.1554,
+      "step": 6350
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8545738458633423,
+      "learning_rate": 4.534056910035724e-06,
+      "loss": 0.1512,
+      "step": 6400
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.859368085861206,
+      "learning_rate": 4.5268751456461605e-06,
+      "loss": 0.148,
+      "step": 6450
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9073702692985535,
+      "learning_rate": 4.5196442356717526e-06,
+      "loss": 0.1522,
+      "step": 6500
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8107109069824219,
+      "learning_rate": 4.512364355440172e-06,
+      "loss": 0.1506,
+      "step": 6550
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6515960693359375,
+      "learning_rate": 4.505035681466472e-06,
+      "loss": 0.1459,
+      "step": 6600
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8400324583053589,
+      "learning_rate": 4.497658391448803e-06,
+      "loss": 0.1486,
+      "step": 6650
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8535892367362976,
+      "learning_rate": 4.49023266426411e-06,
+      "loss": 0.1533,
+      "step": 6700
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8867340087890625,
+      "learning_rate": 4.482758679963792e-06,
+      "loss": 0.1546,
+      "step": 6750
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8694542050361633,
+      "learning_rate": 4.475236619769336e-06,
+      "loss": 0.1462,
+      "step": 6800
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8599735498428345,
+      "learning_rate": 4.4676666660679265e-06,
+      "loss": 0.1472,
+      "step": 6850
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8738076686859131,
+      "learning_rate": 4.460049002408018e-06,
+      "loss": 0.1521,
+      "step": 6900
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9029892683029175,
+      "learning_rate": 4.452383813494887e-06,
+      "loss": 0.1534,
+      "step": 6950
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.828575849533081,
+      "learning_rate": 4.444671285186155e-06,
+      "loss": 0.1487,
+      "step": 7000
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7742441892623901,
+      "learning_rate": 4.4369116044872786e-06,
+      "loss": 0.1449,
+      "step": 7050
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8644819855690002,
+      "learning_rate": 4.42910495954702e-06,
+      "loss": 0.1532,
+      "step": 7100
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.955912172794342,
+      "learning_rate": 4.421251539652879e-06,
+      "loss": 0.1479,
+      "step": 7150
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8100504875183105,
+      "learning_rate": 4.413351535226507e-06,
+      "loss": 0.1515,
+      "step": 7200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8034056425094604,
+      "learning_rate": 4.4054051378190915e-06,
+      "loss": 0.1509,
+      "step": 7250
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0016452074050903,
+      "learning_rate": 4.397412540106707e-06,
+      "loss": 0.1477,
+      "step": 7300
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7266576290130615,
+      "learning_rate": 4.3893739358856465e-06,
+      "loss": 0.1382,
+      "step": 7350
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8133652210235596,
+      "learning_rate": 4.38128952006772e-06,
+      "loss": 0.148,
+      "step": 7400
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8336594104766846,
+      "learning_rate": 4.373159488675533e-06,
+      "loss": 0.1487,
+      "step": 7450
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7953880429267883,
+      "learning_rate": 4.364984038837727e-06,
+      "loss": 0.1471,
+      "step": 7500
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8772010803222656,
+      "learning_rate": 4.356763368784207e-06,
+      "loss": 0.1509,
+      "step": 7550
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8125504851341248,
+      "learning_rate": 4.348497677841328e-06,
+      "loss": 0.1477,
+      "step": 7600
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8239701390266418,
+      "learning_rate": 4.340187166427067e-06,
+      "loss": 0.1439,
+      "step": 7650
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8961694240570068,
+      "learning_rate": 4.331832036046162e-06,
+      "loss": 0.1484,
+      "step": 7700
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8788717985153198,
+      "learning_rate": 4.323432489285223e-06,
+      "loss": 0.1459,
+      "step": 7750
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8486537337303162,
+      "learning_rate": 4.3149887298078275e-06,
+      "loss": 0.1457,
+      "step": 7800
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.914941132068634,
+      "learning_rate": 4.306500962349573e-06,
+      "loss": 0.144,
+      "step": 7850
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8333587050437927,
+      "learning_rate": 4.2979693927131205e-06,
+      "loss": 0.1434,
+      "step": 7900
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8071913719177246,
+      "learning_rate": 4.289394227763199e-06,
+      "loss": 0.1498,
+      "step": 7950
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7681615948677063,
+      "learning_rate": 4.2807756754215926e-06,
+      "loss": 0.1461,
+      "step": 8000
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8993908762931824,
+      "learning_rate": 4.272113944662099e-06,
+      "loss": 0.146,
+      "step": 8050
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8620333671569824,
+      "learning_rate": 4.263409245505461e-06,
+      "loss": 0.1444,
+      "step": 8100
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8208089470863342,
+      "learning_rate": 4.254661789014274e-06,
+      "loss": 0.1415,
+      "step": 8150
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7570805549621582,
+      "learning_rate": 4.2458717872878715e-06,
+      "loss": 0.1406,
+      "step": 8200
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8517532348632812,
+      "learning_rate": 4.237039453457179e-06,
+      "loss": 0.1439,
+      "step": 8250
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7504327893257141,
+      "learning_rate": 4.228165001679547e-06,
+      "loss": 0.1408,
+      "step": 8300
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7522333264350891,
+      "learning_rate": 4.219248647133559e-06,
+      "loss": 0.1434,
+      "step": 8350
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9496443271636963,
+      "learning_rate": 4.210290606013813e-06,
+      "loss": 0.1472,
+      "step": 8400
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0091819763183594,
+      "learning_rate": 4.2012910955256825e-06,
+      "loss": 0.1502,
+      "step": 8450
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9098952412605286,
+      "learning_rate": 4.192250333880045e-06,
+      "loss": 0.1488,
+      "step": 8500
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6374802589416504,
+      "learning_rate": 4.183168540287995e-06,
+      "loss": 0.1443,
+      "step": 8550
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7613967657089233,
+      "learning_rate": 4.174045934955527e-06,
+      "loss": 0.149,
+      "step": 8600
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8770780563354492,
+      "learning_rate": 4.164882739078197e-06,
+      "loss": 0.1432,
+      "step": 8650
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8647637963294983,
+      "learning_rate": 4.155679174835758e-06,
+      "loss": 0.1453,
+      "step": 8700
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8997677564620972,
+      "learning_rate": 4.146435465386776e-06,
+      "loss": 0.1429,
+      "step": 8750
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.702021598815918,
+      "learning_rate": 4.137151834863213e-06,
+      "loss": 0.1448,
+      "step": 8800
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8877437710762024,
+      "learning_rate": 4.1278285083649985e-06,
+      "loss": 0.1449,
+      "step": 8850
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8944686055183411,
+      "learning_rate": 4.11846571195457e-06,
+      "loss": 0.1522,
+      "step": 8900
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9450408816337585,
+      "learning_rate": 4.1090636726513875e-06,
+      "loss": 0.1453,
+      "step": 8950
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7170203328132629,
+      "learning_rate": 4.0996226184264355e-06,
+      "loss": 0.1469,
+      "step": 9000
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8969573974609375,
+      "learning_rate": 4.090142778196692e-06,
+      "loss": 0.1494,
+      "step": 9050
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9128451347351074,
+      "learning_rate": 4.080624381819577e-06,
+      "loss": 0.1413,
+      "step": 9100
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8542819023132324,
+      "learning_rate": 4.071067660087379e-06,
+      "loss": 0.1418,
+      "step": 9150
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8860799670219421,
+      "learning_rate": 4.061472844721664e-06,
+      "loss": 0.1408,
+      "step": 9200
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8985223770141602,
+      "learning_rate": 4.05184016836765e-06,
+      "loss": 0.1444,
+      "step": 9250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7234982252120972,
+      "learning_rate": 4.042169864588571e-06,
+      "loss": 0.1469,
+      "step": 9300
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7840049266815186,
+      "learning_rate": 4.032462167860012e-06,
+      "loss": 0.1367,
+      "step": 9350
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8657572269439697,
+      "learning_rate": 4.022717313564223e-06,
+      "loss": 0.1456,
+      "step": 9400
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8249016404151917,
+      "learning_rate": 4.012935537984414e-06,
+      "loss": 0.1398,
+      "step": 9450
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9602368474006653,
+      "learning_rate": 4.0031170782990214e-06,
+      "loss": 0.1465,
+      "step": 9500
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7444416284561157,
+      "learning_rate": 3.993262172575962e-06,
+      "loss": 0.1415,
+      "step": 9550
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8414493799209595,
+      "learning_rate": 3.983371059766862e-06,
+      "loss": 0.1487,
+      "step": 9600
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8294157385826111,
+      "learning_rate": 3.973443979701252e-06,
+      "loss": 0.1406,
+      "step": 9650
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8009164333343506,
+      "learning_rate": 3.963481173080768e-06,
+      "loss": 0.1445,
+      "step": 9700
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8331242203712463,
+      "learning_rate": 3.9534828814733e-06,
+      "loss": 0.1388,
+      "step": 9750
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8543001413345337,
+      "learning_rate": 3.943449347307146e-06,
+      "loss": 0.1399,
+      "step": 9800
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.625985324382782,
+      "learning_rate": 3.9333808138651265e-06,
+      "loss": 0.1376,
+      "step": 9850
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7723961472511292,
+      "learning_rate": 3.923277525278691e-06,
+      "loss": 0.1443,
+      "step": 9900
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8138018846511841,
+      "learning_rate": 3.913139726521993e-06,
+      "loss": 0.1358,
+      "step": 9950
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7179055213928223,
+      "learning_rate": 3.9029676634059565e-06,
+      "loss": 0.1376,
+      "step": 10000
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7832974195480347,
+      "learning_rate": 3.89276158257231e-06,
+      "loss": 0.1391,
+      "step": 10050
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6319305300712585,
+      "learning_rate": 3.882521731487609e-06,
+      "loss": 0.1394,
+      "step": 10100
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8627734184265137,
+      "learning_rate": 3.872248358437236e-06,
+      "loss": 0.1437,
+      "step": 10150
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8356693387031555,
+      "learning_rate": 3.861941712519379e-06,
+      "loss": 0.1395,
+      "step": 10200
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9000275135040283,
+      "learning_rate": 3.8516020436389945e-06,
+      "loss": 0.1415,
+      "step": 10250
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8124093413352966,
+      "learning_rate": 3.841229602501742e-06,
+      "loss": 0.1425,
+      "step": 10300
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7685266137123108,
+      "learning_rate": 3.8308246406079116e-06,
+      "loss": 0.1444,
+      "step": 10350
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.858231782913208,
+      "learning_rate": 3.820387410246324e-06,
+      "loss": 0.1408,
+      "step": 10400
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7910386323928833,
+      "learning_rate": 3.809918164488208e-06,
+      "loss": 0.1404,
+      "step": 10450
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7267195582389832,
+      "learning_rate": 3.7994171571810756e-06,
+      "loss": 0.1444,
+      "step": 10500
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9443105459213257,
+      "learning_rate": 3.788884642942555e-06,
+      "loss": 0.1427,
+      "step": 10550
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8015535473823547,
+      "learning_rate": 3.7783208771542237e-06,
+      "loss": 0.1386,
+      "step": 10600
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8959324955940247,
+      "learning_rate": 3.7677261159554145e-06,
+      "loss": 0.1428,
+      "step": 10650
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8010550737380981,
+      "learning_rate": 3.757100616237006e-06,
+      "loss": 0.1407,
+      "step": 10700
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7638524174690247,
+      "learning_rate": 3.746444635635191e-06,
+      "loss": 0.1413,
+      "step": 10750
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8806131482124329,
+      "learning_rate": 3.735758432525234e-06,
+      "loss": 0.1392,
+      "step": 10800
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.763507068157196,
+      "learning_rate": 3.725042266015201e-06,
+      "loss": 0.1414,
+      "step": 10850
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7763858437538147,
+      "learning_rate": 3.7142963959396805e-06,
+      "loss": 0.142,
+      "step": 10900
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8945651054382324,
+      "learning_rate": 3.7035210828534846e-06,
+      "loss": 0.1444,
+      "step": 10950
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8169758915901184,
+      "learning_rate": 3.692716588025327e-06,
+      "loss": 0.1392,
+      "step": 11000
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8331303596496582,
+      "learning_rate": 3.68188317343149e-06,
+      "loss": 0.1408,
+      "step": 11050
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7105779051780701,
+      "learning_rate": 3.671021101749476e-06,
+      "loss": 0.1369,
+      "step": 11100
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8562634587287903,
+      "learning_rate": 3.6601306363516297e-06,
+      "loss": 0.1476,
+      "step": 11150
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.831794798374176,
+      "learning_rate": 3.649212041298763e-06,
+      "loss": 0.141,
+      "step": 11200
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7753794193267822,
+      "learning_rate": 3.638265581333742e-06,
+      "loss": 0.1396,
+      "step": 11250
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8438231348991394,
+      "learning_rate": 3.627291521875076e-06,
+      "loss": 0.1411,
+      "step": 11300
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8898864388465881,
+      "learning_rate": 3.616290129010476e-06,
+      "loss": 0.1363,
+      "step": 11350
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8584467768669128,
+      "learning_rate": 3.605261669490407e-06,
+      "loss": 0.1437,
+      "step": 11400
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8061217665672302,
+      "learning_rate": 3.5942064107216183e-06,
+      "loss": 0.1408,
+      "step": 11450
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7721576690673828,
+      "learning_rate": 3.5831246207606597e-06,
+      "loss": 0.14,
+      "step": 11500
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7443546056747437,
+      "learning_rate": 3.57201656830738e-06,
+      "loss": 0.1416,
+      "step": 11550
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8907784819602966,
+      "learning_rate": 3.5608825226984168e-06,
+      "loss": 0.1409,
+      "step": 11600
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8339740633964539,
+      "learning_rate": 3.549722753900662e-06,
+      "loss": 0.1403,
+      "step": 11650
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.844826340675354,
+      "learning_rate": 3.5385375325047167e-06,
+      "loss": 0.135,
+      "step": 11700
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.875929057598114,
+      "learning_rate": 3.5273271297183302e-06,
+      "loss": 0.1404,
+      "step": 11750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7786159515380859,
+      "learning_rate": 3.516091817359825e-06,
+      "loss": 0.1372,
+      "step": 11800
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9039524793624878,
+      "learning_rate": 3.5048318678515052e-06,
+      "loss": 0.1383,
+      "step": 11850
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8629227876663208,
+      "learning_rate": 3.493547554213051e-06,
+      "loss": 0.1365,
+      "step": 11900
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9458798170089722,
+      "learning_rate": 3.482239150054898e-06,
+      "loss": 0.1377,
+      "step": 11950
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7950788736343384,
+      "learning_rate": 3.470906929571605e-06,
+      "loss": 0.1415,
+      "step": 12000
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9919633865356445,
+      "learning_rate": 3.459551167535205e-06,
+      "loss": 0.1375,
+      "step": 12050
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8692936301231384,
+      "learning_rate": 3.4481721392885415e-06,
+      "loss": 0.1406,
+      "step": 12100
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7293336391448975,
+      "learning_rate": 3.4367701207385944e-06,
+      "loss": 0.1379,
+      "step": 12150
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9827872514724731,
+      "learning_rate": 3.425345388349787e-06,
+      "loss": 0.1406,
+      "step": 12200
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8083829283714294,
+      "learning_rate": 3.4138982191372838e-06,
+      "loss": 0.1451,
+      "step": 12250
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8368284106254578,
+      "learning_rate": 3.402428890660279e-06,
+      "loss": 0.1371,
+      "step": 12300
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7933076024055481,
+      "learning_rate": 3.390937681015256e-06,
+      "loss": 0.136,
+      "step": 12350
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7652096748352051,
+      "learning_rate": 3.379424868829254e-06,
+      "loss": 0.1349,
+      "step": 12400
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9007067084312439,
+      "learning_rate": 3.367890733253108e-06,
+      "loss": 0.1457,
+      "step": 12450
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.744931161403656,
+      "learning_rate": 3.3563355539546795e-06,
+      "loss": 0.1413,
+      "step": 12500
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8155565857887268,
+      "learning_rate": 3.3447596111120767e-06,
+      "loss": 0.1332,
+      "step": 12550
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9961397647857666,
+      "learning_rate": 3.333163185406861e-06,
+      "loss": 0.1396,
+      "step": 12600
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7819132208824158,
+      "learning_rate": 3.321546558017243e-06,
+      "loss": 0.1355,
+      "step": 12650
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7664303779602051,
+      "learning_rate": 3.309910010611259e-06,
+      "loss": 0.1379,
+      "step": 12700
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7041974067687988,
+      "learning_rate": 3.29825382533995e-06,
+      "loss": 0.1365,
+      "step": 12750
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8403826355934143,
+      "learning_rate": 3.286578284830513e-06,
+      "loss": 0.1435,
+      "step": 12800
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7580990791320801,
+      "learning_rate": 3.2748836721794514e-06,
+      "loss": 0.1364,
+      "step": 12850
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7609158158302307,
+      "learning_rate": 3.263170270945709e-06,
+      "loss": 0.1374,
+      "step": 12900
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8184811472892761,
+      "learning_rate": 3.2514383651437987e-06,
+      "loss": 0.1396,
+      "step": 12950
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.948807954788208,
+      "learning_rate": 3.239688239236911e-06,
+      "loss": 0.1339,
+      "step": 13000
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8595229983329773,
+      "learning_rate": 3.2279201781300206e-06,
+      "loss": 0.1374,
+      "step": 13050
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8162684440612793,
+      "learning_rate": 3.2161344671629736e-06,
+      "loss": 0.135,
+      "step": 13100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7258612513542175,
+      "learning_rate": 3.2043313921035747e-06,
+      "loss": 0.1376,
+      "step": 13150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8645867705345154,
+      "learning_rate": 3.1925112391406534e-06,
+      "loss": 0.142,
+      "step": 13200
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7972207069396973,
+      "learning_rate": 3.1806742948771276e-06,
+      "loss": 0.1351,
+      "step": 13250
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9695426225662231,
+      "learning_rate": 3.168820846323053e-06,
+      "loss": 0.1403,
+      "step": 13300
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6566038727760315,
+      "learning_rate": 3.1569511808886633e-06,
+      "loss": 0.1342,
+      "step": 13350
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8785290122032166,
+      "learning_rate": 3.1450655863774053e-06,
+      "loss": 0.1391,
+      "step": 13400
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7746759653091431,
+      "learning_rate": 3.1331643509789553e-06,
+      "loss": 0.1385,
+      "step": 13450
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7730455994606018,
+      "learning_rate": 3.121247763262235e-06,
+      "loss": 0.1336,
+      "step": 13500
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.864236056804657,
+      "learning_rate": 3.1093161121684118e-06,
+      "loss": 0.1384,
+      "step": 13550
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9014241099357605,
+      "learning_rate": 3.097369687003896e-06,
+      "loss": 0.1302,
+      "step": 13600
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8704994916915894,
+      "learning_rate": 3.085408777433323e-06,
+      "loss": 0.1347,
+      "step": 13650
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.793929934501648,
+      "learning_rate": 3.0734336734725327e-06,
+      "loss": 0.1396,
+      "step": 13700
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7899141907691956,
+      "learning_rate": 3.0614446654815346e-06,
+      "loss": 0.1331,
+      "step": 13750
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7553165555000305,
+      "learning_rate": 3.049442044157469e-06,
+      "loss": 0.1326,
+      "step": 13800
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7848679423332214,
+      "learning_rate": 3.0374261005275606e-06,
+      "loss": 0.1379,
+      "step": 13850
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9563062191009521,
+      "learning_rate": 3.025397125942056e-06,
+      "loss": 0.1357,
+      "step": 13900
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7703486084938049,
+      "learning_rate": 3.0133554120671653e-06,
+      "loss": 0.1317,
+      "step": 13950
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8696340322494507,
+      "learning_rate": 3.001301250877987e-06,
+      "loss": 0.1309,
+      "step": 14000
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7226243615150452,
+      "learning_rate": 2.9892349346514306e-06,
+      "loss": 0.1346,
+      "step": 14050
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8346455097198486,
+      "learning_rate": 2.977156755959126e-06,
+      "loss": 0.1421,
+      "step": 14100
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9874286651611328,
+      "learning_rate": 2.9650670076603342e-06,
+      "loss": 0.137,
+      "step": 14150
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7056658267974854,
+      "learning_rate": 2.952965982894844e-06,
+      "loss": 0.1355,
+      "step": 14200
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8198770880699158,
+      "learning_rate": 2.9408539750758625e-06,
+      "loss": 0.1353,
+      "step": 14250
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8781945109367371,
+      "learning_rate": 2.9287312778829047e-06,
+      "loss": 0.1333,
+      "step": 14300
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8908482193946838,
+      "learning_rate": 2.9165981852546688e-06,
+      "loss": 0.14,
+      "step": 14350
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.872043251991272,
+      "learning_rate": 2.9044549913819125e-06,
+      "loss": 0.1327,
+      "step": 14400
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8322109580039978,
+      "learning_rate": 2.892301990700316e-06,
+      "loss": 0.1391,
+      "step": 14450
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8258897662162781,
+      "learning_rate": 2.8801394778833475e-06,
+      "loss": 0.1381,
+      "step": 14500
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9012707471847534,
+      "learning_rate": 2.8679677478351147e-06,
+      "loss": 0.1336,
+      "step": 14550
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9831441044807434,
+      "learning_rate": 2.8557870956832135e-06,
+      "loss": 0.135,
+      "step": 14600
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8136524558067322,
+      "learning_rate": 2.8435978167715753e-06,
+      "loss": 0.1367,
+      "step": 14650
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7816210985183716,
+      "learning_rate": 2.8314002066533053e-06,
+      "loss": 0.13,
+      "step": 14700
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8131623268127441,
+      "learning_rate": 2.8191945610835138e-06,
+      "loss": 0.1387,
+      "step": 14750
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7747004628181458,
+      "learning_rate": 2.8069811760121463e-06,
+      "loss": 0.1369,
+      "step": 14800
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.715416669845581,
+      "learning_rate": 2.794760347576809e-06,
+      "loss": 0.1353,
+      "step": 14850
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8204655051231384,
+      "learning_rate": 2.7825323720955853e-06,
+      "loss": 0.1377,
+      "step": 14900
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6974436044692993,
+      "learning_rate": 2.7702975460598545e-06,
+      "loss": 0.1352,
+      "step": 14950
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8297330141067505,
+      "learning_rate": 2.7580561661271015e-06,
+      "loss": 0.1385,
+      "step": 15000
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7207173705101013,
+      "learning_rate": 2.7458085291137213e-06,
+      "loss": 0.1352,
+      "step": 15050
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7374753355979919,
+      "learning_rate": 2.733554931987825e-06,
+      "loss": 0.1357,
+      "step": 15100
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8534835577011108,
+      "learning_rate": 2.7212956718620404e-06,
+      "loss": 0.1286,
+      "step": 15150
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8803167939186096,
+      "learning_rate": 2.709031045986302e-06,
+      "loss": 0.1302,
+      "step": 15200
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7593905925750732,
+      "learning_rate": 2.6967613517406514e-06,
+      "loss": 0.1344,
+      "step": 15250
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8493793606758118,
+      "learning_rate": 2.68448688662802e-06,
+      "loss": 0.133,
+      "step": 15300
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7607067823410034,
+      "learning_rate": 2.6722079482670196e-06,
+      "loss": 0.1355,
+      "step": 15350
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6717010736465454,
+      "learning_rate": 2.6599248343847243e-06,
+      "loss": 0.1337,
+      "step": 15400
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8807181715965271,
+      "learning_rate": 2.6476378428094523e-06,
+      "loss": 0.1308,
+      "step": 15450
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7503638863563538,
+      "learning_rate": 2.6353472714635443e-06,
+      "loss": 0.1358,
+      "step": 15500
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8454340100288391,
+      "learning_rate": 2.6230534183561385e-06,
+      "loss": 0.1349,
+      "step": 15550
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7447136640548706,
+      "learning_rate": 2.6107565815759473e-06,
+      "loss": 0.1366,
+      "step": 15600
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.891941249370575,
+      "learning_rate": 2.598457059284027e-06,
+      "loss": 0.1387,
+      "step": 15650
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8272932767868042,
+      "learning_rate": 2.5861551497065497e-06,
+      "loss": 0.1324,
+      "step": 15700
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8277787566184998,
+      "learning_rate": 2.5738511511275716e-06,
+      "loss": 0.1344,
+      "step": 15750
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7462809681892395,
+      "learning_rate": 2.5615453618818033e-06,
+      "loss": 0.1289,
+      "step": 15800
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6402873396873474,
+      "learning_rate": 2.5492380803473705e-06,
+      "loss": 0.1296,
+      "step": 15850
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8029218912124634,
+      "learning_rate": 2.5369296049385837e-06,
+      "loss": 0.1354,
+      "step": 15900
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.912257969379425,
+      "learning_rate": 2.5246202340987004e-06,
+      "loss": 0.129,
+      "step": 15950
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8321228623390198,
+      "learning_rate": 2.5123102662926912e-06,
+      "loss": 0.1346,
+      "step": 16000
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7897807955741882,
+      "learning_rate": 2.5e-06,
+      "loss": 0.1343,
+      "step": 16050
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8118634223937988,
+      "learning_rate": 2.4876897337073105e-06,
+      "loss": 0.1316,
+      "step": 16100
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8614838123321533,
+      "learning_rate": 2.475379765901301e-06,
+      "loss": 0.1356,
+      "step": 16150
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8548163175582886,
+      "learning_rate": 2.4630703950614176e-06,
+      "loss": 0.1342,
+      "step": 16200
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8339961171150208,
+      "learning_rate": 2.45076191965263e-06,
+      "loss": 0.1253,
+      "step": 16250
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7520368695259094,
+      "learning_rate": 2.4384546381181975e-06,
+      "loss": 0.1281,
+      "step": 16300
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8519024848937988,
+      "learning_rate": 2.4261488488724284e-06,
+      "loss": 0.1339,
+      "step": 16350
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6523202061653137,
+      "learning_rate": 2.413844850293451e-06,
+      "loss": 0.1354,
+      "step": 16400
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8685447573661804,
+      "learning_rate": 2.4015429407159746e-06,
+      "loss": 0.134,
+      "step": 16450
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7570633292198181,
+      "learning_rate": 2.3892434184240536e-06,
+      "loss": 0.131,
+      "step": 16500
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8787110447883606,
+      "learning_rate": 2.3769465816438627e-06,
+      "loss": 0.1305,
+      "step": 16550
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8211491107940674,
+      "learning_rate": 2.3646527285364565e-06,
+      "loss": 0.133,
+      "step": 16600
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8259795308113098,
+      "learning_rate": 2.3523621571905485e-06,
+      "loss": 0.1324,
+      "step": 16650
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8258258700370789,
+      "learning_rate": 2.340075165615276e-06,
+      "loss": 0.1305,
+      "step": 16700
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8116273283958435,
+      "learning_rate": 2.3277920517329813e-06,
+      "loss": 0.1341,
+      "step": 16750
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8958277702331543,
+      "learning_rate": 2.315513113371981e-06,
+      "loss": 0.1274,
+      "step": 16800
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7367030382156372,
+      "learning_rate": 2.303238648259349e-06,
+      "loss": 0.1276,
+      "step": 16850
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7800854444503784,
+      "learning_rate": 2.2909689540136986e-06,
+      "loss": 0.1347,
+      "step": 16900
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7913332581520081,
+      "learning_rate": 2.27870432813796e-06,
+      "loss": 0.1293,
+      "step": 16950
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7659708261489868,
+      "learning_rate": 2.2664450680121757e-06,
+      "loss": 0.1334,
+      "step": 17000
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7981456518173218,
+      "learning_rate": 2.254191470886279e-06,
+      "loss": 0.1356,
+      "step": 17050
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8289954662322998,
+      "learning_rate": 2.241943833872899e-06,
+      "loss": 0.1305,
+      "step": 17100
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8367072343826294,
+      "learning_rate": 2.2297024539401463e-06,
+      "loss": 0.1307,
+      "step": 17150
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8345316648483276,
+      "learning_rate": 2.2174676279044155e-06,
+      "loss": 0.1283,
+      "step": 17200
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8068302869796753,
+      "learning_rate": 2.2052396524231924e-06,
+      "loss": 0.1342,
+      "step": 17250
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.811231791973114,
+      "learning_rate": 2.193018823987854e-06,
+      "loss": 0.1374,
+      "step": 17300
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8022611141204834,
+      "learning_rate": 2.180805438916487e-06,
+      "loss": 0.1296,
+      "step": 17350
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8078568577766418,
+      "learning_rate": 2.1685997933466947e-06,
+      "loss": 0.1273,
+      "step": 17400
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.749242901802063,
+      "learning_rate": 2.1564021832284255e-06,
+      "loss": 0.1302,
+      "step": 17450
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8693391680717468,
+      "learning_rate": 2.1442129043167877e-06,
+      "loss": 0.1296,
+      "step": 17500
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9399083852767944,
+      "learning_rate": 2.1320322521648857e-06,
+      "loss": 0.127,
+      "step": 17550
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8048396110534668,
+      "learning_rate": 2.119860522116653e-06,
+      "loss": 0.128,
+      "step": 17600
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7869442701339722,
+      "learning_rate": 2.1076980092996845e-06,
+      "loss": 0.1324,
+      "step": 17650
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8741517066955566,
+      "learning_rate": 2.0955450086180883e-06,
+      "loss": 0.1351,
+      "step": 17700
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9408276081085205,
+      "learning_rate": 2.083401814745332e-06,
+      "loss": 0.1331,
+      "step": 17750
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8211988210678101,
+      "learning_rate": 2.071268722117096e-06,
+      "loss": 0.1281,
+      "step": 17800
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8505455851554871,
+      "learning_rate": 2.0591460249241383e-06,
+      "loss": 0.1296,
+      "step": 17850
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8733687400817871,
+      "learning_rate": 2.0470340171051567e-06,
+      "loss": 0.1339,
+      "step": 17900
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8266494274139404,
+      "learning_rate": 2.034932992339666e-06,
+      "loss": 0.1283,
+      "step": 17950
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7836053967475891,
+      "learning_rate": 2.022843244040874e-06,
+      "loss": 0.1243,
+      "step": 18000
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9171106815338135,
+      "learning_rate": 2.0107650653485707e-06,
+      "loss": 0.1341,
+      "step": 18050
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.796025812625885,
+      "learning_rate": 1.998698749122014e-06,
+      "loss": 0.1336,
+      "step": 18100
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8582082986831665,
+      "learning_rate": 1.986644587932835e-06,
+      "loss": 0.1311,
+      "step": 18150
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7605078220367432,
+      "learning_rate": 1.9746028740579453e-06,
+      "loss": 0.1288,
+      "step": 18200
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8088697195053101,
+      "learning_rate": 1.96257389947244e-06,
+      "loss": 0.1285,
+      "step": 18250
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8578440546989441,
+      "learning_rate": 1.9505579558425315e-06,
+      "loss": 0.1346,
+      "step": 18300
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7750033140182495,
+      "learning_rate": 1.938555334518466e-06,
+      "loss": 0.1306,
+      "step": 18350
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8363983631134033,
+      "learning_rate": 1.926566326527468e-06,
+      "loss": 0.1313,
+      "step": 18400
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7722070813179016,
+      "learning_rate": 1.914591222566678e-06,
+      "loss": 0.1337,
+      "step": 18450
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8674912452697754,
+      "learning_rate": 1.9026303129961049e-06,
+      "loss": 0.1268,
+      "step": 18500
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0055360794067383,
+      "learning_rate": 1.8906838878315886e-06,
+      "loss": 0.1299,
+      "step": 18550
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.797717273235321,
+      "learning_rate": 1.878752236737765e-06,
+      "loss": 0.1261,
+      "step": 18600
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8092938661575317,
+      "learning_rate": 1.8668356490210449e-06,
+      "loss": 0.1362,
+      "step": 18650
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8348585367202759,
+      "learning_rate": 1.8549344136225946e-06,
+      "loss": 0.1276,
+      "step": 18700
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8768981099128723,
+      "learning_rate": 1.8430488191113373e-06,
+      "loss": 0.1332,
+      "step": 18750
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7733373045921326,
+      "learning_rate": 1.8311791536769485e-06,
+      "loss": 0.134,
+      "step": 18800
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8418094515800476,
+      "learning_rate": 1.819325705122873e-06,
+      "loss": 0.133,
+      "step": 18850
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8411191701889038,
+      "learning_rate": 1.8074887608593477e-06,
+      "loss": 0.1304,
+      "step": 18900
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9010744094848633,
+      "learning_rate": 1.7956686078964257e-06,
+      "loss": 0.131,
+      "step": 18950
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8585550785064697,
+      "learning_rate": 1.7838655328370268e-06,
+      "loss": 0.1308,
+      "step": 19000
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8417059183120728,
+      "learning_rate": 1.7720798218699798e-06,
+      "loss": 0.1282,
+      "step": 19050
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8636181354522705,
+      "learning_rate": 1.7603117607630892e-06,
+      "loss": 0.1256,
+      "step": 19100
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8818506598472595,
+      "learning_rate": 1.7485616348562023e-06,
+      "loss": 0.1306,
+      "step": 19150
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8800364136695862,
+      "learning_rate": 1.7368297290542918e-06,
+      "loss": 0.1303,
+      "step": 19200
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8845418095588684,
+      "learning_rate": 1.72511632782055e-06,
+      "loss": 0.131,
+      "step": 19250
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9156243205070496,
+      "learning_rate": 1.7134217151694873e-06,
+      "loss": 0.1269,
+      "step": 19300
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.772602915763855,
+      "learning_rate": 1.7017461746600506e-06,
+      "loss": 0.1321,
+      "step": 19350
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7767558097839355,
+      "learning_rate": 1.690089989388741e-06,
+      "loss": 0.1302,
+      "step": 19400
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.816466212272644,
+      "learning_rate": 1.678453441982758e-06,
+      "loss": 0.132,
+      "step": 19450
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.746376633644104,
+      "learning_rate": 1.66683681459314e-06,
+      "loss": 0.1321,
+      "step": 19500
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9455410242080688,
+      "learning_rate": 1.6552403888879243e-06,
+      "loss": 0.1265,
+      "step": 19550
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9499367475509644,
+      "learning_rate": 1.6436644460453218e-06,
+      "loss": 0.1266,
+      "step": 19600
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8672183156013489,
+      "learning_rate": 1.6321092667468926e-06,
+      "loss": 0.1288,
+      "step": 19650
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7655101418495178,
+      "learning_rate": 1.6205751311707463e-06,
+      "loss": 0.1346,
+      "step": 19700
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7775772213935852,
+      "learning_rate": 1.6090623189847443e-06,
+      "loss": 0.1276,
+      "step": 19750
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8449127078056335,
+      "learning_rate": 1.5975711093397223e-06,
+      "loss": 0.1323,
+      "step": 19800
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9375335574150085,
+      "learning_rate": 1.5861017808627167e-06,
+      "loss": 0.1338,
+      "step": 19850
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7458611130714417,
+      "learning_rate": 1.574654611650214e-06,
+      "loss": 0.1307,
+      "step": 19900
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.749284029006958,
+      "learning_rate": 1.5632298792614064e-06,
+      "loss": 0.1216,
+      "step": 19950
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8348965644836426,
+      "learning_rate": 1.5518278607114585e-06,
+      "loss": 0.1253,
+      "step": 20000
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9566074013710022,
+      "learning_rate": 1.540448832464796e-06,
+      "loss": 0.1289,
+      "step": 20050
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8038182854652405,
+      "learning_rate": 1.5290930704283953e-06,
+      "loss": 0.1266,
+      "step": 20100
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8434205055236816,
+      "learning_rate": 1.517760849945103e-06,
+      "loss": 0.1276,
+      "step": 20150
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7553625702857971,
+      "learning_rate": 1.5064524457869506e-06,
+      "loss": 0.128,
+      "step": 20200
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7524191737174988,
+      "learning_rate": 1.4951681321484952e-06,
+      "loss": 0.1313,
+      "step": 20250
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8314839601516724,
+      "learning_rate": 1.4839081826401756e-06,
+      "loss": 0.1255,
+      "step": 20300
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8954294323921204,
+      "learning_rate": 1.47267287028167e-06,
+      "loss": 0.1259,
+      "step": 20350
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7757198810577393,
+      "learning_rate": 1.4614624674952843e-06,
+      "loss": 0.1288,
+      "step": 20400
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9685442447662354,
+      "learning_rate": 1.4502772460993387e-06,
+      "loss": 0.1313,
+      "step": 20450
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8876562118530273,
+      "learning_rate": 1.4391174773015836e-06,
+      "loss": 0.1326,
+      "step": 20500
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0486761331558228,
+      "learning_rate": 1.4279834316926217e-06,
+      "loss": 0.126,
+      "step": 20550
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.806884765625,
+      "learning_rate": 1.4168753792393413e-06,
+      "loss": 0.1286,
+      "step": 20600
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8478713631629944,
+      "learning_rate": 1.405793589278382e-06,
+      "loss": 0.1248,
+      "step": 20650
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0112961530685425,
+      "learning_rate": 1.394738330509593e-06,
+      "loss": 0.1265,
+      "step": 20700
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9463728666305542,
+      "learning_rate": 1.3837098709895246e-06,
+      "loss": 0.1274,
+      "step": 20750
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8236443996429443,
+      "learning_rate": 1.3727084781249251e-06,
+      "loss": 0.1287,
+      "step": 20800
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8493944406509399,
+      "learning_rate": 1.3617344186662585e-06,
+      "loss": 0.1252,
+      "step": 20850
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7850596904754639,
+      "learning_rate": 1.3507879587012378e-06,
+      "loss": 0.1199,
+      "step": 20900
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.119826316833496,
+      "learning_rate": 1.3398693636483707e-06,
+      "loss": 0.1259,
+      "step": 20950
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8658459782600403,
+      "learning_rate": 1.328978898250525e-06,
+      "loss": 0.1276,
+      "step": 21000
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8346827030181885,
+      "learning_rate": 1.31811682656851e-06,
+      "loss": 0.126,
+      "step": 21050
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0707381963729858,
+      "learning_rate": 1.307283411974674e-06,
+      "loss": 0.1266,
+      "step": 21100
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8419653177261353,
+      "learning_rate": 1.2964789171465164e-06,
+      "loss": 0.1253,
+      "step": 21150
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6659058332443237,
+      "learning_rate": 1.2857036040603204e-06,
+      "loss": 0.1251,
+      "step": 21200
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9063161015510559,
+      "learning_rate": 1.2749577339848007e-06,
+      "loss": 0.1245,
+      "step": 21250
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9174826741218567,
+      "learning_rate": 1.2642415674747675e-06,
+      "loss": 0.1322,
+      "step": 21300
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8082279562950134,
+      "learning_rate": 1.25355536436481e-06,
+      "loss": 0.1294,
+      "step": 21350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9118791222572327,
+      "learning_rate": 1.2428993837629943e-06,
+      "loss": 0.1277,
+      "step": 21400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9657939076423645,
+      "learning_rate": 1.2322738840445867e-06,
+      "loss": 0.1281,
+      "step": 21450
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8594872355461121,
+      "learning_rate": 1.2216791228457778e-06,
+      "loss": 0.1193,
+      "step": 21500
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0133215188980103,
+      "learning_rate": 1.2111153570574454e-06,
+      "loss": 0.1275,
+      "step": 21550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8617738485336304,
+      "learning_rate": 1.2005828428189256e-06,
+      "loss": 0.1269,
+      "step": 21600
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7881460189819336,
+      "learning_rate": 1.1900818355117918e-06,
+      "loss": 0.1251,
+      "step": 21650
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8519832491874695,
+      "learning_rate": 1.1796125897536782e-06,
+      "loss": 0.1311,
+      "step": 21700
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8465796113014221,
+      "learning_rate": 1.1691753593920884e-06,
+      "loss": 0.1293,
+      "step": 21750
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.892995297908783,
+      "learning_rate": 1.1587703974982583e-06,
+      "loss": 0.1272,
+      "step": 21800
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8453019261360168,
+      "learning_rate": 1.148397956361007e-06,
+      "loss": 0.1267,
+      "step": 21850
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7927883863449097,
+      "learning_rate": 1.1380582874806208e-06,
+      "loss": 0.1274,
+      "step": 21900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.848616898059845,
+      "learning_rate": 1.127751641562765e-06,
+      "loss": 0.1318,
+      "step": 21950
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8871294260025024,
+      "learning_rate": 1.1174782685123919e-06,
+      "loss": 0.1278,
+      "step": 22000
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9262788891792297,
+      "learning_rate": 1.107238417427691e-06,
+      "loss": 0.1242,
+      "step": 22050
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0166550874710083,
+      "learning_rate": 1.0970323365940443e-06,
+      "loss": 0.1271,
+      "step": 22100
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9763181805610657,
+      "learning_rate": 1.0868602734780075e-06,
+      "loss": 0.1266,
+      "step": 22150
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8121241927146912,
+      "learning_rate": 1.0767224747213102e-06,
+      "loss": 0.1286,
+      "step": 22200
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7165303230285645,
+      "learning_rate": 1.0666191861348741e-06,
+      "loss": 0.1275,
+      "step": 22250
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0297795534133911,
+      "learning_rate": 1.0565506526928548e-06,
+      "loss": 0.1251,
+      "step": 22300
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8937394022941589,
+      "learning_rate": 1.0465171185267007e-06,
+      "loss": 0.1297,
+      "step": 22350
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9162909388542175,
+      "learning_rate": 1.036518826919233e-06,
+      "loss": 0.1325,
+      "step": 22400
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8973715305328369,
+      "learning_rate": 1.0265560202987474e-06,
+      "loss": 0.1296,
+      "step": 22450
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8906801342964172,
+      "learning_rate": 1.0166289402331391e-06,
+      "loss": 0.128,
+      "step": 22500
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7446568608283997,
+      "learning_rate": 1.006737827424038e-06,
+      "loss": 0.1238,
+      "step": 22550
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0760340690612793,
+      "learning_rate": 9.9688292170098e-07,
+      "loss": 0.129,
+      "step": 22600
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9284375309944153,
+      "learning_rate": 9.870644620155878e-07,
+      "loss": 0.125,
+      "step": 22650
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9257674217224121,
+      "learning_rate": 9.77282686435777e-07,
+      "loss": 0.1272,
+      "step": 22700
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7764928936958313,
+      "learning_rate": 9.67537832139989e-07,
+      "loss": 0.1224,
+      "step": 22750
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7755612730979919,
+      "learning_rate": 9.578301354114292e-07,
+      "loss": 0.1279,
+      "step": 22800
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.847436785697937,
+      "learning_rate": 9.481598316323504e-07,
+      "loss": 0.1273,
+      "step": 22850
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8615244030952454,
+      "learning_rate": 9.385271552783376e-07,
+      "loss": 0.1267,
+      "step": 22900
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8775952458381653,
+      "learning_rate": 9.289323399126216e-07,
+      "loss": 0.1276,
+      "step": 22950
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9326611161231995,
+      "learning_rate": 9.193756181804248e-07,
+      "loss": 0.1296,
+      "step": 23000
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6743093132972717,
+      "learning_rate": 9.098572218033084e-07,
+      "loss": 0.1279,
+      "step": 23050
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9707579612731934,
+      "learning_rate": 9.003773815735644e-07,
+      "loss": 0.1283,
+      "step": 23100
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8160597681999207,
+      "learning_rate": 8.90936327348613e-07,
+      "loss": 0.127,
+      "step": 23150
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.070728063583374,
+      "learning_rate": 8.815342880454312e-07,
+      "loss": 0.1265,
+      "step": 23200
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.945746123790741,
+      "learning_rate": 8.721714916350019e-07,
+      "loss": 0.1241,
+      "step": 23250
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.844608724117279,
+      "learning_rate": 8.628481651367876e-07,
+      "loss": 0.1271,
+      "step": 23300
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8871841430664062,
+      "learning_rate": 8.535645346132246e-07,
+      "loss": 0.1313,
+      "step": 23350
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.791871964931488,
+      "learning_rate": 8.443208251642418e-07,
+      "loss": 0.1242,
+      "step": 23400
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9160509705543518,
+      "learning_rate": 8.351172609218033e-07,
+      "loss": 0.1217,
+      "step": 23450
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7962935566902161,
+      "learning_rate": 8.259540650444736e-07,
+      "loss": 0.128,
+      "step": 23500
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7934704422950745,
+      "learning_rate": 8.168314597120059e-07,
+      "loss": 0.1217,
+      "step": 23550
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7720403671264648,
+      "learning_rate": 8.077496661199557e-07,
+      "loss": 0.1252,
+      "step": 23600
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.806203305721283,
+      "learning_rate": 7.987089044743182e-07,
+      "loss": 0.1276,
+      "step": 23650
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9752463698387146,
+      "learning_rate": 7.897093939861878e-07,
+      "loss": 0.1288,
+      "step": 23700
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8070717453956604,
+      "learning_rate": 7.807513528664415e-07,
+      "loss": 0.1272,
+      "step": 23750
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.773583710193634,
+      "learning_rate": 7.71834998320454e-07,
+      "loss": 0.1192,
+      "step": 23800
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7892506122589111,
+      "learning_rate": 7.629605465428211e-07,
+      "loss": 0.1217,
+      "step": 23850
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6606497764587402,
+      "learning_rate": 7.541282127121291e-07,
+      "loss": 0.1222,
+      "step": 23900
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8119127154350281,
+      "learning_rate": 7.453382109857269e-07,
+      "loss": 0.1241,
+      "step": 23950
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7261886596679688,
+      "learning_rate": 7.365907544945398e-07,
+      "loss": 0.1294,
+      "step": 24000
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8544095754623413,
+      "learning_rate": 7.27886055337902e-07,
+      "loss": 0.126,
+      "step": 24050
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9304544925689697,
+      "learning_rate": 7.192243245784075e-07,
+      "loss": 0.1296,
+      "step": 24100
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8993602395057678,
+      "learning_rate": 7.106057722368012e-07,
+      "loss": 0.1225,
+      "step": 24150
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7909960150718689,
+      "learning_rate": 7.020306072868804e-07,
+      "loss": 0.1261,
+      "step": 24200
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8330161571502686,
+      "learning_rate": 6.934990376504269e-07,
+      "loss": 0.1242,
+      "step": 24250
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9058279395103455,
+      "learning_rate": 6.850112701921735e-07,
+      "loss": 0.1224,
+      "step": 24300
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9598106741905212,
+      "learning_rate": 6.76567510714777e-07,
+      "loss": 0.1261,
+      "step": 24350
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8675400018692017,
+      "learning_rate": 6.681679639538388e-07,
+      "loss": 0.1275,
+      "step": 24400
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7243128418922424,
+      "learning_rate": 6.598128335729332e-07,
+      "loss": 0.1301,
+      "step": 24450
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.866690456867218,
+      "learning_rate": 6.515023221586722e-07,
+      "loss": 0.1276,
+      "step": 24500
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8871277570724487,
+      "learning_rate": 6.432366312157933e-07,
+      "loss": 0.1285,
+      "step": 24550
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6913353800773621,
+      "learning_rate": 6.35015961162273e-07,
+      "loss": 0.1207,
+      "step": 24600
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8807539939880371,
+      "learning_rate": 6.268405113244677e-07,
+      "loss": 0.1227,
+      "step": 24650
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9127259850502014,
+      "learning_rate": 6.187104799322805e-07,
+      "loss": 0.1234,
+      "step": 24700
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0802431106567383,
+      "learning_rate": 6.106260641143547e-07,
+      "loss": 0.1257,
+      "step": 24750
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8640583157539368,
+      "learning_rate": 6.025874598932937e-07,
+      "loss": 0.1195,
+      "step": 24800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8064579963684082,
+      "learning_rate": 5.945948621809092e-07,
+      "loss": 0.1266,
+      "step": 24850
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8165306448936462,
+      "learning_rate": 5.866484647734935e-07,
+      "loss": 0.1229,
+      "step": 24900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.885006308555603,
+      "learning_rate": 5.787484603471221e-07,
+      "loss": 0.1215,
+      "step": 24950
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0080711841583252,
+      "learning_rate": 5.708950404529812e-07,
+      "loss": 0.1233,
+      "step": 25000
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8697314858436584,
+      "learning_rate": 5.630883955127211e-07,
+      "loss": 0.1215,
+      "step": 25050
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8583015203475952,
+      "learning_rate": 5.553287148138462e-07,
+      "loss": 0.1267,
+      "step": 25100
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.771619439125061,
+      "learning_rate": 5.47616186505113e-07,
+      "loss": 0.1185,
+      "step": 25150
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8931272625923157,
+      "learning_rate": 5.399509975919828e-07,
+      "loss": 0.1232,
+      "step": 25200
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9257284998893738,
+      "learning_rate": 5.323333339320739e-07,
+      "loss": 0.126,
+      "step": 25250
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7422381639480591,
+      "learning_rate": 5.247633802306637e-07,
+      "loss": 0.1258,
+      "step": 25300
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8282724022865295,
+      "learning_rate": 5.172413200362092e-07,
+      "loss": 0.1244,
+      "step": 25350
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9066148996353149,
+      "learning_rate": 5.097673357358906e-07,
+      "loss": 0.1227,
+      "step": 25400
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0274531841278076,
+      "learning_rate": 5.023416085511976e-07,
+      "loss": 0.1206,
+      "step": 25450
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8062939047813416,
+      "learning_rate": 4.949643185335288e-07,
+      "loss": 0.121,
+      "step": 25500
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9104852080345154,
+      "learning_rate": 4.876356445598279e-07,
+      "loss": 0.1318,
+      "step": 25550
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9484885334968567,
+      "learning_rate": 4.803557643282486e-07,
+      "loss": 0.1215,
+      "step": 25600
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0123718976974487,
+      "learning_rate": 4.731248543538405e-07,
+      "loss": 0.1232,
+      "step": 25650
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7490533590316772,
+      "learning_rate": 4.6594308996427696e-07,
+      "loss": 0.1226,
+      "step": 25700
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.867254912853241,
+      "learning_rate": 4.588106452955973e-07,
+      "loss": 0.1183,
+      "step": 25750
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.889826238155365,
+      "learning_rate": 4.517276932879877e-07,
+      "loss": 0.1229,
+      "step": 25800
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8290361166000366,
+      "learning_rate": 4.446944056815866e-07,
+      "loss": 0.123,
+      "step": 25850
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8148057460784912,
+      "learning_rate": 4.377109530123216e-07,
+      "loss": 0.1275,
+      "step": 25900
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8523454666137695,
+      "learning_rate": 4.307775046077739e-07,
+      "loss": 0.1227,
+      "step": 25950
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8022044897079468,
+      "learning_rate": 4.2389422858307244e-07,
+      "loss": 0.1273,
+      "step": 26000
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7871876955032349,
+      "learning_rate": 4.1706129183681834e-07,
+      "loss": 0.1281,
+      "step": 26050
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8694405555725098,
+      "learning_rate": 4.10278860047037e-07,
+      "loss": 0.1202,
+      "step": 26100
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9443354606628418,
+      "learning_rate": 4.035470976671621e-07,
+      "loss": 0.1194,
+      "step": 26150
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8417434096336365,
+      "learning_rate": 3.9686616792204677e-07,
+      "loss": 0.1213,
+      "step": 26200
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.855648398399353,
+      "learning_rate": 3.902362328040091e-07,
+      "loss": 0.1272,
+      "step": 26250
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8079158067703247,
+      "learning_rate": 3.836574530688983e-07,
+      "loss": 0.1237,
+      "step": 26300
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9010571837425232,
+      "learning_rate": 3.7712998823220243e-07,
+      "loss": 0.1268,
+      "step": 26350
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9768967032432556,
+      "learning_rate": 3.7065399656517955e-07,
+      "loss": 0.1226,
+      "step": 26400
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.720100462436676,
+      "learning_rate": 3.6422963509101626e-07,
+      "loss": 0.1212,
+      "step": 26450
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8054412603378296,
+      "learning_rate": 3.578570595810274e-07,
+      "loss": 0.12,
+      "step": 26500
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7987850904464722,
+      "learning_rate": 3.515364245508704e-07,
+      "loss": 0.1205,
+      "step": 26550
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7818375825881958,
+      "learning_rate": 3.452678832568071e-07,
+      "loss": 0.1301,
+      "step": 26600
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8323056101799011,
+      "learning_rate": 3.390515876919831e-07,
+      "loss": 0.1208,
+      "step": 26650
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8122744560241699,
+      "learning_rate": 3.328876885827406e-07,
+      "loss": 0.121,
+      "step": 26700
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9000005125999451,
+      "learning_rate": 3.267763353849704e-07,
+      "loss": 0.1187,
+      "step": 26750
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7364488840103149,
+      "learning_rate": 3.207176762804814e-07,
+      "loss": 0.1228,
+      "step": 26800
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9550331234931946,
+      "learning_rate": 3.1471185817341153e-07,
+      "loss": 0.1197,
+      "step": 26850
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8683730959892273,
+      "learning_rate": 3.0875902668666386e-07,
+      "loss": 0.1279,
+      "step": 26900
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8578702807426453,
+      "learning_rate": 3.0285932615837646e-07,
+      "loss": 0.1212,
+      "step": 26950
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9135186076164246,
+      "learning_rate": 2.970128996384228e-07,
+      "loss": 0.1219,
+      "step": 27000
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8612381815910339,
+      "learning_rate": 2.9121988888494297e-07,
+      "loss": 0.1274,
+      "step": 27050
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8324002027511597,
+      "learning_rate": 2.854804343609058e-07,
+      "loss": 0.1282,
+      "step": 27100
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8618267774581909,
+      "learning_rate": 2.7979467523070484e-07,
+      "loss": 0.1219,
+      "step": 27150
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8603401780128479,
+      "learning_rate": 2.741627493567822e-07,
+      "loss": 0.1225,
+      "step": 27200
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0212452411651611,
+      "learning_rate": 2.685847932962868e-07,
+      "loss": 0.1225,
+      "step": 27250
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7207959294319153,
+      "learning_rate": 2.630609422977623e-07,
+      "loss": 0.1266,
+      "step": 27300
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9578468203544617,
+      "learning_rate": 2.575913302978697e-07,
+      "loss": 0.1274,
+      "step": 27350
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.033157229423523,
+      "learning_rate": 2.5217608991813774e-07,
+      "loss": 0.116,
+      "step": 27400
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8012121915817261,
+      "learning_rate": 2.468153524617478e-07,
+      "loss": 0.121,
+      "step": 27450
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7251663208007812,
+      "learning_rate": 2.4150924791035037e-07,
+      "loss": 0.1213,
+      "step": 27500
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8772590160369873,
+      "learning_rate": 2.3625790492091544e-07,
+      "loss": 0.1205,
+      "step": 27550
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9037516713142395,
+      "learning_rate": 2.3106145082260777e-07,
+      "loss": 0.1234,
+      "step": 27600
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8488568067550659,
+      "learning_rate": 2.2592001161370392e-07,
+      "loss": 0.1203,
+      "step": 27650
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8486872911453247,
+      "learning_rate": 2.2083371195853797e-07,
+      "loss": 0.1309,
+      "step": 27700
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.964940071105957,
+      "learning_rate": 2.158026751844733e-07,
+      "loss": 0.1228,
+      "step": 27750
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8965867161750793,
+      "learning_rate": 2.1082702327891918e-07,
+      "loss": 0.13,
+      "step": 27800
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8563171029090881,
+      "learning_rate": 2.0590687688636619e-07,
+      "loss": 0.1187,
+      "step": 27850
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.000481367111206,
+      "learning_rate": 2.0104235530546745e-07,
+      "loss": 0.1267,
+      "step": 27900
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.793648898601532,
+      "learning_rate": 1.9623357648614088e-07,
+      "loss": 0.1225,
+      "step": 27950
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8979523181915283,
+      "learning_rate": 1.914806570267111e-07,
+      "loss": 0.1234,
+      "step": 28000
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9331807494163513,
+      "learning_rate": 1.8678371217108387e-07,
+      "loss": 0.1197,
+      "step": 28050
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8871297240257263,
+      "learning_rate": 1.821428558059493e-07,
+      "loss": 0.1159,
+      "step": 28100
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0340665578842163,
+      "learning_rate": 1.7755820045802146e-07,
+      "loss": 0.1266,
+      "step": 28150
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.055516004562378,
+      "learning_rate": 1.7302985729131e-07,
+      "loss": 0.1255,
+      "step": 28200
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8390622138977051,
+      "learning_rate": 1.6855793610442484e-07,
+      "loss": 0.1256,
+      "step": 28250
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8009083867073059,
+      "learning_rate": 1.6414254532791357e-07,
+      "loss": 0.1215,
+      "step": 28300
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0313583612442017,
+      "learning_rate": 1.5978379202163275e-07,
+      "loss": 0.1274,
+      "step": 28350
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9737227559089661,
+      "learning_rate": 1.554817818721513e-07,
+      "loss": 0.1248,
+      "step": 28400
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0153248310089111,
+      "learning_rate": 1.51236619190189e-07,
+      "loss": 0.1247,
+      "step": 28450
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8174459338188171,
+      "learning_rate": 1.4704840690808658e-07,
+      "loss": 0.122,
+      "step": 28500
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9065647125244141,
+      "learning_rate": 1.4291724657730904e-07,
+      "loss": 0.1245,
+      "step": 28550
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9131068587303162,
+      "learning_rate": 1.3884323836598656e-07,
+      "loss": 0.1199,
+      "step": 28600
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8409487009048462,
+      "learning_rate": 1.348264810564809e-07,
+      "loss": 0.1282,
+      "step": 28650
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8395945429801941,
+      "learning_rate": 1.3086707204299415e-07,
+      "loss": 0.1264,
+      "step": 28700
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8718541860580444,
+      "learning_rate": 1.269651073292058e-07,
+      "loss": 0.1204,
+      "step": 28750
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8746705055236816,
+      "learning_rate": 1.2312068152594448e-07,
+      "loss": 0.1244,
+      "step": 28800
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8849207758903503,
+      "learning_rate": 1.1933388784889617e-07,
+      "loss": 0.1215,
+      "step": 28850
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9171212315559387,
+      "learning_rate": 1.1560481811633911e-07,
+      "loss": 0.1274,
+      "step": 28900
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9320688247680664,
+      "learning_rate": 1.1193356274692424e-07,
+      "loss": 0.1211,
+      "step": 28950
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7832902073860168,
+      "learning_rate": 1.0832021075747712e-07,
+      "loss": 0.1257,
+      "step": 29000
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8894676566123962,
+      "learning_rate": 1.047648497608414e-07,
+      "loss": 0.1268,
+      "step": 29050
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7472261786460876,
+      "learning_rate": 1.0126756596375687e-07,
+      "loss": 0.1222,
+      "step": 29100
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9430793523788452,
+      "learning_rate": 9.782844416476423e-08,
+      "loss": 0.1219,
+      "step": 29150
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8395842909812927,
+      "learning_rate": 9.444756775215446e-08,
+      "loss": 0.1212,
+      "step": 29200
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8106001019477844,
+      "learning_rate": 9.112501870194273e-08,
+      "loss": 0.124,
+      "step": 29250
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9083417654037476,
+      "learning_rate": 8.786087757588269e-08,
+      "loss": 0.1269,
+      "step": 29300
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9193884134292603,
+      "learning_rate": 8.465522351951305e-08,
+      "loss": 0.1207,
+      "step": 29350
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9699887633323669,
+      "learning_rate": 8.150813426023752e-08,
+      "loss": 0.1245,
+      "step": 29400
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.984941303730011,
+      "learning_rate": 7.841968610544121e-08,
+      "loss": 0.1228,
+      "step": 29450
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0322271585464478,
+      "learning_rate": 7.538995394063996e-08,
+      "loss": 0.1247,
+      "step": 29500
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9415401220321655,
+      "learning_rate": 7.241901122766515e-08,
+      "loss": 0.1209,
+      "step": 29550
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8878545165061951,
+      "learning_rate": 6.950693000288056e-08,
+      "loss": 0.1268,
+      "step": 29600
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8881374001502991,
+      "learning_rate": 6.665378087543889e-08,
+      "loss": 0.1239,
+      "step": 29650
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8045355081558228,
+      "learning_rate": 6.385963302556642e-08,
+      "loss": 0.1197,
+      "step": 29700
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8438431620597839,
+      "learning_rate": 6.112455420288821e-08,
+      "loss": 0.1242,
+      "step": 29750
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8916789293289185,
+      "learning_rate": 5.844861072478336e-08,
+      "loss": 0.1244,
+      "step": 29800
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8392596244812012,
+      "learning_rate": 5.583186747477848e-08,
+      "loss": 0.1234,
+      "step": 29850
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8782594203948975,
+      "learning_rate": 5.32743879009745e-08,
+      "loss": 0.1225,
+      "step": 29900
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8895407319068909,
+      "learning_rate": 5.077623401450599e-08,
+      "loss": 0.1212,
+      "step": 29950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9181334376335144,
+      "learning_rate": 4.8337466388040935e-08,
+      "loss": 0.1203,
+      "step": 30000
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7273566126823425,
+      "learning_rate": 4.595814415430916e-08,
+      "loss": 0.1235,
+      "step": 30050
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8395570516586304,
+      "learning_rate": 4.3638325004670134e-08,
+      "loss": 0.1275,
+      "step": 30100
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9657353758811951,
+      "learning_rate": 4.1378065187714365e-08,
+      "loss": 0.1194,
+      "step": 30150
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8621689081192017,
+      "learning_rate": 3.917741950789727e-08,
+      "loss": 0.1219,
+      "step": 30200
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8802511692047119,
+      "learning_rate": 3.703644132421386e-08,
+      "loss": 0.1246,
+      "step": 30250
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8545549511909485,
+      "learning_rate": 3.4955182548901956e-08,
+      "loss": 0.1236,
+      "step": 30300
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9442453384399414,
+      "learning_rate": 3.293369364618465e-08,
+      "loss": 0.1255,
+      "step": 30350
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8238327503204346,
+      "learning_rate": 3.097202363104679e-08,
+      "loss": 0.1223,
+      "step": 30400
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9696635007858276,
+      "learning_rate": 2.9070220068045663e-08,
+      "loss": 0.1203,
+      "step": 30450
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.760059118270874,
+      "learning_rate": 2.722832907015971e-08,
+      "loss": 0.1241,
+      "step": 30500
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9633733034133911,
+      "learning_rate": 2.544639529766829e-08,
+      "loss": 0.1282,
+      "step": 30550
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.949769914150238,
+      "learning_rate": 2.3724461957068955e-08,
+      "loss": 0.1244,
+      "step": 30600
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8858135342597961,
+      "learning_rate": 2.206257080003188e-08,
+      "loss": 0.1241,
+      "step": 30650
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8067658543586731,
+      "learning_rate": 2.0460762122385124e-08,
+      "loss": 0.1162,
+      "step": 30700
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.874151885509491,
+      "learning_rate": 1.8919074763138757e-08,
+      "loss": 0.1246,
+      "step": 30750
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8034783601760864,
+      "learning_rate": 1.7437546103542814e-08,
+      "loss": 0.1246,
+      "step": 30800
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9454385042190552,
+      "learning_rate": 1.6016212066181368e-08,
+      "loss": 0.122,
+      "step": 30850
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8052737712860107,
+      "learning_rate": 1.4655107114101008e-08,
+      "loss": 0.1256,
+      "step": 30900
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9731265902519226,
+      "learning_rate": 1.3354264249975379e-08,
+      "loss": 0.1246,
+      "step": 30950
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8946937918663025,
+      "learning_rate": 1.2113715015304728e-08,
+      "loss": 0.1245,
+      "step": 31000
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8165032863616943,
+      "learning_rate": 1.0933489489651783e-08,
+      "loss": 0.12,
+      "step": 31050
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8956162333488464,
+      "learning_rate": 9.81361628991151e-09,
+      "loss": 0.1226,
+      "step": 31100
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9210562705993652,
+      "learning_rate": 8.754122569618329e-09,
+      "loss": 0.123,
+      "step": 31150
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9154261946678162,
+      "learning_rate": 7.755034018286644e-09,
+      "loss": 0.1242,
+      "step": 31200
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7360591888427734,
+      "learning_rate": 6.816374860788566e-09,
+      "loss": 0.1215,
+      "step": 31250
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.83289635181427,
+      "learning_rate": 5.938167856766319e-09,
+      "loss": 0.1224,
+      "step": 31300
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7892395257949829,
+      "learning_rate": 5.120434300080745e-09,
+      "loss": 0.1207,
+      "step": 31350
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9500446915626526,
+      "learning_rate": 4.363194018293937e-09,
+      "loss": 0.13,
+      "step": 31400
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8159528970718384,
+      "learning_rate": 3.666465372190453e-09,
+      "loss": 0.1224,
+      "step": 31450
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7910547852516174,
+      "learning_rate": 3.030265255329623e-09,
+      "loss": 0.1228,
+      "step": 31500
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8346930742263794,
+      "learning_rate": 2.4546090936383717e-09,
+      "loss": 0.1239,
+      "step": 31550
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8468943238258362,
+      "learning_rate": 1.9395108450351308e-09,
+      "loss": 0.119,
+      "step": 31600
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8092018961906433,
+      "learning_rate": 1.4849829990931653e-09,
+      "loss": 0.1221,
+      "step": 31650
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8936166763305664,
+      "learning_rate": 1.0910365767358155e-09,
+      "loss": 0.1218,
+      "step": 31700
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.029935598373413,
+      "learning_rate": 7.576811299714326e-10,
+      "loss": 0.1292,
+      "step": 31750
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8104945421218872,
+      "learning_rate": 4.849247416599534e-10,
+      "loss": 0.1189,
+      "step": 31800
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7662272453308105,
+      "learning_rate": 2.727740253177791e-10,
+      "loss": 0.1243,
+      "step": 31850
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8399808406829834,
+      "learning_rate": 1.2123412495762543e-10,
+      "loss": 0.1229,
+      "step": 31900
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7275273203849792,
+      "learning_rate": 3.0308714963067644e-11,
+      "loss": 0.1182,
+      "step": 31950
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9211012721061707,
+      "learning_rate": 0.0,
+      "loss": 0.1208,
+      "step": 32000
+    }
+  ],
+  "logging_steps": 50,
+  "max_steps": 32000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 3200,
+  "total_flos": 2.2216620291373056e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}