diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12715 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999447544334568,
+  "eval_steps": 500,
+  "global_step": 9050,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0001104911330865698,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.2099447513812156e-07,
+      "loss": 2.1213,
+      "step": 1
+    },
+    {
+      "epoch": 0.000552455665432849,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.1049723756906078e-06,
+      "loss": 2.3029,
+      "step": 5
+    },
+    {
+      "epoch": 0.001104911330865698,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.2099447513812157e-06,
+      "loss": 2.2939,
+      "step": 10
+    },
+    {
+      "epoch": 0.001657366996298547,
+      "grad_norm": 1.828125,
+      "learning_rate": 3.314917127071823e-06,
+      "loss": 2.2427,
+      "step": 15
+    },
+    {
+      "epoch": 0.002209822661731396,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.419889502762431e-06,
+      "loss": 2.0646,
+      "step": 20
+    },
+    {
+      "epoch": 0.002762278327164245,
+      "grad_norm": 3.65625,
+      "learning_rate": 5.524861878453038e-06,
+      "loss": 2.1394,
+      "step": 25
+    },
+    {
+      "epoch": 0.003314733992597094,
+      "grad_norm": 4.375,
+      "learning_rate": 6.629834254143646e-06,
+      "loss": 2.137,
+      "step": 30
+    },
+    {
+      "epoch": 0.003867189658029943,
+      "grad_norm": 3.890625,
+      "learning_rate": 7.734806629834254e-06,
+      "loss": 2.0171,
+      "step": 35
+    },
+    {
+      "epoch": 0.004419645323462792,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.839779005524863e-06,
+      "loss": 1.9479,
+      "step": 40
+    },
+    {
+      "epoch": 0.004972100988895641,
+      "grad_norm": 1.34375,
+      "learning_rate": 9.94475138121547e-06,
+      "loss": 1.847,
+      "step": 45
+    },
+    {
+      "epoch": 0.00552455665432849,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.1049723756906077e-05,
+      "loss": 1.8616,
+      "step": 50
+    },
+    {
+      "epoch": 0.0060770123197613395,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.2154696132596685e-05,
+      "loss": 1.761,
+      "step": 55
+    },
+    {
+      "epoch": 0.006629467985194188,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.3259668508287292e-05,
+      "loss": 1.7377,
+      "step": 60
+    },
+    {
+      "epoch": 0.007181923650627037,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.4364640883977901e-05,
+      "loss": 1.6858,
+      "step": 65
+    },
+    {
+      "epoch": 0.007734379316059886,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.5469613259668508e-05,
+      "loss": 1.6879,
+      "step": 70
+    },
+    {
+      "epoch": 0.008286834981492736,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.6574585635359117e-05,
+      "loss": 1.6719,
+      "step": 75
+    },
+    {
+      "epoch": 0.008839290646925584,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.7679558011049725e-05,
+      "loss": 1.6673,
+      "step": 80
+    },
+    {
+      "epoch": 0.009391746312358434,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.878453038674033e-05,
+      "loss": 1.5939,
+      "step": 85
+    },
+    {
+      "epoch": 0.009944201977791282,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.988950276243094e-05,
+      "loss": 1.4704,
+      "step": 90
+    },
+    {
+      "epoch": 0.010496657643224131,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.0994475138121548e-05,
+      "loss": 1.6011,
+      "step": 95
+    },
+    {
+      "epoch": 0.01104911330865698,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.2099447513812153e-05,
+      "loss": 1.5358,
+      "step": 100
+    },
+    {
+      "epoch": 0.01160156897408983,
+      "grad_norm": 0.466796875,
+      "learning_rate": 2.3204419889502762e-05,
+      "loss": 1.5575,
+      "step": 105
+    },
+    {
+      "epoch": 0.012154024639522679,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.430939226519337e-05,
+      "loss": 1.5238,
+      "step": 110
+    },
+    {
+      "epoch": 0.012706480304955527,
+      "grad_norm": 0.365234375,
+      "learning_rate": 2.541436464088398e-05,
+      "loss": 1.5489,
+      "step": 115
+    },
+    {
+      "epoch": 0.013258935970388377,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.6519337016574585e-05,
+      "loss": 1.5478,
+      "step": 120
+    },
+    {
+      "epoch": 0.013811391635821225,
+      "grad_norm": 0.373046875,
+      "learning_rate": 2.7624309392265197e-05,
+      "loss": 1.4958,
+      "step": 125
+    },
+    {
+      "epoch": 0.014363847301254075,
+      "grad_norm": 0.37109375,
+      "learning_rate": 2.8729281767955802e-05,
+      "loss": 1.5456,
+      "step": 130
+    },
+    {
+      "epoch": 0.014916302966686923,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.983425414364641e-05,
+      "loss": 1.48,
+      "step": 135
+    },
+    {
+      "epoch": 0.015468758632119772,
+      "grad_norm": 0.384765625,
+      "learning_rate": 3.0939226519337016e-05,
+      "loss": 1.5851,
+      "step": 140
+    },
+    {
+      "epoch": 0.01602121429755262,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.2044198895027625e-05,
+      "loss": 1.5424,
+      "step": 145
+    },
+    {
+      "epoch": 0.016573669962985472,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.3149171270718233e-05,
+      "loss": 1.5361,
+      "step": 150
+    },
+    {
+      "epoch": 0.01712612562841832,
+      "grad_norm": 0.375,
+      "learning_rate": 3.425414364640884e-05,
+      "loss": 1.5748,
+      "step": 155
+    },
+    {
+      "epoch": 0.017678581293851168,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.535911602209945e-05,
+      "loss": 1.4981,
+      "step": 160
+    },
+    {
+      "epoch": 0.018231036959284016,
+      "grad_norm": 0.388671875,
+      "learning_rate": 3.646408839779006e-05,
+      "loss": 1.4862,
+      "step": 165
+    },
+    {
+      "epoch": 0.018783492624716867,
+      "grad_norm": 0.384765625,
+      "learning_rate": 3.756906077348066e-05,
+      "loss": 1.4597,
+      "step": 170
+    },
+    {
+      "epoch": 0.019335948290149715,
+      "grad_norm": 0.45703125,
+      "learning_rate": 3.867403314917128e-05,
+      "loss": 1.5168,
+      "step": 175
+    },
+    {
+      "epoch": 0.019888403955582563,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.977900552486188e-05,
+      "loss": 1.4759,
+      "step": 180
+    },
+    {
+      "epoch": 0.020440859621015415,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.088397790055249e-05,
+      "loss": 1.511,
+      "step": 185
+    },
+    {
+      "epoch": 0.020993315286448263,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.1988950276243096e-05,
+      "loss": 1.4943,
+      "step": 190
+    },
+    {
+      "epoch": 0.02154577095188111,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.3093922651933705e-05,
+      "loss": 1.472,
+      "step": 195
+    },
+    {
+      "epoch": 0.02209822661731396,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.419889502762431e-05,
+      "loss": 1.4631,
+      "step": 200
+    },
+    {
+      "epoch": 0.02265068228274681,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.530386740331492e-05,
+      "loss": 1.5233,
+      "step": 205
+    },
+    {
+      "epoch": 0.02320313794817966,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.6408839779005524e-05,
+      "loss": 1.5071,
+      "step": 210
+    },
+    {
+      "epoch": 0.023755593613612506,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.751381215469613e-05,
+      "loss": 1.4828,
+      "step": 215
+    },
+    {
+      "epoch": 0.024308049279045358,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.861878453038674e-05,
+      "loss": 1.4937,
+      "step": 220
+    },
+    {
+      "epoch": 0.024860504944478206,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.972375690607735e-05,
+      "loss": 1.6039,
+      "step": 225
+    },
+    {
+      "epoch": 0.025412960609911054,
+      "grad_norm": 0.421875,
+      "learning_rate": 5.082872928176796e-05,
+      "loss": 1.4386,
+      "step": 230
+    },
+    {
+      "epoch": 0.025965416275343902,
+      "grad_norm": 0.451171875,
+      "learning_rate": 5.193370165745857e-05,
+      "loss": 1.5466,
+      "step": 235
+    },
+    {
+      "epoch": 0.026517871940776754,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.303867403314917e-05,
+      "loss": 1.4743,
+      "step": 240
+    },
+    {
+      "epoch": 0.0270703276062096,
+      "grad_norm": 0.455078125,
+      "learning_rate": 5.414364640883978e-05,
+      "loss": 1.4524,
+      "step": 245
+    },
+    {
+      "epoch": 0.02762278327164245,
+      "grad_norm": 0.427734375,
+      "learning_rate": 5.5248618784530394e-05,
+      "loss": 1.4057,
+      "step": 250
+    },
+    {
+      "epoch": 0.0281752389370753,
+      "grad_norm": 0.453125,
+      "learning_rate": 5.6353591160220996e-05,
+      "loss": 1.4469,
+      "step": 255
+    },
+    {
+      "epoch": 0.02872769460250815,
+      "grad_norm": 0.4921875,
+      "learning_rate": 5.7458563535911604e-05,
+      "loss": 1.4177,
+      "step": 260
+    },
+    {
+      "epoch": 0.029280150267940997,
+      "grad_norm": 0.5078125,
+      "learning_rate": 5.8563535911602206e-05,
+      "loss": 1.4479,
+      "step": 265
+    },
+    {
+      "epoch": 0.029832605933373845,
+      "grad_norm": 0.47265625,
+      "learning_rate": 5.966850828729282e-05,
+      "loss": 1.427,
+      "step": 270
+    },
+    {
+      "epoch": 0.030385061598806697,
+      "grad_norm": 0.470703125,
+      "learning_rate": 6.077348066298343e-05,
+      "loss": 1.4548,
+      "step": 275
+    },
+    {
+      "epoch": 0.030937517264239545,
+      "grad_norm": 0.486328125,
+      "learning_rate": 6.187845303867403e-05,
+      "loss": 1.5069,
+      "step": 280
+    },
+    {
+      "epoch": 0.03148997292967239,
+      "grad_norm": 0.46484375,
+      "learning_rate": 6.298342541436464e-05,
+      "loss": 1.4171,
+      "step": 285
+    },
+    {
+      "epoch": 0.03204242859510524,
+      "grad_norm": 0.462890625,
+      "learning_rate": 6.408839779005525e-05,
+      "loss": 1.4631,
+      "step": 290
+    },
+    {
+      "epoch": 0.03259488426053809,
+      "grad_norm": 0.466796875,
+      "learning_rate": 6.519337016574586e-05,
+      "loss": 1.4806,
+      "step": 295
+    },
+    {
+      "epoch": 0.033147339925970944,
+      "grad_norm": 0.478515625,
+      "learning_rate": 6.629834254143647e-05,
+      "loss": 1.4121,
+      "step": 300
+    },
+    {
+      "epoch": 0.03369979559140379,
+      "grad_norm": 0.490234375,
+      "learning_rate": 6.740331491712708e-05,
+      "loss": 1.4418,
+      "step": 305
+    },
+    {
+      "epoch": 0.03425225125683664,
+      "grad_norm": 0.5,
+      "learning_rate": 6.850828729281768e-05,
+      "loss": 1.4879,
+      "step": 310
+    },
+    {
+      "epoch": 0.03480470692226949,
+      "grad_norm": 0.4765625,
+      "learning_rate": 6.961325966850829e-05,
+      "loss": 1.4641,
+      "step": 315
+    },
+    {
+      "epoch": 0.035357162587702336,
+      "grad_norm": 0.4765625,
+      "learning_rate": 7.07182320441989e-05,
+      "loss": 1.495,
+      "step": 320
+    },
+    {
+      "epoch": 0.035909618253135184,
+      "grad_norm": 0.474609375,
+      "learning_rate": 7.182320441988951e-05,
+      "loss": 1.4523,
+      "step": 325
+    },
+    {
+      "epoch": 0.03646207391856803,
+      "grad_norm": 0.46875,
+      "learning_rate": 7.292817679558012e-05,
+      "loss": 1.4367,
+      "step": 330
+    },
+    {
+      "epoch": 0.03701452958400089,
+      "grad_norm": 0.451171875,
+      "learning_rate": 7.403314917127073e-05,
+      "loss": 1.47,
+      "step": 335
+    },
+    {
+      "epoch": 0.037566985249433735,
+      "grad_norm": 0.486328125,
+      "learning_rate": 7.513812154696132e-05,
+      "loss": 1.5092,
+      "step": 340
+    },
+    {
+      "epoch": 0.03811944091486658,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.624309392265195e-05,
+      "loss": 1.4718,
+      "step": 345
+    },
+    {
+      "epoch": 0.03867189658029943,
+      "grad_norm": 0.423828125,
+      "learning_rate": 7.734806629834255e-05,
+      "loss": 1.4461,
+      "step": 350
+    },
+    {
+      "epoch": 0.03922435224573228,
+      "grad_norm": 0.458984375,
+      "learning_rate": 7.845303867403315e-05,
+      "loss": 1.4724,
+      "step": 355
+    },
+    {
+      "epoch": 0.03977680791116513,
+      "grad_norm": 0.47265625,
+      "learning_rate": 7.955801104972376e-05,
+      "loss": 1.4095,
+      "step": 360
+    },
+    {
+      "epoch": 0.040329263576597975,
+      "grad_norm": 0.494140625,
+      "learning_rate": 8.066298342541438e-05,
+      "loss": 1.4511,
+      "step": 365
+    },
+    {
+      "epoch": 0.04088171924203083,
+      "grad_norm": 0.44140625,
+      "learning_rate": 8.176795580110498e-05,
+      "loss": 1.49,
+      "step": 370
+    },
+    {
+      "epoch": 0.04143417490746368,
+      "grad_norm": 0.458984375,
+      "learning_rate": 8.287292817679558e-05,
+      "loss": 1.4709,
+      "step": 375
+    },
+    {
+      "epoch": 0.041986630572896526,
+      "grad_norm": 0.447265625,
+      "learning_rate": 8.397790055248619e-05,
+      "loss": 1.46,
+      "step": 380
+    },
+    {
+      "epoch": 0.042539086238329374,
+      "grad_norm": 0.435546875,
+      "learning_rate": 8.50828729281768e-05,
+      "loss": 1.5146,
+      "step": 385
+    },
+    {
+      "epoch": 0.04309154190376222,
+      "grad_norm": 0.4765625,
+      "learning_rate": 8.618784530386741e-05,
+      "loss": 1.4947,
+      "step": 390
+    },
+    {
+      "epoch": 0.04364399756919507,
+      "grad_norm": 0.447265625,
+      "learning_rate": 8.729281767955802e-05,
+      "loss": 1.4608,
+      "step": 395
+    },
+    {
+      "epoch": 0.04419645323462792,
+      "grad_norm": 0.443359375,
+      "learning_rate": 8.839779005524861e-05,
+      "loss": 1.3874,
+      "step": 400
+    },
+    {
+      "epoch": 0.04474890890006077,
+      "grad_norm": 0.447265625,
+      "learning_rate": 8.950276243093924e-05,
+      "loss": 1.4436,
+      "step": 405
+    },
+    {
+      "epoch": 0.04530136456549362,
+      "grad_norm": 0.458984375,
+      "learning_rate": 9.060773480662984e-05,
+      "loss": 1.3886,
+      "step": 410
+    },
+    {
+      "epoch": 0.04585382023092647,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.171270718232044e-05,
+      "loss": 1.4823,
+      "step": 415
+    },
+    {
+      "epoch": 0.04640627589635932,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.281767955801105e-05,
+      "loss": 1.4335,
+      "step": 420
+    },
+    {
+      "epoch": 0.046958731561792165,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.392265193370167e-05,
+      "loss": 1.4415,
+      "step": 425
+    },
+    {
+      "epoch": 0.04751118722722501,
+      "grad_norm": 0.4765625,
+      "learning_rate": 9.502762430939227e-05,
+      "loss": 1.4289,
+      "step": 430
+    },
+    {
+      "epoch": 0.04806364289265786,
+      "grad_norm": 0.4609375,
+      "learning_rate": 9.613259668508287e-05,
+      "loss": 1.4191,
+      "step": 435
+    },
+    {
+      "epoch": 0.048616098558090716,
+      "grad_norm": 0.451171875,
+      "learning_rate": 9.723756906077348e-05,
+      "loss": 1.3955,
+      "step": 440
+    },
+    {
+      "epoch": 0.049168554223523564,
+      "grad_norm": 0.427734375,
+      "learning_rate": 9.834254143646409e-05,
+      "loss": 1.4465,
+      "step": 445
+    },
+    {
+      "epoch": 0.04972100988895641,
+      "grad_norm": 0.5,
+      "learning_rate": 9.94475138121547e-05,
+      "loss": 1.4598,
+      "step": 450
+    },
+    {
+      "epoch": 0.05027346555438926,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00010055248618784532,
+      "loss": 1.4326,
+      "step": 455
+    },
+    {
+      "epoch": 0.05082592121982211,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00010165745856353592,
+      "loss": 1.3476,
+      "step": 460
+    },
+    {
+      "epoch": 0.051378376885254956,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00010276243093922653,
+      "loss": 1.4352,
+      "step": 465
+    },
+    {
+      "epoch": 0.051930832550687804,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00010386740331491714,
+      "loss": 1.4684,
+      "step": 470
+    },
+    {
+      "epoch": 0.05248328821612066,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00010497237569060774,
+      "loss": 1.441,
+      "step": 475
+    },
+    {
+      "epoch": 0.05303574388155351,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00010607734806629834,
+      "loss": 1.4258,
+      "step": 480
+    },
+    {
+      "epoch": 0.053588199546986355,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00010718232044198895,
+      "loss": 1.419,
+      "step": 485
+    },
+    {
+      "epoch": 0.0541406552124192,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00010828729281767956,
+      "loss": 1.3802,
+      "step": 490
+    },
+    {
+      "epoch": 0.05469311087785205,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00010939226519337018,
+      "loss": 1.3694,
+      "step": 495
+    },
+    {
+      "epoch": 0.0552455665432849,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00011049723756906079,
+      "loss": 1.4704,
+      "step": 500
+    },
+    {
+      "epoch": 0.05579802220871775,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001116022099447514,
+      "loss": 1.3262,
+      "step": 505
+    },
+    {
+      "epoch": 0.0563504778741506,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00011270718232044199,
+      "loss": 1.4695,
+      "step": 510
+    },
+    {
+      "epoch": 0.05690293353958345,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001138121546961326,
+      "loss": 1.4465,
+      "step": 515
+    },
+    {
+      "epoch": 0.0574553892050163,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00011491712707182321,
+      "loss": 1.4442,
+      "step": 520
+    },
+    {
+      "epoch": 0.058007844870449146,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0001160220994475138,
+      "loss": 1.4143,
+      "step": 525
+    },
+    {
+      "epoch": 0.058560300535881994,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00011712707182320441,
+      "loss": 1.4964,
+      "step": 530
+    },
+    {
+      "epoch": 0.05911275620131484,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00011823204419889505,
+      "loss": 1.4598,
+      "step": 535
+    },
+    {
+      "epoch": 0.05966521186674769,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00011933701657458564,
+      "loss": 1.348,
+      "step": 540
+    },
+    {
+      "epoch": 0.060217667532180545,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00012044198895027625,
+      "loss": 1.3783,
+      "step": 545
+    },
+    {
+      "epoch": 0.06077012319761339,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00012154696132596686,
+      "loss": 1.3765,
+      "step": 550
+    },
+    {
+      "epoch": 0.06132257886304624,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00012265193370165746,
+      "loss": 1.4892,
+      "step": 555
+    },
+    {
+      "epoch": 0.06187503452847909,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00012375690607734806,
+      "loss": 1.4185,
+      "step": 560
+    },
+    {
+      "epoch": 0.06242749019391194,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00012486187845303867,
+      "loss": 1.3789,
+      "step": 565
+    },
+    {
+      "epoch": 0.06297994585934479,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00012596685082872928,
+      "loss": 1.4294,
+      "step": 570
+    },
+    {
+      "epoch": 0.06353240152477764,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00012707182320441992,
+      "loss": 1.3654,
+      "step": 575
+    },
+    {
+      "epoch": 0.06408485719021048,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001281767955801105,
+      "loss": 1.3842,
+      "step": 580
+    },
+    {
+      "epoch": 0.06463731285564334,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001292817679558011,
+      "loss": 1.422,
+      "step": 585
+    },
+    {
+      "epoch": 0.06518976852107618,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00013038674033149172,
+      "loss": 1.4607,
+      "step": 590
+    },
+    {
+      "epoch": 0.06574222418650903,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00013149171270718233,
+      "loss": 1.4292,
+      "step": 595
+    },
+    {
+      "epoch": 0.06629467985194189,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00013259668508287293,
+      "loss": 1.4353,
+      "step": 600
+    },
+    {
+      "epoch": 0.06684713551737473,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00013370165745856354,
+      "loss": 1.4552,
+      "step": 605
+    },
+    {
+      "epoch": 0.06739959118280758,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00013480662983425415,
+      "loss": 1.4496,
+      "step": 610
+    },
+    {
+      "epoch": 0.06795204684824042,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00013591160220994476,
+      "loss": 1.4456,
+      "step": 615
+    },
+    {
+      "epoch": 0.06850450251367328,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00013701657458563537,
+      "loss": 1.4358,
+      "step": 620
+    },
+    {
+      "epoch": 0.06905695817910612,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00013812154696132598,
+      "loss": 1.473,
+      "step": 625
+    },
+    {
+      "epoch": 0.06960941384453898,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00013922651933701659,
+      "loss": 1.3869,
+      "step": 630
+    },
+    {
+      "epoch": 0.07016186950997183,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001403314917127072,
+      "loss": 1.3762,
+      "step": 635
+    },
+    {
+      "epoch": 0.07071432517540467,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001414364640883978,
+      "loss": 1.3875,
+      "step": 640
+    },
+    {
+      "epoch": 0.07126678084083753,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00014254143646408839,
+      "loss": 1.3839,
+      "step": 645
+    },
+    {
+      "epoch": 0.07181923650627037,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00014364640883977902,
+      "loss": 1.4395,
+      "step": 650
+    },
+    {
+      "epoch": 0.07237169217170322,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00014475138121546963,
+      "loss": 1.4761,
+      "step": 655
+    },
+    {
+      "epoch": 0.07292414783713606,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00014585635359116024,
+      "loss": 1.4714,
+      "step": 660
+    },
+    {
+      "epoch": 0.07347660350256892,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00014696132596685085,
+      "loss": 1.403,
+      "step": 665
+    },
+    {
+      "epoch": 0.07402905916800177,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00014806629834254146,
+      "loss": 1.4563,
+      "step": 670
+    },
+    {
+      "epoch": 0.07458151483343461,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00014917127071823204,
+      "loss": 1.4246,
+      "step": 675
+    },
+    {
+      "epoch": 0.07513397049886747,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00015027624309392265,
+      "loss": 1.4217,
+      "step": 680
+    },
+    {
+      "epoch": 0.07568642616430031,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00015138121546961325,
+      "loss": 1.4074,
+      "step": 685
+    },
+    {
+      "epoch": 0.07623888182973317,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0001524861878453039,
+      "loss": 1.4298,
+      "step": 690
+    },
+    {
+      "epoch": 0.076791337495166,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0001535911602209945,
+      "loss": 1.4207,
+      "step": 695
+    },
+    {
+      "epoch": 0.07734379316059886,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0001546961325966851,
+      "loss": 1.4175,
+      "step": 700
+    },
+    {
+      "epoch": 0.07789624882603172,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0001558011049723757,
+      "loss": 1.4153,
+      "step": 705
+    },
+    {
+      "epoch": 0.07844870449146456,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0001569060773480663,
+      "loss": 1.2493,
+      "step": 710
+    },
+    {
+      "epoch": 0.07900116015689741,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0001580110497237569,
+      "loss": 1.2664,
+      "step": 715
+    },
+    {
+      "epoch": 0.07955361582233025,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00015911602209944752,
+      "loss": 1.3678,
+      "step": 720
+    },
+    {
+      "epoch": 0.08010607148776311,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00016022099447513812,
+      "loss": 1.395,
+      "step": 725
+    },
+    {
+      "epoch": 0.08065852715319595,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00016132596685082876,
+      "loss": 1.4848,
+      "step": 730
+    },
+    {
+      "epoch": 0.0812109828186288,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00016243093922651934,
+      "loss": 1.303,
+      "step": 735
+    },
+    {
+      "epoch": 0.08176343848406166,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00016353591160220995,
+      "loss": 1.3469,
+      "step": 740
+    },
+    {
+      "epoch": 0.0823158941494945,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00016464088397790056,
+      "loss": 1.407,
+      "step": 745
+    },
+    {
+      "epoch": 0.08286834981492736,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00016574585635359117,
+      "loss": 1.4167,
+      "step": 750
+    },
+    {
+      "epoch": 0.0834208054803602,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00016685082872928178,
+      "loss": 1.3591,
+      "step": 755
+    },
+    {
+      "epoch": 0.08397326114579305,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00016795580110497238,
+      "loss": 1.4427,
+      "step": 760
+    },
+    {
+      "epoch": 0.08452571681122589,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.000169060773480663,
+      "loss": 1.407,
+      "step": 765
+    },
+    {
+      "epoch": 0.08507817247665875,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0001701657458563536,
+      "loss": 1.4429,
+      "step": 770
+    },
+    {
+      "epoch": 0.0856306281420916,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0001712707182320442,
+      "loss": 1.3577,
+      "step": 775
+    },
+    {
+      "epoch": 0.08618308380752444,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00017237569060773482,
+      "loss": 1.3686,
+      "step": 780
+    },
+    {
+      "epoch": 0.0867355394729573,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00017348066298342543,
+      "loss": 1.3666,
+      "step": 785
+    },
+    {
+      "epoch": 0.08728799513839014,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00017458563535911604,
+      "loss": 1.4079,
+      "step": 790
+    },
+    {
+      "epoch": 0.087840450803823,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00017569060773480665,
+      "loss": 1.4109,
+      "step": 795
+    },
+    {
+      "epoch": 0.08839290646925584,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00017679558011049723,
+      "loss": 1.3654,
+      "step": 800
+    },
+    {
+      "epoch": 0.08894536213468869,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00017790055248618784,
+      "loss": 1.4313,
+      "step": 805
+    },
+    {
+      "epoch": 0.08949781780012155,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00017900552486187847,
+      "loss": 1.3809,
+      "step": 810
+    },
+    {
+      "epoch": 0.09005027346555439,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00018011049723756908,
+      "loss": 1.4881,
+      "step": 815
+    },
+    {
+      "epoch": 0.09060272913098724,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0001812154696132597,
+      "loss": 1.4934,
+      "step": 820
+    },
+    {
+      "epoch": 0.09115518479642008,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0001823204419889503,
+      "loss": 1.4283,
+      "step": 825
+    },
+    {
+      "epoch": 0.09170764046185294,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00018342541436464088,
+      "loss": 1.3682,
+      "step": 830
+    },
+    {
+      "epoch": 0.09226009612728578,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0001845303867403315,
+      "loss": 1.3781,
+      "step": 835
+    },
+    {
+      "epoch": 0.09281255179271863,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0001856353591160221,
+      "loss": 1.4479,
+      "step": 840
+    },
+    {
+      "epoch": 0.09336500745815149,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001867403314917127,
+      "loss": 1.4122,
+      "step": 845
+    },
+    {
+      "epoch": 0.09391746312358433,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00018784530386740334,
+      "loss": 1.3943,
+      "step": 850
+    },
+    {
+      "epoch": 0.09446991878901718,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00018895027624309395,
+      "loss": 1.3659,
+      "step": 855
+    },
+    {
+      "epoch": 0.09502237445445003,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019005524861878453,
+      "loss": 1.3628,
+      "step": 860
+    },
+    {
+      "epoch": 0.09557483011988288,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019116022099447514,
+      "loss": 1.3492,
+      "step": 865
+    },
+    {
+      "epoch": 0.09612728578531572,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019226519337016575,
+      "loss": 1.4476,
+      "step": 870
+    },
+    {
+      "epoch": 0.09667974145074858,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019337016574585636,
+      "loss": 1.3911,
+      "step": 875
+    },
+    {
+      "epoch": 0.09723219711618143,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019447513812154697,
+      "loss": 1.3648,
+      "step": 880
+    },
+    {
+      "epoch": 0.09778465278161427,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019558011049723757,
+      "loss": 1.3301,
+      "step": 885
+    },
+    {
+      "epoch": 0.09833710844704713,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019668508287292818,
+      "loss": 1.4445,
+      "step": 890
+    },
+    {
+      "epoch": 0.09888956411247997,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001977900552486188,
+      "loss": 1.4455,
+      "step": 895
+    },
+    {
+      "epoch": 0.09944201977791282,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0001988950276243094,
+      "loss": 1.3361,
+      "step": 900
+    },
+    {
+      "epoch": 0.09999447544334567,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0002,
+      "loss": 1.4043,
+      "step": 905
+    },
+    {
+      "epoch": 0.10054693110877852,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019999981403661345,
+      "loss": 1.375,
+      "step": 910
+    },
+    {
+      "epoch": 0.10109938677421137,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019999925614714537,
+      "loss": 1.3785,
+      "step": 915
+    },
+    {
+      "epoch": 0.10165184243964422,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019999832633367076,
+      "loss": 1.2666,
+      "step": 920
+    },
+    {
+      "epoch": 0.10220429810507707,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0001999970245996478,
+      "loss": 1.4409,
+      "step": 925
+    },
+    {
+      "epoch": 0.10275675377050991,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019999535094991798,
+      "loss": 1.3618,
+      "step": 930
+    },
+    {
+      "epoch": 0.10330920943594277,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019999330539070617,
+      "loss": 1.4307,
+      "step": 935
+    },
+    {
+      "epoch": 0.10386166510137561,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019999088792962017,
+      "loss": 1.2785,
+      "step": 940
+    },
+    {
+      "epoch": 0.10441412076680846,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019998809857565131,
+      "loss": 1.3679,
+      "step": 945
+    },
+    {
+      "epoch": 0.10496657643224132,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019998493733917384,
+      "loss": 1.3698,
+      "step": 950
+    },
+    {
+      "epoch": 0.10551903209767416,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019998140423194534,
+      "loss": 1.3474,
+      "step": 955
+    },
+    {
+      "epoch": 0.10607148776310701,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0001999774992671063,
+      "loss": 1.4296,
+      "step": 960
+    },
+    {
+      "epoch": 0.10662394342853986,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0001999732224591804,
+      "loss": 1.3559,
+      "step": 965
+    },
+    {
+      "epoch": 0.10717639909397271,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001999685738240742,
+      "loss": 1.41,
+      "step": 970
+    },
+    {
+      "epoch": 0.10772885475940555,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00019996355337907718,
+      "loss": 1.3789,
+      "step": 975
+    },
+    {
+      "epoch": 0.1082813104248384,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00019995816114286178,
+      "loss": 1.4477,
+      "step": 980
+    },
+    {
+      "epoch": 0.10883376609027126,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019995239713548318,
+      "loss": 1.4113,
+      "step": 985
+    },
+    {
+      "epoch": 0.1093862217557041,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019994626137837917,
+      "loss": 1.3787,
+      "step": 990
+    },
+    {
+      "epoch": 0.10993867742113696,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00019993975389437038,
+      "loss": 1.3096,
+      "step": 995
+    },
+    {
+      "epoch": 0.1104911330865698,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00019993287470765984,
+      "loss": 1.359,
+      "step": 1000
+    },
+    {
+      "epoch": 0.11104358875200265,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019992562384383309,
+      "loss": 1.4228,
+      "step": 1005
+    },
+    {
+      "epoch": 0.1115960444174355,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00019991800132985804,
+      "loss": 1.4003,
+      "step": 1010
+    },
+    {
+      "epoch": 0.11214850008286835,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00019991000719408485,
+      "loss": 1.3181,
+      "step": 1015
+    },
+    {
+      "epoch": 0.1127009557483012,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019990164146624584,
+      "loss": 1.3981,
+      "step": 1020
+    },
+    {
+      "epoch": 0.11325341141373405,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00019989290417745542,
+      "loss": 1.3001,
+      "step": 1025
+    },
+    {
+      "epoch": 0.1138058670791669,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019988379536020986,
+      "loss": 1.4415,
+      "step": 1030
+    },
+    {
+      "epoch": 0.11435832274459974,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00019987431504838737,
+      "loss": 1.3646,
+      "step": 1035
+    },
+    {
+      "epoch": 0.1149107784100326,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0001998644632772477,
+      "loss": 1.3522,
+      "step": 1040
+    },
+    {
+      "epoch": 0.11546323407546544,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019985424008343226,
+      "loss": 1.3817,
+      "step": 1045
+    },
+    {
+      "epoch": 0.11601568974089829,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019984364550496383,
+      "loss": 1.3372,
+      "step": 1050
+    },
+    {
+      "epoch": 0.11656814540633115,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019983267958124644,
+      "loss": 1.3866,
+      "step": 1055
+    },
+    {
+      "epoch": 0.11712060107176399,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001998213423530654,
+      "loss": 1.4264,
+      "step": 1060
+    },
+    {
+      "epoch": 0.11767305673719684,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019980963386258683,
+      "loss": 1.4479,
+      "step": 1065
+    },
+    {
+      "epoch": 0.11822551240262968,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001997975541533577,
+      "loss": 1.4479,
+      "step": 1070
+    },
+    {
+      "epoch": 0.11877796806806254,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019978510327030579,
+      "loss": 1.3959,
+      "step": 1075
+    },
+    {
+      "epoch": 0.11933042373349538,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019977228125973916,
+      "loss": 1.3914,
+      "step": 1080
+    },
+    {
+      "epoch": 0.11988287939892824,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0001997590881693464,
+      "loss": 1.3624,
+      "step": 1085
+    },
+    {
+      "epoch": 0.12043533506436109,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019974552404819607,
+      "loss": 1.3704,
+      "step": 1090
+    },
+    {
+      "epoch": 0.12098779072979393,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0001997315889467368,
+      "loss": 1.2807,
+      "step": 1095
+    },
+    {
+      "epoch": 0.12154024639522679,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0001997172829167969,
+      "loss": 1.419,
+      "step": 1100
+    },
+    {
+      "epoch": 0.12209270206065963,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00019970260601158443,
+      "loss": 1.377,
+      "step": 1105
+    },
+    {
+      "epoch": 0.12264515772609248,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00019968755828568668,
+      "loss": 1.3901,
+      "step": 1110
+    },
+    {
+      "epoch": 0.12319761339152532,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019967213979507018,
+      "loss": 1.3007,
+      "step": 1115
+    },
+    {
+      "epoch": 0.12375006905695818,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001996563505970804,
+      "loss": 1.335,
+      "step": 1120
+    },
+    {
+      "epoch": 0.12430252472239103,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019964019075044163,
+      "loss": 1.2959,
+      "step": 1125
+    },
+    {
+      "epoch": 0.12485498038782387,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019962366031525664,
+      "loss": 1.3547,
+      "step": 1130
+    },
+    {
+      "epoch": 0.12540743605325672,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0001996067593530065,
+      "loss": 1.356,
+      "step": 1135
+    },
+    {
+      "epoch": 0.12595989171868957,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019958948792655055,
+      "loss": 1.3994,
+      "step": 1140
+    },
+    {
+      "epoch": 0.12651234738412243,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0001995718461001257,
+      "loss": 1.3695,
+      "step": 1145
+    },
+    {
+      "epoch": 0.12706480304955528,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00019955383393934674,
+      "loss": 1.3279,
+      "step": 1150
+    },
+    {
+      "epoch": 0.12761725871498814,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00019953545151120565,
+      "loss": 1.3996,
+      "step": 1155
+    },
+    {
+      "epoch": 0.12816971438042096,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001995166988840716,
+      "loss": 1.3476,
+      "step": 1160
+    },
+    {
+      "epoch": 0.12872217004585382,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019949757612769067,
+      "loss": 1.3534,
+      "step": 1165
+    },
+    {
+      "epoch": 0.12927462571128667,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001994780833131855,
+      "loss": 1.4534,
+      "step": 1170
+    },
+    {
+      "epoch": 0.12982708137671953,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00019945822051305507,
+      "loss": 1.3668,
+      "step": 1175
+    },
+    {
+      "epoch": 0.13037953704215235,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019943798780117447,
+      "loss": 1.4138,
+      "step": 1180
+    },
+    {
+      "epoch": 0.1309319927075852,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00019941738525279453,
+      "loss": 1.3172,
+      "step": 1185
+    },
+    {
+      "epoch": 0.13148444837301806,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019939641294454172,
+      "loss": 1.394,
+      "step": 1190
+    },
+    {
+      "epoch": 0.13203690403845092,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0001993750709544176,
+      "loss": 1.3651,
+      "step": 1195
+    },
+    {
+      "epoch": 0.13258935970388377,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019935335936179874,
+      "loss": 1.4192,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1331418153693166,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019933127824743645,
+      "loss": 1.3419,
+      "step": 1205
+    },
+    {
+      "epoch": 0.13369427103474946,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019930882769345624,
+      "loss": 1.3274,
+      "step": 1210
+    },
+    {
+      "epoch": 0.1342467267001823,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019928600778335774,
+      "loss": 1.4188,
+      "step": 1215
+    },
+    {
+      "epoch": 0.13479918236561517,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001992628186020143,
+      "loss": 1.4108,
+      "step": 1220
+    },
+    {
+      "epoch": 0.13535163803104802,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0001992392602356727,
+      "loss": 1.4137,
+      "step": 1225
+    },
+    {
+      "epoch": 0.13590409369648085,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019921533277195283,
+      "loss": 1.3725,
+      "step": 1230
+    },
+    {
+      "epoch": 0.1364565493619137,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019919103629984728,
+      "loss": 1.3933,
+      "step": 1235
+    },
+    {
+      "epoch": 0.13700900502734656,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0001991663709097212,
+      "loss": 1.347,
+      "step": 1240
+    },
+    {
+      "epoch": 0.1375614606927794,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019914133669331175,
+      "loss": 1.3592,
+      "step": 1245
+    },
+    {
+      "epoch": 0.13811391635821224,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00019911593374372788,
+      "loss": 1.3287,
+      "step": 1250
+    },
+    {
+      "epoch": 0.1386663720236451,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019909016215544998,
+      "loss": 1.418,
+      "step": 1255
+    },
+    {
+      "epoch": 0.13921882768907795,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019906402202432945,
+      "loss": 1.3322,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1397712833545108,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019903751344758848,
+      "loss": 1.3598,
+      "step": 1265
+    },
+    {
+      "epoch": 0.14032373901994366,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019901063652381953,
+      "loss": 1.3468,
+      "step": 1270
+    },
+    {
+      "epoch": 0.1408761946853765,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00019898339135298508,
+      "loss": 1.2972,
+      "step": 1275
+    },
+    {
+      "epoch": 0.14142865035080934,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019895577803641726,
+      "loss": 1.3682,
+      "step": 1280
+    },
+    {
+      "epoch": 0.1419811060162422,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019892779667681732,
+      "loss": 1.3667,
+      "step": 1285
+    },
+    {
+      "epoch": 0.14253356168167505,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019889944737825545,
+      "loss": 1.4319,
+      "step": 1290
+    },
+    {
+      "epoch": 0.1430860173471079,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001988707302461703,
+      "loss": 1.3237,
+      "step": 1295
+    },
+    {
+      "epoch": 0.14363847301254073,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019884164538736858,
+      "loss": 1.3517,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1441909286779736,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019881219291002463,
+      "loss": 1.3761,
+      "step": 1305
+    },
+    {
+      "epoch": 0.14474338434340644,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019878237292368013,
+      "loss": 1.421,
+      "step": 1310
+    },
+    {
+      "epoch": 0.1452958400088393,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019875218553924357,
+      "loss": 1.3595,
+      "step": 1315
+    },
+    {
+      "epoch": 0.14584829567427213,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019872163086898993,
+      "loss": 1.404,
+      "step": 1320
+    },
+    {
+      "epoch": 0.14640075133970498,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00019869070902656018,
+      "loss": 1.3475,
+      "step": 1325
+    },
+    {
+      "epoch": 0.14695320700513784,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019865942012696098,
+      "loss": 1.4619,
+      "step": 1330
+    },
+    {
+      "epoch": 0.1475056626705707,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0001986277642865641,
+      "loss": 1.3436,
+      "step": 1335
+    },
+    {
+      "epoch": 0.14805811833600355,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019859574162310608,
+      "loss": 1.3013,
+      "step": 1340
+    },
+    {
+      "epoch": 0.14861057400143637,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0001985633522556878,
+      "loss": 1.3682,
+      "step": 1345
+    },
+    {
+      "epoch": 0.14916302966686923,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019853059630477396,
+      "loss": 1.3098,
+      "step": 1350
+    },
+    {
+      "epoch": 0.14971548533230208,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019849747389219272,
+      "loss": 1.3603,
+      "step": 1355
+    },
+    {
+      "epoch": 0.15026794099773494,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0001984639851411352,
+      "loss": 1.3294,
+      "step": 1360
+    },
+    {
+      "epoch": 0.1508203966631678,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019843013017615505,
+      "loss": 1.3643,
+      "step": 1365
+    },
+    {
+      "epoch": 0.15137285232860062,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019839590912316791,
+      "loss": 1.3405,
+      "step": 1370
+    },
+    {
+      "epoch": 0.15192530799403348,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0001983613221094511,
+      "loss": 1.3499,
+      "step": 1375
+    },
+    {
+      "epoch": 0.15247776365946633,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001983263692636429,
+      "loss": 1.3834,
+      "step": 1380
+    },
+    {
+      "epoch": 0.15303021932489919,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001982910507157424,
+      "loss": 1.3726,
+      "step": 1385
+    },
+    {
+      "epoch": 0.153582674990332,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019825536659710867,
+      "loss": 1.3209,
+      "step": 1390
+    },
+    {
+      "epoch": 0.15413513065576487,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00019821931704046047,
+      "loss": 1.331,
+      "step": 1395
+    },
+    {
+      "epoch": 0.15468758632119772,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00019818290217987587,
+      "loss": 1.3082,
+      "step": 1400
+    },
+    {
+      "epoch": 0.15524004198663058,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0001981461221507914,
+      "loss": 1.335,
+      "step": 1405
+    },
+    {
+      "epoch": 0.15579249765206343,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0001981089770900018,
+      "loss": 1.3439,
+      "step": 1410
+    },
+    {
+      "epoch": 0.15634495331749626,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019807146713565955,
+      "loss": 1.3955,
+      "step": 1415
+    },
+    {
+      "epoch": 0.15689740898292912,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019803359242727425,
+      "loss": 1.3849,
+      "step": 1420
+    },
+    {
+      "epoch": 0.15744986464836197,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019799535310571203,
+      "loss": 1.351,
+      "step": 1425
+    },
+    {
+      "epoch": 0.15800232031379483,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00019795674931319515,
+      "loss": 1.3958,
+      "step": 1430
+    },
+    {
+      "epoch": 0.15855477597922768,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0001979177811933015,
+      "loss": 1.2827,
+      "step": 1435
+    },
+    {
+      "epoch": 0.1591072316446605,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0001978784488909639,
+      "loss": 1.3038,
+      "step": 1440
+    },
+    {
+      "epoch": 0.15965968731009336,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019783875255246973,
+      "loss": 1.3968,
+      "step": 1445
+    },
+    {
+      "epoch": 0.16021214297552622,
+      "grad_norm": 11.0,
+      "learning_rate": 0.00019779869232546034,
+      "loss": 1.387,
+      "step": 1450
+    },
+    {
+      "epoch": 0.16076459864095907,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001977582683589304,
+      "loss": 1.427,
+      "step": 1455
+    },
+    {
+      "epoch": 0.1613170543063919,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00019771748080322745,
+      "loss": 1.2821,
+      "step": 1460
+    },
+    {
+      "epoch": 0.16186950997182475,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00019767632981005138,
+      "loss": 1.3751,
+      "step": 1465
+    },
+    {
+      "epoch": 0.1624219656372576,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0001976348155324537,
+      "loss": 1.3822,
+      "step": 1470
+    },
+    {
+      "epoch": 0.16297442130269046,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019759293812483713,
+      "loss": 1.4484,
+      "step": 1475
+    },
+    {
+      "epoch": 0.16352687696812332,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.000197550697742955,
+      "loss": 1.3884,
+      "step": 1480
+    },
+    {
+      "epoch": 0.16407933263355615,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0001975080945439106,
+      "loss": 1.3392,
+      "step": 1485
+    },
+    {
+      "epoch": 0.164631788298989,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019746512868615656,
+      "loss": 1.3039,
+      "step": 1490
+    },
+    {
+      "epoch": 0.16518424396442186,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001974218003294945,
+      "loss": 1.3382,
+      "step": 1495
+    },
+    {
+      "epoch": 0.1657366996298547,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0001973781096350741,
+      "loss": 1.2985,
+      "step": 1500
+    },
+    {
+      "epoch": 0.16628915529528757,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0001973340567653928,
+      "loss": 1.321,
+      "step": 1505
+    },
+    {
+      "epoch": 0.1668416109607204,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019728964188429503,
+      "loss": 1.3347,
+      "step": 1510
+    },
+    {
+      "epoch": 0.16739406662615325,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019724486515697155,
+      "loss": 1.3552,
+      "step": 1515
+    },
+    {
+      "epoch": 0.1679465222915861,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019719972674995905,
+      "loss": 1.3776,
+      "step": 1520
+    },
+    {
+      "epoch": 0.16849897795701896,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00019715422683113938,
+      "loss": 1.4107,
+      "step": 1525
+    },
+    {
+      "epoch": 0.16905143362245179,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019710836556973885,
+      "loss": 1.3361,
+      "step": 1530
+    },
+    {
+      "epoch": 0.16960388928788464,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019706214313632784,
+      "loss": 1.3714,
+      "step": 1535
+    },
+    {
+      "epoch": 0.1701563449533175,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00019701555970281988,
+      "loss": 1.2905,
+      "step": 1540
+    },
+    {
+      "epoch": 0.17070880061875035,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0001969686154424713,
+      "loss": 1.3327,
+      "step": 1545
+    },
+    {
+      "epoch": 0.1712612562841832,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00019692131052988034,
+      "loss": 1.3297,
+      "step": 1550
+    },
+    {
+      "epoch": 0.17181371194961603,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019687364514098664,
+      "loss": 1.3781,
+      "step": 1555
+    },
+    {
+      "epoch": 0.1723661676150489,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019682561945307052,
+      "loss": 1.2793,
+      "step": 1560
+    },
+    {
+      "epoch": 0.17291862328048174,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019677723364475237,
+      "loss": 1.3794,
+      "step": 1565
+    },
+    {
+      "epoch": 0.1734710789459146,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019672848789599204,
+      "loss": 1.3247,
+      "step": 1570
+    },
+    {
+      "epoch": 0.17402353461134745,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019667938238808797,
+      "loss": 1.393,
+      "step": 1575
+    },
+    {
+      "epoch": 0.17457599027678028,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019662991730367663,
+      "loss": 1.3854,
+      "step": 1580
+    },
+    {
+      "epoch": 0.17512844594221313,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00019658009282673202,
+      "loss": 1.3966,
+      "step": 1585
+    },
+    {
+      "epoch": 0.175680901607646,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00019652990914256467,
+      "loss": 1.3691,
+      "step": 1590
+    },
+    {
+      "epoch": 0.17623335727307884,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019647936643782109,
+      "loss": 1.4073,
+      "step": 1595
+    },
+    {
+      "epoch": 0.17678581293851167,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0001964284649004832,
+      "loss": 1.373,
+      "step": 1600
+    },
+    {
+      "epoch": 0.17733826860394453,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019637720471986735,
+      "loss": 1.3112,
+      "step": 1605
+    },
+    {
+      "epoch": 0.17789072426937738,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00019632558608662402,
+      "loss": 1.2798,
+      "step": 1610
+    },
+    {
+      "epoch": 0.17844317993481024,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0001962736091927366,
+      "loss": 1.258,
+      "step": 1615
+    },
+    {
+      "epoch": 0.1789956356002431,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019622127423152112,
+      "loss": 1.308,
+      "step": 1620
+    },
+    {
+      "epoch": 0.17954809126567592,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019616858139762534,
+      "loss": 1.3554,
+      "step": 1625
+    },
+    {
+      "epoch": 0.18010054693110877,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00019611553088702798,
+      "loss": 1.3128,
+      "step": 1630
+    },
+    {
+      "epoch": 0.18065300259654163,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0001960621228970381,
+      "loss": 1.3385,
+      "step": 1635
+    },
+    {
+      "epoch": 0.18120545826197448,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0001960083576262943,
+      "loss": 1.396,
+      "step": 1640
+    },
+    {
+      "epoch": 0.18175791392740734,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019595423527476405,
+      "loss": 1.3687,
+      "step": 1645
+    },
+    {
+      "epoch": 0.18231036959284017,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00019589975604374286,
+      "loss": 1.4066,
+      "step": 1650
+    },
+    {
+      "epoch": 0.18286282525827302,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019584492013585355,
+      "loss": 1.3031,
+      "step": 1655
+    },
+    {
+      "epoch": 0.18341528092370588,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00019578972775504555,
+      "loss": 1.3094,
+      "step": 1660
+    },
+    {
+      "epoch": 0.18396773658913873,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019573417910659412,
+      "loss": 1.3321,
+      "step": 1665
+    },
+    {
+      "epoch": 0.18452019225457156,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00019567827439709954,
+      "loss": 1.3338,
+      "step": 1670
+    },
+    {
+      "epoch": 0.1850726479200044,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00019562201383448638,
+      "loss": 1.2639,
+      "step": 1675
+    },
+    {
+      "epoch": 0.18562510358543727,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019556539762800276,
+      "loss": 1.335,
+      "step": 1680
+    },
+    {
+      "epoch": 0.18617755925087012,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00019550842598821952,
+      "loss": 1.3488,
+      "step": 1685
+    },
+    {
+      "epoch": 0.18673001491630298,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001954510991270294,
+      "loss": 1.3334,
+      "step": 1690
+    },
+    {
+      "epoch": 0.1872824705817358,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00019539341725764638,
+      "loss": 1.3627,
+      "step": 1695
+    },
+    {
+      "epoch": 0.18783492624716866,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00019533538059460475,
+      "loss": 1.3849,
+      "step": 1700
+    },
+    {
+      "epoch": 0.18838738191260151,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0001952769893537584,
+      "loss": 1.352,
+      "step": 1705
+    },
+    {
+      "epoch": 0.18893983757803437,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019521824375228004,
+      "loss": 1.2987,
+      "step": 1710
+    },
+    {
+      "epoch": 0.18949229324346722,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001951591440086602,
+      "loss": 1.3289,
+      "step": 1715
+    },
+    {
+      "epoch": 0.19004474890890005,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00019509969034270673,
+      "loss": 1.2721,
+      "step": 1720
+    },
+    {
+      "epoch": 0.1905972045743329,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0001950398829755437,
+      "loss": 1.3934,
+      "step": 1725
+    },
+    {
+      "epoch": 0.19114966023976576,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0001949797221296107,
+      "loss": 1.329,
+      "step": 1730
+    },
+    {
+      "epoch": 0.19170211590519862,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00019491920802866205,
+      "loss": 1.4029,
+      "step": 1735
+    },
+    {
+      "epoch": 0.19225457157063144,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00019485834089776586,
+      "loss": 1.3859,
+      "step": 1740
+    },
+    {
+      "epoch": 0.1928070272360643,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00019479712096330336,
+      "loss": 1.3165,
+      "step": 1745
+    },
+    {
+      "epoch": 0.19335948290149715,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0001947355484529678,
+      "loss": 1.2785,
+      "step": 1750
+    },
+    {
+      "epoch": 0.19391193856693,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00019467362359576386,
+      "loss": 1.2629,
+      "step": 1755
+    },
+    {
+      "epoch": 0.19446439423236286,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019461134662200668,
+      "loss": 1.308,
+      "step": 1760
+    },
+    {
+      "epoch": 0.1950168498977957,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019454871776332095,
+      "loss": 1.3369,
+      "step": 1765
+    },
+    {
+      "epoch": 0.19556930556322855,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00019448573725264022,
+      "loss": 1.2883,
+      "step": 1770
+    },
+    {
+      "epoch": 0.1961217612286614,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019442240532420584,
+      "loss": 1.409,
+      "step": 1775
+    },
+    {
+      "epoch": 0.19667421689409426,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0001943587222135662,
+      "loss": 1.2601,
+      "step": 1780
+    },
+    {
+      "epoch": 0.1972266725595271,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019429468815757587,
+      "loss": 1.2943,
+      "step": 1785
+    },
+    {
+      "epoch": 0.19777912822495994,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019423030339439464,
+      "loss": 1.2805,
+      "step": 1790
+    },
+    {
+      "epoch": 0.1983315838903928,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00019416556816348663,
+      "loss": 1.3609,
+      "step": 1795
+    },
+    {
+      "epoch": 0.19888403955582565,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019410048270561956,
+      "loss": 1.3424,
+      "step": 1800
+    },
+    {
+      "epoch": 0.1994364952212585,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0001940350472628637,
+      "loss": 1.3322,
+      "step": 1805
+    },
+    {
+      "epoch": 0.19998895088669133,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019396926207859084,
+      "loss": 1.3729,
+      "step": 1810
+    },
+    {
+      "epoch": 0.20054140655212419,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00019390312739747385,
+      "loss": 1.3087,
+      "step": 1815
+    },
+    {
+      "epoch": 0.20109386221755704,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0001938366434654852,
+      "loss": 1.2203,
+      "step": 1820
+    },
+    {
+      "epoch": 0.2016463178829899,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019376981052989653,
+      "loss": 1.294,
+      "step": 1825
+    },
+    {
+      "epoch": 0.20219877354842275,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019370262883927733,
+      "loss": 1.314,
+      "step": 1830
+    },
+    {
+      "epoch": 0.20275122921385558,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00019363509864349436,
+      "loss": 1.3179,
+      "step": 1835
+    },
+    {
+      "epoch": 0.20330368487928843,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001935672201937105,
+      "loss": 1.3179,
+      "step": 1840
+    },
+    {
+      "epoch": 0.2038561405447213,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019349899374238383,
+      "loss": 1.2649,
+      "step": 1845
+    },
+    {
+      "epoch": 0.20440859621015414,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0001934304195432668,
+      "loss": 1.326,
+      "step": 1850
+    },
+    {
+      "epoch": 0.204961051875587,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00019336149785140525,
+      "loss": 1.3869,
+      "step": 1855
+    },
+    {
+      "epoch": 0.20551350754101982,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00019329222892313736,
+      "loss": 1.3094,
+      "step": 1860
+    },
+    {
+      "epoch": 0.20606596320645268,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00019322261301609286,
+      "loss": 1.3309,
+      "step": 1865
+    },
+    {
+      "epoch": 0.20661841887188553,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00019315265038919192,
+      "loss": 1.3306,
+      "step": 1870
+    },
+    {
+      "epoch": 0.2071708745373184,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00019308234130264431,
+      "loss": 1.3068,
+      "step": 1875
+    },
+    {
+      "epoch": 0.20772333020275122,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0001930116860179483,
+      "loss": 1.332,
+      "step": 1880
+    },
+    {
+      "epoch": 0.20827578586818407,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00019294068479788984,
+      "loss": 1.3728,
+      "step": 1885
+    },
+    {
+      "epoch": 0.20882824153361693,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00019286933790654148,
+      "loss": 1.3674,
+      "step": 1890
+    },
+    {
+      "epoch": 0.20938069719904978,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00019279764560926142,
+      "loss": 1.3572,
+      "step": 1895
+    },
+    {
+      "epoch": 0.20993315286448264,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00019272560817269247,
+      "loss": 1.4198,
+      "step": 1900
+    },
+    {
+      "epoch": 0.21048560852991546,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00019265322586476118,
+      "loss": 1.4785,
+      "step": 1905
+    },
+    {
+      "epoch": 0.21103806419534832,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0001925804989546767,
+      "loss": 1.3762,
+      "step": 1910
+    },
+    {
+      "epoch": 0.21159051986078117,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0001925074277129299,
+      "loss": 1.3075,
+      "step": 1915
+    },
+    {
+      "epoch": 0.21214297552621403,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001924340124112923,
+      "loss": 1.3402,
+      "step": 1920
+    },
+    {
+      "epoch": 0.21269543119164688,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00019236025332281507,
+      "loss": 1.3579,
+      "step": 1925
+    },
+    {
+      "epoch": 0.2132478868570797,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019228615072182796,
+      "loss": 1.3567,
+      "step": 1930
+    },
+    {
+      "epoch": 0.21380034252251257,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019221170488393843,
+      "loss": 1.3228,
+      "step": 1935
+    },
+    {
+      "epoch": 0.21435279818794542,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019213691608603047,
+      "loss": 1.3398,
+      "step": 1940
+    },
+    {
+      "epoch": 0.21490525385337828,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001920617846062636,
+      "loss": 1.2582,
+      "step": 1945
+    },
+    {
+      "epoch": 0.2154577095188111,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.000191986310724072,
+      "loss": 1.3183,
+      "step": 1950
+    },
+    {
+      "epoch": 0.21601016518424396,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00019191049472016313,
+      "loss": 1.2936,
+      "step": 1955
+    },
+    {
+      "epoch": 0.2165626208496768,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001918343368765171,
+      "loss": 1.3027,
+      "step": 1960
+    },
+    {
+      "epoch": 0.21711507651510967,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0001917578374763853,
+      "loss": 1.285,
+      "step": 1965
+    },
+    {
+      "epoch": 0.21766753218054252,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00019168099680428943,
+      "loss": 1.3343,
+      "step": 1970
+    },
+    {
+      "epoch": 0.21821998784597535,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001916038151460206,
+      "loss": 1.3197,
+      "step": 1975
+    },
+    {
+      "epoch": 0.2187724435114082,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.000191526292788638,
+      "loss": 1.3124,
+      "step": 1980
+    },
+    {
+      "epoch": 0.21932489917684106,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00019144843002046806,
+      "loss": 1.2425,
+      "step": 1985
+    },
+    {
+      "epoch": 0.21987735484227391,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00019137022713110324,
+      "loss": 1.2814,
+      "step": 1990
+    },
+    {
+      "epoch": 0.22042981050770677,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00019129168441140104,
+      "loss": 1.3209,
+      "step": 1995
+    },
+    {
+      "epoch": 0.2209822661731396,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019121280215348286,
+      "loss": 1.2622,
+      "step": 2000
+    },
+    {
+      "epoch": 0.22153472183857245,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00019113358065073297,
+      "loss": 1.248,
+      "step": 2005
+    },
+    {
+      "epoch": 0.2220871775040053,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00019105402019779728,
+      "loss": 1.3935,
+      "step": 2010
+    },
+    {
+      "epoch": 0.22263963316943816,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00019097412109058247,
+      "loss": 1.3131,
+      "step": 2015
+    },
+    {
+      "epoch": 0.223192088834871,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00019089388362625466,
+      "loss": 1.2621,
+      "step": 2020
+    },
+    {
+      "epoch": 0.22374454450030384,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00019081330810323852,
+      "loss": 1.3199,
+      "step": 2025
+    },
+    {
+      "epoch": 0.2242970001657367,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.000190732394821216,
+      "loss": 1.3163,
+      "step": 2030
+    },
+    {
+      "epoch": 0.22484945583116955,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00019065114408112517,
+      "loss": 1.3475,
+      "step": 2035
+    },
+    {
+      "epoch": 0.2254019114966024,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019056955618515934,
+      "loss": 1.2857,
+      "step": 2040
+    },
+    {
+      "epoch": 0.22595436716203524,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019048763143676578,
+      "loss": 1.3265,
+      "step": 2045
+    },
+    {
+      "epoch": 0.2265068228274681,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001904053701406445,
+      "loss": 1.3118,
+      "step": 2050
+    },
+    {
+      "epoch": 0.22705927849290095,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001903227726027473,
+      "loss": 1.2948,
+      "step": 2055
+    },
+    {
+      "epoch": 0.2276117341583338,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00019023983913027655,
+      "loss": 1.3374,
+      "step": 2060
+    },
+    {
+      "epoch": 0.22816418982376666,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00019015657003168405,
+      "loss": 1.4124,
+      "step": 2065
+    },
+    {
+      "epoch": 0.22871664548919948,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00019007296561666985,
+      "loss": 1.2983,
+      "step": 2070
+    },
+    {
+      "epoch": 0.22926910115463234,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00018998902619618116,
+      "loss": 1.3912,
+      "step": 2075
+    },
+    {
+      "epoch": 0.2298215568200652,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00018990475208241115,
+      "loss": 1.2852,
+      "step": 2080
+    },
+    {
+      "epoch": 0.23037401248549805,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0001898201435887978,
+      "loss": 1.3782,
+      "step": 2085
+    },
+    {
+      "epoch": 0.23092646815093087,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00018973520103002277,
+      "loss": 1.3141,
+      "step": 2090
+    },
+    {
+      "epoch": 0.23147892381636373,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0001896499247220102,
+      "loss": 1.2974,
+      "step": 2095
+    },
+    {
+      "epoch": 0.23203137948179658,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00018956431498192547,
+      "loss": 1.3403,
+      "step": 2100
+    },
+    {
+      "epoch": 0.23258383514722944,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00018947837212817415,
+      "loss": 1.3873,
+      "step": 2105
+    },
+    {
+      "epoch": 0.2331362908126623,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00018939209648040071,
+      "loss": 1.402,
+      "step": 2110
+    },
+    {
+      "epoch": 0.23368874647809512,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00018930548835948736,
+      "loss": 1.3259,
+      "step": 2115
+    },
+    {
+      "epoch": 0.23424120214352798,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00018921854808755294,
+      "loss": 1.3384,
+      "step": 2120
+    },
+    {
+      "epoch": 0.23479365780896083,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00018913127598795156,
+      "loss": 1.2929,
+      "step": 2125
+    },
+    {
+      "epoch": 0.2353461134743937,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00018904367238527155,
+      "loss": 1.4016,
+      "step": 2130
+    },
+    {
+      "epoch": 0.23589856913982654,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00018895573760533413,
+      "loss": 1.3384,
+      "step": 2135
+    },
+    {
+      "epoch": 0.23645102480525937,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00018886747197519233,
+      "loss": 1.3469,
+      "step": 2140
+    },
+    {
+      "epoch": 0.23700348047069222,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001887788758231296,
+      "loss": 1.3132,
+      "step": 2145
+    },
+    {
+      "epoch": 0.23755593613612508,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00018868994947865883,
+      "loss": 1.4067,
+      "step": 2150
+    },
+    {
+      "epoch": 0.23810839180155793,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00018860069327252086,
+      "loss": 1.3245,
+      "step": 2155
+    },
+    {
+      "epoch": 0.23866084746699076,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0001885111075366834,
+      "loss": 1.3196,
+      "step": 2160
+    },
+    {
+      "epoch": 0.23921330313242362,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00018842119260433982,
+      "loss": 1.329,
+      "step": 2165
+    },
+    {
+      "epoch": 0.23976575879785647,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001883309488099078,
+      "loss": 1.333,
+      "step": 2170
+    },
+    {
+      "epoch": 0.24031821446328933,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00018824037648902819,
+      "loss": 1.3282,
+      "step": 2175
+    },
+    {
+      "epoch": 0.24087067012872218,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00018814947597856367,
+      "loss": 1.2719,
+      "step": 2180
+    },
+    {
+      "epoch": 0.241423125794155,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00018805824761659764,
+      "loss": 1.3457,
+      "step": 2185
+    },
+    {
+      "epoch": 0.24197558145958786,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00018796669174243273,
+      "loss": 1.3,
+      "step": 2190
+    },
+    {
+      "epoch": 0.24252803712502072,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001878748086965898,
+      "loss": 1.2966,
+      "step": 2195
+    },
+    {
+      "epoch": 0.24308049279045357,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0001877825988208065,
+      "loss": 1.2928,
+      "step": 2200
+    },
+    {
+      "epoch": 0.24363294845588643,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00018769006245803596,
+      "loss": 1.2629,
+      "step": 2205
+    },
+    {
+      "epoch": 0.24418540412131925,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0001875971999524458,
+      "loss": 1.2908,
+      "step": 2210
+    },
+    {
+      "epoch": 0.2447378597867521,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0001875040116494165,
+      "loss": 1.2242,
+      "step": 2215
+    },
+    {
+      "epoch": 0.24529031545218496,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00018741049789554028,
+      "loss": 1.3844,
+      "step": 2220
+    },
+    {
+      "epoch": 0.24584277111761782,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00018731665903861985,
+      "loss": 1.2795,
+      "step": 2225
+    },
+    {
+      "epoch": 0.24639522678305065,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00018722249542766703,
+      "loss": 1.242,
+      "step": 2230
+    },
+    {
+      "epoch": 0.2469476824484835,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00018712800741290154,
+      "loss": 1.2709,
+      "step": 2235
+    },
+    {
+      "epoch": 0.24750013811391636,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00018703319534574956,
+      "loss": 1.2267,
+      "step": 2240
+    },
+    {
+      "epoch": 0.2480525937793492,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00018693805957884258,
+      "loss": 1.2533,
+      "step": 2245
+    },
+    {
+      "epoch": 0.24860504944478207,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00018684260046601594,
+      "loss": 1.3512,
+      "step": 2250
+    },
+    {
+      "epoch": 0.2491575051102149,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0001867468183623077,
+      "loss": 1.265,
+      "step": 2255
+    },
+    {
+      "epoch": 0.24970996077564775,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0001866507136239571,
+      "loss": 1.2883,
+      "step": 2260
+    },
+    {
+      "epoch": 0.2502624164410806,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00018655428660840345,
+      "loss": 1.2778,
+      "step": 2265
+    },
+    {
+      "epoch": 0.25081487210651343,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00018645753767428458,
+      "loss": 1.3121,
+      "step": 2270
+    },
+    {
+      "epoch": 0.2513673277719463,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0001863604671814357,
+      "loss": 1.3,
+      "step": 2275
+    },
+    {
+      "epoch": 0.25191978343737914,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00018626307549088792,
+      "loss": 1.2828,
+      "step": 2280
+    },
+    {
+      "epoch": 0.252472239102812,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0001861653629648671,
+      "loss": 1.3681,
+      "step": 2285
+    },
+    {
+      "epoch": 0.25302469476824485,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00018606732996679224,
+      "loss": 1.3393,
+      "step": 2290
+    },
+    {
+      "epoch": 0.2535771504336777,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00018596897686127428,
+      "loss": 1.3363,
+      "step": 2295
+    },
+    {
+      "epoch": 0.25412960609911056,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0001858703040141148,
+      "loss": 1.3374,
+      "step": 2300
+    },
+    {
+      "epoch": 0.2546820617645434,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00018577131179230448,
+      "loss": 1.2833,
+      "step": 2305
+    },
+    {
+      "epoch": 0.25523451742997627,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00018567200056402195,
+      "loss": 1.3367,
+      "step": 2310
+    },
+    {
+      "epoch": 0.2557869730954091,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00018557237069863222,
+      "loss": 1.277,
+      "step": 2315
+    },
+    {
+      "epoch": 0.2563394287608419,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00018547242256668548,
+      "loss": 1.3652,
+      "step": 2320
+    },
+    {
+      "epoch": 0.2568918844262748,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00018537215653991552,
+      "loss": 1.368,
+      "step": 2325
+    },
+    {
+      "epoch": 0.25744434009170764,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0001852715729912386,
+      "loss": 1.3541,
+      "step": 2330
+    },
+    {
+      "epoch": 0.25799679575714046,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00018517067229475184,
+      "loss": 1.3887,
+      "step": 2335
+    },
+    {
+      "epoch": 0.25854925142257335,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00018506945482573195,
+      "loss": 1.3053,
+      "step": 2340
+    },
+    {
+      "epoch": 0.2591017070880062,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0001849679209606338,
+      "loss": 1.3297,
+      "step": 2345
+    },
+    {
+      "epoch": 0.25965416275343906,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.000184866071077089,
+      "loss": 1.3666,
+      "step": 2350
+    },
+    {
+      "epoch": 0.2602066184188719,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00018476390555390457,
+      "loss": 1.292,
+      "step": 2355
+    },
+    {
+      "epoch": 0.2607590740843047,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0001846614247710614,
+      "loss": 1.3541,
+      "step": 2360
+    },
+    {
+      "epoch": 0.2613115297497376,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.000184558629109713,
+      "loss": 1.3721,
+      "step": 2365
+    },
+    {
+      "epoch": 0.2618639854151704,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00018445551895218394,
+      "loss": 1.2935,
+      "step": 2370
+    },
+    {
+      "epoch": 0.2624164410806033,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00018435209468196847,
+      "loss": 1.3716,
+      "step": 2375
+    },
+    {
+      "epoch": 0.26296889674603613,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00018424835668372919,
+      "loss": 1.3419,
+      "step": 2380
+    },
+    {
+      "epoch": 0.26352135241146896,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0001841443053432955,
+      "loss": 1.2971,
+      "step": 2385
+    },
+    {
+      "epoch": 0.26407380807690184,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00018403994104766212,
+      "loss": 1.3557,
+      "step": 2390
+    },
+    {
+      "epoch": 0.26462626374233467,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00018393526418498786,
+      "loss": 1.296,
+      "step": 2395
+    },
+    {
+      "epoch": 0.26517871940776755,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00018383027514459402,
+      "loss": 1.2831,
+      "step": 2400
+    },
+    {
+      "epoch": 0.2657311750732004,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00018372497431696288,
+      "loss": 1.316,
+      "step": 2405
+    },
+    {
+      "epoch": 0.2662836307386332,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00018361936209373644,
+      "loss": 1.2525,
+      "step": 2410
+    },
+    {
+      "epoch": 0.2668360864040661,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00018351343886771488,
+      "loss": 1.3607,
+      "step": 2415
+    },
+    {
+      "epoch": 0.2673885420694989,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00018340720503285497,
+      "loss": 1.3376,
+      "step": 2420
+    },
+    {
+      "epoch": 0.2679409977349318,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00018330066098426882,
+      "loss": 1.3497,
+      "step": 2425
+    },
+    {
+      "epoch": 0.2684934534003646,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00018319380711822225,
+      "loss": 1.3144,
+      "step": 2430
+    },
+    {
+      "epoch": 0.26904590906579745,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00018308664383213344,
+      "loss": 1.4368,
+      "step": 2435
+    },
+    {
+      "epoch": 0.26959836473123033,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00018297917152457126,
+      "loss": 1.3755,
+      "step": 2440
+    },
+    {
+      "epoch": 0.27015082039666316,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00018287139059525412,
+      "loss": 1.2929,
+      "step": 2445
+    },
+    {
+      "epoch": 0.27070327606209604,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00018276330144504803,
+      "loss": 1.3217,
+      "step": 2450
+    },
+    {
+      "epoch": 0.27125573172752887,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0001826549044759655,
+      "loss": 1.3161,
+      "step": 2455
+    },
+    {
+      "epoch": 0.2718081873929617,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00018254620009116397,
+      "loss": 1.2695,
+      "step": 2460
+    },
+    {
+      "epoch": 0.2723606430583946,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00018243718869494408,
+      "loss": 1.3496,
+      "step": 2465
+    },
+    {
+      "epoch": 0.2729130987238274,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001823278706927484,
+      "loss": 1.3295,
+      "step": 2470
+    },
+    {
+      "epoch": 0.27346555438926023,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00018221824649115984,
+      "loss": 1.3566,
+      "step": 2475
+    },
+    {
+      "epoch": 0.2740180100546931,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00018210831649790018,
+      "loss": 1.2789,
+      "step": 2480
+    },
+    {
+      "epoch": 0.27457046572012594,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00018199808112182847,
+      "loss": 1.2588,
+      "step": 2485
+    },
+    {
+      "epoch": 0.2751229213855588,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00018188754077293963,
+      "loss": 1.2579,
+      "step": 2490
+    },
+    {
+      "epoch": 0.27567537705099165,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00018177669586236277,
+      "loss": 1.2771,
+      "step": 2495
+    },
+    {
+      "epoch": 0.2762278327164245,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001816655468023598,
+      "loss": 1.3111,
+      "step": 2500
+    },
+    {
+      "epoch": 0.27678028838185736,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00018155409400632386,
+      "loss": 1.3309,
+      "step": 2505
+    },
+    {
+      "epoch": 0.2773327440472902,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001814423378887777,
+      "loss": 1.2619,
+      "step": 2510
+    },
+    {
+      "epoch": 0.2778851997127231,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00018133027886537225,
+      "loss": 1.2475,
+      "step": 2515
+    },
+    {
+      "epoch": 0.2784376553781559,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00018121791735288504,
+      "loss": 1.2622,
+      "step": 2520
+    },
+    {
+      "epoch": 0.27899011104358873,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00018110525376921862,
+      "loss": 1.2824,
+      "step": 2525
+    },
+    {
+      "epoch": 0.2795425667090216,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00018099228853339901,
+      "loss": 1.2897,
+      "step": 2530
+    },
+    {
+      "epoch": 0.28009502237445444,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00018087902206557411,
+      "loss": 1.4139,
+      "step": 2535
+    },
+    {
+      "epoch": 0.2806474780398873,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00018076545478701235,
+      "loss": 1.2527,
+      "step": 2540
+    },
+    {
+      "epoch": 0.28119993370532015,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00018065158712010076,
+      "loss": 1.2459,
+      "step": 2545
+    },
+    {
+      "epoch": 0.281752389370753,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0001805374194883437,
+      "loss": 1.4089,
+      "step": 2550
+    },
+    {
+      "epoch": 0.28230484503618586,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00018042295231636115,
+      "loss": 1.4094,
+      "step": 2555
+    },
+    {
+      "epoch": 0.2828573007016187,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0001803081860298872,
+      "loss": 1.3252,
+      "step": 2560
+    },
+    {
+      "epoch": 0.28340975636705157,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0001801931210557684,
+      "loss": 1.3043,
+      "step": 2565
+    },
+    {
+      "epoch": 0.2839622120324844,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00018007775782196214,
+      "loss": 1.2738,
+      "step": 2570
+    },
+    {
+      "epoch": 0.2845146676979172,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00017996209675753523,
+      "loss": 1.3325,
+      "step": 2575
+    },
+    {
+      "epoch": 0.2850671233633501,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0001798461382926621,
+      "loss": 1.2705,
+      "step": 2580
+    },
+    {
+      "epoch": 0.28561957902878293,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00017972988285862337,
+      "loss": 1.3578,
+      "step": 2585
+    },
+    {
+      "epoch": 0.2861720346942158,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00017961333088780404,
+      "loss": 1.2751,
+      "step": 2590
+    },
+    {
+      "epoch": 0.28672449035964864,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00017949648281369217,
+      "loss": 1.3509,
+      "step": 2595
+    },
+    {
+      "epoch": 0.28727694602508147,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00017937933907087703,
+      "loss": 1.2251,
+      "step": 2600
+    },
+    {
+      "epoch": 0.28782940169051435,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00017926190009504752,
+      "loss": 1.3057,
+      "step": 2605
+    },
+    {
+      "epoch": 0.2883818573559472,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00017914416632299065,
+      "loss": 1.2594,
+      "step": 2610
+    },
+    {
+      "epoch": 0.28893431302138,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00017902613819258985,
+      "loss": 1.2807,
+      "step": 2615
+    },
+    {
+      "epoch": 0.2894867686868129,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0001789078161428233,
+      "loss": 1.2231,
+      "step": 2620
+    },
+    {
+      "epoch": 0.2900392243522457,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00017878920061376247,
+      "loss": 1.2458,
+      "step": 2625
+    },
+    {
+      "epoch": 0.2905916800176786,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0001786702920465702,
+      "loss": 1.2969,
+      "step": 2630
+    },
+    {
+      "epoch": 0.2911441356831114,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00017855109088349926,
+      "loss": 1.2793,
+      "step": 2635
+    },
+    {
+      "epoch": 0.29169659134854425,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00017843159756789076,
+      "loss": 1.3146,
+      "step": 2640
+    },
+    {
+      "epoch": 0.29224904701397714,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00017831181254417228,
+      "loss": 1.3319,
+      "step": 2645
+    },
+    {
+      "epoch": 0.29280150267940996,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00017819173625785643,
+      "loss": 1.3275,
+      "step": 2650
+    },
+    {
+      "epoch": 0.29335395834484285,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00017807136915553903,
+      "loss": 1.3764,
+      "step": 2655
+    },
+    {
+      "epoch": 0.2939064140102757,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0001779507116848976,
+      "loss": 1.3263,
+      "step": 2660
+    },
+    {
+      "epoch": 0.2944588696757085,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00017782976429468956,
+      "loss": 1.2854,
+      "step": 2665
+    },
+    {
+      "epoch": 0.2950113253411414,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00017770852743475066,
+      "loss": 1.2836,
+      "step": 2670
+    },
+    {
+      "epoch": 0.2955637810065742,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00017758700155599317,
+      "loss": 1.3247,
+      "step": 2675
+    },
+    {
+      "epoch": 0.2961162366720071,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00017746518711040442,
+      "loss": 1.293,
+      "step": 2680
+    },
+    {
+      "epoch": 0.2966686923374399,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00017734308455104496,
+      "loss": 1.2325,
+      "step": 2685
+    },
+    {
+      "epoch": 0.29722114800287275,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00017722069433204687,
+      "loss": 1.3441,
+      "step": 2690
+    },
+    {
+      "epoch": 0.29777360366830563,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00017709801690861214,
+      "loss": 1.3705,
+      "step": 2695
+    },
+    {
+      "epoch": 0.29832605933373846,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00017697505273701097,
+      "loss": 1.2979,
+      "step": 2700
+    },
+    {
+      "epoch": 0.29887851499917134,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00017685180227458003,
+      "loss": 1.3489,
+      "step": 2705
+    },
+    {
+      "epoch": 0.29943097066460417,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001767282659797208,
+      "loss": 1.2235,
+      "step": 2710
+    },
+    {
+      "epoch": 0.299983426330037,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001766044443118978,
+      "loss": 1.3033,
+      "step": 2715
+    },
+    {
+      "epoch": 0.3005358819954699,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.000176480337731637,
+      "loss": 1.1975,
+      "step": 2720
+    },
+    {
+      "epoch": 0.3010883376609027,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.000176355946700524,
+      "loss": 1.2594,
+      "step": 2725
+    },
+    {
+      "epoch": 0.3016407933263356,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00017623127168120233,
+      "loss": 1.3387,
+      "step": 2730
+    },
+    {
+      "epoch": 0.3021932489917684,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00017610631313737173,
+      "loss": 1.3067,
+      "step": 2735
+    },
+    {
+      "epoch": 0.30274570465720124,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00017598107153378657,
+      "loss": 1.3715,
+      "step": 2740
+    },
+    {
+      "epoch": 0.3032981603226341,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00017585554733625384,
+      "loss": 1.3346,
+      "step": 2745
+    },
+    {
+      "epoch": 0.30385061598806695,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00017572974101163165,
+      "loss": 1.2732,
+      "step": 2750
+    },
+    {
+      "epoch": 0.3044030716534998,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00017560365302782738,
+      "loss": 1.3209,
+      "step": 2755
+    },
+    {
+      "epoch": 0.30495552731893266,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00017547728385379605,
+      "loss": 1.2767,
+      "step": 2760
+    },
+    {
+      "epoch": 0.3055079829843655,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0001753506339595384,
+      "loss": 1.3191,
+      "step": 2765
+    },
+    {
+      "epoch": 0.30606043864979837,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00017522370381609935,
+      "loss": 1.2933,
+      "step": 2770
+    },
+    {
+      "epoch": 0.3066128943152312,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00017509649389556607,
+      "loss": 1.3383,
+      "step": 2775
+    },
+    {
+      "epoch": 0.307165349980664,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00017496900467106627,
+      "loss": 1.2884,
+      "step": 2780
+    },
+    {
+      "epoch": 0.3077178056460969,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00017484123661676656,
+      "loss": 1.2966,
+      "step": 2785
+    },
+    {
+      "epoch": 0.30827026131152974,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0001747131902078705,
+      "loss": 1.2811,
+      "step": 2790
+    },
+    {
+      "epoch": 0.3088227169769626,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00017458486592061704,
+      "loss": 1.2886,
+      "step": 2795
+    },
+    {
+      "epoch": 0.30937517264239545,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00017445626423227844,
+      "loss": 1.2527,
+      "step": 2800
+    },
+    {
+      "epoch": 0.3099276283078283,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001743273856211589,
+      "loss": 1.2855,
+      "step": 2805
+    },
+    {
+      "epoch": 0.31048008397326116,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00017419823056659243,
+      "loss": 1.3601,
+      "step": 2810
+    },
+    {
+      "epoch": 0.311032539638694,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00017406879954894134,
+      "loss": 1.2841,
+      "step": 2815
+    },
+    {
+      "epoch": 0.31158499530412687,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00017393909304959414,
+      "loss": 1.3243,
+      "step": 2820
+    },
+    {
+      "epoch": 0.3121374509695597,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00017380911155096408,
+      "loss": 1.3383,
+      "step": 2825
+    },
+    {
+      "epoch": 0.3126899066349925,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00017367885553648717,
+      "loss": 1.3337,
+      "step": 2830
+    },
+    {
+      "epoch": 0.3132423623004254,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00017354832549062034,
+      "loss": 1.2695,
+      "step": 2835
+    },
+    {
+      "epoch": 0.31379481796585823,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00017341752189883983,
+      "loss": 1.3181,
+      "step": 2840
+    },
+    {
+      "epoch": 0.3143472736312911,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0001732864452476392,
+      "loss": 1.3033,
+      "step": 2845
+    },
+    {
+      "epoch": 0.31489972929672394,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001731550960245276,
+      "loss": 1.3264,
+      "step": 2850
+    },
+    {
+      "epoch": 0.31545218496215677,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00017302347471802798,
+      "loss": 1.317,
+      "step": 2855
+    },
+    {
+      "epoch": 0.31600464062758965,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00017289158181767517,
+      "loss": 1.2907,
+      "step": 2860
+    },
+    {
+      "epoch": 0.3165570962930225,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00017275941781401427,
+      "loss": 1.3268,
+      "step": 2865
+    },
+    {
+      "epoch": 0.31710955195845536,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00017262698319859846,
+      "loss": 1.4696,
+      "step": 2870
+    },
+    {
+      "epoch": 0.3176620076238882,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00017249427846398766,
+      "loss": 1.2852,
+      "step": 2875
+    },
+    {
+      "epoch": 0.318214463289321,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00017236130410374625,
+      "loss": 1.3287,
+      "step": 2880
+    },
+    {
+      "epoch": 0.3187669189547539,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0001722280606124415,
+      "loss": 1.2764,
+      "step": 2885
+    },
+    {
+      "epoch": 0.3193193746201867,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00017209454848564156,
+      "loss": 1.2118,
+      "step": 2890
+    },
+    {
+      "epoch": 0.31987183028561955,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00017196076821991384,
+      "loss": 1.3399,
+      "step": 2895
+    },
+    {
+      "epoch": 0.32042428595105243,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00017182672031282296,
+      "loss": 1.2984,
+      "step": 2900
+    },
+    {
+      "epoch": 0.32097674161648526,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00017169240526292896,
+      "loss": 1.435,
+      "step": 2905
+    },
+    {
+      "epoch": 0.32152919728191814,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0001715578235697855,
+      "loss": 1.2973,
+      "step": 2910
+    },
+    {
+      "epoch": 0.32208165294735097,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001714229757339379,
+      "loss": 1.3144,
+      "step": 2915
+    },
+    {
+      "epoch": 0.3226341086127838,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00017128786225692136,
+      "loss": 1.3097,
+      "step": 2920
+    },
+    {
+      "epoch": 0.3231865642782167,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00017115248364125906,
+      "loss": 1.2551,
+      "step": 2925
+    },
+    {
+      "epoch": 0.3237390199436495,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00017101684039046036,
+      "loss": 1.3567,
+      "step": 2930
+    },
+    {
+      "epoch": 0.3242914756090824,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00017088093300901883,
+      "loss": 1.394,
+      "step": 2935
+    },
+    {
+      "epoch": 0.3248439312745152,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00017074476200241035,
+      "loss": 1.2932,
+      "step": 2940
+    },
+    {
+      "epoch": 0.32539638693994805,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001706083278770914,
+      "loss": 1.315,
+      "step": 2945
+    },
+    {
+      "epoch": 0.32594884260538093,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00017047163114049702,
+      "loss": 1.3137,
+      "step": 2950
+    },
+    {
+      "epoch": 0.32650129827081376,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00017033467230103894,
+      "loss": 1.2952,
+      "step": 2955
+    },
+    {
+      "epoch": 0.32705375393624664,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00017019745186810378,
+      "loss": 1.3501,
+      "step": 2960
+    },
+    {
+      "epoch": 0.32760620960167947,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0001700599703520511,
+      "loss": 1.3543,
+      "step": 2965
+    },
+    {
+      "epoch": 0.3281586652671123,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00016992222826421133,
+      "loss": 1.2998,
+      "step": 2970
+    },
+    {
+      "epoch": 0.3287111209325452,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0001697842261168843,
+      "loss": 1.3317,
+      "step": 2975
+    },
+    {
+      "epoch": 0.329263576597978,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00016964596442333696,
+      "loss": 1.218,
+      "step": 2980
+    },
+    {
+      "epoch": 0.3298160322634109,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00016950744369780148,
+      "loss": 1.2898,
+      "step": 2985
+    },
+    {
+      "epoch": 0.3303684879288437,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00016936866445547353,
+      "loss": 1.2866,
+      "step": 2990
+    },
+    {
+      "epoch": 0.33092094359427654,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00016922962721251038,
+      "loss": 1.3173,
+      "step": 2995
+    },
+    {
+      "epoch": 0.3314733992597094,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001690903324860286,
+      "loss": 1.2364,
+      "step": 3000
+    },
+    {
+      "epoch": 0.33202585492514225,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0001689507807941027,
+      "loss": 1.3763,
+      "step": 3005
+    },
+    {
+      "epoch": 0.33257831059057513,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0001688109726557627,
+      "loss": 1.2882,
+      "step": 3010
+    },
+    {
+      "epoch": 0.33313076625600796,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00016867090859099256,
+      "loss": 1.2957,
+      "step": 3015
+    },
+    {
+      "epoch": 0.3336832219214408,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00016853058912072802,
+      "loss": 1.2837,
+      "step": 3020
+    },
+    {
+      "epoch": 0.33423567758687367,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0001683900147668547,
+      "loss": 1.3043,
+      "step": 3025
+    },
+    {
+      "epoch": 0.3347881332523065,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0001682491860522063,
+      "loss": 1.2826,
+      "step": 3030
+    },
+    {
+      "epoch": 0.3353405889177393,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001681081035005626,
+      "loss": 1.339,
+      "step": 3035
+    },
+    {
+      "epoch": 0.3358930445831722,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00016796676763664725,
+      "loss": 1.3558,
+      "step": 3040
+    },
+    {
+      "epoch": 0.33644550024860503,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00016782517898612619,
+      "loss": 1.2467,
+      "step": 3045
+    },
+    {
+      "epoch": 0.3369979559140379,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00016768333807560558,
+      "loss": 1.2127,
+      "step": 3050
+    },
+    {
+      "epoch": 0.33755041157947074,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00016754124543262973,
+      "loss": 1.3179,
+      "step": 3055
+    },
+    {
+      "epoch": 0.33810286724490357,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00016739890158567916,
+      "loss": 1.2387,
+      "step": 3060
+    },
+    {
+      "epoch": 0.33865532291033645,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001672563070641688,
+      "loss": 1.3292,
+      "step": 3065
+    },
+    {
+      "epoch": 0.3392077785757693,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00016711346239844588,
+      "loss": 1.3224,
+      "step": 3070
+    },
+    {
+      "epoch": 0.33976023424120216,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00016697036811978786,
+      "loss": 1.2961,
+      "step": 3075
+    },
+    {
+      "epoch": 0.340312689906635,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00016682702476040077,
+      "loss": 1.3446,
+      "step": 3080
+    },
+    {
+      "epoch": 0.3408651455720678,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00016668343285341686,
+      "loss": 1.33,
+      "step": 3085
+    },
+    {
+      "epoch": 0.3414176012375007,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00016653959293289297,
+      "loss": 1.3379,
+      "step": 3090
+    },
+    {
+      "epoch": 0.34197005690293353,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00016639550553380818,
+      "loss": 1.3434,
+      "step": 3095
+    },
+    {
+      "epoch": 0.3425225125683664,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00016625117119206214,
+      "loss": 1.3041,
+      "step": 3100
+    },
+    {
+      "epoch": 0.34307496823379924,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00016610659044447298,
+      "loss": 1.2399,
+      "step": 3105
+    },
+    {
+      "epoch": 0.34362742389923207,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00016596176382877506,
+      "loss": 1.313,
+      "step": 3110
+    },
+    {
+      "epoch": 0.34417987956466495,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00016581669188361748,
+      "loss": 1.2899,
+      "step": 3115
+    },
+    {
+      "epoch": 0.3447323352300978,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00016567137514856154,
+      "loss": 1.3798,
+      "step": 3120
+    },
+    {
+      "epoch": 0.34528479089553066,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00016552581416407917,
+      "loss": 1.3817,
+      "step": 3125
+    },
+    {
+      "epoch": 0.3458372465609635,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00016538000947155062,
+      "loss": 1.2274,
+      "step": 3130
+    },
+    {
+      "epoch": 0.3463897022263963,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0001652339616132625,
+      "loss": 1.266,
+      "step": 3135
+    },
+    {
+      "epoch": 0.3469421578918292,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00016508767113240598,
+      "loss": 1.3188,
+      "step": 3140
+    },
+    {
+      "epoch": 0.347494613557262,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00016494113857307453,
+      "loss": 1.2402,
+      "step": 3145
+    },
+    {
+      "epoch": 0.3480470692226949,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00016479436448026195,
+      "loss": 1.291,
+      "step": 3150
+    },
+    {
+      "epoch": 0.34859952488812773,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00016464734939986036,
+      "loss": 1.2659,
+      "step": 3155
+    },
+    {
+      "epoch": 0.34915198055356056,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00016450009387865822,
+      "loss": 1.3496,
+      "step": 3160
+    },
+    {
+      "epoch": 0.34970443621899344,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00016435259846433824,
+      "loss": 1.2812,
+      "step": 3165
+    },
+    {
+      "epoch": 0.35025689188442627,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00016420486370547537,
+      "loss": 1.2947,
+      "step": 3170
+    },
+    {
+      "epoch": 0.3508093475498591,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00016405689015153472,
+      "loss": 1.2675,
+      "step": 3175
+    },
+    {
+      "epoch": 0.351361803215292,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00016390867835286953,
+      "loss": 1.3378,
+      "step": 3180
+    },
+    {
+      "epoch": 0.3519142588807248,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0001637602288607192,
+      "loss": 1.2887,
+      "step": 3185
+    },
+    {
+      "epoch": 0.3524667145461577,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00016361154222720715,
+      "loss": 1.2812,
+      "step": 3190
+    },
+    {
+      "epoch": 0.3530191702115905,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00016346261900533867,
+      "loss": 1.3614,
+      "step": 3195
+    },
+    {
+      "epoch": 0.35357162587702334,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00016331345974899923,
+      "loss": 1.2825,
+      "step": 3200
+    },
+    {
+      "epoch": 0.3541240815424562,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00016316406501295198,
+      "loss": 1.2426,
+      "step": 3205
+    },
+    {
+      "epoch": 0.35467653720788905,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0001630144353528359,
+      "loss": 1.2618,
+      "step": 3210
+    },
+    {
+      "epoch": 0.35522899287332194,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00016286457132516383,
+      "loss": 1.3534,
+      "step": 3215
+    },
+    {
+      "epoch": 0.35578144853875476,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0001627144734873202,
+      "loss": 1.2713,
+      "step": 3220
+    },
+    {
+      "epoch": 0.3563339042041876,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00016256414239755902,
+      "loss": 1.2645,
+      "step": 3225
+    },
+    {
+      "epoch": 0.3568863598696205,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00016241357861500184,
+      "loss": 1.2915,
+      "step": 3230
+    },
+    {
+      "epoch": 0.3574388155350533,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00016226278269963578,
+      "loss": 1.3063,
+      "step": 3235
+    },
+    {
+      "epoch": 0.3579912712004862,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00016211175521231108,
+      "loss": 1.2619,
+      "step": 3240
+    },
+    {
+      "epoch": 0.358543726865919,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00016196049671473954,
+      "loss": 1.2738,
+      "step": 3245
+    },
+    {
+      "epoch": 0.35909618253135184,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0001618090077694919,
+      "loss": 1.2864,
+      "step": 3250
+    },
+    {
+      "epoch": 0.3596486381967847,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00016165728893999617,
+      "loss": 1.3055,
+      "step": 3255
+    },
+    {
+      "epoch": 0.36020109386221755,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00016150534079053527,
+      "loss": 1.2808,
+      "step": 3260
+    },
+    {
+      "epoch": 0.36075354952765043,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00016135316388624505,
+      "loss": 1.3164,
+      "step": 3265
+    },
+    {
+      "epoch": 0.36130600519308326,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0001612007587931122,
+      "loss": 1.3134,
+      "step": 3270
+    },
+    {
+      "epoch": 0.3618584608585161,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00016104812607797202,
+      "loss": 1.2804,
+      "step": 3275
+    },
+    {
+      "epoch": 0.36241091652394897,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00016089526630850643,
+      "loss": 1.2536,
+      "step": 3280
+    },
+    {
+      "epoch": 0.3629633721893818,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0001607421800532419,
+      "loss": 1.2638,
+      "step": 3285
+    },
+    {
+      "epoch": 0.3635158278548147,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00016058886788154712,
+      "loss": 1.2628,
+      "step": 3290
+    },
+    {
+      "epoch": 0.3640682835202475,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00016043533036363115,
+      "loss": 1.2148,
+      "step": 3295
+    },
+    {
+      "epoch": 0.36462073918568033,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00016028156807054112,
+      "loss": 1.3718,
+      "step": 3300
+    },
+    {
+      "epoch": 0.3651731948511132,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0001601275815741602,
+      "loss": 1.2852,
+      "step": 3305
+    },
+    {
+      "epoch": 0.36572565051654604,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00015997337144720532,
+      "loss": 1.3223,
+      "step": 3310
+    },
+    {
+      "epoch": 0.36627810618197887,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00015981893826322527,
+      "loss": 1.2156,
+      "step": 3315
+    },
+    {
+      "epoch": 0.36683056184741175,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00015966428259659845,
+      "loss": 1.3567,
+      "step": 3320
+    },
+    {
+      "epoch": 0.3673830175128446,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00015950940502253063,
+      "loss": 1.2521,
+      "step": 3325
+    },
+    {
+      "epoch": 0.36793547317827746,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00015935430611705296,
+      "loss": 1.2567,
+      "step": 3330
+    },
+    {
+      "epoch": 0.3684879288437103,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0001591989864570199,
+      "loss": 1.3041,
+      "step": 3335
+    },
+    {
+      "epoch": 0.3690403845091431,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00015904344662010672,
+      "loss": 1.3083,
+      "step": 3340
+    },
+    {
+      "epoch": 0.369592840174576,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00015888768718480778,
+      "loss": 1.319,
+      "step": 3345
+    },
+    {
+      "epoch": 0.3701452958400088,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00015873170873043413,
+      "loss": 1.2914,
+      "step": 3350
+    },
+    {
+      "epoch": 0.3706977515054417,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00015857551183711137,
+      "loss": 1.3141,
+      "step": 3355
+    },
+    {
+      "epoch": 0.37125020717087454,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0001584190970857776,
+      "loss": 1.3307,
+      "step": 3360
+    },
+    {
+      "epoch": 0.37180266283630736,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00015826246505818112,
+      "loss": 1.3437,
+      "step": 3365
+    },
+    {
+      "epoch": 0.37235511850174025,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00015810561633687842,
+      "loss": 1.282,
+      "step": 3370
+    },
+    {
+      "epoch": 0.3729075741671731,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00015794855150523182,
+      "loss": 1.2792,
+      "step": 3375
+    },
+    {
+      "epoch": 0.37346002983260596,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00015779127114740757,
+      "loss": 1.301,
+      "step": 3380
+    },
+    {
+      "epoch": 0.3740124854980388,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00015763377584837335,
+      "loss": 1.3113,
+      "step": 3385
+    },
+    {
+      "epoch": 0.3745649411634716,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0001574760661938964,
+      "loss": 1.2551,
+      "step": 3390
+    },
+    {
+      "epoch": 0.3751173968289045,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00015731814277054112,
+      "loss": 1.2969,
+      "step": 3395
+    },
+    {
+      "epoch": 0.3756698524943373,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00015716000616566698,
+      "loss": 1.3498,
+      "step": 3400
+    },
+    {
+      "epoch": 0.3762223081597702,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001570016569674264,
+      "loss": 1.3202,
+      "step": 3405
+    },
+    {
+      "epoch": 0.37677476382520303,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00015684309576476246,
+      "loss": 1.3306,
+      "step": 3410
+    },
+    {
+      "epoch": 0.37732721949063586,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00015668432314740663,
+      "loss": 1.2447,
+      "step": 3415
+    },
+    {
+      "epoch": 0.37787967515606874,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00015652533970587687,
+      "loss": 1.3604,
+      "step": 3420
+    },
+    {
+      "epoch": 0.37843213082150157,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00015636614603147512,
+      "loss": 1.2636,
+      "step": 3425
+    },
+    {
+      "epoch": 0.37898458648693445,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0001562067427162853,
+      "loss": 1.2856,
+      "step": 3430
+    },
+    {
+      "epoch": 0.3795370421523673,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00015604713035317097,
+      "loss": 1.2489,
+      "step": 3435
+    },
+    {
+      "epoch": 0.3800894978178001,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00015588730953577335,
+      "loss": 1.3487,
+      "step": 3440
+    },
+    {
+      "epoch": 0.380641953483233,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0001557272808585087,
+      "loss": 1.3533,
+      "step": 3445
+    },
+    {
+      "epoch": 0.3811944091486658,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00015556704491656665,
+      "loss": 1.3122,
+      "step": 3450
+    },
+    {
+      "epoch": 0.38174686481409864,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00015540660230590748,
+      "loss": 1.2188,
+      "step": 3455
+    },
+    {
+      "epoch": 0.3822993204795315,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00015524595362326025,
+      "loss": 1.2876,
+      "step": 3460
+    },
+    {
+      "epoch": 0.38285177614496435,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00015508509946612044,
+      "loss": 1.2888,
+      "step": 3465
+    },
+    {
+      "epoch": 0.38340423181039723,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0001549240404327477,
+      "loss": 1.2401,
+      "step": 3470
+    },
+    {
+      "epoch": 0.38395668747583006,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00015476277712216365,
+      "loss": 1.3932,
+      "step": 3475
+    },
+    {
+      "epoch": 0.3845091431412629,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00015460131013414979,
+      "loss": 1.3198,
+      "step": 3480
+    },
+    {
+      "epoch": 0.38506159880669577,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00015443964006924509,
+      "loss": 1.2714,
+      "step": 3485
+    },
+    {
+      "epoch": 0.3856140544721286,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00015427776752874371,
+      "loss": 1.3115,
+      "step": 3490
+    },
+    {
+      "epoch": 0.3861665101375615,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00015411569311469308,
+      "loss": 1.317,
+      "step": 3495
+    },
+    {
+      "epoch": 0.3867189658029943,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00015395341742989124,
+      "loss": 1.2624,
+      "step": 3500
+    },
+    {
+      "epoch": 0.38727142146842713,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00015379094107788497,
+      "loss": 1.3209,
+      "step": 3505
+    },
+    {
+      "epoch": 0.38782387713386,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00015362826466296732,
+      "loss": 1.2344,
+      "step": 3510
+    },
+    {
+      "epoch": 0.38837633279929284,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0001534653887901754,
+      "loss": 1.3466,
+      "step": 3515
+    },
+    {
+      "epoch": 0.3889287884647257,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0001533023140652882,
+      "loss": 1.3633,
+      "step": 3520
+    },
+    {
+      "epoch": 0.38948124413015855,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00015313904109482432,
+      "loss": 1.2775,
+      "step": 3525
+    },
+    {
+      "epoch": 0.3900336997955914,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0001529755704860396,
+      "loss": 1.2686,
+      "step": 3530
+    },
+    {
+      "epoch": 0.39058615546102426,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.000152811902846925,
+      "loss": 1.3165,
+      "step": 3535
+    },
+    {
+      "epoch": 0.3911386111264571,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0001526480387862043,
+      "loss": 1.2956,
+      "step": 3540
+    },
+    {
+      "epoch": 0.39169106679189,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00015248397891333185,
+      "loss": 1.3655,
+      "step": 3545
+    },
+    {
+      "epoch": 0.3922435224573228,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00015231972383849017,
+      "loss": 1.2996,
+      "step": 3550
+    },
+    {
+      "epoch": 0.39279597812275563,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00015215527417258794,
+      "loss": 1.2721,
+      "step": 3555
+    },
+    {
+      "epoch": 0.3933484337881885,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00015199063052725745,
+      "loss": 1.2413,
+      "step": 3560
+    },
+    {
+      "epoch": 0.39390088945362134,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00015182579351485248,
+      "loss": 1.2568,
+      "step": 3565
+    },
+    {
+      "epoch": 0.3944533451190542,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00015166076374844605,
+      "loss": 1.2275,
+      "step": 3570
+    },
+    {
+      "epoch": 0.39500580078448705,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00015149554184182802,
+      "loss": 1.3508,
+      "step": 3575
+    },
+    {
+      "epoch": 0.3955582564499199,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00015133012840950292,
+      "loss": 1.27,
+      "step": 3580
+    },
+    {
+      "epoch": 0.39611071211535276,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00015116452406668758,
+      "loss": 1.2897,
+      "step": 3585
+    },
+    {
+      "epoch": 0.3966631677807856,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00015099872942930887,
+      "loss": 1.2945,
+      "step": 3590
+    },
+    {
+      "epoch": 0.3972156234462184,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00015083274511400142,
+      "loss": 1.2058,
+      "step": 3595
+    },
+    {
+      "epoch": 0.3977680791116513,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001506665717381054,
+      "loss": 1.2996,
+      "step": 3600
+    },
+    {
+      "epoch": 0.3983205347770841,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00015050020991966406,
+      "loss": 1.2427,
+      "step": 3605
+    },
+    {
+      "epoch": 0.398872990442517,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00015033366027742155,
+      "loss": 1.2581,
+      "step": 3610
+    },
+    {
+      "epoch": 0.39942544610794983,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00015016692343082052,
+      "loss": 1.2623,
+      "step": 3615
+    },
+    {
+      "epoch": 0.39997790177338266,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 1.2438,
+      "step": 3620
+    },
+    {
+      "epoch": 0.40053035743881554,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00014983289060579294,
+      "loss": 1.3467,
+      "step": 3625
+    },
+    {
+      "epoch": 0.40108281310424837,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00014966559586972387,
+      "loss": 1.3014,
+      "step": 3630
+    },
+    {
+      "epoch": 0.40163526876968125,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0001494981164140067,
+      "loss": 1.3843,
+      "step": 3635
+    },
+    {
+      "epoch": 0.4021877244351141,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001493304528615424,
+      "loss": 1.3016,
+      "step": 3640
+    },
+    {
+      "epoch": 0.4027401801005469,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00014916260583591658,
+      "loss": 1.2421,
+      "step": 3645
+    },
+    {
+      "epoch": 0.4032926357659798,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00014899457596139729,
+      "loss": 1.3111,
+      "step": 3650
+    },
+    {
+      "epoch": 0.4038450914314126,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0001488263638629326,
+      "loss": 1.3026,
+      "step": 3655
+    },
+    {
+      "epoch": 0.4043975470968455,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00014865797016614838,
+      "loss": 1.3204,
+      "step": 3660
+    },
+    {
+      "epoch": 0.4049500027622783,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001484893954973458,
+      "loss": 1.3437,
+      "step": 3665
+    },
+    {
+      "epoch": 0.40550245842771115,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00014832064048349926,
+      "loss": 1.3389,
+      "step": 3670
+    },
+    {
+      "epoch": 0.40605491409314404,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00014815170575225382,
+      "loss": 1.2688,
+      "step": 3675
+    },
+    {
+      "epoch": 0.40660736975857686,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00014798259193192297,
+      "loss": 1.2756,
+      "step": 3680
+    },
+    {
+      "epoch": 0.40715982542400975,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00014781329965148624,
+      "loss": 1.267,
+      "step": 3685
+    },
+    {
+      "epoch": 0.4077122810894426,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.000147643829540587,
+      "loss": 1.2677,
+      "step": 3690
+    },
+    {
+      "epoch": 0.4082647367548754,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00014747418222952995,
+      "loss": 1.2098,
+      "step": 3695
+    },
+    {
+      "epoch": 0.4088171924203083,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00014730435834927884,
+      "loss": 1.2054,
+      "step": 3700
+    },
+    {
+      "epoch": 0.4093696480857411,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001471343585314542,
+      "loss": 1.2668,
+      "step": 3705
+    },
+    {
+      "epoch": 0.409922103751174,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001469641834083308,
+      "loss": 1.2696,
+      "step": 3710
+    },
+    {
+      "epoch": 0.4104745594166068,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00014679383361283554,
+      "loss": 1.2778,
+      "step": 3715
+    },
+    {
+      "epoch": 0.41102701508203965,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00014662330977854488,
+      "loss": 1.3343,
+      "step": 3720
+    },
+    {
+      "epoch": 0.41157947074747253,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00014645261253968262,
+      "loss": 1.2976,
+      "step": 3725
+    },
+    {
+      "epoch": 0.41213192641290536,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00014628174253111752,
+      "loss": 1.2054,
+      "step": 3730
+    },
+    {
+      "epoch": 0.41268438207833824,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00014611070038836083,
+      "loss": 1.2902,
+      "step": 3735
+    },
+    {
+      "epoch": 0.41323683774377107,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00014593948674756417,
+      "loss": 1.2411,
+      "step": 3740
+    },
+    {
+      "epoch": 0.4137892934092039,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00014576810224551683,
+      "loss": 1.2742,
+      "step": 3745
+    },
+    {
+      "epoch": 0.4143417490746368,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00014559654751964364,
+      "loss": 1.3241,
+      "step": 3750
+    },
+    {
+      "epoch": 0.4148942047400696,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00014542482320800264,
+      "loss": 1.3757,
+      "step": 3755
+    },
+    {
+      "epoch": 0.41544666040550243,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00014525292994928247,
+      "loss": 1.3469,
+      "step": 3760
+    },
+    {
+      "epoch": 0.4159991160709353,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00014508086838280017,
+      "loss": 1.354,
+      "step": 3765
+    },
+    {
+      "epoch": 0.41655157173636814,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0001449086391484988,
+      "loss": 1.3268,
+      "step": 3770
+    },
+    {
+      "epoch": 0.417104027401801,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00014473624288694498,
+      "loss": 1.3431,
+      "step": 3775
+    },
+    {
+      "epoch": 0.41765648306723385,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00014456368023932657,
+      "loss": 1.2649,
+      "step": 3780
+    },
+    {
+      "epoch": 0.4182089387326667,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00014439095184745024,
+      "loss": 1.2472,
+      "step": 3785
+    },
+    {
+      "epoch": 0.41876139439809956,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00014421805835373915,
+      "loss": 1.3079,
+      "step": 3790
+    },
+    {
+      "epoch": 0.4193138500635324,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0001440450004012305,
+      "loss": 1.3912,
+      "step": 3795
+    },
+    {
+      "epoch": 0.4198663057289653,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00014387177863357307,
+      "loss": 1.2799,
+      "step": 3800
+    },
+    {
+      "epoch": 0.4204187613943981,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00014369839369502506,
+      "loss": 1.2929,
+      "step": 3805
+    },
+    {
+      "epoch": 0.4209712170598309,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00014352484623045148,
+      "loss": 1.3207,
+      "step": 3810
+    },
+    {
+      "epoch": 0.4215236727252638,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00014335113688532182,
+      "loss": 1.3164,
+      "step": 3815
+    },
+    {
+      "epoch": 0.42207612839069664,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0001431772663057076,
+      "loss": 1.2804,
+      "step": 3820
+    },
+    {
+      "epoch": 0.4226285840561295,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00014300323513828008,
+      "loss": 1.3599,
+      "step": 3825
+    },
+    {
+      "epoch": 0.42318103972156235,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00014282904403030772,
+      "loss": 1.2706,
+      "step": 3830
+    },
+    {
+      "epoch": 0.4237334953869952,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.000142654693629654,
+      "loss": 1.3047,
+      "step": 3835
+    },
+    {
+      "epoch": 0.42428595105242806,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00014248018458477463,
+      "loss": 1.2786,
+      "step": 3840
+    },
+    {
+      "epoch": 0.4248384067178609,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00014230551754471554,
+      "loss": 1.3008,
+      "step": 3845
+    },
+    {
+      "epoch": 0.42539086238329377,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00014213069315911013,
+      "loss": 1.3216,
+      "step": 3850
+    },
+    {
+      "epoch": 0.4259433180487266,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0001419557120781772,
+      "loss": 1.278,
+      "step": 3855
+    },
+    {
+      "epoch": 0.4264957737141594,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00014178057495271815,
+      "loss": 1.2674,
+      "step": 3860
+    },
+    {
+      "epoch": 0.4270482293795923,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00014160528243411494,
+      "loss": 1.3296,
+      "step": 3865
+    },
+    {
+      "epoch": 0.42760068504502513,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00014142983517432723,
+      "loss": 1.2329,
+      "step": 3870
+    },
+    {
+      "epoch": 0.428153140710458,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00014125423382589048,
+      "loss": 1.3433,
+      "step": 3875
+    },
+    {
+      "epoch": 0.42870559637589084,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0001410784790419131,
+      "loss": 1.2509,
+      "step": 3880
+    },
+    {
+      "epoch": 0.42925805204132367,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00014090257147607413,
+      "loss": 1.2614,
+      "step": 3885
+    },
+    {
+      "epoch": 0.42981050770675655,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00014072651178262096,
+      "loss": 1.3318,
+      "step": 3890
+    },
+    {
+      "epoch": 0.4303629633721894,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00014055030061636668,
+      "loss": 1.2845,
+      "step": 3895
+    },
+    {
+      "epoch": 0.4309154190376222,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00014037393863268783,
+      "loss": 1.2372,
+      "step": 3900
+    },
+    {
+      "epoch": 0.4314678747030551,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00014019742648752184,
+      "loss": 1.2611,
+      "step": 3905
+    },
+    {
+      "epoch": 0.4320203303684879,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001400207648373646,
+      "loss": 1.3005,
+      "step": 3910
+    },
+    {
+      "epoch": 0.4325727860339208,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00013984395433926816,
+      "loss": 1.3007,
+      "step": 3915
+    },
+    {
+      "epoch": 0.4331252416993536,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00013966699565083802,
+      "loss": 1.2965,
+      "step": 3920
+    },
+    {
+      "epoch": 0.43367769736478645,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00013948988943023096,
+      "loss": 1.3147,
+      "step": 3925
+    },
+    {
+      "epoch": 0.43423015303021933,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00013931263633615241,
+      "loss": 1.3699,
+      "step": 3930
+    },
+    {
+      "epoch": 0.43478260869565216,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0001391352370278541,
+      "loss": 1.3404,
+      "step": 3935
+    },
+    {
+      "epoch": 0.43533506436108504,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00013895769216513157,
+      "loss": 1.3351,
+      "step": 3940
+    },
+    {
+      "epoch": 0.43588752002651787,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00013878000240832167,
+      "loss": 1.2524,
+      "step": 3945
+    },
+    {
+      "epoch": 0.4364399756919507,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00013860216841830018,
+      "loss": 1.2973,
+      "step": 3950
+    },
+    {
+      "epoch": 0.4369924313573836,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00013842419085647933,
+      "loss": 1.3037,
+      "step": 3955
+    },
+    {
+      "epoch": 0.4375448870228164,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00013824607038480532,
+      "loss": 1.2613,
+      "step": 3960
+    },
+    {
+      "epoch": 0.4380973426882493,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00013806780766575588,
+      "loss": 1.2013,
+      "step": 3965
+    },
+    {
+      "epoch": 0.4386497983536821,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001378894033623378,
+      "loss": 1.3508,
+      "step": 3970
+    },
+    {
+      "epoch": 0.43920225401911495,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00013771085813808442,
+      "loss": 1.3022,
+      "step": 3975
+    },
+    {
+      "epoch": 0.43975470968454783,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00013753217265705323,
+      "loss": 1.3253,
+      "step": 3980
+    },
+    {
+      "epoch": 0.44030716534998066,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0001373533475838234,
+      "loss": 1.2449,
+      "step": 3985
+    },
+    {
+      "epoch": 0.44085962101541354,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0001371743835834932,
+      "loss": 1.2731,
+      "step": 3990
+    },
+    {
+      "epoch": 0.44141207668084637,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00013699528132167776,
+      "loss": 1.3044,
+      "step": 3995
+    },
+    {
+      "epoch": 0.4419645323462792,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00013681604146450625,
+      "loss": 1.2405,
+      "step": 4000
+    },
+    {
+      "epoch": 0.4425169880117121,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00013663666467861972,
+      "loss": 1.3248,
+      "step": 4005
+    },
+    {
+      "epoch": 0.4430694436771449,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00013645715163116846,
+      "loss": 1.2828,
+      "step": 4010
+    },
+    {
+      "epoch": 0.4436218993425778,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0001362775029898096,
+      "loss": 1.3126,
+      "step": 4015
+    },
+    {
+      "epoch": 0.4441743550080106,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00013609771942270444,
+      "loss": 1.2572,
+      "step": 4020
+    },
+    {
+      "epoch": 0.44472681067344344,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001359178015985163,
+      "loss": 1.3412,
+      "step": 4025
+    },
+    {
+      "epoch": 0.4452792663388763,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00013573775018640766,
+      "loss": 1.2685,
+      "step": 4030
+    },
+    {
+      "epoch": 0.44583172200430915,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00013555756585603793,
+      "loss": 1.3055,
+      "step": 4035
+    },
+    {
+      "epoch": 0.446384177669742,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00013537724927756094,
+      "loss": 1.2442,
+      "step": 4040
+    },
+    {
+      "epoch": 0.44693663333517486,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001351968011216223,
+      "loss": 1.3003,
+      "step": 4045
+    },
+    {
+      "epoch": 0.4474890890006077,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00013501622205935697,
+      "loss": 1.3651,
+      "step": 4050
+    },
+    {
+      "epoch": 0.44804154466604057,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0001348355127623869,
+      "loss": 1.2699,
+      "step": 4055
+    },
+    {
+      "epoch": 0.4485940003314734,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00013465467390281826,
+      "loss": 1.2554,
+      "step": 4060
+    },
+    {
+      "epoch": 0.4491464559969062,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00013447370615323923,
+      "loss": 1.2467,
+      "step": 4065
+    },
+    {
+      "epoch": 0.4496989116623391,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00013429261018671734,
+      "loss": 1.298,
+      "step": 4070
+    },
+    {
+      "epoch": 0.45025136732777193,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00013411138667679696,
+      "loss": 1.1604,
+      "step": 4075
+    },
+    {
+      "epoch": 0.4508038229932048,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00013393003629749684,
+      "loss": 1.2926,
+      "step": 4080
+    },
+    {
+      "epoch": 0.45135627865863764,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00013374855972330757,
+      "loss": 1.2703,
+      "step": 4085
+    },
+    {
+      "epoch": 0.45190873432407047,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00013356695762918914,
+      "loss": 1.3145,
+      "step": 4090
+    },
+    {
+      "epoch": 0.45246118998950335,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001333852306905684,
+      "loss": 1.3115,
+      "step": 4095
+    },
+    {
+      "epoch": 0.4530136456549362,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0001332033795833364,
+      "loss": 1.3531,
+      "step": 4100
+    },
+    {
+      "epoch": 0.45356610132036906,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00013302140498384617,
+      "loss": 1.3463,
+      "step": 4105
+    },
+    {
+      "epoch": 0.4541185569858019,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0001328393075689099,
+      "loss": 1.2554,
+      "step": 4110
+    },
+    {
+      "epoch": 0.4546710126512347,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001326570880157967,
+      "loss": 1.2158,
+      "step": 4115
+    },
+    {
+      "epoch": 0.4552234683166676,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001324747470022298,
+      "loss": 1.3134,
+      "step": 4120
+    },
+    {
+      "epoch": 0.45577592398210043,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00013229228520638436,
+      "loss": 1.3496,
+      "step": 4125
+    },
+    {
+      "epoch": 0.4563283796475333,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00013210970330688454,
+      "loss": 1.3995,
+      "step": 4130
+    },
+    {
+      "epoch": 0.45688083531296614,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001319270019828013,
+      "loss": 1.3097,
+      "step": 4135
+    },
+    {
+      "epoch": 0.45743329097839897,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00013174418191364988,
+      "loss": 1.2007,
+      "step": 4140
+    },
+    {
+      "epoch": 0.45798574664383185,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00013156124377938699,
+      "loss": 1.297,
+      "step": 4145
+    },
+    {
+      "epoch": 0.4585382023092647,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00013137818826040854,
+      "loss": 1.3482,
+      "step": 4150
+    },
+    {
+      "epoch": 0.45909065797469756,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00013119501603754704,
+      "loss": 1.3098,
+      "step": 4155
+    },
+    {
+      "epoch": 0.4596431136401304,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.000131011727792069,
+      "loss": 1.2378,
+      "step": 4160
+    },
+    {
+      "epoch": 0.4601955693055632,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00013082832420567247,
+      "loss": 1.2998,
+      "step": 4165
+    },
+    {
+      "epoch": 0.4607480249709961,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00013064480596048454,
+      "loss": 1.271,
+      "step": 4170
+    },
+    {
+      "epoch": 0.4613004806364289,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00013046117373905866,
+      "loss": 1.282,
+      "step": 4175
+    },
+    {
+      "epoch": 0.46185293630186175,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001302774282243722,
+      "loss": 1.2486,
+      "step": 4180
+    },
+    {
+      "epoch": 0.46240539196729463,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00013009357009982397,
+      "loss": 1.2647,
+      "step": 4185
+    },
+    {
+      "epoch": 0.46295784763272746,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00012990960004923154,
+      "loss": 1.2852,
+      "step": 4190
+    },
+    {
+      "epoch": 0.46351030329816034,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0001297255187568288,
+      "loss": 1.2556,
+      "step": 4195
+    },
+    {
+      "epoch": 0.46406275896359317,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00012954132690726336,
+      "loss": 1.3166,
+      "step": 4200
+    },
+    {
+      "epoch": 0.464615214629026,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00012935702518559398,
+      "loss": 1.2838,
+      "step": 4205
+    },
+    {
+      "epoch": 0.4651676702944589,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00012917261427728815,
+      "loss": 1.2858,
+      "step": 4210
+    },
+    {
+      "epoch": 0.4657201259598917,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0001289880948682194,
+      "loss": 1.2866,
+      "step": 4215
+    },
+    {
+      "epoch": 0.4662725816253246,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0001288034676446648,
+      "loss": 1.2379,
+      "step": 4220
+    },
+    {
+      "epoch": 0.4668250372907574,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00012861873329330248,
+      "loss": 1.2144,
+      "step": 4225
+    },
+    {
+      "epoch": 0.46737749295619024,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00012843389250120885,
+      "loss": 1.3324,
+      "step": 4230
+    },
+    {
+      "epoch": 0.4679299486216231,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00012824894595585637,
+      "loss": 1.2816,
+      "step": 4235
+    },
+    {
+      "epoch": 0.46848240428705595,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00012806389434511076,
+      "loss": 1.2862,
+      "step": 4240
+    },
+    {
+      "epoch": 0.46903485995248884,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0001278787383572285,
+      "loss": 1.2216,
+      "step": 4245
+    },
+    {
+      "epoch": 0.46958731561792166,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00012769347868085427,
+      "loss": 1.2242,
+      "step": 4250
+    },
+    {
+      "epoch": 0.4701397712833545,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00012750811600501842,
+      "loss": 1.3282,
+      "step": 4255
+    },
+    {
+      "epoch": 0.4706922269487874,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00012732265101913435,
+      "loss": 1.2393,
+      "step": 4260
+    },
+    {
+      "epoch": 0.4712446826142202,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.000127137084412996,
+      "loss": 1.2319,
+      "step": 4265
+    },
+    {
+      "epoch": 0.4717971382796531,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00012695141687677527,
+      "loss": 1.2498,
+      "step": 4270
+    },
+    {
+      "epoch": 0.4723495939450859,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00012676564910101947,
+      "loss": 1.3246,
+      "step": 4275
+    },
+    {
+      "epoch": 0.47290204961051874,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001265797817766486,
+      "loss": 1.3203,
+      "step": 4280
+    },
+    {
+      "epoch": 0.4734545052759516,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0001263938155949531,
+      "loss": 1.2485,
+      "step": 4285
+    },
+    {
+      "epoch": 0.47400696094138445,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00012620775124759092,
+      "loss": 1.2922,
+      "step": 4290
+    },
+    {
+      "epoch": 0.47455941660681733,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001260215894265852,
+      "loss": 1.3289,
+      "step": 4295
+    },
+    {
+      "epoch": 0.47511187227225016,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0001258353308243217,
+      "loss": 1.2854,
+      "step": 4300
+    },
+    {
+      "epoch": 0.475664327937683,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00012564897613354586,
+      "loss": 1.2822,
+      "step": 4305
+    },
+    {
+      "epoch": 0.47621678360311587,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00012546252604736074,
+      "loss": 1.2535,
+      "step": 4310
+    },
+    {
+      "epoch": 0.4767692392685487,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00012527598125922413,
+      "loss": 1.2759,
+      "step": 4315
+    },
+    {
+      "epoch": 0.4773216949339815,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00012508934246294604,
+      "loss": 1.3582,
+      "step": 4320
+    },
+    {
+      "epoch": 0.4778741505994144,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00012490261035268612,
+      "loss": 1.2938,
+      "step": 4325
+    },
+    {
+      "epoch": 0.47842660626484723,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00012471578562295115,
+      "loss": 1.2608,
+      "step": 4330
+    },
+    {
+      "epoch": 0.4789790619302801,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0001245288689685922,
+      "loss": 1.3062,
+      "step": 4335
+    },
+    {
+      "epoch": 0.47953151759571294,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0001243418610848024,
+      "loss": 1.228,
+      "step": 4340
+    },
+    {
+      "epoch": 0.48008397326114577,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00012415476266711413,
+      "loss": 1.3653,
+      "step": 4345
+    },
+    {
+      "epoch": 0.48063642892657865,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00012396757441139654,
+      "loss": 1.2368,
+      "step": 4350
+    },
+    {
+      "epoch": 0.4811888845920115,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00012378029701385287,
+      "loss": 1.2846,
+      "step": 4355
+    },
+    {
+      "epoch": 0.48174134025744436,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00012359293117101782,
+      "loss": 1.2454,
+      "step": 4360
+    },
+    {
+      "epoch": 0.4822937959228772,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0001234054775797552,
+      "loss": 1.2893,
+      "step": 4365
+    },
+    {
+      "epoch": 0.48284625158831,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00012321793693725509,
+      "loss": 1.2866,
+      "step": 4370
+    },
+    {
+      "epoch": 0.4833987072537429,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00012303030994103133,
+      "loss": 1.3267,
+      "step": 4375
+    },
+    {
+      "epoch": 0.4839511629191757,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00012284259728891897,
+      "loss": 1.3041,
+      "step": 4380
+    },
+    {
+      "epoch": 0.4845036185846086,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00012265479967907159,
+      "loss": 1.2848,
+      "step": 4385
+    },
+    {
+      "epoch": 0.48505607425004144,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00012246691780995881,
+      "loss": 1.2127,
+      "step": 4390
+    },
+    {
+      "epoch": 0.48560852991547426,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0001222789523803636,
+      "loss": 1.2426,
+      "step": 4395
+    },
+    {
+      "epoch": 0.48616098558090715,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00012209090408937971,
+      "loss": 1.2206,
+      "step": 4400
+    },
+    {
+      "epoch": 0.48671344124634,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00012190277363640907,
+      "loss": 1.3072,
+      "step": 4405
+    },
+    {
+      "epoch": 0.48726589691177286,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00012171456172115923,
+      "loss": 1.3464,
+      "step": 4410
+    },
+    {
+      "epoch": 0.4878183525772057,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00012152626904364067,
+      "loss": 1.2361,
+      "step": 4415
+    },
+    {
+      "epoch": 0.4883708082426385,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00012133789630416425,
+      "loss": 1.3115,
+      "step": 4420
+    },
+    {
+      "epoch": 0.4889232639080714,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00012114944420333869,
+      "loss": 1.2941,
+      "step": 4425
+    },
+    {
+      "epoch": 0.4894757195735042,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00012096091344206777,
+      "loss": 1.2627,
+      "step": 4430
+    },
+    {
+      "epoch": 0.4900281752389371,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00012077230472154786,
+      "loss": 1.3617,
+      "step": 4435
+    },
+    {
+      "epoch": 0.49058063090436993,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00012058361874326526,
+      "loss": 1.2621,
+      "step": 4440
+    },
+    {
+      "epoch": 0.49113308656980276,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00012039485620899369,
+      "loss": 1.2315,
+      "step": 4445
+    },
+    {
+      "epoch": 0.49168554223523564,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00012020601782079155,
+      "loss": 1.2041,
+      "step": 4450
+    },
+    {
+      "epoch": 0.49223799790066847,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00012001710428099935,
+      "loss": 1.2634,
+      "step": 4455
+    },
+    {
+      "epoch": 0.4927904535661013,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00011982811629223709,
+      "loss": 1.3279,
+      "step": 4460
+    },
+    {
+      "epoch": 0.4933429092315342,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00011963905455740177,
+      "loss": 1.2847,
+      "step": 4465
+    },
+    {
+      "epoch": 0.493895364896967,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00011944991977966452,
+      "loss": 1.2713,
+      "step": 4470
+    },
+    {
+      "epoch": 0.4944478205623999,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00011926071266246826,
+      "loss": 1.2448,
+      "step": 4475
+    },
+    {
+      "epoch": 0.4950002762278327,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00011907143390952493,
+      "loss": 1.2263,
+      "step": 4480
+    },
+    {
+      "epoch": 0.49555273189326554,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00011888208422481287,
+      "loss": 1.3221,
+      "step": 4485
+    },
+    {
+      "epoch": 0.4961051875586984,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00011869266431257422,
+      "loss": 1.2537,
+      "step": 4490
+    },
+    {
+      "epoch": 0.49665764322413125,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00011850317487731239,
+      "loss": 1.2193,
+      "step": 4495
+    },
+    {
+      "epoch": 0.49721009888956413,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00011831361662378932,
+      "loss": 1.3306,
+      "step": 4500
+    },
+    {
+      "epoch": 0.49776255455499696,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001181239902570229,
+      "loss": 1.3171,
+      "step": 4505
+    },
+    {
+      "epoch": 0.4983150102204298,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00011793429648228436,
+      "loss": 1.2719,
+      "step": 4510
+    },
+    {
+      "epoch": 0.49886746588586267,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00011774453600509559,
+      "loss": 1.2798,
+      "step": 4515
+    },
+    {
+      "epoch": 0.4994199215512955,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00011755470953122667,
+      "loss": 1.2583,
+      "step": 4520
+    },
+    {
+      "epoch": 0.4999723772167284,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00011736481776669306,
+      "loss": 1.372,
+      "step": 4525
+    },
+    {
+      "epoch": 0.5005248328821612,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00011717486141775305,
+      "loss": 1.2945,
+      "step": 4530
+    },
+    {
+      "epoch": 0.501077288547594,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00011698484119090518,
+      "loss": 1.2232,
+      "step": 4535
+    },
+    {
+      "epoch": 0.5016297442130269,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00011679475779288555,
+      "loss": 1.2779,
+      "step": 4540
+    },
+    {
+      "epoch": 0.5021821998784598,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00011660461193066521,
+      "loss": 1.2376,
+      "step": 4545
+    },
+    {
+      "epoch": 0.5027346555438926,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0001164144043114475,
+      "loss": 1.328,
+      "step": 4550
+    },
+    {
+      "epoch": 0.5032871112093255,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00011622413564266555,
+      "loss": 1.2981,
+      "step": 4555
+    },
+    {
+      "epoch": 0.5038395668747583,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0001160338066319794,
+      "loss": 1.2478,
+      "step": 4560
+    },
+    {
+      "epoch": 0.5043920225401911,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00011584341798727365,
+      "loss": 1.3444,
+      "step": 4565
+    },
+    {
+      "epoch": 0.504944478205624,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001156529704166546,
+      "loss": 1.3496,
+      "step": 4570
+    },
+    {
+      "epoch": 0.5054969338710569,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00011546246462844779,
+      "loss": 1.2273,
+      "step": 4575
+    },
+    {
+      "epoch": 0.5060493895364897,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00011527190133119526,
+      "loss": 1.2409,
+      "step": 4580
+    },
+    {
+      "epoch": 0.5066018452019225,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001150812812336529,
+      "loss": 1.2615,
+      "step": 4585
+    },
+    {
+      "epoch": 0.5071543008673554,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00011489060504478788,
+      "loss": 1.2579,
+      "step": 4590
+    },
+    {
+      "epoch": 0.5077067565327883,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00011469987347377602,
+      "loss": 1.343,
+      "step": 4595
+    },
+    {
+      "epoch": 0.5082592121982211,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00011450908722999909,
+      "loss": 1.1766,
+      "step": 4600
+    },
+    {
+      "epoch": 0.508811667863654,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00011431824702304221,
+      "loss": 1.2905,
+      "step": 4605
+    },
+    {
+      "epoch": 0.5093641235290868,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00011412735356269124,
+      "loss": 1.2963,
+      "step": 4610
+    },
+    {
+      "epoch": 0.5099165791945196,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00011393640755893002,
+      "loss": 1.3127,
+      "step": 4615
+    },
+    {
+      "epoch": 0.5104690348599525,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00011374540972193786,
+      "loss": 1.2661,
+      "step": 4620
+    },
+    {
+      "epoch": 0.5110214905253854,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00011355436076208687,
+      "loss": 1.2579,
+      "step": 4625
+    },
+    {
+      "epoch": 0.5115739461908182,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00011336326138993927,
+      "loss": 1.2682,
+      "step": 4630
+    },
+    {
+      "epoch": 0.512126401856251,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00011317211231624483,
+      "loss": 1.2175,
+      "step": 4635
+    },
+    {
+      "epoch": 0.5126788575216839,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00011298091425193806,
+      "loss": 1.2072,
+      "step": 4640
+    },
+    {
+      "epoch": 0.5132313131871168,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00011278966790813582,
+      "loss": 1.2484,
+      "step": 4645
+    },
+    {
+      "epoch": 0.5137837688525496,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001125983739961344,
+      "loss": 1.333,
+      "step": 4650
+    },
+    {
+      "epoch": 0.5143362245179824,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00011240703322740711,
+      "loss": 1.3642,
+      "step": 4655
+    },
+    {
+      "epoch": 0.5148886801834153,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00011221564631360152,
+      "loss": 1.3279,
+      "step": 4660
+    },
+    {
+      "epoch": 0.5154411358488481,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00011202421396653677,
+      "loss": 1.2654,
+      "step": 4665
+    },
+    {
+      "epoch": 0.5159935915142809,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.000111832736898201,
+      "loss": 1.2242,
+      "step": 4670
+    },
+    {
+      "epoch": 0.5165460471797139,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00011164121582074873,
+      "loss": 1.3846,
+      "step": 4675
+    },
+    {
+      "epoch": 0.5170985028451467,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00011144965144649809,
+      "loss": 1.302,
+      "step": 4680
+    },
+    {
+      "epoch": 0.5176509585105795,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00011125804448792831,
+      "loss": 1.2522,
+      "step": 4685
+    },
+    {
+      "epoch": 0.5182034141760123,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00011106639565767692,
+      "loss": 1.2961,
+      "step": 4690
+    },
+    {
+      "epoch": 0.5187558698414452,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00011087470566853726,
+      "loss": 1.2395,
+      "step": 4695
+    },
+    {
+      "epoch": 0.5193083255068781,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00011068297523345573,
+      "loss": 1.3415,
+      "step": 4700
+    },
+    {
+      "epoch": 0.5198607811723109,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00011049120506552913,
+      "loss": 1.2696,
+      "step": 4705
+    },
+    {
+      "epoch": 0.5204132368377438,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00011029939587800206,
+      "loss": 1.2383,
+      "step": 4710
+    },
+    {
+      "epoch": 0.5209656925031766,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00011010754838426428,
+      "loss": 1.404,
+      "step": 4715
+    },
+    {
+      "epoch": 0.5215181481686094,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0001099156632978479,
+      "loss": 1.3234,
+      "step": 4720
+    },
+    {
+      "epoch": 0.5220706038340424,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00010972374133242502,
+      "loss": 1.2487,
+      "step": 4725
+    },
+    {
+      "epoch": 0.5226230594994752,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00010953178320180475,
+      "loss": 1.2092,
+      "step": 4730
+    },
+    {
+      "epoch": 0.523175515164908,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00010933978961993083,
+      "loss": 1.3322,
+      "step": 4735
+    },
+    {
+      "epoch": 0.5237279708303408,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00010914776130087873,
+      "loss": 1.2356,
+      "step": 4740
+    },
+    {
+      "epoch": 0.5242804264957737,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0001089556989588532,
+      "loss": 1.2565,
+      "step": 4745
+    },
+    {
+      "epoch": 0.5248328821612066,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00010876360330818553,
+      "loss": 1.2775,
+      "step": 4750
+    },
+    {
+      "epoch": 0.5253853378266394,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00010857147506333088,
+      "loss": 1.2035,
+      "step": 4755
+    },
+    {
+      "epoch": 0.5259377934920723,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00010837931493886562,
+      "loss": 1.3017,
+      "step": 4760
+    },
+    {
+      "epoch": 0.5264902491575051,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0001081871236494847,
+      "loss": 1.2652,
+      "step": 4765
+    },
+    {
+      "epoch": 0.5270427048229379,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00010799490190999892,
+      "loss": 1.2914,
+      "step": 4770
+    },
+    {
+      "epoch": 0.5275951604883709,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0001078026504353325,
+      "loss": 1.2566,
+      "step": 4775
+    },
+    {
+      "epoch": 0.5281476161538037,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00010761036994052008,
+      "loss": 1.3195,
+      "step": 4780
+    },
+    {
+      "epoch": 0.5287000718192365,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00010741806114070434,
+      "loss": 1.3088,
+      "step": 4785
+    },
+    {
+      "epoch": 0.5292525274846693,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00010722572475113315,
+      "loss": 1.2676,
+      "step": 4790
+    },
+    {
+      "epoch": 0.5298049831501022,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00010703336148715705,
+      "loss": 1.3551,
+      "step": 4795
+    },
+    {
+      "epoch": 0.5303574388155351,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00010684097206422654,
+      "loss": 1.2925,
+      "step": 4800
+    },
+    {
+      "epoch": 0.5309098944809679,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00010664855719788935,
+      "loss": 1.281,
+      "step": 4805
+    },
+    {
+      "epoch": 0.5314623501464008,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00010645611760378795,
+      "loss": 1.3131,
+      "step": 4810
+    },
+    {
+      "epoch": 0.5320148058118336,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00010626365399765667,
+      "loss": 1.3253,
+      "step": 4815
+    },
+    {
+      "epoch": 0.5325672614772664,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00010607116709531918,
+      "loss": 1.2705,
+      "step": 4820
+    },
+    {
+      "epoch": 0.5331197171426993,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00010587865761268583,
+      "loss": 1.3432,
+      "step": 4825
+    },
+    {
+      "epoch": 0.5336721728081322,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00010568612626575092,
+      "loss": 1.2824,
+      "step": 4830
+    },
+    {
+      "epoch": 0.534224628473565,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00010549357377059006,
+      "loss": 1.2724,
+      "step": 4835
+    },
+    {
+      "epoch": 0.5347770841389978,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00010530100084335758,
+      "loss": 1.2218,
+      "step": 4840
+    },
+    {
+      "epoch": 0.5353295398044307,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0001051084082002837,
+      "loss": 1.4052,
+      "step": 4845
+    },
+    {
+      "epoch": 0.5358819954698636,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00010491579655767203,
+      "loss": 1.3135,
+      "step": 4850
+    },
+    {
+      "epoch": 0.5364344511352964,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00010472316663189683,
+      "loss": 1.2908,
+      "step": 4855
+    },
+    {
+      "epoch": 0.5369869068007292,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00010453051913940042,
+      "loss": 1.2162,
+      "step": 4860
+    },
+    {
+      "epoch": 0.5375393624661621,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00010433785479669038,
+      "loss": 1.3188,
+      "step": 4865
+    },
+    {
+      "epoch": 0.5380918181315949,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00010414517432033695,
+      "loss": 1.251,
+      "step": 4870
+    },
+    {
+      "epoch": 0.5386442737970278,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001039524784269704,
+      "loss": 1.2296,
+      "step": 4875
+    },
+    {
+      "epoch": 0.5391967294624607,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00010375976783327841,
+      "loss": 1.2543,
+      "step": 4880
+    },
+    {
+      "epoch": 0.5397491851278935,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00010356704325600324,
+      "loss": 1.2968,
+      "step": 4885
+    },
+    {
+      "epoch": 0.5403016407933263,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00010337430541193918,
+      "loss": 1.2856,
+      "step": 4890
+    },
+    {
+      "epoch": 0.5408540964587591,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00010318155501792988,
+      "loss": 1.2595,
+      "step": 4895
+    },
+    {
+      "epoch": 0.5414065521241921,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00010298879279086568,
+      "loss": 1.2911,
+      "step": 4900
+    },
+    {
+      "epoch": 0.5419590077896249,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00010279601944768089,
+      "loss": 1.2429,
+      "step": 4905
+    },
+    {
+      "epoch": 0.5425114634550577,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001026032357053512,
+      "loss": 1.3281,
+      "step": 4910
+    },
+    {
+      "epoch": 0.5430639191204906,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00010241044228089096,
+      "loss": 1.2073,
+      "step": 4915
+    },
+    {
+      "epoch": 0.5436163747859234,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00010221763989135052,
+      "loss": 1.3138,
+      "step": 4920
+    },
+    {
+      "epoch": 0.5441688304513563,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00010202482925381358,
+      "loss": 1.2066,
+      "step": 4925
+    },
+    {
+      "epoch": 0.5447212861167892,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00010183201108539453,
+      "loss": 1.2961,
+      "step": 4930
+    },
+    {
+      "epoch": 0.545273741782222,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00010163918610323579,
+      "loss": 1.3148,
+      "step": 4935
+    },
+    {
+      "epoch": 0.5458261974476548,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00010144635502450508,
+      "loss": 1.2551,
+      "step": 4940
+    },
+    {
+      "epoch": 0.5463786531130876,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00010125351856639278,
+      "loss": 1.2512,
+      "step": 4945
+    },
+    {
+      "epoch": 0.5469311087785205,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00010106067744610933,
+      "loss": 1.3768,
+      "step": 4950
+    },
+    {
+      "epoch": 0.5474835644439534,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00010086783238088244,
+      "loss": 1.307,
+      "step": 4955
+    },
+    {
+      "epoch": 0.5480360201093862,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00010067498408795462,
+      "loss": 1.3127,
+      "step": 4960
+    },
+    {
+      "epoch": 0.5485884757748191,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00010048213328458028,
+      "loss": 1.2862,
+      "step": 4965
+    },
+    {
+      "epoch": 0.5491409314402519,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00010028928068802315,
+      "loss": 1.171,
+      "step": 4970
+    },
+    {
+      "epoch": 0.5496933871056847,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00010009642701555368,
+      "loss": 1.2546,
+      "step": 4975
+    },
+    {
+      "epoch": 0.5502458427711177,
+      "grad_norm": 0.3828125,
+      "learning_rate": 9.990357298444632e-05,
+      "loss": 1.1777,
+      "step": 4980
+    },
+    {
+      "epoch": 0.5507982984365505,
+      "grad_norm": 0.390625,
+      "learning_rate": 9.971071931197685e-05,
+      "loss": 1.2886,
+      "step": 4985
+    },
+    {
+      "epoch": 0.5513507541019833,
+      "grad_norm": 0.40234375,
+      "learning_rate": 9.951786671541973e-05,
+      "loss": 1.2069,
+      "step": 4990
+    },
+    {
+      "epoch": 0.5519032097674161,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.932501591204536e-05,
+      "loss": 1.3094,
+      "step": 4995
+    },
+    {
+      "epoch": 0.552455665432849,
+      "grad_norm": 0.40234375,
+      "learning_rate": 9.913216761911755e-05,
+      "loss": 1.3084,
+      "step": 5000
+    },
+    {
+      "epoch": 0.5530081210982819,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.893932255389068e-05,
+      "loss": 1.2617,
+      "step": 5005
+    },
+    {
+      "epoch": 0.5535605767637147,
+      "grad_norm": 0.40234375,
+      "learning_rate": 9.874648143360723e-05,
+      "loss": 1.2746,
+      "step": 5010
+    },
+    {
+      "epoch": 0.5541130324291476,
+      "grad_norm": 0.416015625,
+      "learning_rate": 9.855364497549496e-05,
+      "loss": 1.2916,
+      "step": 5015
+    },
+    {
+      "epoch": 0.5546654880945804,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.836081389676422e-05,
+      "loss": 1.3503,
+      "step": 5020
+    },
+    {
+      "epoch": 0.5552179437600132,
+      "grad_norm": 0.390625,
+      "learning_rate": 9.816798891460546e-05,
+      "loss": 1.2901,
+      "step": 5025
+    },
+    {
+      "epoch": 0.5557703994254461,
+      "grad_norm": 0.376953125,
+      "learning_rate": 9.797517074618642e-05,
+      "loss": 1.245,
+      "step": 5030
+    },
+    {
+      "epoch": 0.556322855090879,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.778236010864949e-05,
+      "loss": 1.2161,
+      "step": 5035
+    },
+    {
+      "epoch": 0.5568753107563118,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.758955771910906e-05,
+      "loss": 1.2126,
+      "step": 5040
+    },
+    {
+      "epoch": 0.5574277664217446,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.739676429464881e-05,
+      "loss": 1.2392,
+      "step": 5045
+    },
+    {
+      "epoch": 0.5579802220871775,
+      "grad_norm": 0.39453125,
+      "learning_rate": 9.720398055231911e-05,
+      "loss": 1.2809,
+      "step": 5050
+    },
+    {
+      "epoch": 0.5585326777526104,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.701120720913433e-05,
+      "loss": 1.2809,
+      "step": 5055
+    },
+    {
+      "epoch": 0.5590851334180432,
+      "grad_norm": 0.3984375,
+      "learning_rate": 9.68184449820701e-05,
+      "loss": 1.2276,
+      "step": 5060
+    },
+    {
+      "epoch": 0.559637589083476,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.662569458806085e-05,
+      "loss": 1.1839,
+      "step": 5065
+    },
+    {
+      "epoch": 0.5601900447489089,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.64329567439968e-05,
+      "loss": 1.294,
+      "step": 5070
+    },
+    {
+      "epoch": 0.5607425004143417,
+      "grad_norm": 0.384765625,
+      "learning_rate": 9.624023216672161e-05,
+      "loss": 1.2283,
+      "step": 5075
+    },
+    {
+      "epoch": 0.5612949560797746,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.604752157302961e-05,
+      "loss": 1.3297,
+      "step": 5080
+    },
+    {
+      "epoch": 0.5618474117452075,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.585482567966309e-05,
+      "loss": 1.2957,
+      "step": 5085
+    },
+    {
+      "epoch": 0.5623998674106403,
+      "grad_norm": 0.41015625,
+      "learning_rate": 9.566214520330966e-05,
+      "loss": 1.2493,
+      "step": 5090
+    },
+    {
+      "epoch": 0.5629523230760731,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.54694808605996e-05,
+      "loss": 1.2427,
+      "step": 5095
+    },
+    {
+      "epoch": 0.563504778741506,
+      "grad_norm": 0.392578125,
+      "learning_rate": 9.527683336810318e-05,
+      "loss": 1.1977,
+      "step": 5100
+    },
+    {
+      "epoch": 0.5640572344069389,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.5084203442328e-05,
+      "loss": 1.3098,
+      "step": 5105
+    },
+    {
+      "epoch": 0.5646096900723717,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.489159179971632e-05,
+      "loss": 1.2023,
+      "step": 5110
+    },
+    {
+      "epoch": 0.5651621457378045,
+      "grad_norm": 0.3828125,
+      "learning_rate": 9.469899915664244e-05,
+      "loss": 1.2855,
+      "step": 5115
+    },
+    {
+      "epoch": 0.5657146014032374,
+      "grad_norm": 0.37890625,
+      "learning_rate": 9.450642622940995e-05,
+      "loss": 1.2757,
+      "step": 5120
+    },
+    {
+      "epoch": 0.5662670570686702,
+      "grad_norm": 0.392578125,
+      "learning_rate": 9.43138737342491e-05,
+      "loss": 1.2366,
+      "step": 5125
+    },
+    {
+      "epoch": 0.5668195127341031,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.412134238731418e-05,
+      "loss": 1.4092,
+      "step": 5130
+    },
+    {
+      "epoch": 0.567371968399536,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.392883290468083e-05,
+      "loss": 1.3325,
+      "step": 5135
+    },
+    {
+      "epoch": 0.5679244240649688,
+      "grad_norm": 0.40234375,
+      "learning_rate": 9.373634600234334e-05,
+      "loss": 1.3355,
+      "step": 5140
+    },
+    {
+      "epoch": 0.5684768797304016,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.354388239621208e-05,
+      "loss": 1.2945,
+      "step": 5145
+    },
+    {
+      "epoch": 0.5690293353958344,
+      "grad_norm": 0.419921875,
+      "learning_rate": 9.335144280211066e-05,
+      "loss": 1.2847,
+      "step": 5150
+    },
+    {
+      "epoch": 0.5695817910612674,
+      "grad_norm": 0.408203125,
+      "learning_rate": 9.31590279357735e-05,
+      "loss": 1.2663,
+      "step": 5155
+    },
+    {
+      "epoch": 0.5701342467267002,
+      "grad_norm": 0.408203125,
+      "learning_rate": 9.296663851284297e-05,
+      "loss": 1.3607,
+      "step": 5160
+    },
+    {
+      "epoch": 0.570686702392133,
+      "grad_norm": 0.3984375,
+      "learning_rate": 9.277427524886689e-05,
+      "loss": 1.3177,
+      "step": 5165
+    },
+    {
+      "epoch": 0.5712391580575659,
+      "grad_norm": 0.40625,
+      "learning_rate": 9.258193885929569e-05,
+      "loss": 1.2937,
+      "step": 5170
+    },
+    {
+      "epoch": 0.5717916137229987,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.238963005947993e-05,
+      "loss": 1.2867,
+      "step": 5175
+    },
+    {
+      "epoch": 0.5723440693884316,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.219734956466752e-05,
+      "loss": 1.2528,
+      "step": 5180
+    },
+    {
+      "epoch": 0.5728965250538645,
+      "grad_norm": 0.388671875,
+      "learning_rate": 9.200509809000108e-05,
+      "loss": 1.3146,
+      "step": 5185
+    },
+    {
+      "epoch": 0.5734489807192973,
+      "grad_norm": 0.373046875,
+      "learning_rate": 9.181287635051534e-05,
+      "loss": 1.2696,
+      "step": 5190
+    },
+    {
+      "epoch": 0.5740014363847301,
+      "grad_norm": 0.388671875,
+      "learning_rate": 9.16206850611344e-05,
+      "loss": 1.3678,
+      "step": 5195
+    },
+    {
+      "epoch": 0.5745538920501629,
+      "grad_norm": 0.416015625,
+      "learning_rate": 9.142852493666914e-05,
+      "loss": 1.3089,
+      "step": 5200
+    },
+    {
+      "epoch": 0.5751063477155959,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.123639669181448e-05,
+      "loss": 1.2477,
+      "step": 5205
+    },
+    {
+      "epoch": 0.5756588033810287,
+      "grad_norm": 0.41015625,
+      "learning_rate": 9.104430104114681e-05,
+      "loss": 1.2444,
+      "step": 5210
+    },
+    {
+      "epoch": 0.5762112590464615,
+      "grad_norm": 0.39453125,
+      "learning_rate": 9.085223869912129e-05,
+      "loss": 1.2129,
+      "step": 5215
+    },
+    {
+      "epoch": 0.5767637147118944,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.066021038006919e-05,
+      "loss": 1.3529,
+      "step": 5220
+    },
+    {
+      "epoch": 0.5773161703773272,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.046821679819527e-05,
+      "loss": 1.2836,
+      "step": 5225
+    },
+    {
+      "epoch": 0.57786862604276,
+      "grad_norm": 0.40625,
+      "learning_rate": 9.0276258667575e-05,
+      "loss": 1.2341,
+      "step": 5230
+    },
+    {
+      "epoch": 0.578421081708193,
+      "grad_norm": 0.392578125,
+      "learning_rate": 9.008433670215212e-05,
+      "loss": 1.2246,
+      "step": 5235
+    },
+    {
+      "epoch": 0.5789735373736258,
+      "grad_norm": 0.421875,
+      "learning_rate": 8.989245161573576e-05,
+      "loss": 1.2864,
+      "step": 5240
+    },
+    {
+      "epoch": 0.5795259930390586,
+      "grad_norm": 0.390625,
+      "learning_rate": 8.970060412199795e-05,
+      "loss": 1.2539,
+      "step": 5245
+    },
+    {
+      "epoch": 0.5800784487044914,
+      "grad_norm": 0.40625,
+      "learning_rate": 8.950879493447091e-05,
+      "loss": 1.3407,
+      "step": 5250
+    },
+    {
+      "epoch": 0.5806309043699243,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.931702476654431e-05,
+      "loss": 1.2563,
+      "step": 5255
+    },
+    {
+      "epoch": 0.5811833600353572,
+      "grad_norm": 0.396484375,
+      "learning_rate": 8.912529433146278e-05,
+      "loss": 1.2357,
+      "step": 5260
+    },
+    {
+      "epoch": 0.58173581570079,
+      "grad_norm": 0.421875,
+      "learning_rate": 8.893360434232312e-05,
+      "loss": 1.2888,
+      "step": 5265
+    },
+    {
+      "epoch": 0.5822882713662229,
+      "grad_norm": 0.396484375,
+      "learning_rate": 8.874195551207174e-05,
+      "loss": 1.2673,
+      "step": 5270
+    },
+    {
+      "epoch": 0.5828407270316557,
+      "grad_norm": 0.388671875,
+      "learning_rate": 8.855034855350194e-05,
+      "loss": 1.3072,
+      "step": 5275
+    },
+    {
+      "epoch": 0.5833931826970885,
+      "grad_norm": 0.404296875,
+      "learning_rate": 8.835878417925132e-05,
+      "loss": 1.2163,
+      "step": 5280
+    },
+    {
+      "epoch": 0.5839456383625214,
+      "grad_norm": 0.392578125,
+      "learning_rate": 8.816726310179904e-05,
+      "loss": 1.296,
+      "step": 5285
+    },
+    {
+      "epoch": 0.5844980940279543,
+      "grad_norm": 0.392578125,
+      "learning_rate": 8.797578603346328e-05,
+      "loss": 1.2156,
+      "step": 5290
+    },
+    {
+      "epoch": 0.5850505496933871,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.778435368639851e-05,
+      "loss": 1.1998,
+      "step": 5295
+    },
+    {
+      "epoch": 0.5856030053588199,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.759296677259291e-05,
+      "loss": 1.3334,
+      "step": 5300
+    },
+    {
+      "epoch": 0.5861554610242528,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.740162600386565e-05,
+      "loss": 1.2662,
+      "step": 5305
+    },
+    {
+      "epoch": 0.5867079166896857,
+      "grad_norm": 0.388671875,
+      "learning_rate": 8.721033209186425e-05,
+      "loss": 1.3158,
+      "step": 5310
+    },
+    {
+      "epoch": 0.5872603723551185,
+      "grad_norm": 0.40625,
+      "learning_rate": 8.701908574806197e-05,
+      "loss": 1.2605,
+      "step": 5315
+    },
+    {
+      "epoch": 0.5878128280205513,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.682788768375521e-05,
+      "loss": 1.2702,
+      "step": 5320
+    },
+    {
+      "epoch": 0.5883652836859842,
+      "grad_norm": 0.376953125,
+      "learning_rate": 8.663673861006074e-05,
+      "loss": 1.2213,
+      "step": 5325
+    },
+    {
+      "epoch": 0.588917739351417,
+      "grad_norm": 0.38671875,
+      "learning_rate": 8.644563923791318e-05,
+      "loss": 1.1698,
+      "step": 5330
+    },
+    {
+      "epoch": 0.5894701950168499,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.625459027806214e-05,
+      "loss": 1.2026,
+      "step": 5335
+    },
+    {
+      "epoch": 0.5900226506822828,
+      "grad_norm": 0.40625,
+      "learning_rate": 8.606359244106999e-05,
+      "loss": 1.3359,
+      "step": 5340
+    },
+    {
+      "epoch": 0.5905751063477156,
+      "grad_norm": 0.408203125,
+      "learning_rate": 8.587264643730877e-05,
+      "loss": 1.3318,
+      "step": 5345
+    },
+    {
+      "epoch": 0.5911275620131484,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.568175297695777e-05,
+      "loss": 1.2001,
+      "step": 5350
+    },
+    {
+      "epoch": 0.5916800176785812,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.549091277000091e-05,
+      "loss": 1.2584,
+      "step": 5355
+    },
+    {
+      "epoch": 0.5922324733440142,
+      "grad_norm": 0.404296875,
+      "learning_rate": 8.530012652622397e-05,
+      "loss": 1.2887,
+      "step": 5360
+    },
+    {
+      "epoch": 0.592784929009447,
+      "grad_norm": 0.408203125,
+      "learning_rate": 8.510939495521213e-05,
+      "loss": 1.2336,
+      "step": 5365
+    },
+    {
+      "epoch": 0.5933373846748798,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.491871876634712e-05,
+      "loss": 1.2963,
+      "step": 5370
+    },
+    {
+      "epoch": 0.5938898403403127,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.472809866880475e-05,
+      "loss": 1.2591,
+      "step": 5375
+    },
+    {
+      "epoch": 0.5944422960057455,
+      "grad_norm": 0.408203125,
+      "learning_rate": 8.45375353715522e-05,
+      "loss": 1.2437,
+      "step": 5380
+    },
+    {
+      "epoch": 0.5949947516711784,
+      "grad_norm": 0.423828125,
+      "learning_rate": 8.434702958334539e-05,
+      "loss": 1.2978,
+      "step": 5385
+    },
+    {
+      "epoch": 0.5955472073366113,
+      "grad_norm": 0.41015625,
+      "learning_rate": 8.415658201272636e-05,
+      "loss": 1.2626,
+      "step": 5390
+    },
+    {
+      "epoch": 0.5960996630020441,
+      "grad_norm": 0.408203125,
+      "learning_rate": 8.39661933680206e-05,
+      "loss": 1.2776,
+      "step": 5395
+    },
+    {
+      "epoch": 0.5966521186674769,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.377586435733446e-05,
+      "loss": 1.2243,
+      "step": 5400
+    },
+    {
+      "epoch": 0.5972045743329097,
+      "grad_norm": 0.388671875,
+      "learning_rate": 8.358559568855249e-05,
+      "loss": 1.3193,
+      "step": 5405
+    },
+    {
+      "epoch": 0.5977570299983427,
+      "grad_norm": 0.416015625,
+      "learning_rate": 8.33953880693348e-05,
+      "loss": 1.3303,
+      "step": 5410
+    },
+    {
+      "epoch": 0.5983094856637755,
+      "grad_norm": 0.388671875,
+      "learning_rate": 8.320524220711446e-05,
+      "loss": 1.2865,
+      "step": 5415
+    },
+    {
+      "epoch": 0.5988619413292083,
+      "grad_norm": 0.40234375,
+      "learning_rate": 8.301515880909481e-05,
+      "loss": 1.2883,
+      "step": 5420
+    },
+    {
+      "epoch": 0.5994143969946412,
+      "grad_norm": 0.376953125,
+      "learning_rate": 8.282513858224698e-05,
+      "loss": 1.2033,
+      "step": 5425
+    },
+    {
+      "epoch": 0.599966852660074,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.263518223330697e-05,
+      "loss": 1.3027,
+      "step": 5430
+    },
+    {
+      "epoch": 0.6005193083255069,
+      "grad_norm": 0.4375,
+      "learning_rate": 8.244529046877336e-05,
+      "loss": 1.2636,
+      "step": 5435
+    },
+    {
+      "epoch": 0.6010717639909398,
+      "grad_norm": 0.4375,
+      "learning_rate": 8.225546399490442e-05,
+      "loss": 1.3465,
+      "step": 5440
+    },
+    {
+      "epoch": 0.6016242196563726,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.206570351771568e-05,
+      "loss": 1.2668,
+      "step": 5445
+    },
+    {
+      "epoch": 0.6021766753218054,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.187600974297714e-05,
+      "loss": 1.2356,
+      "step": 5450
+    },
+    {
+      "epoch": 0.6027291309872382,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.16863833762107e-05,
+      "loss": 1.3551,
+      "step": 5455
+    },
+    {
+      "epoch": 0.6032815866526712,
+      "grad_norm": 0.390625,
+      "learning_rate": 8.149682512268763e-05,
+      "loss": 1.2005,
+      "step": 5460
+    },
+    {
+      "epoch": 0.603834042318104,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.130733568742579e-05,
+      "loss": 1.2591,
+      "step": 5465
+    },
+    {
+      "epoch": 0.6043864979835368,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.111791577518716e-05,
+      "loss": 1.2687,
+      "step": 5470
+    },
+    {
+      "epoch": 0.6049389536489697,
+      "grad_norm": 0.41015625,
+      "learning_rate": 8.092856609047508e-05,
+      "loss": 1.215,
+      "step": 5475
+    },
+    {
+      "epoch": 0.6054914093144025,
+      "grad_norm": 0.400390625,
+      "learning_rate": 8.073928733753175e-05,
+      "loss": 1.313,
+      "step": 5480
+    },
+    {
+      "epoch": 0.6060438649798354,
+      "grad_norm": 0.41015625,
+      "learning_rate": 8.055008022033551e-05,
+      "loss": 1.2682,
+      "step": 5485
+    },
+    {
+      "epoch": 0.6065963206452682,
+      "grad_norm": 0.400390625,
+      "learning_rate": 8.036094544259827e-05,
+      "loss": 1.3116,
+      "step": 5490
+    },
+    {
+      "epoch": 0.6071487763107011,
+      "grad_norm": 0.4453125,
+      "learning_rate": 8.017188370776292e-05,
+      "loss": 1.2642,
+      "step": 5495
+    },
+    {
+      "epoch": 0.6077012319761339,
+      "grad_norm": 0.40625,
+      "learning_rate": 7.998289571900067e-05,
+      "loss": 1.2091,
+      "step": 5500
+    },
+    {
+      "epoch": 0.6082536876415667,
+      "grad_norm": 0.44140625,
+      "learning_rate": 7.979398217920849e-05,
+      "loss": 1.3081,
+      "step": 5505
+    },
+    {
+      "epoch": 0.6088061433069996,
+      "grad_norm": 0.443359375,
+      "learning_rate": 7.960514379100632e-05,
+      "loss": 1.2977,
+      "step": 5510
+    },
+    {
+      "epoch": 0.6093585989724325,
+      "grad_norm": 0.400390625,
+      "learning_rate": 7.941638125673475e-05,
+      "loss": 1.2382,
+      "step": 5515
+    },
+    {
+      "epoch": 0.6099110546378653,
+      "grad_norm": 0.3828125,
+      "learning_rate": 7.922769527845217e-05,
+      "loss": 1.3764,
+      "step": 5520
+    },
+    {
+      "epoch": 0.6104635103032982,
+      "grad_norm": 0.38671875,
+      "learning_rate": 7.903908655793224e-05,
+      "loss": 1.2624,
+      "step": 5525
+    },
+    {
+      "epoch": 0.611015965968731,
+      "grad_norm": 0.38671875,
+      "learning_rate": 7.885055579666133e-05,
+      "loss": 1.2123,
+      "step": 5530
+    },
+    {
+      "epoch": 0.6115684216341638,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.866210369583576e-05,
+      "loss": 1.2908,
+      "step": 5535
+    },
+    {
+      "epoch": 0.6121208772995967,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.847373095635937e-05,
+      "loss": 1.2451,
+      "step": 5540
+    },
+    {
+      "epoch": 0.6126733329650296,
+      "grad_norm": 0.3984375,
+      "learning_rate": 7.82854382788408e-05,
+      "loss": 1.2879,
+      "step": 5545
+    },
+    {
+      "epoch": 0.6132257886304624,
+      "grad_norm": 0.423828125,
+      "learning_rate": 7.809722636359095e-05,
+      "loss": 1.3246,
+      "step": 5550
+    },
+    {
+      "epoch": 0.6137782442958952,
+      "grad_norm": 0.388671875,
+      "learning_rate": 7.790909591062032e-05,
+      "loss": 1.2816,
+      "step": 5555
+    },
+    {
+      "epoch": 0.614330699961328,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.772104761963645e-05,
+      "loss": 1.2925,
+      "step": 5560
+    },
+    {
+      "epoch": 0.614883155626761,
+      "grad_norm": 0.400390625,
+      "learning_rate": 7.753308219004122e-05,
+      "loss": 1.3182,
+      "step": 5565
+    },
+    {
+      "epoch": 0.6154356112921938,
+      "grad_norm": 0.400390625,
+      "learning_rate": 7.734520032092845e-05,
+      "loss": 1.3615,
+      "step": 5570
+    },
+    {
+      "epoch": 0.6159880669576266,
+      "grad_norm": 0.39453125,
+      "learning_rate": 7.715740271108107e-05,
+      "loss": 1.2766,
+      "step": 5575
+    },
+    {
+      "epoch": 0.6165405226230595,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.69696900589687e-05,
+      "loss": 1.2598,
+      "step": 5580
+    },
+    {
+      "epoch": 0.6170929782884923,
+      "grad_norm": 0.38671875,
+      "learning_rate": 7.678206306274495e-05,
+      "loss": 1.3825,
+      "step": 5585
+    },
+    {
+      "epoch": 0.6176454339539252,
+      "grad_norm": 0.40625,
+      "learning_rate": 7.659452242024482e-05,
+      "loss": 1.2055,
+      "step": 5590
+    },
+    {
+      "epoch": 0.6181978896193581,
+      "grad_norm": 0.423828125,
+      "learning_rate": 7.64070688289822e-05,
+      "loss": 1.217,
+      "step": 5595
+    },
+    {
+      "epoch": 0.6187503452847909,
+      "grad_norm": 0.3984375,
+      "learning_rate": 7.621970298614717e-05,
+      "loss": 1.363,
+      "step": 5600
+    },
+    {
+      "epoch": 0.6193028009502237,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.603242558860347e-05,
+      "loss": 1.2693,
+      "step": 5605
+    },
+    {
+      "epoch": 0.6198552566156565,
+      "grad_norm": 0.38671875,
+      "learning_rate": 7.584523733288589e-05,
+      "loss": 1.3201,
+      "step": 5610
+    },
+    {
+      "epoch": 0.6204077122810895,
+      "grad_norm": 0.43359375,
+      "learning_rate": 7.565813891519765e-05,
+      "loss": 1.3271,
+      "step": 5615
+    },
+    {
+      "epoch": 0.6209601679465223,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.547113103140786e-05,
+      "loss": 1.3233,
+      "step": 5620
+    },
+    {
+      "epoch": 0.6215126236119551,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.528421437704892e-05,
+      "loss": 1.2494,
+      "step": 5625
+    },
+    {
+      "epoch": 0.622065079277388,
+      "grad_norm": 0.396484375,
+      "learning_rate": 7.509738964731389e-05,
+      "loss": 1.2124,
+      "step": 5630
+    },
+    {
+      "epoch": 0.6226175349428208,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.491065753705399e-05,
+      "loss": 1.2642,
+      "step": 5635
+    },
+    {
+      "epoch": 0.6231699906082537,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.472401874077592e-05,
+      "loss": 1.3271,
+      "step": 5640
+    },
+    {
+      "epoch": 0.6237224462736866,
+      "grad_norm": 0.41015625,
+      "learning_rate": 7.453747395263931e-05,
+      "loss": 1.2386,
+      "step": 5645
+    },
+    {
+      "epoch": 0.6242749019391194,
+      "grad_norm": 0.39453125,
+      "learning_rate": 7.43510238664542e-05,
+      "loss": 1.2424,
+      "step": 5650
+    },
+    {
+      "epoch": 0.6248273576045522,
+      "grad_norm": 0.40625,
+      "learning_rate": 7.416466917567837e-05,
+      "loss": 1.3081,
+      "step": 5655
+    },
+    {
+      "epoch": 0.625379813269985,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.397841057341479e-05,
+      "loss": 1.2392,
+      "step": 5660
+    },
+    {
+      "epoch": 0.625932268935418,
+      "grad_norm": 0.41015625,
+      "learning_rate": 7.379224875240911e-05,
+      "loss": 1.3145,
+      "step": 5665
+    },
+    {
+      "epoch": 0.6264847246008508,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.360618440504694e-05,
+      "loss": 1.2563,
+      "step": 5670
+    },
+    {
+      "epoch": 0.6270371802662836,
+      "grad_norm": 0.412109375,
+      "learning_rate": 7.342021822335143e-05,
+      "loss": 1.2278,
+      "step": 5675
+    },
+    {
+      "epoch": 0.6275896359317165,
+      "grad_norm": 0.392578125,
+      "learning_rate": 7.323435089898059e-05,
+      "loss": 1.3552,
+      "step": 5680
+    },
+    {
+      "epoch": 0.6281420915971493,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.304858312322475e-05,
+      "loss": 1.205,
+      "step": 5685
+    },
+    {
+      "epoch": 0.6286945472625822,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.2862915587004e-05,
+      "loss": 1.1788,
+      "step": 5690
+    },
+    {
+      "epoch": 0.629247002928015,
+      "grad_norm": 0.396484375,
+      "learning_rate": 7.267734898086564e-05,
+      "loss": 1.3006,
+      "step": 5695
+    },
+    {
+      "epoch": 0.6297994585934479,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.249188399498158e-05,
+      "loss": 1.269,
+      "step": 5700
+    },
+    {
+      "epoch": 0.6303519142588807,
+      "grad_norm": 0.41796875,
+      "learning_rate": 7.230652131914574e-05,
+      "loss": 1.2548,
+      "step": 5705
+    },
+    {
+      "epoch": 0.6309043699243135,
+      "grad_norm": 0.39453125,
+      "learning_rate": 7.21212616427715e-05,
+      "loss": 1.3417,
+      "step": 5710
+    },
+    {
+      "epoch": 0.6314568255897465,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.193610565488924e-05,
+      "loss": 1.314,
+      "step": 5715
+    },
+    {
+      "epoch": 0.6320092812551793,
+      "grad_norm": 0.412109375,
+      "learning_rate": 7.175105404414362e-05,
+      "loss": 1.2951,
+      "step": 5720
+    },
+    {
+      "epoch": 0.6325617369206121,
+      "grad_norm": 0.4140625,
+      "learning_rate": 7.156610749879116e-05,
+      "loss": 1.2535,
+      "step": 5725
+    },
+    {
+      "epoch": 0.633114192586045,
+      "grad_norm": 0.380859375,
+      "learning_rate": 7.138126670669755e-05,
+      "loss": 1.2447,
+      "step": 5730
+    },
+    {
+      "epoch": 0.6336666482514778,
+      "grad_norm": 0.412109375,
+      "learning_rate": 7.11965323553352e-05,
+      "loss": 1.2694,
+      "step": 5735
+    },
+    {
+      "epoch": 0.6342191039169107,
+      "grad_norm": 0.41015625,
+      "learning_rate": 7.10119051317806e-05,
+      "loss": 1.1998,
+      "step": 5740
+    },
+    {
+      "epoch": 0.6347715595823435,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.082738572271185e-05,
+      "loss": 1.1994,
+      "step": 5745
+    },
+    {
+      "epoch": 0.6353240152477764,
+      "grad_norm": 0.4140625,
+      "learning_rate": 7.0642974814406e-05,
+      "loss": 1.2666,
+      "step": 5750
+    },
+    {
+      "epoch": 0.6358764709132092,
+      "grad_norm": 0.416015625,
+      "learning_rate": 7.045867309273664e-05,
+      "loss": 1.297,
+      "step": 5755
+    },
+    {
+      "epoch": 0.636428926578642,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.027448124317119e-05,
+      "loss": 1.2803,
+      "step": 5760
+    },
+    {
+      "epoch": 0.636981382244075,
+      "grad_norm": 0.40625,
+      "learning_rate": 7.009039995076844e-05,
+      "loss": 1.3067,
+      "step": 5765
+    },
+    {
+      "epoch": 0.6375338379095078,
+      "grad_norm": 0.37890625,
+      "learning_rate": 6.990642990017602e-05,
+      "loss": 1.2694,
+      "step": 5770
+    },
+    {
+      "epoch": 0.6380862935749406,
+      "grad_norm": 0.384765625,
+      "learning_rate": 6.97225717756278e-05,
+      "loss": 1.3048,
+      "step": 5775
+    },
+    {
+      "epoch": 0.6386387492403734,
+      "grad_norm": 0.388671875,
+      "learning_rate": 6.953882626094136e-05,
+      "loss": 1.2704,
+      "step": 5780
+    },
+    {
+      "epoch": 0.6391912049058063,
+      "grad_norm": 0.376953125,
+      "learning_rate": 6.93551940395155e-05,
+      "loss": 1.3077,
+      "step": 5785
+    },
+    {
+      "epoch": 0.6397436605712391,
+      "grad_norm": 0.3984375,
+      "learning_rate": 6.917167579432753e-05,
+      "loss": 1.2559,
+      "step": 5790
+    },
+    {
+      "epoch": 0.640296116236672,
+      "grad_norm": 0.392578125,
+      "learning_rate": 6.898827220793103e-05,
+      "loss": 1.3388,
+      "step": 5795
+    },
+    {
+      "epoch": 0.6408485719021049,
+      "grad_norm": 0.390625,
+      "learning_rate": 6.880498396245298e-05,
+      "loss": 1.2705,
+      "step": 5800
+    },
+    {
+      "epoch": 0.6414010275675377,
+      "grad_norm": 0.408203125,
+      "learning_rate": 6.862181173959146e-05,
+      "loss": 1.2947,
+      "step": 5805
+    },
+    {
+      "epoch": 0.6419534832329705,
+      "grad_norm": 0.4375,
+      "learning_rate": 6.843875622061304e-05,
+      "loss": 1.2814,
+      "step": 5810
+    },
+    {
+      "epoch": 0.6425059388984034,
+      "grad_norm": 0.40234375,
+      "learning_rate": 6.825581808635016e-05,
+      "loss": 1.2504,
+      "step": 5815
+    },
+    {
+      "epoch": 0.6430583945638363,
+      "grad_norm": 0.404296875,
+      "learning_rate": 6.80729980171987e-05,
+      "loss": 1.2417,
+      "step": 5820
+    },
+    {
+      "epoch": 0.6436108502292691,
+      "grad_norm": 0.42578125,
+      "learning_rate": 6.789029669311551e-05,
+      "loss": 1.3356,
+      "step": 5825
+    },
+    {
+      "epoch": 0.6441633058947019,
+      "grad_norm": 0.412109375,
+      "learning_rate": 6.770771479361568e-05,
+      "loss": 1.2996,
+      "step": 5830
+    },
+    {
+      "epoch": 0.6447157615601348,
+      "grad_norm": 0.408203125,
+      "learning_rate": 6.752525299777021e-05,
+      "loss": 1.2249,
+      "step": 5835
+    },
+    {
+      "epoch": 0.6452682172255676,
+      "grad_norm": 0.416015625,
+      "learning_rate": 6.734291198420333e-05,
+      "loss": 1.3227,
+      "step": 5840
+    },
+    {
+      "epoch": 0.6458206728910005,
+      "grad_norm": 0.376953125,
+      "learning_rate": 6.716069243109011e-05,
+      "loss": 1.1827,
+      "step": 5845
+    },
+    {
+      "epoch": 0.6463731285564334,
+      "grad_norm": 0.40234375,
+      "learning_rate": 6.697859501615387e-05,
+      "loss": 1.3241,
+      "step": 5850
+    },
+    {
+      "epoch": 0.6469255842218662,
+      "grad_norm": 0.412109375,
+      "learning_rate": 6.679662041666362e-05,
+      "loss": 1.2363,
+      "step": 5855
+    },
+    {
+      "epoch": 0.647478039887299,
+      "grad_norm": 0.400390625,
+      "learning_rate": 6.661476930943163e-05,
+      "loss": 1.2711,
+      "step": 5860
+    },
+    {
+      "epoch": 0.6480304955527318,
+      "grad_norm": 0.40625,
+      "learning_rate": 6.643304237081087e-05,
+      "loss": 1.2842,
+      "step": 5865
+    },
+    {
+      "epoch": 0.6485829512181648,
+      "grad_norm": 0.44140625,
+      "learning_rate": 6.625144027669245e-05,
+      "loss": 1.249,
+      "step": 5870
+    },
+    {
+      "epoch": 0.6491354068835976,
+      "grad_norm": 0.421875,
+      "learning_rate": 6.60699637025032e-05,
+      "loss": 1.2837,
+      "step": 5875
+    },
+    {
+      "epoch": 0.6496878625490304,
+      "grad_norm": 0.41796875,
+      "learning_rate": 6.588861332320306e-05,
+      "loss": 1.3581,
+      "step": 5880
+    },
+    {
+      "epoch": 0.6502403182144633,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.570738981328266e-05,
+      "loss": 1.2256,
+      "step": 5885
+    },
+    {
+      "epoch": 0.6507927738798961,
+      "grad_norm": 0.3828125,
+      "learning_rate": 6.552629384676079e-05,
+      "loss": 1.2203,
+      "step": 5890
+    },
+    {
+      "epoch": 0.651345229545329,
+      "grad_norm": 0.388671875,
+      "learning_rate": 6.534532609718177e-05,
+      "loss": 1.309,
+      "step": 5895
+    },
+    {
+      "epoch": 0.6518976852107619,
+      "grad_norm": 0.404296875,
+      "learning_rate": 6.516448723761315e-05,
+      "loss": 1.2835,
+      "step": 5900
+    },
+    {
+      "epoch": 0.6524501408761947,
+      "grad_norm": 0.400390625,
+      "learning_rate": 6.498377794064303e-05,
+      "loss": 1.3349,
+      "step": 5905
+    },
+    {
+      "epoch": 0.6530025965416275,
+      "grad_norm": 0.419921875,
+      "learning_rate": 6.480319887837771e-05,
+      "loss": 1.2314,
+      "step": 5910
+    },
+    {
+      "epoch": 0.6535550522070603,
+      "grad_norm": 0.458984375,
+      "learning_rate": 6.462275072243908e-05,
+      "loss": 1.3439,
+      "step": 5915
+    },
+    {
+      "epoch": 0.6541075078724933,
+      "grad_norm": 0.412109375,
+      "learning_rate": 6.444243414396208e-05,
+      "loss": 1.2487,
+      "step": 5920
+    },
+    {
+      "epoch": 0.6546599635379261,
+      "grad_norm": 0.40234375,
+      "learning_rate": 6.426224981359238e-05,
+      "loss": 1.2491,
+      "step": 5925
+    },
+    {
+      "epoch": 0.6552124192033589,
+      "grad_norm": 0.38671875,
+      "learning_rate": 6.408219840148375e-05,
+      "loss": 1.2978,
+      "step": 5930
+    },
+    {
+      "epoch": 0.6557648748687918,
+      "grad_norm": 1.2890625,
+      "learning_rate": 6.390228057729557e-05,
+      "loss": 1.2105,
+      "step": 5935
+    },
+    {
+      "epoch": 0.6563173305342246,
+      "grad_norm": 0.384765625,
+      "learning_rate": 6.372249701019045e-05,
+      "loss": 1.2386,
+      "step": 5940
+    },
+    {
+      "epoch": 0.6568697861996575,
+      "grad_norm": 0.416015625,
+      "learning_rate": 6.354284836883156e-05,
+      "loss": 1.2054,
+      "step": 5945
+    },
+    {
+      "epoch": 0.6574222418650904,
+      "grad_norm": 0.416015625,
+      "learning_rate": 6.336333532138032e-05,
+      "loss": 1.1582,
+      "step": 5950
+    },
+    {
+      "epoch": 0.6579746975305232,
+      "grad_norm": 0.421875,
+      "learning_rate": 6.31839585354938e-05,
+      "loss": 1.3335,
+      "step": 5955
+    },
+    {
+      "epoch": 0.658527153195956,
+      "grad_norm": 0.40234375,
+      "learning_rate": 6.300471867832229e-05,
+      "loss": 1.2848,
+      "step": 5960
+    },
+    {
+      "epoch": 0.6590796088613888,
+      "grad_norm": 0.4375,
+      "learning_rate": 6.282561641650682e-05,
+      "loss": 1.3053,
+      "step": 5965
+    },
+    {
+      "epoch": 0.6596320645268218,
+      "grad_norm": 0.3828125,
+      "learning_rate": 6.264665241617666e-05,
+      "loss": 1.2261,
+      "step": 5970
+    },
+    {
+      "epoch": 0.6601845201922546,
+      "grad_norm": 0.458984375,
+      "learning_rate": 6.246782734294683e-05,
+      "loss": 1.3282,
+      "step": 5975
+    },
+    {
+      "epoch": 0.6607369758576874,
+      "grad_norm": 0.384765625,
+      "learning_rate": 6.228914186191563e-05,
+      "loss": 1.2368,
+      "step": 5980
+    },
+    {
+      "epoch": 0.6612894315231203,
+      "grad_norm": 0.4296875,
+      "learning_rate": 6.211059663766224e-05,
+      "loss": 1.3354,
+      "step": 5985
+    },
+    {
+      "epoch": 0.6618418871885531,
+      "grad_norm": 0.4140625,
+      "learning_rate": 6.193219233424414e-05,
+      "loss": 1.2843,
+      "step": 5990
+    },
+    {
+      "epoch": 0.662394342853986,
+      "grad_norm": 0.423828125,
+      "learning_rate": 6.175392961519471e-05,
+      "loss": 1.289,
+      "step": 5995
+    },
+    {
+      "epoch": 0.6629467985194188,
+      "grad_norm": 0.40625,
+      "learning_rate": 6.15758091435207e-05,
+      "loss": 1.266,
+      "step": 6000
+    },
+    {
+      "epoch": 0.6634992541848517,
+      "grad_norm": 0.392578125,
+      "learning_rate": 6.139783158169984e-05,
+      "loss": 1.3227,
+      "step": 6005
+    },
+    {
+      "epoch": 0.6640517098502845,
+      "grad_norm": 0.384765625,
+      "learning_rate": 6.121999759167837e-05,
+      "loss": 1.2908,
+      "step": 6010
+    },
+    {
+      "epoch": 0.6646041655157173,
+      "grad_norm": 0.39453125,
+      "learning_rate": 6.104230783486847e-05,
+      "loss": 1.1879,
+      "step": 6015
+    },
+    {
+      "epoch": 0.6651566211811503,
+      "grad_norm": 0.400390625,
+      "learning_rate": 6.086476297214594e-05,
+      "loss": 1.2822,
+      "step": 6020
+    },
+    {
+      "epoch": 0.6657090768465831,
+      "grad_norm": 0.4140625,
+      "learning_rate": 6.068736366384764e-05,
+      "loss": 1.2844,
+      "step": 6025
+    },
+    {
+      "epoch": 0.6662615325120159,
+      "grad_norm": 0.384765625,
+      "learning_rate": 6.0510110569769095e-05,
+      "loss": 1.1896,
+      "step": 6030
+    },
+    {
+      "epoch": 0.6668139881774487,
+      "grad_norm": 0.42578125,
+      "learning_rate": 6.033300434916203e-05,
+      "loss": 1.3519,
+      "step": 6035
+    },
+    {
+      "epoch": 0.6673664438428816,
+      "grad_norm": 0.412109375,
+      "learning_rate": 6.0156045660731873e-05,
+      "loss": 1.2742,
+      "step": 6040
+    },
+    {
+      "epoch": 0.6679188995083145,
+      "grad_norm": 0.384765625,
+      "learning_rate": 5.9979235162635394e-05,
+      "loss": 1.1947,
+      "step": 6045
+    },
+    {
+      "epoch": 0.6684713551737473,
+      "grad_norm": 0.4140625,
+      "learning_rate": 5.980257351247818e-05,
+      "loss": 1.2286,
+      "step": 6050
+    },
+    {
+      "epoch": 0.6690238108391802,
+      "grad_norm": 0.439453125,
+      "learning_rate": 5.9626061367312166e-05,
+      "loss": 1.2746,
+      "step": 6055
+    },
+    {
+      "epoch": 0.669576266504613,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.9449699383633316e-05,
+      "loss": 1.2367,
+      "step": 6060
+    },
+    {
+      "epoch": 0.6701287221700458,
+      "grad_norm": 0.416015625,
+      "learning_rate": 5.927348821737906e-05,
+      "loss": 1.3109,
+      "step": 6065
+    },
+    {
+      "epoch": 0.6706811778354786,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.909742852392587e-05,
+      "loss": 1.2512,
+      "step": 6070
+    },
+    {
+      "epoch": 0.6712336335009116,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.8921520958086905e-05,
+      "loss": 1.2713,
+      "step": 6075
+    },
+    {
+      "epoch": 0.6717860891663444,
+      "grad_norm": 0.41796875,
+      "learning_rate": 5.8745766174109495e-05,
+      "loss": 1.341,
+      "step": 6080
+    },
+    {
+      "epoch": 0.6723385448317772,
+      "grad_norm": 0.396484375,
+      "learning_rate": 5.857016482567275e-05,
+      "loss": 1.3013,
+      "step": 6085
+    },
+    {
+      "epoch": 0.6728910004972101,
+      "grad_norm": 0.423828125,
+      "learning_rate": 5.8394717565885106e-05,
+      "loss": 1.3698,
+      "step": 6090
+    },
+    {
+      "epoch": 0.6734434561626429,
+      "grad_norm": 0.412109375,
+      "learning_rate": 5.821942504728183e-05,
+      "loss": 1.2323,
+      "step": 6095
+    },
+    {
+      "epoch": 0.6739959118280758,
+      "grad_norm": 0.412109375,
+      "learning_rate": 5.804428792182279e-05,
+      "loss": 1.2904,
+      "step": 6100
+    },
+    {
+      "epoch": 0.6745483674935087,
+      "grad_norm": 0.412109375,
+      "learning_rate": 5.786930684088988e-05,
+      "loss": 1.3443,
+      "step": 6105
+    },
+    {
+      "epoch": 0.6751008231589415,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.7694482455284504e-05,
+      "loss": 1.2083,
+      "step": 6110
+    },
+    {
+      "epoch": 0.6756532788243743,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.751981541522539e-05,
+      "loss": 1.1749,
+      "step": 6115
+    },
+    {
+      "epoch": 0.6762057344898071,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.734530637034603e-05,
+      "loss": 1.3391,
+      "step": 6120
+    },
+    {
+      "epoch": 0.6767581901552401,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.7170955969692265e-05,
+      "loss": 1.2439,
+      "step": 6125
+    },
+    {
+      "epoch": 0.6773106458206729,
+      "grad_norm": 0.400390625,
+      "learning_rate": 5.699676486171994e-05,
+      "loss": 1.1964,
+      "step": 6130
+    },
+    {
+      "epoch": 0.6778631014861057,
+      "grad_norm": 0.412109375,
+      "learning_rate": 5.6822733694292427e-05,
+      "loss": 1.2591,
+      "step": 6135
+    },
+    {
+      "epoch": 0.6784155571515386,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.664886311467821e-05,
+      "loss": 1.2042,
+      "step": 6140
+    },
+    {
+      "epoch": 0.6789680128169714,
+      "grad_norm": 0.4375,
+      "learning_rate": 5.647515376954852e-05,
+      "loss": 1.2,
+      "step": 6145
+    },
+    {
+      "epoch": 0.6795204684824043,
+      "grad_norm": 0.416015625,
+      "learning_rate": 5.630160630497493e-05,
+      "loss": 1.2134,
+      "step": 6150
+    },
+    {
+      "epoch": 0.6800729241478372,
+      "grad_norm": 0.40625,
+      "learning_rate": 5.612822136642697e-05,
+      "loss": 1.2775,
+      "step": 6155
+    },
+    {
+      "epoch": 0.68062537981327,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.5954999598769575e-05,
+      "loss": 1.2292,
+      "step": 6160
+    },
+    {
+      "epoch": 0.6811778354787028,
+      "grad_norm": 0.41015625,
+      "learning_rate": 5.578194164626089e-05,
+      "loss": 1.2832,
+      "step": 6165
+    },
+    {
+      "epoch": 0.6817302911441356,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.5609048152549794e-05,
+      "loss": 1.2754,
+      "step": 6170
+    },
+    {
+      "epoch": 0.6822827468095686,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.543631976067345e-05,
+      "loss": 1.2711,
+      "step": 6175
+    },
+    {
+      "epoch": 0.6828352024750014,
+      "grad_norm": 0.396484375,
+      "learning_rate": 5.526375711305504e-05,
+      "loss": 1.292,
+      "step": 6180
+    },
+    {
+      "epoch": 0.6833876581404342,
+      "grad_norm": 0.408203125,
+      "learning_rate": 5.509136085150122e-05,
+      "loss": 1.2672,
+      "step": 6185
+    },
+    {
+      "epoch": 0.6839401138058671,
+      "grad_norm": 0.41796875,
+      "learning_rate": 5.491913161719984e-05,
+      "loss": 1.2646,
+      "step": 6190
+    },
+    {
+      "epoch": 0.6844925694712999,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.4747070050717556e-05,
+      "loss": 1.2929,
+      "step": 6195
+    },
+    {
+      "epoch": 0.6850450251367328,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.457517679199736e-05,
+      "loss": 1.3068,
+      "step": 6200
+    },
+    {
+      "epoch": 0.6855974808021656,
+      "grad_norm": 0.400390625,
+      "learning_rate": 5.4403452480356346e-05,
+      "loss": 1.2772,
+      "step": 6205
+    },
+    {
+      "epoch": 0.6861499364675985,
+      "grad_norm": 0.41015625,
+      "learning_rate": 5.423189775448323e-05,
+      "loss": 1.3216,
+      "step": 6210
+    },
+    {
+      "epoch": 0.6867023921330313,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.406051325243586e-05,
+      "loss": 1.1877,
+      "step": 6215
+    },
+    {
+      "epoch": 0.6872548477984641,
+      "grad_norm": 0.427734375,
+      "learning_rate": 5.3889299611639174e-05,
+      "loss": 1.2715,
+      "step": 6220
+    },
+    {
+      "epoch": 0.6878073034638971,
+      "grad_norm": 0.416015625,
+      "learning_rate": 5.371825746888251e-05,
+      "loss": 1.3154,
+      "step": 6225
+    },
+    {
+      "epoch": 0.6883597591293299,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.35473874603174e-05,
+      "loss": 1.2977,
+      "step": 6230
+    },
+    {
+      "epoch": 0.6889122147947627,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.337669022145515e-05,
+      "loss": 1.2979,
+      "step": 6235
+    },
+    {
+      "epoch": 0.6894646704601956,
+      "grad_norm": 0.421875,
+      "learning_rate": 5.320616638716448e-05,
+      "loss": 1.3065,
+      "step": 6240
+    },
+    {
+      "epoch": 0.6900171261256284,
+      "grad_norm": 0.431640625,
+      "learning_rate": 5.3035816591669205e-05,
+      "loss": 1.3274,
+      "step": 6245
+    },
+    {
+      "epoch": 0.6905695817910613,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.286564146854581e-05,
+      "loss": 1.2623,
+      "step": 6250
+    },
+    {
+      "epoch": 0.6911220374564941,
+      "grad_norm": 0.408203125,
+      "learning_rate": 5.269564165072115e-05,
+      "loss": 1.2494,
+      "step": 6255
+    },
+    {
+      "epoch": 0.691674493121927,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.2525817770470084e-05,
+      "loss": 1.2287,
+      "step": 6260
+    },
+    {
+      "epoch": 0.6922269487873598,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.2356170459413035e-05,
+      "loss": 1.2245,
+      "step": 6265
+    },
+    {
+      "epoch": 0.6927794044527926,
+      "grad_norm": 0.408203125,
+      "learning_rate": 5.2186700348513786e-05,
+      "loss": 1.3329,
+      "step": 6270
+    },
+    {
+      "epoch": 0.6933318601182256,
+      "grad_norm": 0.408203125,
+      "learning_rate": 5.2017408068077064e-05,
+      "loss": 1.2252,
+      "step": 6275
+    },
+    {
+      "epoch": 0.6938843157836584,
+      "grad_norm": 0.380859375,
+      "learning_rate": 5.18482942477462e-05,
+      "loss": 1.172,
+      "step": 6280
+    },
+    {
+      "epoch": 0.6944367714490912,
+      "grad_norm": 0.4140625,
+      "learning_rate": 5.1679359516500735e-05,
+      "loss": 1.2943,
+      "step": 6285
+    },
+    {
+      "epoch": 0.694989227114524,
+      "grad_norm": 0.41015625,
+      "learning_rate": 5.151060450265419e-05,
+      "loss": 1.3527,
+      "step": 6290
+    },
+    {
+      "epoch": 0.6955416827799569,
+      "grad_norm": 0.40625,
+      "learning_rate": 5.1342029833851634e-05,
+      "loss": 1.2435,
+      "step": 6295
+    },
+    {
+      "epoch": 0.6960941384453898,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.1173636137067406e-05,
+      "loss": 1.2147,
+      "step": 6300
+    },
+    {
+      "epoch": 0.6966465941108226,
+      "grad_norm": 0.41015625,
+      "learning_rate": 5.1005424038602724e-05,
+      "loss": 1.2178,
+      "step": 6305
+    },
+    {
+      "epoch": 0.6971990497762555,
+      "grad_norm": 0.41015625,
+      "learning_rate": 5.083739416408343e-05,
+      "loss": 1.2232,
+      "step": 6310
+    },
+    {
+      "epoch": 0.6977515054416883,
+      "grad_norm": 0.408203125,
+      "learning_rate": 5.066954713845766e-05,
+      "loss": 1.2486,
+      "step": 6315
+    },
+    {
+      "epoch": 0.6983039611071211,
+      "grad_norm": 0.41015625,
+      "learning_rate": 5.050188358599335e-05,
+      "loss": 1.2738,
+      "step": 6320
+    },
+    {
+      "epoch": 0.6988564167725541,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.033440413027619e-05,
+      "loss": 1.2159,
+      "step": 6325
+    },
+    {
+      "epoch": 0.6994088724379869,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.01671093942071e-05,
+      "loss": 1.1586,
+      "step": 6330
+    },
+    {
+      "epoch": 0.6999613281034197,
+      "grad_norm": 0.41015625,
+      "learning_rate": 5.000000000000002e-05,
+      "loss": 1.2886,
+      "step": 6335
+    },
+    {
+      "epoch": 0.7005137837688525,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.9833076569179506e-05,
+      "loss": 1.2738,
+      "step": 6340
+    },
+    {
+      "epoch": 0.7010662394342854,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4.9666339722578494e-05,
+      "loss": 1.2559,
+      "step": 6345
+    },
+    {
+      "epoch": 0.7016186950997182,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.949979008033596e-05,
+      "loss": 1.2716,
+      "step": 6350
+    },
+    {
+      "epoch": 0.7021711507651511,
+      "grad_norm": 0.38671875,
+      "learning_rate": 4.93334282618946e-05,
+      "loss": 1.2993,
+      "step": 6355
+    },
+    {
+      "epoch": 0.702723606430584,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.9167254885998584e-05,
+      "loss": 1.2113,
+      "step": 6360
+    },
+    {
+      "epoch": 0.7032760620960168,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.900127057069116e-05,
+      "loss": 1.3193,
+      "step": 6365
+    },
+    {
+      "epoch": 0.7038285177614496,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.883547593331248e-05,
+      "loss": 1.2175,
+      "step": 6370
+    },
+    {
+      "epoch": 0.7043809734268824,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.866987159049713e-05,
+      "loss": 1.2761,
+      "step": 6375
+    },
+    {
+      "epoch": 0.7049334290923154,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.850445815817202e-05,
+      "loss": 1.2263,
+      "step": 6380
+    },
+    {
+      "epoch": 0.7054858847577482,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.833923625155399e-05,
+      "loss": 1.3204,
+      "step": 6385
+    },
+    {
+      "epoch": 0.706038340423181,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4.817420648514755e-05,
+      "loss": 1.234,
+      "step": 6390
+    },
+    {
+      "epoch": 0.7065907960886139,
+      "grad_norm": 0.390625,
+      "learning_rate": 4.800936947274255e-05,
+      "loss": 1.2804,
+      "step": 6395
+    },
+    {
+      "epoch": 0.7071432517540467,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.7844725827412054e-05,
+      "loss": 1.3185,
+      "step": 6400
+    },
+    {
+      "epoch": 0.7076957074194796,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.7680276161509795e-05,
+      "loss": 1.1644,
+      "step": 6405
+    },
+    {
+      "epoch": 0.7082481630849125,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.751602108666818e-05,
+      "loss": 1.2911,
+      "step": 6410
+    },
+    {
+      "epoch": 0.7088006187503453,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.735196121379571e-05,
+      "loss": 1.2112,
+      "step": 6415
+    },
+    {
+      "epoch": 0.7093530744157781,
+      "grad_norm": 0.390625,
+      "learning_rate": 4.7188097153075017e-05,
+      "loss": 1.2008,
+      "step": 6420
+    },
+    {
+      "epoch": 0.7099055300812109,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.7024429513960425e-05,
+      "loss": 1.2918,
+      "step": 6425
+    },
+    {
+      "epoch": 0.7104579857466439,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.686095890517569e-05,
+      "loss": 1.2994,
+      "step": 6430
+    },
+    {
+      "epoch": 0.7110104414120767,
+      "grad_norm": 0.3671875,
+      "learning_rate": 4.6697685934711785e-05,
+      "loss": 1.1452,
+      "step": 6435
+    },
+    {
+      "epoch": 0.7115628970775095,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.65346112098246e-05,
+      "loss": 1.2731,
+      "step": 6440
+    },
+    {
+      "epoch": 0.7121153527429424,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.637173533703267e-05,
+      "loss": 1.2484,
+      "step": 6445
+    },
+    {
+      "epoch": 0.7126678084083752,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.6209058922115015e-05,
+      "loss": 1.3127,
+      "step": 6450
+    },
+    {
+      "epoch": 0.7132202640738081,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.6046582570108744e-05,
+      "loss": 1.2954,
+      "step": 6455
+    },
+    {
+      "epoch": 0.713772719739241,
+      "grad_norm": 0.388671875,
+      "learning_rate": 4.588430688530696e-05,
+      "loss": 1.2205,
+      "step": 6460
+    },
+    {
+      "epoch": 0.7143251754046738,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.5722232471256296e-05,
+      "loss": 1.2443,
+      "step": 6465
+    },
+    {
+      "epoch": 0.7148776310701066,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.556035993075495e-05,
+      "loss": 1.3188,
+      "step": 6470
+    },
+    {
+      "epoch": 0.7154300867355394,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.539868986585022e-05,
+      "loss": 1.3274,
+      "step": 6475
+    },
+    {
+      "epoch": 0.7159825424009724,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.523722287783636e-05,
+      "loss": 1.2691,
+      "step": 6480
+    },
+    {
+      "epoch": 0.7165349980664052,
+      "grad_norm": 0.41015625,
+      "learning_rate": 4.5075959567252335e-05,
+      "loss": 1.3686,
+      "step": 6485
+    },
+    {
+      "epoch": 0.717087453731838,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.491490053387958e-05,
+      "loss": 1.2057,
+      "step": 6490
+    },
+    {
+      "epoch": 0.7176399093972708,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.475404637673974e-05,
+      "loss": 1.2386,
+      "step": 6495
+    },
+    {
+      "epoch": 0.7181923650627037,
+      "grad_norm": 0.416015625,
+      "learning_rate": 4.459339769409252e-05,
+      "loss": 1.241,
+      "step": 6500
+    },
+    {
+      "epoch": 0.7187448207281366,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.443295508343336e-05,
+      "loss": 1.2884,
+      "step": 6505
+    },
+    {
+      "epoch": 0.7192972763935694,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.427271914149128e-05,
+      "loss": 1.3309,
+      "step": 6510
+    },
+    {
+      "epoch": 0.7198497320590023,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.41126904642267e-05,
+      "loss": 1.2217,
+      "step": 6515
+    },
+    {
+      "epoch": 0.7204021877244351,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.395286964682903e-05,
+      "loss": 1.2936,
+      "step": 6520
+    },
+    {
+      "epoch": 0.7209546433898679,
+      "grad_norm": 0.404296875,
+      "learning_rate": 4.379325728371473e-05,
+      "loss": 1.2449,
+      "step": 6525
+    },
+    {
+      "epoch": 0.7215070990553009,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.363385396852491e-05,
+      "loss": 1.2352,
+      "step": 6530
+    },
+    {
+      "epoch": 0.7220595547207337,
+      "grad_norm": 0.404296875,
+      "learning_rate": 4.347466029412316e-05,
+      "loss": 1.3162,
+      "step": 6535
+    },
+    {
+      "epoch": 0.7226120103861665,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.331567685259338e-05,
+      "loss": 1.2229,
+      "step": 6540
+    },
+    {
+      "epoch": 0.7231644660515993,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.3156904235237574e-05,
+      "loss": 1.288,
+      "step": 6545
+    },
+    {
+      "epoch": 0.7237169217170322,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.2998343032573596e-05,
+      "loss": 1.3225,
+      "step": 6550
+    },
+    {
+      "epoch": 0.7242693773824651,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.2839993834333014e-05,
+      "loss": 1.3295,
+      "step": 6555
+    },
+    {
+      "epoch": 0.7248218330478979,
+      "grad_norm": 0.416015625,
+      "learning_rate": 4.2681857229458885e-05,
+      "loss": 1.2258,
+      "step": 6560
+    },
+    {
+      "epoch": 0.7253742887133308,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.25239338061036e-05,
+      "loss": 1.3371,
+      "step": 6565
+    },
+    {
+      "epoch": 0.7259267443787636,
+      "grad_norm": 0.390625,
+      "learning_rate": 4.236622415162668e-05,
+      "loss": 1.3193,
+      "step": 6570
+    },
+    {
+      "epoch": 0.7264792000441964,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.220872885259247e-05,
+      "loss": 1.2166,
+      "step": 6575
+    },
+    {
+      "epoch": 0.7270316557096294,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.20514484947682e-05,
+      "loss": 1.204,
+      "step": 6580
+    },
+    {
+      "epoch": 0.7275841113750622,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.189438366312162e-05,
+      "loss": 1.3055,
+      "step": 6585
+    },
+    {
+      "epoch": 0.728136567040495,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.17375349418189e-05,
+      "loss": 1.2411,
+      "step": 6590
+    },
+    {
+      "epoch": 0.7286890227059278,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.158090291422243e-05,
+      "loss": 1.2973,
+      "step": 6595
+    },
+    {
+      "epoch": 0.7292414783713607,
+      "grad_norm": 0.404296875,
+      "learning_rate": 4.142448816288864e-05,
+      "loss": 1.1524,
+      "step": 6600
+    },
+    {
+      "epoch": 0.7297939340367936,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.1268291269565885e-05,
+      "loss": 1.2506,
+      "step": 6605
+    },
+    {
+      "epoch": 0.7303463897022264,
+      "grad_norm": 0.3828125,
+      "learning_rate": 4.111231281519222e-05,
+      "loss": 1.2727,
+      "step": 6610
+    },
+    {
+      "epoch": 0.7308988453676593,
+      "grad_norm": 0.376953125,
+      "learning_rate": 4.095655337989329e-05,
+      "loss": 1.2792,
+      "step": 6615
+    },
+    {
+      "epoch": 0.7314513010330921,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.080101354298016e-05,
+      "loss": 1.2532,
+      "step": 6620
+    },
+    {
+      "epoch": 0.7320037566985249,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.0645693882947046e-05,
+      "loss": 1.2731,
+      "step": 6625
+    },
+    {
+      "epoch": 0.7325562123639577,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.0490594977469406e-05,
+      "loss": 1.226,
+      "step": 6630
+    },
+    {
+      "epoch": 0.7331086680293907,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.0335717403401576e-05,
+      "loss": 1.3171,
+      "step": 6635
+    },
+    {
+      "epoch": 0.7336611236948235,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.018106173677473e-05,
+      "loss": 1.2098,
+      "step": 6640
+    },
+    {
+      "epoch": 0.7342135793602563,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.00266285527947e-05,
+      "loss": 1.2317,
+      "step": 6645
+    },
+    {
+      "epoch": 0.7347660350256892,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.987241842583983e-05,
+      "loss": 1.2911,
+      "step": 6650
+    },
+    {
+      "epoch": 0.735318490691122,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.971843192945889e-05,
+      "loss": 1.2999,
+      "step": 6655
+    },
+    {
+      "epoch": 0.7358709463565549,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.9564669636368866e-05,
+      "loss": 1.3028,
+      "step": 6660
+    },
+    {
+      "epoch": 0.7364234020219877,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.9411132118452896e-05,
+      "loss": 1.2787,
+      "step": 6665
+    },
+    {
+      "epoch": 0.7369758576874206,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.9257819946758135e-05,
+      "loss": 1.2632,
+      "step": 6670
+    },
+    {
+      "epoch": 0.7375283133528534,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.910473369149361e-05,
+      "loss": 1.257,
+      "step": 6675
+    },
+    {
+      "epoch": 0.7380807690182862,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.895187392202804e-05,
+      "loss": 1.318,
+      "step": 6680
+    },
+    {
+      "epoch": 0.7386332246837192,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.8799241206887836e-05,
+      "loss": 1.2326,
+      "step": 6685
+    },
+    {
+      "epoch": 0.739185680349152,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.864683611375497e-05,
+      "loss": 1.2301,
+      "step": 6690
+    },
+    {
+      "epoch": 0.7397381360145848,
+      "grad_norm": 0.38671875,
+      "learning_rate": 3.849465920946475e-05,
+      "loss": 1.2722,
+      "step": 6695
+    },
+    {
+      "epoch": 0.7402905916800177,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.834271106000385e-05,
+      "loss": 1.2496,
+      "step": 6700
+    },
+    {
+      "epoch": 0.7408430473454505,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.819099223050813e-05,
+      "loss": 1.2122,
+      "step": 6705
+    },
+    {
+      "epoch": 0.7413955030108834,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.8039503285260506e-05,
+      "loss": 1.2146,
+      "step": 6710
+    },
+    {
+      "epoch": 0.7419479586763162,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.788824478768893e-05,
+      "loss": 1.228,
+      "step": 6715
+    },
+    {
+      "epoch": 0.7425004143417491,
+      "grad_norm": 0.3828125,
+      "learning_rate": 3.773721730036426e-05,
+      "loss": 1.2635,
+      "step": 6720
+    },
+    {
+      "epoch": 0.7430528700071819,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.758642138499819e-05,
+      "loss": 1.2543,
+      "step": 6725
+    },
+    {
+      "epoch": 0.7436053256726147,
+      "grad_norm": 0.37109375,
+      "learning_rate": 3.743585760244104e-05,
+      "loss": 1.2001,
+      "step": 6730
+    },
+    {
+      "epoch": 0.7441577813380477,
+      "grad_norm": 0.388671875,
+      "learning_rate": 3.728552651267985e-05,
+      "loss": 1.2411,
+      "step": 6735
+    },
+    {
+      "epoch": 0.7447102370034805,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.7135428674836184e-05,
+      "loss": 1.2956,
+      "step": 6740
+    },
+    {
+      "epoch": 0.7452626926689133,
+      "grad_norm": 0.388671875,
+      "learning_rate": 3.698556464716411e-05,
+      "loss": 1.2976,
+      "step": 6745
+    },
+    {
+      "epoch": 0.7458151483343461,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.683593498704801e-05,
+      "loss": 1.2107,
+      "step": 6750
+    },
+    {
+      "epoch": 0.746367603999779,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.6686540251000756e-05,
+      "loss": 1.3122,
+      "step": 6755
+    },
+    {
+      "epoch": 0.7469200596652119,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.6537380994661295e-05,
+      "loss": 1.2732,
+      "step": 6760
+    },
+    {
+      "epoch": 0.7474725153306447,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.638845777279286e-05,
+      "loss": 1.2454,
+      "step": 6765
+    },
+    {
+      "epoch": 0.7480249709960776,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.623977113928081e-05,
+      "loss": 1.2426,
+      "step": 6770
+    },
+    {
+      "epoch": 0.7485774266615104,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.6091321647130484e-05,
+      "loss": 1.2851,
+      "step": 6775
+    },
+    {
+      "epoch": 0.7491298823269432,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.59431098484653e-05,
+      "loss": 1.216,
+      "step": 6780
+    },
+    {
+      "epoch": 0.7496823379923762,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.579513629452464e-05,
+      "loss": 1.2255,
+      "step": 6785
+    },
+    {
+      "epoch": 0.750234793657809,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.564740153566176e-05,
+      "loss": 1.2575,
+      "step": 6790
+    },
+    {
+      "epoch": 0.7507872493232418,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.5499906121341785e-05,
+      "loss": 1.2037,
+      "step": 6795
+    },
+    {
+      "epoch": 0.7513397049886746,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.535265060013965e-05,
+      "loss": 1.2699,
+      "step": 6800
+    },
+    {
+      "epoch": 0.7518921606541075,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.520563551973806e-05,
+      "loss": 1.2491,
+      "step": 6805
+    },
+    {
+      "epoch": 0.7524446163195404,
+      "grad_norm": 0.458984375,
+      "learning_rate": 3.5058861426925447e-05,
+      "loss": 1.2921,
+      "step": 6810
+    },
+    {
+      "epoch": 0.7529970719849732,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.491232886759398e-05,
+      "loss": 1.2566,
+      "step": 6815
+    },
+    {
+      "epoch": 0.7535495276504061,
+      "grad_norm": 0.38671875,
+      "learning_rate": 3.4766038386737506e-05,
+      "loss": 1.303,
+      "step": 6820
+    },
+    {
+      "epoch": 0.7541019833158389,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.461999052844942e-05,
+      "loss": 1.3304,
+      "step": 6825
+    },
+    {
+      "epoch": 0.7546544389812717,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.447418583592084e-05,
+      "loss": 1.2985,
+      "step": 6830
+    },
+    {
+      "epoch": 0.7552068946467047,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.432862485143846e-05,
+      "loss": 1.2463,
+      "step": 6835
+    },
+    {
+      "epoch": 0.7557593503121375,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.418330811638255e-05,
+      "loss": 1.311,
+      "step": 6840
+    },
+    {
+      "epoch": 0.7563118059775703,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.4038236171224946e-05,
+      "loss": 1.2148,
+      "step": 6845
+    },
+    {
+      "epoch": 0.7568642616430031,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.389340955552707e-05,
+      "loss": 1.2363,
+      "step": 6850
+    },
+    {
+      "epoch": 0.757416717308436,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.374882880793785e-05,
+      "loss": 1.2539,
+      "step": 6855
+    },
+    {
+      "epoch": 0.7579691729738689,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.360449446619183e-05,
+      "loss": 1.2374,
+      "step": 6860
+    },
+    {
+      "epoch": 0.7585216286393017,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.346040706710705e-05,
+      "loss": 1.2242,
+      "step": 6865
+    },
+    {
+      "epoch": 0.7590740843047346,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.331656714658313e-05,
+      "loss": 1.283,
+      "step": 6870
+    },
+    {
+      "epoch": 0.7596265399701674,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.317297523959927e-05,
+      "loss": 1.3228,
+      "step": 6875
+    },
+    {
+      "epoch": 0.7601789956356002,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.302963188021216e-05,
+      "loss": 1.2234,
+      "step": 6880
+    },
+    {
+      "epoch": 0.7607314513010331,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.2886537601554165e-05,
+      "loss": 1.2781,
+      "step": 6885
+    },
+    {
+      "epoch": 0.761283906966466,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.274369293583121e-05,
+      "loss": 1.2353,
+      "step": 6890
+    },
+    {
+      "epoch": 0.7618363626318988,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.260109841432085e-05,
+      "loss": 1.238,
+      "step": 6895
+    },
+    {
+      "epoch": 0.7623888182973316,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.24587545673703e-05,
+      "loss": 1.2044,
+      "step": 6900
+    },
+    {
+      "epoch": 0.7629412739627645,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.231666192439442e-05,
+      "loss": 1.2474,
+      "step": 6905
+    },
+    {
+      "epoch": 0.7634937296281973,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.217482101387381e-05,
+      "loss": 1.3326,
+      "step": 6910
+    },
+    {
+      "epoch": 0.7640461852936302,
+      "grad_norm": 0.462890625,
+      "learning_rate": 3.203323236335277e-05,
+      "loss": 1.2619,
+      "step": 6915
+    },
+    {
+      "epoch": 0.764598640959063,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.189189649943743e-05,
+      "loss": 1.3013,
+      "step": 6920
+    },
+    {
+      "epoch": 0.7651510966244959,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.17508139477937e-05,
+      "loss": 1.2109,
+      "step": 6925
+    },
+    {
+      "epoch": 0.7657035522899287,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.1609985233145334e-05,
+      "loss": 1.2275,
+      "step": 6930
+    },
+    {
+      "epoch": 0.7662560079553615,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.146941087927203e-05,
+      "loss": 1.3199,
+      "step": 6935
+    },
+    {
+      "epoch": 0.7668084636207945,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.132909140900746e-05,
+      "loss": 1.2437,
+      "step": 6940
+    },
+    {
+      "epoch": 0.7673609192862273,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.118902734423731e-05,
+      "loss": 1.2415,
+      "step": 6945
+    },
+    {
+      "epoch": 0.7679133749516601,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.104921920589733e-05,
+      "loss": 1.2964,
+      "step": 6950
+    },
+    {
+      "epoch": 0.768465830617093,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.0909667513971396e-05,
+      "loss": 1.2566,
+      "step": 6955
+    },
+    {
+      "epoch": 0.7690182862825258,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.077037278748965e-05,
+      "loss": 1.3033,
+      "step": 6960
+    },
+    {
+      "epoch": 0.7695707419479587,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.063133554452645e-05,
+      "loss": 1.212,
+      "step": 6965
+    },
+    {
+      "epoch": 0.7701231976133915,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.0492556302198526e-05,
+      "loss": 1.2162,
+      "step": 6970
+    },
+    {
+      "epoch": 0.7706756532788244,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.0354035576663043e-05,
+      "loss": 1.3013,
+      "step": 6975
+    },
+    {
+      "epoch": 0.7712281089442572,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.0215773883115706e-05,
+      "loss": 1.2962,
+      "step": 6980
+    },
+    {
+      "epoch": 0.77178056460969,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.0077771735788684e-05,
+      "loss": 1.2916,
+      "step": 6985
+    },
+    {
+      "epoch": 0.772333020275123,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.9940029647948963e-05,
+      "loss": 1.2644,
+      "step": 6990
+    },
+    {
+      "epoch": 0.7728854759405558,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.9802548131896236e-05,
+      "loss": 1.3068,
+      "step": 6995
+    },
+    {
+      "epoch": 0.7734379316059886,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.9665327698961077e-05,
+      "loss": 1.3421,
+      "step": 7000
+    },
+    {
+      "epoch": 0.7739903872714214,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.9528368859502996e-05,
+      "loss": 1.2997,
+      "step": 7005
+    },
+    {
+      "epoch": 0.7745428429368543,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.93916721229086e-05,
+      "loss": 1.1679,
+      "step": 7010
+    },
+    {
+      "epoch": 0.7750952986022872,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.9255237997589657e-05,
+      "loss": 1.2539,
+      "step": 7015
+    },
+    {
+      "epoch": 0.77564775426772,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.9119066990981193e-05,
+      "loss": 1.1873,
+      "step": 7020
+    },
+    {
+      "epoch": 0.7762002099331529,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.8983159609539635e-05,
+      "loss": 1.2057,
+      "step": 7025
+    },
+    {
+      "epoch": 0.7767526655985857,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.884751635874098e-05,
+      "loss": 1.2382,
+      "step": 7030
+    },
+    {
+      "epoch": 0.7773051212640185,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.8712137743078695e-05,
+      "loss": 1.1869,
+      "step": 7035
+    },
+    {
+      "epoch": 0.7778575769294515,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.8577024266062146e-05,
+      "loss": 1.2406,
+      "step": 7040
+    },
+    {
+      "epoch": 0.7784100325948843,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.844217643021454e-05,
+      "loss": 1.2458,
+      "step": 7045
+    },
+    {
+      "epoch": 0.7789624882603171,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.8307594737071052e-05,
+      "loss": 1.2471,
+      "step": 7050
+    },
+    {
+      "epoch": 0.7795149439257499,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.8173279687177057e-05,
+      "loss": 1.2669,
+      "step": 7055
+    },
+    {
+      "epoch": 0.7800673995911828,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.8039231780086183e-05,
+      "loss": 1.2655,
+      "step": 7060
+    },
+    {
+      "epoch": 0.7806198552566157,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.7905451514358472e-05,
+      "loss": 1.2019,
+      "step": 7065
+    },
+    {
+      "epoch": 0.7811723109220485,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.7771939387558554e-05,
+      "loss": 1.2506,
+      "step": 7070
+    },
+    {
+      "epoch": 0.7817247665874814,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.7638695896253774e-05,
+      "loss": 1.2125,
+      "step": 7075
+    },
+    {
+      "epoch": 0.7822772222529142,
+      "grad_norm": 0.458984375,
+      "learning_rate": 2.7505721536012353e-05,
+      "loss": 1.2211,
+      "step": 7080
+    },
+    {
+      "epoch": 0.782829677918347,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.7373016801401576e-05,
+      "loss": 1.2609,
+      "step": 7085
+    },
+    {
+      "epoch": 0.78338213358378,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.7240582185985798e-05,
+      "loss": 1.3198,
+      "step": 7090
+    },
+    {
+      "epoch": 0.7839345892492128,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.7108418182324857e-05,
+      "loss": 1.2867,
+      "step": 7095
+    },
+    {
+      "epoch": 0.7844870449146456,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.6976525281972078e-05,
+      "loss": 1.2847,
+      "step": 7100
+    },
+    {
+      "epoch": 0.7850395005800784,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.68449039754724e-05,
+      "loss": 1.3226,
+      "step": 7105
+    },
+    {
+      "epoch": 0.7855919562455113,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.6713554752360802e-05,
+      "loss": 1.2922,
+      "step": 7110
+    },
+    {
+      "epoch": 0.7861444119109442,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.6582478101160167e-05,
+      "loss": 1.2294,
+      "step": 7115
+    },
+    {
+      "epoch": 0.786696867576377,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.6451674509379643e-05,
+      "loss": 1.2535,
+      "step": 7120
+    },
+    {
+      "epoch": 0.7872493232418099,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.632114446351286e-05,
+      "loss": 1.3219,
+      "step": 7125
+    },
+    {
+      "epoch": 0.7878017789072427,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.619088844903592e-05,
+      "loss": 1.2972,
+      "step": 7130
+    },
+    {
+      "epoch": 0.7883542345726755,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.606090695040586e-05,
+      "loss": 1.3204,
+      "step": 7135
+    },
+    {
+      "epoch": 0.7889066902381084,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.5931200451058678e-05,
+      "loss": 1.2019,
+      "step": 7140
+    },
+    {
+      "epoch": 0.7894591459035413,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.5801769433407565e-05,
+      "loss": 1.2714,
+      "step": 7145
+    },
+    {
+      "epoch": 0.7900116015689741,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.567261437884112e-05,
+      "loss": 1.1561,
+      "step": 7150
+    },
+    {
+      "epoch": 0.7905640572344069,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.5543735767721576e-05,
+      "loss": 1.2677,
+      "step": 7155
+    },
+    {
+      "epoch": 0.7911165128998398,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.5415134079383006e-05,
+      "loss": 1.2254,
+      "step": 7160
+    },
+    {
+      "epoch": 0.7916689685652727,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.5286809792129496e-05,
+      "loss": 1.2255,
+      "step": 7165
+    },
+    {
+      "epoch": 0.7922214242307055,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.5158763383233443e-05,
+      "loss": 1.2183,
+      "step": 7170
+    },
+    {
+      "epoch": 0.7927738798961383,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.5030995328933726e-05,
+      "loss": 1.2753,
+      "step": 7175
+    },
+    {
+      "epoch": 0.7933263355615712,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.490350610443396e-05,
+      "loss": 1.2647,
+      "step": 7180
+    },
+    {
+      "epoch": 0.793878791227004,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.477629618390066e-05,
+      "loss": 1.2645,
+      "step": 7185
+    },
+    {
+      "epoch": 0.7944312468924368,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.4649366040461597e-05,
+      "loss": 1.2241,
+      "step": 7190
+    },
+    {
+      "epoch": 0.7949837025578698,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.4522716146203974e-05,
+      "loss": 1.2617,
+      "step": 7195
+    },
+    {
+      "epoch": 0.7955361582233026,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.4396346972172634e-05,
+      "loss": 1.1895,
+      "step": 7200
+    },
+    {
+      "epoch": 0.7960886138887354,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.4270258988368376e-05,
+      "loss": 1.2278,
+      "step": 7205
+    },
+    {
+      "epoch": 0.7966410695541682,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.4144452663746176e-05,
+      "loss": 1.2745,
+      "step": 7210
+    },
+    {
+      "epoch": 0.7971935252196011,
+      "grad_norm": 0.384765625,
+      "learning_rate": 2.401892846621344e-05,
+      "loss": 1.2702,
+      "step": 7215
+    },
+    {
+      "epoch": 0.797745980885034,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.3893686862628263e-05,
+      "loss": 1.1883,
+      "step": 7220
+    },
+    {
+      "epoch": 0.7982984365504668,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.3768728318797684e-05,
+      "loss": 1.3128,
+      "step": 7225
+    },
+    {
+      "epoch": 0.7988508922158997,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.364405329947603e-05,
+      "loss": 1.2534,
+      "step": 7230
+    },
+    {
+      "epoch": 0.7994033478813325,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.3519662268363006e-05,
+      "loss": 1.2878,
+      "step": 7235
+    },
+    {
+      "epoch": 0.7999558035467653,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.339555568810221e-05,
+      "loss": 1.176,
+      "step": 7240
+    },
+    {
+      "epoch": 0.8005082592121983,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.3271734020279225e-05,
+      "loss": 1.3506,
+      "step": 7245
+    },
+    {
+      "epoch": 0.8010607148776311,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.3148197725419983e-05,
+      "loss": 1.355,
+      "step": 7250
+    },
+    {
+      "epoch": 0.8016131705430639,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.3024947262989038e-05,
+      "loss": 1.2662,
+      "step": 7255
+    },
+    {
+      "epoch": 0.8021656262084967,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.2901983091387867e-05,
+      "loss": 1.2308,
+      "step": 7260
+    },
+    {
+      "epoch": 0.8027180818739296,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.2779305667953154e-05,
+      "loss": 1.263,
+      "step": 7265
+    },
+    {
+      "epoch": 0.8032705375393625,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.2656915448955053e-05,
+      "loss": 1.3141,
+      "step": 7270
+    },
+    {
+      "epoch": 0.8038229932047953,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.253481288959558e-05,
+      "loss": 1.2824,
+      "step": 7275
+    },
+    {
+      "epoch": 0.8043754488702282,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.241299844400684e-05,
+      "loss": 1.2874,
+      "step": 7280
+    },
+    {
+      "epoch": 0.804927904535661,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.2291472565249384e-05,
+      "loss": 1.0976,
+      "step": 7285
+    },
+    {
+      "epoch": 0.8054803602010938,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.217023570531045e-05,
+      "loss": 1.2376,
+      "step": 7290
+    },
+    {
+      "epoch": 0.8060328158665268,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.2049288315102412e-05,
+      "loss": 1.2104,
+      "step": 7295
+    },
+    {
+      "epoch": 0.8065852715319596,
+      "grad_norm": 0.3828125,
+      "learning_rate": 2.1928630844460973e-05,
+      "loss": 1.2477,
+      "step": 7300
+    },
+    {
+      "epoch": 0.8071377271973924,
+      "grad_norm": 0.3828125,
+      "learning_rate": 2.1808263742143585e-05,
+      "loss": 1.2546,
+      "step": 7305
+    },
+    {
+      "epoch": 0.8076901828628252,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.1688187455827736e-05,
+      "loss": 1.2315,
+      "step": 7310
+    },
+    {
+      "epoch": 0.8082426385282581,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.1568402432109257e-05,
+      "loss": 1.3435,
+      "step": 7315
+    },
+    {
+      "epoch": 0.808795094193691,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.1448909116500747e-05,
+      "loss": 1.2878,
+      "step": 7320
+    },
+    {
+      "epoch": 0.8093475498591238,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.1329707953429822e-05,
+      "loss": 1.2644,
+      "step": 7325
+    },
+    {
+      "epoch": 0.8099000055245567,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.1210799386237535e-05,
+      "loss": 1.3027,
+      "step": 7330
+    },
+    {
+      "epoch": 0.8104524611899895,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.1092183857176683e-05,
+      "loss": 1.3232,
+      "step": 7335
+    },
+    {
+      "epoch": 0.8110049168554223,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.097386180741019e-05,
+      "loss": 1.3211,
+      "step": 7340
+    },
+    {
+      "epoch": 0.8115573725208552,
+      "grad_norm": 0.443359375,
+      "learning_rate": 2.0855833677009384e-05,
+      "loss": 1.2968,
+      "step": 7345
+    },
+    {
+      "epoch": 0.8121098281862881,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.0738099904952512e-05,
+      "loss": 1.2396,
+      "step": 7350
+    },
+    {
+      "epoch": 0.8126622838517209,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.0620660929123004e-05,
+      "loss": 1.334,
+      "step": 7355
+    },
+    {
+      "epoch": 0.8132147395171537,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.0503517186307842e-05,
+      "loss": 1.2061,
+      "step": 7360
+    },
+    {
+      "epoch": 0.8137671951825866,
+      "grad_norm": 0.3828125,
+      "learning_rate": 2.0386669112195976e-05,
+      "loss": 1.1548,
+      "step": 7365
+    },
+    {
+      "epoch": 0.8143196508480195,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.0270117141376664e-05,
+      "loss": 1.2279,
+      "step": 7370
+    },
+    {
+      "epoch": 0.8148721065134523,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.0153861707337906e-05,
+      "loss": 1.2821,
+      "step": 7375
+    },
+    {
+      "epoch": 0.8154245621788851,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.0037903242464785e-05,
+      "loss": 1.2913,
+      "step": 7380
+    },
+    {
+      "epoch": 0.815977017844318,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.9922242178037864e-05,
+      "loss": 1.1988,
+      "step": 7385
+    },
+    {
+      "epoch": 0.8165294735097508,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.9806878944231643e-05,
+      "loss": 1.3184,
+      "step": 7390
+    },
+    {
+      "epoch": 0.8170819291751837,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.9691813970112827e-05,
+      "loss": 1.2636,
+      "step": 7395
+    },
+    {
+      "epoch": 0.8176343848406166,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.9577047683638873e-05,
+      "loss": 1.265,
+      "step": 7400
+    },
+    {
+      "epoch": 0.8181868405060494,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.9462580511656338e-05,
+      "loss": 1.3265,
+      "step": 7405
+    },
+    {
+      "epoch": 0.8187392961714822,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.934841287989928e-05,
+      "loss": 1.2263,
+      "step": 7410
+    },
+    {
+      "epoch": 0.819291751836915,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.9234545212987688e-05,
+      "loss": 1.2812,
+      "step": 7415
+    },
+    {
+      "epoch": 0.819844207502348,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.91209779344259e-05,
+      "loss": 1.2624,
+      "step": 7420
+    },
+    {
+      "epoch": 0.8203966631677808,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.900771146660103e-05,
+      "loss": 1.2781,
+      "step": 7425
+    },
+    {
+      "epoch": 0.8209491188332136,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.88947462307814e-05,
+      "loss": 1.3008,
+      "step": 7430
+    },
+    {
+      "epoch": 0.8215015744986465,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.8782082647114962e-05,
+      "loss": 1.2242,
+      "step": 7435
+    },
+    {
+      "epoch": 0.8220540301640793,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.8669721134627748e-05,
+      "loss": 1.2445,
+      "step": 7440
+    },
+    {
+      "epoch": 0.8226064858295122,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.855766211122234e-05,
+      "loss": 1.194,
+      "step": 7445
+    },
+    {
+      "epoch": 0.8231589414949451,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.8445905993676183e-05,
+      "loss": 1.3307,
+      "step": 7450
+    },
+    {
+      "epoch": 0.8237113971603779,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.8334453197640224e-05,
+      "loss": 1.2829,
+      "step": 7455
+    },
+    {
+      "epoch": 0.8242638528258107,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.8223304137637243e-05,
+      "loss": 1.2545,
+      "step": 7460
+    },
+    {
+      "epoch": 0.8248163084912435,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.8112459227060386e-05,
+      "loss": 1.2735,
+      "step": 7465
+    },
+    {
+      "epoch": 0.8253687641566765,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.8001918878171532e-05,
+      "loss": 1.2054,
+      "step": 7470
+    },
+    {
+      "epoch": 0.8259212198221093,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.789168350209983e-05,
+      "loss": 1.2966,
+      "step": 7475
+    },
+    {
+      "epoch": 0.8264736754875421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.778175350884016e-05,
+      "loss": 1.321,
+      "step": 7480
+    },
+    {
+      "epoch": 0.827026131152975,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.767212930725163e-05,
+      "loss": 1.249,
+      "step": 7485
+    },
+    {
+      "epoch": 0.8275785868184078,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.756281130505595e-05,
+      "loss": 1.2322,
+      "step": 7490
+    },
+    {
+      "epoch": 0.8281310424838406,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.745379990883603e-05,
+      "loss": 1.2847,
+      "step": 7495
+    },
+    {
+      "epoch": 0.8286834981492736,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.7345095524034484e-05,
+      "loss": 1.2884,
+      "step": 7500
+    },
+    {
+      "epoch": 0.8292359538147064,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.723669855495199e-05,
+      "loss": 1.2099,
+      "step": 7505
+    },
+    {
+      "epoch": 0.8297884094801392,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.712860940474591e-05,
+      "loss": 1.2257,
+      "step": 7510
+    },
+    {
+      "epoch": 0.830340865145572,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.702082847542873e-05,
+      "loss": 1.2523,
+      "step": 7515
+    },
+    {
+      "epoch": 0.8308933208110049,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.6913356167866578e-05,
+      "loss": 1.2422,
+      "step": 7520
+    },
+    {
+      "epoch": 0.8314457764764378,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.680619288177775e-05,
+      "loss": 1.2598,
+      "step": 7525
+    },
+    {
+      "epoch": 0.8319982321418706,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.6699339015731185e-05,
+      "loss": 1.2946,
+      "step": 7530
+    },
+    {
+      "epoch": 0.8325506878073035,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.659279496714503e-05,
+      "loss": 1.2523,
+      "step": 7535
+    },
+    {
+      "epoch": 0.8331031434727363,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.648656113228515e-05,
+      "loss": 1.3067,
+      "step": 7540
+    },
+    {
+      "epoch": 0.8336555991381691,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.6380637906263574e-05,
+      "loss": 1.2586,
+      "step": 7545
+    },
+    {
+      "epoch": 0.834208054803602,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.6275025683037148e-05,
+      "loss": 1.2271,
+      "step": 7550
+    },
+    {
+      "epoch": 0.8347605104690349,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.616972485540601e-05,
+      "loss": 1.3128,
+      "step": 7555
+    },
+    {
+      "epoch": 0.8353129661344677,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.6064735815012145e-05,
+      "loss": 1.3026,
+      "step": 7560
+    },
+    {
+      "epoch": 0.8358654217999005,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.5960058952337887e-05,
+      "loss": 1.2414,
+      "step": 7565
+    },
+    {
+      "epoch": 0.8364178774653334,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.585569465670451e-05,
+      "loss": 1.324,
+      "step": 7570
+    },
+    {
+      "epoch": 0.8369703331307663,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.575164331627079e-05,
+      "loss": 1.2407,
+      "step": 7575
+    },
+    {
+      "epoch": 0.8375227887961991,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5647905318031507e-05,
+      "loss": 1.2253,
+      "step": 7580
+    },
+    {
+      "epoch": 0.838075244461632,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.554448104781606e-05,
+      "loss": 1.3098,
+      "step": 7585
+    },
+    {
+      "epoch": 0.8386277001270648,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.5441370890287022e-05,
+      "loss": 1.223,
+      "step": 7590
+    },
+    {
+      "epoch": 0.8391801557924976,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.5338575228938614e-05,
+      "loss": 1.2348,
+      "step": 7595
+    },
+    {
+      "epoch": 0.8397326114579305,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.523609444609545e-05,
+      "loss": 1.3541,
+      "step": 7600
+    },
+    {
+      "epoch": 0.8402850671233634,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.5133928922911012e-05,
+      "loss": 1.1748,
+      "step": 7605
+    },
+    {
+      "epoch": 0.8408375227887962,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.5032079039366209e-05,
+      "loss": 1.3035,
+      "step": 7610
+    },
+    {
+      "epoch": 0.841389978454229,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.4930545174268062e-05,
+      "loss": 1.3263,
+      "step": 7615
+    },
+    {
+      "epoch": 0.8419424341196619,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.4829327705248164e-05,
+      "loss": 1.2054,
+      "step": 7620
+    },
+    {
+      "epoch": 0.8424948897850948,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.4728427008761402e-05,
+      "loss": 1.3129,
+      "step": 7625
+    },
+    {
+      "epoch": 0.8430473454505276,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.4627843460084478e-05,
+      "loss": 1.345,
+      "step": 7630
+    },
+    {
+      "epoch": 0.8435998011159604,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.4527577433314532e-05,
+      "loss": 1.2276,
+      "step": 7635
+    },
+    {
+      "epoch": 0.8441522567813933,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.4427629301367773e-05,
+      "loss": 1.296,
+      "step": 7640
+    },
+    {
+      "epoch": 0.8447047124468261,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.4327999435978068e-05,
+      "loss": 1.2482,
+      "step": 7645
+    },
+    {
+      "epoch": 0.845257168112259,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.422868820769554e-05,
+      "loss": 1.2892,
+      "step": 7650
+    },
+    {
+      "epoch": 0.8458096237776919,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.4129695985885228e-05,
+      "loss": 1.2172,
+      "step": 7655
+    },
+    {
+      "epoch": 0.8463620794431247,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.403102313872573e-05,
+      "loss": 1.2775,
+      "step": 7660
+    },
+    {
+      "epoch": 0.8469145351085575,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.3932670033207784e-05,
+      "loss": 1.2685,
+      "step": 7665
+    },
+    {
+      "epoch": 0.8474669907739903,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.3834637035132903e-05,
+      "loss": 1.3008,
+      "step": 7670
+    },
+    {
+      "epoch": 0.8480194464394233,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.373692450911207e-05,
+      "loss": 1.2244,
+      "step": 7675
+    },
+    {
+      "epoch": 0.8485719021048561,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.3639532818564327e-05,
+      "loss": 1.1604,
+      "step": 7680
+    },
+    {
+      "epoch": 0.8491243577702889,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.3542462325715443e-05,
+      "loss": 1.2167,
+      "step": 7685
+    },
+    {
+      "epoch": 0.8496768134357218,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.344571339159657e-05,
+      "loss": 1.2944,
+      "step": 7690
+    },
+    {
+      "epoch": 0.8502292691011546,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.3349286376042914e-05,
+      "loss": 1.2022,
+      "step": 7695
+    },
+    {
+      "epoch": 0.8507817247665875,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.3253181637692324e-05,
+      "loss": 1.2281,
+      "step": 7700
+    },
+    {
+      "epoch": 0.8513341804320204,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.3157399533984082e-05,
+      "loss": 1.2718,
+      "step": 7705
+    },
+    {
+      "epoch": 0.8518866360974532,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.3061940421157459e-05,
+      "loss": 1.1907,
+      "step": 7710
+    },
+    {
+      "epoch": 0.852439091762886,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.2966804654250465e-05,
+      "loss": 1.1876,
+      "step": 7715
+    },
+    {
+      "epoch": 0.8529915474283188,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.287199258709848e-05,
+      "loss": 1.2103,
+      "step": 7720
+    },
+    {
+      "epoch": 0.8535440030937518,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.2777504572332976e-05,
+      "loss": 1.2198,
+      "step": 7725
+    },
+    {
+      "epoch": 0.8540964587591846,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2683340961380163e-05,
+      "loss": 1.2929,
+      "step": 7730
+    },
+    {
+      "epoch": 0.8546489144246174,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.2589502104459738e-05,
+      "loss": 1.2406,
+      "step": 7735
+    },
+    {
+      "epoch": 0.8552013700900503,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.249598835058352e-05,
+      "loss": 1.2474,
+      "step": 7740
+    },
+    {
+      "epoch": 0.8557538257554831,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.2402800047554208e-05,
+      "loss": 1.3239,
+      "step": 7745
+    },
+    {
+      "epoch": 0.856306281420916,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.2309937541964057e-05,
+      "loss": 1.2417,
+      "step": 7750
+    },
+    {
+      "epoch": 0.8568587370863489,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.2217401179193556e-05,
+      "loss": 1.2904,
+      "step": 7755
+    },
+    {
+      "epoch": 0.8574111927517817,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.2125191303410221e-05,
+      "loss": 1.2472,
+      "step": 7760
+    },
+    {
+      "epoch": 0.8579636484172145,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.2033308257567289e-05,
+      "loss": 1.2792,
+      "step": 7765
+    },
+    {
+      "epoch": 0.8585161040826473,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.1941752383402394e-05,
+      "loss": 1.2666,
+      "step": 7770
+    },
+    {
+      "epoch": 0.8590685597480802,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.1850524021436337e-05,
+      "loss": 1.2452,
+      "step": 7775
+    },
+    {
+      "epoch": 0.8596210154135131,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.175962351097184e-05,
+      "loss": 1.2691,
+      "step": 7780
+    },
+    {
+      "epoch": 0.8601734710789459,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.166905119009223e-05,
+      "loss": 1.2344,
+      "step": 7785
+    },
+    {
+      "epoch": 0.8607259267443788,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.1578807395660207e-05,
+      "loss": 1.3321,
+      "step": 7790
+    },
+    {
+      "epoch": 0.8612783824098116,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.1488892463316615e-05,
+      "loss": 1.2145,
+      "step": 7795
+    },
+    {
+      "epoch": 0.8618308380752444,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.1399306727479164e-05,
+      "loss": 1.205,
+      "step": 7800
+    },
+    {
+      "epoch": 0.8623832937406773,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.1310050521341198e-05,
+      "loss": 1.3174,
+      "step": 7805
+    },
+    {
+      "epoch": 0.8629357494061102,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.1221124176870412e-05,
+      "loss": 1.2388,
+      "step": 7810
+    },
+    {
+      "epoch": 0.863488205071543,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.1132528024807686e-05,
+      "loss": 1.2616,
+      "step": 7815
+    },
+    {
+      "epoch": 0.8640406607369758,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.1044262394665872e-05,
+      "loss": 1.3381,
+      "step": 7820
+    },
+    {
+      "epoch": 0.8645931164024087,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.0956327614728457e-05,
+      "loss": 1.2389,
+      "step": 7825
+    },
+    {
+      "epoch": 0.8651455720678416,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.0868724012048438e-05,
+      "loss": 1.2653,
+      "step": 7830
+    },
+    {
+      "epoch": 0.8656980277332744,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.078145191244706e-05,
+      "loss": 1.2068,
+      "step": 7835
+    },
+    {
+      "epoch": 0.8662504833987072,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.069451164051264e-05,
+      "loss": 1.2396,
+      "step": 7840
+    },
+    {
+      "epoch": 0.8668029390641401,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.0607903519599328e-05,
+      "loss": 1.2372,
+      "step": 7845
+    },
+    {
+      "epoch": 0.8673553947295729,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.052162787182588e-05,
+      "loss": 1.2051,
+      "step": 7850
+    },
+    {
+      "epoch": 0.8679078503950058,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.0435685018074548e-05,
+      "loss": 1.3016,
+      "step": 7855
+    },
+    {
+      "epoch": 0.8684603060604387,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.0350075277989812e-05,
+      "loss": 1.2742,
+      "step": 7860
+    },
+    {
+      "epoch": 0.8690127617258715,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.026479896997723e-05,
+      "loss": 1.2405,
+      "step": 7865
+    },
+    {
+      "epoch": 0.8695652173913043,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.0179856411202204e-05,
+      "loss": 1.3052,
+      "step": 7870
+    },
+    {
+      "epoch": 0.8701176730567372,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.0095247917588869e-05,
+      "loss": 1.3322,
+      "step": 7875
+    },
+    {
+      "epoch": 0.8706701287221701,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.0010973803818857e-05,
+      "loss": 1.215,
+      "step": 7880
+    },
+    {
+      "epoch": 0.8712225843876029,
+      "grad_norm": 0.38671875,
+      "learning_rate": 9.927034383330159e-06,
+      "loss": 1.1596,
+      "step": 7885
+    },
+    {
+      "epoch": 0.8717750400530357,
+      "grad_norm": 0.392578125,
+      "learning_rate": 9.843429968315943e-06,
+      "loss": 1.2819,
+      "step": 7890
+    },
+    {
+      "epoch": 0.8723274957184686,
+      "grad_norm": 0.435546875,
+      "learning_rate": 9.760160869723456e-06,
+      "loss": 1.3613,
+      "step": 7895
+    },
+    {
+      "epoch": 0.8728799513839014,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.677227397252708e-06,
+      "loss": 1.293,
+      "step": 7900
+    },
+    {
+      "epoch": 0.8734324070493343,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.594629859355519e-06,
+      "loss": 1.2734,
+      "step": 7905
+    },
+    {
+      "epoch": 0.8739848627147672,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.512368563234241e-06,
+      "loss": 1.2817,
+      "step": 7910
+    },
+    {
+      "epoch": 0.8745373183802,
+      "grad_norm": 0.3984375,
+      "learning_rate": 9.430443814840662e-06,
+      "loss": 1.2749,
+      "step": 7915
+    },
+    {
+      "epoch": 0.8750897740456328,
+      "grad_norm": 0.384765625,
+      "learning_rate": 9.348855918874844e-06,
+      "loss": 1.2575,
+      "step": 7920
+    },
+    {
+      "epoch": 0.8756422297110656,
+      "grad_norm": 0.39453125,
+      "learning_rate": 9.267605178784033e-06,
+      "loss": 1.2978,
+      "step": 7925
+    },
+    {
+      "epoch": 0.8761946853764986,
+      "grad_norm": 0.41015625,
+      "learning_rate": 9.186691896761479e-06,
+      "loss": 1.1692,
+      "step": 7930
+    },
+    {
+      "epoch": 0.8767471410419314,
+      "grad_norm": 0.3828125,
+      "learning_rate": 9.106116373745332e-06,
+      "loss": 1.3137,
+      "step": 7935
+    },
+    {
+      "epoch": 0.8772995967073642,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.025878909417552e-06,
+      "loss": 1.2491,
+      "step": 7940
+    },
+    {
+      "epoch": 0.8778520523727971,
+      "grad_norm": 0.40625,
+      "learning_rate": 8.94597980220273e-06,
+      "loss": 1.2099,
+      "step": 7945
+    },
+    {
+      "epoch": 0.8784045080382299,
+      "grad_norm": 0.396484375,
+      "learning_rate": 8.866419349267064e-06,
+      "loss": 1.2653,
+      "step": 7950
+    },
+    {
+      "epoch": 0.8789569637036628,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.787197846517148e-06,
+      "loss": 1.255,
+      "step": 7955
+    },
+    {
+      "epoch": 0.8795094193690957,
+      "grad_norm": 0.388671875,
+      "learning_rate": 8.70831558859897e-06,
+      "loss": 1.2871,
+      "step": 7960
+    },
+    {
+      "epoch": 0.8800618750345285,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.629772868896779e-06,
+      "loss": 1.2383,
+      "step": 7965
+    },
+    {
+      "epoch": 0.8806143306999613,
+      "grad_norm": 0.41015625,
+      "learning_rate": 8.55156997953197e-06,
+      "loss": 1.2273,
+      "step": 7970
+    },
+    {
+      "epoch": 0.8811667863653941,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.473707211362026e-06,
+      "loss": 1.2977,
+      "step": 7975
+    },
+    {
+      "epoch": 0.8817192420308271,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.396184853979416e-06,
+      "loss": 1.2394,
+      "step": 7980
+    },
+    {
+      "epoch": 0.8822716976962599,
+      "grad_norm": 0.396484375,
+      "learning_rate": 8.319003195710574e-06,
+      "loss": 1.3483,
+      "step": 7985
+    },
+    {
+      "epoch": 0.8828241533616927,
+      "grad_norm": 0.40234375,
+      "learning_rate": 8.242162523614716e-06,
+      "loss": 1.2118,
+      "step": 7990
+    },
+    {
+      "epoch": 0.8833766090271256,
+      "grad_norm": 0.38671875,
+      "learning_rate": 8.165663123482903e-06,
+      "loss": 1.2282,
+      "step": 7995
+    },
+    {
+      "epoch": 0.8839290646925584,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.089505279836873e-06,
+      "loss": 1.2785,
+      "step": 8000
+    },
+    {
+      "epoch": 0.8844815203579913,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.013689275928037e-06,
+      "loss": 1.24,
+      "step": 8005
+    },
+    {
+      "epoch": 0.8850339760234242,
+      "grad_norm": 0.423828125,
+      "learning_rate": 7.938215393736414e-06,
+      "loss": 1.2231,
+      "step": 8010
+    },
+    {
+      "epoch": 0.885586431688857,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.86308391396956e-06,
+      "loss": 1.2741,
+      "step": 8015
+    },
+    {
+      "epoch": 0.8861388873542898,
+      "grad_norm": 0.3828125,
+      "learning_rate": 7.788295116061584e-06,
+      "loss": 1.1967,
+      "step": 8020
+    },
+    {
+      "epoch": 0.8866913430197226,
+      "grad_norm": 0.416015625,
+      "learning_rate": 7.713849278172047e-06,
+      "loss": 1.2404,
+      "step": 8025
+    },
+    {
+      "epoch": 0.8872437986851556,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.639746677184945e-06,
+      "loss": 1.317,
+      "step": 8030
+    },
+    {
+      "epoch": 0.8877962543505884,
+      "grad_norm": 0.40625,
+      "learning_rate": 7.5659875887076905e-06,
+      "loss": 1.2615,
+      "step": 8035
+    },
+    {
+      "epoch": 0.8883487100160212,
+      "grad_norm": 0.435546875,
+      "learning_rate": 7.492572287070088e-06,
+      "loss": 1.2847,
+      "step": 8040
+    },
+    {
+      "epoch": 0.888901165681454,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.419501045323296e-06,
+      "loss": 1.3259,
+      "step": 8045
+    },
+    {
+      "epoch": 0.8894536213468869,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.346774135238832e-06,
+      "loss": 1.2395,
+      "step": 8050
+    },
+    {
+      "epoch": 0.8900060770123197,
+      "grad_norm": 0.388671875,
+      "learning_rate": 7.274391827307547e-06,
+      "loss": 1.284,
+      "step": 8055
+    },
+    {
+      "epoch": 0.8905585326777526,
+      "grad_norm": 0.359375,
+      "learning_rate": 7.202354390738608e-06,
+      "loss": 1.2127,
+      "step": 8060
+    },
+    {
+      "epoch": 0.8911109883431855,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.130662093458529e-06,
+      "loss": 1.3602,
+      "step": 8065
+    },
+    {
+      "epoch": 0.8916634440086183,
+      "grad_norm": 0.3828125,
+      "learning_rate": 7.059315202110173e-06,
+      "loss": 1.3355,
+      "step": 8070
+    },
+    {
+      "epoch": 0.8922158996740511,
+      "grad_norm": 0.412109375,
+      "learning_rate": 6.98831398205172e-06,
+      "loss": 1.3255,
+      "step": 8075
+    },
+    {
+      "epoch": 0.892768355339484,
+      "grad_norm": 0.423828125,
+      "learning_rate": 6.917658697355722e-06,
+      "loss": 1.2978,
+      "step": 8080
+    },
+    {
+      "epoch": 0.8933208110049169,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.8473496108080845e-06,
+      "loss": 1.1923,
+      "step": 8085
+    },
+    {
+      "epoch": 0.8938732666703497,
+      "grad_norm": 0.39453125,
+      "learning_rate": 6.777386983907152e-06,
+      "loss": 1.2953,
+      "step": 8090
+    },
+    {
+      "epoch": 0.8944257223357825,
+      "grad_norm": 0.41796875,
+      "learning_rate": 6.7077710768626455e-06,
+      "loss": 1.3616,
+      "step": 8095
+    },
+    {
+      "epoch": 0.8949781780012154,
+      "grad_norm": 0.404296875,
+      "learning_rate": 6.638502148594772e-06,
+      "loss": 1.2919,
+      "step": 8100
+    },
+    {
+      "epoch": 0.8955306336666482,
+      "grad_norm": 0.380859375,
+      "learning_rate": 6.5695804567332044e-06,
+      "loss": 1.2795,
+      "step": 8105
+    },
+    {
+      "epoch": 0.8960830893320811,
+      "grad_norm": 0.41796875,
+      "learning_rate": 6.501006257616205e-06,
+      "loss": 1.2792,
+      "step": 8110
+    },
+    {
+      "epoch": 0.896635544997514,
+      "grad_norm": 0.439453125,
+      "learning_rate": 6.432779806289535e-06,
+      "loss": 1.3561,
+      "step": 8115
+    },
+    {
+      "epoch": 0.8971880006629468,
+      "grad_norm": 0.412109375,
+      "learning_rate": 6.364901356505648e-06,
+      "loss": 1.2155,
+      "step": 8120
+    },
+    {
+      "epoch": 0.8977404563283796,
+      "grad_norm": 0.416015625,
+      "learning_rate": 6.297371160722676e-06,
+      "loss": 1.2476,
+      "step": 8125
+    },
+    {
+      "epoch": 0.8982929119938124,
+      "grad_norm": 0.404296875,
+      "learning_rate": 6.230189470103498e-06,
+      "loss": 1.2504,
+      "step": 8130
+    },
+    {
+      "epoch": 0.8988453676592454,
+      "grad_norm": 0.40234375,
+      "learning_rate": 6.163356534514808e-06,
+      "loss": 1.2738,
+      "step": 8135
+    },
+    {
+      "epoch": 0.8993978233246782,
+      "grad_norm": 0.412109375,
+      "learning_rate": 6.096872602526182e-06,
+      "loss": 1.3324,
+      "step": 8140
+    },
+    {
+      "epoch": 0.899950278990111,
+      "grad_norm": 0.404296875,
+      "learning_rate": 6.030737921409169e-06,
+      "loss": 1.2297,
+      "step": 8145
+    },
+    {
+      "epoch": 0.9005027346555439,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.964952737136353e-06,
+      "loss": 1.2203,
+      "step": 8150
+    },
+    {
+      "epoch": 0.9010551903209767,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.899517294380441e-06,
+      "loss": 1.2453,
+      "step": 8155
+    },
+    {
+      "epoch": 0.9016076459864096,
+      "grad_norm": 0.412109375,
+      "learning_rate": 5.834431836513388e-06,
+      "loss": 1.3345,
+      "step": 8160
+    },
+    {
+      "epoch": 0.9021601016518425,
+      "grad_norm": 0.431640625,
+      "learning_rate": 5.769696605605379e-06,
+      "loss": 1.2634,
+      "step": 8165
+    },
+    {
+      "epoch": 0.9027125573172753,
+      "grad_norm": 0.396484375,
+      "learning_rate": 5.705311842424133e-06,
+      "loss": 1.2929,
+      "step": 8170
+    },
+    {
+      "epoch": 0.9032650129827081,
+      "grad_norm": 0.408203125,
+      "learning_rate": 5.641277786433796e-06,
+      "loss": 1.2304,
+      "step": 8175
+    },
+    {
+      "epoch": 0.9038174686481409,
+      "grad_norm": 0.4140625,
+      "learning_rate": 5.577594675794162e-06,
+      "loss": 1.3047,
+      "step": 8180
+    },
+    {
+      "epoch": 0.9043699243135739,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.514262747359778e-06,
+      "loss": 1.2051,
+      "step": 8185
+    },
+    {
+      "epoch": 0.9049223799790067,
+      "grad_norm": 0.412109375,
+      "learning_rate": 5.451282236679045e-06,
+      "loss": 1.2079,
+      "step": 8190
+    },
+    {
+      "epoch": 0.9054748356444395,
+      "grad_norm": 0.45703125,
+      "learning_rate": 5.388653377993324e-06,
+      "loss": 1.2679,
+      "step": 8195
+    },
+    {
+      "epoch": 0.9060272913098724,
+      "grad_norm": 0.412109375,
+      "learning_rate": 5.326376404236133e-06,
+      "loss": 1.3061,
+      "step": 8200
+    },
+    {
+      "epoch": 0.9065797469753052,
+      "grad_norm": 0.400390625,
+      "learning_rate": 5.264451547032212e-06,
+      "loss": 1.1406,
+      "step": 8205
+    },
+    {
+      "epoch": 0.9071322026407381,
+      "grad_norm": 0.41015625,
+      "learning_rate": 5.202879036696662e-06,
+      "loss": 1.2934,
+      "step": 8210
+    },
+    {
+      "epoch": 0.907684658306171,
+      "grad_norm": 0.427734375,
+      "learning_rate": 5.141659102234131e-06,
+      "loss": 1.2391,
+      "step": 8215
+    },
+    {
+      "epoch": 0.9082371139716038,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.080791971337972e-06,
+      "loss": 1.2339,
+      "step": 8220
+    },
+    {
+      "epoch": 0.9087895696370366,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.020277870389312e-06,
+      "loss": 1.2473,
+      "step": 8225
+    },
+    {
+      "epoch": 0.9093420253024694,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.960117024456323e-06,
+      "loss": 1.2827,
+      "step": 8230
+    },
+    {
+      "epoch": 0.9098944809679024,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.9003096572932785e-06,
+      "loss": 1.2659,
+      "step": 8235
+    },
+    {
+      "epoch": 0.9104469366333352,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.840855991339799e-06,
+      "loss": 1.155,
+      "step": 8240
+    },
+    {
+      "epoch": 0.910999392298768,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.781756247719982e-06,
+      "loss": 1.3278,
+      "step": 8245
+    },
+    {
+      "epoch": 0.9115518479642009,
+      "grad_norm": 0.375,
+      "learning_rate": 4.7230106462415876e-06,
+      "loss": 1.3132,
+      "step": 8250
+    },
+    {
+      "epoch": 0.9121043036296337,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.6646194053952656e-06,
+      "loss": 1.3053,
+      "step": 8255
+    },
+    {
+      "epoch": 0.9126567592950666,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.6065827423536375e-06,
+      "loss": 1.3756,
+      "step": 8260
+    },
+    {
+      "epoch": 0.9132092149604994,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.548900872970607e-06,
+      "loss": 1.299,
+      "step": 8265
+    },
+    {
+      "epoch": 0.9137616706259323,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.491574011780497e-06,
+      "loss": 1.3044,
+      "step": 8270
+    },
+    {
+      "epoch": 0.9143141262913651,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.434602371997243e-06,
+      "loss": 1.2712,
+      "step": 8275
+    },
+    {
+      "epoch": 0.9148665819567979,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.3779861655136255e-06,
+      "loss": 1.2805,
+      "step": 8280
+    },
+    {
+      "epoch": 0.9154190376222309,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.321725602900473e-06,
+      "loss": 1.3191,
+      "step": 8285
+    },
+    {
+      "epoch": 0.9159714932876637,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.265820893405892e-06,
+      "loss": 1.2231,
+      "step": 8290
+    },
+    {
+      "epoch": 0.9165239489530965,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.210272244954449e-06,
+      "loss": 1.3134,
+      "step": 8295
+    },
+    {
+      "epoch": 0.9170764046185294,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4.1550798641464605e-06,
+      "loss": 1.2432,
+      "step": 8300
+    },
+    {
+      "epoch": 0.9176288602839622,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.100243956257144e-06,
+      "loss": 1.2339,
+      "step": 8305
+    },
+    {
+      "epoch": 0.9181813159493951,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.045764725235956e-06,
+      "loss": 1.2767,
+      "step": 8310
+    },
+    {
+      "epoch": 0.9187337716148279,
+      "grad_norm": 0.38671875,
+      "learning_rate": 3.991642373705695e-06,
+      "loss": 1.2005,
+      "step": 8315
+    },
+    {
+      "epoch": 0.9192862272802608,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.937877102961918e-06,
+      "loss": 1.2877,
+      "step": 8320
+    },
+    {
+      "epoch": 0.9198386829456936,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.884469112972033e-06,
+      "loss": 1.3861,
+      "step": 8325
+    },
+    {
+      "epoch": 0.9203911386111264,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.83141860237467e-06,
+      "loss": 1.3506,
+      "step": 8330
+    },
+    {
+      "epoch": 0.9209435942765593,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.7787257684788745e-06,
+      "loss": 1.1882,
+      "step": 8335
+    },
+    {
+      "epoch": 0.9214960499419922,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.7263908072634025e-06,
+      "loss": 1.2829,
+      "step": 8340
+    },
+    {
+      "epoch": 0.922048505607425,
+      "grad_norm": 0.384765625,
+      "learning_rate": 3.6744139133759957e-06,
+      "loss": 1.2689,
+      "step": 8345
+    },
+    {
+      "epoch": 0.9226009612728578,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.6227952801326404e-06,
+      "loss": 1.3579,
+      "step": 8350
+    },
+    {
+      "epoch": 0.9231534169382907,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.571535099516832e-06,
+      "loss": 1.2262,
+      "step": 8355
+    },
+    {
+      "epoch": 0.9237058726037235,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.520633562178932e-06,
+      "loss": 1.2218,
+      "step": 8360
+    },
+    {
+      "epoch": 0.9242583282691564,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.470090857435371e-06,
+      "loss": 1.2624,
+      "step": 8365
+    },
+    {
+      "epoch": 0.9248107839345893,
+      "grad_norm": 0.45703125,
+      "learning_rate": 3.419907173268e-06,
+      "loss": 1.2848,
+      "step": 8370
+    },
+    {
+      "epoch": 0.9253632396000221,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.3700826963233735e-06,
+      "loss": 1.2628,
+      "step": 8375
+    },
+    {
+      "epoch": 0.9259156952654549,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.320617611912069e-06,
+      "loss": 1.3411,
+      "step": 8380
+    },
+    {
+      "epoch": 0.9264681509308877,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.271512104007979e-06,
+      "loss": 1.2467,
+      "step": 8385
+    },
+    {
+      "epoch": 0.9270206065963207,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.2227663552476194e-06,
+      "loss": 1.2752,
+      "step": 8390
+    },
+    {
+      "epoch": 0.9275730622617535,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.174380546929501e-06,
+      "loss": 1.2332,
+      "step": 8395
+    },
+    {
+      "epoch": 0.9281255179271863,
+      "grad_norm": 0.38671875,
+      "learning_rate": 3.1263548590133917e-06,
+      "loss": 1.2656,
+      "step": 8400
+    },
+    {
+      "epoch": 0.9286779735926192,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.0786894701196777e-06,
+      "loss": 1.2307,
+      "step": 8405
+    },
+    {
+      "epoch": 0.929230429258052,
+      "grad_norm": 0.451171875,
+      "learning_rate": 3.031384557528716e-06,
+      "loss": 1.2646,
+      "step": 8410
+    },
+    {
+      "epoch": 0.9297828849234849,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.9844402971801242e-06,
+      "loss": 1.2306,
+      "step": 8415
+    },
+    {
+      "epoch": 0.9303353405889178,
+      "grad_norm": 0.369140625,
+      "learning_rate": 2.9378568636721835e-06,
+      "loss": 1.1948,
+      "step": 8420
+    },
+    {
+      "epoch": 0.9308877962543506,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.8916344302611586e-06,
+      "loss": 1.2318,
+      "step": 8425
+    },
+    {
+      "epoch": 0.9314402519197834,
+      "grad_norm": 0.3828125,
+      "learning_rate": 2.845773168860644e-06,
+      "loss": 1.2282,
+      "step": 8430
+    },
+    {
+      "epoch": 0.9319927075852162,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.800273250040952e-06,
+      "loss": 1.2427,
+      "step": 8435
+    },
+    {
+      "epoch": 0.9325451632506492,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.755134843028462e-06,
+      "loss": 1.3636,
+      "step": 8440
+    },
+    {
+      "epoch": 0.933097618916082,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.710358115705003e-06,
+      "loss": 1.2182,
+      "step": 8445
+    },
+    {
+      "epoch": 0.9336500745815148,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.6659432346072156e-06,
+      "loss": 1.3193,
+      "step": 8450
+    },
+    {
+      "epoch": 0.9342025302469477,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.6218903649259163e-06,
+      "loss": 1.2126,
+      "step": 8455
+    },
+    {
+      "epoch": 0.9347549859123805,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.578199670505532e-06,
+      "loss": 1.3129,
+      "step": 8460
+    },
+    {
+      "epoch": 0.9353074415778134,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.5348713138434564e-06,
+      "loss": 1.2968,
+      "step": 8465
+    },
+    {
+      "epoch": 0.9358598972432463,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.4919054560894383e-06,
+      "loss": 1.2434,
+      "step": 8470
+    },
+    {
+      "epoch": 0.9364123529086791,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.4493022570450164e-06,
+      "loss": 1.3078,
+      "step": 8475
+    },
+    {
+      "epoch": 0.9369648085741119,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.4070618751628748e-06,
+      "loss": 1.3117,
+      "step": 8480
+    },
+    {
+      "epoch": 0.9375172642395447,
+      "grad_norm": 0.3828125,
+      "learning_rate": 2.365184467546333e-06,
+      "loss": 1.1773,
+      "step": 8485
+    },
+    {
+      "epoch": 0.9380697199049777,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.3236701899486566e-06,
+      "loss": 1.3479,
+      "step": 8490
+    },
+    {
+      "epoch": 0.9386221755704105,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.28251919677257e-06,
+      "loss": 1.1983,
+      "step": 8495
+    },
+    {
+      "epoch": 0.9391746312358433,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.2417316410696333e-06,
+      "loss": 1.3284,
+      "step": 8500
+    },
+    {
+      "epoch": 0.9397270869012762,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.2013076745396765e-06,
+      "loss": 1.2226,
+      "step": 8505
+    },
+    {
+      "epoch": 0.940279542566709,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.161247447530268e-06,
+      "loss": 1.2644,
+      "step": 8510
+    },
+    {
+      "epoch": 0.9408319982321419,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.121551109036124e-06,
+      "loss": 1.2459,
+      "step": 8515
+    },
+    {
+      "epoch": 0.9413844538975747,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.0822188066985214e-06,
+      "loss": 1.2049,
+      "step": 8520
+    },
+    {
+      "epoch": 0.9419369095630076,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.043250686804865e-06,
+      "loss": 1.3138,
+      "step": 8525
+    },
+    {
+      "epoch": 0.9424893652284404,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.004646894287987e-06,
+      "loss": 1.3041,
+      "step": 8530
+    },
+    {
+      "epoch": 0.9430418208938732,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.9664075727257593e-06,
+      "loss": 1.2856,
+      "step": 8535
+    },
+    {
+      "epoch": 0.9435942765593062,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.928532864340438e-06,
+      "loss": 1.2399,
+      "step": 8540
+    },
+    {
+      "epoch": 0.944146732224739,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.891022909998208e-06,
+      "loss": 1.2144,
+      "step": 8545
+    },
+    {
+      "epoch": 0.9446991878901718,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.8538778492086407e-06,
+      "loss": 1.239,
+      "step": 8550
+    },
+    {
+      "epoch": 0.9452516435556046,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.8170978201241474e-06,
+      "loss": 1.214,
+      "step": 8555
+    },
+    {
+      "epoch": 0.9458040992210375,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.7806829595395147e-06,
+      "loss": 1.2959,
+      "step": 8560
+    },
+    {
+      "epoch": 0.9463565548864704,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.7446334028913491e-06,
+      "loss": 1.2792,
+      "step": 8565
+    },
+    {
+      "epoch": 0.9469090105519032,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.7089492842576106e-06,
+      "loss": 1.2391,
+      "step": 8570
+    },
+    {
+      "epoch": 0.9474614662173361,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.6736307363570903e-06,
+      "loss": 1.1879,
+      "step": 8575
+    },
+    {
+      "epoch": 0.9480139218827689,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.638677890548912e-06,
+      "loss": 1.2647,
+      "step": 8580
+    },
+    {
+      "epoch": 0.9485663775482017,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.6040908768320872e-06,
+      "loss": 1.2889,
+      "step": 8585
+    },
+    {
+      "epoch": 0.9491188332136347,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.5698698238449716e-06,
+      "loss": 1.2407,
+      "step": 8590
+    },
+    {
+      "epoch": 0.9496712888790675,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.5360148588648093e-06,
+      "loss": 1.1342,
+      "step": 8595
+    },
+    {
+      "epoch": 0.9502237445445003,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.5025261078073005e-06,
+      "loss": 1.3172,
+      "step": 8600
+    },
+    {
+      "epoch": 0.9507762002099331,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.469403695226057e-06,
+      "loss": 1.2656,
+      "step": 8605
+    },
+    {
+      "epoch": 0.951328655875366,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.436647744312214e-06,
+      "loss": 1.2591,
+      "step": 8610
+    },
+    {
+      "epoch": 0.9518811115407988,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.4042583768939298e-06,
+      "loss": 1.211,
+      "step": 8615
+    },
+    {
+      "epoch": 0.9524335672062317,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.3722357134359099e-06,
+      "loss": 1.2279,
+      "step": 8620
+    },
+    {
+      "epoch": 0.9529860228716646,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.3405798730390273e-06,
+      "loss": 1.3243,
+      "step": 8625
+    },
+    {
+      "epoch": 0.9535384785370974,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3092909734398251e-06,
+      "loss": 1.3247,
+      "step": 8630
+    },
+    {
+      "epoch": 0.9540909342025302,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.278369131010093e-06,
+      "loss": 1.2227,
+      "step": 8635
+    },
+    {
+      "epoch": 0.954643389867963,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.2478144607564469e-06,
+      "loss": 1.2665,
+      "step": 8640
+    },
+    {
+      "epoch": 0.955195845533396,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2176270763198828e-06,
+      "loss": 1.2682,
+      "step": 8645
+    },
+    {
+      "epoch": 0.9557483011988288,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.187807089975379e-06,
+      "loss": 1.1739,
+      "step": 8650
+    },
+    {
+      "epoch": 0.9563007568642616,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.1583546126314293e-06,
+      "loss": 1.2203,
+      "step": 8655
+    },
+    {
+      "epoch": 0.9568532125296945,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.1292697538296982e-06,
+      "loss": 1.2483,
+      "step": 8660
+    },
+    {
+      "epoch": 0.9574056681951273,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.100552621744555e-06,
+      "loss": 1.219,
+      "step": 8665
+    },
+    {
+      "epoch": 0.9579581238605602,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.0722033231826967e-06,
+      "loss": 1.2911,
+      "step": 8670
+    },
+    {
+      "epoch": 0.9585105795259931,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.0442219635827587e-06,
+      "loss": 1.2531,
+      "step": 8675
+    },
+    {
+      "epoch": 0.9590630351914259,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.016608647014916e-06,
+      "loss": 1.2883,
+      "step": 8680
+    },
+    {
+      "epoch": 0.9596154908568587,
+      "grad_norm": 0.390625,
+      "learning_rate": 9.893634761804827e-07,
+      "loss": 1.3147,
+      "step": 8685
+    },
+    {
+      "epoch": 0.9601679465222915,
+      "grad_norm": 0.419921875,
+      "learning_rate": 9.624865524115346e-07,
+      "loss": 1.256,
+      "step": 8690
+    },
+    {
+      "epoch": 0.9607204021877245,
+      "grad_norm": 0.40625,
+      "learning_rate": 9.359779756705544e-07,
+      "loss": 1.4092,
+      "step": 8695
+    },
+    {
+      "epoch": 0.9612728578531573,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.098378445500322e-07,
+      "loss": 1.2386,
+      "step": 8700
+    },
+    {
+      "epoch": 0.9618253135185901,
+      "grad_norm": 0.41015625,
+      "learning_rate": 8.840662562721314e-07,
+      "loss": 1.3856,
+      "step": 8705
+    },
+    {
+      "epoch": 0.962377769184023,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.586633066882565e-07,
+      "loss": 1.2555,
+      "step": 8710
+    },
+    {
+      "epoch": 0.9629302248494558,
+      "grad_norm": 0.41015625,
+      "learning_rate": 8.336290902788091e-07,
+      "loss": 1.2677,
+      "step": 8715
+    },
+    {
+      "epoch": 0.9634826805148887,
+      "grad_norm": 0.423828125,
+      "learning_rate": 8.089637001527317e-07,
+      "loss": 1.2327,
+      "step": 8720
+    },
+    {
+      "epoch": 0.9640351361803216,
+      "grad_norm": 0.388671875,
+      "learning_rate": 7.84667228047209e-07,
+      "loss": 1.2806,
+      "step": 8725
+    },
+    {
+      "epoch": 0.9645875918457544,
+      "grad_norm": 0.396484375,
+      "learning_rate": 7.607397643273229e-07,
+      "loss": 1.2801,
+      "step": 8730
+    },
+    {
+      "epoch": 0.9651400475111872,
+      "grad_norm": 0.427734375,
+      "learning_rate": 7.371813979857312e-07,
+      "loss": 1.3325,
+      "step": 8735
+    },
+    {
+      "epoch": 0.96569250317662,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.139922166422896e-07,
+      "loss": 1.2013,
+      "step": 8740
+    },
+    {
+      "epoch": 0.966244958842053,
+      "grad_norm": 0.408203125,
+      "learning_rate": 6.911723065437858e-07,
+      "loss": 1.2617,
+      "step": 8745
+    },
+    {
+      "epoch": 0.9667974145074858,
+      "grad_norm": 0.392578125,
+      "learning_rate": 6.687217525635614e-07,
+      "loss": 1.2695,
+      "step": 8750
+    },
+    {
+      "epoch": 0.9673498701729186,
+      "grad_norm": 0.412109375,
+      "learning_rate": 6.466406382012457e-07,
+      "loss": 1.1961,
+      "step": 8755
+    },
+    {
+      "epoch": 0.9679023258383515,
+      "grad_norm": 0.421875,
+      "learning_rate": 6.249290455824231e-07,
+      "loss": 1.2049,
+      "step": 8760
+    },
+    {
+      "epoch": 0.9684547815037843,
+      "grad_norm": 0.40625,
+      "learning_rate": 6.03587055458299e-07,
+      "loss": 1.3271,
+      "step": 8765
+    },
+    {
+      "epoch": 0.9690072371692172,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.826147472054677e-07,
+      "loss": 1.2651,
+      "step": 8770
+    },
+    {
+      "epoch": 0.96955969283465,
+      "grad_norm": 0.416015625,
+      "learning_rate": 5.620121988255567e-07,
+      "loss": 1.2567,
+      "step": 8775
+    },
+    {
+      "epoch": 0.9701121485000829,
+      "grad_norm": 0.400390625,
+      "learning_rate": 5.417794869449377e-07,
+      "loss": 1.3054,
+      "step": 8780
+    },
+    {
+      "epoch": 0.9706646041655157,
+      "grad_norm": 0.396484375,
+      "learning_rate": 5.219166868145164e-07,
+      "loss": 1.2261,
+      "step": 8785
+    },
+    {
+      "epoch": 0.9712170598309485,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.024238723093322e-07,
+      "loss": 1.2853,
+      "step": 8790
+    },
+    {
+      "epoch": 0.9717695154963815,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.833011159284029e-07,
+      "loss": 1.1949,
+      "step": 8795
+    },
+    {
+      "epoch": 0.9723219711618143,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.645484887943696e-07,
+      "loss": 1.2838,
+      "step": 8800
+    },
+    {
+      "epoch": 0.9728744268272471,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.461660606532747e-07,
+      "loss": 1.2869,
+      "step": 8805
+    },
+    {
+      "epoch": 0.97342688249268,
+      "grad_norm": 0.357421875,
+      "learning_rate": 4.281538998742951e-07,
+      "loss": 1.2944,
+      "step": 8810
+    },
+    {
+      "epoch": 0.9739793381581128,
+      "grad_norm": 0.41015625,
+      "learning_rate": 4.1051207344946495e-07,
+      "loss": 1.2441,
+      "step": 8815
+    },
+    {
+      "epoch": 0.9745317938235457,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.9324064699348686e-07,
+      "loss": 1.2831,
+      "step": 8820
+    },
+    {
+      "epoch": 0.9750842494889785,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.763396847433875e-07,
+      "loss": 1.2034,
+      "step": 8825
+    },
+    {
+      "epoch": 0.9756367051544114,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.59809249558396e-07,
+      "loss": 1.3094,
+      "step": 8830
+    },
+    {
+      "epoch": 0.9761891608198442,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.436494029196102e-07,
+      "loss": 1.222,
+      "step": 8835
+    },
+    {
+      "epoch": 0.976741616485277,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.278602049298418e-07,
+      "loss": 1.3205,
+      "step": 8840
+    },
+    {
+      "epoch": 0.97729407215071,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.1244171431332737e-07,
+      "loss": 1.2384,
+      "step": 8845
+    },
+    {
+      "epoch": 0.9778465278161428,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.9739398841557295e-07,
+      "loss": 1.3438,
+      "step": 8850
+    },
+    {
+      "epoch": 0.9783989834815756,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.8271708320309895e-07,
+      "loss": 1.2542,
+      "step": 8855
+    },
+    {
+      "epoch": 0.9789514391470084,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.6841105326324e-07,
+      "loss": 1.2843,
+      "step": 8860
+    },
+    {
+      "epoch": 0.9795038948124413,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.544759518039674e-07,
+      "loss": 1.2062,
+      "step": 8865
+    },
+    {
+      "epoch": 0.9800563504778742,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.409118306536229e-07,
+      "loss": 1.3051,
+      "step": 8870
+    },
+    {
+      "epoch": 0.980608806143307,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.277187402608405e-07,
+      "loss": 1.3078,
+      "step": 8875
+    },
+    {
+      "epoch": 0.9811612618087399,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.1489672969423614e-07,
+      "loss": 1.2903,
+      "step": 8880
+    },
+    {
+      "epoch": 0.9817137174741727,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.0244584664229628e-07,
+      "loss": 1.2979,
+      "step": 8885
+    },
+    {
+      "epoch": 0.9822661731396055,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.9036613741320043e-07,
+      "loss": 1.3551,
+      "step": 8890
+    },
+    {
+      "epoch": 0.9828186288050383,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.7865764693461017e-07,
+      "loss": 1.1832,
+      "step": 8895
+    },
+    {
+      "epoch": 0.9833710844704713,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.6732041875354709e-07,
+      "loss": 1.2429,
+      "step": 8900
+    },
+    {
+      "epoch": 0.9839235401359041,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.563544950361928e-07,
+      "loss": 1.254,
+      "step": 8905
+    },
+    {
+      "epoch": 0.9844759958013369,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.457599165677559e-07,
+      "loss": 1.1677,
+      "step": 8910
+    },
+    {
+      "epoch": 0.9850284514667698,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.3553672275230523e-07,
+      "loss": 1.1959,
+      "step": 8915
+    },
+    {
+      "epoch": 0.9855809071322026,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.2568495161264793e-07,
+      "loss": 1.2862,
+      "step": 8920
+    },
+    {
+      "epoch": 0.9861333627976355,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.1620463979014062e-07,
+      "loss": 1.2759,
+      "step": 8925
+    },
+    {
+      "epoch": 0.9866858184630684,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.070958225446006e-07,
+      "loss": 1.3177,
+      "step": 8930
+    },
+    {
+      "epoch": 0.9872382741285012,
+      "grad_norm": 0.390625,
+      "learning_rate": 9.835853375418368e-08,
+      "loss": 1.1833,
+      "step": 8935
+    },
+    {
+      "epoch": 0.987790729793934,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.999280591518444e-08,
+      "loss": 1.2862,
+      "step": 8940
+    },
+    {
+      "epoch": 0.9883431854593668,
+      "grad_norm": 0.37109375,
+      "learning_rate": 8.199867014198059e-08,
+      "loss": 1.3404,
+      "step": 8945
+    },
+    {
+      "epoch": 0.9888956411247998,
+      "grad_norm": 0.3984375,
+      "learning_rate": 7.437615616692206e-08,
+      "loss": 1.2988,
+      "step": 8950
+    },
+    {
+      "epoch": 0.9894480967902326,
+      "grad_norm": 0.400390625,
+      "learning_rate": 6.712529234016441e-08,
+      "loss": 1.2459,
+      "step": 8955
+    },
+    {
+      "epoch": 0.9900005524556654,
+      "grad_norm": 0.38671875,
+      "learning_rate": 6.024610562962441e-08,
+      "loss": 1.2718,
+      "step": 8960
+    },
+    {
+      "epoch": 0.9905530081210983,
+      "grad_norm": 0.40625,
+      "learning_rate": 5.3738621620824657e-08,
+      "loss": 1.2696,
+      "step": 8965
+    },
+    {
+      "epoch": 0.9911054637865311,
+      "grad_norm": 0.41015625,
+      "learning_rate": 4.7602864516849144e-08,
+      "loss": 1.2373,
+      "step": 8970
+    },
+    {
+      "epoch": 0.991657919451964,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4.183885713822111e-08,
+      "loss": 1.2656,
+      "step": 8975
+    },
+    {
+      "epoch": 0.9922103751173968,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.6446620922825356e-08,
+      "loss": 1.2292,
+      "step": 8980
+    },
+    {
+      "epoch": 0.9927628307828297,
+      "grad_norm": 0.4453125,
+      "learning_rate": 3.142617592583052e-08,
+      "loss": 1.2738,
+      "step": 8985
+    },
+    {
+      "epoch": 0.9933152864482625,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.677754081961137e-08,
+      "loss": 1.2733,
+      "step": 8990
+    },
+    {
+      "epoch": 0.9938677421136953,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.2500732893704358e-08,
+      "loss": 1.3335,
+      "step": 8995
+    },
+    {
+      "epoch": 0.9944201977791283,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.8595768054674444e-08,
+      "loss": 1.2119,
+      "step": 9000
+    },
+    {
+      "epoch": 0.9949726534445611,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.506266082615948e-08,
+      "loss": 1.2262,
+      "step": 9005
+    },
+    {
+      "epoch": 0.9955251091099939,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.1901424348714774e-08,
+      "loss": 1.2112,
+      "step": 9010
+    },
+    {
+      "epoch": 0.9960775647754267,
+      "grad_norm": 0.392578125,
+      "learning_rate": 9.112070379835302e-09,
+      "loss": 1.1947,
+      "step": 9015
+    },
+    {
+      "epoch": 0.9966300204408596,
+      "grad_norm": 0.416015625,
+      "learning_rate": 6.6946092938668934e-09,
+      "loss": 1.3042,
+      "step": 9020
+    },
+    {
+      "epoch": 0.9971824761062925,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.649050082006223e-09,
+      "loss": 1.2832,
+      "step": 9025
+    },
+    {
+      "epoch": 0.9977349317717253,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.975400352211999e-09,
+      "loss": 1.2572,
+      "step": 9030
+    },
+    {
+      "epoch": 0.9982873874371582,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.6736663292604704e-09,
+      "loss": 1.2085,
+      "step": 9035
+    },
+    {
+      "epoch": 0.998839843102591,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.438528546344082e-10,
+      "loss": 1.2474,
+      "step": 9040
+    },
+    {
+      "epoch": 0.9993922987680238,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.8596338656751145e-10,
+      "loss": 1.2125,
+      "step": 9045
+    },
+    {
+      "epoch": 0.9999447544334568,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0,
+      "loss": 1.2875,
+      "step": 9050
+    },
+    {
+      "epoch": 0.9999447544334568,
+      "eval_loss": 1.266111135482788,
+      "eval_runtime": 1890.6789,
+      "eval_samples_per_second": 3.812,
+      "eval_steps_per_second": 0.477,
+      "step": 9050
+    },
+    {
+      "epoch": 0.9999447544334568,
+      "step": 9050,
+      "total_flos": 6.363322205866033e+18,
+      "train_loss": 0.5663559655985121,
+      "train_runtime": 33167.9656,
+      "train_samples_per_second": 2.183,
+      "train_steps_per_second": 0.273
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 9050,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "total_flos": 6.363322205866033e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}