diff --git "a/checkpoint-99500/trainer_state.json" "b/checkpoint-99500/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-99500/trainer_state.json"
@@ -0,0 +1,13964 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.99,
+  "eval_steps": 500,
+  "global_step": 99500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.001,
+      "grad_norm": 2.5878946781158447,
+      "learning_rate": 4.99755e-05,
+      "loss": 9.5648,
+      "step": 50
+    },
+    {
+      "epoch": 0.002,
+      "grad_norm": 2.326439619064331,
+      "learning_rate": 4.99505e-05,
+      "loss": 8.7873,
+      "step": 100
+    },
+    {
+      "epoch": 0.003,
+      "grad_norm": 2.3391146659851074,
+      "learning_rate": 4.99255e-05,
+      "loss": 8.1288,
+      "step": 150
+    },
+    {
+      "epoch": 0.004,
+      "grad_norm": 2.398390531539917,
+      "learning_rate": 4.99005e-05,
+      "loss": 7.5247,
+      "step": 200
+    },
+    {
+      "epoch": 0.005,
+      "grad_norm": 2.4866814613342285,
+      "learning_rate": 4.98755e-05,
+      "loss": 7.0484,
+      "step": 250
+    },
+    {
+      "epoch": 0.006,
+      "grad_norm": 2.6657838821411133,
+      "learning_rate": 4.98505e-05,
+      "loss": 6.7936,
+      "step": 300
+    },
+    {
+      "epoch": 0.007,
+      "grad_norm": 2.2808291912078857,
+      "learning_rate": 4.9825500000000006e-05,
+      "loss": 6.5912,
+      "step": 350
+    },
+    {
+      "epoch": 0.008,
+      "grad_norm": 2.881570816040039,
+      "learning_rate": 4.98005e-05,
+      "loss": 6.5094,
+      "step": 400
+    },
+    {
+      "epoch": 0.009,
+      "grad_norm": 2.216550588607788,
+      "learning_rate": 4.9775500000000005e-05,
+      "loss": 6.4908,
+      "step": 450
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7933621406555176,
+      "learning_rate": 4.97505e-05,
+      "loss": 6.3012,
+      "step": 500
+    },
+    {
+      "epoch": 0.011,
+      "grad_norm": 2.3578381538391113,
+      "learning_rate": 4.9725500000000004e-05,
+      "loss": 6.2162,
+      "step": 550
+    },
+    {
+      "epoch": 0.012,
+      "grad_norm": 2.922057867050171,
+      "learning_rate": 4.97005e-05,
+      "loss": 6.1724,
+      "step": 600
+    },
+    {
+      "epoch": 0.013,
+      "grad_norm": 3.2069950103759766,
+      "learning_rate": 4.96755e-05,
+      "loss": 6.0117,
+      "step": 650
+    },
+    {
+      "epoch": 0.014,
+      "grad_norm": 2.830580234527588,
+      "learning_rate": 4.96505e-05,
+      "loss": 6.2267,
+      "step": 700
+    },
+    {
+      "epoch": 0.015,
+      "grad_norm": 3.050321102142334,
+      "learning_rate": 4.96255e-05,
+      "loss": 6.0236,
+      "step": 750
+    },
+    {
+      "epoch": 0.016,
+      "grad_norm": 3.1756136417388916,
+      "learning_rate": 4.96005e-05,
+      "loss": 6.0656,
+      "step": 800
+    },
+    {
+      "epoch": 0.017,
+      "grad_norm": 2.805255889892578,
+      "learning_rate": 4.95755e-05,
+      "loss": 6.0215,
+      "step": 850
+    },
+    {
+      "epoch": 0.018,
+      "grad_norm": 3.590785026550293,
+      "learning_rate": 4.95505e-05,
+      "loss": 6.0075,
+      "step": 900
+    },
+    {
+      "epoch": 0.019,
+      "grad_norm": 3.177901268005371,
+      "learning_rate": 4.9525500000000005e-05,
+      "loss": 5.8187,
+      "step": 950
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7860662937164307,
+      "learning_rate": 4.95005e-05,
+      "loss": 5.9114,
+      "step": 1000
+    },
+    {
+      "epoch": 0.021,
+      "grad_norm": 2.8016605377197266,
+      "learning_rate": 4.9475500000000004e-05,
+      "loss": 5.8214,
+      "step": 1050
+    },
+    {
+      "epoch": 0.022,
+      "grad_norm": 3.046123743057251,
+      "learning_rate": 4.94505e-05,
+      "loss": 5.9951,
+      "step": 1100
+    },
+    {
+      "epoch": 0.023,
+      "grad_norm": 3.1473944187164307,
+      "learning_rate": 4.94255e-05,
+      "loss": 5.8583,
+      "step": 1150
+    },
+    {
+      "epoch": 0.024,
+      "grad_norm": 3.6134166717529297,
+      "learning_rate": 4.94005e-05,
+      "loss": 5.8007,
+      "step": 1200
+    },
+    {
+      "epoch": 0.025,
+      "grad_norm": 3.606462001800537,
+      "learning_rate": 4.93755e-05,
+      "loss": 5.668,
+      "step": 1250
+    },
+    {
+      "epoch": 0.026,
+      "grad_norm": 3.616032838821411,
+      "learning_rate": 4.935050000000001e-05,
+      "loss": 5.7571,
+      "step": 1300
+    },
+    {
+      "epoch": 0.027,
+      "grad_norm": 5.09005069732666,
+      "learning_rate": 4.93255e-05,
+      "loss": 5.6571,
+      "step": 1350
+    },
+    {
+      "epoch": 0.028,
+      "grad_norm": 3.366626501083374,
+      "learning_rate": 4.930050000000001e-05,
+      "loss": 5.7309,
+      "step": 1400
+    },
+    {
+      "epoch": 0.029,
+      "grad_norm": 3.0361785888671875,
+      "learning_rate": 4.92755e-05,
+      "loss": 5.7198,
+      "step": 1450
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8341948986053467,
+      "learning_rate": 4.9250500000000006e-05,
+      "loss": 5.6426,
+      "step": 1500
+    },
+    {
+      "epoch": 0.031,
+      "grad_norm": 2.939432382583618,
+      "learning_rate": 4.9225500000000005e-05,
+      "loss": 5.5296,
+      "step": 1550
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 5.241086483001709,
+      "learning_rate": 4.9200500000000004e-05,
+      "loss": 5.5357,
+      "step": 1600
+    },
+    {
+      "epoch": 0.033,
+      "grad_norm": 4.518623352050781,
+      "learning_rate": 4.9175500000000004e-05,
+      "loss": 5.676,
+      "step": 1650
+    },
+    {
+      "epoch": 0.034,
+      "grad_norm": 2.8882062435150146,
+      "learning_rate": 4.91505e-05,
+      "loss": 5.6926,
+      "step": 1700
+    },
+    {
+      "epoch": 0.035,
+      "grad_norm": 3.0241808891296387,
+      "learning_rate": 4.91255e-05,
+      "loss": 5.5071,
+      "step": 1750
+    },
+    {
+      "epoch": 0.036,
+      "grad_norm": 3.798578977584839,
+      "learning_rate": 4.91005e-05,
+      "loss": 5.3687,
+      "step": 1800
+    },
+    {
+      "epoch": 0.037,
+      "grad_norm": 3.716325044631958,
+      "learning_rate": 4.90755e-05,
+      "loss": 5.5141,
+      "step": 1850
+    },
+    {
+      "epoch": 0.038,
+      "grad_norm": 3.375710964202881,
+      "learning_rate": 4.905050000000001e-05,
+      "loss": 5.5613,
+      "step": 1900
+    },
+    {
+      "epoch": 0.039,
+      "grad_norm": 3.86002516746521,
+      "learning_rate": 4.90255e-05,
+      "loss": 5.4685,
+      "step": 1950
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.336085796356201,
+      "learning_rate": 4.9000500000000006e-05,
+      "loss": 5.4872,
+      "step": 2000
+    },
+    {
+      "epoch": 0.041,
+      "grad_norm": 3.3040659427642822,
+      "learning_rate": 4.89755e-05,
+      "loss": 5.55,
+      "step": 2050
+    },
+    {
+      "epoch": 0.042,
+      "grad_norm": 4.6140313148498535,
+      "learning_rate": 4.8950500000000005e-05,
+      "loss": 5.4541,
+      "step": 2100
+    },
+    {
+      "epoch": 0.043,
+      "grad_norm": 3.5962140560150146,
+      "learning_rate": 4.8926e-05,
+      "loss": 5.503,
+      "step": 2150
+    },
+    {
+      "epoch": 0.044,
+      "grad_norm": 3.8882644176483154,
+      "learning_rate": 4.8901000000000005e-05,
+      "loss": 5.4016,
+      "step": 2200
+    },
+    {
+      "epoch": 0.045,
+      "grad_norm": 3.2127366065979004,
+      "learning_rate": 4.8876e-05,
+      "loss": 5.3505,
+      "step": 2250
+    },
+    {
+      "epoch": 0.046,
+      "grad_norm": 3.812075614929199,
+      "learning_rate": 4.8851000000000004e-05,
+      "loss": 5.308,
+      "step": 2300
+    },
+    {
+      "epoch": 0.047,
+      "grad_norm": 4.519869804382324,
+      "learning_rate": 4.8826e-05,
+      "loss": 5.426,
+      "step": 2350
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 3.7040762901306152,
+      "learning_rate": 4.8801e-05,
+      "loss": 5.3255,
+      "step": 2400
+    },
+    {
+      "epoch": 0.049,
+      "grad_norm": 3.347390651702881,
+      "learning_rate": 4.8776e-05,
+      "loss": 5.3165,
+      "step": 2450
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.786517858505249,
+      "learning_rate": 4.8751e-05,
+      "loss": 5.2374,
+      "step": 2500
+    },
+    {
+      "epoch": 0.051,
+      "grad_norm": 4.848645210266113,
+      "learning_rate": 4.872600000000001e-05,
+      "loss": 5.2179,
+      "step": 2550
+    },
+    {
+      "epoch": 0.052,
+      "grad_norm": 3.5423243045806885,
+      "learning_rate": 4.8701e-05,
+      "loss": 5.2099,
+      "step": 2600
+    },
+    {
+      "epoch": 0.053,
+      "grad_norm": 3.162454128265381,
+      "learning_rate": 4.8676000000000006e-05,
+      "loss": 5.2194,
+      "step": 2650
+    },
+    {
+      "epoch": 0.054,
+      "grad_norm": 4.749024868011475,
+      "learning_rate": 4.8651e-05,
+      "loss": 5.3041,
+      "step": 2700
+    },
+    {
+      "epoch": 0.055,
+      "grad_norm": 4.379371643066406,
+      "learning_rate": 4.8626000000000005e-05,
+      "loss": 5.3414,
+      "step": 2750
+    },
+    {
+      "epoch": 0.056,
+      "grad_norm": 4.969504356384277,
+      "learning_rate": 4.8601000000000004e-05,
+      "loss": 5.3592,
+      "step": 2800
+    },
+    {
+      "epoch": 0.057,
+      "grad_norm": 3.9636573791503906,
+      "learning_rate": 4.8576000000000004e-05,
+      "loss": 5.0791,
+      "step": 2850
+    },
+    {
+      "epoch": 0.058,
+      "grad_norm": 4.6978631019592285,
+      "learning_rate": 4.8551e-05,
+      "loss": 5.2218,
+      "step": 2900
+    },
+    {
+      "epoch": 0.059,
+      "grad_norm": 3.546217203140259,
+      "learning_rate": 4.8526e-05,
+      "loss": 5.1096,
+      "step": 2950
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.556810140609741,
+      "learning_rate": 4.8501e-05,
+      "loss": 5.2164,
+      "step": 3000
+    },
+    {
+      "epoch": 0.061,
+      "grad_norm": 4.784888744354248,
+      "learning_rate": 4.8476e-05,
+      "loss": 5.1909,
+      "step": 3050
+    },
+    {
+      "epoch": 0.062,
+      "grad_norm": 3.073183059692383,
+      "learning_rate": 4.8451e-05,
+      "loss": 5.0549,
+      "step": 3100
+    },
+    {
+      "epoch": 0.063,
+      "grad_norm": 4.279170513153076,
+      "learning_rate": 4.842600000000001e-05,
+      "loss": 5.2979,
+      "step": 3150
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 3.607466220855713,
+      "learning_rate": 4.8401e-05,
+      "loss": 4.9936,
+      "step": 3200
+    },
+    {
+      "epoch": 0.065,
+      "grad_norm": 4.333967685699463,
+      "learning_rate": 4.8376000000000006e-05,
+      "loss": 5.1944,
+      "step": 3250
+    },
+    {
+      "epoch": 0.066,
+      "grad_norm": 4.101257801055908,
+      "learning_rate": 4.8351000000000005e-05,
+      "loss": 5.1698,
+      "step": 3300
+    },
+    {
+      "epoch": 0.067,
+      "grad_norm": 5.113340377807617,
+      "learning_rate": 4.8326000000000005e-05,
+      "loss": 4.9471,
+      "step": 3350
+    },
+    {
+      "epoch": 0.068,
+      "grad_norm": 4.2900614738464355,
+      "learning_rate": 4.8301000000000004e-05,
+      "loss": 5.0118,
+      "step": 3400
+    },
+    {
+      "epoch": 0.069,
+      "grad_norm": 4.718325138092041,
+      "learning_rate": 4.8276e-05,
+      "loss": 4.983,
+      "step": 3450
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.2561187744140625,
+      "learning_rate": 4.8251e-05,
+      "loss": 5.0226,
+      "step": 3500
+    },
+    {
+      "epoch": 0.071,
+      "grad_norm": 4.311369895935059,
+      "learning_rate": 4.8226e-05,
+      "loss": 4.7971,
+      "step": 3550
+    },
+    {
+      "epoch": 0.072,
+      "grad_norm": 3.74359130859375,
+      "learning_rate": 4.8201e-05,
+      "loss": 4.8312,
+      "step": 3600
+    },
+    {
+      "epoch": 0.073,
+      "grad_norm": 4.252142429351807,
+      "learning_rate": 4.817600000000001e-05,
+      "loss": 4.9795,
+      "step": 3650
+    },
+    {
+      "epoch": 0.074,
+      "grad_norm": 6.2820024490356445,
+      "learning_rate": 4.8151e-05,
+      "loss": 5.2044,
+      "step": 3700
+    },
+    {
+      "epoch": 0.075,
+      "grad_norm": 4.392597198486328,
+      "learning_rate": 4.8126000000000006e-05,
+      "loss": 5.0104,
+      "step": 3750
+    },
+    {
+      "epoch": 0.076,
+      "grad_norm": 3.9632086753845215,
+      "learning_rate": 4.8101e-05,
+      "loss": 4.9215,
+      "step": 3800
+    },
+    {
+      "epoch": 0.077,
+      "grad_norm": 3.8012332916259766,
+      "learning_rate": 4.8076000000000005e-05,
+      "loss": 5.1204,
+      "step": 3850
+    },
+    {
+      "epoch": 0.078,
+      "grad_norm": 3.9559926986694336,
+      "learning_rate": 4.8051000000000005e-05,
+      "loss": 4.9711,
+      "step": 3900
+    },
+    {
+      "epoch": 0.079,
+      "grad_norm": 4.333497047424316,
+      "learning_rate": 4.8026000000000004e-05,
+      "loss": 4.9412,
+      "step": 3950
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.0923027992248535,
+      "learning_rate": 4.8001000000000003e-05,
+      "loss": 5.0554,
+      "step": 4000
+    },
+    {
+      "epoch": 0.081,
+      "grad_norm": 5.284310340881348,
+      "learning_rate": 4.7976e-05,
+      "loss": 5.008,
+      "step": 4050
+    },
+    {
+      "epoch": 0.082,
+      "grad_norm": 4.0718793869018555,
+      "learning_rate": 4.7951e-05,
+      "loss": 4.8831,
+      "step": 4100
+    },
+    {
+      "epoch": 0.083,
+      "grad_norm": 4.595408916473389,
+      "learning_rate": 4.7926e-05,
+      "loss": 4.8901,
+      "step": 4150
+    },
+    {
+      "epoch": 0.084,
+      "grad_norm": 4.26055383682251,
+      "learning_rate": 4.79015e-05,
+      "loss": 4.9658,
+      "step": 4200
+    },
+    {
+      "epoch": 0.085,
+      "grad_norm": 5.186002731323242,
+      "learning_rate": 4.78765e-05,
+      "loss": 4.8566,
+      "step": 4250
+    },
+    {
+      "epoch": 0.086,
+      "grad_norm": 4.462579250335693,
+      "learning_rate": 4.785150000000001e-05,
+      "loss": 4.7892,
+      "step": 4300
+    },
+    {
+      "epoch": 0.087,
+      "grad_norm": 5.473343849182129,
+      "learning_rate": 4.78265e-05,
+      "loss": 5.0057,
+      "step": 4350
+    },
+    {
+      "epoch": 0.088,
+      "grad_norm": 6.329896450042725,
+      "learning_rate": 4.7801500000000006e-05,
+      "loss": 4.841,
+      "step": 4400
+    },
+    {
+      "epoch": 0.089,
+      "grad_norm": 4.852380752563477,
+      "learning_rate": 4.77765e-05,
+      "loss": 4.7658,
+      "step": 4450
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.577455520629883,
+      "learning_rate": 4.7751500000000005e-05,
+      "loss": 4.9862,
+      "step": 4500
+    },
+    {
+      "epoch": 0.091,
+      "grad_norm": 4.643036365509033,
+      "learning_rate": 4.7726500000000005e-05,
+      "loss": 4.8218,
+      "step": 4550
+    },
+    {
+      "epoch": 0.092,
+      "grad_norm": 4.697981834411621,
+      "learning_rate": 4.7701500000000004e-05,
+      "loss": 4.8586,
+      "step": 4600
+    },
+    {
+      "epoch": 0.093,
+      "grad_norm": 4.388003826141357,
+      "learning_rate": 4.76765e-05,
+      "loss": 4.9039,
+      "step": 4650
+    },
+    {
+      "epoch": 0.094,
+      "grad_norm": 4.59213399887085,
+      "learning_rate": 4.76515e-05,
+      "loss": 4.8719,
+      "step": 4700
+    },
+    {
+      "epoch": 0.095,
+      "grad_norm": 3.707645893096924,
+      "learning_rate": 4.76265e-05,
+      "loss": 4.7647,
+      "step": 4750
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 4.055088520050049,
+      "learning_rate": 4.76015e-05,
+      "loss": 4.9086,
+      "step": 4800
+    },
+    {
+      "epoch": 0.097,
+      "grad_norm": 3.311262845993042,
+      "learning_rate": 4.75765e-05,
+      "loss": 4.8342,
+      "step": 4850
+    },
+    {
+      "epoch": 0.098,
+      "grad_norm": 4.882483005523682,
+      "learning_rate": 4.755150000000001e-05,
+      "loss": 4.8363,
+      "step": 4900
+    },
+    {
+      "epoch": 0.099,
+      "grad_norm": 4.083950042724609,
+      "learning_rate": 4.75265e-05,
+      "loss": 4.8207,
+      "step": 4950
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.274576187133789,
+      "learning_rate": 4.7501500000000006e-05,
+      "loss": 4.7845,
+      "step": 5000
+    },
+    {
+      "epoch": 0.101,
+      "grad_norm": 4.489608287811279,
+      "learning_rate": 4.74765e-05,
+      "loss": 4.705,
+      "step": 5050
+    },
+    {
+      "epoch": 0.102,
+      "grad_norm": 4.044312953948975,
+      "learning_rate": 4.7451500000000005e-05,
+      "loss": 4.6666,
+      "step": 5100
+    },
+    {
+      "epoch": 0.103,
+      "grad_norm": 4.765976905822754,
+      "learning_rate": 4.7426500000000004e-05,
+      "loss": 4.9063,
+      "step": 5150
+    },
+    {
+      "epoch": 0.104,
+      "grad_norm": 3.8264522552490234,
+      "learning_rate": 4.7401500000000003e-05,
+      "loss": 4.6931,
+      "step": 5200
+    },
+    {
+      "epoch": 0.105,
+      "grad_norm": 4.612431049346924,
+      "learning_rate": 4.73765e-05,
+      "loss": 4.7997,
+      "step": 5250
+    },
+    {
+      "epoch": 0.106,
+      "grad_norm": 3.871309280395508,
+      "learning_rate": 4.73515e-05,
+      "loss": 4.8181,
+      "step": 5300
+    },
+    {
+      "epoch": 0.107,
+      "grad_norm": 5.149219512939453,
+      "learning_rate": 4.73265e-05,
+      "loss": 4.7243,
+      "step": 5350
+    },
+    {
+      "epoch": 0.108,
+      "grad_norm": 5.467377185821533,
+      "learning_rate": 4.73015e-05,
+      "loss": 4.743,
+      "step": 5400
+    },
+    {
+      "epoch": 0.109,
+      "grad_norm": 4.169463157653809,
+      "learning_rate": 4.72765e-05,
+      "loss": 4.6821,
+      "step": 5450
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.368409633636475,
+      "learning_rate": 4.7251500000000007e-05,
+      "loss": 4.6557,
+      "step": 5500
+    },
+    {
+      "epoch": 0.111,
+      "grad_norm": 4.478771209716797,
+      "learning_rate": 4.72265e-05,
+      "loss": 4.7211,
+      "step": 5550
+    },
+    {
+      "epoch": 0.112,
+      "grad_norm": 4.097254753112793,
+      "learning_rate": 4.7201500000000005e-05,
+      "loss": 4.5969,
+      "step": 5600
+    },
+    {
+      "epoch": 0.113,
+      "grad_norm": 5.199203014373779,
+      "learning_rate": 4.71765e-05,
+      "loss": 4.6524,
+      "step": 5650
+    },
+    {
+      "epoch": 0.114,
+      "grad_norm": 4.438806056976318,
+      "learning_rate": 4.7151500000000004e-05,
+      "loss": 4.8195,
+      "step": 5700
+    },
+    {
+      "epoch": 0.115,
+      "grad_norm": 3.9345571994781494,
+      "learning_rate": 4.7126500000000004e-05,
+      "loss": 4.727,
+      "step": 5750
+    },
+    {
+      "epoch": 0.116,
+      "grad_norm": 4.215692520141602,
+      "learning_rate": 4.71015e-05,
+      "loss": 4.6921,
+      "step": 5800
+    },
+    {
+      "epoch": 0.117,
+      "grad_norm": 4.601874828338623,
+      "learning_rate": 4.70765e-05,
+      "loss": 4.6495,
+      "step": 5850
+    },
+    {
+      "epoch": 0.118,
+      "grad_norm": 4.633566856384277,
+      "learning_rate": 4.70515e-05,
+      "loss": 4.698,
+      "step": 5900
+    },
+    {
+      "epoch": 0.119,
+      "grad_norm": 4.730185031890869,
+      "learning_rate": 4.70265e-05,
+      "loss": 4.4599,
+      "step": 5950
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.261397838592529,
+      "learning_rate": 4.70015e-05,
+      "loss": 4.6781,
+      "step": 6000
+    },
+    {
+      "epoch": 0.121,
+      "grad_norm": 4.527892112731934,
+      "learning_rate": 4.69765e-05,
+      "loss": 4.6323,
+      "step": 6050
+    },
+    {
+      "epoch": 0.122,
+      "grad_norm": 4.5726237297058105,
+      "learning_rate": 4.6951500000000006e-05,
+      "loss": 4.6572,
+      "step": 6100
+    },
+    {
+      "epoch": 0.123,
+      "grad_norm": 6.083981037139893,
+      "learning_rate": 4.69265e-05,
+      "loss": 4.709,
+      "step": 6150
+    },
+    {
+      "epoch": 0.124,
+      "grad_norm": 6.871278762817383,
+      "learning_rate": 4.6901500000000005e-05,
+      "loss": 4.5572,
+      "step": 6200
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 5.033217430114746,
+      "learning_rate": 4.6876500000000004e-05,
+      "loss": 4.8283,
+      "step": 6250
+    },
+    {
+      "epoch": 0.126,
+      "grad_norm": 5.237866401672363,
+      "learning_rate": 4.6851500000000004e-05,
+      "loss": 4.6801,
+      "step": 6300
+    },
+    {
+      "epoch": 0.127,
+      "grad_norm": 4.246436595916748,
+      "learning_rate": 4.68265e-05,
+      "loss": 4.3712,
+      "step": 6350
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 4.287390232086182,
+      "learning_rate": 4.68015e-05,
+      "loss": 4.549,
+      "step": 6400
+    },
+    {
+      "epoch": 0.129,
+      "grad_norm": 5.145524978637695,
+      "learning_rate": 4.67765e-05,
+      "loss": 4.6266,
+      "step": 6450
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.119638919830322,
+      "learning_rate": 4.67515e-05,
+      "loss": 4.5634,
+      "step": 6500
+    },
+    {
+      "epoch": 0.131,
+      "grad_norm": 5.78672981262207,
+      "learning_rate": 4.67265e-05,
+      "loss": 4.402,
+      "step": 6550
+    },
+    {
+      "epoch": 0.132,
+      "grad_norm": 4.446723461151123,
+      "learning_rate": 4.670150000000001e-05,
+      "loss": 4.425,
+      "step": 6600
+    },
+    {
+      "epoch": 0.133,
+      "grad_norm": 6.311605453491211,
+      "learning_rate": 4.667700000000001e-05,
+      "loss": 4.6741,
+      "step": 6650
+    },
+    {
+      "epoch": 0.134,
+      "grad_norm": 4.3736772537231445,
+      "learning_rate": 4.6652e-05,
+      "loss": 4.532,
+      "step": 6700
+    },
+    {
+      "epoch": 0.135,
+      "grad_norm": 4.863807678222656,
+      "learning_rate": 4.6627000000000006e-05,
+      "loss": 4.3248,
+      "step": 6750
+    },
+    {
+      "epoch": 0.136,
+      "grad_norm": 4.457269668579102,
+      "learning_rate": 4.6602e-05,
+      "loss": 4.498,
+      "step": 6800
+    },
+    {
+      "epoch": 0.137,
+      "grad_norm": 5.396666526794434,
+      "learning_rate": 4.6577000000000005e-05,
+      "loss": 4.4462,
+      "step": 6850
+    },
+    {
+      "epoch": 0.138,
+      "grad_norm": 4.7764153480529785,
+      "learning_rate": 4.6552000000000004e-05,
+      "loss": 4.6494,
+      "step": 6900
+    },
+    {
+      "epoch": 0.139,
+      "grad_norm": 3.6046454906463623,
+      "learning_rate": 4.6527000000000004e-05,
+      "loss": 4.5615,
+      "step": 6950
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.655783653259277,
+      "learning_rate": 4.6502e-05,
+      "loss": 4.6513,
+      "step": 7000
+    },
+    {
+      "epoch": 0.141,
+      "grad_norm": 4.3254218101501465,
+      "learning_rate": 4.6477e-05,
+      "loss": 4.66,
+      "step": 7050
+    },
+    {
+      "epoch": 0.142,
+      "grad_norm": 4.6337785720825195,
+      "learning_rate": 4.6452e-05,
+      "loss": 4.6969,
+      "step": 7100
+    },
+    {
+      "epoch": 0.143,
+      "grad_norm": 4.757678985595703,
+      "learning_rate": 4.6427e-05,
+      "loss": 4.5222,
+      "step": 7150
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 5.13063907623291,
+      "learning_rate": 4.6402e-05,
+      "loss": 4.6098,
+      "step": 7200
+    },
+    {
+      "epoch": 0.145,
+      "grad_norm": 4.38551664352417,
+      "learning_rate": 4.637700000000001e-05,
+      "loss": 4.3308,
+      "step": 7250
+    },
+    {
+      "epoch": 0.146,
+      "grad_norm": 5.627738952636719,
+      "learning_rate": 4.6352e-05,
+      "loss": 4.3211,
+      "step": 7300
+    },
+    {
+      "epoch": 0.147,
+      "grad_norm": 5.648825168609619,
+      "learning_rate": 4.6327000000000006e-05,
+      "loss": 4.5699,
+      "step": 7350
+    },
+    {
+      "epoch": 0.148,
+      "grad_norm": 4.3480072021484375,
+      "learning_rate": 4.6302e-05,
+      "loss": 4.4976,
+      "step": 7400
+    },
+    {
+      "epoch": 0.149,
+      "grad_norm": 5.155121803283691,
+      "learning_rate": 4.6277000000000004e-05,
+      "loss": 4.5547,
+      "step": 7450
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.101211071014404,
+      "learning_rate": 4.6252000000000004e-05,
+      "loss": 4.5195,
+      "step": 7500
+    },
+    {
+      "epoch": 0.151,
+      "grad_norm": 4.691388130187988,
+      "learning_rate": 4.6227e-05,
+      "loss": 4.3268,
+      "step": 7550
+    },
+    {
+      "epoch": 0.152,
+      "grad_norm": 4.447563648223877,
+      "learning_rate": 4.6202e-05,
+      "loss": 4.4686,
+      "step": 7600
+    },
+    {
+      "epoch": 0.153,
+      "grad_norm": 5.302483081817627,
+      "learning_rate": 4.6177e-05,
+      "loss": 4.4724,
+      "step": 7650
+    },
+    {
+      "epoch": 0.154,
+      "grad_norm": 5.505932807922363,
+      "learning_rate": 4.6152e-05,
+      "loss": 4.5188,
+      "step": 7700
+    },
+    {
+      "epoch": 0.155,
+      "grad_norm": 4.236738681793213,
+      "learning_rate": 4.6127e-05,
+      "loss": 4.3862,
+      "step": 7750
+    },
+    {
+      "epoch": 0.156,
+      "grad_norm": 4.4891839027404785,
+      "learning_rate": 4.6102e-05,
+      "loss": 4.3585,
+      "step": 7800
+    },
+    {
+      "epoch": 0.157,
+      "grad_norm": 5.605619430541992,
+      "learning_rate": 4.6077000000000006e-05,
+      "loss": 4.4001,
+      "step": 7850
+    },
+    {
+      "epoch": 0.158,
+      "grad_norm": 5.551383018493652,
+      "learning_rate": 4.6052e-05,
+      "loss": 4.6582,
+      "step": 7900
+    },
+    {
+      "epoch": 0.159,
+      "grad_norm": 6.318504810333252,
+      "learning_rate": 4.6027000000000005e-05,
+      "loss": 4.4838,
+      "step": 7950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.358963489532471,
+      "learning_rate": 4.6002e-05,
+      "loss": 4.4965,
+      "step": 8000
+    },
+    {
+      "epoch": 0.161,
+      "grad_norm": 4.824229717254639,
+      "learning_rate": 4.5977000000000004e-05,
+      "loss": 4.4741,
+      "step": 8050
+    },
+    {
+      "epoch": 0.162,
+      "grad_norm": 4.499610424041748,
+      "learning_rate": 4.5952e-05,
+      "loss": 4.4009,
+      "step": 8100
+    },
+    {
+      "epoch": 0.163,
+      "grad_norm": 4.207203388214111,
+      "learning_rate": 4.5927e-05,
+      "loss": 4.5216,
+      "step": 8150
+    },
+    {
+      "epoch": 0.164,
+      "grad_norm": 5.642760276794434,
+      "learning_rate": 4.5902e-05,
+      "loss": 4.3556,
+      "step": 8200
+    },
+    {
+      "epoch": 0.165,
+      "grad_norm": 4.778339385986328,
+      "learning_rate": 4.5877e-05,
+      "loss": 4.4773,
+      "step": 8250
+    },
+    {
+      "epoch": 0.166,
+      "grad_norm": 5.704320907592773,
+      "learning_rate": 4.5852e-05,
+      "loss": 4.2456,
+      "step": 8300
+    },
+    {
+      "epoch": 0.167,
+      "grad_norm": 4.367884635925293,
+      "learning_rate": 4.5827e-05,
+      "loss": 4.4405,
+      "step": 8350
+    },
+    {
+      "epoch": 0.168,
+      "grad_norm": 4.443906784057617,
+      "learning_rate": 4.5802e-05,
+      "loss": 4.4384,
+      "step": 8400
+    },
+    {
+      "epoch": 0.169,
+      "grad_norm": 4.177926540374756,
+      "learning_rate": 4.5777000000000006e-05,
+      "loss": 4.4162,
+      "step": 8450
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.305122375488281,
+      "learning_rate": 4.5752e-05,
+      "loss": 4.4163,
+      "step": 8500
+    },
+    {
+      "epoch": 0.171,
+      "grad_norm": 4.545613765716553,
+      "learning_rate": 4.5727000000000005e-05,
+      "loss": 4.4573,
+      "step": 8550
+    },
+    {
+      "epoch": 0.172,
+      "grad_norm": 5.365312099456787,
+      "learning_rate": 4.5702e-05,
+      "loss": 4.4307,
+      "step": 8600
+    },
+    {
+      "epoch": 0.173,
+      "grad_norm": 5.722323417663574,
+      "learning_rate": 4.5677e-05,
+      "loss": 4.5031,
+      "step": 8650
+    },
+    {
+      "epoch": 0.174,
+      "grad_norm": 4.399505615234375,
+      "learning_rate": 4.5652e-05,
+      "loss": 4.0916,
+      "step": 8700
+    },
+    {
+      "epoch": 0.175,
+      "grad_norm": 3.8715851306915283,
+      "learning_rate": 4.5627e-05,
+      "loss": 4.1975,
+      "step": 8750
+    },
+    {
+      "epoch": 0.176,
+      "grad_norm": 4.734847068786621,
+      "learning_rate": 4.5602e-05,
+      "loss": 4.3628,
+      "step": 8800
+    },
+    {
+      "epoch": 0.177,
+      "grad_norm": 5.130662441253662,
+      "learning_rate": 4.55775e-05,
+      "loss": 4.3021,
+      "step": 8850
+    },
+    {
+      "epoch": 0.178,
+      "grad_norm": 7.081720352172852,
+      "learning_rate": 4.55525e-05,
+      "loss": 4.5149,
+      "step": 8900
+    },
+    {
+      "epoch": 0.179,
+      "grad_norm": 4.4592108726501465,
+      "learning_rate": 4.55275e-05,
+      "loss": 4.3799,
+      "step": 8950
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.49852991104126,
+      "learning_rate": 4.55025e-05,
+      "loss": 4.3414,
+      "step": 9000
+    },
+    {
+      "epoch": 0.181,
+      "grad_norm": 4.531373023986816,
+      "learning_rate": 4.54775e-05,
+      "loss": 4.6448,
+      "step": 9050
+    },
+    {
+      "epoch": 0.182,
+      "grad_norm": 5.069887638092041,
+      "learning_rate": 4.5452500000000006e-05,
+      "loss": 4.4109,
+      "step": 9100
+    },
+    {
+      "epoch": 0.183,
+      "grad_norm": 6.900569438934326,
+      "learning_rate": 4.54275e-05,
+      "loss": 4.1422,
+      "step": 9150
+    },
+    {
+      "epoch": 0.184,
+      "grad_norm": 5.383849620819092,
+      "learning_rate": 4.5402500000000004e-05,
+      "loss": 4.1877,
+      "step": 9200
+    },
+    {
+      "epoch": 0.185,
+      "grad_norm": 4.181257247924805,
+      "learning_rate": 4.5377500000000004e-05,
+      "loss": 4.2277,
+      "step": 9250
+    },
+    {
+      "epoch": 0.186,
+      "grad_norm": 5.585279941558838,
+      "learning_rate": 4.53525e-05,
+      "loss": 4.3192,
+      "step": 9300
+    },
+    {
+      "epoch": 0.187,
+      "grad_norm": 4.499333381652832,
+      "learning_rate": 4.53275e-05,
+      "loss": 4.4794,
+      "step": 9350
+    },
+    {
+      "epoch": 0.188,
+      "grad_norm": 4.638943672180176,
+      "learning_rate": 4.53025e-05,
+      "loss": 4.342,
+      "step": 9400
+    },
+    {
+      "epoch": 0.189,
+      "grad_norm": 4.7845330238342285,
+      "learning_rate": 4.52775e-05,
+      "loss": 4.2032,
+      "step": 9450
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.759675979614258,
+      "learning_rate": 4.52525e-05,
+      "loss": 4.2937,
+      "step": 9500
+    },
+    {
+      "epoch": 0.191,
+      "grad_norm": 5.508966445922852,
+      "learning_rate": 4.52275e-05,
+      "loss": 4.2401,
+      "step": 9550
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 4.584382057189941,
+      "learning_rate": 4.5202500000000006e-05,
+      "loss": 4.3531,
+      "step": 9600
+    },
+    {
+      "epoch": 0.193,
+      "grad_norm": 6.405841827392578,
+      "learning_rate": 4.51775e-05,
+      "loss": 4.3412,
+      "step": 9650
+    },
+    {
+      "epoch": 0.194,
+      "grad_norm": 4.769068241119385,
+      "learning_rate": 4.5152500000000005e-05,
+      "loss": 4.2218,
+      "step": 9700
+    },
+    {
+      "epoch": 0.195,
+      "grad_norm": 5.489443778991699,
+      "learning_rate": 4.51275e-05,
+      "loss": 4.1926,
+      "step": 9750
+    },
+    {
+      "epoch": 0.196,
+      "grad_norm": 4.364252090454102,
+      "learning_rate": 4.5102500000000004e-05,
+      "loss": 4.2831,
+      "step": 9800
+    },
+    {
+      "epoch": 0.197,
+      "grad_norm": 5.687471389770508,
+      "learning_rate": 4.50775e-05,
+      "loss": 4.1691,
+      "step": 9850
+    },
+    {
+      "epoch": 0.198,
+      "grad_norm": 4.656460762023926,
+      "learning_rate": 4.50525e-05,
+      "loss": 4.376,
+      "step": 9900
+    },
+    {
+      "epoch": 0.199,
+      "grad_norm": 6.34237813949585,
+      "learning_rate": 4.50275e-05,
+      "loss": 4.2811,
+      "step": 9950
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.788524150848389,
+      "learning_rate": 4.50025e-05,
+      "loss": 4.3582,
+      "step": 10000
+    },
+    {
+      "epoch": 0.201,
+      "grad_norm": 4.443533420562744,
+      "learning_rate": 4.49775e-05,
+      "loss": 4.201,
+      "step": 10050
+    },
+    {
+      "epoch": 0.202,
+      "grad_norm": 5.163551330566406,
+      "learning_rate": 4.49525e-05,
+      "loss": 4.3056,
+      "step": 10100
+    },
+    {
+      "epoch": 0.203,
+      "grad_norm": 4.789438247680664,
+      "learning_rate": 4.49275e-05,
+      "loss": 4.1726,
+      "step": 10150
+    },
+    {
+      "epoch": 0.204,
+      "grad_norm": 4.787554740905762,
+      "learning_rate": 4.4902500000000006e-05,
+      "loss": 4.2856,
+      "step": 10200
+    },
+    {
+      "epoch": 0.205,
+      "grad_norm": 5.560729503631592,
+      "learning_rate": 4.48775e-05,
+      "loss": 4.1345,
+      "step": 10250
+    },
+    {
+      "epoch": 0.206,
+      "grad_norm": 4.729394435882568,
+      "learning_rate": 4.4852500000000005e-05,
+      "loss": 4.1033,
+      "step": 10300
+    },
+    {
+      "epoch": 0.207,
+      "grad_norm": 6.309693336486816,
+      "learning_rate": 4.48275e-05,
+      "loss": 4.2375,
+      "step": 10350
+    },
+    {
+      "epoch": 0.208,
+      "grad_norm": 5.436013698577881,
+      "learning_rate": 4.4802500000000003e-05,
+      "loss": 4.0883,
+      "step": 10400
+    },
+    {
+      "epoch": 0.209,
+      "grad_norm": 5.470222473144531,
+      "learning_rate": 4.47775e-05,
+      "loss": 4.1247,
+      "step": 10450
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.284532070159912,
+      "learning_rate": 4.47525e-05,
+      "loss": 4.2951,
+      "step": 10500
+    },
+    {
+      "epoch": 0.211,
+      "grad_norm": 5.210220813751221,
+      "learning_rate": 4.47275e-05,
+      "loss": 4.1732,
+      "step": 10550
+    },
+    {
+      "epoch": 0.212,
+      "grad_norm": 4.622988224029541,
+      "learning_rate": 4.47025e-05,
+      "loss": 4.3512,
+      "step": 10600
+    },
+    {
+      "epoch": 0.213,
+      "grad_norm": 4.809802055358887,
+      "learning_rate": 4.46775e-05,
+      "loss": 4.206,
+      "step": 10650
+    },
+    {
+      "epoch": 0.214,
+      "grad_norm": 5.238707542419434,
+      "learning_rate": 4.46525e-05,
+      "loss": 4.146,
+      "step": 10700
+    },
+    {
+      "epoch": 0.215,
+      "grad_norm": 5.506088733673096,
+      "learning_rate": 4.46275e-05,
+      "loss": 4.1784,
+      "step": 10750
+    },
+    {
+      "epoch": 0.216,
+      "grad_norm": 5.379574775695801,
+      "learning_rate": 4.4602500000000005e-05,
+      "loss": 4.1822,
+      "step": 10800
+    },
+    {
+      "epoch": 0.217,
+      "grad_norm": 6.162681579589844,
+      "learning_rate": 4.4578000000000006e-05,
+      "loss": 4.221,
+      "step": 10850
+    },
+    {
+      "epoch": 0.218,
+      "grad_norm": 4.8035759925842285,
+      "learning_rate": 4.4553e-05,
+      "loss": 4.1525,
+      "step": 10900
+    },
+    {
+      "epoch": 0.219,
+      "grad_norm": 4.488569736480713,
+      "learning_rate": 4.4528000000000005e-05,
+      "loss": 4.092,
+      "step": 10950
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.557890892028809,
+      "learning_rate": 4.4503e-05,
+      "loss": 4.2091,
+      "step": 11000
+    },
+    {
+      "epoch": 0.221,
+      "grad_norm": 5.737027168273926,
+      "learning_rate": 4.4478000000000003e-05,
+      "loss": 4.2103,
+      "step": 11050
+    },
+    {
+      "epoch": 0.222,
+      "grad_norm": 5.317434787750244,
+      "learning_rate": 4.4453e-05,
+      "loss": 4.1134,
+      "step": 11100
+    },
+    {
+      "epoch": 0.223,
+      "grad_norm": 5.508534908294678,
+      "learning_rate": 4.4428e-05,
+      "loss": 4.2158,
+      "step": 11150
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 5.144506931304932,
+      "learning_rate": 4.4403e-05,
+      "loss": 4.0672,
+      "step": 11200
+    },
+    {
+      "epoch": 0.225,
+      "grad_norm": 5.047402858734131,
+      "learning_rate": 4.4378e-05,
+      "loss": 4.1603,
+      "step": 11250
+    },
+    {
+      "epoch": 0.226,
+      "grad_norm": 6.11688232421875,
+      "learning_rate": 4.4353e-05,
+      "loss": 4.1523,
+      "step": 11300
+    },
+    {
+      "epoch": 0.227,
+      "grad_norm": 4.963644981384277,
+      "learning_rate": 4.4328e-05,
+      "loss": 4.1287,
+      "step": 11350
+    },
+    {
+      "epoch": 0.228,
+      "grad_norm": 6.210315704345703,
+      "learning_rate": 4.4303e-05,
+      "loss": 4.0335,
+      "step": 11400
+    },
+    {
+      "epoch": 0.229,
+      "grad_norm": 4.604816913604736,
+      "learning_rate": 4.4278000000000005e-05,
+      "loss": 4.151,
+      "step": 11450
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.679665565490723,
+      "learning_rate": 4.4253e-05,
+      "loss": 4.0733,
+      "step": 11500
+    },
+    {
+      "epoch": 0.231,
+      "grad_norm": 5.566308975219727,
+      "learning_rate": 4.4228000000000004e-05,
+      "loss": 4.2364,
+      "step": 11550
+    },
+    {
+      "epoch": 0.232,
+      "grad_norm": 6.232782363891602,
+      "learning_rate": 4.4203e-05,
+      "loss": 4.1428,
+      "step": 11600
+    },
+    {
+      "epoch": 0.233,
+      "grad_norm": 5.726539611816406,
+      "learning_rate": 4.4178e-05,
+      "loss": 4.243,
+      "step": 11650
+    },
+    {
+      "epoch": 0.234,
+      "grad_norm": 7.133997440338135,
+      "learning_rate": 4.4153e-05,
+      "loss": 4.2425,
+      "step": 11700
+    },
+    {
+      "epoch": 0.235,
+      "grad_norm": 5.223475933074951,
+      "learning_rate": 4.4128e-05,
+      "loss": 3.8897,
+      "step": 11750
+    },
+    {
+      "epoch": 0.236,
+      "grad_norm": 5.071702480316162,
+      "learning_rate": 4.4103e-05,
+      "loss": 4.1501,
+      "step": 11800
+    },
+    {
+      "epoch": 0.237,
+      "grad_norm": 6.633401870727539,
+      "learning_rate": 4.4078e-05,
+      "loss": 4.2129,
+      "step": 11850
+    },
+    {
+      "epoch": 0.238,
+      "grad_norm": 5.5083465576171875,
+      "learning_rate": 4.4053e-05,
+      "loss": 4.0511,
+      "step": 11900
+    },
+    {
+      "epoch": 0.239,
+      "grad_norm": 5.211723327636719,
+      "learning_rate": 4.4028e-05,
+      "loss": 4.2103,
+      "step": 11950
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.270011901855469,
+      "learning_rate": 4.4003e-05,
+      "loss": 3.9425,
+      "step": 12000
+    },
+    {
+      "epoch": 0.241,
+      "grad_norm": 6.336957931518555,
+      "learning_rate": 4.3978000000000005e-05,
+      "loss": 4.1376,
+      "step": 12050
+    },
+    {
+      "epoch": 0.242,
+      "grad_norm": 6.871969223022461,
+      "learning_rate": 4.3953e-05,
+      "loss": 3.9205,
+      "step": 12100
+    },
+    {
+      "epoch": 0.243,
+      "grad_norm": 6.551147937774658,
+      "learning_rate": 4.3928000000000004e-05,
+      "loss": 4.0799,
+      "step": 12150
+    },
+    {
+      "epoch": 0.244,
+      "grad_norm": 5.746366500854492,
+      "learning_rate": 4.3903e-05,
+      "loss": 4.0542,
+      "step": 12200
+    },
+    {
+      "epoch": 0.245,
+      "grad_norm": 5.873757362365723,
+      "learning_rate": 4.3878e-05,
+      "loss": 4.1353,
+      "step": 12250
+    },
+    {
+      "epoch": 0.246,
+      "grad_norm": 5.338042736053467,
+      "learning_rate": 4.3853e-05,
+      "loss": 3.8595,
+      "step": 12300
+    },
+    {
+      "epoch": 0.247,
+      "grad_norm": 4.573357105255127,
+      "learning_rate": 4.3828e-05,
+      "loss": 4.2164,
+      "step": 12350
+    },
+    {
+      "epoch": 0.248,
+      "grad_norm": 6.519650459289551,
+      "learning_rate": 4.3803e-05,
+      "loss": 4.0502,
+      "step": 12400
+    },
+    {
+      "epoch": 0.249,
+      "grad_norm": 5.267230987548828,
+      "learning_rate": 4.3778e-05,
+      "loss": 4.1534,
+      "step": 12450
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.070836544036865,
+      "learning_rate": 4.3753e-05,
+      "loss": 3.7346,
+      "step": 12500
+    },
+    {
+      "epoch": 0.251,
+      "grad_norm": 4.947108745574951,
+      "learning_rate": 4.3728000000000006e-05,
+      "loss": 3.976,
+      "step": 12550
+    },
+    {
+      "epoch": 0.252,
+      "grad_norm": 5.57907247543335,
+      "learning_rate": 4.3703e-05,
+      "loss": 3.925,
+      "step": 12600
+    },
+    {
+      "epoch": 0.253,
+      "grad_norm": 5.960052490234375,
+      "learning_rate": 4.3678000000000004e-05,
+      "loss": 4.1336,
+      "step": 12650
+    },
+    {
+      "epoch": 0.254,
+      "grad_norm": 5.7773756980896,
+      "learning_rate": 4.3653e-05,
+      "loss": 4.055,
+      "step": 12700
+    },
+    {
+      "epoch": 0.255,
+      "grad_norm": 5.2892584800720215,
+      "learning_rate": 4.3628e-05,
+      "loss": 4.0579,
+      "step": 12750
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 5.7825798988342285,
+      "learning_rate": 4.3603e-05,
+      "loss": 4.0154,
+      "step": 12800
+    },
+    {
+      "epoch": 0.257,
+      "grad_norm": 4.2211432456970215,
+      "learning_rate": 4.35785e-05,
+      "loss": 4.1892,
+      "step": 12850
+    },
+    {
+      "epoch": 0.258,
+      "grad_norm": 5.150039196014404,
+      "learning_rate": 4.35535e-05,
+      "loss": 4.1491,
+      "step": 12900
+    },
+    {
+      "epoch": 0.259,
+      "grad_norm": 4.940394401550293,
+      "learning_rate": 4.35285e-05,
+      "loss": 4.1802,
+      "step": 12950
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.550903797149658,
+      "learning_rate": 4.35035e-05,
+      "loss": 4.1659,
+      "step": 13000
+    },
+    {
+      "epoch": 0.261,
+      "grad_norm": 5.111509799957275,
+      "learning_rate": 4.34785e-05,
+      "loss": 3.9146,
+      "step": 13050
+    },
+    {
+      "epoch": 0.262,
+      "grad_norm": 5.713197231292725,
+      "learning_rate": 4.34535e-05,
+      "loss": 4.0872,
+      "step": 13100
+    },
+    {
+      "epoch": 0.263,
+      "grad_norm": 4.943962574005127,
+      "learning_rate": 4.34285e-05,
+      "loss": 3.8049,
+      "step": 13150
+    },
+    {
+      "epoch": 0.264,
+      "grad_norm": 5.962152481079102,
+      "learning_rate": 4.3403500000000005e-05,
+      "loss": 3.8484,
+      "step": 13200
+    },
+    {
+      "epoch": 0.265,
+      "grad_norm": 5.06650972366333,
+      "learning_rate": 4.33785e-05,
+      "loss": 3.9027,
+      "step": 13250
+    },
+    {
+      "epoch": 0.266,
+      "grad_norm": 5.804838180541992,
+      "learning_rate": 4.3353500000000004e-05,
+      "loss": 3.8695,
+      "step": 13300
+    },
+    {
+      "epoch": 0.267,
+      "grad_norm": 4.750165939331055,
+      "learning_rate": 4.33285e-05,
+      "loss": 3.9193,
+      "step": 13350
+    },
+    {
+      "epoch": 0.268,
+      "grad_norm": 5.648144721984863,
+      "learning_rate": 4.33035e-05,
+      "loss": 4.1735,
+      "step": 13400
+    },
+    {
+      "epoch": 0.269,
+      "grad_norm": 5.493870735168457,
+      "learning_rate": 4.32785e-05,
+      "loss": 3.9941,
+      "step": 13450
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.866032123565674,
+      "learning_rate": 4.32535e-05,
+      "loss": 3.8839,
+      "step": 13500
+    },
+    {
+      "epoch": 0.271,
+      "grad_norm": 5.975167274475098,
+      "learning_rate": 4.32285e-05,
+      "loss": 4.1467,
+      "step": 13550
+    },
+    {
+      "epoch": 0.272,
+      "grad_norm": 4.195262432098389,
+      "learning_rate": 4.32035e-05,
+      "loss": 3.8118,
+      "step": 13600
+    },
+    {
+      "epoch": 0.273,
+      "grad_norm": 5.3393378257751465,
+      "learning_rate": 4.31785e-05,
+      "loss": 3.9583,
+      "step": 13650
+    },
+    {
+      "epoch": 0.274,
+      "grad_norm": 5.666850566864014,
+      "learning_rate": 4.31535e-05,
+      "loss": 3.8644,
+      "step": 13700
+    },
+    {
+      "epoch": 0.275,
+      "grad_norm": 5.188211917877197,
+      "learning_rate": 4.31285e-05,
+      "loss": 3.9014,
+      "step": 13750
+    },
+    {
+      "epoch": 0.276,
+      "grad_norm": 8.021430969238281,
+      "learning_rate": 4.3103500000000005e-05,
+      "loss": 3.9939,
+      "step": 13800
+    },
+    {
+      "epoch": 0.277,
+      "grad_norm": 6.547386169433594,
+      "learning_rate": 4.30785e-05,
+      "loss": 3.9064,
+      "step": 13850
+    },
+    {
+      "epoch": 0.278,
+      "grad_norm": 6.097850799560547,
+      "learning_rate": 4.3053500000000004e-05,
+      "loss": 3.9909,
+      "step": 13900
+    },
+    {
+      "epoch": 0.279,
+      "grad_norm": 5.55470085144043,
+      "learning_rate": 4.3028499999999996e-05,
+      "loss": 4.0055,
+      "step": 13950
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.464843273162842,
+      "learning_rate": 4.30035e-05,
+      "loss": 3.9649,
+      "step": 14000
+    },
+    {
+      "epoch": 0.281,
+      "grad_norm": 5.489349365234375,
+      "learning_rate": 4.29785e-05,
+      "loss": 4.1088,
+      "step": 14050
+    },
+    {
+      "epoch": 0.282,
+      "grad_norm": 5.301501274108887,
+      "learning_rate": 4.29535e-05,
+      "loss": 3.9677,
+      "step": 14100
+    },
+    {
+      "epoch": 0.283,
+      "grad_norm": 6.240916728973389,
+      "learning_rate": 4.292850000000001e-05,
+      "loss": 3.8114,
+      "step": 14150
+    },
+    {
+      "epoch": 0.284,
+      "grad_norm": 5.443836212158203,
+      "learning_rate": 4.29035e-05,
+      "loss": 4.0816,
+      "step": 14200
+    },
+    {
+      "epoch": 0.285,
+      "grad_norm": 5.28709602355957,
+      "learning_rate": 4.2878500000000006e-05,
+      "loss": 3.829,
+      "step": 14250
+    },
+    {
+      "epoch": 0.286,
+      "grad_norm": 4.40752649307251,
+      "learning_rate": 4.28535e-05,
+      "loss": 3.8467,
+      "step": 14300
+    },
+    {
+      "epoch": 0.287,
+      "grad_norm": 6.202967643737793,
+      "learning_rate": 4.2828500000000005e-05,
+      "loss": 3.9803,
+      "step": 14350
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 5.116330623626709,
+      "learning_rate": 4.2803500000000004e-05,
+      "loss": 3.8783,
+      "step": 14400
+    },
+    {
+      "epoch": 0.289,
+      "grad_norm": 5.7756781578063965,
+      "learning_rate": 4.2778500000000004e-05,
+      "loss": 3.6101,
+      "step": 14450
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.229415416717529,
+      "learning_rate": 4.27535e-05,
+      "loss": 4.0489,
+      "step": 14500
+    },
+    {
+      "epoch": 0.291,
+      "grad_norm": 6.183515548706055,
+      "learning_rate": 4.27285e-05,
+      "loss": 3.8464,
+      "step": 14550
+    },
+    {
+      "epoch": 0.292,
+      "grad_norm": 5.619703769683838,
+      "learning_rate": 4.27035e-05,
+      "loss": 4.0055,
+      "step": 14600
+    },
+    {
+      "epoch": 0.293,
+      "grad_norm": 5.139617919921875,
+      "learning_rate": 4.26785e-05,
+      "loss": 4.0818,
+      "step": 14650
+    },
+    {
+      "epoch": 0.294,
+      "grad_norm": 5.552958011627197,
+      "learning_rate": 4.26535e-05,
+      "loss": 3.9061,
+      "step": 14700
+    },
+    {
+      "epoch": 0.295,
+      "grad_norm": 6.033257007598877,
+      "learning_rate": 4.262850000000001e-05,
+      "loss": 3.9266,
+      "step": 14750
+    },
+    {
+      "epoch": 0.296,
+      "grad_norm": 5.819438934326172,
+      "learning_rate": 4.26035e-05,
+      "loss": 3.9946,
+      "step": 14800
+    },
+    {
+      "epoch": 0.297,
+      "grad_norm": 5.418524742126465,
+      "learning_rate": 4.2578500000000006e-05,
+      "loss": 3.6531,
+      "step": 14850
+    },
+    {
+      "epoch": 0.298,
+      "grad_norm": 6.2832560539245605,
+      "learning_rate": 4.2554e-05,
+      "loss": 3.9278,
+      "step": 14900
+    },
+    {
+      "epoch": 0.299,
+      "grad_norm": 7.28920841217041,
+      "learning_rate": 4.2529e-05,
+      "loss": 3.8046,
+      "step": 14950
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.926489353179932,
+      "learning_rate": 4.2504e-05,
+      "loss": 3.8063,
+      "step": 15000
+    },
+    {
+      "epoch": 0.301,
+      "grad_norm": 5.459051609039307,
+      "learning_rate": 4.2479000000000004e-05,
+      "loss": 3.8941,
+      "step": 15050
+    },
+    {
+      "epoch": 0.302,
+      "grad_norm": 5.072142124176025,
+      "learning_rate": 4.2454e-05,
+      "loss": 3.8465,
+      "step": 15100
+    },
+    {
+      "epoch": 0.303,
+      "grad_norm": 5.166111469268799,
+      "learning_rate": 4.2429e-05,
+      "loss": 4.0484,
+      "step": 15150
+    },
+    {
+      "epoch": 0.304,
+      "grad_norm": 5.6650471687316895,
+      "learning_rate": 4.2404e-05,
+      "loss": 3.7839,
+      "step": 15200
+    },
+    {
+      "epoch": 0.305,
+      "grad_norm": 6.4663801193237305,
+      "learning_rate": 4.2379e-05,
+      "loss": 3.7774,
+      "step": 15250
+    },
+    {
+      "epoch": 0.306,
+      "grad_norm": 5.610959529876709,
+      "learning_rate": 4.2354e-05,
+      "loss": 3.9454,
+      "step": 15300
+    },
+    {
+      "epoch": 0.307,
+      "grad_norm": 4.7916259765625,
+      "learning_rate": 4.2329e-05,
+      "loss": 3.8588,
+      "step": 15350
+    },
+    {
+      "epoch": 0.308,
+      "grad_norm": 4.65864372253418,
+      "learning_rate": 4.230400000000001e-05,
+      "loss": 4.0363,
+      "step": 15400
+    },
+    {
+      "epoch": 0.309,
+      "grad_norm": 5.171529293060303,
+      "learning_rate": 4.2279e-05,
+      "loss": 3.7952,
+      "step": 15450
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.2722578048706055,
+      "learning_rate": 4.2254000000000006e-05,
+      "loss": 3.8943,
+      "step": 15500
+    },
+    {
+      "epoch": 0.311,
+      "grad_norm": 6.563569068908691,
+      "learning_rate": 4.2229000000000005e-05,
+      "loss": 4.1681,
+      "step": 15550
+    },
+    {
+      "epoch": 0.312,
+      "grad_norm": 4.982414245605469,
+      "learning_rate": 4.2204000000000005e-05,
+      "loss": 3.648,
+      "step": 15600
+    },
+    {
+      "epoch": 0.313,
+      "grad_norm": 5.668543338775635,
+      "learning_rate": 4.2179000000000004e-05,
+      "loss": 3.9033,
+      "step": 15650
+    },
+    {
+      "epoch": 0.314,
+      "grad_norm": 5.404988765716553,
+      "learning_rate": 4.2154e-05,
+      "loss": 3.6414,
+      "step": 15700
+    },
+    {
+      "epoch": 0.315,
+      "grad_norm": 5.084871292114258,
+      "learning_rate": 4.2129e-05,
+      "loss": 3.88,
+      "step": 15750
+    },
+    {
+      "epoch": 0.316,
+      "grad_norm": 5.059499263763428,
+      "learning_rate": 4.2104e-05,
+      "loss": 3.8755,
+      "step": 15800
+    },
+    {
+      "epoch": 0.317,
+      "grad_norm": 5.1128010749816895,
+      "learning_rate": 4.2079e-05,
+      "loss": 3.946,
+      "step": 15850
+    },
+    {
+      "epoch": 0.318,
+      "grad_norm": 5.611748218536377,
+      "learning_rate": 4.205400000000001e-05,
+      "loss": 3.8578,
+      "step": 15900
+    },
+    {
+      "epoch": 0.319,
+      "grad_norm": 6.221893787384033,
+      "learning_rate": 4.2029e-05,
+      "loss": 4.0508,
+      "step": 15950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.107609272003174,
+      "learning_rate": 4.2004000000000006e-05,
+      "loss": 3.8917,
+      "step": 16000
+    },
+    {
+      "epoch": 0.321,
+      "grad_norm": 5.290852069854736,
+      "learning_rate": 4.1979e-05,
+      "loss": 3.8707,
+      "step": 16050
+    },
+    {
+      "epoch": 0.322,
+      "grad_norm": 6.895267486572266,
+      "learning_rate": 4.1954000000000005e-05,
+      "loss": 4.0213,
+      "step": 16100
+    },
+    {
+      "epoch": 0.323,
+      "grad_norm": 4.469918727874756,
+      "learning_rate": 4.1929000000000005e-05,
+      "loss": 3.7341,
+      "step": 16150
+    },
+    {
+      "epoch": 0.324,
+      "grad_norm": 8.303849220275879,
+      "learning_rate": 4.1904000000000004e-05,
+      "loss": 3.6815,
+      "step": 16200
+    },
+    {
+      "epoch": 0.325,
+      "grad_norm": 5.290466785430908,
+      "learning_rate": 4.1879000000000003e-05,
+      "loss": 3.8443,
+      "step": 16250
+    },
+    {
+      "epoch": 0.326,
+      "grad_norm": 6.372802734375,
+      "learning_rate": 4.1854e-05,
+      "loss": 3.6625,
+      "step": 16300
+    },
+    {
+      "epoch": 0.327,
+      "grad_norm": 5.735015392303467,
+      "learning_rate": 4.1829e-05,
+      "loss": 3.9276,
+      "step": 16350
+    },
+    {
+      "epoch": 0.328,
+      "grad_norm": 5.2200093269348145,
+      "learning_rate": 4.1804e-05,
+      "loss": 3.9428,
+      "step": 16400
+    },
+    {
+      "epoch": 0.329,
+      "grad_norm": 6.257447719573975,
+      "learning_rate": 4.1779e-05,
+      "loss": 3.9017,
+      "step": 16450
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.63900899887085,
+      "learning_rate": 4.175400000000001e-05,
+      "loss": 3.8419,
+      "step": 16500
+    },
+    {
+      "epoch": 0.331,
+      "grad_norm": 4.8323469161987305,
+      "learning_rate": 4.1729e-05,
+      "loss": 3.7655,
+      "step": 16550
+    },
+    {
+      "epoch": 0.332,
+      "grad_norm": 4.395730972290039,
+      "learning_rate": 4.1704000000000006e-05,
+      "loss": 3.8917,
+      "step": 16600
+    },
+    {
+      "epoch": 0.333,
+      "grad_norm": 5.6126484870910645,
+      "learning_rate": 4.1679e-05,
+      "loss": 3.9089,
+      "step": 16650
+    },
+    {
+      "epoch": 0.334,
+      "grad_norm": 7.912436485290527,
+      "learning_rate": 4.1654000000000005e-05,
+      "loss": 3.8822,
+      "step": 16700
+    },
+    {
+      "epoch": 0.335,
+      "grad_norm": 6.455771446228027,
+      "learning_rate": 4.1629000000000004e-05,
+      "loss": 3.8588,
+      "step": 16750
+    },
+    {
+      "epoch": 0.336,
+      "grad_norm": 5.0589447021484375,
+      "learning_rate": 4.1604000000000004e-05,
+      "loss": 3.8178,
+      "step": 16800
+    },
+    {
+      "epoch": 0.337,
+      "grad_norm": 5.345367431640625,
+      "learning_rate": 4.1579e-05,
+      "loss": 3.7995,
+      "step": 16850
+    },
+    {
+      "epoch": 0.338,
+      "grad_norm": 5.080057621002197,
+      "learning_rate": 4.1554e-05,
+      "loss": 3.8185,
+      "step": 16900
+    },
+    {
+      "epoch": 0.339,
+      "grad_norm": 5.360045433044434,
+      "learning_rate": 4.15295e-05,
+      "loss": 3.8565,
+      "step": 16950
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.614706039428711,
+      "learning_rate": 4.15045e-05,
+      "loss": 3.9084,
+      "step": 17000
+    },
+    {
+      "epoch": 0.341,
+      "grad_norm": 6.911114692687988,
+      "learning_rate": 4.14795e-05,
+      "loss": 3.9076,
+      "step": 17050
+    },
+    {
+      "epoch": 0.342,
+      "grad_norm": 6.049564361572266,
+      "learning_rate": 4.14545e-05,
+      "loss": 3.8488,
+      "step": 17100
+    },
+    {
+      "epoch": 0.343,
+      "grad_norm": 5.199458599090576,
+      "learning_rate": 4.142950000000001e-05,
+      "loss": 3.7354,
+      "step": 17150
+    },
+    {
+      "epoch": 0.344,
+      "grad_norm": 5.313460826873779,
+      "learning_rate": 4.14045e-05,
+      "loss": 3.7896,
+      "step": 17200
+    },
+    {
+      "epoch": 0.345,
+      "grad_norm": 6.561459064483643,
+      "learning_rate": 4.1379500000000006e-05,
+      "loss": 3.9219,
+      "step": 17250
+    },
+    {
+      "epoch": 0.346,
+      "grad_norm": 6.4212565422058105,
+      "learning_rate": 4.13545e-05,
+      "loss": 3.7611,
+      "step": 17300
+    },
+    {
+      "epoch": 0.347,
+      "grad_norm": 5.5424275398254395,
+      "learning_rate": 4.1329500000000005e-05,
+      "loss": 3.7,
+      "step": 17350
+    },
+    {
+      "epoch": 0.348,
+      "grad_norm": 5.535832405090332,
+      "learning_rate": 4.1304500000000004e-05,
+      "loss": 3.8373,
+      "step": 17400
+    },
+    {
+      "epoch": 0.349,
+      "grad_norm": 6.890970230102539,
+      "learning_rate": 4.1279500000000003e-05,
+      "loss": 3.8647,
+      "step": 17450
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.52531623840332,
+      "learning_rate": 4.12545e-05,
+      "loss": 3.9338,
+      "step": 17500
+    },
+    {
+      "epoch": 0.351,
+      "grad_norm": 5.829474925994873,
+      "learning_rate": 4.12295e-05,
+      "loss": 3.7462,
+      "step": 17550
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 5.539413928985596,
+      "learning_rate": 4.12045e-05,
+      "loss": 3.9142,
+      "step": 17600
+    },
+    {
+      "epoch": 0.353,
+      "grad_norm": 5.760298252105713,
+      "learning_rate": 4.11795e-05,
+      "loss": 3.7838,
+      "step": 17650
+    },
+    {
+      "epoch": 0.354,
+      "grad_norm": 5.714521408081055,
+      "learning_rate": 4.11545e-05,
+      "loss": 3.5264,
+      "step": 17700
+    },
+    {
+      "epoch": 0.355,
+      "grad_norm": 7.143244743347168,
+      "learning_rate": 4.1129500000000007e-05,
+      "loss": 3.5885,
+      "step": 17750
+    },
+    {
+      "epoch": 0.356,
+      "grad_norm": 5.311776161193848,
+      "learning_rate": 4.11045e-05,
+      "loss": 3.8021,
+      "step": 17800
+    },
+    {
+      "epoch": 0.357,
+      "grad_norm": 5.300187587738037,
+      "learning_rate": 4.1079500000000005e-05,
+      "loss": 3.7987,
+      "step": 17850
+    },
+    {
+      "epoch": 0.358,
+      "grad_norm": 4.707153797149658,
+      "learning_rate": 4.10545e-05,
+      "loss": 3.7739,
+      "step": 17900
+    },
+    {
+      "epoch": 0.359,
+      "grad_norm": 6.6488261222839355,
+      "learning_rate": 4.1029500000000004e-05,
+      "loss": 3.6998,
+      "step": 17950
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.836310386657715,
+      "learning_rate": 4.1004500000000004e-05,
+      "loss": 3.8338,
+      "step": 18000
+    },
+    {
+      "epoch": 0.361,
+      "grad_norm": 4.986599922180176,
+      "learning_rate": 4.09795e-05,
+      "loss": 3.8181,
+      "step": 18050
+    },
+    {
+      "epoch": 0.362,
+      "grad_norm": 6.216142654418945,
+      "learning_rate": 4.09545e-05,
+      "loss": 3.7621,
+      "step": 18100
+    },
+    {
+      "epoch": 0.363,
+      "grad_norm": 5.541645050048828,
+      "learning_rate": 4.09295e-05,
+      "loss": 3.6238,
+      "step": 18150
+    },
+    {
+      "epoch": 0.364,
+      "grad_norm": 5.436352252960205,
+      "learning_rate": 4.09045e-05,
+      "loss": 3.5689,
+      "step": 18200
+    },
+    {
+      "epoch": 0.365,
+      "grad_norm": 5.6476335525512695,
+      "learning_rate": 4.08795e-05,
+      "loss": 3.884,
+      "step": 18250
+    },
+    {
+      "epoch": 0.366,
+      "grad_norm": 5.718410968780518,
+      "learning_rate": 4.08545e-05,
+      "loss": 3.6267,
+      "step": 18300
+    },
+    {
+      "epoch": 0.367,
+      "grad_norm": 5.955104827880859,
+      "learning_rate": 4.0829500000000006e-05,
+      "loss": 3.7001,
+      "step": 18350
+    },
+    {
+      "epoch": 0.368,
+      "grad_norm": 7.096001148223877,
+      "learning_rate": 4.08045e-05,
+      "loss": 3.7513,
+      "step": 18400
+    },
+    {
+      "epoch": 0.369,
+      "grad_norm": 5.519197463989258,
+      "learning_rate": 4.0779500000000005e-05,
+      "loss": 3.6899,
+      "step": 18450
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.956851482391357,
+      "learning_rate": 4.0754500000000004e-05,
+      "loss": 3.5223,
+      "step": 18500
+    },
+    {
+      "epoch": 0.371,
+      "grad_norm": 7.059772491455078,
+      "learning_rate": 4.0729500000000004e-05,
+      "loss": 3.6623,
+      "step": 18550
+    },
+    {
+      "epoch": 0.372,
+      "grad_norm": 5.00886344909668,
+      "learning_rate": 4.07045e-05,
+      "loss": 3.6135,
+      "step": 18600
+    },
+    {
+      "epoch": 0.373,
+      "grad_norm": 9.62269115447998,
+      "learning_rate": 4.06795e-05,
+      "loss": 3.5963,
+      "step": 18650
+    },
+    {
+      "epoch": 0.374,
+      "grad_norm": 5.832387924194336,
+      "learning_rate": 4.06545e-05,
+      "loss": 3.8358,
+      "step": 18700
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 5.544480323791504,
+      "learning_rate": 4.06295e-05,
+      "loss": 3.7966,
+      "step": 18750
+    },
+    {
+      "epoch": 0.376,
+      "grad_norm": 7.148738384246826,
+      "learning_rate": 4.06045e-05,
+      "loss": 3.6962,
+      "step": 18800
+    },
+    {
+      "epoch": 0.377,
+      "grad_norm": 6.446701526641846,
+      "learning_rate": 4.057950000000001e-05,
+      "loss": 3.6023,
+      "step": 18850
+    },
+    {
+      "epoch": 0.378,
+      "grad_norm": 5.0687713623046875,
+      "learning_rate": 4.05545e-05,
+      "loss": 3.6317,
+      "step": 18900
+    },
+    {
+      "epoch": 0.379,
+      "grad_norm": 6.391605854034424,
+      "learning_rate": 4.0529500000000006e-05,
+      "loss": 3.7784,
+      "step": 18950
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.7233076095581055,
+      "learning_rate": 4.05045e-05,
+      "loss": 3.8169,
+      "step": 19000
+    },
+    {
+      "epoch": 0.381,
+      "grad_norm": 5.0900678634643555,
+      "learning_rate": 4.048e-05,
+      "loss": 3.9733,
+      "step": 19050
+    },
+    {
+      "epoch": 0.382,
+      "grad_norm": 7.224969387054443,
+      "learning_rate": 4.0455000000000005e-05,
+      "loss": 3.9576,
+      "step": 19100
+    },
+    {
+      "epoch": 0.383,
+      "grad_norm": 5.281608581542969,
+      "learning_rate": 4.0430000000000004e-05,
+      "loss": 3.8107,
+      "step": 19150
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 5.875776290893555,
+      "learning_rate": 4.0405000000000004e-05,
+      "loss": 3.8561,
+      "step": 19200
+    },
+    {
+      "epoch": 0.385,
+      "grad_norm": 5.377148628234863,
+      "learning_rate": 4.038e-05,
+      "loss": 3.8033,
+      "step": 19250
+    },
+    {
+      "epoch": 0.386,
+      "grad_norm": 6.931129455566406,
+      "learning_rate": 4.0355e-05,
+      "loss": 3.6839,
+      "step": 19300
+    },
+    {
+      "epoch": 0.387,
+      "grad_norm": 6.245096206665039,
+      "learning_rate": 4.033e-05,
+      "loss": 3.6428,
+      "step": 19350
+    },
+    {
+      "epoch": 0.388,
+      "grad_norm": 5.333400249481201,
+      "learning_rate": 4.0305e-05,
+      "loss": 3.5137,
+      "step": 19400
+    },
+    {
+      "epoch": 0.389,
+      "grad_norm": 5.533875942230225,
+      "learning_rate": 4.028e-05,
+      "loss": 3.8904,
+      "step": 19450
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.073616981506348,
+      "learning_rate": 4.025500000000001e-05,
+      "loss": 3.7149,
+      "step": 19500
+    },
+    {
+      "epoch": 0.391,
+      "grad_norm": 5.462928771972656,
+      "learning_rate": 4.023e-05,
+      "loss": 3.7531,
+      "step": 19550
+    },
+    {
+      "epoch": 0.392,
+      "grad_norm": 5.310408115386963,
+      "learning_rate": 4.0205000000000006e-05,
+      "loss": 3.6716,
+      "step": 19600
+    },
+    {
+      "epoch": 0.393,
+      "grad_norm": 9.851359367370605,
+      "learning_rate": 4.018e-05,
+      "loss": 3.7123,
+      "step": 19650
+    },
+    {
+      "epoch": 0.394,
+      "grad_norm": 6.514659881591797,
+      "learning_rate": 4.0155000000000004e-05,
+      "loss": 3.8305,
+      "step": 19700
+    },
+    {
+      "epoch": 0.395,
+      "grad_norm": 6.6378374099731445,
+      "learning_rate": 4.0130000000000004e-05,
+      "loss": 3.6212,
+      "step": 19750
+    },
+    {
+      "epoch": 0.396,
+      "grad_norm": 7.522112846374512,
+      "learning_rate": 4.0105e-05,
+      "loss": 3.7809,
+      "step": 19800
+    },
+    {
+      "epoch": 0.397,
+      "grad_norm": 6.330406665802002,
+      "learning_rate": 4.008e-05,
+      "loss": 3.6223,
+      "step": 19850
+    },
+    {
+      "epoch": 0.398,
+      "grad_norm": 5.559330463409424,
+      "learning_rate": 4.0055e-05,
+      "loss": 3.6856,
+      "step": 19900
+    },
+    {
+      "epoch": 0.399,
+      "grad_norm": 5.872786045074463,
+      "learning_rate": 4.003e-05,
+      "loss": 3.6335,
+      "step": 19950
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.806800842285156,
+      "learning_rate": 4.0005e-05,
+      "loss": 3.6443,
+      "step": 20000
+    },
+    {
+      "epoch": 0.401,
+      "grad_norm": 5.577113151550293,
+      "learning_rate": 3.998e-05,
+      "loss": 3.7707,
+      "step": 20050
+    },
+    {
+      "epoch": 0.402,
+      "grad_norm": 6.950751304626465,
+      "learning_rate": 3.9955000000000006e-05,
+      "loss": 3.6968,
+      "step": 20100
+    },
+    {
+      "epoch": 0.403,
+      "grad_norm": 6.79304838180542,
+      "learning_rate": 3.993e-05,
+      "loss": 3.676,
+      "step": 20150
+    },
+    {
+      "epoch": 0.404,
+      "grad_norm": 5.125970840454102,
+      "learning_rate": 3.9905000000000005e-05,
+      "loss": 3.6477,
+      "step": 20200
+    },
+    {
+      "epoch": 0.405,
+      "grad_norm": 5.478579998016357,
+      "learning_rate": 3.988e-05,
+      "loss": 3.7483,
+      "step": 20250
+    },
+    {
+      "epoch": 0.406,
+      "grad_norm": 10.339896202087402,
+      "learning_rate": 3.9855000000000004e-05,
+      "loss": 3.6422,
+      "step": 20300
+    },
+    {
+      "epoch": 0.407,
+      "grad_norm": 5.875314712524414,
+      "learning_rate": 3.983e-05,
+      "loss": 3.6645,
+      "step": 20350
+    },
+    {
+      "epoch": 0.408,
+      "grad_norm": 7.035477638244629,
+      "learning_rate": 3.9805e-05,
+      "loss": 3.9642,
+      "step": 20400
+    },
+    {
+      "epoch": 0.409,
+      "grad_norm": 5.8045125007629395,
+      "learning_rate": 3.978e-05,
+      "loss": 3.6628,
+      "step": 20450
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.441490173339844,
+      "learning_rate": 3.9755e-05,
+      "loss": 3.6702,
+      "step": 20500
+    },
+    {
+      "epoch": 0.411,
+      "grad_norm": 9.806413650512695,
+      "learning_rate": 3.973e-05,
+      "loss": 3.8382,
+      "step": 20550
+    },
+    {
+      "epoch": 0.412,
+      "grad_norm": 5.094761848449707,
+      "learning_rate": 3.9705e-05,
+      "loss": 3.4913,
+      "step": 20600
+    },
+    {
+      "epoch": 0.413,
+      "grad_norm": 5.1229248046875,
+      "learning_rate": 3.968e-05,
+      "loss": 3.6758,
+      "step": 20650
+    },
+    {
+      "epoch": 0.414,
+      "grad_norm": 5.905995845794678,
+      "learning_rate": 3.9655000000000006e-05,
+      "loss": 3.5157,
+      "step": 20700
+    },
+    {
+      "epoch": 0.415,
+      "grad_norm": 5.98081636428833,
+      "learning_rate": 3.963e-05,
+      "loss": 3.7878,
+      "step": 20750
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 5.739871501922607,
+      "learning_rate": 3.9605000000000005e-05,
+      "loss": 3.6435,
+      "step": 20800
+    },
+    {
+      "epoch": 0.417,
+      "grad_norm": 7.51310396194458,
+      "learning_rate": 3.958e-05,
+      "loss": 3.6459,
+      "step": 20850
+    },
+    {
+      "epoch": 0.418,
+      "grad_norm": 5.59712553024292,
+      "learning_rate": 3.9555e-05,
+      "loss": 3.5044,
+      "step": 20900
+    },
+    {
+      "epoch": 0.419,
+      "grad_norm": 6.285709381103516,
+      "learning_rate": 3.953e-05,
+      "loss": 3.5614,
+      "step": 20950
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.389233589172363,
+      "learning_rate": 3.9505e-05,
+      "loss": 3.6717,
+      "step": 21000
+    },
+    {
+      "epoch": 0.421,
+      "grad_norm": 5.242571830749512,
+      "learning_rate": 3.948e-05,
+      "loss": 3.7844,
+      "step": 21050
+    },
+    {
+      "epoch": 0.422,
+      "grad_norm": 5.075847148895264,
+      "learning_rate": 3.94555e-05,
+      "loss": 3.7471,
+      "step": 21100
+    },
+    {
+      "epoch": 0.423,
+      "grad_norm": 7.122387409210205,
+      "learning_rate": 3.94305e-05,
+      "loss": 3.6374,
+      "step": 21150
+    },
+    {
+      "epoch": 0.424,
+      "grad_norm": 6.306674003601074,
+      "learning_rate": 3.94055e-05,
+      "loss": 3.6265,
+      "step": 21200
+    },
+    {
+      "epoch": 0.425,
+      "grad_norm": 5.332902908325195,
+      "learning_rate": 3.93805e-05,
+      "loss": 3.4958,
+      "step": 21250
+    },
+    {
+      "epoch": 0.426,
+      "grad_norm": 5.518804550170898,
+      "learning_rate": 3.93555e-05,
+      "loss": 3.7738,
+      "step": 21300
+    },
+    {
+      "epoch": 0.427,
+      "grad_norm": 6.586755275726318,
+      "learning_rate": 3.9330500000000006e-05,
+      "loss": 3.777,
+      "step": 21350
+    },
+    {
+      "epoch": 0.428,
+      "grad_norm": 6.083924293518066,
+      "learning_rate": 3.93055e-05,
+      "loss": 3.4348,
+      "step": 21400
+    },
+    {
+      "epoch": 0.429,
+      "grad_norm": 6.382724761962891,
+      "learning_rate": 3.9280500000000004e-05,
+      "loss": 3.4485,
+      "step": 21450
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.780561923980713,
+      "learning_rate": 3.9255500000000004e-05,
+      "loss": 3.7721,
+      "step": 21500
+    },
+    {
+      "epoch": 0.431,
+      "grad_norm": 7.1107683181762695,
+      "learning_rate": 3.92305e-05,
+      "loss": 3.7841,
+      "step": 21550
+    },
+    {
+      "epoch": 0.432,
+      "grad_norm": 5.783836364746094,
+      "learning_rate": 3.92055e-05,
+      "loss": 3.8264,
+      "step": 21600
+    },
+    {
+      "epoch": 0.433,
+      "grad_norm": 6.417442321777344,
+      "learning_rate": 3.91805e-05,
+      "loss": 3.5419,
+      "step": 21650
+    },
+    {
+      "epoch": 0.434,
+      "grad_norm": 6.642063140869141,
+      "learning_rate": 3.91555e-05,
+      "loss": 3.6673,
+      "step": 21700
+    },
+    {
+      "epoch": 0.435,
+      "grad_norm": 6.779355525970459,
+      "learning_rate": 3.91305e-05,
+      "loss": 3.3835,
+      "step": 21750
+    },
+    {
+      "epoch": 0.436,
+      "grad_norm": 6.05160665512085,
+      "learning_rate": 3.91055e-05,
+      "loss": 3.3663,
+      "step": 21800
+    },
+    {
+      "epoch": 0.437,
+      "grad_norm": 6.938116073608398,
+      "learning_rate": 3.9080500000000006e-05,
+      "loss": 3.5852,
+      "step": 21850
+    },
+    {
+      "epoch": 0.438,
+      "grad_norm": 5.694869041442871,
+      "learning_rate": 3.90555e-05,
+      "loss": 3.4862,
+      "step": 21900
+    },
+    {
+      "epoch": 0.439,
+      "grad_norm": 6.541887283325195,
+      "learning_rate": 3.9030500000000005e-05,
+      "loss": 3.452,
+      "step": 21950
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.991393566131592,
+      "learning_rate": 3.90055e-05,
+      "loss": 3.4754,
+      "step": 22000
+    },
+    {
+      "epoch": 0.441,
+      "grad_norm": 5.7733025550842285,
+      "learning_rate": 3.8980500000000004e-05,
+      "loss": 3.6772,
+      "step": 22050
+    },
+    {
+      "epoch": 0.442,
+      "grad_norm": 5.654860496520996,
+      "learning_rate": 3.89555e-05,
+      "loss": 3.5211,
+      "step": 22100
+    },
+    {
+      "epoch": 0.443,
+      "grad_norm": 7.81252908706665,
+      "learning_rate": 3.89305e-05,
+      "loss": 3.5404,
+      "step": 22150
+    },
+    {
+      "epoch": 0.444,
+      "grad_norm": 5.495180606842041,
+      "learning_rate": 3.89055e-05,
+      "loss": 3.6398,
+      "step": 22200
+    },
+    {
+      "epoch": 0.445,
+      "grad_norm": 7.39019775390625,
+      "learning_rate": 3.88805e-05,
+      "loss": 3.6327,
+      "step": 22250
+    },
+    {
+      "epoch": 0.446,
+      "grad_norm": 6.242217063903809,
+      "learning_rate": 3.88555e-05,
+      "loss": 3.7173,
+      "step": 22300
+    },
+    {
+      "epoch": 0.447,
+      "grad_norm": 5.690585136413574,
+      "learning_rate": 3.88305e-05,
+      "loss": 3.3506,
+      "step": 22350
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 5.454275131225586,
+      "learning_rate": 3.88055e-05,
+      "loss": 3.6386,
+      "step": 22400
+    },
+    {
+      "epoch": 0.449,
+      "grad_norm": 10.499303817749023,
+      "learning_rate": 3.8780500000000006e-05,
+      "loss": 3.6395,
+      "step": 22450
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.501960277557373,
+      "learning_rate": 3.87555e-05,
+      "loss": 3.6292,
+      "step": 22500
+    },
+    {
+      "epoch": 0.451,
+      "grad_norm": 6.991559028625488,
+      "learning_rate": 3.8730500000000005e-05,
+      "loss": 3.6887,
+      "step": 22550
+    },
+    {
+      "epoch": 0.452,
+      "grad_norm": 6.130929946899414,
+      "learning_rate": 3.87055e-05,
+      "loss": 3.7495,
+      "step": 22600
+    },
+    {
+      "epoch": 0.453,
+      "grad_norm": 6.396117210388184,
+      "learning_rate": 3.8680500000000003e-05,
+      "loss": 3.77,
+      "step": 22650
+    },
+    {
+      "epoch": 0.454,
+      "grad_norm": 5.2830939292907715,
+      "learning_rate": 3.86555e-05,
+      "loss": 3.2854,
+      "step": 22700
+    },
+    {
+      "epoch": 0.455,
+      "grad_norm": 5.548758506774902,
+      "learning_rate": 3.86305e-05,
+      "loss": 3.4142,
+      "step": 22750
+    },
+    {
+      "epoch": 0.456,
+      "grad_norm": 8.338722229003906,
+      "learning_rate": 3.86055e-05,
+      "loss": 3.5935,
+      "step": 22800
+    },
+    {
+      "epoch": 0.457,
+      "grad_norm": 5.559829235076904,
+      "learning_rate": 3.85805e-05,
+      "loss": 3.539,
+      "step": 22850
+    },
+    {
+      "epoch": 0.458,
+      "grad_norm": 6.5197601318359375,
+      "learning_rate": 3.85555e-05,
+      "loss": 3.5664,
+      "step": 22900
+    },
+    {
+      "epoch": 0.459,
+      "grad_norm": 6.871826648712158,
+      "learning_rate": 3.85305e-05,
+      "loss": 3.4904,
+      "step": 22950
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.459062099456787,
+      "learning_rate": 3.85055e-05,
+      "loss": 3.5906,
+      "step": 23000
+    },
+    {
+      "epoch": 0.461,
+      "grad_norm": 7.952236652374268,
+      "learning_rate": 3.8480500000000005e-05,
+      "loss": 3.4198,
+      "step": 23050
+    },
+    {
+      "epoch": 0.462,
+      "grad_norm": 5.091039180755615,
+      "learning_rate": 3.8456000000000006e-05,
+      "loss": 3.6821,
+      "step": 23100
+    },
+    {
+      "epoch": 0.463,
+      "grad_norm": 5.670896530151367,
+      "learning_rate": 3.8431e-05,
+      "loss": 3.4976,
+      "step": 23150
+    },
+    {
+      "epoch": 0.464,
+      "grad_norm": 5.132435321807861,
+      "learning_rate": 3.8406000000000005e-05,
+      "loss": 3.7298,
+      "step": 23200
+    },
+    {
+      "epoch": 0.465,
+      "grad_norm": 7.798178195953369,
+      "learning_rate": 3.8381e-05,
+      "loss": 3.6912,
+      "step": 23250
+    },
+    {
+      "epoch": 0.466,
+      "grad_norm": 7.2740983963012695,
+      "learning_rate": 3.8356000000000003e-05,
+      "loss": 3.4549,
+      "step": 23300
+    },
+    {
+      "epoch": 0.467,
+      "grad_norm": 5.059585094451904,
+      "learning_rate": 3.8331e-05,
+      "loss": 3.4792,
+      "step": 23350
+    },
+    {
+      "epoch": 0.468,
+      "grad_norm": 7.0619893074035645,
+      "learning_rate": 3.8306e-05,
+      "loss": 3.7425,
+      "step": 23400
+    },
+    {
+      "epoch": 0.469,
+      "grad_norm": 6.161375522613525,
+      "learning_rate": 3.8281e-05,
+      "loss": 3.5272,
+      "step": 23450
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.16375207901001,
+      "learning_rate": 3.8256e-05,
+      "loss": 3.669,
+      "step": 23500
+    },
+    {
+      "epoch": 0.471,
+      "grad_norm": 5.802226543426514,
+      "learning_rate": 3.8231e-05,
+      "loss": 3.6014,
+      "step": 23550
+    },
+    {
+      "epoch": 0.472,
+      "grad_norm": 4.789316177368164,
+      "learning_rate": 3.8206e-05,
+      "loss": 3.4649,
+      "step": 23600
+    },
+    {
+      "epoch": 0.473,
+      "grad_norm": 6.274521827697754,
+      "learning_rate": 3.8181e-05,
+      "loss": 3.4292,
+      "step": 23650
+    },
+    {
+      "epoch": 0.474,
+      "grad_norm": 5.716911792755127,
+      "learning_rate": 3.8156000000000005e-05,
+      "loss": 3.3089,
+      "step": 23700
+    },
+    {
+      "epoch": 0.475,
+      "grad_norm": 6.505302429199219,
+      "learning_rate": 3.8131e-05,
+      "loss": 3.6164,
+      "step": 23750
+    },
+    {
+      "epoch": 0.476,
+      "grad_norm": 6.222828388214111,
+      "learning_rate": 3.8106000000000004e-05,
+      "loss": 3.4026,
+      "step": 23800
+    },
+    {
+      "epoch": 0.477,
+      "grad_norm": 6.24766206741333,
+      "learning_rate": 3.8081e-05,
+      "loss": 3.443,
+      "step": 23850
+    },
+    {
+      "epoch": 0.478,
+      "grad_norm": 7.223185062408447,
+      "learning_rate": 3.8056e-05,
+      "loss": 3.6329,
+      "step": 23900
+    },
+    {
+      "epoch": 0.479,
+      "grad_norm": 6.246890544891357,
+      "learning_rate": 3.8031e-05,
+      "loss": 3.3726,
+      "step": 23950
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.942480564117432,
+      "learning_rate": 3.8006e-05,
+      "loss": 3.5571,
+      "step": 24000
+    },
+    {
+      "epoch": 0.481,
+      "grad_norm": 5.436431407928467,
+      "learning_rate": 3.7981e-05,
+      "loss": 3.4648,
+      "step": 24050
+    },
+    {
+      "epoch": 0.482,
+      "grad_norm": 6.082589626312256,
+      "learning_rate": 3.7956e-05,
+      "loss": 3.5571,
+      "step": 24100
+    },
+    {
+      "epoch": 0.483,
+      "grad_norm": 7.166791915893555,
+      "learning_rate": 3.7931e-05,
+      "loss": 3.5147,
+      "step": 24150
+    },
+    {
+      "epoch": 0.484,
+      "grad_norm": 6.249309539794922,
+      "learning_rate": 3.7906e-05,
+      "loss": 3.6697,
+      "step": 24200
+    },
+    {
+      "epoch": 0.485,
+      "grad_norm": 7.935781478881836,
+      "learning_rate": 3.7881e-05,
+      "loss": 3.4738,
+      "step": 24250
+    },
+    {
+      "epoch": 0.486,
+      "grad_norm": 5.446045875549316,
+      "learning_rate": 3.7856000000000005e-05,
+      "loss": 3.3919,
+      "step": 24300
+    },
+    {
+      "epoch": 0.487,
+      "grad_norm": 6.253237724304199,
+      "learning_rate": 3.7831e-05,
+      "loss": 3.6019,
+      "step": 24350
+    },
+    {
+      "epoch": 0.488,
+      "grad_norm": 13.48686695098877,
+      "learning_rate": 3.7806000000000004e-05,
+      "loss": 3.5479,
+      "step": 24400
+    },
+    {
+      "epoch": 0.489,
+      "grad_norm": 7.377110481262207,
+      "learning_rate": 3.7781e-05,
+      "loss": 3.3436,
+      "step": 24450
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.141451835632324,
+      "learning_rate": 3.7756e-05,
+      "loss": 3.4679,
+      "step": 24500
+    },
+    {
+      "epoch": 0.491,
+      "grad_norm": 5.041921615600586,
+      "learning_rate": 3.7731e-05,
+      "loss": 3.5736,
+      "step": 24550
+    },
+    {
+      "epoch": 0.492,
+      "grad_norm": 6.365973949432373,
+      "learning_rate": 3.7706e-05,
+      "loss": 3.5077,
+      "step": 24600
+    },
+    {
+      "epoch": 0.493,
+      "grad_norm": 5.532159805297852,
+      "learning_rate": 3.7681e-05,
+      "loss": 3.4038,
+      "step": 24650
+    },
+    {
+      "epoch": 0.494,
+      "grad_norm": 6.18168830871582,
+      "learning_rate": 3.7656e-05,
+      "loss": 3.36,
+      "step": 24700
+    },
+    {
+      "epoch": 0.495,
+      "grad_norm": 7.991237163543701,
+      "learning_rate": 3.7631e-05,
+      "loss": 3.4926,
+      "step": 24750
+    },
+    {
+      "epoch": 0.496,
+      "grad_norm": 7.126778602600098,
+      "learning_rate": 3.7606000000000006e-05,
+      "loss": 3.6006,
+      "step": 24800
+    },
+    {
+      "epoch": 0.497,
+      "grad_norm": 8.774191856384277,
+      "learning_rate": 3.7581e-05,
+      "loss": 3.551,
+      "step": 24850
+    },
+    {
+      "epoch": 0.498,
+      "grad_norm": 7.611205101013184,
+      "learning_rate": 3.7556000000000004e-05,
+      "loss": 3.3755,
+      "step": 24900
+    },
+    {
+      "epoch": 0.499,
+      "grad_norm": 8.540602684020996,
+      "learning_rate": 3.7531e-05,
+      "loss": 3.3131,
+      "step": 24950
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.313963890075684,
+      "learning_rate": 3.7506e-05,
+      "loss": 3.4656,
+      "step": 25000
+    },
+    {
+      "epoch": 0.501,
+      "grad_norm": 5.498029708862305,
+      "learning_rate": 3.7481e-05,
+      "loss": 3.4094,
+      "step": 25050
+    },
+    {
+      "epoch": 0.502,
+      "grad_norm": 6.962924480438232,
+      "learning_rate": 3.7456e-05,
+      "loss": 3.6499,
+      "step": 25100
+    },
+    {
+      "epoch": 0.503,
+      "grad_norm": 7.007199764251709,
+      "learning_rate": 3.7431e-05,
+      "loss": 3.4367,
+      "step": 25150
+    },
+    {
+      "epoch": 0.504,
+      "grad_norm": 7.297113418579102,
+      "learning_rate": 3.74065e-05,
+      "loss": 3.4985,
+      "step": 25200
+    },
+    {
+      "epoch": 0.505,
+      "grad_norm": 6.30457878112793,
+      "learning_rate": 3.73815e-05,
+      "loss": 3.4324,
+      "step": 25250
+    },
+    {
+      "epoch": 0.506,
+      "grad_norm": 6.83005952835083,
+      "learning_rate": 3.73565e-05,
+      "loss": 3.3904,
+      "step": 25300
+    },
+    {
+      "epoch": 0.507,
+      "grad_norm": 7.69933557510376,
+      "learning_rate": 3.73315e-05,
+      "loss": 3.6869,
+      "step": 25350
+    },
+    {
+      "epoch": 0.508,
+      "grad_norm": 6.051611423492432,
+      "learning_rate": 3.73065e-05,
+      "loss": 3.5676,
+      "step": 25400
+    },
+    {
+      "epoch": 0.509,
+      "grad_norm": 6.487821578979492,
+      "learning_rate": 3.7281500000000005e-05,
+      "loss": 3.5218,
+      "step": 25450
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.36344575881958,
+      "learning_rate": 3.72565e-05,
+      "loss": 3.4862,
+      "step": 25500
+    },
+    {
+      "epoch": 0.511,
+      "grad_norm": 5.60974645614624,
+      "learning_rate": 3.7231500000000004e-05,
+      "loss": 3.304,
+      "step": 25550
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 5.354685306549072,
+      "learning_rate": 3.72065e-05,
+      "loss": 3.4184,
+      "step": 25600
+    },
+    {
+      "epoch": 0.513,
+      "grad_norm": 6.2862420082092285,
+      "learning_rate": 3.71815e-05,
+      "loss": 3.6906,
+      "step": 25650
+    },
+    {
+      "epoch": 0.514,
+      "grad_norm": 7.004101753234863,
+      "learning_rate": 3.71565e-05,
+      "loss": 3.4951,
+      "step": 25700
+    },
+    {
+      "epoch": 0.515,
+      "grad_norm": 11.132220268249512,
+      "learning_rate": 3.71315e-05,
+      "loss": 3.4476,
+      "step": 25750
+    },
+    {
+      "epoch": 0.516,
+      "grad_norm": 10.14500617980957,
+      "learning_rate": 3.71065e-05,
+      "loss": 3.5392,
+      "step": 25800
+    },
+    {
+      "epoch": 0.517,
+      "grad_norm": 5.003336429595947,
+      "learning_rate": 3.70815e-05,
+      "loss": 3.5125,
+      "step": 25850
+    },
+    {
+      "epoch": 0.518,
+      "grad_norm": 5.623690605163574,
+      "learning_rate": 3.70565e-05,
+      "loss": 3.4813,
+      "step": 25900
+    },
+    {
+      "epoch": 0.519,
+      "grad_norm": 6.5926432609558105,
+      "learning_rate": 3.70315e-05,
+      "loss": 3.5556,
+      "step": 25950
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.16590690612793,
+      "learning_rate": 3.70065e-05,
+      "loss": 3.4139,
+      "step": 26000
+    },
+    {
+      "epoch": 0.521,
+      "grad_norm": 5.917094707489014,
+      "learning_rate": 3.6981500000000005e-05,
+      "loss": 3.5458,
+      "step": 26050
+    },
+    {
+      "epoch": 0.522,
+      "grad_norm": 6.361861705780029,
+      "learning_rate": 3.69565e-05,
+      "loss": 3.4147,
+      "step": 26100
+    },
+    {
+      "epoch": 0.523,
+      "grad_norm": 6.118766784667969,
+      "learning_rate": 3.6931500000000004e-05,
+      "loss": 3.668,
+      "step": 26150
+    },
+    {
+      "epoch": 0.524,
+      "grad_norm": 7.089566230773926,
+      "learning_rate": 3.6906499999999996e-05,
+      "loss": 3.3806,
+      "step": 26200
+    },
+    {
+      "epoch": 0.525,
+      "grad_norm": 5.34354829788208,
+      "learning_rate": 3.68815e-05,
+      "loss": 3.4482,
+      "step": 26250
+    },
+    {
+      "epoch": 0.526,
+      "grad_norm": 8.42415714263916,
+      "learning_rate": 3.68565e-05,
+      "loss": 3.5294,
+      "step": 26300
+    },
+    {
+      "epoch": 0.527,
+      "grad_norm": 6.866034984588623,
+      "learning_rate": 3.68315e-05,
+      "loss": 3.5308,
+      "step": 26350
+    },
+    {
+      "epoch": 0.528,
+      "grad_norm": 6.633973121643066,
+      "learning_rate": 3.68065e-05,
+      "loss": 3.598,
+      "step": 26400
+    },
+    {
+      "epoch": 0.529,
+      "grad_norm": 5.965973854064941,
+      "learning_rate": 3.67815e-05,
+      "loss": 3.6788,
+      "step": 26450
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.220425605773926,
+      "learning_rate": 3.67565e-05,
+      "loss": 3.2695,
+      "step": 26500
+    },
+    {
+      "epoch": 0.531,
+      "grad_norm": 5.58171272277832,
+      "learning_rate": 3.67315e-05,
+      "loss": 3.4882,
+      "step": 26550
+    },
+    {
+      "epoch": 0.532,
+      "grad_norm": 6.113706111907959,
+      "learning_rate": 3.67065e-05,
+      "loss": 3.5617,
+      "step": 26600
+    },
+    {
+      "epoch": 0.533,
+      "grad_norm": 6.026816368103027,
+      "learning_rate": 3.6681500000000004e-05,
+      "loss": 3.5393,
+      "step": 26650
+    },
+    {
+      "epoch": 0.534,
+      "grad_norm": 7.218907356262207,
+      "learning_rate": 3.66565e-05,
+      "loss": 3.5193,
+      "step": 26700
+    },
+    {
+      "epoch": 0.535,
+      "grad_norm": 6.00888729095459,
+      "learning_rate": 3.66315e-05,
+      "loss": 3.4065,
+      "step": 26750
+    },
+    {
+      "epoch": 0.536,
+      "grad_norm": 7.0676679611206055,
+      "learning_rate": 3.6606499999999996e-05,
+      "loss": 3.2652,
+      "step": 26800
+    },
+    {
+      "epoch": 0.537,
+      "grad_norm": 5.561299800872803,
+      "learning_rate": 3.65815e-05,
+      "loss": 3.6312,
+      "step": 26850
+    },
+    {
+      "epoch": 0.538,
+      "grad_norm": 6.713068008422852,
+      "learning_rate": 3.65565e-05,
+      "loss": 3.6427,
+      "step": 26900
+    },
+    {
+      "epoch": 0.539,
+      "grad_norm": 5.804120063781738,
+      "learning_rate": 3.65315e-05,
+      "loss": 3.1819,
+      "step": 26950
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.488461017608643,
+      "learning_rate": 3.65065e-05,
+      "loss": 3.4673,
+      "step": 27000
+    },
+    {
+      "epoch": 0.541,
+      "grad_norm": 6.780620098114014,
+      "learning_rate": 3.64815e-05,
+      "loss": 3.4887,
+      "step": 27050
+    },
+    {
+      "epoch": 0.542,
+      "grad_norm": 6.49476957321167,
+      "learning_rate": 3.64565e-05,
+      "loss": 3.6123,
+      "step": 27100
+    },
+    {
+      "epoch": 0.543,
+      "grad_norm": 6.915765285491943,
+      "learning_rate": 3.64315e-05,
+      "loss": 3.4393,
+      "step": 27150
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 6.593012809753418,
+      "learning_rate": 3.6406500000000005e-05,
+      "loss": 3.5151,
+      "step": 27200
+    },
+    {
+      "epoch": 0.545,
+      "grad_norm": 6.4061760902404785,
+      "learning_rate": 3.6382e-05,
+      "loss": 3.4307,
+      "step": 27250
+    },
+    {
+      "epoch": 0.546,
+      "grad_norm": 6.907027721405029,
+      "learning_rate": 3.6357000000000004e-05,
+      "loss": 3.4679,
+      "step": 27300
+    },
+    {
+      "epoch": 0.547,
+      "grad_norm": 5.774991989135742,
+      "learning_rate": 3.6332e-05,
+      "loss": 3.3226,
+      "step": 27350
+    },
+    {
+      "epoch": 0.548,
+      "grad_norm": 5.766547679901123,
+      "learning_rate": 3.6307e-05,
+      "loss": 3.4283,
+      "step": 27400
+    },
+    {
+      "epoch": 0.549,
+      "grad_norm": 7.284669876098633,
+      "learning_rate": 3.6282e-05,
+      "loss": 3.4481,
+      "step": 27450
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.108999729156494,
+      "learning_rate": 3.6257e-05,
+      "loss": 3.5764,
+      "step": 27500
+    },
+    {
+      "epoch": 0.551,
+      "grad_norm": 8.253243446350098,
+      "learning_rate": 3.62325e-05,
+      "loss": 3.4882,
+      "step": 27550
+    },
+    {
+      "epoch": 0.552,
+      "grad_norm": 8.250670433044434,
+      "learning_rate": 3.62075e-05,
+      "loss": 3.3237,
+      "step": 27600
+    },
+    {
+      "epoch": 0.553,
+      "grad_norm": 7.141805648803711,
+      "learning_rate": 3.61825e-05,
+      "loss": 3.5175,
+      "step": 27650
+    },
+    {
+      "epoch": 0.554,
+      "grad_norm": 7.745401859283447,
+      "learning_rate": 3.61575e-05,
+      "loss": 3.3908,
+      "step": 27700
+    },
+    {
+      "epoch": 0.555,
+      "grad_norm": 5.771395683288574,
+      "learning_rate": 3.61325e-05,
+      "loss": 3.2266,
+      "step": 27750
+    },
+    {
+      "epoch": 0.556,
+      "grad_norm": 7.92910623550415,
+      "learning_rate": 3.61075e-05,
+      "loss": 3.1754,
+      "step": 27800
+    },
+    {
+      "epoch": 0.557,
+      "grad_norm": 7.521282196044922,
+      "learning_rate": 3.6082500000000006e-05,
+      "loss": 3.5141,
+      "step": 27850
+    },
+    {
+      "epoch": 0.558,
+      "grad_norm": 6.065182209014893,
+      "learning_rate": 3.60575e-05,
+      "loss": 3.5914,
+      "step": 27900
+    },
+    {
+      "epoch": 0.559,
+      "grad_norm": 7.105879306793213,
+      "learning_rate": 3.6032500000000004e-05,
+      "loss": 3.2975,
+      "step": 27950
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.2546515464782715,
+      "learning_rate": 3.60075e-05,
+      "loss": 3.5552,
+      "step": 28000
+    },
+    {
+      "epoch": 0.561,
+      "grad_norm": 7.344614505767822,
+      "learning_rate": 3.59825e-05,
+      "loss": 3.3722,
+      "step": 28050
+    },
+    {
+      "epoch": 0.562,
+      "grad_norm": 5.996744155883789,
+      "learning_rate": 3.59575e-05,
+      "loss": 3.4005,
+      "step": 28100
+    },
+    {
+      "epoch": 0.563,
+      "grad_norm": 5.243777751922607,
+      "learning_rate": 3.59325e-05,
+      "loss": 3.5161,
+      "step": 28150
+    },
+    {
+      "epoch": 0.564,
+      "grad_norm": 5.685503959655762,
+      "learning_rate": 3.59075e-05,
+      "loss": 3.4874,
+      "step": 28200
+    },
+    {
+      "epoch": 0.565,
+      "grad_norm": 6.744235038757324,
+      "learning_rate": 3.58825e-05,
+      "loss": 3.4842,
+      "step": 28250
+    },
+    {
+      "epoch": 0.566,
+      "grad_norm": 6.83828592300415,
+      "learning_rate": 3.58575e-05,
+      "loss": 3.3306,
+      "step": 28300
+    },
+    {
+      "epoch": 0.567,
+      "grad_norm": 6.369629859924316,
+      "learning_rate": 3.58325e-05,
+      "loss": 3.4862,
+      "step": 28350
+    },
+    {
+      "epoch": 0.568,
+      "grad_norm": 6.583117485046387,
+      "learning_rate": 3.58075e-05,
+      "loss": 3.2907,
+      "step": 28400
+    },
+    {
+      "epoch": 0.569,
+      "grad_norm": 8.253291130065918,
+      "learning_rate": 3.5782500000000005e-05,
+      "loss": 3.3894,
+      "step": 28450
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.858260154724121,
+      "learning_rate": 3.57575e-05,
+      "loss": 3.4645,
+      "step": 28500
+    },
+    {
+      "epoch": 0.571,
+      "grad_norm": 6.5981221199035645,
+      "learning_rate": 3.5732500000000004e-05,
+      "loss": 3.5327,
+      "step": 28550
+    },
+    {
+      "epoch": 0.572,
+      "grad_norm": 6.859750270843506,
+      "learning_rate": 3.5707499999999997e-05,
+      "loss": 3.3362,
+      "step": 28600
+    },
+    {
+      "epoch": 0.573,
+      "grad_norm": 7.866117000579834,
+      "learning_rate": 3.56825e-05,
+      "loss": 3.6273,
+      "step": 28650
+    },
+    {
+      "epoch": 0.574,
+      "grad_norm": 5.867875576019287,
+      "learning_rate": 3.56575e-05,
+      "loss": 3.3157,
+      "step": 28700
+    },
+    {
+      "epoch": 0.575,
+      "grad_norm": 6.136321544647217,
+      "learning_rate": 3.56325e-05,
+      "loss": 3.345,
+      "step": 28750
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 6.829803466796875,
+      "learning_rate": 3.56075e-05,
+      "loss": 3.3066,
+      "step": 28800
+    },
+    {
+      "epoch": 0.577,
+      "grad_norm": 6.177272319793701,
+      "learning_rate": 3.55825e-05,
+      "loss": 3.4357,
+      "step": 28850
+    },
+    {
+      "epoch": 0.578,
+      "grad_norm": 6.678221225738525,
+      "learning_rate": 3.55575e-05,
+      "loss": 3.4919,
+      "step": 28900
+    },
+    {
+      "epoch": 0.579,
+      "grad_norm": 5.453572750091553,
+      "learning_rate": 3.55325e-05,
+      "loss": 3.3439,
+      "step": 28950
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.180026054382324,
+      "learning_rate": 3.55075e-05,
+      "loss": 3.4776,
+      "step": 29000
+    },
+    {
+      "epoch": 0.581,
+      "grad_norm": 6.252904415130615,
+      "learning_rate": 3.5482500000000005e-05,
+      "loss": 3.4176,
+      "step": 29050
+    },
+    {
+      "epoch": 0.582,
+      "grad_norm": 7.99691104888916,
+      "learning_rate": 3.54575e-05,
+      "loss": 3.3812,
+      "step": 29100
+    },
+    {
+      "epoch": 0.583,
+      "grad_norm": 6.05564022064209,
+      "learning_rate": 3.54325e-05,
+      "loss": 3.5308,
+      "step": 29150
+    },
+    {
+      "epoch": 0.584,
+      "grad_norm": 6.284146308898926,
+      "learning_rate": 3.5407499999999996e-05,
+      "loss": 3.3615,
+      "step": 29200
+    },
+    {
+      "epoch": 0.585,
+      "grad_norm": 10.308747291564941,
+      "learning_rate": 3.53825e-05,
+      "loss": 3.2485,
+      "step": 29250
+    },
+    {
+      "epoch": 0.586,
+      "grad_norm": 5.75729513168335,
+      "learning_rate": 3.53575e-05,
+      "loss": 3.3904,
+      "step": 29300
+    },
+    {
+      "epoch": 0.587,
+      "grad_norm": 6.623788356781006,
+      "learning_rate": 3.53325e-05,
+      "loss": 3.4116,
+      "step": 29350
+    },
+    {
+      "epoch": 0.588,
+      "grad_norm": 5.757374286651611,
+      "learning_rate": 3.530750000000001e-05,
+      "loss": 3.4118,
+      "step": 29400
+    },
+    {
+      "epoch": 0.589,
+      "grad_norm": 6.021833896636963,
+      "learning_rate": 3.52825e-05,
+      "loss": 3.3119,
+      "step": 29450
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.991232395172119,
+      "learning_rate": 3.5257500000000006e-05,
+      "loss": 3.4385,
+      "step": 29500
+    },
+    {
+      "epoch": 0.591,
+      "grad_norm": 6.228766441345215,
+      "learning_rate": 3.52325e-05,
+      "loss": 3.3973,
+      "step": 29550
+    },
+    {
+      "epoch": 0.592,
+      "grad_norm": 6.7667131423950195,
+      "learning_rate": 3.5207500000000005e-05,
+      "loss": 3.492,
+      "step": 29600
+    },
+    {
+      "epoch": 0.593,
+      "grad_norm": 7.218946933746338,
+      "learning_rate": 3.5182500000000004e-05,
+      "loss": 3.4921,
+      "step": 29650
+    },
+    {
+      "epoch": 0.594,
+      "grad_norm": 6.98897123336792,
+      "learning_rate": 3.5157500000000003e-05,
+      "loss": 3.432,
+      "step": 29700
+    },
+    {
+      "epoch": 0.595,
+      "grad_norm": 7.497246265411377,
+      "learning_rate": 3.51325e-05,
+      "loss": 3.2246,
+      "step": 29750
+    },
+    {
+      "epoch": 0.596,
+      "grad_norm": 6.9698686599731445,
+      "learning_rate": 3.51075e-05,
+      "loss": 3.4781,
+      "step": 29800
+    },
+    {
+      "epoch": 0.597,
+      "grad_norm": 5.121500492095947,
+      "learning_rate": 3.50825e-05,
+      "loss": 3.4499,
+      "step": 29850
+    },
+    {
+      "epoch": 0.598,
+      "grad_norm": 6.432731628417969,
+      "learning_rate": 3.50575e-05,
+      "loss": 3.2076,
+      "step": 29900
+    },
+    {
+      "epoch": 0.599,
+      "grad_norm": 5.066705703735352,
+      "learning_rate": 3.50325e-05,
+      "loss": 3.4198,
+      "step": 29950
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.430960655212402,
+      "learning_rate": 3.5007500000000007e-05,
+      "loss": 3.4104,
+      "step": 30000
+    },
+    {
+      "epoch": 0.601,
+      "grad_norm": 9.220848083496094,
+      "learning_rate": 3.49825e-05,
+      "loss": 3.4114,
+      "step": 30050
+    },
+    {
+      "epoch": 0.602,
+      "grad_norm": 6.596940517425537,
+      "learning_rate": 3.4957500000000005e-05,
+      "loss": 3.3112,
+      "step": 30100
+    },
+    {
+      "epoch": 0.603,
+      "grad_norm": 8.433357238769531,
+      "learning_rate": 3.49325e-05,
+      "loss": 3.499,
+      "step": 30150
+    },
+    {
+      "epoch": 0.604,
+      "grad_norm": 7.3862738609313965,
+      "learning_rate": 3.4907500000000004e-05,
+      "loss": 3.2682,
+      "step": 30200
+    },
+    {
+      "epoch": 0.605,
+      "grad_norm": 7.238501071929932,
+      "learning_rate": 3.4882500000000004e-05,
+      "loss": 3.4221,
+      "step": 30250
+    },
+    {
+      "epoch": 0.606,
+      "grad_norm": 6.758235931396484,
+      "learning_rate": 3.48575e-05,
+      "loss": 3.3936,
+      "step": 30300
+    },
+    {
+      "epoch": 0.607,
+      "grad_norm": 6.463011741638184,
+      "learning_rate": 3.48325e-05,
+      "loss": 3.5434,
+      "step": 30350
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 8.514334678649902,
+      "learning_rate": 3.48075e-05,
+      "loss": 3.3497,
+      "step": 30400
+    },
+    {
+      "epoch": 0.609,
+      "grad_norm": 5.6331915855407715,
+      "learning_rate": 3.47825e-05,
+      "loss": 3.2981,
+      "step": 30450
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.361815929412842,
+      "learning_rate": 3.475750000000001e-05,
+      "loss": 3.2069,
+      "step": 30500
+    },
+    {
+      "epoch": 0.611,
+      "grad_norm": 5.87959098815918,
+      "learning_rate": 3.47325e-05,
+      "loss": 3.3483,
+      "step": 30550
+    },
+    {
+      "epoch": 0.612,
+      "grad_norm": 7.229435443878174,
+      "learning_rate": 3.4707500000000006e-05,
+      "loss": 3.3584,
+      "step": 30600
+    },
+    {
+      "epoch": 0.613,
+      "grad_norm": 5.749815940856934,
+      "learning_rate": 3.46825e-05,
+      "loss": 3.5367,
+      "step": 30650
+    },
+    {
+      "epoch": 0.614,
+      "grad_norm": 7.94840145111084,
+      "learning_rate": 3.4657500000000005e-05,
+      "loss": 3.3931,
+      "step": 30700
+    },
+    {
+      "epoch": 0.615,
+      "grad_norm": 6.358616828918457,
+      "learning_rate": 3.4632500000000004e-05,
+      "loss": 3.2705,
+      "step": 30750
+    },
+    {
+      "epoch": 0.616,
+      "grad_norm": 6.277470111846924,
+      "learning_rate": 3.4607500000000004e-05,
+      "loss": 3.4044,
+      "step": 30800
+    },
+    {
+      "epoch": 0.617,
+      "grad_norm": 6.311099529266357,
+      "learning_rate": 3.45825e-05,
+      "loss": 3.3239,
+      "step": 30850
+    },
+    {
+      "epoch": 0.618,
+      "grad_norm": 7.405715465545654,
+      "learning_rate": 3.45575e-05,
+      "loss": 3.3971,
+      "step": 30900
+    },
+    {
+      "epoch": 0.619,
+      "grad_norm": 6.301971912384033,
+      "learning_rate": 3.45325e-05,
+      "loss": 3.4035,
+      "step": 30950
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.6296844482421875,
+      "learning_rate": 3.45075e-05,
+      "loss": 3.4517,
+      "step": 31000
+    },
+    {
+      "epoch": 0.621,
+      "grad_norm": 6.209890842437744,
+      "learning_rate": 3.44825e-05,
+      "loss": 3.3755,
+      "step": 31050
+    },
+    {
+      "epoch": 0.622,
+      "grad_norm": 5.518044471740723,
+      "learning_rate": 3.445750000000001e-05,
+      "loss": 3.2734,
+      "step": 31100
+    },
+    {
+      "epoch": 0.623,
+      "grad_norm": 7.205735206604004,
+      "learning_rate": 3.44325e-05,
+      "loss": 3.3552,
+      "step": 31150
+    },
+    {
+      "epoch": 0.624,
+      "grad_norm": 6.634133815765381,
+      "learning_rate": 3.4407500000000006e-05,
+      "loss": 3.2855,
+      "step": 31200
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 6.324268341064453,
+      "learning_rate": 3.43825e-05,
+      "loss": 3.4877,
+      "step": 31250
+    },
+    {
+      "epoch": 0.626,
+      "grad_norm": 5.836503505706787,
+      "learning_rate": 3.4357500000000004e-05,
+      "loss": 3.492,
+      "step": 31300
+    },
+    {
+      "epoch": 0.627,
+      "grad_norm": 6.615384101867676,
+      "learning_rate": 3.4332500000000004e-05,
+      "loss": 3.5195,
+      "step": 31350
+    },
+    {
+      "epoch": 0.628,
+      "grad_norm": 7.209743499755859,
+      "learning_rate": 3.43075e-05,
+      "loss": 3.4772,
+      "step": 31400
+    },
+    {
+      "epoch": 0.629,
+      "grad_norm": 8.48714828491211,
+      "learning_rate": 3.42825e-05,
+      "loss": 3.5931,
+      "step": 31450
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.982743740081787,
+      "learning_rate": 3.42575e-05,
+      "loss": 3.5209,
+      "step": 31500
+    },
+    {
+      "epoch": 0.631,
+      "grad_norm": 6.397634029388428,
+      "learning_rate": 3.4233e-05,
+      "loss": 3.2492,
+      "step": 31550
+    },
+    {
+      "epoch": 0.632,
+      "grad_norm": 6.8665642738342285,
+      "learning_rate": 3.4208e-05,
+      "loss": 3.4509,
+      "step": 31600
+    },
+    {
+      "epoch": 0.633,
+      "grad_norm": 6.930782794952393,
+      "learning_rate": 3.4183e-05,
+      "loss": 3.2818,
+      "step": 31650
+    },
+    {
+      "epoch": 0.634,
+      "grad_norm": 5.9751296043396,
+      "learning_rate": 3.4158e-05,
+      "loss": 3.2229,
+      "step": 31700
+    },
+    {
+      "epoch": 0.635,
+      "grad_norm": 6.509876728057861,
+      "learning_rate": 3.413300000000001e-05,
+      "loss": 3.2482,
+      "step": 31750
+    },
+    {
+      "epoch": 0.636,
+      "grad_norm": 6.303530693054199,
+      "learning_rate": 3.4108e-05,
+      "loss": 3.3513,
+      "step": 31800
+    },
+    {
+      "epoch": 0.637,
+      "grad_norm": 5.964878559112549,
+      "learning_rate": 3.4083000000000006e-05,
+      "loss": 3.1939,
+      "step": 31850
+    },
+    {
+      "epoch": 0.638,
+      "grad_norm": 5.7284321784973145,
+      "learning_rate": 3.4058e-05,
+      "loss": 3.3163,
+      "step": 31900
+    },
+    {
+      "epoch": 0.639,
+      "grad_norm": 5.236353397369385,
+      "learning_rate": 3.4033000000000004e-05,
+      "loss": 3.3796,
+      "step": 31950
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.227442741394043,
+      "learning_rate": 3.4008000000000004e-05,
+      "loss": 3.2888,
+      "step": 32000
+    },
+    {
+      "epoch": 0.641,
+      "grad_norm": 9.655413627624512,
+      "learning_rate": 3.3983e-05,
+      "loss": 3.5678,
+      "step": 32050
+    },
+    {
+      "epoch": 0.642,
+      "grad_norm": 5.565030574798584,
+      "learning_rate": 3.3958e-05,
+      "loss": 3.2497,
+      "step": 32100
+    },
+    {
+      "epoch": 0.643,
+      "grad_norm": 11.382092475891113,
+      "learning_rate": 3.3933e-05,
+      "loss": 3.2119,
+      "step": 32150
+    },
+    {
+      "epoch": 0.644,
+      "grad_norm": 6.986525058746338,
+      "learning_rate": 3.3908e-05,
+      "loss": 3.3342,
+      "step": 32200
+    },
+    {
+      "epoch": 0.645,
+      "grad_norm": 5.240586757659912,
+      "learning_rate": 3.3883e-05,
+      "loss": 3.2399,
+      "step": 32250
+    },
+    {
+      "epoch": 0.646,
+      "grad_norm": 6.879291534423828,
+      "learning_rate": 3.3858e-05,
+      "loss": 3.4391,
+      "step": 32300
+    },
+    {
+      "epoch": 0.647,
+      "grad_norm": 7.895921230316162,
+      "learning_rate": 3.3833000000000006e-05,
+      "loss": 3.3315,
+      "step": 32350
+    },
+    {
+      "epoch": 0.648,
+      "grad_norm": 6.591862678527832,
+      "learning_rate": 3.3808e-05,
+      "loss": 3.2601,
+      "step": 32400
+    },
+    {
+      "epoch": 0.649,
+      "grad_norm": 8.269440650939941,
+      "learning_rate": 3.3783000000000005e-05,
+      "loss": 3.1949,
+      "step": 32450
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.437238693237305,
+      "learning_rate": 3.3758e-05,
+      "loss": 3.5878,
+      "step": 32500
+    },
+    {
+      "epoch": 0.651,
+      "grad_norm": 5.902334690093994,
+      "learning_rate": 3.3733000000000004e-05,
+      "loss": 3.2744,
+      "step": 32550
+    },
+    {
+      "epoch": 0.652,
+      "grad_norm": 8.443634986877441,
+      "learning_rate": 3.3708e-05,
+      "loss": 3.0574,
+      "step": 32600
+    },
+    {
+      "epoch": 0.653,
+      "grad_norm": 6.006880283355713,
+      "learning_rate": 3.3683e-05,
+      "loss": 3.2878,
+      "step": 32650
+    },
+    {
+      "epoch": 0.654,
+      "grad_norm": 6.461446762084961,
+      "learning_rate": 3.3658e-05,
+      "loss": 3.2513,
+      "step": 32700
+    },
+    {
+      "epoch": 0.655,
+      "grad_norm": 7.1918487548828125,
+      "learning_rate": 3.3633e-05,
+      "loss": 3.2637,
+      "step": 32750
+    },
+    {
+      "epoch": 0.656,
+      "grad_norm": 7.551942825317383,
+      "learning_rate": 3.3608e-05,
+      "loss": 3.3599,
+      "step": 32800
+    },
+    {
+      "epoch": 0.657,
+      "grad_norm": 6.817183971405029,
+      "learning_rate": 3.3583e-05,
+      "loss": 3.5034,
+      "step": 32850
+    },
+    {
+      "epoch": 0.658,
+      "grad_norm": 6.113791465759277,
+      "learning_rate": 3.3558e-05,
+      "loss": 3.2996,
+      "step": 32900
+    },
+    {
+      "epoch": 0.659,
+      "grad_norm": 6.195582389831543,
+      "learning_rate": 3.3533000000000006e-05,
+      "loss": 3.3898,
+      "step": 32950
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.301150798797607,
+      "learning_rate": 3.3508e-05,
+      "loss": 3.2412,
+      "step": 33000
+    },
+    {
+      "epoch": 0.661,
+      "grad_norm": 5.854180812835693,
+      "learning_rate": 3.34835e-05,
+      "loss": 3.4688,
+      "step": 33050
+    },
+    {
+      "epoch": 0.662,
+      "grad_norm": 9.124200820922852,
+      "learning_rate": 3.3458500000000005e-05,
+      "loss": 3.3273,
+      "step": 33100
+    },
+    {
+      "epoch": 0.663,
+      "grad_norm": 7.126101970672607,
+      "learning_rate": 3.34335e-05,
+      "loss": 3.2921,
+      "step": 33150
+    },
+    {
+      "epoch": 0.664,
+      "grad_norm": 7.3969526290893555,
+      "learning_rate": 3.3408500000000004e-05,
+      "loss": 3.2588,
+      "step": 33200
+    },
+    {
+      "epoch": 0.665,
+      "grad_norm": 6.349921703338623,
+      "learning_rate": 3.33835e-05,
+      "loss": 3.1005,
+      "step": 33250
+    },
+    {
+      "epoch": 0.666,
+      "grad_norm": 7.069624900817871,
+      "learning_rate": 3.33585e-05,
+      "loss": 3.3315,
+      "step": 33300
+    },
+    {
+      "epoch": 0.667,
+      "grad_norm": 6.957205772399902,
+      "learning_rate": 3.33335e-05,
+      "loss": 3.0547,
+      "step": 33350
+    },
+    {
+      "epoch": 0.668,
+      "grad_norm": 6.064169406890869,
+      "learning_rate": 3.33085e-05,
+      "loss": 3.4743,
+      "step": 33400
+    },
+    {
+      "epoch": 0.669,
+      "grad_norm": 8.175496101379395,
+      "learning_rate": 3.32835e-05,
+      "loss": 3.1822,
+      "step": 33450
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.576241493225098,
+      "learning_rate": 3.325850000000001e-05,
+      "loss": 3.1038,
+      "step": 33500
+    },
+    {
+      "epoch": 0.671,
+      "grad_norm": 6.801853179931641,
+      "learning_rate": 3.32335e-05,
+      "loss": 3.1863,
+      "step": 33550
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 5.9864983558654785,
+      "learning_rate": 3.3208500000000006e-05,
+      "loss": 3.4432,
+      "step": 33600
+    },
+    {
+      "epoch": 0.673,
+      "grad_norm": 9.82041072845459,
+      "learning_rate": 3.31835e-05,
+      "loss": 3.3234,
+      "step": 33650
+    },
+    {
+      "epoch": 0.674,
+      "grad_norm": 6.94479513168335,
+      "learning_rate": 3.3158500000000004e-05,
+      "loss": 3.3856,
+      "step": 33700
+    },
+    {
+      "epoch": 0.675,
+      "grad_norm": 7.436299800872803,
+      "learning_rate": 3.3133500000000004e-05,
+      "loss": 3.4031,
+      "step": 33750
+    },
+    {
+      "epoch": 0.676,
+      "grad_norm": 7.951491832733154,
+      "learning_rate": 3.31085e-05,
+      "loss": 3.1939,
+      "step": 33800
+    },
+    {
+      "epoch": 0.677,
+      "grad_norm": 5.3420000076293945,
+      "learning_rate": 3.30835e-05,
+      "loss": 3.277,
+      "step": 33850
+    },
+    {
+      "epoch": 0.678,
+      "grad_norm": 6.747936725616455,
+      "learning_rate": 3.30585e-05,
+      "loss": 3.457,
+      "step": 33900
+    },
+    {
+      "epoch": 0.679,
+      "grad_norm": 7.466944217681885,
+      "learning_rate": 3.30335e-05,
+      "loss": 3.4629,
+      "step": 33950
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.159815311431885,
+      "learning_rate": 3.30085e-05,
+      "loss": 3.3527,
+      "step": 34000
+    },
+    {
+      "epoch": 0.681,
+      "grad_norm": 6.799013137817383,
+      "learning_rate": 3.29835e-05,
+      "loss": 3.2996,
+      "step": 34050
+    },
+    {
+      "epoch": 0.682,
+      "grad_norm": 7.7746262550354,
+      "learning_rate": 3.2958500000000006e-05,
+      "loss": 3.2661,
+      "step": 34100
+    },
+    {
+      "epoch": 0.683,
+      "grad_norm": 6.598784923553467,
+      "learning_rate": 3.29335e-05,
+      "loss": 3.3371,
+      "step": 34150
+    },
+    {
+      "epoch": 0.684,
+      "grad_norm": 6.144473075866699,
+      "learning_rate": 3.2908500000000005e-05,
+      "loss": 3.2466,
+      "step": 34200
+    },
+    {
+      "epoch": 0.685,
+      "grad_norm": 7.488722324371338,
+      "learning_rate": 3.28835e-05,
+      "loss": 3.2008,
+      "step": 34250
+    },
+    {
+      "epoch": 0.686,
+      "grad_norm": 5.943243980407715,
+      "learning_rate": 3.2858500000000004e-05,
+      "loss": 3.1717,
+      "step": 34300
+    },
+    {
+      "epoch": 0.687,
+      "grad_norm": 7.279262542724609,
+      "learning_rate": 3.28335e-05,
+      "loss": 3.2673,
+      "step": 34350
+    },
+    {
+      "epoch": 0.688,
+      "grad_norm": 7.265892028808594,
+      "learning_rate": 3.28085e-05,
+      "loss": 3.3252,
+      "step": 34400
+    },
+    {
+      "epoch": 0.689,
+      "grad_norm": 7.290516376495361,
+      "learning_rate": 3.27835e-05,
+      "loss": 3.1889,
+      "step": 34450
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.666197776794434,
+      "learning_rate": 3.27585e-05,
+      "loss": 3.2569,
+      "step": 34500
+    },
+    {
+      "epoch": 0.691,
+      "grad_norm": 6.722541809082031,
+      "learning_rate": 3.27335e-05,
+      "loss": 3.0594,
+      "step": 34550
+    },
+    {
+      "epoch": 0.692,
+      "grad_norm": 6.768825531005859,
+      "learning_rate": 3.27085e-05,
+      "loss": 3.2594,
+      "step": 34600
+    },
+    {
+      "epoch": 0.693,
+      "grad_norm": 6.368803024291992,
+      "learning_rate": 3.26835e-05,
+      "loss": 3.1851,
+      "step": 34650
+    },
+    {
+      "epoch": 0.694,
+      "grad_norm": 8.27947998046875,
+      "learning_rate": 3.2658500000000006e-05,
+      "loss": 3.2759,
+      "step": 34700
+    },
+    {
+      "epoch": 0.695,
+      "grad_norm": 6.843438625335693,
+      "learning_rate": 3.26335e-05,
+      "loss": 3.1578,
+      "step": 34750
+    },
+    {
+      "epoch": 0.696,
+      "grad_norm": 8.467235565185547,
+      "learning_rate": 3.2608500000000005e-05,
+      "loss": 3.2223,
+      "step": 34800
+    },
+    {
+      "epoch": 0.697,
+      "grad_norm": 5.828463554382324,
+      "learning_rate": 3.25835e-05,
+      "loss": 3.2224,
+      "step": 34850
+    },
+    {
+      "epoch": 0.698,
+      "grad_norm": 6.603434085845947,
+      "learning_rate": 3.2558500000000003e-05,
+      "loss": 3.3885,
+      "step": 34900
+    },
+    {
+      "epoch": 0.699,
+      "grad_norm": 6.739859580993652,
+      "learning_rate": 3.25335e-05,
+      "loss": 3.1141,
+      "step": 34950
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.304507255554199,
+      "learning_rate": 3.25085e-05,
+      "loss": 3.2604,
+      "step": 35000
+    },
+    {
+      "epoch": 0.701,
+      "grad_norm": 7.21973180770874,
+      "learning_rate": 3.24835e-05,
+      "loss": 3.2029,
+      "step": 35050
+    },
+    {
+      "epoch": 0.702,
+      "grad_norm": 6.183377742767334,
+      "learning_rate": 3.24585e-05,
+      "loss": 3.3226,
+      "step": 35100
+    },
+    {
+      "epoch": 0.703,
+      "grad_norm": 7.0609540939331055,
+      "learning_rate": 3.24335e-05,
+      "loss": 3.186,
+      "step": 35150
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 6.175162315368652,
+      "learning_rate": 3.24085e-05,
+      "loss": 3.2179,
+      "step": 35200
+    },
+    {
+      "epoch": 0.705,
+      "grad_norm": 5.934025764465332,
+      "learning_rate": 3.23835e-05,
+      "loss": 3.2335,
+      "step": 35250
+    },
+    {
+      "epoch": 0.706,
+      "grad_norm": 8.728473663330078,
+      "learning_rate": 3.2358500000000005e-05,
+      "loss": 3.2551,
+      "step": 35300
+    },
+    {
+      "epoch": 0.707,
+      "grad_norm": 7.110035419464111,
+      "learning_rate": 3.23335e-05,
+      "loss": 3.2937,
+      "step": 35350
+    },
+    {
+      "epoch": 0.708,
+      "grad_norm": 6.184364318847656,
+      "learning_rate": 3.2308500000000004e-05,
+      "loss": 3.0839,
+      "step": 35400
+    },
+    {
+      "epoch": 0.709,
+      "grad_norm": 6.328525066375732,
+      "learning_rate": 3.22835e-05,
+      "loss": 3.1893,
+      "step": 35450
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.684478759765625,
+      "learning_rate": 3.22585e-05,
+      "loss": 3.1691,
+      "step": 35500
+    },
+    {
+      "epoch": 0.711,
+      "grad_norm": 5.313352584838867,
+      "learning_rate": 3.22335e-05,
+      "loss": 3.4243,
+      "step": 35550
+    },
+    {
+      "epoch": 0.712,
+      "grad_norm": 6.211653709411621,
+      "learning_rate": 3.22085e-05,
+      "loss": 3.5204,
+      "step": 35600
+    },
+    {
+      "epoch": 0.713,
+      "grad_norm": 5.643269062042236,
+      "learning_rate": 3.21835e-05,
+      "loss": 3.4294,
+      "step": 35650
+    },
+    {
+      "epoch": 0.714,
+      "grad_norm": 7.590113639831543,
+      "learning_rate": 3.21585e-05,
+      "loss": 3.4033,
+      "step": 35700
+    },
+    {
+      "epoch": 0.715,
+      "grad_norm": 7.15920352935791,
+      "learning_rate": 3.21335e-05,
+      "loss": 3.1821,
+      "step": 35750
+    },
+    {
+      "epoch": 0.716,
+      "grad_norm": 6.130990505218506,
+      "learning_rate": 3.21085e-05,
+      "loss": 3.2556,
+      "step": 35800
+    },
+    {
+      "epoch": 0.717,
+      "grad_norm": 6.418147563934326,
+      "learning_rate": 3.20835e-05,
+      "loss": 3.1433,
+      "step": 35850
+    },
+    {
+      "epoch": 0.718,
+      "grad_norm": 7.010365009307861,
+      "learning_rate": 3.2058500000000005e-05,
+      "loss": 3.3129,
+      "step": 35900
+    },
+    {
+      "epoch": 0.719,
+      "grad_norm": 6.206649303436279,
+      "learning_rate": 3.20335e-05,
+      "loss": 3.5057,
+      "step": 35950
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.868955612182617,
+      "learning_rate": 3.2008500000000004e-05,
+      "loss": 3.2482,
+      "step": 36000
+    },
+    {
+      "epoch": 0.721,
+      "grad_norm": 6.5969343185424805,
+      "learning_rate": 3.1983499999999996e-05,
+      "loss": 3.2539,
+      "step": 36050
+    },
+    {
+      "epoch": 0.722,
+      "grad_norm": 6.978048801422119,
+      "learning_rate": 3.19585e-05,
+      "loss": 3.1573,
+      "step": 36100
+    },
+    {
+      "epoch": 0.723,
+      "grad_norm": 6.406214714050293,
+      "learning_rate": 3.19335e-05,
+      "loss": 3.3721,
+      "step": 36150
+    },
+    {
+      "epoch": 0.724,
+      "grad_norm": 9.857353210449219,
+      "learning_rate": 3.19085e-05,
+      "loss": 3.3179,
+      "step": 36200
+    },
+    {
+      "epoch": 0.725,
+      "grad_norm": 6.391420841217041,
+      "learning_rate": 3.18835e-05,
+      "loss": 3.3233,
+      "step": 36250
+    },
+    {
+      "epoch": 0.726,
+      "grad_norm": 8.015118598937988,
+      "learning_rate": 3.18585e-05,
+      "loss": 3.0956,
+      "step": 36300
+    },
+    {
+      "epoch": 0.727,
+      "grad_norm": 5.823697566986084,
+      "learning_rate": 3.18335e-05,
+      "loss": 3.1998,
+      "step": 36350
+    },
+    {
+      "epoch": 0.728,
+      "grad_norm": 7.559394359588623,
+      "learning_rate": 3.1808500000000006e-05,
+      "loss": 3.2908,
+      "step": 36400
+    },
+    {
+      "epoch": 0.729,
+      "grad_norm": 6.329631328582764,
+      "learning_rate": 3.17835e-05,
+      "loss": 3.2908,
+      "step": 36450
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.789269924163818,
+      "learning_rate": 3.1758500000000004e-05,
+      "loss": 3.3667,
+      "step": 36500
+    },
+    {
+      "epoch": 0.731,
+      "grad_norm": 5.785370349884033,
+      "learning_rate": 3.17335e-05,
+      "loss": 3.1064,
+      "step": 36550
+    },
+    {
+      "epoch": 0.732,
+      "grad_norm": 7.6929755210876465,
+      "learning_rate": 3.17085e-05,
+      "loss": 2.9547,
+      "step": 36600
+    },
+    {
+      "epoch": 0.733,
+      "grad_norm": 9.009896278381348,
+      "learning_rate": 3.16835e-05,
+      "loss": 3.0989,
+      "step": 36650
+    },
+    {
+      "epoch": 0.734,
+      "grad_norm": 6.26234245300293,
+      "learning_rate": 3.16585e-05,
+      "loss": 3.2564,
+      "step": 36700
+    },
+    {
+      "epoch": 0.735,
+      "grad_norm": 5.810740947723389,
+      "learning_rate": 3.16335e-05,
+      "loss": 3.0043,
+      "step": 36750
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 6.051753520965576,
+      "learning_rate": 3.16085e-05,
+      "loss": 3.1809,
+      "step": 36800
+    },
+    {
+      "epoch": 0.737,
+      "grad_norm": 7.068359851837158,
+      "learning_rate": 3.15835e-05,
+      "loss": 3.2625,
+      "step": 36850
+    },
+    {
+      "epoch": 0.738,
+      "grad_norm": 8.538505554199219,
+      "learning_rate": 3.15585e-05,
+      "loss": 3.1945,
+      "step": 36900
+    },
+    {
+      "epoch": 0.739,
+      "grad_norm": 8.019240379333496,
+      "learning_rate": 3.15335e-05,
+      "loss": 3.1444,
+      "step": 36950
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.253057479858398,
+      "learning_rate": 3.1508500000000005e-05,
+      "loss": 2.8749,
+      "step": 37000
+    },
+    {
+      "epoch": 0.741,
+      "grad_norm": 6.1826491355896,
+      "learning_rate": 3.14835e-05,
+      "loss": 3.1814,
+      "step": 37050
+    },
+    {
+      "epoch": 0.742,
+      "grad_norm": 6.367573261260986,
+      "learning_rate": 3.1458500000000004e-05,
+      "loss": 3.3143,
+      "step": 37100
+    },
+    {
+      "epoch": 0.743,
+      "grad_norm": 7.386603832244873,
+      "learning_rate": 3.1433499999999996e-05,
+      "loss": 3.4178,
+      "step": 37150
+    },
+    {
+      "epoch": 0.744,
+      "grad_norm": 7.538274765014648,
+      "learning_rate": 3.1409e-05,
+      "loss": 3.2942,
+      "step": 37200
+    },
+    {
+      "epoch": 0.745,
+      "grad_norm": 8.587214469909668,
+      "learning_rate": 3.1384e-05,
+      "loss": 3.1972,
+      "step": 37250
+    },
+    {
+      "epoch": 0.746,
+      "grad_norm": 6.307685375213623,
+      "learning_rate": 3.1359e-05,
+      "loss": 3.2638,
+      "step": 37300
+    },
+    {
+      "epoch": 0.747,
+      "grad_norm": 7.901178359985352,
+      "learning_rate": 3.1334e-05,
+      "loss": 3.212,
+      "step": 37350
+    },
+    {
+      "epoch": 0.748,
+      "grad_norm": 7.572473526000977,
+      "learning_rate": 3.1309e-05,
+      "loss": 2.9569,
+      "step": 37400
+    },
+    {
+      "epoch": 0.749,
+      "grad_norm": 6.52233362197876,
+      "learning_rate": 3.1284e-05,
+      "loss": 3.3622,
+      "step": 37450
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.932403564453125,
+      "learning_rate": 3.1259e-05,
+      "loss": 3.1381,
+      "step": 37500
+    },
+    {
+      "epoch": 0.751,
+      "grad_norm": 6.504958152770996,
+      "learning_rate": 3.1234e-05,
+      "loss": 3.0979,
+      "step": 37550
+    },
+    {
+      "epoch": 0.752,
+      "grad_norm": 8.710679054260254,
+      "learning_rate": 3.1209e-05,
+      "loss": 3.2664,
+      "step": 37600
+    },
+    {
+      "epoch": 0.753,
+      "grad_norm": 5.697458744049072,
+      "learning_rate": 3.1184000000000005e-05,
+      "loss": 3.3793,
+      "step": 37650
+    },
+    {
+      "epoch": 0.754,
+      "grad_norm": 7.049675941467285,
+      "learning_rate": 3.1159e-05,
+      "loss": 3.3189,
+      "step": 37700
+    },
+    {
+      "epoch": 0.755,
+      "grad_norm": 6.419721603393555,
+      "learning_rate": 3.1134000000000004e-05,
+      "loss": 3.1938,
+      "step": 37750
+    },
+    {
+      "epoch": 0.756,
+      "grad_norm": 10.290610313415527,
+      "learning_rate": 3.1108999999999996e-05,
+      "loss": 3.2407,
+      "step": 37800
+    },
+    {
+      "epoch": 0.757,
+      "grad_norm": 7.655055999755859,
+      "learning_rate": 3.1084e-05,
+      "loss": 3.1777,
+      "step": 37850
+    },
+    {
+      "epoch": 0.758,
+      "grad_norm": 7.728888034820557,
+      "learning_rate": 3.1059e-05,
+      "loss": 3.0303,
+      "step": 37900
+    },
+    {
+      "epoch": 0.759,
+      "grad_norm": 7.557064533233643,
+      "learning_rate": 3.1034e-05,
+      "loss": 3.1784,
+      "step": 37950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.15919303894043,
+      "learning_rate": 3.1009e-05,
+      "loss": 3.2102,
+      "step": 38000
+    },
+    {
+      "epoch": 0.761,
+      "grad_norm": 5.564249038696289,
+      "learning_rate": 3.0984e-05,
+      "loss": 3.3053,
+      "step": 38050
+    },
+    {
+      "epoch": 0.762,
+      "grad_norm": 8.526185035705566,
+      "learning_rate": 3.0959e-05,
+      "loss": 3.4698,
+      "step": 38100
+    },
+    {
+      "epoch": 0.763,
+      "grad_norm": 7.014064311981201,
+      "learning_rate": 3.0934e-05,
+      "loss": 3.1349,
+      "step": 38150
+    },
+    {
+      "epoch": 0.764,
+      "grad_norm": 6.149911880493164,
+      "learning_rate": 3.0909e-05,
+      "loss": 3.3659,
+      "step": 38200
+    },
+    {
+      "epoch": 0.765,
+      "grad_norm": 7.0216064453125,
+      "learning_rate": 3.0884000000000004e-05,
+      "loss": 3.3658,
+      "step": 38250
+    },
+    {
+      "epoch": 0.766,
+      "grad_norm": 6.487388610839844,
+      "learning_rate": 3.0859e-05,
+      "loss": 3.1718,
+      "step": 38300
+    },
+    {
+      "epoch": 0.767,
+      "grad_norm": 7.192234039306641,
+      "learning_rate": 3.0834e-05,
+      "loss": 3.1795,
+      "step": 38350
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 6.610414981842041,
+      "learning_rate": 3.0808999999999996e-05,
+      "loss": 3.0035,
+      "step": 38400
+    },
+    {
+      "epoch": 0.769,
+      "grad_norm": 5.408251762390137,
+      "learning_rate": 3.0784e-05,
+      "loss": 3.4079,
+      "step": 38450
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.19015645980835,
+      "learning_rate": 3.0759e-05,
+      "loss": 3.2097,
+      "step": 38500
+    },
+    {
+      "epoch": 0.771,
+      "grad_norm": 8.879510879516602,
+      "learning_rate": 3.0734e-05,
+      "loss": 3.1956,
+      "step": 38550
+    },
+    {
+      "epoch": 0.772,
+      "grad_norm": 6.8085198402404785,
+      "learning_rate": 3.0709e-05,
+      "loss": 3.191,
+      "step": 38600
+    },
+    {
+      "epoch": 0.773,
+      "grad_norm": 6.9796671867370605,
+      "learning_rate": 3.0684e-05,
+      "loss": 3.0009,
+      "step": 38650
+    },
+    {
+      "epoch": 0.774,
+      "grad_norm": 6.667499542236328,
+      "learning_rate": 3.0659e-05,
+      "loss": 3.1651,
+      "step": 38700
+    },
+    {
+      "epoch": 0.775,
+      "grad_norm": 7.873382091522217,
+      "learning_rate": 3.0634e-05,
+      "loss": 3.2115,
+      "step": 38750
+    },
+    {
+      "epoch": 0.776,
+      "grad_norm": 7.049197673797607,
+      "learning_rate": 3.0609e-05,
+      "loss": 3.1057,
+      "step": 38800
+    },
+    {
+      "epoch": 0.777,
+      "grad_norm": 5.474276065826416,
+      "learning_rate": 3.0584000000000004e-05,
+      "loss": 3.205,
+      "step": 38850
+    },
+    {
+      "epoch": 0.778,
+      "grad_norm": 7.194087028503418,
+      "learning_rate": 3.0558999999999997e-05,
+      "loss": 3.1913,
+      "step": 38900
+    },
+    {
+      "epoch": 0.779,
+      "grad_norm": 8.143173217773438,
+      "learning_rate": 3.0534e-05,
+      "loss": 3.3131,
+      "step": 38950
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.855649948120117,
+      "learning_rate": 3.0509e-05,
+      "loss": 3.0642,
+      "step": 39000
+    },
+    {
+      "epoch": 0.781,
+      "grad_norm": 5.941763401031494,
+      "learning_rate": 3.0484e-05,
+      "loss": 3.0076,
+      "step": 39050
+    },
+    {
+      "epoch": 0.782,
+      "grad_norm": 5.700584888458252,
+      "learning_rate": 3.0459000000000004e-05,
+      "loss": 3.1064,
+      "step": 39100
+    },
+    {
+      "epoch": 0.783,
+      "grad_norm": 7.375488758087158,
+      "learning_rate": 3.0434e-05,
+      "loss": 3.4071,
+      "step": 39150
+    },
+    {
+      "epoch": 0.784,
+      "grad_norm": 7.246762752532959,
+      "learning_rate": 3.0409000000000003e-05,
+      "loss": 3.1041,
+      "step": 39200
+    },
+    {
+      "epoch": 0.785,
+      "grad_norm": 6.453239917755127,
+      "learning_rate": 3.03845e-05,
+      "loss": 3.3334,
+      "step": 39250
+    },
+    {
+      "epoch": 0.786,
+      "grad_norm": 6.580065727233887,
+      "learning_rate": 3.0359500000000003e-05,
+      "loss": 3.0096,
+      "step": 39300
+    },
+    {
+      "epoch": 0.787,
+      "grad_norm": 6.784661293029785,
+      "learning_rate": 3.03345e-05,
+      "loss": 3.345,
+      "step": 39350
+    },
+    {
+      "epoch": 0.788,
+      "grad_norm": 5.282735824584961,
+      "learning_rate": 3.0309500000000002e-05,
+      "loss": 3.2078,
+      "step": 39400
+    },
+    {
+      "epoch": 0.789,
+      "grad_norm": 6.952493190765381,
+      "learning_rate": 3.0284499999999998e-05,
+      "loss": 3.2125,
+      "step": 39450
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.221185207366943,
+      "learning_rate": 3.02595e-05,
+      "loss": 3.1933,
+      "step": 39500
+    },
+    {
+      "epoch": 0.791,
+      "grad_norm": 6.6697564125061035,
+      "learning_rate": 3.02345e-05,
+      "loss": 3.2683,
+      "step": 39550
+    },
+    {
+      "epoch": 0.792,
+      "grad_norm": 7.7430243492126465,
+      "learning_rate": 3.0209500000000003e-05,
+      "loss": 3.2786,
+      "step": 39600
+    },
+    {
+      "epoch": 0.793,
+      "grad_norm": 7.560022830963135,
+      "learning_rate": 3.01845e-05,
+      "loss": 3.1485,
+      "step": 39650
+    },
+    {
+      "epoch": 0.794,
+      "grad_norm": 9.309632301330566,
+      "learning_rate": 3.01595e-05,
+      "loss": 3.3837,
+      "step": 39700
+    },
+    {
+      "epoch": 0.795,
+      "grad_norm": 6.466885089874268,
+      "learning_rate": 3.0134499999999998e-05,
+      "loss": 3.0223,
+      "step": 39750
+    },
+    {
+      "epoch": 0.796,
+      "grad_norm": 7.028127670288086,
+      "learning_rate": 3.01095e-05,
+      "loss": 3.2119,
+      "step": 39800
+    },
+    {
+      "epoch": 0.797,
+      "grad_norm": 6.102771282196045,
+      "learning_rate": 3.00845e-05,
+      "loss": 3.1982,
+      "step": 39850
+    },
+    {
+      "epoch": 0.798,
+      "grad_norm": 6.1855387687683105,
+      "learning_rate": 3.0059500000000002e-05,
+      "loss": 3.1221,
+      "step": 39900
+    },
+    {
+      "epoch": 0.799,
+      "grad_norm": 6.996290683746338,
+      "learning_rate": 3.00345e-05,
+      "loss": 3.1688,
+      "step": 39950
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.62674617767334,
+      "learning_rate": 3.00095e-05,
+      "loss": 2.9733,
+      "step": 40000
+    },
+    {
+      "epoch": 0.801,
+      "grad_norm": 6.106110572814941,
+      "learning_rate": 2.9984499999999997e-05,
+      "loss": 3.1656,
+      "step": 40050
+    },
+    {
+      "epoch": 0.802,
+      "grad_norm": 6.850236415863037,
+      "learning_rate": 2.99595e-05,
+      "loss": 3.1764,
+      "step": 40100
+    },
+    {
+      "epoch": 0.803,
+      "grad_norm": 7.164166450500488,
+      "learning_rate": 2.9934500000000003e-05,
+      "loss": 3.1813,
+      "step": 40150
+    },
+    {
+      "epoch": 0.804,
+      "grad_norm": 8.257601737976074,
+      "learning_rate": 2.9909500000000002e-05,
+      "loss": 3.0087,
+      "step": 40200
+    },
+    {
+      "epoch": 0.805,
+      "grad_norm": 6.853578090667725,
+      "learning_rate": 2.9884500000000005e-05,
+      "loss": 3.1218,
+      "step": 40250
+    },
+    {
+      "epoch": 0.806,
+      "grad_norm": 6.926779270172119,
+      "learning_rate": 2.98595e-05,
+      "loss": 3.2044,
+      "step": 40300
+    },
+    {
+      "epoch": 0.807,
+      "grad_norm": 6.643428802490234,
+      "learning_rate": 2.9834500000000004e-05,
+      "loss": 3.2425,
+      "step": 40350
+    },
+    {
+      "epoch": 0.808,
+      "grad_norm": 7.304353713989258,
+      "learning_rate": 2.98095e-05,
+      "loss": 3.2175,
+      "step": 40400
+    },
+    {
+      "epoch": 0.809,
+      "grad_norm": 6.92768669128418,
+      "learning_rate": 2.9784500000000003e-05,
+      "loss": 3.2095,
+      "step": 40450
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.586690425872803,
+      "learning_rate": 2.9759500000000002e-05,
+      "loss": 3.2259,
+      "step": 40500
+    },
+    {
+      "epoch": 0.811,
+      "grad_norm": 6.485148906707764,
+      "learning_rate": 2.9734500000000005e-05,
+      "loss": 3.1969,
+      "step": 40550
+    },
+    {
+      "epoch": 0.812,
+      "grad_norm": 7.481224060058594,
+      "learning_rate": 2.97095e-05,
+      "loss": 3.1681,
+      "step": 40600
+    },
+    {
+      "epoch": 0.813,
+      "grad_norm": 6.11465311050415,
+      "learning_rate": 2.9684500000000004e-05,
+      "loss": 3.2654,
+      "step": 40650
+    },
+    {
+      "epoch": 0.814,
+      "grad_norm": 7.569796562194824,
+      "learning_rate": 2.96595e-05,
+      "loss": 3.2122,
+      "step": 40700
+    },
+    {
+      "epoch": 0.815,
+      "grad_norm": 6.672354698181152,
+      "learning_rate": 2.9634500000000002e-05,
+      "loss": 3.0291,
+      "step": 40750
+    },
+    {
+      "epoch": 0.816,
+      "grad_norm": 14.324336051940918,
+      "learning_rate": 2.9609500000000002e-05,
+      "loss": 3.2593,
+      "step": 40800
+    },
+    {
+      "epoch": 0.817,
+      "grad_norm": 6.56348991394043,
+      "learning_rate": 2.9584500000000004e-05,
+      "loss": 3.2292,
+      "step": 40850
+    },
+    {
+      "epoch": 0.818,
+      "grad_norm": 7.05709981918335,
+      "learning_rate": 2.95595e-05,
+      "loss": 3.1446,
+      "step": 40900
+    },
+    {
+      "epoch": 0.819,
+      "grad_norm": 6.213717937469482,
+      "learning_rate": 2.9534500000000003e-05,
+      "loss": 3.1543,
+      "step": 40950
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.178669929504395,
+      "learning_rate": 2.9509500000000003e-05,
+      "loss": 3.0867,
+      "step": 41000
+    },
+    {
+      "epoch": 0.821,
+      "grad_norm": 7.343588829040527,
+      "learning_rate": 2.9484500000000005e-05,
+      "loss": 3.0533,
+      "step": 41050
+    },
+    {
+      "epoch": 0.822,
+      "grad_norm": 8.41699504852295,
+      "learning_rate": 2.94595e-05,
+      "loss": 2.9823,
+      "step": 41100
+    },
+    {
+      "epoch": 0.823,
+      "grad_norm": 6.462010860443115,
+      "learning_rate": 2.9434500000000004e-05,
+      "loss": 3.134,
+      "step": 41150
+    },
+    {
+      "epoch": 0.824,
+      "grad_norm": 7.462036609649658,
+      "learning_rate": 2.94095e-05,
+      "loss": 3.2841,
+      "step": 41200
+    },
+    {
+      "epoch": 0.825,
+      "grad_norm": 5.845949649810791,
+      "learning_rate": 2.9385e-05,
+      "loss": 3.1668,
+      "step": 41250
+    },
+    {
+      "epoch": 0.826,
+      "grad_norm": 8.043434143066406,
+      "learning_rate": 2.9360000000000003e-05,
+      "loss": 3.092,
+      "step": 41300
+    },
+    {
+      "epoch": 0.827,
+      "grad_norm": 6.7618842124938965,
+      "learning_rate": 2.9335000000000003e-05,
+      "loss": 3.2226,
+      "step": 41350
+    },
+    {
+      "epoch": 0.828,
+      "grad_norm": 5.803843021392822,
+      "learning_rate": 2.9310000000000006e-05,
+      "loss": 3.0952,
+      "step": 41400
+    },
+    {
+      "epoch": 0.829,
+      "grad_norm": 8.092726707458496,
+      "learning_rate": 2.9285e-05,
+      "loss": 3.338,
+      "step": 41450
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.214277267456055,
+      "learning_rate": 2.9260000000000004e-05,
+      "loss": 3.2117,
+      "step": 41500
+    },
+    {
+      "epoch": 0.831,
+      "grad_norm": 6.212014675140381,
+      "learning_rate": 2.9235e-05,
+      "loss": 3.1922,
+      "step": 41550
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 6.231184482574463,
+      "learning_rate": 2.9210000000000003e-05,
+      "loss": 3.1626,
+      "step": 41600
+    },
+    {
+      "epoch": 0.833,
+      "grad_norm": 8.141480445861816,
+      "learning_rate": 2.9185000000000003e-05,
+      "loss": 3.2031,
+      "step": 41650
+    },
+    {
+      "epoch": 0.834,
+      "grad_norm": 5.561233043670654,
+      "learning_rate": 2.9160000000000005e-05,
+      "loss": 3.2058,
+      "step": 41700
+    },
+    {
+      "epoch": 0.835,
+      "grad_norm": 6.850268363952637,
+      "learning_rate": 2.9135e-05,
+      "loss": 3.1208,
+      "step": 41750
+    },
+    {
+      "epoch": 0.836,
+      "grad_norm": 6.550021171569824,
+      "learning_rate": 2.9110000000000004e-05,
+      "loss": 3.1654,
+      "step": 41800
+    },
+    {
+      "epoch": 0.837,
+      "grad_norm": 6.2863359451293945,
+      "learning_rate": 2.9085e-05,
+      "loss": 2.9412,
+      "step": 41850
+    },
+    {
+      "epoch": 0.838,
+      "grad_norm": 7.536694526672363,
+      "learning_rate": 2.9060000000000003e-05,
+      "loss": 3.3024,
+      "step": 41900
+    },
+    {
+      "epoch": 0.839,
+      "grad_norm": 6.780869007110596,
+      "learning_rate": 2.9035000000000002e-05,
+      "loss": 3.0517,
+      "step": 41950
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.739259243011475,
+      "learning_rate": 2.9010000000000005e-05,
+      "loss": 3.2168,
+      "step": 42000
+    },
+    {
+      "epoch": 0.841,
+      "grad_norm": 6.765299320220947,
+      "learning_rate": 2.8985e-05,
+      "loss": 3.2832,
+      "step": 42050
+    },
+    {
+      "epoch": 0.842,
+      "grad_norm": 9.602505683898926,
+      "learning_rate": 2.8960000000000004e-05,
+      "loss": 3.1125,
+      "step": 42100
+    },
+    {
+      "epoch": 0.843,
+      "grad_norm": 8.207633972167969,
+      "learning_rate": 2.8935e-05,
+      "loss": 3.1905,
+      "step": 42150
+    },
+    {
+      "epoch": 0.844,
+      "grad_norm": 6.844604015350342,
+      "learning_rate": 2.8910000000000003e-05,
+      "loss": 3.1824,
+      "step": 42200
+    },
+    {
+      "epoch": 0.845,
+      "grad_norm": 7.132315158843994,
+      "learning_rate": 2.8885000000000002e-05,
+      "loss": 3.197,
+      "step": 42250
+    },
+    {
+      "epoch": 0.846,
+      "grad_norm": 8.794624328613281,
+      "learning_rate": 2.8860000000000005e-05,
+      "loss": 3.074,
+      "step": 42300
+    },
+    {
+      "epoch": 0.847,
+      "grad_norm": 6.943325042724609,
+      "learning_rate": 2.8835e-05,
+      "loss": 3.1916,
+      "step": 42350
+    },
+    {
+      "epoch": 0.848,
+      "grad_norm": 6.533550262451172,
+      "learning_rate": 2.8810000000000004e-05,
+      "loss": 3.1118,
+      "step": 42400
+    },
+    {
+      "epoch": 0.849,
+      "grad_norm": 10.443181991577148,
+      "learning_rate": 2.8785e-05,
+      "loss": 3.0385,
+      "step": 42450
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.8528642654418945,
+      "learning_rate": 2.8760000000000002e-05,
+      "loss": 3.301,
+      "step": 42500
+    },
+    {
+      "epoch": 0.851,
+      "grad_norm": 6.201817512512207,
+      "learning_rate": 2.8735000000000002e-05,
+      "loss": 3.1458,
+      "step": 42550
+    },
+    {
+      "epoch": 0.852,
+      "grad_norm": 5.921964645385742,
+      "learning_rate": 2.8710000000000005e-05,
+      "loss": 3.0353,
+      "step": 42600
+    },
+    {
+      "epoch": 0.853,
+      "grad_norm": 5.661495685577393,
+      "learning_rate": 2.8685e-05,
+      "loss": 3.1163,
+      "step": 42650
+    },
+    {
+      "epoch": 0.854,
+      "grad_norm": 6.364812850952148,
+      "learning_rate": 2.8660000000000003e-05,
+      "loss": 3.1634,
+      "step": 42700
+    },
+    {
+      "epoch": 0.855,
+      "grad_norm": 7.702561378479004,
+      "learning_rate": 2.8635e-05,
+      "loss": 3.2006,
+      "step": 42750
+    },
+    {
+      "epoch": 0.856,
+      "grad_norm": 7.330812454223633,
+      "learning_rate": 2.8610000000000002e-05,
+      "loss": 3.0284,
+      "step": 42800
+    },
+    {
+      "epoch": 0.857,
+      "grad_norm": 7.044076919555664,
+      "learning_rate": 2.8585e-05,
+      "loss": 2.958,
+      "step": 42850
+    },
+    {
+      "epoch": 0.858,
+      "grad_norm": 9.742399215698242,
+      "learning_rate": 2.8560000000000004e-05,
+      "loss": 3.0941,
+      "step": 42900
+    },
+    {
+      "epoch": 0.859,
+      "grad_norm": 6.272192001342773,
+      "learning_rate": 2.8535e-05,
+      "loss": 3.2671,
+      "step": 42950
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.866217136383057,
+      "learning_rate": 2.8510000000000003e-05,
+      "loss": 3.099,
+      "step": 43000
+    },
+    {
+      "epoch": 0.861,
+      "grad_norm": 6.47220516204834,
+      "learning_rate": 2.8485e-05,
+      "loss": 3.1561,
+      "step": 43050
+    },
+    {
+      "epoch": 0.862,
+      "grad_norm": 7.465406894683838,
+      "learning_rate": 2.8460000000000002e-05,
+      "loss": 2.9667,
+      "step": 43100
+    },
+    {
+      "epoch": 0.863,
+      "grad_norm": 6.800641059875488,
+      "learning_rate": 2.8435e-05,
+      "loss": 3.1791,
+      "step": 43150
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 7.830578804016113,
+      "learning_rate": 2.8410000000000004e-05,
+      "loss": 2.9727,
+      "step": 43200
+    },
+    {
+      "epoch": 0.865,
+      "grad_norm": 6.879873275756836,
+      "learning_rate": 2.8385500000000005e-05,
+      "loss": 3.1912,
+      "step": 43250
+    },
+    {
+      "epoch": 0.866,
+      "grad_norm": 6.994164943695068,
+      "learning_rate": 2.83605e-05,
+      "loss": 3.1734,
+      "step": 43300
+    },
+    {
+      "epoch": 0.867,
+      "grad_norm": 8.127955436706543,
+      "learning_rate": 2.8335500000000003e-05,
+      "loss": 3.1977,
+      "step": 43350
+    },
+    {
+      "epoch": 0.868,
+      "grad_norm": 6.18363618850708,
+      "learning_rate": 2.83105e-05,
+      "loss": 2.9296,
+      "step": 43400
+    },
+    {
+      "epoch": 0.869,
+      "grad_norm": 6.684689044952393,
+      "learning_rate": 2.8285500000000002e-05,
+      "loss": 3.0261,
+      "step": 43450
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.57409954071045,
+      "learning_rate": 2.82605e-05,
+      "loss": 3.2812,
+      "step": 43500
+    },
+    {
+      "epoch": 0.871,
+      "grad_norm": 6.665359020233154,
+      "learning_rate": 2.8235500000000004e-05,
+      "loss": 3.1191,
+      "step": 43550
+    },
+    {
+      "epoch": 0.872,
+      "grad_norm": 7.3514533042907715,
+      "learning_rate": 2.82105e-05,
+      "loss": 3.0988,
+      "step": 43600
+    },
+    {
+      "epoch": 0.873,
+      "grad_norm": 7.701603889465332,
+      "learning_rate": 2.8185500000000003e-05,
+      "loss": 3.2266,
+      "step": 43650
+    },
+    {
+      "epoch": 0.874,
+      "grad_norm": 6.6192851066589355,
+      "learning_rate": 2.81605e-05,
+      "loss": 3.2071,
+      "step": 43700
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 7.822471618652344,
+      "learning_rate": 2.8135500000000002e-05,
+      "loss": 3.2628,
+      "step": 43750
+    },
+    {
+      "epoch": 0.876,
+      "grad_norm": 6.103046417236328,
+      "learning_rate": 2.81105e-05,
+      "loss": 3.1475,
+      "step": 43800
+    },
+    {
+      "epoch": 0.877,
+      "grad_norm": 9.693628311157227,
+      "learning_rate": 2.8085500000000004e-05,
+      "loss": 3.2774,
+      "step": 43850
+    },
+    {
+      "epoch": 0.878,
+      "grad_norm": 8.551675796508789,
+      "learning_rate": 2.8061000000000005e-05,
+      "loss": 3.2343,
+      "step": 43900
+    },
+    {
+      "epoch": 0.879,
+      "grad_norm": 7.0858025550842285,
+      "learning_rate": 2.8036e-05,
+      "loss": 3.2596,
+      "step": 43950
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.832662105560303,
+      "learning_rate": 2.8011000000000003e-05,
+      "loss": 3.1012,
+      "step": 44000
+    },
+    {
+      "epoch": 0.881,
+      "grad_norm": 8.389840126037598,
+      "learning_rate": 2.7986000000000003e-05,
+      "loss": 3.1796,
+      "step": 44050
+    },
+    {
+      "epoch": 0.882,
+      "grad_norm": 10.351311683654785,
+      "learning_rate": 2.7961000000000006e-05,
+      "loss": 2.9221,
+      "step": 44100
+    },
+    {
+      "epoch": 0.883,
+      "grad_norm": 6.427529811859131,
+      "learning_rate": 2.7936e-05,
+      "loss": 2.9733,
+      "step": 44150
+    },
+    {
+      "epoch": 0.884,
+      "grad_norm": 6.574181079864502,
+      "learning_rate": 2.7911000000000004e-05,
+      "loss": 3.1097,
+      "step": 44200
+    },
+    {
+      "epoch": 0.885,
+      "grad_norm": 7.591128826141357,
+      "learning_rate": 2.7886e-05,
+      "loss": 3.1482,
+      "step": 44250
+    },
+    {
+      "epoch": 0.886,
+      "grad_norm": 7.332815647125244,
+      "learning_rate": 2.7861000000000003e-05,
+      "loss": 3.1352,
+      "step": 44300
+    },
+    {
+      "epoch": 0.887,
+      "grad_norm": 8.689337730407715,
+      "learning_rate": 2.7836000000000003e-05,
+      "loss": 3.1673,
+      "step": 44350
+    },
+    {
+      "epoch": 0.888,
+      "grad_norm": 6.476491451263428,
+      "learning_rate": 2.7811000000000005e-05,
+      "loss": 3.0232,
+      "step": 44400
+    },
+    {
+      "epoch": 0.889,
+      "grad_norm": 5.135951519012451,
+      "learning_rate": 2.7786e-05,
+      "loss": 3.1024,
+      "step": 44450
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.735162734985352,
+      "learning_rate": 2.7761000000000004e-05,
+      "loss": 3.1889,
+      "step": 44500
+    },
+    {
+      "epoch": 0.891,
+      "grad_norm": 7.481966972351074,
+      "learning_rate": 2.7736e-05,
+      "loss": 3.0462,
+      "step": 44550
+    },
+    {
+      "epoch": 0.892,
+      "grad_norm": 7.0432844161987305,
+      "learning_rate": 2.7711000000000003e-05,
+      "loss": 3.1846,
+      "step": 44600
+    },
+    {
+      "epoch": 0.893,
+      "grad_norm": 9.478955268859863,
+      "learning_rate": 2.7686000000000002e-05,
+      "loss": 3.2704,
+      "step": 44650
+    },
+    {
+      "epoch": 0.894,
+      "grad_norm": 6.461492538452148,
+      "learning_rate": 2.7661000000000005e-05,
+      "loss": 3.1446,
+      "step": 44700
+    },
+    {
+      "epoch": 0.895,
+      "grad_norm": 7.2203168869018555,
+      "learning_rate": 2.7636e-05,
+      "loss": 3.2167,
+      "step": 44750
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 9.509076118469238,
+      "learning_rate": 2.7611000000000004e-05,
+      "loss": 2.979,
+      "step": 44800
+    },
+    {
+      "epoch": 0.897,
+      "grad_norm": 6.559743404388428,
+      "learning_rate": 2.7586e-05,
+      "loss": 3.1303,
+      "step": 44850
+    },
+    {
+      "epoch": 0.898,
+      "grad_norm": 9.804462432861328,
+      "learning_rate": 2.7561000000000003e-05,
+      "loss": 3.1583,
+      "step": 44900
+    },
+    {
+      "epoch": 0.899,
+      "grad_norm": 7.840511322021484,
+      "learning_rate": 2.7536000000000002e-05,
+      "loss": 3.1117,
+      "step": 44950
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.799412250518799,
+      "learning_rate": 2.7511000000000005e-05,
+      "loss": 2.9047,
+      "step": 45000
+    },
+    {
+      "epoch": 0.901,
+      "grad_norm": 7.972790241241455,
+      "learning_rate": 2.7486e-05,
+      "loss": 3.0451,
+      "step": 45050
+    },
+    {
+      "epoch": 0.902,
+      "grad_norm": 10.409866333007812,
+      "learning_rate": 2.7461000000000004e-05,
+      "loss": 3.0812,
+      "step": 45100
+    },
+    {
+      "epoch": 0.903,
+      "grad_norm": 6.936646938323975,
+      "learning_rate": 2.7436e-05,
+      "loss": 3.1419,
+      "step": 45150
+    },
+    {
+      "epoch": 0.904,
+      "grad_norm": 10.039634704589844,
+      "learning_rate": 2.7411000000000002e-05,
+      "loss": 3.0438,
+      "step": 45200
+    },
+    {
+      "epoch": 0.905,
+      "grad_norm": 6.061675548553467,
+      "learning_rate": 2.7386000000000002e-05,
+      "loss": 3.0916,
+      "step": 45250
+    },
+    {
+      "epoch": 0.906,
+      "grad_norm": 8.285578727722168,
+      "learning_rate": 2.7361000000000005e-05,
+      "loss": 3.1964,
+      "step": 45300
+    },
+    {
+      "epoch": 0.907,
+      "grad_norm": 7.144978046417236,
+      "learning_rate": 2.7336e-05,
+      "loss": 2.9567,
+      "step": 45350
+    },
+    {
+      "epoch": 0.908,
+      "grad_norm": 8.918584823608398,
+      "learning_rate": 2.7311000000000003e-05,
+      "loss": 2.9987,
+      "step": 45400
+    },
+    {
+      "epoch": 0.909,
+      "grad_norm": 6.627841949462891,
+      "learning_rate": 2.7286e-05,
+      "loss": 3.1431,
+      "step": 45450
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.455625057220459,
+      "learning_rate": 2.7261000000000002e-05,
+      "loss": 3.102,
+      "step": 45500
+    },
+    {
+      "epoch": 0.911,
+      "grad_norm": 8.264837265014648,
+      "learning_rate": 2.7236e-05,
+      "loss": 3.1442,
+      "step": 45550
+    },
+    {
+      "epoch": 0.912,
+      "grad_norm": 9.39218521118164,
+      "learning_rate": 2.7211000000000004e-05,
+      "loss": 3.0657,
+      "step": 45600
+    },
+    {
+      "epoch": 0.913,
+      "grad_norm": 8.049072265625,
+      "learning_rate": 2.7186e-05,
+      "loss": 3.0677,
+      "step": 45650
+    },
+    {
+      "epoch": 0.914,
+      "grad_norm": 6.083390235900879,
+      "learning_rate": 2.7161000000000003e-05,
+      "loss": 3.018,
+      "step": 45700
+    },
+    {
+      "epoch": 0.915,
+      "grad_norm": 6.058326244354248,
+      "learning_rate": 2.7136e-05,
+      "loss": 2.9001,
+      "step": 45750
+    },
+    {
+      "epoch": 0.916,
+      "grad_norm": 6.957967281341553,
+      "learning_rate": 2.7111000000000002e-05,
+      "loss": 3.1806,
+      "step": 45800
+    },
+    {
+      "epoch": 0.917,
+      "grad_norm": 9.043895721435547,
+      "learning_rate": 2.7086e-05,
+      "loss": 3.0095,
+      "step": 45850
+    },
+    {
+      "epoch": 0.918,
+      "grad_norm": 8.315635681152344,
+      "learning_rate": 2.7061000000000004e-05,
+      "loss": 3.2231,
+      "step": 45900
+    },
+    {
+      "epoch": 0.919,
+      "grad_norm": 6.859596252441406,
+      "learning_rate": 2.7036e-05,
+      "loss": 3.1016,
+      "step": 45950
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.305853843688965,
+      "learning_rate": 2.7011000000000003e-05,
+      "loss": 3.1801,
+      "step": 46000
+    },
+    {
+      "epoch": 0.921,
+      "grad_norm": 7.9762797355651855,
+      "learning_rate": 2.6986e-05,
+      "loss": 2.9616,
+      "step": 46050
+    },
+    {
+      "epoch": 0.922,
+      "grad_norm": 6.378198623657227,
+      "learning_rate": 2.6961e-05,
+      "loss": 3.1772,
+      "step": 46100
+    },
+    {
+      "epoch": 0.923,
+      "grad_norm": 7.276421070098877,
+      "learning_rate": 2.6936e-05,
+      "loss": 3.1461,
+      "step": 46150
+    },
+    {
+      "epoch": 0.924,
+      "grad_norm": 7.234610080718994,
+      "learning_rate": 2.6911000000000004e-05,
+      "loss": 2.9373,
+      "step": 46200
+    },
+    {
+      "epoch": 0.925,
+      "grad_norm": 6.928109169006348,
+      "learning_rate": 2.6886e-05,
+      "loss": 3.0292,
+      "step": 46250
+    },
+    {
+      "epoch": 0.926,
+      "grad_norm": 8.218812942504883,
+      "learning_rate": 2.6861000000000003e-05,
+      "loss": 3.0692,
+      "step": 46300
+    },
+    {
+      "epoch": 0.927,
+      "grad_norm": 6.951169490814209,
+      "learning_rate": 2.6836e-05,
+      "loss": 2.983,
+      "step": 46350
+    },
+    {
+      "epoch": 0.928,
+      "grad_norm": 8.088844299316406,
+      "learning_rate": 2.6811e-05,
+      "loss": 3.0935,
+      "step": 46400
+    },
+    {
+      "epoch": 0.929,
+      "grad_norm": 6.722114086151123,
+      "learning_rate": 2.6786e-05,
+      "loss": 3.0447,
+      "step": 46450
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.844832420349121,
+      "learning_rate": 2.6761000000000003e-05,
+      "loss": 3.1378,
+      "step": 46500
+    },
+    {
+      "epoch": 0.931,
+      "grad_norm": 5.869866371154785,
+      "learning_rate": 2.6736e-05,
+      "loss": 3.0179,
+      "step": 46550
+    },
+    {
+      "epoch": 0.932,
+      "grad_norm": 6.477950096130371,
+      "learning_rate": 2.6711000000000002e-05,
+      "loss": 2.9879,
+      "step": 46600
+    },
+    {
+      "epoch": 0.933,
+      "grad_norm": 8.349437713623047,
+      "learning_rate": 2.6685999999999998e-05,
+      "loss": 3.0,
+      "step": 46650
+    },
+    {
+      "epoch": 0.934,
+      "grad_norm": 6.469250679016113,
+      "learning_rate": 2.6661e-05,
+      "loss": 3.1057,
+      "step": 46700
+    },
+    {
+      "epoch": 0.935,
+      "grad_norm": 6.10085391998291,
+      "learning_rate": 2.6636e-05,
+      "loss": 3.0944,
+      "step": 46750
+    },
+    {
+      "epoch": 0.936,
+      "grad_norm": 7.693483352661133,
+      "learning_rate": 2.6611000000000003e-05,
+      "loss": 3.1472,
+      "step": 46800
+    },
+    {
+      "epoch": 0.937,
+      "grad_norm": 6.2554030418396,
+      "learning_rate": 2.6586e-05,
+      "loss": 3.1175,
+      "step": 46850
+    },
+    {
+      "epoch": 0.938,
+      "grad_norm": 10.338006973266602,
+      "learning_rate": 2.6561000000000002e-05,
+      "loss": 2.9336,
+      "step": 46900
+    },
+    {
+      "epoch": 0.939,
+      "grad_norm": 6.1178765296936035,
+      "learning_rate": 2.6536e-05,
+      "loss": 3.1405,
+      "step": 46950
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.21044635772705,
+      "learning_rate": 2.6511000000000004e-05,
+      "loss": 3.046,
+      "step": 47000
+    },
+    {
+      "epoch": 0.941,
+      "grad_norm": 7.216266632080078,
+      "learning_rate": 2.6486e-05,
+      "loss": 3.129,
+      "step": 47050
+    },
+    {
+      "epoch": 0.942,
+      "grad_norm": 7.356600761413574,
+      "learning_rate": 2.6461000000000003e-05,
+      "loss": 3.0074,
+      "step": 47100
+    },
+    {
+      "epoch": 0.943,
+      "grad_norm": 7.9352264404296875,
+      "learning_rate": 2.6436e-05,
+      "loss": 2.9326,
+      "step": 47150
+    },
+    {
+      "epoch": 0.944,
+      "grad_norm": 6.63361120223999,
+      "learning_rate": 2.6411000000000002e-05,
+      "loss": 3.1064,
+      "step": 47200
+    },
+    {
+      "epoch": 0.945,
+      "grad_norm": 7.0836286544799805,
+      "learning_rate": 2.6386e-05,
+      "loss": 2.981,
+      "step": 47250
+    },
+    {
+      "epoch": 0.946,
+      "grad_norm": 6.541557312011719,
+      "learning_rate": 2.6361000000000004e-05,
+      "loss": 3.0636,
+      "step": 47300
+    },
+    {
+      "epoch": 0.947,
+      "grad_norm": 9.376893043518066,
+      "learning_rate": 2.6336e-05,
+      "loss": 2.9022,
+      "step": 47350
+    },
+    {
+      "epoch": 0.948,
+      "grad_norm": 6.6397929191589355,
+      "learning_rate": 2.6311000000000003e-05,
+      "loss": 3.1219,
+      "step": 47400
+    },
+    {
+      "epoch": 0.949,
+      "grad_norm": 7.550232410430908,
+      "learning_rate": 2.6286e-05,
+      "loss": 3.3295,
+      "step": 47450
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.387320041656494,
+      "learning_rate": 2.6261e-05,
+      "loss": 3.1274,
+      "step": 47500
+    },
+    {
+      "epoch": 0.951,
+      "grad_norm": 6.654208183288574,
+      "learning_rate": 2.6236e-05,
+      "loss": 2.8878,
+      "step": 47550
+    },
+    {
+      "epoch": 0.952,
+      "grad_norm": 7.0465497970581055,
+      "learning_rate": 2.6211000000000004e-05,
+      "loss": 3.3068,
+      "step": 47600
+    },
+    {
+      "epoch": 0.953,
+      "grad_norm": 8.581622123718262,
+      "learning_rate": 2.6186e-05,
+      "loss": 2.9106,
+      "step": 47650
+    },
+    {
+      "epoch": 0.954,
+      "grad_norm": 7.229891777038574,
+      "learning_rate": 2.6161000000000002e-05,
+      "loss": 2.9841,
+      "step": 47700
+    },
+    {
+      "epoch": 0.955,
+      "grad_norm": 7.919389247894287,
+      "learning_rate": 2.6136e-05,
+      "loss": 2.9297,
+      "step": 47750
+    },
+    {
+      "epoch": 0.956,
+      "grad_norm": 7.118837833404541,
+      "learning_rate": 2.6111e-05,
+      "loss": 2.8457,
+      "step": 47800
+    },
+    {
+      "epoch": 0.957,
+      "grad_norm": 6.714677333831787,
+      "learning_rate": 2.6086e-05,
+      "loss": 2.9976,
+      "step": 47850
+    },
+    {
+      "epoch": 0.958,
+      "grad_norm": 9.282464027404785,
+      "learning_rate": 2.6061000000000003e-05,
+      "loss": 3.017,
+      "step": 47900
+    },
+    {
+      "epoch": 0.959,
+      "grad_norm": 6.635936260223389,
+      "learning_rate": 2.6036500000000004e-05,
+      "loss": 3.248,
+      "step": 47950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.59296178817749,
+      "learning_rate": 2.60115e-05,
+      "loss": 3.0149,
+      "step": 48000
+    },
+    {
+      "epoch": 0.961,
+      "grad_norm": 6.398571014404297,
+      "learning_rate": 2.5986500000000003e-05,
+      "loss": 3.2064,
+      "step": 48050
+    },
+    {
+      "epoch": 0.962,
+      "grad_norm": 8.443436622619629,
+      "learning_rate": 2.59615e-05,
+      "loss": 3.086,
+      "step": 48100
+    },
+    {
+      "epoch": 0.963,
+      "grad_norm": 6.965825080871582,
+      "learning_rate": 2.59365e-05,
+      "loss": 3.0896,
+      "step": 48150
+    },
+    {
+      "epoch": 0.964,
+      "grad_norm": 7.949388027191162,
+      "learning_rate": 2.59115e-05,
+      "loss": 3.1704,
+      "step": 48200
+    },
+    {
+      "epoch": 0.965,
+      "grad_norm": 8.066429138183594,
+      "learning_rate": 2.5886500000000004e-05,
+      "loss": 3.0161,
+      "step": 48250
+    },
+    {
+      "epoch": 0.966,
+      "grad_norm": 8.622272491455078,
+      "learning_rate": 2.58615e-05,
+      "loss": 3.1115,
+      "step": 48300
+    },
+    {
+      "epoch": 0.967,
+      "grad_norm": 6.805866241455078,
+      "learning_rate": 2.5836500000000002e-05,
+      "loss": 3.2632,
+      "step": 48350
+    },
+    {
+      "epoch": 0.968,
+      "grad_norm": 6.861439228057861,
+      "learning_rate": 2.58115e-05,
+      "loss": 3.139,
+      "step": 48400
+    },
+    {
+      "epoch": 0.969,
+      "grad_norm": 8.325223922729492,
+      "learning_rate": 2.57865e-05,
+      "loss": 3.0783,
+      "step": 48450
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.219725131988525,
+      "learning_rate": 2.57615e-05,
+      "loss": 3.1166,
+      "step": 48500
+    },
+    {
+      "epoch": 0.971,
+      "grad_norm": 7.012028217315674,
+      "learning_rate": 2.5736500000000003e-05,
+      "loss": 3.1029,
+      "step": 48550
+    },
+    {
+      "epoch": 0.972,
+      "grad_norm": 7.103126525878906,
+      "learning_rate": 2.57115e-05,
+      "loss": 3.0094,
+      "step": 48600
+    },
+    {
+      "epoch": 0.973,
+      "grad_norm": 8.089792251586914,
+      "learning_rate": 2.5686500000000002e-05,
+      "loss": 3.0782,
+      "step": 48650
+    },
+    {
+      "epoch": 0.974,
+      "grad_norm": 8.976265907287598,
+      "learning_rate": 2.5661499999999998e-05,
+      "loss": 3.0816,
+      "step": 48700
+    },
+    {
+      "epoch": 0.975,
+      "grad_norm": 5.711278915405273,
+      "learning_rate": 2.56365e-05,
+      "loss": 3.0422,
+      "step": 48750
+    },
+    {
+      "epoch": 0.976,
+      "grad_norm": 8.107563972473145,
+      "learning_rate": 2.56115e-05,
+      "loss": 2.931,
+      "step": 48800
+    },
+    {
+      "epoch": 0.977,
+      "grad_norm": 6.930409908294678,
+      "learning_rate": 2.5586500000000003e-05,
+      "loss": 3.1535,
+      "step": 48850
+    },
+    {
+      "epoch": 0.978,
+      "grad_norm": 8.81825065612793,
+      "learning_rate": 2.55615e-05,
+      "loss": 3.1463,
+      "step": 48900
+    },
+    {
+      "epoch": 0.979,
+      "grad_norm": 7.318758010864258,
+      "learning_rate": 2.5536500000000002e-05,
+      "loss": 3.0381,
+      "step": 48950
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.180883884429932,
+      "learning_rate": 2.5511499999999998e-05,
+      "loss": 3.0863,
+      "step": 49000
+    },
+    {
+      "epoch": 0.981,
+      "grad_norm": 8.624285697937012,
+      "learning_rate": 2.54865e-05,
+      "loss": 3.1038,
+      "step": 49050
+    },
+    {
+      "epoch": 0.982,
+      "grad_norm": 7.267908573150635,
+      "learning_rate": 2.54615e-05,
+      "loss": 2.8967,
+      "step": 49100
+    },
+    {
+      "epoch": 0.983,
+      "grad_norm": 7.4274773597717285,
+      "learning_rate": 2.5436500000000003e-05,
+      "loss": 2.9911,
+      "step": 49150
+    },
+    {
+      "epoch": 0.984,
+      "grad_norm": 8.504932403564453,
+      "learning_rate": 2.54115e-05,
+      "loss": 3.2724,
+      "step": 49200
+    },
+    {
+      "epoch": 0.985,
+      "grad_norm": 7.094997882843018,
+      "learning_rate": 2.53865e-05,
+      "loss": 3.1187,
+      "step": 49250
+    },
+    {
+      "epoch": 0.986,
+      "grad_norm": 6.276993751525879,
+      "learning_rate": 2.5361499999999998e-05,
+      "loss": 3.0965,
+      "step": 49300
+    },
+    {
+      "epoch": 0.987,
+      "grad_norm": 8.081740379333496,
+      "learning_rate": 2.53365e-05,
+      "loss": 3.0091,
+      "step": 49350
+    },
+    {
+      "epoch": 0.988,
+      "grad_norm": 7.968372821807861,
+      "learning_rate": 2.53115e-05,
+      "loss": 3.2251,
+      "step": 49400
+    },
+    {
+      "epoch": 0.989,
+      "grad_norm": 8.792092323303223,
+      "learning_rate": 2.5286500000000003e-05,
+      "loss": 2.912,
+      "step": 49450
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.958602428436279,
+      "learning_rate": 2.52615e-05,
+      "loss": 3.1384,
+      "step": 49500
+    },
+    {
+      "epoch": 0.991,
+      "grad_norm": 7.0767059326171875,
+      "learning_rate": 2.52365e-05,
+      "loss": 3.0584,
+      "step": 49550
+    },
+    {
+      "epoch": 0.992,
+      "grad_norm": 7.639810085296631,
+      "learning_rate": 2.5211499999999997e-05,
+      "loss": 3.2435,
+      "step": 49600
+    },
+    {
+      "epoch": 0.993,
+      "grad_norm": 9.285743713378906,
+      "learning_rate": 2.51865e-05,
+      "loss": 3.1184,
+      "step": 49650
+    },
+    {
+      "epoch": 0.994,
+      "grad_norm": 6.950512409210205,
+      "learning_rate": 2.51615e-05,
+      "loss": 3.0376,
+      "step": 49700
+    },
+    {
+      "epoch": 0.995,
+      "grad_norm": 8.000362396240234,
+      "learning_rate": 2.5136500000000002e-05,
+      "loss": 3.1514,
+      "step": 49750
+    },
+    {
+      "epoch": 0.996,
+      "grad_norm": 7.425972938537598,
+      "learning_rate": 2.51115e-05,
+      "loss": 3.0469,
+      "step": 49800
+    },
+    {
+      "epoch": 0.997,
+      "grad_norm": 5.315979480743408,
+      "learning_rate": 2.50865e-05,
+      "loss": 3.0874,
+      "step": 49850
+    },
+    {
+      "epoch": 0.998,
+      "grad_norm": 6.839958190917969,
+      "learning_rate": 2.50615e-05,
+      "loss": 3.1059,
+      "step": 49900
+    },
+    {
+      "epoch": 0.999,
+      "grad_norm": 7.220871925354004,
+      "learning_rate": 2.5037e-05,
+      "loss": 3.0,
+      "step": 49950
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.067781448364258,
+      "learning_rate": 2.5012e-05,
+      "loss": 3.1688,
+      "step": 50000
+    },
+    {
+      "epoch": 1.001,
+      "grad_norm": 6.941652297973633,
+      "learning_rate": 2.4987000000000003e-05,
+      "loss": 3.0667,
+      "step": 50050
+    },
+    {
+      "epoch": 1.002,
+      "grad_norm": 6.937021732330322,
+      "learning_rate": 2.4962000000000003e-05,
+      "loss": 3.0059,
+      "step": 50100
+    },
+    {
+      "epoch": 1.003,
+      "grad_norm": 7.390977382659912,
+      "learning_rate": 2.4937000000000002e-05,
+      "loss": 2.9124,
+      "step": 50150
+    },
+    {
+      "epoch": 1.004,
+      "grad_norm": 8.082268714904785,
+      "learning_rate": 2.4912e-05,
+      "loss": 3.1063,
+      "step": 50200
+    },
+    {
+      "epoch": 1.005,
+      "grad_norm": 10.308112144470215,
+      "learning_rate": 2.4887e-05,
+      "loss": 3.0824,
+      "step": 50250
+    },
+    {
+      "epoch": 1.006,
+      "grad_norm": 5.8875298500061035,
+      "learning_rate": 2.4862000000000004e-05,
+      "loss": 2.9386,
+      "step": 50300
+    },
+    {
+      "epoch": 1.007,
+      "grad_norm": 8.643927574157715,
+      "learning_rate": 2.4837000000000003e-05,
+      "loss": 3.0654,
+      "step": 50350
+    },
+    {
+      "epoch": 1.008,
+      "grad_norm": 8.515275955200195,
+      "learning_rate": 2.4812000000000002e-05,
+      "loss": 2.8866,
+      "step": 50400
+    },
+    {
+      "epoch": 1.009,
+      "grad_norm": 7.331372261047363,
+      "learning_rate": 2.4787000000000002e-05,
+      "loss": 2.9803,
+      "step": 50450
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 8.064435958862305,
+      "learning_rate": 2.4762e-05,
+      "loss": 3.043,
+      "step": 50500
+    },
+    {
+      "epoch": 1.011,
+      "grad_norm": 6.975559711456299,
+      "learning_rate": 2.4737e-05,
+      "loss": 3.0534,
+      "step": 50550
+    },
+    {
+      "epoch": 1.012,
+      "grad_norm": 6.67125129699707,
+      "learning_rate": 2.4712000000000003e-05,
+      "loss": 3.022,
+      "step": 50600
+    },
+    {
+      "epoch": 1.013,
+      "grad_norm": 8.514412879943848,
+      "learning_rate": 2.4687000000000003e-05,
+      "loss": 3.0496,
+      "step": 50650
+    },
+    {
+      "epoch": 1.014,
+      "grad_norm": 6.08818244934082,
+      "learning_rate": 2.4662000000000002e-05,
+      "loss": 2.8703,
+      "step": 50700
+    },
+    {
+      "epoch": 1.015,
+      "grad_norm": 6.377881050109863,
+      "learning_rate": 2.4637e-05,
+      "loss": 2.8455,
+      "step": 50750
+    },
+    {
+      "epoch": 1.016,
+      "grad_norm": 6.975340366363525,
+      "learning_rate": 2.4612e-05,
+      "loss": 3.0349,
+      "step": 50800
+    },
+    {
+      "epoch": 1.017,
+      "grad_norm": 7.518290996551514,
+      "learning_rate": 2.4587e-05,
+      "loss": 2.9862,
+      "step": 50850
+    },
+    {
+      "epoch": 1.018,
+      "grad_norm": 6.725731372833252,
+      "learning_rate": 2.4562000000000003e-05,
+      "loss": 3.0017,
+      "step": 50900
+    },
+    {
+      "epoch": 1.019,
+      "grad_norm": 10.636852264404297,
+      "learning_rate": 2.4537000000000002e-05,
+      "loss": 2.911,
+      "step": 50950
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 12.086158752441406,
+      "learning_rate": 2.4512000000000002e-05,
+      "loss": 3.0008,
+      "step": 51000
+    },
+    {
+      "epoch": 1.021,
+      "grad_norm": 6.38085412979126,
+      "learning_rate": 2.4487e-05,
+      "loss": 3.1455,
+      "step": 51050
+    },
+    {
+      "epoch": 1.022,
+      "grad_norm": 5.900729179382324,
+      "learning_rate": 2.4462e-05,
+      "loss": 2.9867,
+      "step": 51100
+    },
+    {
+      "epoch": 1.023,
+      "grad_norm": 6.688613414764404,
+      "learning_rate": 2.4437e-05,
+      "loss": 2.8085,
+      "step": 51150
+    },
+    {
+      "epoch": 1.024,
+      "grad_norm": 7.4456281661987305,
+      "learning_rate": 2.4412000000000003e-05,
+      "loss": 3.0203,
+      "step": 51200
+    },
+    {
+      "epoch": 1.025,
+      "grad_norm": 7.295360088348389,
+      "learning_rate": 2.4387000000000002e-05,
+      "loss": 3.0503,
+      "step": 51250
+    },
+    {
+      "epoch": 1.026,
+      "grad_norm": 6.788741111755371,
+      "learning_rate": 2.4362e-05,
+      "loss": 3.059,
+      "step": 51300
+    },
+    {
+      "epoch": 1.027,
+      "grad_norm": 6.674519062042236,
+      "learning_rate": 2.4337e-05,
+      "loss": 3.1171,
+      "step": 51350
+    },
+    {
+      "epoch": 1.028,
+      "grad_norm": 7.097243309020996,
+      "learning_rate": 2.4312e-05,
+      "loss": 2.8427,
+      "step": 51400
+    },
+    {
+      "epoch": 1.029,
+      "grad_norm": 8.2136869430542,
+      "learning_rate": 2.4287000000000003e-05,
+      "loss": 2.9196,
+      "step": 51450
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 7.26527214050293,
+      "learning_rate": 2.4262000000000003e-05,
+      "loss": 3.1156,
+      "step": 51500
+    },
+    {
+      "epoch": 1.031,
+      "grad_norm": 8.166938781738281,
+      "learning_rate": 2.4237000000000002e-05,
+      "loss": 3.1713,
+      "step": 51550
+    },
+    {
+      "epoch": 1.032,
+      "grad_norm": 6.972194671630859,
+      "learning_rate": 2.4212e-05,
+      "loss": 2.8535,
+      "step": 51600
+    },
+    {
+      "epoch": 1.033,
+      "grad_norm": 9.820430755615234,
+      "learning_rate": 2.4187e-05,
+      "loss": 2.9944,
+      "step": 51650
+    },
+    {
+      "epoch": 1.034,
+      "grad_norm": 7.9938530921936035,
+      "learning_rate": 2.4162e-05,
+      "loss": 3.1709,
+      "step": 51700
+    },
+    {
+      "epoch": 1.035,
+      "grad_norm": 6.771377086639404,
+      "learning_rate": 2.4137000000000003e-05,
+      "loss": 3.1786,
+      "step": 51750
+    },
+    {
+      "epoch": 1.036,
+      "grad_norm": 8.01101016998291,
+      "learning_rate": 2.4112000000000002e-05,
+      "loss": 3.0196,
+      "step": 51800
+    },
+    {
+      "epoch": 1.037,
+      "grad_norm": 6.825372219085693,
+      "learning_rate": 2.4087e-05,
+      "loss": 3.1268,
+      "step": 51850
+    },
+    {
+      "epoch": 1.038,
+      "grad_norm": 6.796056747436523,
+      "learning_rate": 2.4062e-05,
+      "loss": 2.9989,
+      "step": 51900
+    },
+    {
+      "epoch": 1.039,
+      "grad_norm": Infinity,
+      "learning_rate": 2.4037e-05,
+      "loss": 2.8302,
+      "step": 51950
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 6.67830228805542,
+      "learning_rate": 2.40125e-05,
+      "loss": 3.0738,
+      "step": 52000
+    },
+    {
+      "epoch": 1.041,
+      "grad_norm": 7.3164143562316895,
+      "learning_rate": 2.39875e-05,
+      "loss": 3.1385,
+      "step": 52050
+    },
+    {
+      "epoch": 1.042,
+      "grad_norm": 8.837868690490723,
+      "learning_rate": 2.3962500000000003e-05,
+      "loss": 2.98,
+      "step": 52100
+    },
+    {
+      "epoch": 1.043,
+      "grad_norm": 7.212620258331299,
+      "learning_rate": 2.3937500000000002e-05,
+      "loss": 2.9773,
+      "step": 52150
+    },
+    {
+      "epoch": 1.044,
+      "grad_norm": 7.9372878074646,
+      "learning_rate": 2.3912500000000002e-05,
+      "loss": 3.0351,
+      "step": 52200
+    },
+    {
+      "epoch": 1.045,
+      "grad_norm": 8.558395385742188,
+      "learning_rate": 2.38875e-05,
+      "loss": 3.0283,
+      "step": 52250
+    },
+    {
+      "epoch": 1.046,
+      "grad_norm": 9.92737102508545,
+      "learning_rate": 2.38625e-05,
+      "loss": 2.715,
+      "step": 52300
+    },
+    {
+      "epoch": 1.047,
+      "grad_norm": 6.824094772338867,
+      "learning_rate": 2.38375e-05,
+      "loss": 2.8445,
+      "step": 52350
+    },
+    {
+      "epoch": 1.048,
+      "grad_norm": 7.356767654418945,
+      "learning_rate": 2.3812500000000003e-05,
+      "loss": 2.9878,
+      "step": 52400
+    },
+    {
+      "epoch": 1.049,
+      "grad_norm": 9.474124908447266,
+      "learning_rate": 2.3787500000000002e-05,
+      "loss": 2.8968,
+      "step": 52450
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 6.359553813934326,
+      "learning_rate": 2.37625e-05,
+      "loss": 2.8757,
+      "step": 52500
+    },
+    {
+      "epoch": 1.051,
+      "grad_norm": 7.467580795288086,
+      "learning_rate": 2.37375e-05,
+      "loss": 3.1544,
+      "step": 52550
+    },
+    {
+      "epoch": 1.052,
+      "grad_norm": 9.67093276977539,
+      "learning_rate": 2.37125e-05,
+      "loss": 3.0963,
+      "step": 52600
+    },
+    {
+      "epoch": 1.053,
+      "grad_norm": 8.854635238647461,
+      "learning_rate": 2.36875e-05,
+      "loss": 2.9702,
+      "step": 52650
+    },
+    {
+      "epoch": 1.054,
+      "grad_norm": 6.981558322906494,
+      "learning_rate": 2.3662500000000003e-05,
+      "loss": 3.143,
+      "step": 52700
+    },
+    {
+      "epoch": 1.055,
+      "grad_norm": 6.7144293785095215,
+      "learning_rate": 2.3637500000000002e-05,
+      "loss": 2.8273,
+      "step": 52750
+    },
+    {
+      "epoch": 1.056,
+      "grad_norm": 6.277219295501709,
+      "learning_rate": 2.36125e-05,
+      "loss": 2.9522,
+      "step": 52800
+    },
+    {
+      "epoch": 1.057,
+      "grad_norm": 9.357563972473145,
+      "learning_rate": 2.35875e-05,
+      "loss": 2.8644,
+      "step": 52850
+    },
+    {
+      "epoch": 1.058,
+      "grad_norm": 6.855267524719238,
+      "learning_rate": 2.35625e-05,
+      "loss": 2.9058,
+      "step": 52900
+    },
+    {
+      "epoch": 1.059,
+      "grad_norm": 6.813782691955566,
+      "learning_rate": 2.35375e-05,
+      "loss": 3.0752,
+      "step": 52950
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 7.318324565887451,
+      "learning_rate": 2.3512500000000002e-05,
+      "loss": 2.9552,
+      "step": 53000
+    },
+    {
+      "epoch": 1.061,
+      "grad_norm": 8.61180305480957,
+      "learning_rate": 2.3487500000000002e-05,
+      "loss": 2.9447,
+      "step": 53050
+    },
+    {
+      "epoch": 1.062,
+      "grad_norm": 7.704663276672363,
+      "learning_rate": 2.34625e-05,
+      "loss": 3.1072,
+      "step": 53100
+    },
+    {
+      "epoch": 1.063,
+      "grad_norm": 9.428034782409668,
+      "learning_rate": 2.34375e-05,
+      "loss": 2.9625,
+      "step": 53150
+    },
+    {
+      "epoch": 1.064,
+      "grad_norm": 8.153590202331543,
+      "learning_rate": 2.34125e-05,
+      "loss": 3.1441,
+      "step": 53200
+    },
+    {
+      "epoch": 1.065,
+      "grad_norm": 7.686233997344971,
+      "learning_rate": 2.3387500000000003e-05,
+      "loss": 2.977,
+      "step": 53250
+    },
+    {
+      "epoch": 1.066,
+      "grad_norm": 7.506038188934326,
+      "learning_rate": 2.3362500000000002e-05,
+      "loss": 2.9525,
+      "step": 53300
+    },
+    {
+      "epoch": 1.067,
+      "grad_norm": 7.652656078338623,
+      "learning_rate": 2.33375e-05,
+      "loss": 3.0001,
+      "step": 53350
+    },
+    {
+      "epoch": 1.068,
+      "grad_norm": 6.947332382202148,
+      "learning_rate": 2.33125e-05,
+      "loss": 2.9472,
+      "step": 53400
+    },
+    {
+      "epoch": 1.069,
+      "grad_norm": 7.201016902923584,
+      "learning_rate": 2.32875e-05,
+      "loss": 2.9012,
+      "step": 53450
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 7.253627777099609,
+      "learning_rate": 2.32625e-05,
+      "loss": 3.228,
+      "step": 53500
+    },
+    {
+      "epoch": 1.071,
+      "grad_norm": 7.386748313903809,
+      "learning_rate": 2.3237500000000002e-05,
+      "loss": 2.9931,
+      "step": 53550
+    },
+    {
+      "epoch": 1.072,
+      "grad_norm": 7.58491849899292,
+      "learning_rate": 2.3212500000000002e-05,
+      "loss": 3.0007,
+      "step": 53600
+    },
+    {
+      "epoch": 1.073,
+      "grad_norm": 8.3165922164917,
+      "learning_rate": 2.31875e-05,
+      "loss": 2.9794,
+      "step": 53650
+    },
+    {
+      "epoch": 1.074,
+      "grad_norm": 7.365329742431641,
+      "learning_rate": 2.31625e-05,
+      "loss": 2.9635,
+      "step": 53700
+    },
+    {
+      "epoch": 1.075,
+      "grad_norm": 6.985613822937012,
+      "learning_rate": 2.31375e-05,
+      "loss": 2.9755,
+      "step": 53750
+    },
+    {
+      "epoch": 1.076,
+      "grad_norm": 9.172526359558105,
+      "learning_rate": 2.31125e-05,
+      "loss": 2.8953,
+      "step": 53800
+    },
+    {
+      "epoch": 1.077,
+      "grad_norm": 9.151043891906738,
+      "learning_rate": 2.3087500000000002e-05,
+      "loss": 3.0558,
+      "step": 53850
+    },
+    {
+      "epoch": 1.078,
+      "grad_norm": 6.541784763336182,
+      "learning_rate": 2.30625e-05,
+      "loss": 2.8525,
+      "step": 53900
+    },
+    {
+      "epoch": 1.079,
+      "grad_norm": 8.261955261230469,
+      "learning_rate": 2.30375e-05,
+      "loss": 3.0542,
+      "step": 53950
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.869956970214844,
+      "learning_rate": 2.3013e-05,
+      "loss": 2.939,
+      "step": 54000
+    },
+    {
+      "epoch": 1.081,
+      "grad_norm": 7.964324474334717,
+      "learning_rate": 2.2988e-05,
+      "loss": 2.9387,
+      "step": 54050
+    },
+    {
+      "epoch": 1.082,
+      "grad_norm": 7.4598517417907715,
+      "learning_rate": 2.2963e-05,
+      "loss": 3.1512,
+      "step": 54100
+    },
+    {
+      "epoch": 1.083,
+      "grad_norm": 6.513663291931152,
+      "learning_rate": 2.2938e-05,
+      "loss": 2.9179,
+      "step": 54150
+    },
+    {
+      "epoch": 1.084,
+      "grad_norm": 9.144841194152832,
+      "learning_rate": 2.2913000000000002e-05,
+      "loss": 2.8925,
+      "step": 54200
+    },
+    {
+      "epoch": 1.085,
+      "grad_norm": 6.455650329589844,
+      "learning_rate": 2.2888000000000002e-05,
+      "loss": 2.8863,
+      "step": 54250
+    },
+    {
+      "epoch": 1.086,
+      "grad_norm": 7.622753620147705,
+      "learning_rate": 2.2863e-05,
+      "loss": 2.9365,
+      "step": 54300
+    },
+    {
+      "epoch": 1.087,
+      "grad_norm": 8.794417381286621,
+      "learning_rate": 2.2838e-05,
+      "loss": 2.9299,
+      "step": 54350
+    },
+    {
+      "epoch": 1.088,
+      "grad_norm": 7.10263204574585,
+      "learning_rate": 2.2813e-05,
+      "loss": 2.8352,
+      "step": 54400
+    },
+    {
+      "epoch": 1.089,
+      "grad_norm": 7.9989142417907715,
+      "learning_rate": 2.2788e-05,
+      "loss": 2.7885,
+      "step": 54450
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 7.510954856872559,
+      "learning_rate": 2.2763000000000002e-05,
+      "loss": 3.0037,
+      "step": 54500
+    },
+    {
+      "epoch": 1.091,
+      "grad_norm": 7.432366371154785,
+      "learning_rate": 2.2738e-05,
+      "loss": 3.1301,
+      "step": 54550
+    },
+    {
+      "epoch": 1.092,
+      "grad_norm": 7.562934398651123,
+      "learning_rate": 2.2713e-05,
+      "loss": 3.112,
+      "step": 54600
+    },
+    {
+      "epoch": 1.093,
+      "grad_norm": 7.627951622009277,
+      "learning_rate": 2.2688e-05,
+      "loss": 2.9647,
+      "step": 54650
+    },
+    {
+      "epoch": 1.094,
+      "grad_norm": 8.570850372314453,
+      "learning_rate": 2.2663e-05,
+      "loss": 2.9396,
+      "step": 54700
+    },
+    {
+      "epoch": 1.095,
+      "grad_norm": 7.013098239898682,
+      "learning_rate": 2.2638000000000002e-05,
+      "loss": 2.8855,
+      "step": 54750
+    },
+    {
+      "epoch": 1.096,
+      "grad_norm": 6.925095558166504,
+      "learning_rate": 2.2613000000000002e-05,
+      "loss": 2.8753,
+      "step": 54800
+    },
+    {
+      "epoch": 1.097,
+      "grad_norm": 8.075315475463867,
+      "learning_rate": 2.2588e-05,
+      "loss": 3.0105,
+      "step": 54850
+    },
+    {
+      "epoch": 1.098,
+      "grad_norm": 7.342935085296631,
+      "learning_rate": 2.2563e-05,
+      "loss": 3.0008,
+      "step": 54900
+    },
+    {
+      "epoch": 1.099,
+      "grad_norm": 6.944976806640625,
+      "learning_rate": 2.2538e-05,
+      "loss": 3.0758,
+      "step": 54950
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 7.054170608520508,
+      "learning_rate": 2.2513e-05,
+      "loss": 2.9099,
+      "step": 55000
+    },
+    {
+      "epoch": 1.101,
+      "grad_norm": 8.738309860229492,
+      "learning_rate": 2.2488000000000002e-05,
+      "loss": 3.031,
+      "step": 55050
+    },
+    {
+      "epoch": 1.102,
+      "grad_norm": 7.9822893142700195,
+      "learning_rate": 2.2463e-05,
+      "loss": 2.8642,
+      "step": 55100
+    },
+    {
+      "epoch": 1.103,
+      "grad_norm": 6.917319297790527,
+      "learning_rate": 2.2438e-05,
+      "loss": 3.0214,
+      "step": 55150
+    },
+    {
+      "epoch": 1.104,
+      "grad_norm": 8.734328269958496,
+      "learning_rate": 2.2413e-05,
+      "loss": 3.0149,
+      "step": 55200
+    },
+    {
+      "epoch": 1.105,
+      "grad_norm": 6.920499801635742,
+      "learning_rate": 2.2388e-05,
+      "loss": 2.7432,
+      "step": 55250
+    },
+    {
+      "epoch": 1.106,
+      "grad_norm": 6.7622809410095215,
+      "learning_rate": 2.2363e-05,
+      "loss": 2.7757,
+      "step": 55300
+    },
+    {
+      "epoch": 1.107,
+      "grad_norm": 8.261820793151855,
+      "learning_rate": 2.2338000000000002e-05,
+      "loss": 2.983,
+      "step": 55350
+    },
+    {
+      "epoch": 1.108,
+      "grad_norm": 7.027587890625,
+      "learning_rate": 2.2313e-05,
+      "loss": 2.8707,
+      "step": 55400
+    },
+    {
+      "epoch": 1.109,
+      "grad_norm": 6.786430835723877,
+      "learning_rate": 2.2288e-05,
+      "loss": 3.0307,
+      "step": 55450
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 7.964380264282227,
+      "learning_rate": 2.2263e-05,
+      "loss": 2.8678,
+      "step": 55500
+    },
+    {
+      "epoch": 1.111,
+      "grad_norm": 6.880892753601074,
+      "learning_rate": 2.2238e-05,
+      "loss": 2.9177,
+      "step": 55550
+    },
+    {
+      "epoch": 1.112,
+      "grad_norm": 8.703950881958008,
+      "learning_rate": 2.2213e-05,
+      "loss": 2.7289,
+      "step": 55600
+    },
+    {
+      "epoch": 1.113,
+      "grad_norm": 7.1236090660095215,
+      "learning_rate": 2.2188e-05,
+      "loss": 2.8935,
+      "step": 55650
+    },
+    {
+      "epoch": 1.114,
+      "grad_norm": 7.747103691101074,
+      "learning_rate": 2.2163e-05,
+      "loss": 2.805,
+      "step": 55700
+    },
+    {
+      "epoch": 1.115,
+      "grad_norm": 7.6112260818481445,
+      "learning_rate": 2.2138e-05,
+      "loss": 2.9563,
+      "step": 55750
+    },
+    {
+      "epoch": 1.116,
+      "grad_norm": 6.807455539703369,
+      "learning_rate": 2.2113e-05,
+      "loss": 2.9746,
+      "step": 55800
+    },
+    {
+      "epoch": 1.117,
+      "grad_norm": 7.226533889770508,
+      "learning_rate": 2.2088e-05,
+      "loss": 3.0062,
+      "step": 55850
+    },
+    {
+      "epoch": 1.1179999999999999,
+      "grad_norm": 7.821582794189453,
+      "learning_rate": 2.2063e-05,
+      "loss": 2.8594,
+      "step": 55900
+    },
+    {
+      "epoch": 1.119,
+      "grad_norm": 7.580191135406494,
+      "learning_rate": 2.2038e-05,
+      "loss": 2.9352,
+      "step": 55950
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.944276809692383,
+      "learning_rate": 2.2013500000000002e-05,
+      "loss": 2.8431,
+      "step": 56000
+    },
+    {
+      "epoch": 1.121,
+      "grad_norm": 8.267626762390137,
+      "learning_rate": 2.19885e-05,
+      "loss": 2.9671,
+      "step": 56050
+    },
+    {
+      "epoch": 1.1219999999999999,
+      "grad_norm": 6.602245330810547,
+      "learning_rate": 2.19635e-05,
+      "loss": 2.8601,
+      "step": 56100
+    },
+    {
+      "epoch": 1.123,
+      "grad_norm": 7.23159122467041,
+      "learning_rate": 2.19385e-05,
+      "loss": 3.2612,
+      "step": 56150
+    },
+    {
+      "epoch": 1.124,
+      "grad_norm": 8.431252479553223,
+      "learning_rate": 2.19135e-05,
+      "loss": 2.8907,
+      "step": 56200
+    },
+    {
+      "epoch": 1.125,
+      "grad_norm": 7.482148170471191,
+      "learning_rate": 2.1888500000000002e-05,
+      "loss": 2.8308,
+      "step": 56250
+    },
+    {
+      "epoch": 1.126,
+      "grad_norm": 7.137932300567627,
+      "learning_rate": 2.18635e-05,
+      "loss": 2.9463,
+      "step": 56300
+    },
+    {
+      "epoch": 1.127,
+      "grad_norm": 8.12203311920166,
+      "learning_rate": 2.18385e-05,
+      "loss": 3.0064,
+      "step": 56350
+    },
+    {
+      "epoch": 1.1280000000000001,
+      "grad_norm": 6.96235466003418,
+      "learning_rate": 2.18135e-05,
+      "loss": 2.7659,
+      "step": 56400
+    },
+    {
+      "epoch": 1.129,
+      "grad_norm": 6.59172248840332,
+      "learning_rate": 2.17885e-05,
+      "loss": 2.9026,
+      "step": 56450
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.312923431396484,
+      "learning_rate": 2.17635e-05,
+      "loss": 3.0246,
+      "step": 56500
+    },
+    {
+      "epoch": 1.131,
+      "grad_norm": 7.342343330383301,
+      "learning_rate": 2.1738500000000002e-05,
+      "loss": 2.9613,
+      "step": 56550
+    },
+    {
+      "epoch": 1.1320000000000001,
+      "grad_norm": 8.1234130859375,
+      "learning_rate": 2.17135e-05,
+      "loss": 2.9619,
+      "step": 56600
+    },
+    {
+      "epoch": 1.133,
+      "grad_norm": 7.191281318664551,
+      "learning_rate": 2.16885e-05,
+      "loss": 2.9403,
+      "step": 56650
+    },
+    {
+      "epoch": 1.134,
+      "grad_norm": 8.372027397155762,
+      "learning_rate": 2.16635e-05,
+      "loss": 3.0187,
+      "step": 56700
+    },
+    {
+      "epoch": 1.135,
+      "grad_norm": 6.654972553253174,
+      "learning_rate": 2.16385e-05,
+      "loss": 3.056,
+      "step": 56750
+    },
+    {
+      "epoch": 1.1360000000000001,
+      "grad_norm": 5.851999759674072,
+      "learning_rate": 2.16135e-05,
+      "loss": 2.8419,
+      "step": 56800
+    },
+    {
+      "epoch": 1.137,
+      "grad_norm": 8.843352317810059,
+      "learning_rate": 2.1588500000000002e-05,
+      "loss": 2.9252,
+      "step": 56850
+    },
+    {
+      "epoch": 1.138,
+      "grad_norm": 8.799443244934082,
+      "learning_rate": 2.15635e-05,
+      "loss": 2.9716,
+      "step": 56900
+    },
+    {
+      "epoch": 1.139,
+      "grad_norm": 7.772215843200684,
+      "learning_rate": 2.15385e-05,
+      "loss": 2.9954,
+      "step": 56950
+    },
+    {
+      "epoch": 1.1400000000000001,
+      "grad_norm": 8.679427146911621,
+      "learning_rate": 2.15135e-05,
+      "loss": 2.7281,
+      "step": 57000
+    },
+    {
+      "epoch": 1.141,
+      "grad_norm": 7.537428379058838,
+      "learning_rate": 2.14885e-05,
+      "loss": 3.0369,
+      "step": 57050
+    },
+    {
+      "epoch": 1.142,
+      "grad_norm": 7.318295478820801,
+      "learning_rate": 2.14635e-05,
+      "loss": 3.0333,
+      "step": 57100
+    },
+    {
+      "epoch": 1.143,
+      "grad_norm": 7.262914657592773,
+      "learning_rate": 2.14385e-05,
+      "loss": 2.8413,
+      "step": 57150
+    },
+    {
+      "epoch": 1.144,
+      "grad_norm": 7.274723529815674,
+      "learning_rate": 2.14135e-05,
+      "loss": 2.8387,
+      "step": 57200
+    },
+    {
+      "epoch": 1.145,
+      "grad_norm": 8.164852142333984,
+      "learning_rate": 2.13885e-05,
+      "loss": 2.7594,
+      "step": 57250
+    },
+    {
+      "epoch": 1.146,
+      "grad_norm": 7.805911540985107,
+      "learning_rate": 2.13635e-05,
+      "loss": 3.0005,
+      "step": 57300
+    },
+    {
+      "epoch": 1.147,
+      "grad_norm": 6.96591329574585,
+      "learning_rate": 2.13385e-05,
+      "loss": 2.8565,
+      "step": 57350
+    },
+    {
+      "epoch": 1.148,
+      "grad_norm": 7.270416259765625,
+      "learning_rate": 2.13135e-05,
+      "loss": 3.0091,
+      "step": 57400
+    },
+    {
+      "epoch": 1.149,
+      "grad_norm": 7.759993076324463,
+      "learning_rate": 2.12885e-05,
+      "loss": 2.9593,
+      "step": 57450
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.487631320953369,
+      "learning_rate": 2.12635e-05,
+      "loss": 2.9263,
+      "step": 57500
+    },
+    {
+      "epoch": 1.151,
+      "grad_norm": 6.525301456451416,
+      "learning_rate": 2.12385e-05,
+      "loss": 2.8832,
+      "step": 57550
+    },
+    {
+      "epoch": 1.152,
+      "grad_norm": 7.14619779586792,
+      "learning_rate": 2.12135e-05,
+      "loss": 2.8469,
+      "step": 57600
+    },
+    {
+      "epoch": 1.153,
+      "grad_norm": 8.395821571350098,
+      "learning_rate": 2.11885e-05,
+      "loss": 2.9667,
+      "step": 57650
+    },
+    {
+      "epoch": 1.154,
+      "grad_norm": 7.4564900398254395,
+      "learning_rate": 2.11635e-05,
+      "loss": 2.9125,
+      "step": 57700
+    },
+    {
+      "epoch": 1.155,
+      "grad_norm": 7.6553120613098145,
+      "learning_rate": 2.11385e-05,
+      "loss": 3.046,
+      "step": 57750
+    },
+    {
+      "epoch": 1.156,
+      "grad_norm": 7.374424934387207,
+      "learning_rate": 2.11135e-05,
+      "loss": 2.7586,
+      "step": 57800
+    },
+    {
+      "epoch": 1.157,
+      "grad_norm": 8.02834415435791,
+      "learning_rate": 2.10885e-05,
+      "loss": 2.8078,
+      "step": 57850
+    },
+    {
+      "epoch": 1.158,
+      "grad_norm": 8.8065767288208,
+      "learning_rate": 2.10635e-05,
+      "loss": 2.7314,
+      "step": 57900
+    },
+    {
+      "epoch": 1.159,
+      "grad_norm": 7.677451133728027,
+      "learning_rate": 2.10385e-05,
+      "loss": 2.8307,
+      "step": 57950
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.3158464431762695,
+      "learning_rate": 2.10135e-05,
+      "loss": 2.8309,
+      "step": 58000
+    },
+    {
+      "epoch": 1.161,
+      "grad_norm": 8.057973861694336,
+      "learning_rate": 2.0989000000000002e-05,
+      "loss": 2.8944,
+      "step": 58050
+    },
+    {
+      "epoch": 1.162,
+      "grad_norm": 8.379156112670898,
+      "learning_rate": 2.0964e-05,
+      "loss": 2.9813,
+      "step": 58100
+    },
+    {
+      "epoch": 1.163,
+      "grad_norm": 6.20780611038208,
+      "learning_rate": 2.0939e-05,
+      "loss": 2.8746,
+      "step": 58150
+    },
+    {
+      "epoch": 1.164,
+      "grad_norm": 6.247302532196045,
+      "learning_rate": 2.0914e-05,
+      "loss": 3.1069,
+      "step": 58200
+    },
+    {
+      "epoch": 1.165,
+      "grad_norm": 10.308443069458008,
+      "learning_rate": 2.0889e-05,
+      "loss": 2.9991,
+      "step": 58250
+    },
+    {
+      "epoch": 1.166,
+      "grad_norm": 7.055253982543945,
+      "learning_rate": 2.0864e-05,
+      "loss": 3.0063,
+      "step": 58300
+    },
+    {
+      "epoch": 1.167,
+      "grad_norm": 8.708407402038574,
+      "learning_rate": 2.0839e-05,
+      "loss": 3.0384,
+      "step": 58350
+    },
+    {
+      "epoch": 1.168,
+      "grad_norm": 9.416001319885254,
+      "learning_rate": 2.0814e-05,
+      "loss": 3.0263,
+      "step": 58400
+    },
+    {
+      "epoch": 1.169,
+      "grad_norm": 7.336208820343018,
+      "learning_rate": 2.0789e-05,
+      "loss": 2.8834,
+      "step": 58450
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 7.314346790313721,
+      "learning_rate": 2.0764e-05,
+      "loss": 2.9417,
+      "step": 58500
+    },
+    {
+      "epoch": 1.171,
+      "grad_norm": 7.871457576751709,
+      "learning_rate": 2.0739e-05,
+      "loss": 2.7274,
+      "step": 58550
+    },
+    {
+      "epoch": 1.172,
+      "grad_norm": 7.405073165893555,
+      "learning_rate": 2.0714e-05,
+      "loss": 2.9292,
+      "step": 58600
+    },
+    {
+      "epoch": 1.173,
+      "grad_norm": 7.039045333862305,
+      "learning_rate": 2.0689e-05,
+      "loss": 2.9214,
+      "step": 58650
+    },
+    {
+      "epoch": 1.174,
+      "grad_norm": 8.36693000793457,
+      "learning_rate": 2.0664e-05,
+      "loss": 2.7912,
+      "step": 58700
+    },
+    {
+      "epoch": 1.175,
+      "grad_norm": 6.045433521270752,
+      "learning_rate": 2.0639e-05,
+      "loss": 2.9735,
+      "step": 58750
+    },
+    {
+      "epoch": 1.176,
+      "grad_norm": 6.981970310211182,
+      "learning_rate": 2.0614e-05,
+      "loss": 3.0241,
+      "step": 58800
+    },
+    {
+      "epoch": 1.177,
+      "grad_norm": 8.423614501953125,
+      "learning_rate": 2.0589e-05,
+      "loss": 3.1809,
+      "step": 58850
+    },
+    {
+      "epoch": 1.178,
+      "grad_norm": 7.818522930145264,
+      "learning_rate": 2.0563999999999998e-05,
+      "loss": 2.8685,
+      "step": 58900
+    },
+    {
+      "epoch": 1.179,
+      "grad_norm": 7.669625759124756,
+      "learning_rate": 2.0539e-05,
+      "loss": 2.993,
+      "step": 58950
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 7.713820457458496,
+      "learning_rate": 2.0514e-05,
+      "loss": 2.8745,
+      "step": 59000
+    },
+    {
+      "epoch": 1.181,
+      "grad_norm": 8.473793983459473,
+      "learning_rate": 2.0489e-05,
+      "loss": 2.9275,
+      "step": 59050
+    },
+    {
+      "epoch": 1.182,
+      "grad_norm": 7.552177429199219,
+      "learning_rate": 2.0464e-05,
+      "loss": 2.9823,
+      "step": 59100
+    },
+    {
+      "epoch": 1.183,
+      "grad_norm": 8.08139419555664,
+      "learning_rate": 2.0439e-05,
+      "loss": 2.8557,
+      "step": 59150
+    },
+    {
+      "epoch": 1.184,
+      "grad_norm": 7.447910785675049,
+      "learning_rate": 2.0414e-05,
+      "loss": 2.9044,
+      "step": 59200
+    },
+    {
+      "epoch": 1.185,
+      "grad_norm": 8.74582290649414,
+      "learning_rate": 2.0389e-05,
+      "loss": 2.8841,
+      "step": 59250
+    },
+    {
+      "epoch": 1.186,
+      "grad_norm": 6.812131404876709,
+      "learning_rate": 2.0364e-05,
+      "loss": 2.7652,
+      "step": 59300
+    },
+    {
+      "epoch": 1.187,
+      "grad_norm": 6.931506156921387,
+      "learning_rate": 2.0339e-05,
+      "loss": 2.9071,
+      "step": 59350
+    },
+    {
+      "epoch": 1.188,
+      "grad_norm": 7.589115619659424,
+      "learning_rate": 2.0314e-05,
+      "loss": 3.0316,
+      "step": 59400
+    },
+    {
+      "epoch": 1.189,
+      "grad_norm": 7.810254096984863,
+      "learning_rate": 2.0289e-05,
+      "loss": 2.89,
+      "step": 59450
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.848294258117676,
+      "learning_rate": 2.0264e-05,
+      "loss": 2.938,
+      "step": 59500
+    },
+    {
+      "epoch": 1.191,
+      "grad_norm": 8.115826606750488,
+      "learning_rate": 2.0239e-05,
+      "loss": 2.958,
+      "step": 59550
+    },
+    {
+      "epoch": 1.192,
+      "grad_norm": 7.386075973510742,
+      "learning_rate": 2.0214e-05,
+      "loss": 2.902,
+      "step": 59600
+    },
+    {
+      "epoch": 1.193,
+      "grad_norm": 5.866041660308838,
+      "learning_rate": 2.0189e-05,
+      "loss": 2.883,
+      "step": 59650
+    },
+    {
+      "epoch": 1.194,
+      "grad_norm": 6.591806411743164,
+      "learning_rate": 2.0164000000000002e-05,
+      "loss": 2.7417,
+      "step": 59700
+    },
+    {
+      "epoch": 1.195,
+      "grad_norm": 7.479393482208252,
+      "learning_rate": 2.0139e-05,
+      "loss": 2.8905,
+      "step": 59750
+    },
+    {
+      "epoch": 1.196,
+      "grad_norm": 6.727341651916504,
+      "learning_rate": 2.0114e-05,
+      "loss": 2.9524,
+      "step": 59800
+    },
+    {
+      "epoch": 1.197,
+      "grad_norm": 7.175183296203613,
+      "learning_rate": 2.0089000000000004e-05,
+      "loss": 2.7486,
+      "step": 59850
+    },
+    {
+      "epoch": 1.198,
+      "grad_norm": 8.155020713806152,
+      "learning_rate": 2.0064000000000003e-05,
+      "loss": 3.042,
+      "step": 59900
+    },
+    {
+      "epoch": 1.199,
+      "grad_norm": 6.671029567718506,
+      "learning_rate": 2.0039000000000002e-05,
+      "loss": 3.034,
+      "step": 59950
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.177928924560547,
+      "learning_rate": 2.0014000000000002e-05,
+      "loss": 3.0329,
+      "step": 60000
+    },
+    {
+      "epoch": 1.201,
+      "grad_norm": 8.286439895629883,
+      "learning_rate": 1.9989e-05,
+      "loss": 3.051,
+      "step": 60050
+    },
+    {
+      "epoch": 1.202,
+      "grad_norm": 8.125651359558105,
+      "learning_rate": 1.99645e-05,
+      "loss": 3.1017,
+      "step": 60100
+    },
+    {
+      "epoch": 1.203,
+      "grad_norm": 6.588275909423828,
+      "learning_rate": 1.99395e-05,
+      "loss": 2.9559,
+      "step": 60150
+    },
+    {
+      "epoch": 1.204,
+      "grad_norm": 8.682856559753418,
+      "learning_rate": 1.99145e-05,
+      "loss": 2.799,
+      "step": 60200
+    },
+    {
+      "epoch": 1.205,
+      "grad_norm": 7.333592414855957,
+      "learning_rate": 1.98895e-05,
+      "loss": 2.9984,
+      "step": 60250
+    },
+    {
+      "epoch": 1.206,
+      "grad_norm": 6.971369743347168,
+      "learning_rate": 1.98645e-05,
+      "loss": 2.8116,
+      "step": 60300
+    },
+    {
+      "epoch": 1.207,
+      "grad_norm": 9.593948364257812,
+      "learning_rate": 1.98395e-05,
+      "loss": 2.7672,
+      "step": 60350
+    },
+    {
+      "epoch": 1.208,
+      "grad_norm": 6.947750091552734,
+      "learning_rate": 1.9814499999999998e-05,
+      "loss": 2.8234,
+      "step": 60400
+    },
+    {
+      "epoch": 1.209,
+      "grad_norm": 9.746689796447754,
+      "learning_rate": 1.97895e-05,
+      "loss": 2.8323,
+      "step": 60450
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.871998310089111,
+      "learning_rate": 1.97645e-05,
+      "loss": 3.0075,
+      "step": 60500
+    },
+    {
+      "epoch": 1.211,
+      "grad_norm": 8.099973678588867,
+      "learning_rate": 1.97395e-05,
+      "loss": 2.894,
+      "step": 60550
+    },
+    {
+      "epoch": 1.212,
+      "grad_norm": 9.960383415222168,
+      "learning_rate": 1.97145e-05,
+      "loss": 3.0326,
+      "step": 60600
+    },
+    {
+      "epoch": 1.213,
+      "grad_norm": 7.341196060180664,
+      "learning_rate": 1.96895e-05,
+      "loss": 2.7773,
+      "step": 60650
+    },
+    {
+      "epoch": 1.214,
+      "grad_norm": 7.2196125984191895,
+      "learning_rate": 1.96645e-05,
+      "loss": 2.9141,
+      "step": 60700
+    },
+    {
+      "epoch": 1.215,
+      "grad_norm": 7.776864051818848,
+      "learning_rate": 1.96395e-05,
+      "loss": 3.0412,
+      "step": 60750
+    },
+    {
+      "epoch": 1.216,
+      "grad_norm": 9.311785697937012,
+      "learning_rate": 1.9614500000000003e-05,
+      "loss": 2.9752,
+      "step": 60800
+    },
+    {
+      "epoch": 1.217,
+      "grad_norm": 9.076855659484863,
+      "learning_rate": 1.9589500000000003e-05,
+      "loss": 2.7665,
+      "step": 60850
+    },
+    {
+      "epoch": 1.218,
+      "grad_norm": 7.043571472167969,
+      "learning_rate": 1.9564500000000002e-05,
+      "loss": 2.8502,
+      "step": 60900
+    },
+    {
+      "epoch": 1.219,
+      "grad_norm": 8.467164993286133,
+      "learning_rate": 1.95395e-05,
+      "loss": 2.8783,
+      "step": 60950
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 6.73739767074585,
+      "learning_rate": 1.95145e-05,
+      "loss": 2.7169,
+      "step": 61000
+    },
+    {
+      "epoch": 1.221,
+      "grad_norm": 8.30189037322998,
+      "learning_rate": 1.9489500000000004e-05,
+      "loss": 2.8347,
+      "step": 61050
+    },
+    {
+      "epoch": 1.222,
+      "grad_norm": 8.079895973205566,
+      "learning_rate": 1.9464500000000003e-05,
+      "loss": 3.0235,
+      "step": 61100
+    },
+    {
+      "epoch": 1.223,
+      "grad_norm": 8.46454906463623,
+      "learning_rate": 1.9439500000000003e-05,
+      "loss": 2.7273,
+      "step": 61150
+    },
+    {
+      "epoch": 1.224,
+      "grad_norm": 7.661693096160889,
+      "learning_rate": 1.9414500000000002e-05,
+      "loss": 3.0257,
+      "step": 61200
+    },
+    {
+      "epoch": 1.225,
+      "grad_norm": 8.443622589111328,
+      "learning_rate": 1.93895e-05,
+      "loss": 2.9563,
+      "step": 61250
+    },
+    {
+      "epoch": 1.226,
+      "grad_norm": 6.420770168304443,
+      "learning_rate": 1.93645e-05,
+      "loss": 3.0847,
+      "step": 61300
+    },
+    {
+      "epoch": 1.227,
+      "grad_norm": 7.339363098144531,
+      "learning_rate": 1.9339500000000003e-05,
+      "loss": 2.9745,
+      "step": 61350
+    },
+    {
+      "epoch": 1.228,
+      "grad_norm": 10.756458282470703,
+      "learning_rate": 1.9314500000000003e-05,
+      "loss": 2.8565,
+      "step": 61400
+    },
+    {
+      "epoch": 1.229,
+      "grad_norm": 7.621661186218262,
+      "learning_rate": 1.9289500000000002e-05,
+      "loss": 2.8492,
+      "step": 61450
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 8.242738723754883,
+      "learning_rate": 1.92645e-05,
+      "loss": 2.9573,
+      "step": 61500
+    },
+    {
+      "epoch": 1.231,
+      "grad_norm": 8.807046890258789,
+      "learning_rate": 1.92395e-05,
+      "loss": 2.8937,
+      "step": 61550
+    },
+    {
+      "epoch": 1.232,
+      "grad_norm": 7.321935653686523,
+      "learning_rate": 1.92145e-05,
+      "loss": 2.7142,
+      "step": 61600
+    },
+    {
+      "epoch": 1.233,
+      "grad_norm": 7.530791759490967,
+      "learning_rate": 1.9189500000000003e-05,
+      "loss": 2.8536,
+      "step": 61650
+    },
+    {
+      "epoch": 1.234,
+      "grad_norm": 9.09104061126709,
+      "learning_rate": 1.9164500000000003e-05,
+      "loss": 2.9084,
+      "step": 61700
+    },
+    {
+      "epoch": 1.2349999999999999,
+      "grad_norm": 8.539567947387695,
+      "learning_rate": 1.9139500000000002e-05,
+      "loss": 3.0237,
+      "step": 61750
+    },
+    {
+      "epoch": 1.236,
+      "grad_norm": 7.933023929595947,
+      "learning_rate": 1.91145e-05,
+      "loss": 2.8838,
+      "step": 61800
+    },
+    {
+      "epoch": 1.237,
+      "grad_norm": 7.881004810333252,
+      "learning_rate": 1.90895e-05,
+      "loss": 2.6865,
+      "step": 61850
+    },
+    {
+      "epoch": 1.238,
+      "grad_norm": 7.823519706726074,
+      "learning_rate": 1.90645e-05,
+      "loss": 2.8898,
+      "step": 61900
+    },
+    {
+      "epoch": 1.2389999999999999,
+      "grad_norm": 7.689711570739746,
+      "learning_rate": 1.9039500000000003e-05,
+      "loss": 2.9576,
+      "step": 61950
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.04930305480957,
+      "learning_rate": 1.9014500000000002e-05,
+      "loss": 2.7864,
+      "step": 62000
+    },
+    {
+      "epoch": 1.241,
+      "grad_norm": 7.104748249053955,
+      "learning_rate": 1.8989500000000002e-05,
+      "loss": 2.8336,
+      "step": 62050
+    },
+    {
+      "epoch": 1.242,
+      "grad_norm": 8.643699645996094,
+      "learning_rate": 1.8965000000000002e-05,
+      "loss": 2.9855,
+      "step": 62100
+    },
+    {
+      "epoch": 1.2429999999999999,
+      "grad_norm": 8.03945255279541,
+      "learning_rate": 1.894e-05,
+      "loss": 2.9015,
+      "step": 62150
+    },
+    {
+      "epoch": 1.244,
+      "grad_norm": 8.157134056091309,
+      "learning_rate": 1.8915e-05,
+      "loss": 3.0076,
+      "step": 62200
+    },
+    {
+      "epoch": 1.245,
+      "grad_norm": 7.763930797576904,
+      "learning_rate": 1.8890000000000004e-05,
+      "loss": 2.9506,
+      "step": 62250
+    },
+    {
+      "epoch": 1.246,
+      "grad_norm": 7.697092056274414,
+      "learning_rate": 1.8865000000000003e-05,
+      "loss": 2.8397,
+      "step": 62300
+    },
+    {
+      "epoch": 1.2469999999999999,
+      "grad_norm": 6.57480001449585,
+      "learning_rate": 1.8840000000000003e-05,
+      "loss": 2.7468,
+      "step": 62350
+    },
+    {
+      "epoch": 1.248,
+      "grad_norm": 8.595124244689941,
+      "learning_rate": 1.8815000000000002e-05,
+      "loss": 2.9574,
+      "step": 62400
+    },
+    {
+      "epoch": 1.249,
+      "grad_norm": 7.175886631011963,
+      "learning_rate": 1.879e-05,
+      "loss": 2.9085,
+      "step": 62450
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.397321701049805,
+      "learning_rate": 1.8765e-05,
+      "loss": 2.9286,
+      "step": 62500
+    },
+    {
+      "epoch": 1.251,
+      "grad_norm": 6.980536937713623,
+      "learning_rate": 1.8740000000000004e-05,
+      "loss": 2.747,
+      "step": 62550
+    },
+    {
+      "epoch": 1.252,
+      "grad_norm": 7.271572113037109,
+      "learning_rate": 1.8715000000000003e-05,
+      "loss": 2.7768,
+      "step": 62600
+    },
+    {
+      "epoch": 1.2530000000000001,
+      "grad_norm": 7.7483696937561035,
+      "learning_rate": 1.8690000000000002e-05,
+      "loss": 3.0488,
+      "step": 62650
+    },
+    {
+      "epoch": 1.254,
+      "grad_norm": 11.277056694030762,
+      "learning_rate": 1.8665000000000002e-05,
+      "loss": 3.0649,
+      "step": 62700
+    },
+    {
+      "epoch": 1.255,
+      "grad_norm": 8.678947448730469,
+      "learning_rate": 1.864e-05,
+      "loss": 2.9212,
+      "step": 62750
+    },
+    {
+      "epoch": 1.256,
+      "grad_norm": 7.193722724914551,
+      "learning_rate": 1.8615e-05,
+      "loss": 2.7391,
+      "step": 62800
+    },
+    {
+      "epoch": 1.2570000000000001,
+      "grad_norm": 12.467113494873047,
+      "learning_rate": 1.8590000000000003e-05,
+      "loss": 2.9668,
+      "step": 62850
+    },
+    {
+      "epoch": 1.258,
+      "grad_norm": 7.693551063537598,
+      "learning_rate": 1.8565000000000003e-05,
+      "loss": 2.7936,
+      "step": 62900
+    },
+    {
+      "epoch": 1.259,
+      "grad_norm": 7.1652445793151855,
+      "learning_rate": 1.8540000000000002e-05,
+      "loss": 2.9067,
+      "step": 62950
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 6.875372409820557,
+      "learning_rate": 1.8515e-05,
+      "loss": 2.9379,
+      "step": 63000
+    },
+    {
+      "epoch": 1.2610000000000001,
+      "grad_norm": 8.923434257507324,
+      "learning_rate": 1.849e-05,
+      "loss": 2.7059,
+      "step": 63050
+    },
+    {
+      "epoch": 1.262,
+      "grad_norm": 11.432680130004883,
+      "learning_rate": 1.8465e-05,
+      "loss": 2.9926,
+      "step": 63100
+    },
+    {
+      "epoch": 1.263,
+      "grad_norm": 10.260088920593262,
+      "learning_rate": 1.8440000000000003e-05,
+      "loss": 3.0907,
+      "step": 63150
+    },
+    {
+      "epoch": 1.264,
+      "grad_norm": 8.717737197875977,
+      "learning_rate": 1.8415000000000002e-05,
+      "loss": 2.9059,
+      "step": 63200
+    },
+    {
+      "epoch": 1.2650000000000001,
+      "grad_norm": 8.755745887756348,
+      "learning_rate": 1.8390000000000002e-05,
+      "loss": 2.8056,
+      "step": 63250
+    },
+    {
+      "epoch": 1.266,
+      "grad_norm": 6.1573286056518555,
+      "learning_rate": 1.8365e-05,
+      "loss": 3.0321,
+      "step": 63300
+    },
+    {
+      "epoch": 1.267,
+      "grad_norm": 8.053323745727539,
+      "learning_rate": 1.834e-05,
+      "loss": 2.8293,
+      "step": 63350
+    },
+    {
+      "epoch": 1.268,
+      "grad_norm": 7.611751556396484,
+      "learning_rate": 1.8315e-05,
+      "loss": 2.9095,
+      "step": 63400
+    },
+    {
+      "epoch": 1.2690000000000001,
+      "grad_norm": 9.002436637878418,
+      "learning_rate": 1.8290000000000003e-05,
+      "loss": 2.8894,
+      "step": 63450
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.82766056060791,
+      "learning_rate": 1.8265000000000002e-05,
+      "loss": 3.1194,
+      "step": 63500
+    },
+    {
+      "epoch": 1.271,
+      "grad_norm": 7.163455486297607,
+      "learning_rate": 1.824e-05,
+      "loss": 2.8781,
+      "step": 63550
+    },
+    {
+      "epoch": 1.272,
+      "grad_norm": 7.2303032875061035,
+      "learning_rate": 1.8215e-05,
+      "loss": 2.7364,
+      "step": 63600
+    },
+    {
+      "epoch": 1.2730000000000001,
+      "grad_norm": 7.012924671173096,
+      "learning_rate": 1.819e-05,
+      "loss": 2.7426,
+      "step": 63650
+    },
+    {
+      "epoch": 1.274,
+      "grad_norm": 8.245091438293457,
+      "learning_rate": 1.8165000000000003e-05,
+      "loss": 2.7933,
+      "step": 63700
+    },
+    {
+      "epoch": 1.275,
+      "grad_norm": 8.162280082702637,
+      "learning_rate": 1.8140000000000003e-05,
+      "loss": 2.9715,
+      "step": 63750
+    },
+    {
+      "epoch": 1.276,
+      "grad_norm": 6.616583824157715,
+      "learning_rate": 1.8115000000000002e-05,
+      "loss": 2.7865,
+      "step": 63800
+    },
+    {
+      "epoch": 1.2770000000000001,
+      "grad_norm": 7.844861030578613,
+      "learning_rate": 1.809e-05,
+      "loss": 2.9921,
+      "step": 63850
+    },
+    {
+      "epoch": 1.278,
+      "grad_norm": 5.926692008972168,
+      "learning_rate": 1.8065e-05,
+      "loss": 2.7858,
+      "step": 63900
+    },
+    {
+      "epoch": 1.279,
+      "grad_norm": 8.044204711914062,
+      "learning_rate": 1.804e-05,
+      "loss": 2.9768,
+      "step": 63950
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 7.818645477294922,
+      "learning_rate": 1.8015000000000003e-05,
+      "loss": 2.7761,
+      "step": 64000
+    },
+    {
+      "epoch": 1.2810000000000001,
+      "grad_norm": 6.930302143096924,
+      "learning_rate": 1.7990000000000002e-05,
+      "loss": 2.9969,
+      "step": 64050
+    },
+    {
+      "epoch": 1.282,
+      "grad_norm": 7.224350929260254,
+      "learning_rate": 1.7965e-05,
+      "loss": 2.8126,
+      "step": 64100
+    },
+    {
+      "epoch": 1.283,
+      "grad_norm": 8.74736499786377,
+      "learning_rate": 1.7940500000000002e-05,
+      "loss": 3.0012,
+      "step": 64150
+    },
+    {
+      "epoch": 1.284,
+      "grad_norm": 7.229979038238525,
+      "learning_rate": 1.79155e-05,
+      "loss": 2.7311,
+      "step": 64200
+    },
+    {
+      "epoch": 1.285,
+      "grad_norm": 7.284625053405762,
+      "learning_rate": 1.78905e-05,
+      "loss": 3.009,
+      "step": 64250
+    },
+    {
+      "epoch": 1.286,
+      "grad_norm": 7.718355178833008,
+      "learning_rate": 1.78655e-05,
+      "loss": 2.874,
+      "step": 64300
+    },
+    {
+      "epoch": 1.287,
+      "grad_norm": 7.456071376800537,
+      "learning_rate": 1.7840500000000003e-05,
+      "loss": 2.9934,
+      "step": 64350
+    },
+    {
+      "epoch": 1.288,
+      "grad_norm": 10.048409461975098,
+      "learning_rate": 1.7815500000000002e-05,
+      "loss": 2.7551,
+      "step": 64400
+    },
+    {
+      "epoch": 1.289,
+      "grad_norm": 6.842342853546143,
+      "learning_rate": 1.7790500000000002e-05,
+      "loss": 2.9174,
+      "step": 64450
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.162269592285156,
+      "learning_rate": 1.77655e-05,
+      "loss": 2.9437,
+      "step": 64500
+    },
+    {
+      "epoch": 1.291,
+      "grad_norm": 6.678775787353516,
+      "learning_rate": 1.77405e-05,
+      "loss": 3.0262,
+      "step": 64550
+    },
+    {
+      "epoch": 1.292,
+      "grad_norm": 7.563632011413574,
+      "learning_rate": 1.77155e-05,
+      "loss": 2.8644,
+      "step": 64600
+    },
+    {
+      "epoch": 1.293,
+      "grad_norm": 6.082263946533203,
+      "learning_rate": 1.7690500000000003e-05,
+      "loss": 2.7633,
+      "step": 64650
+    },
+    {
+      "epoch": 1.294,
+      "grad_norm": 6.205733299255371,
+      "learning_rate": 1.7665500000000002e-05,
+      "loss": 2.7663,
+      "step": 64700
+    },
+    {
+      "epoch": 1.295,
+      "grad_norm": 8.632345199584961,
+      "learning_rate": 1.76405e-05,
+      "loss": 2.7141,
+      "step": 64750
+    },
+    {
+      "epoch": 1.296,
+      "grad_norm": 6.342571258544922,
+      "learning_rate": 1.76155e-05,
+      "loss": 2.8406,
+      "step": 64800
+    },
+    {
+      "epoch": 1.297,
+      "grad_norm": 6.610126495361328,
+      "learning_rate": 1.75905e-05,
+      "loss": 2.87,
+      "step": 64850
+    },
+    {
+      "epoch": 1.298,
+      "grad_norm": 7.278533458709717,
+      "learning_rate": 1.75655e-05,
+      "loss": 2.8149,
+      "step": 64900
+    },
+    {
+      "epoch": 1.299,
+      "grad_norm": 8.10925006866455,
+      "learning_rate": 1.7540500000000003e-05,
+      "loss": 2.8956,
+      "step": 64950
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.207983016967773,
+      "learning_rate": 1.7515500000000002e-05,
+      "loss": 2.9087,
+      "step": 65000
+    },
+    {
+      "epoch": 1.301,
+      "grad_norm": 8.794236183166504,
+      "learning_rate": 1.74905e-05,
+      "loss": 3.0564,
+      "step": 65050
+    },
+    {
+      "epoch": 1.302,
+      "grad_norm": 15.703174591064453,
+      "learning_rate": 1.74655e-05,
+      "loss": 2.9178,
+      "step": 65100
+    },
+    {
+      "epoch": 1.303,
+      "grad_norm": 8.117988586425781,
+      "learning_rate": 1.74405e-05,
+      "loss": 2.8114,
+      "step": 65150
+    },
+    {
+      "epoch": 1.304,
+      "grad_norm": 7.884382247924805,
+      "learning_rate": 1.7415500000000003e-05,
+      "loss": 2.7444,
+      "step": 65200
+    },
+    {
+      "epoch": 1.305,
+      "grad_norm": 8.498047828674316,
+      "learning_rate": 1.7390500000000002e-05,
+      "loss": 2.8569,
+      "step": 65250
+    },
+    {
+      "epoch": 1.306,
+      "grad_norm": 7.473427772521973,
+      "learning_rate": 1.7365500000000002e-05,
+      "loss": 2.9345,
+      "step": 65300
+    },
+    {
+      "epoch": 1.307,
+      "grad_norm": 7.29819917678833,
+      "learning_rate": 1.73405e-05,
+      "loss": 3.117,
+      "step": 65350
+    },
+    {
+      "epoch": 1.308,
+      "grad_norm": 8.479053497314453,
+      "learning_rate": 1.73155e-05,
+      "loss": 2.8155,
+      "step": 65400
+    },
+    {
+      "epoch": 1.309,
+      "grad_norm": 8.74229907989502,
+      "learning_rate": 1.72905e-05,
+      "loss": 2.9308,
+      "step": 65450
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.387805938720703,
+      "learning_rate": 1.7265500000000003e-05,
+      "loss": 2.9132,
+      "step": 65500
+    },
+    {
+      "epoch": 1.311,
+      "grad_norm": 8.244154930114746,
+      "learning_rate": 1.7240500000000002e-05,
+      "loss": 2.8277,
+      "step": 65550
+    },
+    {
+      "epoch": 1.312,
+      "grad_norm": 6.698085308074951,
+      "learning_rate": 1.72155e-05,
+      "loss": 3.0865,
+      "step": 65600
+    },
+    {
+      "epoch": 1.313,
+      "grad_norm": 10.616350173950195,
+      "learning_rate": 1.71905e-05,
+      "loss": 2.8149,
+      "step": 65650
+    },
+    {
+      "epoch": 1.314,
+      "grad_norm": 9.140168190002441,
+      "learning_rate": 1.71655e-05,
+      "loss": 2.8297,
+      "step": 65700
+    },
+    {
+      "epoch": 1.315,
+      "grad_norm": 9.049459457397461,
+      "learning_rate": 1.71405e-05,
+      "loss": 2.6896,
+      "step": 65750
+    },
+    {
+      "epoch": 1.316,
+      "grad_norm": 6.275703430175781,
+      "learning_rate": 1.7115500000000002e-05,
+      "loss": 2.8437,
+      "step": 65800
+    },
+    {
+      "epoch": 1.317,
+      "grad_norm": 9.357379913330078,
+      "learning_rate": 1.7090500000000002e-05,
+      "loss": 3.0127,
+      "step": 65850
+    },
+    {
+      "epoch": 1.318,
+      "grad_norm": 7.831468105316162,
+      "learning_rate": 1.70655e-05,
+      "loss": 2.792,
+      "step": 65900
+    },
+    {
+      "epoch": 1.319,
+      "grad_norm": 9.359698295593262,
+      "learning_rate": 1.70405e-05,
+      "loss": 3.0451,
+      "step": 65950
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 6.813355922698975,
+      "learning_rate": 1.70155e-05,
+      "loss": 2.7334,
+      "step": 66000
+    },
+    {
+      "epoch": 1.321,
+      "grad_norm": 8.375040054321289,
+      "learning_rate": 1.69905e-05,
+      "loss": 2.7643,
+      "step": 66050
+    },
+    {
+      "epoch": 1.322,
+      "grad_norm": 6.97425651550293,
+      "learning_rate": 1.6965500000000002e-05,
+      "loss": 2.9207,
+      "step": 66100
+    },
+    {
+      "epoch": 1.323,
+      "grad_norm": 9.489729881286621,
+      "learning_rate": 1.6941000000000003e-05,
+      "loss": 2.8561,
+      "step": 66150
+    },
+    {
+      "epoch": 1.324,
+      "grad_norm": 8.421103477478027,
+      "learning_rate": 1.6916000000000002e-05,
+      "loss": 3.0594,
+      "step": 66200
+    },
+    {
+      "epoch": 1.325,
+      "grad_norm": 7.924808502197266,
+      "learning_rate": 1.6891e-05,
+      "loss": 2.9725,
+      "step": 66250
+    },
+    {
+      "epoch": 1.326,
+      "grad_norm": 8.143972396850586,
+      "learning_rate": 1.6866e-05,
+      "loss": 2.7149,
+      "step": 66300
+    },
+    {
+      "epoch": 1.327,
+      "grad_norm": 11.077095031738281,
+      "learning_rate": 1.6841e-05,
+      "loss": 2.858,
+      "step": 66350
+    },
+    {
+      "epoch": 1.328,
+      "grad_norm": 7.736219882965088,
+      "learning_rate": 1.6816e-05,
+      "loss": 2.9161,
+      "step": 66400
+    },
+    {
+      "epoch": 1.329,
+      "grad_norm": 9.575567245483398,
+      "learning_rate": 1.6791000000000002e-05,
+      "loss": 2.8953,
+      "step": 66450
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 6.864261150360107,
+      "learning_rate": 1.6766000000000002e-05,
+      "loss": 2.7205,
+      "step": 66500
+    },
+    {
+      "epoch": 1.331,
+      "grad_norm": 7.764801502227783,
+      "learning_rate": 1.6741e-05,
+      "loss": 2.9188,
+      "step": 66550
+    },
+    {
+      "epoch": 1.332,
+      "grad_norm": 10.629688262939453,
+      "learning_rate": 1.6716e-05,
+      "loss": 2.9199,
+      "step": 66600
+    },
+    {
+      "epoch": 1.333,
+      "grad_norm": 8.512171745300293,
+      "learning_rate": 1.6691e-05,
+      "loss": 2.8188,
+      "step": 66650
+    },
+    {
+      "epoch": 1.334,
+      "grad_norm": 8.378336906433105,
+      "learning_rate": 1.6666000000000003e-05,
+      "loss": 2.7863,
+      "step": 66700
+    },
+    {
+      "epoch": 1.335,
+      "grad_norm": 8.283653259277344,
+      "learning_rate": 1.6641000000000002e-05,
+      "loss": 2.9134,
+      "step": 66750
+    },
+    {
+      "epoch": 1.336,
+      "grad_norm": 6.067685604095459,
+      "learning_rate": 1.6616e-05,
+      "loss": 2.661,
+      "step": 66800
+    },
+    {
+      "epoch": 1.337,
+      "grad_norm": 7.018548488616943,
+      "learning_rate": 1.6591e-05,
+      "loss": 2.7564,
+      "step": 66850
+    },
+    {
+      "epoch": 1.338,
+      "grad_norm": 8.71909236907959,
+      "learning_rate": 1.6566e-05,
+      "loss": 2.7492,
+      "step": 66900
+    },
+    {
+      "epoch": 1.339,
+      "grad_norm": 7.719536304473877,
+      "learning_rate": 1.6541e-05,
+      "loss": 2.9521,
+      "step": 66950
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.370909690856934,
+      "learning_rate": 1.6516000000000002e-05,
+      "loss": 2.9058,
+      "step": 67000
+    },
+    {
+      "epoch": 1.341,
+      "grad_norm": 6.47285795211792,
+      "learning_rate": 1.6491000000000002e-05,
+      "loss": 2.8832,
+      "step": 67050
+    },
+    {
+      "epoch": 1.342,
+      "grad_norm": 6.0031890869140625,
+      "learning_rate": 1.6466e-05,
+      "loss": 2.6373,
+      "step": 67100
+    },
+    {
+      "epoch": 1.343,
+      "grad_norm": 7.42556095123291,
+      "learning_rate": 1.6441e-05,
+      "loss": 2.9758,
+      "step": 67150
+    },
+    {
+      "epoch": 1.3439999999999999,
+      "grad_norm": 6.864173889160156,
+      "learning_rate": 1.6416e-05,
+      "loss": 2.8131,
+      "step": 67200
+    },
+    {
+      "epoch": 1.345,
+      "grad_norm": 8.983753204345703,
+      "learning_rate": 1.6391e-05,
+      "loss": 2.7171,
+      "step": 67250
+    },
+    {
+      "epoch": 1.346,
+      "grad_norm": 7.330747604370117,
+      "learning_rate": 1.6366000000000002e-05,
+      "loss": 2.7697,
+      "step": 67300
+    },
+    {
+      "epoch": 1.347,
+      "grad_norm": 8.947467803955078,
+      "learning_rate": 1.6341e-05,
+      "loss": 2.836,
+      "step": 67350
+    },
+    {
+      "epoch": 1.3479999999999999,
+      "grad_norm": 8.130457878112793,
+      "learning_rate": 1.6316e-05,
+      "loss": 2.8659,
+      "step": 67400
+    },
+    {
+      "epoch": 1.349,
+      "grad_norm": 7.954550743103027,
+      "learning_rate": 1.6291e-05,
+      "loss": 2.7594,
+      "step": 67450
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 7.327019691467285,
+      "learning_rate": 1.6266e-05,
+      "loss": 2.956,
+      "step": 67500
+    },
+    {
+      "epoch": 1.351,
+      "grad_norm": 8.24224853515625,
+      "learning_rate": 1.6241e-05,
+      "loss": 2.7842,
+      "step": 67550
+    },
+    {
+      "epoch": 1.3519999999999999,
+      "grad_norm": 7.031774997711182,
+      "learning_rate": 1.6216000000000002e-05,
+      "loss": 2.8288,
+      "step": 67600
+    },
+    {
+      "epoch": 1.353,
+      "grad_norm": 17.19927978515625,
+      "learning_rate": 1.6191e-05,
+      "loss": 2.9684,
+      "step": 67650
+    },
+    {
+      "epoch": 1.354,
+      "grad_norm": 9.650681495666504,
+      "learning_rate": 1.6166e-05,
+      "loss": 2.853,
+      "step": 67700
+    },
+    {
+      "epoch": 1.355,
+      "grad_norm": 7.688391208648682,
+      "learning_rate": 1.6141e-05,
+      "loss": 2.9444,
+      "step": 67750
+    },
+    {
+      "epoch": 1.3559999999999999,
+      "grad_norm": 6.4372406005859375,
+      "learning_rate": 1.6116e-05,
+      "loss": 2.7925,
+      "step": 67800
+    },
+    {
+      "epoch": 1.357,
+      "grad_norm": 8.255843162536621,
+      "learning_rate": 1.6091e-05,
+      "loss": 2.9575,
+      "step": 67850
+    },
+    {
+      "epoch": 1.358,
+      "grad_norm": 8.876387596130371,
+      "learning_rate": 1.6066e-05,
+      "loss": 2.8265,
+      "step": 67900
+    },
+    {
+      "epoch": 1.359,
+      "grad_norm": 6.5156941413879395,
+      "learning_rate": 1.6041e-05,
+      "loss": 2.984,
+      "step": 67950
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 10.278947830200195,
+      "learning_rate": 1.6016e-05,
+      "loss": 2.8802,
+      "step": 68000
+    },
+    {
+      "epoch": 1.361,
+      "grad_norm": 8.730320930480957,
+      "learning_rate": 1.5991e-05,
+      "loss": 2.9155,
+      "step": 68050
+    },
+    {
+      "epoch": 1.362,
+      "grad_norm": 7.59114408493042,
+      "learning_rate": 1.5966e-05,
+      "loss": 2.9146,
+      "step": 68100
+    },
+    {
+      "epoch": 1.363,
+      "grad_norm": 7.220095157623291,
+      "learning_rate": 1.5941000000000002e-05,
+      "loss": 2.6534,
+      "step": 68150
+    },
+    {
+      "epoch": 1.3639999999999999,
+      "grad_norm": 8.136276245117188,
+      "learning_rate": 1.5916500000000003e-05,
+      "loss": 2.9809,
+      "step": 68200
+    },
+    {
+      "epoch": 1.365,
+      "grad_norm": 7.093013763427734,
+      "learning_rate": 1.5891500000000002e-05,
+      "loss": 2.7613,
+      "step": 68250
+    },
+    {
+      "epoch": 1.366,
+      "grad_norm": 8.815773963928223,
+      "learning_rate": 1.58665e-05,
+      "loss": 3.066,
+      "step": 68300
+    },
+    {
+      "epoch": 1.367,
+      "grad_norm": 7.6265668869018555,
+      "learning_rate": 1.58415e-05,
+      "loss": 2.8987,
+      "step": 68350
+    },
+    {
+      "epoch": 1.3679999999999999,
+      "grad_norm": 8.445241928100586,
+      "learning_rate": 1.58165e-05,
+      "loss": 2.7399,
+      "step": 68400
+    },
+    {
+      "epoch": 1.369,
+      "grad_norm": 9.16798210144043,
+      "learning_rate": 1.57915e-05,
+      "loss": 2.9594,
+      "step": 68450
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.255063533782959,
+      "learning_rate": 1.5766500000000002e-05,
+      "loss": 2.7959,
+      "step": 68500
+    },
+    {
+      "epoch": 1.371,
+      "grad_norm": 7.769148349761963,
+      "learning_rate": 1.57415e-05,
+      "loss": 2.802,
+      "step": 68550
+    },
+    {
+      "epoch": 1.3719999999999999,
+      "grad_norm": 8.59789752960205,
+      "learning_rate": 1.57165e-05,
+      "loss": 2.7217,
+      "step": 68600
+    },
+    {
+      "epoch": 1.373,
+      "grad_norm": 10.071572303771973,
+      "learning_rate": 1.56915e-05,
+      "loss": 2.9676,
+      "step": 68650
+    },
+    {
+      "epoch": 1.374,
+      "grad_norm": 10.081389427185059,
+      "learning_rate": 1.56665e-05,
+      "loss": 2.7645,
+      "step": 68700
+    },
+    {
+      "epoch": 1.375,
+      "grad_norm": 8.29869270324707,
+      "learning_rate": 1.56415e-05,
+      "loss": 2.9042,
+      "step": 68750
+    },
+    {
+      "epoch": 1.376,
+      "grad_norm": 7.104742527008057,
+      "learning_rate": 1.5616500000000002e-05,
+      "loss": 3.0615,
+      "step": 68800
+    },
+    {
+      "epoch": 1.377,
+      "grad_norm": 6.897149085998535,
+      "learning_rate": 1.55915e-05,
+      "loss": 2.8236,
+      "step": 68850
+    },
+    {
+      "epoch": 1.3780000000000001,
+      "grad_norm": 6.750307083129883,
+      "learning_rate": 1.55665e-05,
+      "loss": 2.8946,
+      "step": 68900
+    },
+    {
+      "epoch": 1.379,
+      "grad_norm": 9.587626457214355,
+      "learning_rate": 1.55415e-05,
+      "loss": 2.8499,
+      "step": 68950
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.258056640625,
+      "learning_rate": 1.55165e-05,
+      "loss": 2.7918,
+      "step": 69000
+    },
+    {
+      "epoch": 1.381,
+      "grad_norm": 7.532637596130371,
+      "learning_rate": 1.54915e-05,
+      "loss": 2.6911,
+      "step": 69050
+    },
+    {
+      "epoch": 1.3820000000000001,
+      "grad_norm": 8.102679252624512,
+      "learning_rate": 1.5466500000000002e-05,
+      "loss": 3.051,
+      "step": 69100
+    },
+    {
+      "epoch": 1.383,
+      "grad_norm": 8.062738418579102,
+      "learning_rate": 1.54415e-05,
+      "loss": 2.9364,
+      "step": 69150
+    },
+    {
+      "epoch": 1.384,
+      "grad_norm": 8.074463844299316,
+      "learning_rate": 1.54165e-05,
+      "loss": 2.9216,
+      "step": 69200
+    },
+    {
+      "epoch": 1.385,
+      "grad_norm": 7.463252544403076,
+      "learning_rate": 1.53915e-05,
+      "loss": 2.7501,
+      "step": 69250
+    },
+    {
+      "epoch": 1.3860000000000001,
+      "grad_norm": 7.215155124664307,
+      "learning_rate": 1.53665e-05,
+      "loss": 2.6774,
+      "step": 69300
+    },
+    {
+      "epoch": 1.387,
+      "grad_norm": 8.170955657958984,
+      "learning_rate": 1.53415e-05,
+      "loss": 2.7927,
+      "step": 69350
+    },
+    {
+      "epoch": 1.388,
+      "grad_norm": 8.302571296691895,
+      "learning_rate": 1.53165e-05,
+      "loss": 2.7675,
+      "step": 69400
+    },
+    {
+      "epoch": 1.389,
+      "grad_norm": 6.916486740112305,
+      "learning_rate": 1.52915e-05,
+      "loss": 2.9605,
+      "step": 69450
+    },
+    {
+      "epoch": 1.3900000000000001,
+      "grad_norm": 6.986796855926514,
+      "learning_rate": 1.52665e-05,
+      "loss": 2.9328,
+      "step": 69500
+    },
+    {
+      "epoch": 1.391,
+      "grad_norm": 7.800340175628662,
+      "learning_rate": 1.52415e-05,
+      "loss": 2.8924,
+      "step": 69550
+    },
+    {
+      "epoch": 1.392,
+      "grad_norm": 9.064175605773926,
+      "learning_rate": 1.52165e-05,
+      "loss": 2.8534,
+      "step": 69600
+    },
+    {
+      "epoch": 1.393,
+      "grad_norm": 8.305397033691406,
+      "learning_rate": 1.51915e-05,
+      "loss": 2.8954,
+      "step": 69650
+    },
+    {
+      "epoch": 1.3940000000000001,
+      "grad_norm": 7.492577075958252,
+      "learning_rate": 1.51665e-05,
+      "loss": 2.8653,
+      "step": 69700
+    },
+    {
+      "epoch": 1.395,
+      "grad_norm": 7.92070198059082,
+      "learning_rate": 1.5142e-05,
+      "loss": 2.7273,
+      "step": 69750
+    },
+    {
+      "epoch": 1.396,
+      "grad_norm": 7.920971393585205,
+      "learning_rate": 1.5117000000000001e-05,
+      "loss": 2.9048,
+      "step": 69800
+    },
+    {
+      "epoch": 1.397,
+      "grad_norm": 9.150144577026367,
+      "learning_rate": 1.5092e-05,
+      "loss": 2.6779,
+      "step": 69850
+    },
+    {
+      "epoch": 1.3980000000000001,
+      "grad_norm": 8.129121780395508,
+      "learning_rate": 1.5067e-05,
+      "loss": 2.9142,
+      "step": 69900
+    },
+    {
+      "epoch": 1.399,
+      "grad_norm": 8.334802627563477,
+      "learning_rate": 1.5042000000000001e-05,
+      "loss": 2.6763,
+      "step": 69950
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.4681806564331055,
+      "learning_rate": 1.5017e-05,
+      "loss": 2.93,
+      "step": 70000
+    },
+    {
+      "epoch": 1.401,
+      "grad_norm": 7.846401214599609,
+      "learning_rate": 1.4992e-05,
+      "loss": 2.7229,
+      "step": 70050
+    },
+    {
+      "epoch": 1.4020000000000001,
+      "grad_norm": 7.523430347442627,
+      "learning_rate": 1.4967000000000001e-05,
+      "loss": 2.8843,
+      "step": 70100
+    },
+    {
+      "epoch": 1.403,
+      "grad_norm": 9.278549194335938,
+      "learning_rate": 1.4942e-05,
+      "loss": 2.7884,
+      "step": 70150
+    },
+    {
+      "epoch": 1.404,
+      "grad_norm": 7.723369121551514,
+      "learning_rate": 1.4917e-05,
+      "loss": 2.853,
+      "step": 70200
+    },
+    {
+      "epoch": 1.405,
+      "grad_norm": 8.88216781616211,
+      "learning_rate": 1.4892e-05,
+      "loss": 2.9649,
+      "step": 70250
+    },
+    {
+      "epoch": 1.4060000000000001,
+      "grad_norm": 6.627724647521973,
+      "learning_rate": 1.4867e-05,
+      "loss": 2.7964,
+      "step": 70300
+    },
+    {
+      "epoch": 1.407,
+      "grad_norm": 8.70402717590332,
+      "learning_rate": 1.4842e-05,
+      "loss": 2.7892,
+      "step": 70350
+    },
+    {
+      "epoch": 1.408,
+      "grad_norm": 7.616790294647217,
+      "learning_rate": 1.4817e-05,
+      "loss": 2.9413,
+      "step": 70400
+    },
+    {
+      "epoch": 1.409,
+      "grad_norm": 8.156023025512695,
+      "learning_rate": 1.4792e-05,
+      "loss": 2.9861,
+      "step": 70450
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.509369850158691,
+      "learning_rate": 1.4767000000000001e-05,
+      "loss": 2.8202,
+      "step": 70500
+    },
+    {
+      "epoch": 1.411,
+      "grad_norm": 7.710995674133301,
+      "learning_rate": 1.4742e-05,
+      "loss": 2.9804,
+      "step": 70550
+    },
+    {
+      "epoch": 1.412,
+      "grad_norm": 6.933726787567139,
+      "learning_rate": 1.4717e-05,
+      "loss": 2.802,
+      "step": 70600
+    },
+    {
+      "epoch": 1.413,
+      "grad_norm": 8.056614875793457,
+      "learning_rate": 1.4692000000000001e-05,
+      "loss": 2.8331,
+      "step": 70650
+    },
+    {
+      "epoch": 1.414,
+      "grad_norm": 7.54365348815918,
+      "learning_rate": 1.4667e-05,
+      "loss": 2.8463,
+      "step": 70700
+    },
+    {
+      "epoch": 1.415,
+      "grad_norm": 7.553271770477295,
+      "learning_rate": 1.4642e-05,
+      "loss": 2.7486,
+      "step": 70750
+    },
+    {
+      "epoch": 1.416,
+      "grad_norm": 8.390814781188965,
+      "learning_rate": 1.4617e-05,
+      "loss": 2.8333,
+      "step": 70800
+    },
+    {
+      "epoch": 1.417,
+      "grad_norm": 7.930309772491455,
+      "learning_rate": 1.4592e-05,
+      "loss": 2.91,
+      "step": 70850
+    },
+    {
+      "epoch": 1.418,
+      "grad_norm": 8.551098823547363,
+      "learning_rate": 1.4567e-05,
+      "loss": 2.8685,
+      "step": 70900
+    },
+    {
+      "epoch": 1.419,
+      "grad_norm": 8.798742294311523,
+      "learning_rate": 1.4542e-05,
+      "loss": 2.8343,
+      "step": 70950
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.70396900177002,
+      "learning_rate": 1.4517e-05,
+      "loss": 2.9567,
+      "step": 71000
+    },
+    {
+      "epoch": 1.421,
+      "grad_norm": 9.252113342285156,
+      "learning_rate": 1.4492e-05,
+      "loss": 2.9396,
+      "step": 71050
+    },
+    {
+      "epoch": 1.422,
+      "grad_norm": 8.939455032348633,
+      "learning_rate": 1.4467e-05,
+      "loss": 2.5287,
+      "step": 71100
+    },
+    {
+      "epoch": 1.423,
+      "grad_norm": 9.436803817749023,
+      "learning_rate": 1.4442e-05,
+      "loss": 2.8976,
+      "step": 71150
+    },
+    {
+      "epoch": 1.424,
+      "grad_norm": 8.383010864257812,
+      "learning_rate": 1.4417e-05,
+      "loss": 2.7318,
+      "step": 71200
+    },
+    {
+      "epoch": 1.425,
+      "grad_norm": 8.16480541229248,
+      "learning_rate": 1.4392e-05,
+      "loss": 2.7685,
+      "step": 71250
+    },
+    {
+      "epoch": 1.426,
+      "grad_norm": 9.6614351272583,
+      "learning_rate": 1.4367e-05,
+      "loss": 2.7841,
+      "step": 71300
+    },
+    {
+      "epoch": 1.427,
+      "grad_norm": 7.361180305480957,
+      "learning_rate": 1.4342e-05,
+      "loss": 2.8441,
+      "step": 71350
+    },
+    {
+      "epoch": 1.428,
+      "grad_norm": 10.832164764404297,
+      "learning_rate": 1.4317e-05,
+      "loss": 2.875,
+      "step": 71400
+    },
+    {
+      "epoch": 1.429,
+      "grad_norm": 6.530023097991943,
+      "learning_rate": 1.4292e-05,
+      "loss": 2.8486,
+      "step": 71450
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.04982852935791,
+      "learning_rate": 1.4266999999999999e-05,
+      "loss": 3.0097,
+      "step": 71500
+    },
+    {
+      "epoch": 1.431,
+      "grad_norm": 7.762668132781982,
+      "learning_rate": 1.4242e-05,
+      "loss": 2.8063,
+      "step": 71550
+    },
+    {
+      "epoch": 1.432,
+      "grad_norm": 6.672072887420654,
+      "learning_rate": 1.4217e-05,
+      "loss": 2.752,
+      "step": 71600
+    },
+    {
+      "epoch": 1.433,
+      "grad_norm": 7.793580532073975,
+      "learning_rate": 1.4191999999999999e-05,
+      "loss": 2.7347,
+      "step": 71650
+    },
+    {
+      "epoch": 1.434,
+      "grad_norm": 7.609199523925781,
+      "learning_rate": 1.4167e-05,
+      "loss": 2.904,
+      "step": 71700
+    },
+    {
+      "epoch": 1.435,
+      "grad_norm": 8.817373275756836,
+      "learning_rate": 1.4142e-05,
+      "loss": 2.7914,
+      "step": 71750
+    },
+    {
+      "epoch": 1.436,
+      "grad_norm": 9.75827693939209,
+      "learning_rate": 1.4116999999999999e-05,
+      "loss": 2.8292,
+      "step": 71800
+    },
+    {
+      "epoch": 1.437,
+      "grad_norm": 6.9642863273620605,
+      "learning_rate": 1.4092e-05,
+      "loss": 2.6669,
+      "step": 71850
+    },
+    {
+      "epoch": 1.438,
+      "grad_norm": 9.276019096374512,
+      "learning_rate": 1.4067e-05,
+      "loss": 2.984,
+      "step": 71900
+    },
+    {
+      "epoch": 1.439,
+      "grad_norm": 7.211697578430176,
+      "learning_rate": 1.4042e-05,
+      "loss": 2.9085,
+      "step": 71950
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 7.778343200683594,
+      "learning_rate": 1.4017e-05,
+      "loss": 2.897,
+      "step": 72000
+    },
+    {
+      "epoch": 1.441,
+      "grad_norm": 8.795747756958008,
+      "learning_rate": 1.3992e-05,
+      "loss": 2.7523,
+      "step": 72050
+    },
+    {
+      "epoch": 1.442,
+      "grad_norm": 7.971023082733154,
+      "learning_rate": 1.3967e-05,
+      "loss": 2.895,
+      "step": 72100
+    },
+    {
+      "epoch": 1.443,
+      "grad_norm": 8.681403160095215,
+      "learning_rate": 1.3942e-05,
+      "loss": 3.0107,
+      "step": 72150
+    },
+    {
+      "epoch": 1.444,
+      "grad_norm": 7.388852596282959,
+      "learning_rate": 1.3916999999999999e-05,
+      "loss": 2.7337,
+      "step": 72200
+    },
+    {
+      "epoch": 1.445,
+      "grad_norm": 6.82358980178833,
+      "learning_rate": 1.3892e-05,
+      "loss": 2.7527,
+      "step": 72250
+    },
+    {
+      "epoch": 1.446,
+      "grad_norm": 7.868986129760742,
+      "learning_rate": 1.3867e-05,
+      "loss": 2.8683,
+      "step": 72300
+    },
+    {
+      "epoch": 1.447,
+      "grad_norm": 7.227389812469482,
+      "learning_rate": 1.3841999999999999e-05,
+      "loss": 2.8543,
+      "step": 72350
+    },
+    {
+      "epoch": 1.448,
+      "grad_norm": 6.622435569763184,
+      "learning_rate": 1.3817e-05,
+      "loss": 2.8388,
+      "step": 72400
+    },
+    {
+      "epoch": 1.449,
+      "grad_norm": 6.9106903076171875,
+      "learning_rate": 1.3792e-05,
+      "loss": 2.7931,
+      "step": 72450
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.011186599731445,
+      "learning_rate": 1.3766999999999999e-05,
+      "loss": 2.9716,
+      "step": 72500
+    },
+    {
+      "epoch": 1.451,
+      "grad_norm": 8.601675987243652,
+      "learning_rate": 1.3742e-05,
+      "loss": 2.8998,
+      "step": 72550
+    },
+    {
+      "epoch": 1.452,
+      "grad_norm": 6.609960079193115,
+      "learning_rate": 1.3717e-05,
+      "loss": 2.5795,
+      "step": 72600
+    },
+    {
+      "epoch": 1.453,
+      "grad_norm": 7.050849914550781,
+      "learning_rate": 1.3691999999999999e-05,
+      "loss": 2.8272,
+      "step": 72650
+    },
+    {
+      "epoch": 1.454,
+      "grad_norm": 9.343693733215332,
+      "learning_rate": 1.3667000000000001e-05,
+      "loss": 2.6923,
+      "step": 72700
+    },
+    {
+      "epoch": 1.455,
+      "grad_norm": 7.410576343536377,
+      "learning_rate": 1.3642000000000003e-05,
+      "loss": 2.7491,
+      "step": 72750
+    },
+    {
+      "epoch": 1.456,
+      "grad_norm": 7.9082159996032715,
+      "learning_rate": 1.3617000000000002e-05,
+      "loss": 3.0432,
+      "step": 72800
+    },
+    {
+      "epoch": 1.457,
+      "grad_norm": 10.245935440063477,
+      "learning_rate": 1.3592000000000001e-05,
+      "loss": 2.7326,
+      "step": 72850
+    },
+    {
+      "epoch": 1.458,
+      "grad_norm": 7.3419318199157715,
+      "learning_rate": 1.3567000000000002e-05,
+      "loss": 2.7541,
+      "step": 72900
+    },
+    {
+      "epoch": 1.459,
+      "grad_norm": 7.791906833648682,
+      "learning_rate": 1.3542000000000002e-05,
+      "loss": 2.8323,
+      "step": 72950
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 8.035292625427246,
+      "learning_rate": 1.3517000000000001e-05,
+      "loss": 2.5501,
+      "step": 73000
+    },
+    {
+      "epoch": 1.461,
+      "grad_norm": 8.372005462646484,
+      "learning_rate": 1.3492000000000002e-05,
+      "loss": 2.7617,
+      "step": 73050
+    },
+    {
+      "epoch": 1.462,
+      "grad_norm": 8.679768562316895,
+      "learning_rate": 1.3467000000000002e-05,
+      "loss": 2.7125,
+      "step": 73100
+    },
+    {
+      "epoch": 1.463,
+      "grad_norm": 15.464470863342285,
+      "learning_rate": 1.3442000000000001e-05,
+      "loss": 2.9627,
+      "step": 73150
+    },
+    {
+      "epoch": 1.464,
+      "grad_norm": 7.9514007568359375,
+      "learning_rate": 1.3417000000000002e-05,
+      "loss": 3.045,
+      "step": 73200
+    },
+    {
+      "epoch": 1.465,
+      "grad_norm": 6.683189868927002,
+      "learning_rate": 1.3392000000000002e-05,
+      "loss": 2.7564,
+      "step": 73250
+    },
+    {
+      "epoch": 1.466,
+      "grad_norm": 11.89483642578125,
+      "learning_rate": 1.3367000000000001e-05,
+      "loss": 2.9175,
+      "step": 73300
+    },
+    {
+      "epoch": 1.467,
+      "grad_norm": 8.024170875549316,
+      "learning_rate": 1.3342000000000002e-05,
+      "loss": 2.6941,
+      "step": 73350
+    },
+    {
+      "epoch": 1.468,
+      "grad_norm": 8.924114227294922,
+      "learning_rate": 1.3317000000000001e-05,
+      "loss": 2.8156,
+      "step": 73400
+    },
+    {
+      "epoch": 1.4689999999999999,
+      "grad_norm": 8.561711311340332,
+      "learning_rate": 1.3292000000000003e-05,
+      "loss": 2.8866,
+      "step": 73450
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.818349838256836,
+      "learning_rate": 1.3267000000000002e-05,
+      "loss": 2.7045,
+      "step": 73500
+    },
+    {
+      "epoch": 1.471,
+      "grad_norm": 9.295713424682617,
+      "learning_rate": 1.3242000000000001e-05,
+      "loss": 2.8804,
+      "step": 73550
+    },
+    {
+      "epoch": 1.472,
+      "grad_norm": 6.626889705657959,
+      "learning_rate": 1.3217000000000002e-05,
+      "loss": 2.8654,
+      "step": 73600
+    },
+    {
+      "epoch": 1.4729999999999999,
+      "grad_norm": 10.385565757751465,
+      "learning_rate": 1.3192000000000002e-05,
+      "loss": 2.8314,
+      "step": 73650
+    },
+    {
+      "epoch": 1.474,
+      "grad_norm": 8.731403350830078,
+      "learning_rate": 1.3167000000000001e-05,
+      "loss": 2.8624,
+      "step": 73700
+    },
+    {
+      "epoch": 1.475,
+      "grad_norm": 7.520509719848633,
+      "learning_rate": 1.31425e-05,
+      "loss": 2.8683,
+      "step": 73750
+    },
+    {
+      "epoch": 1.476,
+      "grad_norm": 8.531784057617188,
+      "learning_rate": 1.3118e-05,
+      "loss": 2.812,
+      "step": 73800
+    },
+    {
+      "epoch": 1.4769999999999999,
+      "grad_norm": 7.280162334442139,
+      "learning_rate": 1.3093e-05,
+      "loss": 2.756,
+      "step": 73850
+    },
+    {
+      "epoch": 1.478,
+      "grad_norm": 8.175580024719238,
+      "learning_rate": 1.3068e-05,
+      "loss": 2.7574,
+      "step": 73900
+    },
+    {
+      "epoch": 1.479,
+      "grad_norm": 7.7262091636657715,
+      "learning_rate": 1.3043e-05,
+      "loss": 2.6983,
+      "step": 73950
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 7.174941062927246,
+      "learning_rate": 1.3018e-05,
+      "loss": 2.8821,
+      "step": 74000
+    },
+    {
+      "epoch": 1.4809999999999999,
+      "grad_norm": 9.21951961517334,
+      "learning_rate": 1.2992999999999999e-05,
+      "loss": 2.8512,
+      "step": 74050
+    },
+    {
+      "epoch": 1.482,
+      "grad_norm": 13.13577651977539,
+      "learning_rate": 1.2968e-05,
+      "loss": 2.7765,
+      "step": 74100
+    },
+    {
+      "epoch": 1.483,
+      "grad_norm": 7.468155384063721,
+      "learning_rate": 1.2943e-05,
+      "loss": 2.8334,
+      "step": 74150
+    },
+    {
+      "epoch": 1.484,
+      "grad_norm": 7.371670722961426,
+      "learning_rate": 1.2917999999999999e-05,
+      "loss": 2.5598,
+      "step": 74200
+    },
+    {
+      "epoch": 1.4849999999999999,
+      "grad_norm": 7.9383745193481445,
+      "learning_rate": 1.2893e-05,
+      "loss": 2.8735,
+      "step": 74250
+    },
+    {
+      "epoch": 1.486,
+      "grad_norm": 7.459591865539551,
+      "learning_rate": 1.2868e-05,
+      "loss": 2.8774,
+      "step": 74300
+    },
+    {
+      "epoch": 1.487,
+      "grad_norm": 9.01418685913086,
+      "learning_rate": 1.2842999999999999e-05,
+      "loss": 2.9084,
+      "step": 74350
+    },
+    {
+      "epoch": 1.488,
+      "grad_norm": 8.918654441833496,
+      "learning_rate": 1.2818e-05,
+      "loss": 2.7615,
+      "step": 74400
+    },
+    {
+      "epoch": 1.4889999999999999,
+      "grad_norm": 8.913796424865723,
+      "learning_rate": 1.2793e-05,
+      "loss": 2.5816,
+      "step": 74450
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 7.444356918334961,
+      "learning_rate": 1.2767999999999999e-05,
+      "loss": 2.7854,
+      "step": 74500
+    },
+    {
+      "epoch": 1.491,
+      "grad_norm": 11.2793550491333,
+      "learning_rate": 1.2743e-05,
+      "loss": 2.7344,
+      "step": 74550
+    },
+    {
+      "epoch": 1.492,
+      "grad_norm": 6.877553939819336,
+      "learning_rate": 1.2718e-05,
+      "loss": 2.81,
+      "step": 74600
+    },
+    {
+      "epoch": 1.4929999999999999,
+      "grad_norm": 8.273252487182617,
+      "learning_rate": 1.2692999999999999e-05,
+      "loss": 2.7148,
+      "step": 74650
+    },
+    {
+      "epoch": 1.494,
+      "grad_norm": 6.85491418838501,
+      "learning_rate": 1.2668e-05,
+      "loss": 2.848,
+      "step": 74700
+    },
+    {
+      "epoch": 1.495,
+      "grad_norm": 7.391516208648682,
+      "learning_rate": 1.2642999999999999e-05,
+      "loss": 2.9437,
+      "step": 74750
+    },
+    {
+      "epoch": 1.496,
+      "grad_norm": 8.175915718078613,
+      "learning_rate": 1.2617999999999998e-05,
+      "loss": 2.5342,
+      "step": 74800
+    },
+    {
+      "epoch": 1.4969999999999999,
+      "grad_norm": 7.925018310546875,
+      "learning_rate": 1.2593e-05,
+      "loss": 2.8257,
+      "step": 74850
+    },
+    {
+      "epoch": 1.498,
+      "grad_norm": 7.637334823608398,
+      "learning_rate": 1.2567999999999999e-05,
+      "loss": 2.7577,
+      "step": 74900
+    },
+    {
+      "epoch": 1.499,
+      "grad_norm": 6.907679557800293,
+      "learning_rate": 1.2543000000000002e-05,
+      "loss": 2.6929,
+      "step": 74950
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 12.479944229125977,
+      "learning_rate": 1.2518000000000001e-05,
+      "loss": 2.6092,
+      "step": 75000
+    },
+    {
+      "epoch": 1.501,
+      "grad_norm": 8.500955581665039,
+      "learning_rate": 1.2493e-05,
+      "loss": 2.9479,
+      "step": 75050
+    },
+    {
+      "epoch": 1.502,
+      "grad_norm": 8.31230640411377,
+      "learning_rate": 1.2468e-05,
+      "loss": 2.7979,
+      "step": 75100
+    },
+    {
+      "epoch": 1.5030000000000001,
+      "grad_norm": 7.513004779815674,
+      "learning_rate": 1.2443000000000001e-05,
+      "loss": 2.8403,
+      "step": 75150
+    },
+    {
+      "epoch": 1.504,
+      "grad_norm": 9.051568984985352,
+      "learning_rate": 1.2418e-05,
+      "loss": 2.5605,
+      "step": 75200
+    },
+    {
+      "epoch": 1.505,
+      "grad_norm": 9.6289644241333,
+      "learning_rate": 1.2393e-05,
+      "loss": 2.91,
+      "step": 75250
+    },
+    {
+      "epoch": 1.506,
+      "grad_norm": 7.6557440757751465,
+      "learning_rate": 1.2368e-05,
+      "loss": 2.7622,
+      "step": 75300
+    },
+    {
+      "epoch": 1.5070000000000001,
+      "grad_norm": 9.397058486938477,
+      "learning_rate": 1.2343e-05,
+      "loss": 2.5693,
+      "step": 75350
+    },
+    {
+      "epoch": 1.508,
+      "grad_norm": 7.9182634353637695,
+      "learning_rate": 1.2318e-05,
+      "loss": 2.7578,
+      "step": 75400
+    },
+    {
+      "epoch": 1.509,
+      "grad_norm": 8.215141296386719,
+      "learning_rate": 1.2293e-05,
+      "loss": 2.8933,
+      "step": 75450
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 6.805432319641113,
+      "learning_rate": 1.2268e-05,
+      "loss": 2.8146,
+      "step": 75500
+    },
+    {
+      "epoch": 1.5110000000000001,
+      "grad_norm": 8.983750343322754,
+      "learning_rate": 1.2243e-05,
+      "loss": 2.8299,
+      "step": 75550
+    },
+    {
+      "epoch": 1.512,
+      "grad_norm": 8.060677528381348,
+      "learning_rate": 1.2218e-05,
+      "loss": 2.8027,
+      "step": 75600
+    },
+    {
+      "epoch": 1.513,
+      "grad_norm": 7.529645919799805,
+      "learning_rate": 1.2193e-05,
+      "loss": 2.752,
+      "step": 75650
+    },
+    {
+      "epoch": 1.514,
+      "grad_norm": 8.330324172973633,
+      "learning_rate": 1.2168000000000001e-05,
+      "loss": 2.8057,
+      "step": 75700
+    },
+    {
+      "epoch": 1.5150000000000001,
+      "grad_norm": 8.606231689453125,
+      "learning_rate": 1.2143e-05,
+      "loss": 2.881,
+      "step": 75750
+    },
+    {
+      "epoch": 1.516,
+      "grad_norm": 7.919892311096191,
+      "learning_rate": 1.2118e-05,
+      "loss": 2.7935,
+      "step": 75800
+    },
+    {
+      "epoch": 1.517,
+      "grad_norm": 7.2072954177856445,
+      "learning_rate": 1.2093000000000001e-05,
+      "loss": 2.6488,
+      "step": 75850
+    },
+    {
+      "epoch": 1.518,
+      "grad_norm": 7.602199077606201,
+      "learning_rate": 1.2068e-05,
+      "loss": 2.7267,
+      "step": 75900
+    },
+    {
+      "epoch": 1.5190000000000001,
+      "grad_norm": 7.478748321533203,
+      "learning_rate": 1.2043e-05,
+      "loss": 2.8507,
+      "step": 75950
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.189718246459961,
+      "learning_rate": 1.2018e-05,
+      "loss": 2.797,
+      "step": 76000
+    },
+    {
+      "epoch": 1.521,
+      "grad_norm": 9.281253814697266,
+      "learning_rate": 1.1993e-05,
+      "loss": 2.7886,
+      "step": 76050
+    },
+    {
+      "epoch": 1.522,
+      "grad_norm": 7.495121002197266,
+      "learning_rate": 1.1968e-05,
+      "loss": 2.8119,
+      "step": 76100
+    },
+    {
+      "epoch": 1.5230000000000001,
+      "grad_norm": 7.031370162963867,
+      "learning_rate": 1.1943e-05,
+      "loss": 2.9879,
+      "step": 76150
+    },
+    {
+      "epoch": 1.524,
+      "grad_norm": 9.18350601196289,
+      "learning_rate": 1.1918e-05,
+      "loss": 2.8631,
+      "step": 76200
+    },
+    {
+      "epoch": 1.525,
+      "grad_norm": 7.201645851135254,
+      "learning_rate": 1.1893e-05,
+      "loss": 2.5649,
+      "step": 76250
+    },
+    {
+      "epoch": 1.526,
+      "grad_norm": 9.630722045898438,
+      "learning_rate": 1.1868e-05,
+      "loss": 2.9442,
+      "step": 76300
+    },
+    {
+      "epoch": 1.5270000000000001,
+      "grad_norm": 8.107288360595703,
+      "learning_rate": 1.1843e-05,
+      "loss": 2.7623,
+      "step": 76350
+    },
+    {
+      "epoch": 1.528,
+      "grad_norm": 7.947506427764893,
+      "learning_rate": 1.1818e-05,
+      "loss": 2.5865,
+      "step": 76400
+    },
+    {
+      "epoch": 1.529,
+      "grad_norm": 10.158778190612793,
+      "learning_rate": 1.1793e-05,
+      "loss": 2.7721,
+      "step": 76450
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 11.925228118896484,
+      "learning_rate": 1.1768000000000002e-05,
+      "loss": 2.8884,
+      "step": 76500
+    },
+    {
+      "epoch": 1.5310000000000001,
+      "grad_norm": 7.677661418914795,
+      "learning_rate": 1.1743000000000001e-05,
+      "loss": 2.5829,
+      "step": 76550
+    },
+    {
+      "epoch": 1.532,
+      "grad_norm": 9.155203819274902,
+      "learning_rate": 1.1718000000000002e-05,
+      "loss": 2.8175,
+      "step": 76600
+    },
+    {
+      "epoch": 1.533,
+      "grad_norm": 8.825461387634277,
+      "learning_rate": 1.1693000000000001e-05,
+      "loss": 2.8015,
+      "step": 76650
+    },
+    {
+      "epoch": 1.534,
+      "grad_norm": 7.908650875091553,
+      "learning_rate": 1.1668e-05,
+      "loss": 2.8491,
+      "step": 76700
+    },
+    {
+      "epoch": 1.5350000000000001,
+      "grad_norm": 8.90393352508545,
+      "learning_rate": 1.1643000000000002e-05,
+      "loss": 2.6554,
+      "step": 76750
+    },
+    {
+      "epoch": 1.536,
+      "grad_norm": 12.58026123046875,
+      "learning_rate": 1.1618000000000001e-05,
+      "loss": 2.9377,
+      "step": 76800
+    },
+    {
+      "epoch": 1.537,
+      "grad_norm": 7.935201644897461,
+      "learning_rate": 1.1593e-05,
+      "loss": 2.9783,
+      "step": 76850
+    },
+    {
+      "epoch": 1.538,
+      "grad_norm": 14.243797302246094,
+      "learning_rate": 1.1568000000000002e-05,
+      "loss": 2.7387,
+      "step": 76900
+    },
+    {
+      "epoch": 1.5390000000000001,
+      "grad_norm": 7.501498699188232,
+      "learning_rate": 1.1543000000000001e-05,
+      "loss": 2.8034,
+      "step": 76950
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.914602279663086,
+      "learning_rate": 1.1518e-05,
+      "loss": 2.9279,
+      "step": 77000
+    },
+    {
+      "epoch": 1.541,
+      "grad_norm": 8.507695198059082,
+      "learning_rate": 1.1493000000000002e-05,
+      "loss": 2.9641,
+      "step": 77050
+    },
+    {
+      "epoch": 1.542,
+      "grad_norm": 7.860219955444336,
+      "learning_rate": 1.1468000000000001e-05,
+      "loss": 2.8456,
+      "step": 77100
+    },
+    {
+      "epoch": 1.5430000000000001,
+      "grad_norm": 8.061058044433594,
+      "learning_rate": 1.1443e-05,
+      "loss": 2.6333,
+      "step": 77150
+    },
+    {
+      "epoch": 1.544,
+      "grad_norm": 9.513175964355469,
+      "learning_rate": 1.1418000000000001e-05,
+      "loss": 2.6512,
+      "step": 77200
+    },
+    {
+      "epoch": 1.545,
+      "grad_norm": 7.933878421783447,
+      "learning_rate": 1.1393000000000001e-05,
+      "loss": 2.8595,
+      "step": 77250
+    },
+    {
+      "epoch": 1.546,
+      "grad_norm": 9.191105842590332,
+      "learning_rate": 1.1368e-05,
+      "loss": 2.9499,
+      "step": 77300
+    },
+    {
+      "epoch": 1.5470000000000002,
+      "grad_norm": 8.587617874145508,
+      "learning_rate": 1.13435e-05,
+      "loss": 2.7946,
+      "step": 77350
+    },
+    {
+      "epoch": 1.548,
+      "grad_norm": 8.40201187133789,
+      "learning_rate": 1.13185e-05,
+      "loss": 2.935,
+      "step": 77400
+    },
+    {
+      "epoch": 1.549,
+      "grad_norm": 9.933843612670898,
+      "learning_rate": 1.12935e-05,
+      "loss": 2.53,
+      "step": 77450
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.645332336425781,
+      "learning_rate": 1.12685e-05,
+      "loss": 2.7493,
+      "step": 77500
+    },
+    {
+      "epoch": 1.5510000000000002,
+      "grad_norm": 8.894145011901855,
+      "learning_rate": 1.12435e-05,
+      "loss": 2.6933,
+      "step": 77550
+    },
+    {
+      "epoch": 1.552,
+      "grad_norm": 7.8815460205078125,
+      "learning_rate": 1.1218500000000001e-05,
+      "loss": 2.9458,
+      "step": 77600
+    },
+    {
+      "epoch": 1.553,
+      "grad_norm": 7.892411231994629,
+      "learning_rate": 1.11935e-05,
+      "loss": 2.6212,
+      "step": 77650
+    },
+    {
+      "epoch": 1.554,
+      "grad_norm": 9.983649253845215,
+      "learning_rate": 1.1168500000000002e-05,
+      "loss": 2.7327,
+      "step": 77700
+    },
+    {
+      "epoch": 1.5550000000000002,
+      "grad_norm": 8.785431861877441,
+      "learning_rate": 1.1143500000000001e-05,
+      "loss": 2.8999,
+      "step": 77750
+    },
+    {
+      "epoch": 1.556,
+      "grad_norm": 7.402805805206299,
+      "learning_rate": 1.11185e-05,
+      "loss": 3.0402,
+      "step": 77800
+    },
+    {
+      "epoch": 1.557,
+      "grad_norm": 8.64169692993164,
+      "learning_rate": 1.1093500000000001e-05,
+      "loss": 2.8152,
+      "step": 77850
+    },
+    {
+      "epoch": 1.558,
+      "grad_norm": 6.422133445739746,
+      "learning_rate": 1.10685e-05,
+      "loss": 2.7819,
+      "step": 77900
+    },
+    {
+      "epoch": 1.5590000000000002,
+      "grad_norm": 8.386688232421875,
+      "learning_rate": 1.10435e-05,
+      "loss": 2.8271,
+      "step": 77950
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.966951370239258,
+      "learning_rate": 1.1018500000000001e-05,
+      "loss": 2.8392,
+      "step": 78000
+    },
+    {
+      "epoch": 1.561,
+      "grad_norm": 8.008108139038086,
+      "learning_rate": 1.09935e-05,
+      "loss": 2.6764,
+      "step": 78050
+    },
+    {
+      "epoch": 1.562,
+      "grad_norm": 6.586634159088135,
+      "learning_rate": 1.0968500000000002e-05,
+      "loss": 2.6548,
+      "step": 78100
+    },
+    {
+      "epoch": 1.563,
+      "grad_norm": 6.59722375869751,
+      "learning_rate": 1.0943500000000001e-05,
+      "loss": 2.9959,
+      "step": 78150
+    },
+    {
+      "epoch": 1.564,
+      "grad_norm": 8.587728500366211,
+      "learning_rate": 1.09185e-05,
+      "loss": 2.8217,
+      "step": 78200
+    },
+    {
+      "epoch": 1.565,
+      "grad_norm": 8.256272315979004,
+      "learning_rate": 1.0893500000000002e-05,
+      "loss": 2.8172,
+      "step": 78250
+    },
+    {
+      "epoch": 1.5659999999999998,
+      "grad_norm": 9.824472427368164,
+      "learning_rate": 1.0868500000000001e-05,
+      "loss": 2.7618,
+      "step": 78300
+    },
+    {
+      "epoch": 1.567,
+      "grad_norm": 6.0259785652160645,
+      "learning_rate": 1.08435e-05,
+      "loss": 2.776,
+      "step": 78350
+    },
+    {
+      "epoch": 1.568,
+      "grad_norm": 8.524801254272461,
+      "learning_rate": 1.0818500000000002e-05,
+      "loss": 2.8886,
+      "step": 78400
+    },
+    {
+      "epoch": 1.569,
+      "grad_norm": 7.844742298126221,
+      "learning_rate": 1.0793500000000001e-05,
+      "loss": 2.6163,
+      "step": 78450
+    },
+    {
+      "epoch": 1.5699999999999998,
+      "grad_norm": 8.0889253616333,
+      "learning_rate": 1.07685e-05,
+      "loss": 2.6713,
+      "step": 78500
+    },
+    {
+      "epoch": 1.571,
+      "grad_norm": 8.610453605651855,
+      "learning_rate": 1.0743500000000001e-05,
+      "loss": 2.6546,
+      "step": 78550
+    },
+    {
+      "epoch": 1.572,
+      "grad_norm": 10.174454689025879,
+      "learning_rate": 1.07185e-05,
+      "loss": 2.8693,
+      "step": 78600
+    },
+    {
+      "epoch": 1.573,
+      "grad_norm": 9.715279579162598,
+      "learning_rate": 1.06935e-05,
+      "loss": 2.8965,
+      "step": 78650
+    },
+    {
+      "epoch": 1.5739999999999998,
+      "grad_norm": 8.34848403930664,
+      "learning_rate": 1.0668500000000001e-05,
+      "loss": 2.7681,
+      "step": 78700
+    },
+    {
+      "epoch": 1.575,
+      "grad_norm": 10.231193542480469,
+      "learning_rate": 1.06435e-05,
+      "loss": 3.1195,
+      "step": 78750
+    },
+    {
+      "epoch": 1.576,
+      "grad_norm": 7.847792625427246,
+      "learning_rate": 1.06185e-05,
+      "loss": 2.9861,
+      "step": 78800
+    },
+    {
+      "epoch": 1.577,
+      "grad_norm": 12.434806823730469,
+      "learning_rate": 1.0593500000000001e-05,
+      "loss": 2.7446,
+      "step": 78850
+    },
+    {
+      "epoch": 1.5779999999999998,
+      "grad_norm": 7.213706970214844,
+      "learning_rate": 1.05685e-05,
+      "loss": 2.8863,
+      "step": 78900
+    },
+    {
+      "epoch": 1.579,
+      "grad_norm": 13.66837215423584,
+      "learning_rate": 1.05435e-05,
+      "loss": 2.8019,
+      "step": 78950
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 12.315549850463867,
+      "learning_rate": 1.0518500000000001e-05,
+      "loss": 2.6926,
+      "step": 79000
+    },
+    {
+      "epoch": 1.581,
+      "grad_norm": 10.095964431762695,
+      "learning_rate": 1.04935e-05,
+      "loss": 2.6185,
+      "step": 79050
+    },
+    {
+      "epoch": 1.5819999999999999,
+      "grad_norm": 7.9094648361206055,
+      "learning_rate": 1.04685e-05,
+      "loss": 2.9309,
+      "step": 79100
+    },
+    {
+      "epoch": 1.583,
+      "grad_norm": 9.682660102844238,
+      "learning_rate": 1.0443500000000001e-05,
+      "loss": 2.7369,
+      "step": 79150
+    },
+    {
+      "epoch": 1.584,
+      "grad_norm": 8.518179893493652,
+      "learning_rate": 1.04185e-05,
+      "loss": 2.6872,
+      "step": 79200
+    },
+    {
+      "epoch": 1.585,
+      "grad_norm": 8.616283416748047,
+      "learning_rate": 1.03935e-05,
+      "loss": 2.9505,
+      "step": 79250
+    },
+    {
+      "epoch": 1.5859999999999999,
+      "grad_norm": 8.388751983642578,
+      "learning_rate": 1.03685e-05,
+      "loss": 3.0227,
+      "step": 79300
+    },
+    {
+      "epoch": 1.587,
+      "grad_norm": 7.643990993499756,
+      "learning_rate": 1.03435e-05,
+      "loss": 2.684,
+      "step": 79350
+    },
+    {
+      "epoch": 1.588,
+      "grad_norm": 7.016181945800781,
+      "learning_rate": 1.03185e-05,
+      "loss": 3.0096,
+      "step": 79400
+    },
+    {
+      "epoch": 1.589,
+      "grad_norm": 7.859565258026123,
+      "learning_rate": 1.02935e-05,
+      "loss": 2.8234,
+      "step": 79450
+    },
+    {
+      "epoch": 1.5899999999999999,
+      "grad_norm": 8.977822303771973,
+      "learning_rate": 1.02685e-05,
+      "loss": 2.8249,
+      "step": 79500
+    },
+    {
+      "epoch": 1.591,
+      "grad_norm": 9.124774932861328,
+      "learning_rate": 1.0243500000000001e-05,
+      "loss": 2.786,
+      "step": 79550
+    },
+    {
+      "epoch": 1.592,
+      "grad_norm": 7.759701728820801,
+      "learning_rate": 1.02185e-05,
+      "loss": 2.6182,
+      "step": 79600
+    },
+    {
+      "epoch": 1.593,
+      "grad_norm": 9.396001815795898,
+      "learning_rate": 1.01935e-05,
+      "loss": 2.7735,
+      "step": 79650
+    },
+    {
+      "epoch": 1.5939999999999999,
+      "grad_norm": 7.322061538696289,
+      "learning_rate": 1.0168500000000001e-05,
+      "loss": 2.749,
+      "step": 79700
+    },
+    {
+      "epoch": 1.595,
+      "grad_norm": 10.232261657714844,
+      "learning_rate": 1.01435e-05,
+      "loss": 2.5499,
+      "step": 79750
+    },
+    {
+      "epoch": 1.596,
+      "grad_norm": 7.829950332641602,
+      "learning_rate": 1.01185e-05,
+      "loss": 2.6453,
+      "step": 79800
+    },
+    {
+      "epoch": 1.597,
+      "grad_norm": 7.685505390167236,
+      "learning_rate": 1.0093500000000001e-05,
+      "loss": 2.7082,
+      "step": 79850
+    },
+    {
+      "epoch": 1.5979999999999999,
+      "grad_norm": 7.936647415161133,
+      "learning_rate": 1.00685e-05,
+      "loss": 2.734,
+      "step": 79900
+    },
+    {
+      "epoch": 1.599,
+      "grad_norm": 8.177151679992676,
+      "learning_rate": 1.00435e-05,
+      "loss": 2.8945,
+      "step": 79950
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 7.7781782150268555,
+      "learning_rate": 1.00185e-05,
+      "loss": 2.8666,
+      "step": 80000
+    },
+    {
+      "epoch": 1.601,
+      "grad_norm": 8.397016525268555,
+      "learning_rate": 9.9935e-06,
+      "loss": 2.7994,
+      "step": 80050
+    },
+    {
+      "epoch": 1.6019999999999999,
+      "grad_norm": 8.55648422241211,
+      "learning_rate": 9.9685e-06,
+      "loss": 2.8561,
+      "step": 80100
+    },
+    {
+      "epoch": 1.603,
+      "grad_norm": 10.180916786193848,
+      "learning_rate": 9.9435e-06,
+      "loss": 2.7149,
+      "step": 80150
+    },
+    {
+      "epoch": 1.604,
+      "grad_norm": 7.665710926055908,
+      "learning_rate": 9.9185e-06,
+      "loss": 2.7395,
+      "step": 80200
+    },
+    {
+      "epoch": 1.605,
+      "grad_norm": 14.644109725952148,
+      "learning_rate": 9.8935e-06,
+      "loss": 2.7051,
+      "step": 80250
+    },
+    {
+      "epoch": 1.6059999999999999,
+      "grad_norm": 9.358149528503418,
+      "learning_rate": 9.8685e-06,
+      "loss": 2.9969,
+      "step": 80300
+    },
+    {
+      "epoch": 1.607,
+      "grad_norm": 9.007706642150879,
+      "learning_rate": 9.8435e-06,
+      "loss": 2.7834,
+      "step": 80350
+    },
+    {
+      "epoch": 1.608,
+      "grad_norm": 7.31819486618042,
+      "learning_rate": 9.8185e-06,
+      "loss": 2.8133,
+      "step": 80400
+    },
+    {
+      "epoch": 1.609,
+      "grad_norm": 10.879515647888184,
+      "learning_rate": 9.7935e-06,
+      "loss": 2.6609,
+      "step": 80450
+    },
+    {
+      "epoch": 1.6099999999999999,
+      "grad_norm": 7.164587020874023,
+      "learning_rate": 9.7685e-06,
+      "loss": 2.7477,
+      "step": 80500
+    },
+    {
+      "epoch": 1.611,
+      "grad_norm": 7.594228744506836,
+      "learning_rate": 9.743499999999999e-06,
+      "loss": 2.6883,
+      "step": 80550
+    },
+    {
+      "epoch": 1.612,
+      "grad_norm": 8.859107971191406,
+      "learning_rate": 9.7185e-06,
+      "loss": 2.7275,
+      "step": 80600
+    },
+    {
+      "epoch": 1.613,
+      "grad_norm": 8.09995174407959,
+      "learning_rate": 9.6935e-06,
+      "loss": 2.6976,
+      "step": 80650
+    },
+    {
+      "epoch": 1.6139999999999999,
+      "grad_norm": 8.347760200500488,
+      "learning_rate": 9.668499999999999e-06,
+      "loss": 2.7801,
+      "step": 80700
+    },
+    {
+      "epoch": 1.615,
+      "grad_norm": 9.736233711242676,
+      "learning_rate": 9.643500000000002e-06,
+      "loss": 2.812,
+      "step": 80750
+    },
+    {
+      "epoch": 1.616,
+      "grad_norm": 8.58044719696045,
+      "learning_rate": 9.618500000000001e-06,
+      "loss": 2.8104,
+      "step": 80800
+    },
+    {
+      "epoch": 1.617,
+      "grad_norm": 8.394573211669922,
+      "learning_rate": 9.5935e-06,
+      "loss": 2.9928,
+      "step": 80850
+    },
+    {
+      "epoch": 1.6179999999999999,
+      "grad_norm": 8.22284984588623,
+      "learning_rate": 9.568500000000002e-06,
+      "loss": 2.7226,
+      "step": 80900
+    },
+    {
+      "epoch": 1.619,
+      "grad_norm": 8.150736808776855,
+      "learning_rate": 9.543500000000001e-06,
+      "loss": 2.6672,
+      "step": 80950
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.5970635414123535,
+      "learning_rate": 9.5185e-06,
+      "loss": 2.8622,
+      "step": 81000
+    },
+    {
+      "epoch": 1.621,
+      "grad_norm": 10.968223571777344,
+      "learning_rate": 9.493500000000002e-06,
+      "loss": 2.6366,
+      "step": 81050
+    },
+    {
+      "epoch": 1.6219999999999999,
+      "grad_norm": 6.821518898010254,
+      "learning_rate": 9.468500000000001e-06,
+      "loss": 2.8406,
+      "step": 81100
+    },
+    {
+      "epoch": 1.623,
+      "grad_norm": 7.469738483428955,
+      "learning_rate": 9.4435e-06,
+      "loss": 2.8246,
+      "step": 81150
+    },
+    {
+      "epoch": 1.624,
+      "grad_norm": 8.006429672241211,
+      "learning_rate": 9.418500000000001e-06,
+      "loss": 2.8267,
+      "step": 81200
+    },
+    {
+      "epoch": 1.625,
+      "grad_norm": 8.809013366699219,
+      "learning_rate": 9.3935e-06,
+      "loss": 2.7887,
+      "step": 81250
+    },
+    {
+      "epoch": 1.626,
+      "grad_norm": 12.106205940246582,
+      "learning_rate": 9.3685e-06,
+      "loss": 2.6218,
+      "step": 81300
+    },
+    {
+      "epoch": 1.627,
+      "grad_norm": 10.05888843536377,
+      "learning_rate": 9.344e-06,
+      "loss": 2.8547,
+      "step": 81350
+    },
+    {
+      "epoch": 1.6280000000000001,
+      "grad_norm": 8.236129760742188,
+      "learning_rate": 9.319e-06,
+      "loss": 2.872,
+      "step": 81400
+    },
+    {
+      "epoch": 1.629,
+      "grad_norm": 8.157784461975098,
+      "learning_rate": 9.294e-06,
+      "loss": 2.701,
+      "step": 81450
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 6.6969218254089355,
+      "learning_rate": 9.269e-06,
+      "loss": 2.9475,
+      "step": 81500
+    },
+    {
+      "epoch": 1.631,
+      "grad_norm": 8.2749662399292,
+      "learning_rate": 9.244e-06,
+      "loss": 2.7482,
+      "step": 81550
+    },
+    {
+      "epoch": 1.6320000000000001,
+      "grad_norm": 9.816543579101562,
+      "learning_rate": 9.219e-06,
+      "loss": 2.8068,
+      "step": 81600
+    },
+    {
+      "epoch": 1.633,
+      "grad_norm": 8.033219337463379,
+      "learning_rate": 9.194e-06,
+      "loss": 2.9093,
+      "step": 81650
+    },
+    {
+      "epoch": 1.634,
+      "grad_norm": 8.458338737487793,
+      "learning_rate": 9.169e-06,
+      "loss": 2.8396,
+      "step": 81700
+    },
+    {
+      "epoch": 1.635,
+      "grad_norm": 8.59599781036377,
+      "learning_rate": 9.144e-06,
+      "loss": 2.9586,
+      "step": 81750
+    },
+    {
+      "epoch": 1.6360000000000001,
+      "grad_norm": 7.180260181427002,
+      "learning_rate": 9.119e-06,
+      "loss": 2.599,
+      "step": 81800
+    },
+    {
+      "epoch": 1.637,
+      "grad_norm": 6.488671779632568,
+      "learning_rate": 9.094e-06,
+      "loss": 2.8477,
+      "step": 81850
+    },
+    {
+      "epoch": 1.638,
+      "grad_norm": 8.595635414123535,
+      "learning_rate": 9.069e-06,
+      "loss": 2.8529,
+      "step": 81900
+    },
+    {
+      "epoch": 1.639,
+      "grad_norm": 7.423949718475342,
+      "learning_rate": 9.044000000000002e-06,
+      "loss": 2.698,
+      "step": 81950
+    },
+    {
+      "epoch": 1.6400000000000001,
+      "grad_norm": 8.720772743225098,
+      "learning_rate": 9.019000000000001e-06,
+      "loss": 2.8964,
+      "step": 82000
+    },
+    {
+      "epoch": 1.641,
+      "grad_norm": 11.61228084564209,
+      "learning_rate": 8.994e-06,
+      "loss": 2.9746,
+      "step": 82050
+    },
+    {
+      "epoch": 1.642,
+      "grad_norm": 6.805294990539551,
+      "learning_rate": 8.969000000000002e-06,
+      "loss": 2.7932,
+      "step": 82100
+    },
+    {
+      "epoch": 1.643,
+      "grad_norm": 9.611919403076172,
+      "learning_rate": 8.944000000000001e-06,
+      "loss": 2.8123,
+      "step": 82150
+    },
+    {
+      "epoch": 1.6440000000000001,
+      "grad_norm": 9.956771850585938,
+      "learning_rate": 8.919e-06,
+      "loss": 2.8532,
+      "step": 82200
+    },
+    {
+      "epoch": 1.645,
+      "grad_norm": 7.504790306091309,
+      "learning_rate": 8.894000000000002e-06,
+      "loss": 2.6689,
+      "step": 82250
+    },
+    {
+      "epoch": 1.646,
+      "grad_norm": 9.280957221984863,
+      "learning_rate": 8.869000000000001e-06,
+      "loss": 2.7067,
+      "step": 82300
+    },
+    {
+      "epoch": 1.647,
+      "grad_norm": 7.816487789154053,
+      "learning_rate": 8.844e-06,
+      "loss": 2.7388,
+      "step": 82350
+    },
+    {
+      "epoch": 1.6480000000000001,
+      "grad_norm": 6.874303340911865,
+      "learning_rate": 8.819000000000001e-06,
+      "loss": 2.742,
+      "step": 82400
+    },
+    {
+      "epoch": 1.649,
+      "grad_norm": 9.157050132751465,
+      "learning_rate": 8.794e-06,
+      "loss": 2.7184,
+      "step": 82450
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.791154861450195,
+      "learning_rate": 8.769e-06,
+      "loss": 2.5517,
+      "step": 82500
+    },
+    {
+      "epoch": 1.651,
+      "grad_norm": 8.14765453338623,
+      "learning_rate": 8.744000000000001e-06,
+      "loss": 2.9349,
+      "step": 82550
+    },
+    {
+      "epoch": 1.6520000000000001,
+      "grad_norm": 7.623929023742676,
+      "learning_rate": 8.719e-06,
+      "loss": 2.6772,
+      "step": 82600
+    },
+    {
+      "epoch": 1.653,
+      "grad_norm": 8.879040718078613,
+      "learning_rate": 8.694e-06,
+      "loss": 2.8483,
+      "step": 82650
+    },
+    {
+      "epoch": 1.654,
+      "grad_norm": 12.571939468383789,
+      "learning_rate": 8.669000000000001e-06,
+      "loss": 2.6692,
+      "step": 82700
+    },
+    {
+      "epoch": 1.655,
+      "grad_norm": 8.578469276428223,
+      "learning_rate": 8.644e-06,
+      "loss": 2.6946,
+      "step": 82750
+    },
+    {
+      "epoch": 1.6560000000000001,
+      "grad_norm": 8.490697860717773,
+      "learning_rate": 8.619e-06,
+      "loss": 2.8827,
+      "step": 82800
+    },
+    {
+      "epoch": 1.657,
+      "grad_norm": 8.043312072753906,
+      "learning_rate": 8.594000000000001e-06,
+      "loss": 2.648,
+      "step": 82850
+    },
+    {
+      "epoch": 1.658,
+      "grad_norm": 8.180224418640137,
+      "learning_rate": 8.569e-06,
+      "loss": 2.6368,
+      "step": 82900
+    },
+    {
+      "epoch": 1.659,
+      "grad_norm": 7.162864685058594,
+      "learning_rate": 8.544e-06,
+      "loss": 2.7297,
+      "step": 82950
+    },
+    {
+      "epoch": 1.6600000000000001,
+      "grad_norm": 8.248373031616211,
+      "learning_rate": 8.519000000000001e-06,
+      "loss": 2.869,
+      "step": 83000
+    },
+    {
+      "epoch": 1.661,
+      "grad_norm": 6.883781909942627,
+      "learning_rate": 8.494e-06,
+      "loss": 2.7607,
+      "step": 83050
+    },
+    {
+      "epoch": 1.662,
+      "grad_norm": 9.59625244140625,
+      "learning_rate": 8.469e-06,
+      "loss": 2.6657,
+      "step": 83100
+    },
+    {
+      "epoch": 1.663,
+      "grad_norm": 8.114121437072754,
+      "learning_rate": 8.444e-06,
+      "loss": 2.7106,
+      "step": 83150
+    },
+    {
+      "epoch": 1.6640000000000001,
+      "grad_norm": 9.549163818359375,
+      "learning_rate": 8.419e-06,
+      "loss": 2.8075,
+      "step": 83200
+    },
+    {
+      "epoch": 1.665,
+      "grad_norm": 8.121686935424805,
+      "learning_rate": 8.394000000000001e-06,
+      "loss": 2.8488,
+      "step": 83250
+    },
+    {
+      "epoch": 1.666,
+      "grad_norm": 10.259820938110352,
+      "learning_rate": 8.369e-06,
+      "loss": 2.7512,
+      "step": 83300
+    },
+    {
+      "epoch": 1.667,
+      "grad_norm": 10.132661819458008,
+      "learning_rate": 8.344500000000001e-06,
+      "loss": 2.7909,
+      "step": 83350
+    },
+    {
+      "epoch": 1.6680000000000001,
+      "grad_norm": 7.976057052612305,
+      "learning_rate": 8.3195e-06,
+      "loss": 2.6088,
+      "step": 83400
+    },
+    {
+      "epoch": 1.669,
+      "grad_norm": 9.32275104522705,
+      "learning_rate": 8.294500000000002e-06,
+      "loss": 2.669,
+      "step": 83450
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.177163124084473,
+      "learning_rate": 8.269500000000001e-06,
+      "loss": 2.7145,
+      "step": 83500
+    },
+    {
+      "epoch": 1.671,
+      "grad_norm": 6.75959587097168,
+      "learning_rate": 8.2445e-06,
+      "loss": 2.8134,
+      "step": 83550
+    },
+    {
+      "epoch": 1.6720000000000002,
+      "grad_norm": 8.769372940063477,
+      "learning_rate": 8.219500000000002e-06,
+      "loss": 2.6065,
+      "step": 83600
+    },
+    {
+      "epoch": 1.673,
+      "grad_norm": 7.8767900466918945,
+      "learning_rate": 8.194500000000001e-06,
+      "loss": 2.7516,
+      "step": 83650
+    },
+    {
+      "epoch": 1.674,
+      "grad_norm": 8.258367538452148,
+      "learning_rate": 8.1695e-06,
+      "loss": 2.8359,
+      "step": 83700
+    },
+    {
+      "epoch": 1.675,
+      "grad_norm": 6.696470260620117,
+      "learning_rate": 8.144500000000001e-06,
+      "loss": 2.7341,
+      "step": 83750
+    },
+    {
+      "epoch": 1.6760000000000002,
+      "grad_norm": 10.333334922790527,
+      "learning_rate": 8.1195e-06,
+      "loss": 2.8067,
+      "step": 83800
+    },
+    {
+      "epoch": 1.677,
+      "grad_norm": 6.307063102722168,
+      "learning_rate": 8.0945e-06,
+      "loss": 2.9026,
+      "step": 83850
+    },
+    {
+      "epoch": 1.678,
+      "grad_norm": 8.203290939331055,
+      "learning_rate": 8.069500000000001e-06,
+      "loss": 2.7649,
+      "step": 83900
+    },
+    {
+      "epoch": 1.679,
+      "grad_norm": 7.663438320159912,
+      "learning_rate": 8.0445e-06,
+      "loss": 2.7698,
+      "step": 83950
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 17.893810272216797,
+      "learning_rate": 8.0195e-06,
+      "loss": 2.7126,
+      "step": 84000
+    },
+    {
+      "epoch": 1.681,
+      "grad_norm": 7.697596549987793,
+      "learning_rate": 7.994500000000001e-06,
+      "loss": 2.7735,
+      "step": 84050
+    },
+    {
+      "epoch": 1.682,
+      "grad_norm": 8.048392295837402,
+      "learning_rate": 7.9695e-06,
+      "loss": 2.8021,
+      "step": 84100
+    },
+    {
+      "epoch": 1.683,
+      "grad_norm": 7.380740642547607,
+      "learning_rate": 7.9445e-06,
+      "loss": 2.856,
+      "step": 84150
+    },
+    {
+      "epoch": 1.6840000000000002,
+      "grad_norm": 7.892630577087402,
+      "learning_rate": 7.919500000000001e-06,
+      "loss": 2.7246,
+      "step": 84200
+    },
+    {
+      "epoch": 1.685,
+      "grad_norm": 8.605746269226074,
+      "learning_rate": 7.8945e-06,
+      "loss": 2.8434,
+      "step": 84250
+    },
+    {
+      "epoch": 1.686,
+      "grad_norm": 9.128484725952148,
+      "learning_rate": 7.8695e-06,
+      "loss": 2.7519,
+      "step": 84300
+    },
+    {
+      "epoch": 1.687,
+      "grad_norm": 10.054672241210938,
+      "learning_rate": 7.844500000000001e-06,
+      "loss": 2.6398,
+      "step": 84350
+    },
+    {
+      "epoch": 1.688,
+      "grad_norm": 9.681634902954102,
+      "learning_rate": 7.8195e-06,
+      "loss": 2.8255,
+      "step": 84400
+    },
+    {
+      "epoch": 1.689,
+      "grad_norm": 10.02967357635498,
+      "learning_rate": 7.7945e-06,
+      "loss": 2.5538,
+      "step": 84450
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.61908483505249,
+      "learning_rate": 7.7695e-06,
+      "loss": 2.7887,
+      "step": 84500
+    },
+    {
+      "epoch": 1.6909999999999998,
+      "grad_norm": 8.514749526977539,
+      "learning_rate": 7.7445e-06,
+      "loss": 2.8793,
+      "step": 84550
+    },
+    {
+      "epoch": 1.692,
+      "grad_norm": 7.921048164367676,
+      "learning_rate": 7.7195e-06,
+      "loss": 2.8279,
+      "step": 84600
+    },
+    {
+      "epoch": 1.693,
+      "grad_norm": 5.613151550292969,
+      "learning_rate": 7.6945e-06,
+      "loss": 2.6869,
+      "step": 84650
+    },
+    {
+      "epoch": 1.694,
+      "grad_norm": 7.689486503601074,
+      "learning_rate": 7.6695e-06,
+      "loss": 2.9914,
+      "step": 84700
+    },
+    {
+      "epoch": 1.6949999999999998,
+      "grad_norm": 8.673030853271484,
+      "learning_rate": 7.644500000000001e-06,
+      "loss": 2.7578,
+      "step": 84750
+    },
+    {
+      "epoch": 1.696,
+      "grad_norm": 7.168893814086914,
+      "learning_rate": 7.6195000000000005e-06,
+      "loss": 2.7542,
+      "step": 84800
+    },
+    {
+      "epoch": 1.697,
+      "grad_norm": 6.62797212600708,
+      "learning_rate": 7.5945e-06,
+      "loss": 2.5553,
+      "step": 84850
+    },
+    {
+      "epoch": 1.698,
+      "grad_norm": 8.141118049621582,
+      "learning_rate": 7.5695e-06,
+      "loss": 2.6875,
+      "step": 84900
+    },
+    {
+      "epoch": 1.6989999999999998,
+      "grad_norm": 9.107691764831543,
+      "learning_rate": 7.5445e-06,
+      "loss": 2.6884,
+      "step": 84950
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.87322998046875,
+      "learning_rate": 7.5195e-06,
+      "loss": 2.8789,
+      "step": 85000
+    },
+    {
+      "epoch": 1.701,
+      "grad_norm": 7.496004104614258,
+      "learning_rate": 7.4945e-06,
+      "loss": 2.6971,
+      "step": 85050
+    },
+    {
+      "epoch": 1.702,
+      "grad_norm": 9.35421371459961,
+      "learning_rate": 7.4695e-06,
+      "loss": 2.5611,
+      "step": 85100
+    },
+    {
+      "epoch": 1.7029999999999998,
+      "grad_norm": 9.931110382080078,
+      "learning_rate": 7.4445000000000005e-06,
+      "loss": 2.9122,
+      "step": 85150
+    },
+    {
+      "epoch": 1.704,
+      "grad_norm": 8.53326416015625,
+      "learning_rate": 7.4195e-06,
+      "loss": 2.8646,
+      "step": 85200
+    },
+    {
+      "epoch": 1.705,
+      "grad_norm": 7.935835838317871,
+      "learning_rate": 7.3945e-06,
+      "loss": 2.7549,
+      "step": 85250
+    },
+    {
+      "epoch": 1.706,
+      "grad_norm": 8.31574821472168,
+      "learning_rate": 7.3695e-06,
+      "loss": 2.7098,
+      "step": 85300
+    },
+    {
+      "epoch": 1.7069999999999999,
+      "grad_norm": 8.184920310974121,
+      "learning_rate": 7.3445e-06,
+      "loss": 2.6436,
+      "step": 85350
+    },
+    {
+      "epoch": 1.708,
+      "grad_norm": 6.512944221496582,
+      "learning_rate": 7.32e-06,
+      "loss": 2.6321,
+      "step": 85400
+    },
+    {
+      "epoch": 1.709,
+      "grad_norm": 7.280327796936035,
+      "learning_rate": 7.2950000000000005e-06,
+      "loss": 2.9698,
+      "step": 85450
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 11.694329261779785,
+      "learning_rate": 7.270000000000001e-06,
+      "loss": 2.8015,
+      "step": 85500
+    },
+    {
+      "epoch": 1.7109999999999999,
+      "grad_norm": 10.256030082702637,
+      "learning_rate": 7.245e-06,
+      "loss": 2.6892,
+      "step": 85550
+    },
+    {
+      "epoch": 1.712,
+      "grad_norm": 8.213347434997559,
+      "learning_rate": 7.22e-06,
+      "loss": 2.7122,
+      "step": 85600
+    },
+    {
+      "epoch": 1.713,
+      "grad_norm": 10.824342727661133,
+      "learning_rate": 7.1950000000000006e-06,
+      "loss": 2.7303,
+      "step": 85650
+    },
+    {
+      "epoch": 1.714,
+      "grad_norm": 10.240262985229492,
+      "learning_rate": 7.17e-06,
+      "loss": 2.8331,
+      "step": 85700
+    },
+    {
+      "epoch": 1.7149999999999999,
+      "grad_norm": 10.220215797424316,
+      "learning_rate": 7.145e-06,
+      "loss": 2.7955,
+      "step": 85750
+    },
+    {
+      "epoch": 1.716,
+      "grad_norm": 11.018972396850586,
+      "learning_rate": 7.1200000000000004e-06,
+      "loss": 2.5736,
+      "step": 85800
+    },
+    {
+      "epoch": 1.717,
+      "grad_norm": 9.33096694946289,
+      "learning_rate": 7.095000000000001e-06,
+      "loss": 2.5607,
+      "step": 85850
+    },
+    {
+      "epoch": 1.718,
+      "grad_norm": 8.596012115478516,
+      "learning_rate": 7.07e-06,
+      "loss": 2.8816,
+      "step": 85900
+    },
+    {
+      "epoch": 1.7189999999999999,
+      "grad_norm": 6.443115234375,
+      "learning_rate": 7.045e-06,
+      "loss": 2.5941,
+      "step": 85950
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.386388778686523,
+      "learning_rate": 7.0200000000000006e-06,
+      "loss": 2.67,
+      "step": 86000
+    },
+    {
+      "epoch": 1.721,
+      "grad_norm": 8.550931930541992,
+      "learning_rate": 6.995e-06,
+      "loss": 2.8201,
+      "step": 86050
+    },
+    {
+      "epoch": 1.722,
+      "grad_norm": 7.8057475090026855,
+      "learning_rate": 6.97e-06,
+      "loss": 2.672,
+      "step": 86100
+    },
+    {
+      "epoch": 1.7229999999999999,
+      "grad_norm": 7.748495578765869,
+      "learning_rate": 6.945e-06,
+      "loss": 2.7135,
+      "step": 86150
+    },
+    {
+      "epoch": 1.724,
+      "grad_norm": 9.472148895263672,
+      "learning_rate": 6.92e-06,
+      "loss": 2.8646,
+      "step": 86200
+    },
+    {
+      "epoch": 1.725,
+      "grad_norm": 8.10639762878418,
+      "learning_rate": 6.895e-06,
+      "loss": 2.6441,
+      "step": 86250
+    },
+    {
+      "epoch": 1.726,
+      "grad_norm": 16.613393783569336,
+      "learning_rate": 6.87e-06,
+      "loss": 2.7735,
+      "step": 86300
+    },
+    {
+      "epoch": 1.7269999999999999,
+      "grad_norm": 8.373213768005371,
+      "learning_rate": 6.845e-06,
+      "loss": 2.7388,
+      "step": 86350
+    },
+    {
+      "epoch": 1.728,
+      "grad_norm": 10.412379264831543,
+      "learning_rate": 6.82e-06,
+      "loss": 2.8095,
+      "step": 86400
+    },
+    {
+      "epoch": 1.729,
+      "grad_norm": 8.624077796936035,
+      "learning_rate": 6.795e-06,
+      "loss": 2.8076,
+      "step": 86450
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.361799240112305,
+      "learning_rate": 6.7699999999999996e-06,
+      "loss": 2.7069,
+      "step": 86500
+    },
+    {
+      "epoch": 1.7309999999999999,
+      "grad_norm": 9.244890213012695,
+      "learning_rate": 6.745e-06,
+      "loss": 2.8612,
+      "step": 86550
+    },
+    {
+      "epoch": 1.732,
+      "grad_norm": 9.704329490661621,
+      "learning_rate": 6.72e-06,
+      "loss": 2.7387,
+      "step": 86600
+    },
+    {
+      "epoch": 1.733,
+      "grad_norm": 6.897952079772949,
+      "learning_rate": 6.695e-06,
+      "loss": 2.7645,
+      "step": 86650
+    },
+    {
+      "epoch": 1.734,
+      "grad_norm": 8.001317024230957,
+      "learning_rate": 6.67e-06,
+      "loss": 2.7385,
+      "step": 86700
+    },
+    {
+      "epoch": 1.7349999999999999,
+      "grad_norm": 7.381739616394043,
+      "learning_rate": 6.645e-06,
+      "loss": 2.862,
+      "step": 86750
+    },
+    {
+      "epoch": 1.736,
+      "grad_norm": 10.89158821105957,
+      "learning_rate": 6.62e-06,
+      "loss": 2.732,
+      "step": 86800
+    },
+    {
+      "epoch": 1.737,
+      "grad_norm": 7.905108451843262,
+      "learning_rate": 6.5949999999999995e-06,
+      "loss": 2.7843,
+      "step": 86850
+    },
+    {
+      "epoch": 1.738,
+      "grad_norm": 6.913460731506348,
+      "learning_rate": 6.57e-06,
+      "loss": 2.7391,
+      "step": 86900
+    },
+    {
+      "epoch": 1.7389999999999999,
+      "grad_norm": 7.739956378936768,
+      "learning_rate": 6.545e-06,
+      "loss": 2.6727,
+      "step": 86950
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.15610122680664,
+      "learning_rate": 6.519999999999999e-06,
+      "loss": 2.686,
+      "step": 87000
+    },
+    {
+      "epoch": 1.741,
+      "grad_norm": 7.165774822235107,
+      "learning_rate": 6.495e-06,
+      "loss": 2.7024,
+      "step": 87050
+    },
+    {
+      "epoch": 1.742,
+      "grad_norm": 7.682003021240234,
+      "learning_rate": 6.47e-06,
+      "loss": 2.6132,
+      "step": 87100
+    },
+    {
+      "epoch": 1.7429999999999999,
+      "grad_norm": 8.491655349731445,
+      "learning_rate": 6.444999999999999e-06,
+      "loss": 2.7714,
+      "step": 87150
+    },
+    {
+      "epoch": 1.744,
+      "grad_norm": 8.800849914550781,
+      "learning_rate": 6.4199999999999995e-06,
+      "loss": 2.7004,
+      "step": 87200
+    },
+    {
+      "epoch": 1.745,
+      "grad_norm": 8.72658920288086,
+      "learning_rate": 6.395000000000001e-06,
+      "loss": 2.8618,
+      "step": 87250
+    },
+    {
+      "epoch": 1.746,
+      "grad_norm": 10.49486255645752,
+      "learning_rate": 6.370000000000001e-06,
+      "loss": 2.9177,
+      "step": 87300
+    },
+    {
+      "epoch": 1.7469999999999999,
+      "grad_norm": 7.540380001068115,
+      "learning_rate": 6.345000000000001e-06,
+      "loss": 2.7355,
+      "step": 87350
+    },
+    {
+      "epoch": 1.748,
+      "grad_norm": Infinity,
+      "learning_rate": 6.320000000000001e-06,
+      "loss": 3.01,
+      "step": 87400
+    },
+    {
+      "epoch": 1.749,
+      "grad_norm": 8.305073738098145,
+      "learning_rate": 6.2955e-06,
+      "loss": 2.8643,
+      "step": 87450
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 7.897730827331543,
+      "learning_rate": 6.2705e-06,
+      "loss": 2.7766,
+      "step": 87500
+    },
+    {
+      "epoch": 1.751,
+      "grad_norm": 7.82082462310791,
+      "learning_rate": 6.2455e-06,
+      "loss": 2.8744,
+      "step": 87550
+    },
+    {
+      "epoch": 1.752,
+      "grad_norm": 8.460726737976074,
+      "learning_rate": 6.220500000000001e-06,
+      "loss": 2.8135,
+      "step": 87600
+    },
+    {
+      "epoch": 1.7530000000000001,
+      "grad_norm": 9.274124145507812,
+      "learning_rate": 6.1955e-06,
+      "loss": 2.6903,
+      "step": 87650
+    },
+    {
+      "epoch": 1.754,
+      "grad_norm": 8.046218872070312,
+      "learning_rate": 6.1705000000000005e-06,
+      "loss": 2.5912,
+      "step": 87700
+    },
+    {
+      "epoch": 1.755,
+      "grad_norm": 8.15789794921875,
+      "learning_rate": 6.145500000000001e-06,
+      "loss": 2.5456,
+      "step": 87750
+    },
+    {
+      "epoch": 1.756,
+      "grad_norm": 11.076677322387695,
+      "learning_rate": 6.120500000000001e-06,
+      "loss": 2.7817,
+      "step": 87800
+    },
+    {
+      "epoch": 1.7570000000000001,
+      "grad_norm": 7.190091609954834,
+      "learning_rate": 6.0955e-06,
+      "loss": 2.775,
+      "step": 87850
+    },
+    {
+      "epoch": 1.758,
+      "grad_norm": 8.741618156433105,
+      "learning_rate": 6.070500000000001e-06,
+      "loss": 2.8138,
+      "step": 87900
+    },
+    {
+      "epoch": 1.759,
+      "grad_norm": 7.122415065765381,
+      "learning_rate": 6.045500000000001e-06,
+      "loss": 2.7268,
+      "step": 87950
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 6.823777198791504,
+      "learning_rate": 6.0205e-06,
+      "loss": 2.6694,
+      "step": 88000
+    },
+    {
+      "epoch": 1.7610000000000001,
+      "grad_norm": 7.419735908508301,
+      "learning_rate": 5.9955000000000004e-06,
+      "loss": 2.8717,
+      "step": 88050
+    },
+    {
+      "epoch": 1.762,
+      "grad_norm": 6.787914276123047,
+      "learning_rate": 5.970500000000001e-06,
+      "loss": 2.7379,
+      "step": 88100
+    },
+    {
+      "epoch": 1.763,
+      "grad_norm": 7.9896979331970215,
+      "learning_rate": 5.9455e-06,
+      "loss": 2.8077,
+      "step": 88150
+    },
+    {
+      "epoch": 1.764,
+      "grad_norm": 8.55472469329834,
+      "learning_rate": 5.9205e-06,
+      "loss": 2.7669,
+      "step": 88200
+    },
+    {
+      "epoch": 1.7650000000000001,
+      "grad_norm": 7.805665969848633,
+      "learning_rate": 5.8955000000000006e-06,
+      "loss": 2.7029,
+      "step": 88250
+    },
+    {
+      "epoch": 1.766,
+      "grad_norm": 7.704306125640869,
+      "learning_rate": 5.8705e-06,
+      "loss": 2.7733,
+      "step": 88300
+    },
+    {
+      "epoch": 1.767,
+      "grad_norm": 6.891770362854004,
+      "learning_rate": 5.8455e-06,
+      "loss": 2.7838,
+      "step": 88350
+    },
+    {
+      "epoch": 1.768,
+      "grad_norm": 7.864100456237793,
+      "learning_rate": 5.8205000000000004e-06,
+      "loss": 2.8568,
+      "step": 88400
+    },
+    {
+      "epoch": 1.7690000000000001,
+      "grad_norm": 10.85545539855957,
+      "learning_rate": 5.7955e-06,
+      "loss": 2.7936,
+      "step": 88450
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.065977573394775,
+      "learning_rate": 5.7705e-06,
+      "loss": 3.0453,
+      "step": 88500
+    },
+    {
+      "epoch": 1.771,
+      "grad_norm": 8.579532623291016,
+      "learning_rate": 5.7455e-06,
+      "loss": 2.9797,
+      "step": 88550
+    },
+    {
+      "epoch": 1.772,
+      "grad_norm": 7.2842183113098145,
+      "learning_rate": 5.7205000000000005e-06,
+      "loss": 2.6284,
+      "step": 88600
+    },
+    {
+      "epoch": 1.7730000000000001,
+      "grad_norm": 9.140216827392578,
+      "learning_rate": 5.6955e-06,
+      "loss": 2.7266,
+      "step": 88650
+    },
+    {
+      "epoch": 1.774,
+      "grad_norm": 7.360970497131348,
+      "learning_rate": 5.6705e-06,
+      "loss": 2.798,
+      "step": 88700
+    },
+    {
+      "epoch": 1.775,
+      "grad_norm": 8.7395658493042,
+      "learning_rate": 5.6455e-06,
+      "loss": 2.8203,
+      "step": 88750
+    },
+    {
+      "epoch": 1.776,
+      "grad_norm": 8.213441848754883,
+      "learning_rate": 5.6205e-06,
+      "loss": 2.8288,
+      "step": 88800
+    },
+    {
+      "epoch": 1.7770000000000001,
+      "grad_norm": 7.257961750030518,
+      "learning_rate": 5.5955e-06,
+      "loss": 2.8347,
+      "step": 88850
+    },
+    {
+      "epoch": 1.778,
+      "grad_norm": 9.789780616760254,
+      "learning_rate": 5.5705e-06,
+      "loss": 2.8285,
+      "step": 88900
+    },
+    {
+      "epoch": 1.779,
+      "grad_norm": 7.806823253631592,
+      "learning_rate": 5.5455e-06,
+      "loss": 2.6556,
+      "step": 88950
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 7.615421772003174,
+      "learning_rate": 5.5205e-06,
+      "loss": 2.6485,
+      "step": 89000
+    },
+    {
+      "epoch": 1.7810000000000001,
+      "grad_norm": 8.010266304016113,
+      "learning_rate": 5.4955e-06,
+      "loss": 2.6579,
+      "step": 89050
+    },
+    {
+      "epoch": 1.782,
+      "grad_norm": 8.119599342346191,
+      "learning_rate": 5.4704999999999995e-06,
+      "loss": 2.6313,
+      "step": 89100
+    },
+    {
+      "epoch": 1.783,
+      "grad_norm": 11.232370376586914,
+      "learning_rate": 5.445500000000001e-06,
+      "loss": 2.6671,
+      "step": 89150
+    },
+    {
+      "epoch": 1.784,
+      "grad_norm": 6.684316635131836,
+      "learning_rate": 5.420500000000001e-06,
+      "loss": 2.707,
+      "step": 89200
+    },
+    {
+      "epoch": 1.7850000000000001,
+      "grad_norm": 8.435420036315918,
+      "learning_rate": 5.3955e-06,
+      "loss": 2.4425,
+      "step": 89250
+    },
+    {
+      "epoch": 1.786,
+      "grad_norm": 7.559298992156982,
+      "learning_rate": 5.3705000000000005e-06,
+      "loss": 2.7967,
+      "step": 89300
+    },
+    {
+      "epoch": 1.787,
+      "grad_norm": 7.989433765411377,
+      "learning_rate": 5.345500000000001e-06,
+      "loss": 2.7226,
+      "step": 89350
+    },
+    {
+      "epoch": 1.788,
+      "grad_norm": 8.866364479064941,
+      "learning_rate": 5.3205e-06,
+      "loss": 2.6818,
+      "step": 89400
+    },
+    {
+      "epoch": 1.7890000000000001,
+      "grad_norm": 8.641900062561035,
+      "learning_rate": 5.296000000000001e-06,
+      "loss": 2.8602,
+      "step": 89450
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.063851356506348,
+      "learning_rate": 5.271e-06,
+      "loss": 2.6172,
+      "step": 89500
+    },
+    {
+      "epoch": 1.791,
+      "grad_norm": 8.536163330078125,
+      "learning_rate": 5.246e-06,
+      "loss": 2.6273,
+      "step": 89550
+    },
+    {
+      "epoch": 1.792,
+      "grad_norm": 8.912287712097168,
+      "learning_rate": 5.2210000000000005e-06,
+      "loss": 2.7513,
+      "step": 89600
+    },
+    {
+      "epoch": 1.7930000000000001,
+      "grad_norm": 8.023194313049316,
+      "learning_rate": 5.196e-06,
+      "loss": 2.5922,
+      "step": 89650
+    },
+    {
+      "epoch": 1.794,
+      "grad_norm": 7.167906761169434,
+      "learning_rate": 5.171e-06,
+      "loss": 2.7784,
+      "step": 89700
+    },
+    {
+      "epoch": 1.795,
+      "grad_norm": 7.181880950927734,
+      "learning_rate": 5.1465e-06,
+      "loss": 2.9942,
+      "step": 89750
+    },
+    {
+      "epoch": 1.796,
+      "grad_norm": 9.94197940826416,
+      "learning_rate": 5.1215e-06,
+      "loss": 2.6009,
+      "step": 89800
+    },
+    {
+      "epoch": 1.7970000000000002,
+      "grad_norm": 8.615551948547363,
+      "learning_rate": 5.0965e-06,
+      "loss": 2.6638,
+      "step": 89850
+    },
+    {
+      "epoch": 1.798,
+      "grad_norm": 8.286895751953125,
+      "learning_rate": 5.0715000000000005e-06,
+      "loss": 2.6442,
+      "step": 89900
+    },
+    {
+      "epoch": 1.799,
+      "grad_norm": 7.661980628967285,
+      "learning_rate": 5.046500000000001e-06,
+      "loss": 2.718,
+      "step": 89950
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.584294319152832,
+      "learning_rate": 5.0215e-06,
+      "loss": 2.7903,
+      "step": 90000
+    },
+    {
+      "epoch": 1.8010000000000002,
+      "grad_norm": 6.850968360900879,
+      "learning_rate": 4.9965e-06,
+      "loss": 2.7178,
+      "step": 90050
+    },
+    {
+      "epoch": 1.802,
+      "grad_norm": 8.273259162902832,
+      "learning_rate": 4.971500000000001e-06,
+      "loss": 2.8497,
+      "step": 90100
+    },
+    {
+      "epoch": 1.803,
+      "grad_norm": 7.811845779418945,
+      "learning_rate": 4.946500000000001e-06,
+      "loss": 2.652,
+      "step": 90150
+    },
+    {
+      "epoch": 1.804,
+      "grad_norm": 6.695546627044678,
+      "learning_rate": 4.9215e-06,
+      "loss": 2.6411,
+      "step": 90200
+    },
+    {
+      "epoch": 1.8050000000000002,
+      "grad_norm": 11.059279441833496,
+      "learning_rate": 4.8965000000000005e-06,
+      "loss": 2.5529,
+      "step": 90250
+    },
+    {
+      "epoch": 1.806,
+      "grad_norm": 9.72866153717041,
+      "learning_rate": 4.871500000000001e-06,
+      "loss": 2.9216,
+      "step": 90300
+    },
+    {
+      "epoch": 1.807,
+      "grad_norm": 10.693723678588867,
+      "learning_rate": 4.8465e-06,
+      "loss": 2.8347,
+      "step": 90350
+    },
+    {
+      "epoch": 1.808,
+      "grad_norm": 10.578742027282715,
+      "learning_rate": 4.8215e-06,
+      "loss": 2.8648,
+      "step": 90400
+    },
+    {
+      "epoch": 1.8090000000000002,
+      "grad_norm": 9.165677070617676,
+      "learning_rate": 4.796500000000001e-06,
+      "loss": 2.5032,
+      "step": 90450
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.2900471687316895,
+      "learning_rate": 4.7715e-06,
+      "loss": 2.7891,
+      "step": 90500
+    },
+    {
+      "epoch": 1.811,
+      "grad_norm": 7.986770153045654,
+      "learning_rate": 4.7465e-06,
+      "loss": 2.8503,
+      "step": 90550
+    },
+    {
+      "epoch": 1.812,
+      "grad_norm": 8.599214553833008,
+      "learning_rate": 4.7215000000000004e-06,
+      "loss": 2.7908,
+      "step": 90600
+    },
+    {
+      "epoch": 1.813,
+      "grad_norm": 8.372736930847168,
+      "learning_rate": 4.6965e-06,
+      "loss": 2.7009,
+      "step": 90650
+    },
+    {
+      "epoch": 1.814,
+      "grad_norm": 8.184940338134766,
+      "learning_rate": 4.6715e-06,
+      "loss": 2.9262,
+      "step": 90700
+    },
+    {
+      "epoch": 1.815,
+      "grad_norm": 8.346232414245605,
+      "learning_rate": 4.6465e-06,
+      "loss": 2.7132,
+      "step": 90750
+    },
+    {
+      "epoch": 1.8159999999999998,
+      "grad_norm": 9.309367179870605,
+      "learning_rate": 4.6215e-06,
+      "loss": 2.7366,
+      "step": 90800
+    },
+    {
+      "epoch": 1.817,
+      "grad_norm": 8.382227897644043,
+      "learning_rate": 4.5965e-06,
+      "loss": 2.6953,
+      "step": 90850
+    },
+    {
+      "epoch": 1.818,
+      "grad_norm": 9.15678882598877,
+      "learning_rate": 4.5715e-06,
+      "loss": 2.5969,
+      "step": 90900
+    },
+    {
+      "epoch": 1.819,
+      "grad_norm": 9.202027320861816,
+      "learning_rate": 4.5465000000000004e-06,
+      "loss": 2.7032,
+      "step": 90950
+    },
+    {
+      "epoch": 1.8199999999999998,
+      "grad_norm": 9.636245727539062,
+      "learning_rate": 4.5215e-06,
+      "loss": 2.6589,
+      "step": 91000
+    },
+    {
+      "epoch": 1.821,
+      "grad_norm": 8.213318824768066,
+      "learning_rate": 4.4965e-06,
+      "loss": 2.667,
+      "step": 91050
+    },
+    {
+      "epoch": 1.822,
+      "grad_norm": 7.809316635131836,
+      "learning_rate": 4.4715e-06,
+      "loss": 2.8199,
+      "step": 91100
+    },
+    {
+      "epoch": 1.823,
+      "grad_norm": 9.173991203308105,
+      "learning_rate": 4.4465e-06,
+      "loss": 2.7688,
+      "step": 91150
+    },
+    {
+      "epoch": 1.8239999999999998,
+      "grad_norm": 8.543388366699219,
+      "learning_rate": 4.4215e-06,
+      "loss": 2.6941,
+      "step": 91200
+    },
+    {
+      "epoch": 1.825,
+      "grad_norm": 8.782526969909668,
+      "learning_rate": 4.3965e-06,
+      "loss": 2.5714,
+      "step": 91250
+    },
+    {
+      "epoch": 1.826,
+      "grad_norm": 8.245438575744629,
+      "learning_rate": 4.3714999999999996e-06,
+      "loss": 2.9752,
+      "step": 91300
+    },
+    {
+      "epoch": 1.827,
+      "grad_norm": 8.939053535461426,
+      "learning_rate": 4.3465e-06,
+      "loss": 2.7425,
+      "step": 91350
+    },
+    {
+      "epoch": 1.8279999999999998,
+      "grad_norm": 9.105463981628418,
+      "learning_rate": 4.321500000000001e-06,
+      "loss": 2.951,
+      "step": 91400
+    },
+    {
+      "epoch": 1.829,
+      "grad_norm": 7.773897171020508,
+      "learning_rate": 4.2965e-06,
+      "loss": 2.8498,
+      "step": 91450
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.115293502807617,
+      "learning_rate": 4.2715000000000005e-06,
+      "loss": 2.984,
+      "step": 91500
+    },
+    {
+      "epoch": 1.831,
+      "grad_norm": 10.776023864746094,
+      "learning_rate": 4.246500000000001e-06,
+      "loss": 2.6162,
+      "step": 91550
+    },
+    {
+      "epoch": 1.8319999999999999,
+      "grad_norm": 10.591742515563965,
+      "learning_rate": 4.2215e-06,
+      "loss": 2.8948,
+      "step": 91600
+    },
+    {
+      "epoch": 1.833,
+      "grad_norm": 7.4504241943359375,
+      "learning_rate": 4.1965e-06,
+      "loss": 2.7548,
+      "step": 91650
+    },
+    {
+      "epoch": 1.834,
+      "grad_norm": 7.637448310852051,
+      "learning_rate": 4.171500000000001e-06,
+      "loss": 2.874,
+      "step": 91700
+    },
+    {
+      "epoch": 1.835,
+      "grad_norm": 9.06761360168457,
+      "learning_rate": 4.1465e-06,
+      "loss": 2.5051,
+      "step": 91750
+    },
+    {
+      "epoch": 1.8359999999999999,
+      "grad_norm": 6.866405487060547,
+      "learning_rate": 4.1215e-06,
+      "loss": 2.5928,
+      "step": 91800
+    },
+    {
+      "epoch": 1.837,
+      "grad_norm": 12.77744197845459,
+      "learning_rate": 4.0965000000000005e-06,
+      "loss": 2.7717,
+      "step": 91850
+    },
+    {
+      "epoch": 1.838,
+      "grad_norm": 9.051816940307617,
+      "learning_rate": 4.0715e-06,
+      "loss": 2.6854,
+      "step": 91900
+    },
+    {
+      "epoch": 1.839,
+      "grad_norm": 8.761739730834961,
+      "learning_rate": 4.0465e-06,
+      "loss": 2.6948,
+      "step": 91950
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 8.750024795532227,
+      "learning_rate": 4.0215e-06,
+      "loss": 2.7667,
+      "step": 92000
+    },
+    {
+      "epoch": 1.841,
+      "grad_norm": 8.970072746276855,
+      "learning_rate": 3.996500000000001e-06,
+      "loss": 2.6667,
+      "step": 92050
+    },
+    {
+      "epoch": 1.842,
+      "grad_norm": 8.1638822555542,
+      "learning_rate": 3.9715e-06,
+      "loss": 2.8341,
+      "step": 92100
+    },
+    {
+      "epoch": 1.843,
+      "grad_norm": 7.813655376434326,
+      "learning_rate": 3.9465e-06,
+      "loss": 2.6207,
+      "step": 92150
+    },
+    {
+      "epoch": 1.8439999999999999,
+      "grad_norm": 10.0947265625,
+      "learning_rate": 3.9215000000000005e-06,
+      "loss": 2.8325,
+      "step": 92200
+    },
+    {
+      "epoch": 1.845,
+      "grad_norm": 7.8281569480896,
+      "learning_rate": 3.8965e-06,
+      "loss": 2.7996,
+      "step": 92250
+    },
+    {
+      "epoch": 1.846,
+      "grad_norm": 8.70777702331543,
+      "learning_rate": 3.8715e-06,
+      "loss": 2.7297,
+      "step": 92300
+    },
+    {
+      "epoch": 1.847,
+      "grad_norm": 7.831836223602295,
+      "learning_rate": 3.8465e-06,
+      "loss": 2.7486,
+      "step": 92350
+    },
+    {
+      "epoch": 1.8479999999999999,
+      "grad_norm": 7.55377197265625,
+      "learning_rate": 3.8215e-06,
+      "loss": 2.802,
+      "step": 92400
+    },
+    {
+      "epoch": 1.849,
+      "grad_norm": 7.586316108703613,
+      "learning_rate": 3.7965e-06,
+      "loss": 2.7082,
+      "step": 92450
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 7.920407772064209,
+      "learning_rate": 3.7715000000000002e-06,
+      "loss": 2.687,
+      "step": 92500
+    },
+    {
+      "epoch": 1.851,
+      "grad_norm": 7.852161884307861,
+      "learning_rate": 3.7465e-06,
+      "loss": 2.7005,
+      "step": 92550
+    },
+    {
+      "epoch": 1.8519999999999999,
+      "grad_norm": 8.755575180053711,
+      "learning_rate": 3.7215e-06,
+      "loss": 2.6285,
+      "step": 92600
+    },
+    {
+      "epoch": 1.853,
+      "grad_norm": 10.1644868850708,
+      "learning_rate": 3.6965e-06,
+      "loss": 2.7896,
+      "step": 92650
+    },
+    {
+      "epoch": 1.854,
+      "grad_norm": 8.134500503540039,
+      "learning_rate": 3.6715e-06,
+      "loss": 2.8607,
+      "step": 92700
+    },
+    {
+      "epoch": 1.855,
+      "grad_norm": 8.263608932495117,
+      "learning_rate": 3.6464999999999997e-06,
+      "loss": 2.5583,
+      "step": 92750
+    },
+    {
+      "epoch": 1.8559999999999999,
+      "grad_norm": 8.864056587219238,
+      "learning_rate": 3.6215e-06,
+      "loss": 2.7813,
+      "step": 92800
+    },
+    {
+      "epoch": 1.857,
+      "grad_norm": 9.94490909576416,
+      "learning_rate": 3.5964999999999998e-06,
+      "loss": 2.6101,
+      "step": 92850
+    },
+    {
+      "epoch": 1.858,
+      "grad_norm": 8.170188903808594,
+      "learning_rate": 3.5715e-06,
+      "loss": 2.9997,
+      "step": 92900
+    },
+    {
+      "epoch": 1.859,
+      "grad_norm": 10.034306526184082,
+      "learning_rate": 3.5465000000000007e-06,
+      "loss": 2.7796,
+      "step": 92950
+    },
+    {
+      "epoch": 1.8599999999999999,
+      "grad_norm": 8.248494148254395,
+      "learning_rate": 3.5215000000000005e-06,
+      "loss": 2.8471,
+      "step": 93000
+    },
+    {
+      "epoch": 1.861,
+      "grad_norm": 8.490750312805176,
+      "learning_rate": 3.4965000000000003e-06,
+      "loss": 2.549,
+      "step": 93050
+    },
+    {
+      "epoch": 1.862,
+      "grad_norm": 7.735034465789795,
+      "learning_rate": 3.4715000000000006e-06,
+      "loss": 2.8037,
+      "step": 93100
+    },
+    {
+      "epoch": 1.863,
+      "grad_norm": 9.583478927612305,
+      "learning_rate": 3.4465000000000004e-06,
+      "loss": 2.7986,
+      "step": 93150
+    },
+    {
+      "epoch": 1.8639999999999999,
+      "grad_norm": 7.504220008850098,
+      "learning_rate": 3.4215e-06,
+      "loss": 2.8001,
+      "step": 93200
+    },
+    {
+      "epoch": 1.865,
+      "grad_norm": 8.959890365600586,
+      "learning_rate": 3.3965000000000004e-06,
+      "loss": 2.5646,
+      "step": 93250
+    },
+    {
+      "epoch": 1.866,
+      "grad_norm": 8.324706077575684,
+      "learning_rate": 3.3715000000000002e-06,
+      "loss": 2.7756,
+      "step": 93300
+    },
+    {
+      "epoch": 1.867,
+      "grad_norm": 9.705889701843262,
+      "learning_rate": 3.3465000000000005e-06,
+      "loss": 2.6537,
+      "step": 93350
+    },
+    {
+      "epoch": 1.8679999999999999,
+      "grad_norm": 12.64687728881836,
+      "learning_rate": 3.3215000000000003e-06,
+      "loss": 2.76,
+      "step": 93400
+    },
+    {
+      "epoch": 1.869,
+      "grad_norm": 8.79395580291748,
+      "learning_rate": 3.2965e-06,
+      "loss": 2.7001,
+      "step": 93450
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.220784187316895,
+      "learning_rate": 3.2715000000000004e-06,
+      "loss": 2.8881,
+      "step": 93500
+    },
+    {
+      "epoch": 1.871,
+      "grad_norm": 9.860733985900879,
+      "learning_rate": 3.2465e-06,
+      "loss": 2.7711,
+      "step": 93550
+    },
+    {
+      "epoch": 1.8719999999999999,
+      "grad_norm": 7.789809703826904,
+      "learning_rate": 3.2215000000000004e-06,
+      "loss": 2.6483,
+      "step": 93600
+    },
+    {
+      "epoch": 1.873,
+      "grad_norm": 7.488524436950684,
+      "learning_rate": 3.1965000000000002e-06,
+      "loss": 2.6206,
+      "step": 93650
+    },
+    {
+      "epoch": 1.874,
+      "grad_norm": 10.260944366455078,
+      "learning_rate": 3.1715e-06,
+      "loss": 2.6502,
+      "step": 93700
+    },
+    {
+      "epoch": 1.875,
+      "grad_norm": 8.944072723388672,
+      "learning_rate": 3.147e-06,
+      "loss": 3.0729,
+      "step": 93750
+    },
+    {
+      "epoch": 1.876,
+      "grad_norm": 8.896251678466797,
+      "learning_rate": 3.1220000000000003e-06,
+      "loss": 2.5911,
+      "step": 93800
+    },
+    {
+      "epoch": 1.877,
+      "grad_norm": 8.604350090026855,
+      "learning_rate": 3.097e-06,
+      "loss": 2.6665,
+      "step": 93850
+    },
+    {
+      "epoch": 1.8780000000000001,
+      "grad_norm": 8.519975662231445,
+      "learning_rate": 3.0720000000000004e-06,
+      "loss": 2.81,
+      "step": 93900
+    },
+    {
+      "epoch": 1.879,
+      "grad_norm": 7.503047943115234,
+      "learning_rate": 3.0470000000000002e-06,
+      "loss": 2.7349,
+      "step": 93950
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.925019264221191,
+      "learning_rate": 3.022e-06,
+      "loss": 2.6914,
+      "step": 94000
+    },
+    {
+      "epoch": 1.881,
+      "grad_norm": 7.482077121734619,
+      "learning_rate": 2.9970000000000003e-06,
+      "loss": 2.7253,
+      "step": 94050
+    },
+    {
+      "epoch": 1.8820000000000001,
+      "grad_norm": 8.396530151367188,
+      "learning_rate": 2.972e-06,
+      "loss": 3.0853,
+      "step": 94100
+    },
+    {
+      "epoch": 1.883,
+      "grad_norm": 10.036365509033203,
+      "learning_rate": 2.947e-06,
+      "loss": 2.811,
+      "step": 94150
+    },
+    {
+      "epoch": 1.884,
+      "grad_norm": 9.501229286193848,
+      "learning_rate": 2.922e-06,
+      "loss": 2.7846,
+      "step": 94200
+    },
+    {
+      "epoch": 1.885,
+      "grad_norm": 7.217312335968018,
+      "learning_rate": 2.897e-06,
+      "loss": 2.7255,
+      "step": 94250
+    },
+    {
+      "epoch": 1.8860000000000001,
+      "grad_norm": 9.453542709350586,
+      "learning_rate": 2.872e-06,
+      "loss": 2.5492,
+      "step": 94300
+    },
+    {
+      "epoch": 1.887,
+      "grad_norm": 17.13677406311035,
+      "learning_rate": 2.847e-06,
+      "loss": 2.8925,
+      "step": 94350
+    },
+    {
+      "epoch": 1.888,
+      "grad_norm": 7.6179962158203125,
+      "learning_rate": 2.822e-06,
+      "loss": 2.5682,
+      "step": 94400
+    },
+    {
+      "epoch": 1.889,
+      "grad_norm": 7.958014488220215,
+      "learning_rate": 2.797e-06,
+      "loss": 2.6691,
+      "step": 94450
+    },
+    {
+      "epoch": 1.8900000000000001,
+      "grad_norm": 9.357193946838379,
+      "learning_rate": 2.7720000000000003e-06,
+      "loss": 2.4874,
+      "step": 94500
+    },
+    {
+      "epoch": 1.891,
+      "grad_norm": 11.814285278320312,
+      "learning_rate": 2.747e-06,
+      "loss": 2.8565,
+      "step": 94550
+    },
+    {
+      "epoch": 1.892,
+      "grad_norm": 8.839566230773926,
+      "learning_rate": 2.7220000000000004e-06,
+      "loss": 2.8531,
+      "step": 94600
+    },
+    {
+      "epoch": 1.893,
+      "grad_norm": 8.386056900024414,
+      "learning_rate": 2.697e-06,
+      "loss": 2.5458,
+      "step": 94650
+    },
+    {
+      "epoch": 1.8940000000000001,
+      "grad_norm": 7.575038433074951,
+      "learning_rate": 2.672e-06,
+      "loss": 2.6183,
+      "step": 94700
+    },
+    {
+      "epoch": 1.895,
+      "grad_norm": 7.668638229370117,
+      "learning_rate": 2.6470000000000002e-06,
+      "loss": 2.6924,
+      "step": 94750
+    },
+    {
+      "epoch": 1.896,
+      "grad_norm": 8.719830513000488,
+      "learning_rate": 2.622e-06,
+      "loss": 2.6432,
+      "step": 94800
+    },
+    {
+      "epoch": 1.897,
+      "grad_norm": 10.851173400878906,
+      "learning_rate": 2.5970000000000003e-06,
+      "loss": 2.9338,
+      "step": 94850
+    },
+    {
+      "epoch": 1.8980000000000001,
+      "grad_norm": 7.968566417694092,
+      "learning_rate": 2.572e-06,
+      "loss": 2.8757,
+      "step": 94900
+    },
+    {
+      "epoch": 1.899,
+      "grad_norm": 7.560086727142334,
+      "learning_rate": 2.547e-06,
+      "loss": 2.7454,
+      "step": 94950
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.198180198669434,
+      "learning_rate": 2.522e-06,
+      "loss": 2.766,
+      "step": 95000
+    },
+    {
+      "epoch": 1.901,
+      "grad_norm": 7.824384689331055,
+      "learning_rate": 2.497e-06,
+      "loss": 2.7285,
+      "step": 95050
+    },
+    {
+      "epoch": 1.9020000000000001,
+      "grad_norm": 7.9695281982421875,
+      "learning_rate": 2.4720000000000002e-06,
+      "loss": 2.6896,
+      "step": 95100
+    },
+    {
+      "epoch": 1.903,
+      "grad_norm": 7.852643013000488,
+      "learning_rate": 2.447e-06,
+      "loss": 2.6197,
+      "step": 95150
+    },
+    {
+      "epoch": 1.904,
+      "grad_norm": 8.303346633911133,
+      "learning_rate": 2.422e-06,
+      "loss": 2.6494,
+      "step": 95200
+    },
+    {
+      "epoch": 1.905,
+      "grad_norm": 11.907207489013672,
+      "learning_rate": 2.397e-06,
+      "loss": 2.7931,
+      "step": 95250
+    },
+    {
+      "epoch": 1.9060000000000001,
+      "grad_norm": 8.8694486618042,
+      "learning_rate": 2.3720000000000003e-06,
+      "loss": 2.644,
+      "step": 95300
+    },
+    {
+      "epoch": 1.907,
+      "grad_norm": 7.289035797119141,
+      "learning_rate": 2.347e-06,
+      "loss": 2.7105,
+      "step": 95350
+    },
+    {
+      "epoch": 1.908,
+      "grad_norm": 8.548935890197754,
+      "learning_rate": 2.3220000000000004e-06,
+      "loss": 2.8733,
+      "step": 95400
+    },
+    {
+      "epoch": 1.909,
+      "grad_norm": 8.640983581542969,
+      "learning_rate": 2.297e-06,
+      "loss": 2.7284,
+      "step": 95450
+    },
+    {
+      "epoch": 1.9100000000000001,
+      "grad_norm": 8.68470573425293,
+      "learning_rate": 2.2725e-06,
+      "loss": 2.7682,
+      "step": 95500
+    },
+    {
+      "epoch": 1.911,
+      "grad_norm": 8.883843421936035,
+      "learning_rate": 2.2475e-06,
+      "loss": 2.7007,
+      "step": 95550
+    },
+    {
+      "epoch": 1.912,
+      "grad_norm": 8.315601348876953,
+      "learning_rate": 2.2225000000000003e-06,
+      "loss": 2.8442,
+      "step": 95600
+    },
+    {
+      "epoch": 1.913,
+      "grad_norm": 7.067404270172119,
+      "learning_rate": 2.1975e-06,
+      "loss": 2.6383,
+      "step": 95650
+    },
+    {
+      "epoch": 1.9140000000000001,
+      "grad_norm": 9.462407112121582,
+      "learning_rate": 2.1725000000000004e-06,
+      "loss": 2.7837,
+      "step": 95700
+    },
+    {
+      "epoch": 1.915,
+      "grad_norm": 7.8980021476745605,
+      "learning_rate": 2.1475e-06,
+      "loss": 2.8218,
+      "step": 95750
+    },
+    {
+      "epoch": 1.916,
+      "grad_norm": 8.796659469604492,
+      "learning_rate": 2.1225e-06,
+      "loss": 2.5969,
+      "step": 95800
+    },
+    {
+      "epoch": 1.917,
+      "grad_norm": 7.042479038238525,
+      "learning_rate": 2.0975000000000002e-06,
+      "loss": 2.7725,
+      "step": 95850
+    },
+    {
+      "epoch": 1.9180000000000001,
+      "grad_norm": 7.105340480804443,
+      "learning_rate": 2.0725e-06,
+      "loss": 2.7451,
+      "step": 95900
+    },
+    {
+      "epoch": 1.919,
+      "grad_norm": 8.806255340576172,
+      "learning_rate": 2.0475000000000003e-06,
+      "loss": 2.7912,
+      "step": 95950
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.678021430969238,
+      "learning_rate": 2.0225e-06,
+      "loss": 2.9036,
+      "step": 96000
+    },
+    {
+      "epoch": 1.921,
+      "grad_norm": 12.134692192077637,
+      "learning_rate": 1.9975e-06,
+      "loss": 2.7814,
+      "step": 96050
+    },
+    {
+      "epoch": 1.9220000000000002,
+      "grad_norm": 9.222649574279785,
+      "learning_rate": 1.9725e-06,
+      "loss": 2.619,
+      "step": 96100
+    },
+    {
+      "epoch": 1.923,
+      "grad_norm": 8.665557861328125,
+      "learning_rate": 1.9475e-06,
+      "loss": 2.7856,
+      "step": 96150
+    },
+    {
+      "epoch": 1.924,
+      "grad_norm": 10.79527759552002,
+      "learning_rate": 1.9225e-06,
+      "loss": 2.7519,
+      "step": 96200
+    },
+    {
+      "epoch": 1.925,
+      "grad_norm": 10.203853607177734,
+      "learning_rate": 1.8975e-06,
+      "loss": 2.605,
+      "step": 96250
+    },
+    {
+      "epoch": 1.9260000000000002,
+      "grad_norm": 8.2529296875,
+      "learning_rate": 1.8724999999999999e-06,
+      "loss": 2.7996,
+      "step": 96300
+    },
+    {
+      "epoch": 1.927,
+      "grad_norm": 7.979179382324219,
+      "learning_rate": 1.8474999999999999e-06,
+      "loss": 2.8076,
+      "step": 96350
+    },
+    {
+      "epoch": 1.928,
+      "grad_norm": 9.023404121398926,
+      "learning_rate": 1.8225000000000003e-06,
+      "loss": 2.5915,
+      "step": 96400
+    },
+    {
+      "epoch": 1.929,
+      "grad_norm": 9.803990364074707,
+      "learning_rate": 1.7975000000000002e-06,
+      "loss": 2.8453,
+      "step": 96450
+    },
+    {
+      "epoch": 1.9300000000000002,
+      "grad_norm": 9.243449211120605,
+      "learning_rate": 1.7725000000000002e-06,
+      "loss": 2.7708,
+      "step": 96500
+    },
+    {
+      "epoch": 1.931,
+      "grad_norm": 8.036003112792969,
+      "learning_rate": 1.7475000000000002e-06,
+      "loss": 2.814,
+      "step": 96550
+    },
+    {
+      "epoch": 1.932,
+      "grad_norm": 7.145360469818115,
+      "learning_rate": 1.7225000000000002e-06,
+      "loss": 2.8895,
+      "step": 96600
+    },
+    {
+      "epoch": 1.933,
+      "grad_norm": 7.846442699432373,
+      "learning_rate": 1.6975e-06,
+      "loss": 2.8026,
+      "step": 96650
+    },
+    {
+      "epoch": 1.9340000000000002,
+      "grad_norm": 10.024850845336914,
+      "learning_rate": 1.6725e-06,
+      "loss": 2.8142,
+      "step": 96700
+    },
+    {
+      "epoch": 1.935,
+      "grad_norm": 8.556451797485352,
+      "learning_rate": 1.6475000000000001e-06,
+      "loss": 2.9857,
+      "step": 96750
+    },
+    {
+      "epoch": 1.936,
+      "grad_norm": 8.042387008666992,
+      "learning_rate": 1.6225000000000001e-06,
+      "loss": 2.6511,
+      "step": 96800
+    },
+    {
+      "epoch": 1.937,
+      "grad_norm": 9.563372611999512,
+      "learning_rate": 1.5975e-06,
+      "loss": 2.7634,
+      "step": 96850
+    },
+    {
+      "epoch": 1.938,
+      "grad_norm": 8.98689079284668,
+      "learning_rate": 1.5725e-06,
+      "loss": 2.6179,
+      "step": 96900
+    },
+    {
+      "epoch": 1.939,
+      "grad_norm": 9.716072082519531,
+      "learning_rate": 1.5475e-06,
+      "loss": 2.7824,
+      "step": 96950
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.737848281860352,
+      "learning_rate": 1.5225000000000002e-06,
+      "loss": 2.6499,
+      "step": 97000
+    },
+    {
+      "epoch": 1.9409999999999998,
+      "grad_norm": 11.896839141845703,
+      "learning_rate": 1.4975e-06,
+      "loss": 2.5394,
+      "step": 97050
+    },
+    {
+      "epoch": 1.942,
+      "grad_norm": 6.5530853271484375,
+      "learning_rate": 1.4725e-06,
+      "loss": 2.6731,
+      "step": 97100
+    },
+    {
+      "epoch": 1.943,
+      "grad_norm": 6.974987030029297,
+      "learning_rate": 1.4475000000000001e-06,
+      "loss": 2.8178,
+      "step": 97150
+    },
+    {
+      "epoch": 1.944,
+      "grad_norm": 6.8209614753723145,
+      "learning_rate": 1.4225000000000001e-06,
+      "loss": 2.6135,
+      "step": 97200
+    },
+    {
+      "epoch": 1.9449999999999998,
+      "grad_norm": 9.504709243774414,
+      "learning_rate": 1.3975e-06,
+      "loss": 2.7069,
+      "step": 97250
+    },
+    {
+      "epoch": 1.946,
+      "grad_norm": 10.736202239990234,
+      "learning_rate": 1.3725e-06,
+      "loss": 2.7967,
+      "step": 97300
+    },
+    {
+      "epoch": 1.947,
+      "grad_norm": 8.686497688293457,
+      "learning_rate": 1.3475000000000002e-06,
+      "loss": 2.6214,
+      "step": 97350
+    },
+    {
+      "epoch": 1.948,
+      "grad_norm": 10.431570053100586,
+      "learning_rate": 1.3225e-06,
+      "loss": 2.6681,
+      "step": 97400
+    },
+    {
+      "epoch": 1.9489999999999998,
+      "grad_norm": 11.108994483947754,
+      "learning_rate": 1.2975e-06,
+      "loss": 2.8571,
+      "step": 97450
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.426375389099121,
+      "learning_rate": 1.2725e-06,
+      "loss": 2.6336,
+      "step": 97500
+    },
+    {
+      "epoch": 1.951,
+      "grad_norm": 9.929731369018555,
+      "learning_rate": 1.2475000000000001e-06,
+      "loss": 2.8664,
+      "step": 97550
+    },
+    {
+      "epoch": 1.952,
+      "grad_norm": 5.834004878997803,
+      "learning_rate": 1.2225e-06,
+      "loss": 2.4785,
+      "step": 97600
+    },
+    {
+      "epoch": 1.9529999999999998,
+      "grad_norm": 10.382954597473145,
+      "learning_rate": 1.1975e-06,
+      "loss": 2.884,
+      "step": 97650
+    },
+    {
+      "epoch": 1.954,
+      "grad_norm": 9.039685249328613,
+      "learning_rate": 1.1725e-06,
+      "loss": 2.7769,
+      "step": 97700
+    },
+    {
+      "epoch": 1.955,
+      "grad_norm": 10.186605453491211,
+      "learning_rate": 1.1475000000000002e-06,
+      "loss": 2.8527,
+      "step": 97750
+    },
+    {
+      "epoch": 1.956,
+      "grad_norm": 7.526618480682373,
+      "learning_rate": 1.1225e-06,
+      "loss": 2.4586,
+      "step": 97800
+    },
+    {
+      "epoch": 1.9569999999999999,
+      "grad_norm": 7.888932704925537,
+      "learning_rate": 1.0975e-06,
+      "loss": 2.7046,
+      "step": 97850
+    },
+    {
+      "epoch": 1.958,
+      "grad_norm": 8.395057678222656,
+      "learning_rate": 1.0725000000000001e-06,
+      "loss": 2.8441,
+      "step": 97900
+    },
+    {
+      "epoch": 1.959,
+      "grad_norm": 7.477360725402832,
+      "learning_rate": 1.0475000000000001e-06,
+      "loss": 2.6665,
+      "step": 97950
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.053496360778809,
+      "learning_rate": 1.0225e-06,
+      "loss": 2.7325,
+      "step": 98000
+    },
+    {
+      "epoch": 1.9609999999999999,
+      "grad_norm": 8.144623756408691,
+      "learning_rate": 9.975e-07,
+      "loss": 2.9861,
+      "step": 98050
+    },
+    {
+      "epoch": 1.962,
+      "grad_norm": 8.730600357055664,
+      "learning_rate": 9.725e-07,
+      "loss": 2.7406,
+      "step": 98100
+    },
+    {
+      "epoch": 1.963,
+      "grad_norm": 8.985459327697754,
+      "learning_rate": 9.475000000000001e-07,
+      "loss": 2.9892,
+      "step": 98150
+    },
+    {
+      "epoch": 1.964,
+      "grad_norm": 11.969583511352539,
+      "learning_rate": 9.225000000000001e-07,
+      "loss": 2.7578,
+      "step": 98200
+    },
+    {
+      "epoch": 1.9649999999999999,
+      "grad_norm": 7.816971778869629,
+      "learning_rate": 8.975000000000001e-07,
+      "loss": 2.6194,
+      "step": 98250
+    },
+    {
+      "epoch": 1.966,
+      "grad_norm": 8.625751495361328,
+      "learning_rate": 8.725e-07,
+      "loss": 2.8595,
+      "step": 98300
+    },
+    {
+      "epoch": 1.967,
+      "grad_norm": 8.965986251831055,
+      "learning_rate": 8.475e-07,
+      "loss": 2.5684,
+      "step": 98350
+    },
+    {
+      "epoch": 1.968,
+      "grad_norm": 8.818077087402344,
+      "learning_rate": 8.225e-07,
+      "loss": 2.789,
+      "step": 98400
+    },
+    {
+      "epoch": 1.9689999999999999,
+      "grad_norm": 8.827529907226562,
+      "learning_rate": 7.975e-07,
+      "loss": 2.7706,
+      "step": 98450
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.630989074707031,
+      "learning_rate": 7.725e-07,
+      "loss": 2.6708,
+      "step": 98500
+    },
+    {
+      "epoch": 1.971,
+      "grad_norm": 7.110955715179443,
+      "learning_rate": 7.475e-07,
+      "loss": 2.628,
+      "step": 98550
+    },
+    {
+      "epoch": 1.972,
+      "grad_norm": 7.431270122528076,
+      "learning_rate": 7.225e-07,
+      "loss": 2.7951,
+      "step": 98600
+    },
+    {
+      "epoch": 1.9729999999999999,
+      "grad_norm": 7.719696521759033,
+      "learning_rate": 6.975000000000001e-07,
+      "loss": 2.9401,
+      "step": 98650
+    },
+    {
+      "epoch": 1.974,
+      "grad_norm": 8.271461486816406,
+      "learning_rate": 6.725e-07,
+      "loss": 2.7169,
+      "step": 98700
+    },
+    {
+      "epoch": 1.975,
+      "grad_norm": 9.32712173461914,
+      "learning_rate": 6.475e-07,
+      "loss": 2.7377,
+      "step": 98750
+    },
+    {
+      "epoch": 1.976,
+      "grad_norm": 7.909132957458496,
+      "learning_rate": 6.225e-07,
+      "loss": 2.7912,
+      "step": 98800
+    },
+    {
+      "epoch": 1.9769999999999999,
+      "grad_norm": 12.058829307556152,
+      "learning_rate": 5.975000000000001e-07,
+      "loss": 2.809,
+      "step": 98850
+    },
+    {
+      "epoch": 1.978,
+      "grad_norm": 9.400949478149414,
+      "learning_rate": 5.725e-07,
+      "loss": 2.6873,
+      "step": 98900
+    },
+    {
+      "epoch": 1.979,
+      "grad_norm": 8.208101272583008,
+      "learning_rate": 5.48e-07,
+      "loss": 2.6407,
+      "step": 98950
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.864774703979492,
+      "learning_rate": 5.23e-07,
+      "loss": 2.8138,
+      "step": 99000
+    },
+    {
+      "epoch": 1.9809999999999999,
+      "grad_norm": 7.672954559326172,
+      "learning_rate": 4.98e-07,
+      "loss": 3.0117,
+      "step": 99050
+    },
+    {
+      "epoch": 1.982,
+      "grad_norm": 8.58277702331543,
+      "learning_rate": 4.73e-07,
+      "loss": 2.7453,
+      "step": 99100
+    },
+    {
+      "epoch": 1.983,
+      "grad_norm": 8.067780494689941,
+      "learning_rate": 4.48e-07,
+      "loss": 2.7547,
+      "step": 99150
+    },
+    {
+      "epoch": 1.984,
+      "grad_norm": 8.917376518249512,
+      "learning_rate": 4.2300000000000007e-07,
+      "loss": 2.7548,
+      "step": 99200
+    },
+    {
+      "epoch": 1.9849999999999999,
+      "grad_norm": 7.516794681549072,
+      "learning_rate": 3.9800000000000004e-07,
+      "loss": 2.7478,
+      "step": 99250
+    },
+    {
+      "epoch": 1.986,
+      "grad_norm": 9.242195129394531,
+      "learning_rate": 3.73e-07,
+      "loss": 2.7344,
+      "step": 99300
+    },
+    {
+      "epoch": 1.987,
+      "grad_norm": 8.246512413024902,
+      "learning_rate": 3.48e-07,
+      "loss": 2.6711,
+      "step": 99350
+    },
+    {
+      "epoch": 1.988,
+      "grad_norm": 8.762199401855469,
+      "learning_rate": 3.23e-07,
+      "loss": 2.6444,
+      "step": 99400
+    },
+    {
+      "epoch": 1.9889999999999999,
+      "grad_norm": 7.332550525665283,
+      "learning_rate": 2.98e-07,
+      "loss": 2.5687,
+      "step": 99450
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 12.712468147277832,
+      "learning_rate": 2.7299999999999997e-07,
+      "loss": 2.8477,
+      "step": 99500
+    }
+  ],
+  "logging_steps": 50,
+  "max_steps": 100000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 482880651264000.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}